July 2018 Vol.17 No.

7 ISSN 1582 - 9596

Environmental
Engineering
and Management
Journal
An International Journal

Founding Editor: Matei Macoveanu

Editor-in-Chief: Maria Gavrilescu

“Gheorghe Asachi” Technical University of Iasi

Environmental
Engineering
and Management
Journal

Environmental Engineering and Management Journal encourages initiatives and actions

concerning the improvement of education, research, marketing and management, in order to achieve
sustainable development. This journal brings valuable opportunities for those offering products,
technologies, services, educational programs or other related activities, creating thus a closer relation
with the request of the market in the fields of environmental engineering, management and education.
This journal address researchers, designers, academic staff, specialists with responsibilities in the field
of environmental protection and management from government organizations (central and local
administrations, environmental protection agencies) or from the private or public companies. Also,
graduates of specialization courses or of the Environmental Engineering and Management profile, as
well as other specialists may find in this journal a direct linkage between the offer and request of the
market concerned with the protection of the environment and the administration of natural resources
in the national and international context.
The journal was conceived as a means to develop scientific and technical relationships between
people who offer and request solutions for environmental protection and conservation of natural
resources, creating thus the premises to enhance the transfer of technology and know-how, the
confirmation and implementation of ecological products and services.
Taking into consideration these aspects, we gladly welcome any persons or companies which
correspond to the above-mentioned purposes and objectives to use our journal to identify potential
collaborators.

_______________________________________________________________________________________________

For subscriptions and orders please contact us at:

OAIMDD (Academic Organization for Environmental Engineering and Sustainable
Development),
73 Mangeron Blvd., PO 10, Box 2111, 700050 Iasi, Romania, CIF: 10150285

Phone/Fax: 0040-232-271759
E-mail: brindusa.sluser.at.gmail.com, mgav_eemj.at.yahoo.com

In LEI:
ALPHA BANK ROMANIA (ABR)
SWIFT Code: BUCUROBU
IBAN: RO87BUCU1032235333663RON
Beneficiary: OAIMDD-EEMJ
Address: PO 10, Box 2111, 700050 Iasi, Romania

In EURO:
ALPHA BANK ROMANIA (ABR)
SWIFT Code: BUCUROBU
IBAN: RO79BUCU1031215940036EUR
Beneficiary: OAIMDD-EEMJ
Address: PO 10, Box 2111, 700050 Iasi, Romania

__________________________________________________________________________________
Department of Environmental Engineering and Management and
EcoZone Publishing House - O.A.I.M.D.D.
73 Prof.Dr.Docent Dimitrie Mangeron Street, 700050-Iasi, Romania
Phone/Fax: 0040-232-271759
 Environmental
Engineering and
Management Journal
EDITORIAL BOARD
Founding Editor:
Matei Macoveanu, Gheorghe Asachi Technical University of Iasi, Romania

Editor-in-Chief:
Maria Gavrilescu, Gheorghe Asachi Technical University of Iasi, Romania

SCIENTIFIC ADVISORY BOARD

Maria Madalena dos Santos Alves Anca Duta Capra José Mondéjar Jiménez
University of Minho, Braga Transilvania University of Brasov University Castilla-La Mancha, Cuenca
Portugal Romania Spain

Abdeltif Amrane Fabio Fava Shin' ichi Nakatsuji

University of Rennes, ENSCR Alma Mater Studiorum University of Bologna University of Hyogo
France Italy Japan

Ecaterina Andronescu Eugenio Campos Ferreira Valentin Nedeff

University Polytechnica of Bucharest University of Minho, Braga, Vasile Alecsandri University of Bacau
Romania Portugal Romania

Robert Armon Cristian Fosalau Alexandru Ozunu

Technion-Israel Institute of Technology, Haifa Gheorghe Asachi Technical University of Iasi Babes-Bolyai University of Cluj-Napoca
Israel Romania Romania

Adisa Azapagic Anton Friedl Yannis A. Phillis

The University of Manchester Vienna University of Technology Technical University of Crete, Chania
United Kingdom Austria Greece

Hamidi Abdul Aziz Anne Giroir Fendler Marcel Ionel Popa

Universiti Sains Malaysia, Penang University Claude Bernard Lyon 1 Gheorghe Asachi Technical University of Iasi
Malaysia France Romania

Pranas Baltrenas Silvia Fiore Marcel Popa

Vilnius Gediminas Technical University Polytechnic University of Turin, Gheorghe Asachi Technical University of Iasi
Lithuania Italy Romania

Hans Bressers Ion Giurma Valentin I. Popa

University of Twente, Enschede GheorgheAsachiTechnicalUniversity of Iasi Gheorghe Asachi Technical University of Iasi
The Netherlands Romania Romania

Han Brezet Yuh-Shan Ho Tudor Prisecaru

Delft University of Technology Peking University University Polytechnica of Bucharest
The Netherlands People's Republic of China Romania

Dan Cascaval Arjen Y. Hoekstra Gabriel-Lucian Radu

Gheorghe Asachi Technical University of Iasi University of Twente, Enschede Polytechnica University of Bucharest
Romania The Netherlands Romania

Aleg Cherp Nicolae Hurduc Ákos Rédey

Central European University, Budapest Gheorghe Asachi Technical University of Iasi PannonUniversity, Veszprém
Hungary Romania Hungary

Yusuf Chisti Ralf Isenmann Joop Schoonman

Massey University, Palmerston North Munich University of Applied Sciences Delft University of Technology
New Zealand Germany The Netherlands

Philippe Corvini Marcel Istrate Dan Scutaru

University of Applied Sciences Northwestern Gheorghe Asachi Technical University of Iasi Gheorghe Asachi Technical University of Iasi
Switzerland, Muttenz, Switzerland Romania Romania

Igor Cretescu Ravi Jain Bogdan C. Simionescu

Gheorghe Asachi Technical University of Iasi University of Pacific, Baun Hall Stockton GheorgheAsachiTechnicalUniversity of Iasi
Romania United States of America Romania

Silvia Curteanu Michael Søgaard Jørgensen Florian Statescu

Gheorghe Asachi Technical University of Iasi Aalborg University Gheorghe Asachi Technical University of Iasi
Romania Denmark Romania

Andrew J. Daugulis Nicolas Kalogerakis Carmen Teodosiu

Queen's University Kingston Technical University of Crete, Chania Gheorghe Asachi Technical University of Iasi
Canada Greece Romania

Valeriu David Gheorghe Lazaroiu Saulius Vasarevicius

Gheorghe Asachi Technical University of Iasi University Polytechnica of Bucharest Vilnius Gediminas Technical University
Romania Romania Lithuania

Katerina Demnerova Thomas Lindhqvist Angheluta Vadineanu

University of Prague International Institute for Industrial Environmental The University of Bucharest
CzechRepublic Economics, Lund University, Sweden Romania

Gheorghe Duca Andreas Paul Loibner Colin Webb

StateUniversity of Moldavia, Kishinew University of Natural Resources and Life Sciences, The University of Manchester
Republic of Moldavia Vienna, Austria United Kingdom

Emil Dumitriu Tudor Lupascu Peter Wilderer

Gheorghe Asachi Technical University of Iasi Academy of Sciences, Institute of Chemistry, Technical University Munich
Romania Kishinev, Republic of Moldavia Germany

Jurek Duszczyk Antonio Marzocchella Petra Winzer

Delft University of Technology University of Naples Federico II, Bergische University Wuppertal
The Netherlands Naples, Italy Germany
Environmental Engineering and
Management Journal
Environmental Engineering and Management Journal is included and indexed in:
CABI
Chemical Abstracts Service/SciFinder (ACS) (since 2002)
EBSCO Database (since 2002)
EVISA
ICAAP (International Consortium for Advancement of Academic Publications)
Index Copernicus Journal Master List (ICV/2015=156.46)
Journal Citation Reports® 2018 Edition (IF=1.334), (Environmental Sciences, Ranked 170 out of 241; 5-Year Impact Factor: 1.021; Article
Influence® Score: 0.086; Immediacy Index: 0.054; Normalized eigenfactor: 0.212; Quartile: Q3; Environment/Ecology, Ranked 158 out of
359, Quartile: Q2)
MedSci
ProQuest (since 2002)
The National University Research Council (RO)
Science Citation Index Expanded™
SJR (SCImago Journal&Country Rank) (Environmental Sciences, SJR index/2017= 0.41; SNIP index/2017 = 1.166; CiteScore/2017=1.26;
Ranked 680 out of 1633, H=29, Environmental Engineering: Ranked 64 out of 219; Management, Monitoring, Policy and Law: Ranked 142
out of 306; Pollution, Ranked 62 out of 160; Quartile: Q3)
SCOPUS (since 2002)
ISI Master Journal List
Web of Science® (H=28)

Home page: http://www.eemj.icpm.tuiasi.ro; http://www.eemj.eu

Full text: http://www.ecozone.ro, http://www.eemj.eu
Founding Editor: Matei Macoveanu,Iasi (RO)
Editor-in-Chief: Maria Gavrilescu, Iasi (RO)
Environmental Engineering and Management Journal is edited by
Gheorghe Asachi Technical University of Iasi and EcoZone Publishing House of the Academic Organization
for Environmental Engineering and Sustainable Development (O.A.I.M.D.D.)
Editorial and Production Office:
Department of Environmental Engineering and Management – Cristofor Simionescu Faculty of Chemical Engineering and
Environmental Protection and EcoZone - O.A.I.M.D.D., 73 Prof.Dr.Docent Dimitrie Mangeron Street, 700050 Iasi, Romania
Phone: +40-232-278680, ext. 2259, 2137
Fax: +40-232-271759
e-mail: eemjournal.at.yahoo.com, eem_journal.at.yahoo.com, eemjeditor.at.yahoo.com, eemj_editor.at.yahoo.com,
eemjournal.at.gmail.com, eemj.editor.at.gmail.com, eemj.office.at.gmail.com
Editorial production assistants and secretariat:
Laura Bulgariu
Elena-Diana Comăniță
Petronela Cozma
Raluca-Maria Hlihor
Mihaela Roșca
Isabela Maria Simion
Camelia Smaranda
Sebastian Ionuț Vasilică

Administrative and financial management:

O.A.I.M.D.D. President, Brîndușa Mihaela Slușer
Published 12 issues per year, under the aegis of the Gheorghe Asachi Technical University of Iasi, Romania
by EcoZone Publishing House of the Academic Organization for Environmental Engineering
and Sustainable Development (O.A.I.M.D.D.), http://www.ecozone.ro
Annual subscription rate 2018 (12 issues)
Print: Electronic: Print+Electronic:
EURO 500 per volume 400 per volume 800 per volume

Individual costs
EURO 50 per issue 40 per issue
Order directly to the Editorial Office, 73 Prof.Dr.Docent Dimitrie Mangeron Street, 700050 Iasi, Romania
Phone/Fax: Fax: +40-232-271759, e-mail: brindusa.sluser.at.gmail.com, eemj.office.at.gmail.com
Electronic, full text: Order or purchase on-line at: www.eemj.eu, www.ecozone.ro
Bank account (EURO):
ALPHA BANK ROMANIA (ABR)
SWIFT Code: BUCUROBU,
IBAN: RO79BUCU1031215940036EUR
Beneficiary: OAIMDD-EEMJ, Address: PO 10, Box 2111, 700050 Iasi, Romania
Bank account (LEI):
ALPHA BANK ROMANIA (ABR), IBAN: RO87BUCU1032235333663RON,
Beneficiary: OAIMDD-EEMJ
All rights reserved, including those of translation into foreign languages. No part of each issue may be
reproduced in any form (photoprint, microfilm, or any other means) nor transmitted or translated without
written permission from the publishers. Only single copies of contributions, or parts thereof, may be
made for personal use. This journal was carefully produced in all its parts. Even so, authors, editors and
publisher do not guarantee the information contained there to be free of errors. Registered names,
trademarks etc. used in this journal, even when not marked as such, are not be considered unprotected
by law.
Environmental Engineering and Management Journal July 2018, Vol.17, No. 7, 1521-1772
http://www.eemj.icpm.tuiasi.ro/; http://www.eemj.eu

“Gheorghe Asachi” Technical University of Iasi, Romania

CONTENTS

Efficiency of an electrocoagulation treatment of water contaminated

by hydrocarbons in a continuous mode powered by photovoltaic solar modules
Kahina Ighilahriz, Mourad Taleb Ahmed, Hayet Djelal, Rachida Maachi............................... 1521

A brief review on recent advances in air-cathode microbial fuel cells

Pritha Chatterjee, Makarand Madhao Ghangrekar, Donal Leech........................................... 1531

Comparing Mamdani and Sugeno hierarchical fuzzy systems

for environmental impact assessment: A pipeline project case study
Sharosin Givargis, Azadeh Dabiri, Negin Hakiminejad,
Mahmood Mahmoodi, Farzam Babaei Semiromi..................................................................... 1545

Dimension and structural traits of soil micropores in cultivations differing

in the duration of organic management
Nikolaos Monokrousos, Efimia M. Papatheodorou, George P. Stamou.................................. 1555

Changes of arsenic speciation during swine manure windrow composting

at full scale
Rui Wang, Yuansong Wei, Yulong Wu, Jiaxi Zheng, Bofu Zheng ............................................. 1563

Integrated use of two microalgal species for the treatment of aquaculture

effluent and biomass production
Katya Velichkova, Ivaylo Sirakov, Yordan Staykov.................................................................. 1575

Environmental performance of sajor-caju mushroom production

based on farm sizes in Thailand
Siriprapa Ueawiwatsakul, Thumrongrut Mungcharoen, Rungnapa Tongpool......................... 1583
Indirect environment-related effects of electric car vehicles use
Bogdan Ovidiu Varga, Florin Mariasiu.................................................................................... 1591

The “insurance effect”: how to increase the investment amount in green

buildings –A model-based approach to reduce the energy efficiency gap
Hans Ulrich Buhl, Tobias Gaugler, Philipp.............................................................................. 1599

Adsorption behavior of gadolinium(III) from aqueous solution by D418 resin

Leilei Pi, Chunhua Xiong, Jianxiong Jiang, Xuming Zheng, Rongrong Zhao,
Caiping Yao, Qing Chen, Yaqin Fu, Yaofeng Zhu.............................................................. 1613

Removal of hexavalent chromium using two innovative adsorbents

Reda Abubeah, Hossam Altaher, Tarek E. Khalil.................................................................... 1621

Vertical distribution and potential mobility of heavy metals

in new and old tailings of a lead/zinc sulfide mine
Malik Anju................................................................................................................................ 1635

Supply chain emission reduction optimization under consumer carbon

sensitivity and carbon tax policy
Lan Bai, Xianliang Shi, Honghu Gao, Qiwen Du...................................................................... 1645

Factors affecting the pattern of vegetation carbon density in a karst

region in Northwest Guangxi, China
Mingyang Zhang, Kelin Wang, Huiyu Liu, Weijian Luo, Jing Wang, Yuemin Yue.................... 1657

Analysis of commercially available bottled water in Poland

Rajmund Michalski, Magdalena Jabłońska-Czapla,
Sebastian Szopa, Aleksandra Łyko........................................................................................... 1667

Survey of industrial wastewater treatment performance with biological

monitoring of protozoa in the industrial park of Abadan
Abotaleb Bay, Mahdi Sadeghi, Hadi Rahimzade, Nasser Behnampour................................... 1679

Application of Response Surface Methodology (RSM) for optimization

of zinc extraction from anaerobic sewage sludge
Vladimír Frišták, Martin Pipíška, Michaela Valovčiaková, Juraj Lesný....................................... 1685

Analysis of the continuous measurements of PM10 and PM2.5

concentrations in Beirut, Lebanon
Wehbeh Farah, Myriam Mrad Nakhlé, Maher Abboud, Nelly Ziade,
Isabella Annesi-Maesano, Rita Zaarour, Nada Saliba, Georges Germanos,
Najat Aoun Saliba, Alan L. Shihadeh, Jocelyne Gerard............................................................ 1693
Contrastive soil properties, microbial structure and soil enzymes in the
rhizosphere of Scirpus triqueter and bulk soil in petroleum-contaminated wetland
Jing Wei, Xiaoyan Liu, Chuanhua Wang, Xueping Chen, Xia Liang, Qian Wang..................... 1701

Study of nitrogen forms in the seasonal dynamics and kinetics of nitrification

and de-nitrification in Prut and Nistru river waters
Petru Spataru, Igor Povar, Tudor Lupascu, Alfredo Carlos Alder, Elena Mosanu................ 1711

Physical networks based on gelatin and azo-polysiloxanes

Elena- Luiza Epure, Elena Resmeriţă, Anca Daniela Rusu,
Bogdana Simionescu, Constanţa Ibănescu………………………………………………………… 1721

Energy sources analysis in the perspective of sustainable development

Tiberiu Rusu, Ciprian Cristea, Tudor Rusu, Ileana-Codruţa Groze, Laurenţiu Stoica............. 1729

Heavy metals content in the sprouts of Glyceria maxima (Hartm.) Holmb.

and in river sediments (Northern Poland)
Agnieszka Parzych, Małgorzata Cymer.................................................................................... 1743

Research on wastewaters bioremediation with aquatic species

for constructed wetlands
Dana Malschi, Liviu Muntean, Ioan Oprea, Carmen Roba, Gabriela Popiţa,
Lucrina Ştefănescu, Bianca Malschi Florian, Elena Rînba..................................................... 1753

Environmentally friendly cellulosic fibers from corn stalks

Ana Maria Cheșcă, Bogdan Marian Tofănică, Adrian Cătălin Puiţel,
Raluca Nicu, Dan Gavrilescu ......... ......................................................................................... 1765
Environmental Engineering and Management Journal July 2018, Vol.17, No. 7, 1521-1529
http://www.eemj.icpm.tuiasi.ro/; http://www.eemj.eu

“Gheorghe Asachi” Technical University of Iasi, Romania

EFFICIENCY OF AN ELECTROCOAGULATION TREATMENT OF

WATER CONTAMINATED BY HYDROCARBONS IN A CONTINUOUS
MODE POWERED BY PHOTOVOLTAIC SOLAR MODULES

Kahina Ighilahriz1, Mourad Taleb Ahmed1∗, Hayet Djelal2,3, Rachida Maachi1

1USTHB / Mechanical Engineering & Process Engineering, Environmental Engineering, 16111 Algeria
2Environmental Trades School, Ker Lann Campus, 35170 Bruz, France
3
European University of Brittany, Rennes, France

Abstract

The aim of this study is the treatment of liquid effluents from a petroleum industry (NAFTAL /Algeria) by an electrochemical
treatment process (electrocoagulation). The electrocoagulation has proved to be useful for the treatment of wastewater, but in order
to enhance its ecological characteristic we found it would be essential to carry out our experiments with ecological energy derived
from a photovoltaic solar module, which would contribute to sustainability objectives. The electrocoagulation process in a
continuous mode using anodes (aluminum and iron) and a graphite cathode was studied. Changes in the Chemical Oxygen Demand
were analyzed which stated the efficiency of the method. The current density, the duration of treatment, the pH and the amount of
electrolyte added in order to optimize these operating parameters lead to a better reduction of the pollution load. After 1h of
electrolysis, optimum conditions were: current density of 100 A/m2, with initial pH 7.4 and a flow rate of 3.94.10-3 L/s. More than
94% of COD were removed with an aluminum anode.

Key words: COD removal, electrocoagulation, electrode, petroleum oil, photovoltaic energy

Received: January, 2014; Revised final: July, 2014; Accepted: September, 2014; Published in final edited form: July 2018

1. Introduction portion of recoverable wastewater comes from

Oil is a mixture of hydrocarbons, which are
combination of carbon and hydrogen molecules. It is
less dense than water, has a characteristic odor and its
color varies from black to brown (Cerqueira and Da
Costa Marques, 2012). Depending on its origin, it may
contain a variety of chemicals such as organic salts,
aliphatic and aromatic hydrocarbons, oils and grease,
metals and radioactive material (Henderson, 1999,
Oliveira et al., 2005). In order to meet the needs of the
growing demand for oil, its extraction has risen
dramatically in recent decades producing huge
amounts of toxic effluents worldwide - 33.6 million
barrels a day (Diya’uddeen et al., 2011). A large
petrochemical activities such as oil extraction and its
subsequent refining processes (Rastegar et al., 2017;
Maretto et al., 2014); the release of these wastewaters
into the environment largely affects the ecosystem. In
addition, the failure to observe safety standards and
current equipment maintenance in the oil industry
aggravates water pollution problems. In order to
protect the environment, the treatment and reuse of
wastewater has become an absolute necessity.
Improved water management in a refinery must reduce
the volume and the coast of raw water used in refinery
operation, but firstly contaminants should be reduced
(Davoudi et al., 2014). For these reasons, innovative
and inexpensive techniques for the treatment of

∗
Author to whom all correspondence should be addressed: e-mail: mdtaleb_fr@yahoo.fr; Phone: +213 21 24 79 19; Fax: +213 21 24 79 19
 Ighilahriz et al./Environmental Engineering and Management Journal 17 (2018), 7, 1521-1529
wastewater in the oil industry should be developed. In
recent years, there has been a growing interest for the
use of electrochemical methods for the treatment of
industrial wastewater especially electrocoagulation
(EC). This process is well known in wastewater
treatments (Martinez-Huitle and Brillas, 2009;
Mohammed et al., 2009). In fact, it has been
successfully used for the treatment of drinking water
(Pop et al., 2017; Vik et al., 1984) and wastewaters
generated from aquaculture (Lin and Chen, 1997),
industries (Ramirez, 1981), textile industry (Lin and
Peng, 1996; Pajootan et al., 2016), tannery industry
(Benhadji et al., 2011), restaurants (Chen et al., 2000),
landfill leachate, municipal sewage (Biwyk et al.,
1980) and urban wastewater (Pouet and Grasmick,
1995). This process was used to remove phenol (Awad
and Abuzaid, 2000) and surfactants (Ciorba et al.,
2000) and even the oil industry (El-Naas et al., 2009;
Hariz et al., 2013) used it to cut oil–water emulsions
(Bensadok et al., 2008), oil suspension used for
machining and drilling operations (Pansares et al.,
2008) and for oil drilling mud treatment (Ighilahriz et
al., 2013). Electrocoagulation (EC) is an effective
method wherein the flocculating agent is produced by
the electro-oxidation of a single-use anode made of
iron or aluminum mostly. This process turns a suitable
pH into an insoluble metal hydroxide that can remove
pollutants. In this process, the treatment is done
without the addition of chemical flocculants or
coagulants, thereby reducing the amount of sludge that
should be removed.
This technique is easy and requires simple
equipment and a small amount of sludge (Mollah et
al., 2004; Yehya et al., 2015) is produced. Its
technology’s efficiency and reliability were stated and
it provides an environmentally compatible method to
reduce a wide variety of pollutants (Ciorba et al.,
2000).
During the electrocoagulation process, the
electrochemical reactions of electrodes in aluminum
are summarized as follows:
At the anode Eq. (1):
3+
𝐴𝐴𝐴𝐴(𝑠𝑠) → 𝐴𝐴𝐴𝐴(𝑎𝑎𝑎𝑎) + 3𝑒𝑒 −
At the cathode:
The reaction occurring at the cathode is dependent on
pH. Indeed, under neutral or basic conditions
hydrogen is produced through Eq. (2), whereas under
acidic conditions, Eq. (3) better describes the
hydrogen evolution:
− −
2𝐻𝐻2 𝑂𝑂(𝑙𝑙) + 2𝑒𝑒 → 𝐻𝐻2(𝑔𝑔) + 2𝑂𝑂𝐻𝐻(𝑎𝑎𝑎𝑎)
+
2𝐻𝐻(𝑎𝑎𝑎𝑎) + 2𝑒𝑒 − → 𝐻𝐻2(𝑔𝑔)
In this study, EC using aluminum and iron
electrodes in continuous mode was studied in the
treatment of an oily wastewater generated from a
petroleum industry at Dar el Beida (Algiers, Algeria).
The first purpose of this study was to observe the
1522
effects of the main operating variables on the
efficiency of the EC process, like the anode materials,
the operating time and the current density in order to
determine the optimal conditions for it to remove
pollutants. The sludge formed was characterized. The
second purpose of the study was to explore a new
solution for treating oily wastewater with a renewable
energy through a photovoltaic solar power which is a
new form of non-polluting, abundant and renewable
energy. Photovoltaic modules last long and involve
low maintenance costs (Valero et al., 2008); they can
produce a high direct current when exposed to the
sunlight and connected in series and/or in parallel.
Based on the above facts, the photovoltaic energy can
be used directly in electrocoagulation to treat
wastewater, and this is becoming a popular trend in the
environmental protection field. The use of the
photovoltaic solar power, an eco-friendly energy can
produce a high direct current when exposed to
sunlight. The photovoltaic module can be installed
near a remote or absent power region (Navarro-Solis
et al., 2010; Valero et al., 2008; Zhang et al., 2013).
2. Material and methods
2.1. Water samples
An effluent was selected from NAFTAL
(Algeria) oil industry which contains kerosene/ water
emulsions taken downstream of the local sewage
treatment plant at the industrial site. The sample
containers were preserved in a refrigerator at 4°C
before use. At the beginning of the study, the main
characteristics of the collected samples were analyzed
and the water temperature was increased to the room
temperature before each electrocoagulation test. Table
1 shows the characteristics of the effluent used during
the process.
Table 1. The main characteristics of oily wastewater
Permissible levels
Parameters Values
[Decree, 2006]
COD (mg O2/L) 292 100
(1) BOD5 (mg O2/L) 30 25
pH 7.50 5.5-8.5
Conductivity
0.36 -
(mS/cm)
Turbidity (NTU) 2.32 -
2.2. The electrocoagulation unit
It is composed of two compartments. The first
(2) consists of a 10-L cylindrical tank. The second
compartment is provided with a cathode of graphite
(3) and anodes of cylindrical shapes made of aluminum
(with a 0.8-cm internal diameter, a 12.1-cm height
with an active area of 30.39 cm2) and iron (with 0.3-
cm internal diameter, a 12.1-cm height with an active
area of 11.39 cm2). The electrodes were placed at the
middle bottom of the reactor, parallel to each other in
the cell at a fixed distance of 1 cm. The EC cell was
fed continuously by a peristaltic pump (Iwaki Magnet
 Efficiency of an electrocoagulation treatment of water contaminated by hydrocarbons in a continuous mode
Pump; 16-19 L/min) with the effluents from a
wastewater tank. The reactor operates as a closed
reactor with full recirculation, the total solution does
not exceed 2-L, and electrodes were connected to a DC
power supply (HEP-613; 0-2 A; 40-V). The cell was
powered by a photovoltaic system. The battery had an
internal resistance of 17 mΩ, a 4.8 A maximum
charging current and a minimum safe end discharge
voltage of 1.60 V/cell. The SOLAR23 solar module
line was especially designed for PV off grid systems
and covered modules in the range from 5 to 100 Wp.
Samples were taken at various time interval. The
experimental equipment is schematically shown in
Fig. 1.
The operation lasted 2 h; samples were taken
from the reactor every 15 min and were analyzed.
Their pH, conductivity and COD were determined.
The pH measurement was carried out with a HANNA
Digital Display model SED pH meter using 12500 V.
The conductivity measurements were made by using a
HANNA.COD conductivity meter. COD was
measured using a thermo ECO reactor refluxing by
titration method following the NFT 90-101 standard.
Boiling under reflux in the presence of mercury sulfate
(II), a test sample, in the presence of a known amount
of potassium dichromate and a silver catalyst, strongly
acidified with sulfuric acid for a period of time during
which a portion of dichromate was reduced by the
oxidizable matter present in the solution. The
dichromate excess was titrated with a standard
solution of iron sulfate (II) and ammonium.
Calculation of COD was done from the amount of the
reduced dichromate. BOD5 analysis was carried out by
a Schrank Aqua Lytic Thermostat following the NF
T90-103 standard. The sample was incubated for 5
days at 20 °C. The microorganisms consumed the
dissolved oxygen and the carbon dioxide released was
trapped by sodium hydroxide. The amount of sludge
generated after the EC treatment under optimal
conditions using aluminum electrodes was
characterized using a FT-IR spectrometer (Perkin-
Elmer N°200139).
3. Results and discussion
The EC process is quite complex and is
affected by several operating parameters such as the
Fig. 1. Electrochemical cell
initial pH, the duration of the EC, the nature of the
electrode material and the current density. In order to
extend the process performance, the effects of those
parameters were studied and the results are stated
below. The solution was prepared with 1 L of pure
effluent and 1 L of distilled water which led us to add
electrolyte support to raise the solution conductivity,
to decrease the voltage between electrodes due to the
decreased resistance of polluted water and to reduce
energy consumption.
3.1. Effect of current density and electrolysis duration
on EC process
The current density is considered as a key
parameter in EC (Khemis et al., 2006; Xu and Zhu,
2004) especially for the reduction kinetics of COD. It
is knowledged that current density determines the
production rate of coagulant as well as the adjustment
of bubble production, and hence it affects the growth
of flocs (Chen et al., 2004; Mollah et al., 2004). In
order to assess the effect of the applied current density,
a set of experiments was planned at an initial pH 7.4
(since it was the natural pH of the solution) and
varying applied current density (50-125 A/m2) for a
concentration of supporting salt NaCl of 0.5 g/L.
The COD reduction kinetics goes fast during
the first 15–30 min of EC and is almost stabilized
thereafter. One can say that when the current density
increases, the processing time decreases thanks to the
strong electrodes dissolution. In fact, when current
density increases the production rate of hydrogen
bubbles increases and their size decreases. This
highlights an effective elimination of the pollutant by
flotation (Ben Hariz et al., 2013). Moreover, the
increase of current density entails the increase in the
amount of coagulant as a consequence of greater
removal rate. At a lower current density, the anode
dissolution does not provide a destabilization of
particles because of the low anode dissolution rate.
However, some COD removal is noticeable
which can be explained by an indirect oxidation of the
oil suspension provided by oxidizing species such as
ClO- allowing the indirect electrochemical treatment
of the wastewater (Adhoum et al., 2004). The most
significant reduction rate (92 %) occurred at a current
density of 100 A/m2 for 1 h.
1523
 Ighilahriz et al./Environmental Engineering and Management Journal 17 (2018), 7, 1521-1529

The fact that for higher current densities the
COD removal decreases can be explained by an
overdose of coagulant, which can carry an excessive
adsorption of cations and reversing the charge of the
particles. The particles could then be stabilized again
(Amirtharajah and O'Melia, 1990). Fig. 2 (a) proves
that duration has effects on experimental change
results in COD reduction for different current
densities.
From Fig. 2 (b) one can infer that the decrease
in conductivity, being related to the current density
during this spell of time, is due to the precipitation of
salts produced during the EC reaction with aluminum
electrodes. For the case of the experiments made at
150 A/m2 the conductivity is lower. The reason of this
finding may be that the increment of current density
causes an excessive presence of ions produced during
EC leading to lower conductivity measured after
decantation. The forthcoming experiments were all
run at the current density of 100 A/m2.
3.2. Effect of initial pH on EC process
It is established that the initial pH is a
significant factor affecting the electrocoagulation
process (Asselin et al., 2008; Yildiz et al., 2008). In
order to study its effect on the COD removal and the
final pH, the treated solution was adjusted between pH
4 and pH 10 with 0.1 M sodium hydroxide or 0.1 M
hydrochloric acid. The current density was set at
100A/m2 and likewise the NaCl concentration was
fixed at 0.5 g/L. Fig. 3 shows that COD removal
increased from 34 % for pH 4 to 92 % for pH 7.4.
Former studies already revealed that EC gave best
results when the initial pH varied from pH 3 to neutral
pH 7 forming a majority of aluminum complexes
which played an active part during the removal of
pollutant through EC (Alaton et al., 2009).
With a pH 8 solution the removal efficiency
decreased because of the amphoteric character of the
aluminum hydroxide Al(OH)3 whose solubility
increases when the solution becomes either more
acidic or alkaline (Bensadok et al., 2008; Koparal and
Ogutveren, 2002; Mollah et al., 2004). Moreover, a
high pH entails the formation of Al(OH)4−, which is
soluble and useless for the removal of oil emulsion
(Koren and Syversen, 1995). Therefore, a further
increase of the influential pH does not help the COD
removal.
During the electrolysis the solution pH changes
as shown in Fig. 4 and thus, the final pH was 6.3, 7.2,
7.3, 8.3 for the initial pH of 4, 5, 6, 7.4 respectively,
which was due to the OH- ions produced in the solution
by the water electrolysis.

Fig.2. Effect of current density on: (a) COD and (b) conductivity vs. time.
pH=7.4, CNaCl=0.5g/L, Q=5.79.10-3 L/s (Aluminum anode)

Fig.3. Effect of initial pH on COD, j= 100 A/m2, t=1h, CNaCl=0.5g/L, Q=5.79.10-3L/s (Aluminum anode)

1524
 Efficiency of an electrocoagulation treatment of water contaminated by hydrocarbons in a continuous mode
We can observe that for a medium alkaline pH
of 8, the final pH was not that different (pH = 8.5). For
a basic pH of 10, the final pH decreased to 9.3 by the
production of H+. On one hand, metal hydroxides
Al(OH)3 are formed and on the other hand, the
oxidation of water to oxygen can help to reduce the pH
of the water to be treated (Chen et al., 2004).
Fig.4. Effect of initial pH on pH, j= 100 A/m2,
CNaCl=0.5g/L, Q=5.79.10-3 L/s (Aluminum anode)
3.3. Effect of electrolyte concentration on EC process
The conductivity of the reaction mixture is one
of the main parameters that influence water treatment
by EC and can be increased by the addition of salt
(Mollah et al., 2004; Holt, 2005). In our study NaCl
was used because it is non-toxic and its cost is
moderate (Calvo et al., 2003). Adding some
electrolyte support would decrease power
consumption and destroy the passivation layer and
hence improve the dissolution rate of Al electrodes
(Lee and Pyun, 1999). Moreover, the
electrochemically generated chlorine was found to be
effective in water disinfection (Muthukumar, 2007).
When there are chlorides in the liquid waste, Cl2 and
OCl- can be produced from anodic discharge of
chloride ions (Chen et al., 2004) by Eqs. (4-6). The
hypochlorite ion OCl- is a strong oxidant which could
oxidize some of the organic molecules present in
wastewater therefore it contributes to the water
treatment.
2𝐶𝐶𝐶𝐶 − + 2𝑒𝑒 − → 𝐶𝐶𝐶𝐶2
− +
𝐶𝐶𝐶𝐶2 + 𝐻𝐻2 𝑂𝑂 → 𝐻𝐻𝐻𝐻𝐻𝐻𝐻𝐻 + 𝐶𝐶𝐶𝐶 + 𝐻𝐻
𝐻𝐻𝐻𝐻𝐻𝐻𝐻𝐻 + 𝐻𝐻2 𝑂𝑂 → 𝑂𝑂𝐶𝐶𝐶𝐶 − + 𝐻𝐻3 𝑂𝑂 + ,
pKa: 7.3 at 25°C
In order to study the effect of NaCl on the COD
removal (Fig. 5), various concentrations were
analyzed (from 0.25 to 1.0 g/L).
3.4. Flow rate effect on EC process
In order to investigate the effect of time on the
COD removal efficiency, the flow rate at start (Q) was
gradually increased from 3.94.10-3 to 16.3.10-3 L/s at
100 A/m2; pH = 7.4 and CNaCl = 0.5 g/L.
Fig.5. Rate abatement as function of electrolyte
concentration: j =100A/m2, t=1h, pH=7.4
(Aluminum anode)
According to Fig. 6, the reduction of COD
decreased with increasing flow rate. We can conclude
that when the flow rate raises the COD reduction
drops. Indeed, the lowest the flow rate the highest the
residence time is. Therefore, the effluent remains
longer time in contact with the coagulant into the
reactor and improves the treatment efficiency, in
agreement with the results previously obtained by
other researchers (Emamjomeh and Sivakumar, 2009).
The optimal inlet flow rate could therefore be
considered to be 3.94.10-3L/s.
(4)
Fig.6. COD as function of flow rate using j=100 A/m2,
pH=7.4, CNaCl=0.5g/L (Aluminum anode)
(5)
(6)
3.5. Effect of anode material on EC process
The composition of the electrode material is a
key element in the electrochemical treatment; the
choice of the adequate electrode material is important
to reach better removal efficiency.
Aluminum and iron are commonly used
because they are cheap. Thus, experiments were
conducted for both aluminum and iron anodes under
the same operational conditions.
1525
 Ighilahriz et al./Environmental Engineering and Management Journal 17 (2018), 7, 1521-1529

𝑈𝑈𝑈𝑈𝑈𝑈
Fig. 7 shows the effect of current density on 𝐸𝐸𝐶𝐶 = × 103 (7)
𝑉𝑉×𝛥𝛥𝛥𝛥𝛥𝛥𝛥𝛥
COD removal. We found out that the rate of COD
reduction at a current density of 100 A/m2 was greater
where U is cell voltage (V), I is the applied electrical
with aluminum electrode whereas it was almost the
current (A), t is the reaction time (h), V is the reactor
same at a current density of 50 A/m2. With a density
volume (L) and ΔCOD is the subtraction of COD (mg
of 125 A/m2, removal was more effective with iron
O2/L) before and after the electrolysis (COD removed
anodes. The trends for COD removal in Fig. 7 by
during electrochemical treatment).
different anode materials (aluminum and iron) did not
have the same trends at different current densities this
can be explained by the generation of various shapes
and multiples species of Aluminum such as
monomeric species including, Al(OH)2+ , Al(OH)+2 ,
Al(OH)- 4 and Al2(OH)4+2 and the polymeric species
such as Al2(OH)+5 , Al6(OH)3+15 , Al13(OH)5+34 ,
Al7(OH)4+17 , Al8(OH)204+ , Al13O4(OH)7+24 (Mollah et
al., 2001).

Fig.8. COD as function of pH using j=100 A/m2 t =1 h,

CNaCl=0.5g/L, Q=3.94.10-3 L/s for aluminum and iron
electrodes

The chemical consumption (CHC) was

calculated in terms of kg of chemicals/kg COD using
Eq. (8):

Fig.7. Rate abatement as function of current density using
pH = 7.4, t=1h, CNaCl=0.5 g/L, Q= 3.94.10-3 L/s
for aluminum and iron electrodes
Both followed the same trends and the scatter
diagram shows that the optimum pH obtained with the
aluminum electrodes (7.4) is not the same as with iron
(6.0). This can be explained by the fact that aluminum
hydroxide has a greater compatibility with the
pollutant compared to iron hydroxide.
Fig. 8 compares the COD reduction according
to pH with both electrodes and shows the worse results
obtained with iron electrodes under the same operating
conditions as the experiments with aluminum - namely
pHi equal to 7.4, Q =3.94.10-3 L/s, j = 100 A/m2 and
[NaCl] = 0.5 g/L. Indeed, a reduction of 95% was
achieved with aluminum electrodes against 69% with
iron.
3.6. Electrical energy consumption
One of the most important factors to be taken
into account for the selection of a water treatment
process is the cost. Electrodes, chemicals and the
energy used during EC add up to the process operating
costs (Ayhan Şengil and Özacar, 2006; Gebologlu,
2010; Kobya et al., 2003, Khandegar and Saroha,
2013). The electrical energy required for the removal
of COD using aluminum electrodes was calculated in
terms of kWh/kg COD using Eq. (7):
𝐶𝐶ℎ𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑢𝑢𝑢𝑢𝑢𝑢𝑢𝑢
𝐶𝐶𝐶𝐶𝐶𝐶 = × 103 (8)
𝑉𝑉×∆𝐶𝐶𝐶𝐶𝐶𝐶
The electrode consumption (kg/kg COD) was
calculated according to Eq. (9):
𝐼𝐼×𝑡𝑡×𝑀𝑀
𝐸𝐸𝐸𝐸𝐸𝐸 = × 103 (9)
𝑛𝑛×𝐹𝐹×∆𝐶𝐶𝐶𝐶𝐶𝐶×𝑉𝑉
where: M is the molar mass (kg ̸mol), n is the valence
number (= 3) and F is the Faraday constant (96485
C.mol−1).
As seen in Fig. 9, a band appeared at 3434
corresponding to O-H stretching (Jewel et al., 2007),
whereas the peak towards 1643 cm–1 is due to the
presence of bending hydroxyl (Drouiche et al., 2009).
The operating cost was calculated from the following
expression Eq. (10) (Khandegar and Saroha, 2013):
𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 = 𝑎𝑎𝑎𝑎𝑐𝑐 + 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 + 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 (10)
where: a is the electrical energy price (0.022 US$ per
kWh), b corresponds to the chemical consumption
price (12.60 US$ per kg of added NaCl) and c is the
electrode material consumption (3.8 US$ per kg of
aluminum). Working at 100 A/m 2 during 1 h of
electrolysis, Q= 3.94.10-3 L/s and CNaCl = 0.5 g/L and
pH 7.4, the operating cost of EC was 0.0761US$ per
kg of COD, with Algeria price electricity. These costs
are in agreement with those given in the literature
(Saeedi and Khalvati-Fahlyani, 2011).

1526
 Efficiency of an electrocoagulation treatment of water contaminated by hydrocarbons in a continuous mode
Fig.9. FT-IR spectrum of the sludge produced in the electrochemical process
3.7. Characterization of sludge obtained with the EC
The sample obtained under the optimal
conditions using aluminum electrodes was filtered to
get the sludge formed during the EC in order to be
analyzed with the Fourier Transform Infrared
Spectroscopy (FT-IR). The characteristic of Al–O-H
bending is represented by the band at 872 cm−1 (Jewel
et al., 2007). These results account for a mechanism in
which the pollutant was removed from wastewater in
the EC process. In fact, the sludge contains aluminum
hydroxide in a major part which proves that the Al was
turned into sludge.
4. Conclusions
This study confirms that the electrocoagulation
process is satisfactory with results reaching 94 % of
oil removal (liquid effluent from a petroleum
industry). The optimum conditions are 1 h of
electrolysis and a current density of 100A/m2 with an
initial pH 7.4 and a 3.94.10-3L/s flow rate with an
aluminum anode.
In the best treatment conditions, the COD
residual (18 mg O2/L) was less than the permissible
level of 100 mg O2/L. The operational cost of EC was
0.0761US$ per kg of COD. Overall, the results
confirm the efficiency of EC process, powered
directly by a photovoltaic module, in removing
pollution from oil wastewater.
References
Adhoum N., Monser L., Bellakhal N., Belgaied J.E., (2004),
Treatment of electroplating wastewater containing
Cu2+, Zn2+and Cr(VI) by electrocoagulation, Journal of
Hazardous Materials, 112, 207-213.
Alaton I.A., Kabdasli I., Vardar B., Tunay O., (2009),
Electrocoagulation of simulated reactive dyebath
effluent with aluminum and stainless-steel electrodes,
Journal of Hazardous Materials, 164, 1586-1594.
Amirtharajah A., O’Mella C.R., (1990), Coagulation
Processes: Destabilization, Mixing, and Flocculation,
In: Water Quality and Treatment: A Handbook of
Community Water Supplies, Pontius F.W. (Eds.),
AWWA, McGraw-Hill Inc., New York, 269-365.
Asselin M., Drogui P., Benmoussa H., Blais J.F., (2008),
Effectiveness of electrocoagulation process in
removing organic compounds from slaughterhouse
wastewater using monopolar and bipolar electrolytic
cells, Chemosphere, 72, 1727-1733.
Ayhan Şengil I., Özacar M., (2006), Treatment of dairy
wastewaters by electrocoagulation using mild steel
electrodes, Journal of Hazardous Materials, 137, 1197-
1205.
Awad Y.M., Abuzaid N.S., (2000), The influence of
residence time on the anodic oxidation of-phenol,
Separation and Purification Technology, 18, 227-236.
Benhadji A., Taleb Ahmed M., Maachi R., (2011),
Electrocoagulation and effect of cathode materials on
the removal of pollutants from tannery wastewater of
Rouïba, Desalination, 277, 128-134.
Ben Hariz I., Halleb A., Adhoum N., Monser L., (2013),
Treatment of petroleum refinery sulfidic spent caustic
wastes by electrocoagulation, Separation and
Purification Technology, 107,150–157.
Bensadok K., Benammar S., Lapicque F., Nezzal G., (2008),
Electrocoagulation of cutting oil emulsions using
aluminum plate electrodes, Journal of Hazardous
Materials, 152, 423-430.
Biwyk A., (1980), Electrocoagulation of Biologically
Treated Sewage, Proc. of 35th of Purdue Industrial
Waste Conference, Lafayette, Indiana, USA.
Calvo L.S., Leclere J.P., Tanguay G., Cames M.C.,
Paternotte G., Valentin G., Rostan A., Lapicque F.,
(2003), An electrocoagulation unit for the purification
of soluble oil wastes of high COD, Environmental
progress, 22, 57-65.
Cerqueira A.A., Da Costa Marques M.R., (2012),
Electrolytic Treatment of Wastewater in the Oil
Industry, In: New Technologies in the Oil and Gas
1527
 Ighilahriz et al./Environmental Engineering and Management Journal 17 (2018), 7, 1521-1529
Industry, Gomes J.S. (Eds.), InTech Publisher, Rijeka,
Croatia, 3-28.
Chen G., Chen X., Yue P.L., (2000), Electrocoagulation and
electroflotation of restaurant wastewater, Journal of
Environmental Engineering, 126, 858-863.
Chen G., (2004), Electrochemical technologies in
wastewater treatment, Separation and Purification
Technology, 38, 11-41.
Ciorba G.A., Radovan C., Vlaicu I., Pitulice L., (2000),
Correlation between organic component and electrode
material: Consequences on removal of surfactants from
wastewater, Electrochemical Acta, 46, 297-303.
Davoudi M., Mottaghi H.R., Samieirad S., Heidari Y.,
(2014), Characteristics, treatments, and management
for disposal of sanitary wastewater in South Pars Gas
processing plants, Journal of Natural Gas Science and
Engineering, 19, 91-97.
Decree, (2006), Executive Decree N. 06-141 of 20 Rabie El
Aouel 1427 corresponding to 19 April 2006 defining
the limits of industrial liquid effluent discharges, Annex
I, On line at: http://www.sante.dz/jms2010/oms/dec06-
141.pdf.
Diya’uddeen B.H., Daud W.M.A.W., Abdul Aziz A.R.,
(2011), Treatment technologies for petroleum refinery
effluents –A review, Process Safety and Environmental
Protection, 89, 95-105.
Drouiche N., Aoudj S., Hecini M., Ghaffour N., Lounici H.,
Mameri N., (2009), Study on the treatment of
photovoltaic wastewater using electrocoagulation:
Fluoride removal with aluminum electrodes-
Characteristics of products, Journal of Hazardous
Materials, 169, 65-69.
Emamjomeh M.M., Sivakumar M., (2009), Fluoride
removal by a continuous flow electrocoagulation
reactor, Journal of Environmental Management, 90,
1204-1212.
El-Naas M.H., Al-Zuhair S., Al-Lobaney A., Makhlouf S.,
(2009), Assessment of electrocoagulation for the
treatment of petroleum refinery wastewater, Journal of
Environmental Management, 91, 180–185.
Gebologlu U., (2010), Removal of arsenic by
electrochemical wastewater treatment process, MSc
Thesis, Gebze Institute of Technology, Gebze, Turkey.
Hariz I.B., Halleb A., Adhoum N., Monser L., (2013),
Treatment of petroleum refinery sulfidic spent caustic
wastes by electrocoagulation, Separation and
Purification Technology, 107, 150-157.
Henderson S.B., (1999), Potential impact of production
chemicals on the toxicity of produced water discharges
from North Sea Oil Platforms, Marine Pollution
Bulletin, 38, 1141-1151.
Holt P.K., Barton G.W., Mitchell C.A., (2005), The future
of electrocoagulation as localized water treatment
technology, Chemosphere, 59, 355-367.
Ighilahriz K., Taleb Ahmed M., Djelal H., Maachi R.,
(2013), Electrocoagulation and electro-oxidation
treatment for the leachate of oil-drilling mud,
Desalination and Water Treatment, 52, 1-7.
Jewel A.G., Daida P., Kesmez M., Weir M., Moreno H.,
Parga J.R., Irwin G., McWhinney H., Grady E.,
Peterson T., Cocke D.L., (2007), Arsenic removal by
electrocoagulation using combined Al–Fe electrode
system and characterization of products, Journal of
Hazardous Materials, 139, 220-231.
Khandegar V., Saroha K.A., (2013), Electrocoagulation for
the treatment of textile industry effluent - A review,
Journal of Environmental Management, 128, 949-963.
Khemis M., Leclere J.P., Tanguy G., Valentin G., Lapicque
F., (2006), Treatment of industrial liquid wastes by
1528
electrocoagulation: experimental investigation and an
overall interpretation model, Chemical Engineering
Science, 61, 3602-3609.
Koparal A.S., Ogutveren U.B., (2002), Removal of nitrate
from water by electroreduction and electrocoagulation,
Journal of Hazardous Materials, 89, 83-94.
Kobya M., Can O.T., Bayramoglu M., (2003), Treatment of
textile wastewaters by electrocoagulation using iron
and aluminum electrodes, Journal of Hazardous
Materials, 100, 163-178.
Koren J.P.F., Syversen U., (1995), State of the art
electroflocculation, Filtration Separation, 32, 153-156.
Lin S.H., Chen M.L., (1997), Treatment of textile
wastewater by chemical methods for reuse, Water
Research, 31, 868-876.
Lin S.H., Peng C.F., (1996), Continuous treatment of textile
wastewater by combined coagulation, Water Research,
30, 587-592.
Lin S.H., Shyu C.T., Sun M.C., (1998), Saline wastewater
treatment by electrochemical-method, Water Research,
32,1059-1067.
Lee W.J., Pyun S.I., (1999), Effects of hydroxide ion
addition on anodic dissolution of pure aluminum
chloride ion-containing solution, Electrochimica Acta,
44, 4041-4049.
Maretto M., Blanchi F., Vignola R., Canepari S., Baric M.,
Iazzoni R., Tagliabue M., Papini M.P., (2014),
Microporous and mesoporous materials for the
treatment of wastewater produced by petrochemical
activities, Journal of Cleaner Production, 44, 22-34.
Martinez-Huitle C.A., Brillas E., (2009), Decontamination
of wastewaters containing synthetic organic dyes by
electrochemical methods: A general review, Applied
Catalysis B: Environmental, 87, 105-145.
Mollah M.Y.A., Schennach R., Parga J.R., Cocke D.L.,
(2001), Electrocoagulation (EC)-science and
applications, Journal of Hazardous Materials, 84, 29-
41.
Mollah M., Pathak S., Patil P., Vayuvegula M., Agrawal T.,
Gomes J., Kesmez M., Cocke D., (2004), Treatment of
orange II azo-dye by electro coagulation technique in a
continuous flow cell using sacrificial iron electrodes,
Journal of Hazardous Materials, 109, 165-171.
Mollah M.Y.A., Morkovsky P., Gomes J.A.G., Kesmez M.,
Parga J., Cocke D.L., (2004), Fundamentals, present
and future perspectives of electrocoagulation, Journal
of Hazardous Materials, 114, 199-210.
Muthukumar M., Karuppiah P.T., Bhaskar Raju G., (2007),
Electrochemical removal of Cl Acid orange 10 from
aqueous solution, Separation and Purification
Technology, 55 ,198-205.
Navarro-Solis I., Villalba-Almendra L., Alvarez-Gallegos
A., (2010), H2 production by PE Melectrolysis, assisted
by textile effluent treatment and a solar photovoltaic
cell, International Journal of Hydrogen Energy, 35,
10833-10841.
Oliveira E., Santelli R., Casella R., (2005), Direct
determination of lead in produced waters from
petroleum exploration by electrothermal atomic
absorption spectrometry X-ray fluorence using Ir-W
permanent modifier combined with hydrofluoric acid,
Analytica Chimica Acta, 545, 85-91.
Pajootan E., Arami M., Bahrami H., (2016), Optimization of
the combined UV/electrocoagulation process for dye
removal from textile wastewater using Response
Surface Methodology, Environmental Engineering and
Management Journal, 15, 189-198.
Pansares F., Martinez F., Jimenez C., Saez C., Rodrigo
M.A., (2008), Coagulation and electrocoagulation of
 Efficiency of an electrocoagulation treatment of water contaminated by hydrocarbons in a continuous mode
oil-in-water emulsion, Journal of Hazardous Materials,
151, 44-51.
Pop A., Baciu A., Wanko G., Bodor K., Vlaicu I., Manea F.,
(2017), Assessment of electrocoagulation process.
Application in arsenic removal from drinking water,
Environmental Engineering and Management Journal,
16, 821-828.
Pouet M.F., Grasmick A., (1995), Urban wastewater
treatment by electrocoagulation and flotation, Water
Research, 31, 275-283.
Ramirez E.R., (1981), Physicochemical Treatment of
Rendering Wastewater by Electro Coagulation, Proc.
of 36th Purdue Industrial Waste Conference, Lafayette,
Indiana, 437-446.
Rastegar S.O., Mousavi S.M., Shojaosadati S.A., Sheibani
S., (2017), Kinetic constants determination of
petroleum refinery effluent treatment in a UASB
reactor using RSM, Environmental Engineering and
Management Journal, 16, 121-130.
Saeedi M., Khalvati-Fahlyani A., (2011), Treatment of oily
wastewater of a gas refinery by electrocoagulation
using aluminum electrodes, Water Environment
research, 83, 256-264.
Valero D., Ortiz J.M., Exposito E., Montiel V., Aldaz A.,
(2008), Electrocoagulation of a synthetic textile
effluent powered by photovoltaic energy without
batteries: direct connection behavior, Solar Energy
Materials and Solar Cells, 92, 291-297.
Vik E.L., Carlson D.A., Ekum A.S., Gjessing E.T., (1984),
Electrocoagulation of potable water, Water Research,
18, 1355-1360.
Xu X., Zhu X., (2004), Treatment of refectory oily
wastewater by electrocoagulation process,
Chemosphere, 56, 889-894.
Yehya T., Favier L., Kadmi Y., Audonnet F., Fayad N.,
Gavrilescu M., Vial C., (2015), Removal of
carbamazepine by electrocoagulation: investigation of
some key operational parameters, Environmental
Engineering and Management Journal, 14, 639-645.
Yildiz Y., Koparal A., Keskinler B., (2008), Effect of initial
pH and supporting electrolte of the treatment of water
containing high concentration of humic substances by
electrocoagulation, Chemical Engineering Journal,
138, 63-72.
Zhang S., Zhang J., Wang W., Li F., Cheng X., (2013),
Removal of phosphate from landscape water using an
electrocoagulation process powered directly by
photovoltaic solar modules, Solar Energy Materials
and Solar Cells, 117, 73-80.
1529
1530
 Environmental Engineering and Management Journal July 2018, Vol.17, No. 7, 1531-1544
http://www.eemj.icpm.tuiasi.ro/; http://www.eemj.eu

“Gheorghe Asachi” Technical University of Iasi, Romania

A BRIEF REVIEW ON RECENT ADVANCES

IN AIR-CATHODE MICROBIAL FUEL CELLS

Pritha Chatterjee1, Makarand Madhao Ghangrekar1∗, Donal Leech2

1Department of Civil Engineering, Indian Institute of Technology, Kharagpur – 721302. India
2School of Chemistry, National University of Ireland Galway, University Road, Galway, Ireland

Abstract

The performance of cathode affects the overall performance of a microbial fuel cell (MFC), and cathode configuration affects the
fabrication cost of this bioelectrochemical system in a large way. Various modifications in cathode structures have been observed
in MFCs. Among them air-cathode MFCs are considered to be the most efficient and sustainable option due to elimination of
aeration, as required in aqueous cathodes which consumes energy and increases operating cost of MFCs. An air cathode usually
consists of (a) a conductive base material, (b) current collectors, (c) catalyst layer, (d) binder layer, and (e) diffusion layer. In this
article the recent advances made in fabrication of air-cathode MFC are reviewed. The advantages and disadvantages of different
materials used for preparation of an air-cathode are analyzed. On-field application of MFC is limited even after a decade of
extensive research. This article also presents the challenges incurred in scaling-up of an MFC for real field application. In addition,
this review will assist to realize the advantages and disadvantages of different materials used in construction of air-cathodes. This
knowledge will help in intelligent selection of materials for fabrication of air-cathode MFC in future and will further aid to achieve
a competitive technology for electricity harvesting while treating wastewater.

Key words: air-cathode, binder, catalyst, current collector, microbial fuel cell

Received: January, 2014; Revised final: July, 2014; Accepted: September, 2014; Published in final edited form: July 2018

1. Introduction concluded that MFCs are capable of treating many

A microbial fuel cell (MFC) represents a
technology that has the potential to address the
problems of energy crisis and environmental pollution
at the same time, as energy is recovered in the form of
direct electricity for onsite use during treatment of
wastewaters (Angenent et al., 2004; Pant et al., 2016).
Habermann and Pommer (1991) did a pioneering work
in using MFC as a wastewater treatment system. A
MFC is a bioreactor incorporated with
electrochemical system that converts chemical energy
into electrical energy (Du et al., 2007). The organic
matter from the wastewater is converted to carbon
dioxide in the anodic chamber under anaerobic
conditions. Pant et al., (2010a) reviewed the different
types of substrates that can be used in MFCs and
complex wastes like pharmaceutical waste,
biorefinery waste etc. During this oxidation of organic
matter, the electrons released reduce the anode, and
migrate through the external circuit towards the
cathode. The device intended to be powered is
attached to this external circuit. Protons move from
anodic chamber to cathodic chamber through a proton
exchange membrane. On the cathode the oxidant,
generally oxygen is reduced to form water. Thus,
carbon neutral energy liberation in the form of direct
electricity is possible using wastewater as a fuel in
MFC.
Although there has been marked improvement
in the performance of MFC through research, large
scale practical applications of MFCs are yet to be
achieved. The main constraints for practical

∗
Author to whom all correspondence should be addressed: e-mail: E-mail: ghangrekar@civil.iitkgp.ac.in; Phone: +91-3222-283440; Fax: +91-
3222-282254
 Chatterjee et al./Environmental Engineering and Management Journal 17 (2018), 7, 1531-1544
application of MFCs are high fabrication cost and low
energy conversion efficiency (Li et al., 2010; Oliveira
et al., 2016; Zhang et al., 2010b). Rozendal et al.,
(2008) estimated capital costs for fabrication of
MFCs, based on materials currently being used in the
laboratory. They showed that the price of electrode
materials and separator can account for the 57% of
total capital cost. Hence to reduce the capital cost, it is
necessary to design efficient electrodes and
search/synthesize cheaper electrode and separator
materials. The material selected for separator should
be effective in proton transfer, should minimize
substrate diffusion from anode to cathode and also
oxygen diffusion from cathode to anode (Alvarez-
Gallego et al., 2012). The material selected for
electrode should offer higher specific surface area, and
cathode material should have lower oxygen reduction
overpotential (Sevda et al., 2013).
There are a number of review articles
describing various aspects of MFC. However, most of
them are based on particular topics like non-platinum
based catalyst (Ben Liew et al., 2014), nano-carbon as
electrode material (Ghasemi et al., 2013a), graphene
based electrodes (Wang et al., 2013), anode
architecture (Kumar et al., 2013), and overall
development of MFC (Bullen et al., 2006; Du et al.,
2007; Logan et al., 2006a; Oliveira et al., 2013a;
Rabaey and Verstraete, 2005). Power output of a MFC
can be increased by improving performance of the
cathode (Liu and Logan, 2004). Researchers have thus
focused on the performance of cathode as the most
challenging aspect of MFC design (Zhang et al.,
2011b). A MFC configuration consisting of an anode
and a cathode chamber separated by a proton exchange
membrane (PEM) is difficult to implement on a large
scale (Kim et al., 2009). Therefore, researchers are
focusing on simpler reactor design to overcome this
problem. For example, single chamber air cathode
MFCs have been studied, where the cathode does not
require placement in water (Ahmed et al., 2012;
Cheng et al., 2006; Duteanu et al., 2010; Feng et al.,
2011; Kim et al., 2009; Liu and Logan, 2004; Logan
et al., 2006b). Elmekawy et al., (2013) reviewed the
different challenges and potential opportunities of
microfluidic microbial fuel cells and mentioned that
these micro-sized MFCs offered lower internal
resistance. In this article, the recent advances in air
cathode MFCs are explored. To overcome the
difficulties in scaling up of MFCs, the challenges that
need to be addressed for practical implementation are
also discussed.
2. Single chamber MFC: Advantages and
disadvantages
In an attempt to minimize the operating cost of
MFCs air breathing single chambered MFCs were
developed by researchers. The major reaction that
takes place at the cathode is reduction of oxygen. The
solubility of oxygen (mole fraction basis) in water is
only 4.6*10-6 at 25 °C, while that in air it is 0.21
(Logan, 2008; Wei et al., 2011). Because of this,
1532
oxygen reduction reaction at aqueous cathode is
restricted due to limited availability of oxygen. The
single-chambered air cathode offers several
advantages over two-chambered systems such as: (a)
the catholyte does not need to be aerated, oxygen in
the air can directly react at the electrode, (b) recycling
or chemical regeneration of the catholyte is not
required, (c) higher volumetric power density is easily
achievable due to smaller cell volume (Liu and Logan,
2004), and (d) the overall volume and footprint of the
MFCs will be less in air cathode configuration due to
elimination of a cathodic chamber.
In spite of these advantages offered by air-
cathode MFCs, there are certain disadvantages related
mostly to the materials used in construction. The
major disadvantages associated with air-cathode
MFCs are (a) fouling of the cathode surface exposed
to air, (b) diffusion of oxygen into anodic chamber, (c)
electrolyte loss due to evaporation (Wei et al., 2011),
(d) necessity of a catalyst on cathode for enhancing
oxygen reduction reaction (Freguia et al., 2007), and
(e) requirement of a binder. This review attempts to
elaborate the advantages and disadvantages of
different materials used in construction of air-
cathodes. This knowledge will help in intelligent use
of materials for making different components of air
cathode MFC in future.
3. Components of air cathode
The separator has anolyte on one side and it is
exposed to air on the other side. An air cathode usually
consists of (a) a conductive base material, (b) current
collector, (c) catalyst layer, (d) binder layer, and (e)
diffusion layer (Wei et al., 2011).
3.1. Conductive base materials
The base material for the electrode provides a
conductive path to the external circuit. To function as
a good electrode, a material should have the following
properties namely (a) good conductivity, (b) high
chemical stability, (c) high mechanical strength, and
(d) low cost (Sharma et al.). Carbon in paper or cloth
form is the most widely used material in air cathodes
(Ahmed et al., 2012; Cheng et al., 2006; Martin et al.,
2011; Shi et al., 2012; Sun et al., 2009; Zhang et al.,
2009b). Although carbon is not very conductive, its
stability and lack of reactivity makes it a suitable
electrode material. However, carbon cloth is
expensive (Zhang et al., 2010a) and longtime stability
of carbon paper is doubtful due to the brittleness of the
material.
Electrodes made of metals and alloys such as
stainless steel (SS) can also be used. For example,
Zhang et al. (2011b) constructed a cathode by
impregnating carbon powder around a stainless steel
mesh and obtained a power density of 1616 mW/m2.
Simple or modified carbon powders are also being
increasingly used as electrode materials (Duteanu et
al., 2010). Zhuang et al., (2010) applied a nickel (Ni)
based paint and manganese dioxide (MnO2) catalyst to
 A brief review on recent advances in air-cathode microbial fuel cells

carbon cloth and obtained a volumetric power density
of 11 W/m3. This carbon cloth served as both a
membrane and the base material for the cathode. It
showed a high proton diffusion rate, reduced the
internal resistance and gave higher power output. Sun
et al., (2009) obtained a power density of 878 mW/m3
using a platinum catalyst on carbon paper as cathode.
Martin et al., (2011) used a platinum (Pt) catalyst on
carbon paper cathode and obtained a power density of
90 W/m3.
This suggests, in spite of same cathodic
material used, the power output obtained is different
due to several reasons like differences in the MFC
configuration, catalyst loading, anolyte composition,
mode of operation, temperature, pH etc. Performance
of some air cathode MFCs using different materials
for cathodes are listed in Table 1 through Table 4. An
air cathode MFC (Fig. 1) is prepared by impregnating
the conductive electrode material, with or without
catalyst, to enhance oxygen reduction reaction on the
exposed surface of membrane separator, provided
with an appropriate current collector.
3.2. Current collectors
Although carbon based electrodes are
extensively used in MFCs, these materials have very
low electrical conductivity. This can contribute to
increased ohmic losses in larger anodic volume
(Zhang et al., 2010b). Rozendal et al., (2008) reported
that the electrical resistance of graphite is 1375 µΩ
cm, while that of the metal titanium is 42 µΩ cm.
Inclusion of a metal current collector on the cathode
surface can ensure proper distribution of current and
also reduces the ohmic losses (Zhang et al., 2010b).
Zhang et al., (2011b) used commercially available
stainless steel (SS) mesh with varying characteristics
as current collectors and found that MFC performance
varies as a function of the current collector. The
performance of MFCs using different current
collectors is compared in Table 1. For example, Zuo
et al. (2008) pressed a SS mesh against the cathode and
the internal resistance decreased from 47 Ω (without
current collector) to 29 Ω.

Fig. 1. Schematic representation of an air-cathode MFC

Table 1. Performance comparison of MFCs using different current collectors

Current Diffusion Max. power Prepared

Cathode Anode Membrane Binder Catalyst
collector layer density by
Zhang et
Carbon Graphite Not Not 2 layers of 1616 ± 25
SS mesh Pt al.,
powder fiber brush mentioned mentioned PDMS mW/m2
(2011b)
Activated Zhang et
Graphite Not 1 layer of
carbon Ni mesh PTFE Pt/Pd 1415 mW/m2 al.,
fiber brush mentioned PTFE
powder (2009b)
574 ± 51
Graphite SS mesh
Graphite Not mW/m2 Zuo et al.,
paint on Nafion Nafion CoTMPP
fiber brush No mentioned 449 ± 35 (2008)
AEM
collector mW/m2

1533
 Chatterjee et al./Environmental Engineering and Management Journal 17 (2018), 7, 1531-1544
The maximum power produced increased by
28% compared to the same system without the mesh.
Other metals like titanium (Duteanu et al., 2010) and
nickel (Zhang et al., 2009b) have also been used. Air
cathode MFCs employing more uniformly spread
current collector like a sock net give more consistent
and dependable results for future use, and it is reported
to produce power density of 3.2 times higher than the
MFC having similar configuration but provided with
SS wire as current collector (Chatterjee and
Ghangrekar, 2014).
3.3. Catalyst
Use of an air cathode eliminates the problem of
low aqueous solubility of oxygen, by direct diffusion
of oxygen to the cathode. The oxygen reduction
reaction (ORR) however has a high overpotential for
reduction to water (Freguia et al., 2007). To overcome
this problem catalysts are often used on cathode
surface. Use of catalyst on the cathode helps in: (a)
increasing the affinity of the cathode to oxygen, and
(b) decreasing the activation energy of ORR (Duteanu
et al., 2010). A catalyst layer is applied on the surface
of cathodes by spray paint gun or brush. The catalysts
generally used in air cathode MFC can be grouped into
three types: noble metal catalyst, non-noble metal
catalyst and activated carbon powder.
3.3.1. Noble and non-noble metal catalysts
Platinum (Pt) is the most widely used cathode
catalyst; however, it is very costly and moreover it has
poor ORR kinetics in neutral pH and at lower
temperature (Duteanu et al., 2010). Some other
cheaper catalysts like Co-tetra-methyl
phenylporphyrin (CoTMPP) (Zuo et al., 2008), iron
phthalocyanine (FePc) (Cheng et al., 2006; Sun et al.,
2009) and manganese dioxide (Zhang et al., 2009b;
Zhuang et al., 2010a) are also used. Zuo et al., (2007)
showed that a tube cathode MFC with two CoTMPP
coated tubes produced slightly less power (8.8 W/m3)
than that obtained with a carbon paper cathode with Pt
catalyst (9.9 W/m3) (Table 2). Zuo et al., (2008)
measured the performance of CoTMPP as catalyst and
found that both power and Coulombic efficiency (CE)
of MFC increased proportionately with the catalyst
loading. Longer operation times were needed as the
catalyst loading was decreased, allowing more oxygen
to diffuse through the cathode into the electrolyte. This
oxygen was used by bacteria for aerobic degradation
of the acetate substrate, lowering the electron recovery
and therefore decreasing the CE (Zuo et al., 2008).
Ghasemi et al., (2013b) used copper
pthalocyanin (CuPc) and Ni nanoparticles and
observed CuPc to be performing almost similar to Pt
catalyst followed by Ni nanoparticles. Use of
nanoparticles in cathode as catalyst increases the
effective surface area for ORR, thus improving its
performance. However, use of such metals like copper
or nickel needs further investigation for long-term
stability to operate MFCs for a long time. There is a
1534
possibility of leaching of these metals into the
catholyte and thus creating toxicity for the microbial
population hence reducing the opportunity and
advantage of biocathode formation in the MFC. A new
type of binder-free gas diffusion electrode made of
cobalt oxide (Co3O4) micro-particles directly grown
on stainless steel mesh (SSM), by using an ammonia-
evaporation-induced method, demonstrated improved
performance in terms of electrocatalytic activity,
selectivity, durability and economics toward ORR in
pH-neutral solution, in comparison with conventional
carbon supported platinum catalyst (Gong et al.,
2014).
3.3.2. Activated carbon powder
Carbon is generally considered an inert
material; however, it has some superficial activity due
to the presence of heteroatoms such as oxygen,
hydrogen, sulphur, nitrogen, etc. The activity of
carbon can be increased by optimizing the proportion
of these atoms on the carbon surface. Activated carbon
powder thus produced can be used in place of other
ORR catalysts (Duteanu et al., 2010; Harnisch et al.,
2009; Pant et al., 2013; Pant et al., 2010b; Zhang et al.,
2009a; Zhang et al., 2011a). These non-metal catalysts
are more stable than the commonly used platinum (Shi
et al., 2012). An air cathode MFC using N-doped
carbon powder prepared by nitric acid (HNO3)
refluxing after heat treatment and acid treatment was
prepared by Shi et al. (2012) and their performance
was compared to a MFC using Pt catalyst on cathode.
Threefold increase in power, comparable with
platinum catalyzed cathode, is reported using treated
carbon as compared to the untreated carbon. Besides
nitric acid refluxing, N-doped carbon powder can also
be synthesized by chemical vapor deposition (Feng et
al., 2011) and sol-gel polymerization (Jin et al., 2011).
Dong et al., (2012) prepared an air cathode by rolling
activated carbon (AC) and polytetreafluoroethylene
(PTFE), combined in different ratios. Maximum
power density was obtained for AC to PTFE ratio of 6
(802 mW/m2) and lowest power density was obtained
for a ratio of 11 (584 mW/m2). Duteanu et al., (2010)
prepared an air cathode by coating a polymer
electrolyte membrane with acid treated carbon powder
and compared its performance in MFCs using
cathodes prepared by coating the same membrane with
untreated carbon with and without Pt catalyst; and
comparable performance in the MFC using acid
treated carbon powder to the Pt/C cathode was
reported. At lower current densities the Pt/C cathode
performed better, while the acid treated carbon
powder cathode had a comparable performance with
Pt/C electrode at higher current densities.
Recently, a porous nitrogen-doped carbon
nanosheet was used as an alternative to Pt catalyst in a
MFC (Wen et al., 2014), which gave a higher power
density than a Pt catalyst. A higher volumetric power
obtained with polypyrrole (PPy) compared with
conventionally used Vulcan XC exhibits that PPy is a
better conductive support to ORR (Khilari et al.,
 A brief review on recent advances in air-cathode microbial fuel cells

2014). Cost effective manganese dioxide material for the replacement of expensive Pt in
nanotube/graphene composites are reported to constructing large-scale MFC for wastewater
demonstrate high ORR activity and high power treatment and bioelectricity production (Khilari et al.,
generation ability, hence making it a potential cathode 2013a) (Table 2).

Table 2. Performance of MFCs using different cathode catalyst

Current Diffusion Max. power Prepared

Cathode Anode Membrane Binder Catalyst
collector layer density by
Carbon No catalyst 102 mW/m3
cloth 466 ±25
MnO2 Zhang et
coated Graphite Not Not mW/m3
CEM PVDF al.,
with granules mentioned mentioned
726 ± 25 (2009b)
graphite Pt
mW/m3
paint
Zhuang
Ni based Graphite Not Not
Carbon cloth PTFE MnO2 11 W/m3 et al.,
paint granules mentioned mentioned
(2010)
Sun et
Carbon Carbon Microfiltration Not Not
PTFE Pt 878 mW/m3 al.,
paper paper membrane mentioned mentioned
(2009)
Graphite Zuo et
Graphite Not Not
fiber CEM CoTMPP 449 mW/m3 al.,
paint mentioned mentioned
brush (2008)
CoTMPP 8.8 mW/m3 Zuo et
Graphite Carbon Ultrafiltration Not Not
Nafion al.,
paint paper membrane mentioned mentioned Pt 9.9 mW/m3
(2007)
Carbon
Pt 480 mW/m2
cloth
Cheng et
Carbon Carbon Not Not
Not mentioned Nafion CoTMPP 369 mW/m2 al.,
cloth cloth mentioned mentioned
(2006)
Carbon
No catalyst 93 mW/m2
cloth
554 ± 27
C-CoOx
mW/m2
412 ± 20
C-FePc Ahmed et
Carbon Carbon Not Not Not mW/m2
Not mentioned al.,
cloth cloth mentioned mentioned mentioned 654 ± 32
C-CoOx-FePc (2012)
mW/m2
Carbon 271 ± 13
powder mW/m2
Untreated
310.8 ± 15
carbon
mW/m2
powder
Carbon Carbon Not 4 PTFE Treated Shi et al.,
Not mentioned Nafion 934.7 ± 5
cloth brush mentioned layers carbon (2012)
mW/m2
powder
980.5 ±7
Pt/C
mW/m2
Pt 90 W/m3
Martin et
Carbon Carbon Not Mn2O3 32 W/m3
J-cloth Nafion PTFE al.,
paper felt mentioned Fe2O3 15 W/m3
(2011)
C 8 W/m3
PTFE+ Dong et
Carbon Not Not
SS mesh Not mentioned Carbon AC + PTFE 802 mW/m2 al.,
mesh mentioned mentioned
powder (2012)
HNO3 treated
170 mW/m2
Poly- carbon Duteanu
Carbon Carbon
electrolyte Ti mesh Nafion PTFE Untreated et al.,
powder felt 51 mW/m2
membrane carbon (2010)
Pt 217 mW/m2
37.9
Carbon black
mW/m2 Ghasemi
Carbon Carbon Not Not Not
Nafion 117 94.4 et al.,
powder Paper mentioned mentioned mentioned Ni/C
mW/m2 (2013b)
Pc/C 56 mW/m2

1535
 Chatterjee et al./Environmental Engineering and Management Journal 17 (2018), 7, 1531-1544

118.2
CuPc/C
mW/m2
120.8
Pt
mW/m2
1159.34 Wen et
PNCN
Carbon Carbon Not mW/m2 al. (2014)
Not mentioned SS net PTFE
powder brush mentioned 858.49
Pt
mW/m2
21.2
Pt
Carbon 4 layer mW/m3 Gong et
SS mesh Not mentioned SS net Nafion
cloth Teflon 17.8 al. (2014)
Co3O4
mW/m3
6.11
MnCo2O4/PPy
mW/m3
Carbon Carbon Not Not Not 4.22 Khilari et
Nafion MnCo2O4/C
cloth cloth mentioned mentioned mentioned mW/m3 al. (2014)
1.77
C
mW/m3

3.4. Binder sulphonate groups to the active sites on the catalyst

To apply the catalyst layer properly to an
electrode surface or to apply carbon powder on the
separator surface, with or without catalyst, a binder is
required. Polymer such as Nafion (Cheng et al., 2006;
Duteanu et al., 2010; Martin et al., 2011; Shi et al.,
2012; Zhang et al., 2009b; Zhuang et al., 2010; Zuo et
al., 2007) is the most commonly used binder. Nafion
is most suitable for use as a binder due to its high
proton conductivity (Zhang et al., 2012). The high cost
of Nafion ($ 667/m2) has resulted in research to find
cheaper alternatives to it (Zhang et al., 2012), such as
polytetrafluoroethylene (PTFE) (Cheng et al., 2006;
Zuo et al., 2008), polyvinylidene fluoride (PVDF)
(Zhang et al., 2009b), polydimethyl siloxane (PDMS)
(Zhang et al., 2012; Zhang et al., 2010a), etc. Cheng et
al., (2006) found that power density in MFCs
decreases with use of PTFE in place of Nafion, but
observed a more stable performance in the MFC using
the PTFE binder. Wang et al., (2010) mixed PTFE and
Nafion in different ratios and used them as binders;
and observed increased power density with an increase
in percentage of Nafion in the mixture. Nafion is a
polymer, which is conductive to protons having both
hydrophilic and hydrophobic domains along with a
transition zone. On the other hand, PTFE is a highly
hydrophobic material that can reduce electrolyte loss
from the anode chamber and have good resistance to
oxygen transfer across the material. It is however not
an electrolyte. As PTFE is highly hydrophobic it also
causes the cathode catalyst to become too dry, thus
limiting effective proton transfer (Cheng et al., 2006).
Zhang et al., (2009b) used a polyvinylidene
fluoride (PVDF) binder and obtained a maximum
power density of 726 mW/m3 using Pt catalyst (Table
2). Saito et al., (2010) explored the possibility of using
non-ionic hydrophobic polyphenylsulphone (PPS) as
binder of cathode catalyst. The PPS material was
sulphonated to different degrees and best performance
was reported in the MFC with the non-ionic non-
sulphonated binder. It was inferred that the presence
of sulphonate groups in the binder reduced the oxygen
reduction capacity of the cathodes by adsorption of the
and also by hampering proton diffusion. Zhang et al.,
(2012) examined hydrophobic polydimethylsiloxane
(PDMS) as an anti-flooding catalyst binder in MFC
cathodes.
Although all these binders used as alternatives
to Nafion initially had a lower performance, they had
almost equivalent performance after long-term use
(Cheng et al., 2006; Saito et al., 2010; Wang et al.,
2010; Zhang et al., 2012). Moreover, as Nafion is an
expensive material it may bring economic limitation
in full-scale application of MFC systems. Although
the cheaper binders, such as PDMS, produce about
23% less power compared to Nafion, their use will
reduce the production cost of MFC, for example
PDMS costs around only 0.23% of that of Nafion
(Zhang et al., 2012).
3.5. Diffusion layer
A hydrophobic layer is usually applied to the
air breathing side of the cathode to reduce diffusion of
oxygen into the anode chamber and also to reduce
electrolyte loss from the anode chamber due to
evaporation (Wei et al., 2011). Cheng et al., (2006)
obtained higher CE and power densities after applying
successive layers of PTFE as a diffusion layer. This
reduced electrolyte loss from the anode chamber.
Whilst successive addition of diffusion layers
increased the CE, the power density diminished after
a certain number of layers due to lower availability of
oxygen at the cathode. Power density was reported to
be increased by 42% and CE by 200% compared to
that for commercially available cathodes.
Zhang et al. (2010a) used PDMS as a diffusion
layer. It is stable, hydrophobic and less expensive than
PTFE. Maximum power density of 1635 mW/m2 was
reported with three layers of PDMS and a carbon cloth
cathode. Generic plastic wrap was also used as
diffusion layer by Hays et al., (2011) to obtain a power
density of 150 mW/m2. They found a problem that a
liquid layer was formed between the cathode and the
diffusion layer. This liquid layer reduced oxygen mass
transfer from gaseous phase to liquid phase, thereby
giving a lower performance of the MFCs.

1536
 A brief review on recent advances in air-cathode microbial fuel cells

Performances of MFCs using different binders
are summarized in Table 3. Best performance (766
mW/m2) was observed using four diffusion layers on
the cathode surface (Table 4). Zhang et al. (2011b)
compared the long-term performance and stability of
MFCs having cathodes with diffusion layers of
varying porosity. They observed that although initially
the power production was greater in the MFC with
higher porosity of diffusion layer, after one year its
performance decreased by 40%, unlike MFCs with
lower diffusion layer porosity whose performance
decreased by only 20%. Liu et al., (2011) prepared a
novel air cathode MFC using a hydrophobic double-
sided cloth as the base material for cathodes without
any diffusion layer. This decreased the cost and
simplified the manufacturing of cathodes. However,
long-term stability of this configuration needs to be
evaluated.

Table 3. Performance of MFCs under different binders

Diffusion Max. power

Cathode Anode Separator Binder Catalyst Prepared by
layer density
Carbon Nafion 1790 mW/m2
Graphite 4 PTFE Saito et al.,
powder on Carbon cloth Pt
fiber brush PPS layers 1660 mW/m2 (2010)
carbon cloth
Carbon Carbon Nafion Not 480 mW/m2 Cheng et
Carbon cloth Pt
powder cloth PTFE mentioned 360 mW/m2 al., (2006)
Nafion 2220 mW/m2
Cu mesh
Graphite PDMS 1710 mW/m2 Zhang et
Pt
fiber brush Nafion 1840 mW/m2 al., (2012)
SS mesh
PDMS 1680 mW/m2
PDMS 1060 mW/m2
PTFE 549 mW/m2
Carbon Not Nafion + 4 PTFE Wang et al.,
Carbon cloth Pt/C 685 mW/m2
cloth mentioned PTFE (3:2) layers (2010)
Nafion +
844 mW/m2
PTFE (2:1)

Table 4. Performance of MFCs under different diffusion layers

Max. power
Cathode Anode Membrane Binder Diffusion layer Catalyst Prepared by
density
2 layers of 150 ± 22
plastic wrap mW/m2
Activated Carbon Pulp laminated Not 1 layer of Not 110 ± 14 Hays et al.,
carbon mesh glass fiber mentioned plastic wrap mentioned mW/m2 (2011)
140 ± 6.4
PDMS
mW/m2
1592 ± 19
1 PDMS layer
mW/m2
1610 ±
SS mesh 2 PDMS layer
56mW/m2
Graphite Not 1010 Zhang et al.,
Not mentioned 3 PDMS layer Pt
fiber brush mentioned mW/m2 (2010b)
1553 ± 19
1 PDMS layer
Carbon mW/m2
cloth 1635 ± 62
3 PDMS layer
mW/m2
4 PTFE layer 766 mW/m2
633 ± 7
Carbon Carbon 8 PTFE layer Cheng et al.,
Carbon cloth Nafion Pt mW/m2
powder cloth (2006)
538 ± 6
No layer
mW/m2
70 % porous 734 ± 18
Activated Graphite Not layer of PTFE Not mW/m2 Zhang et al.,
Not mentioned
carbon fiber brush mentioned 30 % porous mentioned 789 ± 68 (2011b)
layer of PTFE mW/m2
Double 0.7 ± 0.02
No layer Pt-C
sided cloth Carbon Not mW/m2 Liu et al.,
Not mentioned
Carbon cloth mentioned 0.66 ± 0.01 (2011)
4 PTFE layers Pt-C
cloth mW/m2
Carbon Carbon Micro-fiber Not Polyolefine Tugtas et
Pt 750 mW/m2
cloth cloth cleaning cloth mentioned membrane al., (2011)

1537
 Chatterjee et al./Environmental Engineering and Management Journal 17 (2018), 7, 1531-1544
Tugtas et al., (2011) used cheaper spun-
bonded polyolefin sheet as a diffusion layer in place
of PTFE and obtained a power density of 750 mW/m2
using a carbon cloth cathode. Successful application
of such cheaper material will contribute in bringing
down the production cost of MFC and enhancing its
economic viability for full-scale application.
4. Membrane Electrode Assembly (MEA)
In a conventional MFC the anodic and cathodic
compartments are separated by a membrane and
respective electrodes are placed in these
compartments. This increases the distance between
cathode and anode and simultaneously increases the
length of the path required to be traversed by a proton
thus increasing the internal resistance of the system
(Wei et al., 2011). Construction of MFCs where this
distance can be minimized may therefore provide for
devices with higher performance. Reactor designs
with electrodes placed on either side of the membrane
are known as membrane electrode assembly (MEA)
MFCs (Hays et al., 2011; Zhang et al., 2011b).
Implementation of MEA-MFCs have improved the
overall reactor performance and reduced the capital
cost considerably (Kim et al., 2009; Liu and Logan,
2004; Zhuang et al., 2010).
MEAs are generally constructed by the
following two steps: (a) hot pressing the membrane
with the elctrodes to obtain close contact between the
membrane and the electrodes (Prakash et al., 2010;
Zhuang et al., 2010), and (b) applying a conductive
layer, with or without catalyst, to the surface of the
separator by a spray paint gun (Zhuang et al., 2010).
Catalysts are usually applied on that side of the
cathode which is facing the membrane (Prakash et al.,
2010). Complete contact between the electrodes and
the membrane is a strict requirement for MEA
constructed by this method, which makes it difficult
for implementation of larger volume MFCs in field
due to limitations of substrate diffusion (Zhuang et al.,
2010). Moreover, this method is only suitable for
Nafion-electrode assemblies or materials that can
withstand high temperature and not for other cheaper
alternatives where thermal degeneration might occur
(Kim et al., 2009). This increases the cost of the entire
setup as we try to scale up the MFCs. To avoid this
problem Zuo et al., (2007) constructed a new type of
membrane electrode assembly by coating a membrane
with graphite paint and furthur coating it with a
catalyst. Duteanu et al., (2010) also used the same
procedure to prepare a membrane electrode assembly.
Kim et al., (2009) applied a layer of hydrophilic
hydrogel between the membrane and the cathode
which improved the performance by increasing
hydration and maintaining proper contact between the
membrane and electrode.
Zhuang et al., (2010) compared the
performance of different membrane-cathode
assemblies (MCA) and cloth-cathode assemblies
(CCA) in MFCs, to find an optimum cathode
1538
configuration for scaling up of MFC. Carbon cloth
containing catalyst was hot pressed to an ion exchange
membrane to prepare the MCAs. To prepare the CCA
a nickel based conductive paint and a catalyst was
sprayed on a carbon cloth. The CCAs were reported to
perform better in terms of power generation and COD
removal because of their higher proton transfer rate,
establishing CCA as a cost-effective option for scaling
up of MFCs. Composite membranes consisting of
polyvinyl alcohol (PVA), silicotungstic acid (STA),
and graphene oxide (GO) were prepared and evaluated
as proton-conducting membranes in air-cathode
MFCs. These synthesized PVA−STA−GO
membranes showed excellent kinetic properties, better
durability, reduced oxygen crossover and higher
power density compared to the commercially
available Nafion 117 (Khilari et al., 2013b).
Ajayi and Weigele (2012) prepared a low cost
membrane electrode assembly by coating a terracotta-
pot with graphite paint and obtained a maximum
power density of 33.13 mW/m2 using a graphite felt
anode. Efforts are being made over the past few years
to prepare a sustainable low cost wastewater treatment
system using baked earthenware as separators (Behera
et al., 2010a; Behera et al., 2010b; Jana et al., 2010).
This material is quite effective for proton transfer and
has been used successfully as a low cost substitute to
costly proton exchange membranes like Nafion,
Ultrex, etc. It provides for better wastewater treatment
and higher electricity generation than expensive
proton exchange membrane materials (Jana et al.,
2010). The cost of this earthenware membrane is 100
fold less than that of Nafion.
5. Membrane fouling
Membrane fouling is a problem in almost all
membrane processes, which needs an immediate
solution. The common fouling types are organic,
colloidal, biofouling and scaling (Vermaas et al.,
2012). Although performance of fouled membrane in
other membrane processes can be partially restored by
appropriate cleaning method, it will cause operational
difficulties in MFCs and also add to the overall cost of
the system (Ang et al., 2006; Creber et al., 2010; Kang
and Cao, 2012). The major factors that affect
membrane fouling are its physicochemical properties
like water affinity, surface roughness, and electrostatic
charge (Kang and Cao, 2012; Louie et al., 2006).
Fouling of membrane or more specifically cathode
fouling in single chamber MFCs reduces overall
power output by reducing the exposed cathode surface
area and by reducing oxygen diffusion to the active
catalytic sites. Fouling of cathode also increases the
resistance of the system (Vermaas et al., 2012).
Fouling might also cause limitations in cation transfer
thereby leading to losses in cathodic potential (Rikame
et al., 2012). Xu et al., (2012) observed a 32.3%
decrease in power production in MFC due to fouling
of the membrane. Therefore, solutions are sought to
inhibit fouling and also for removing fouling.
 A brief review on recent advances in air-cathode microbial fuel cells
Surface modifications by changing the water
affinity of the surface could be a possible option to
reduce fouling. Researchers have said that hydrophilic
membranes containing hydroxyl, carboxyl or ethylene
oxide groups are better fouling resistant because most
of the potential fouling compounds are hydrophobic in
nature and a thin water layer on the membrane surface
prevents hydrophobic foulants from getting attached
to the membrane surface (Kang and Cao, 2012; Liu et
al., 2012; Rana and Matsuura, 2010; Vermaas et al.,
2012). However, for hydrophilic foulants membrane
hydrophilicity will even increase fouling further
(Kwon et al., 2005). Chen et al., (2012) prepared an
air-cathode microbial fuel cell using hydrophilic
polyvinyl alcohol (PVA) as binder and reported better
coulombic efficiency (94%) and power density (1220
mW/m2) of the MFC compared to the one prepared
using glass fiber separator because of better fouling
mitigation properties of PVA.
Another antifouling mechanism is to decrease
the surface roughness of the membrane because
fouling particles are more likely to be attached to
rough surfaces than to smooth ones (Elimelech et al.,
1997; Sagle et al., 2009; Vermaas et al., 2012;
Vrijenhoek et al., 2001). The surface charge of the
membrane is also an important factor in controlling
fouling, as electrostatic repulsion between the
membrane surface and the foulants will reduce
membrane fouling (Vermaas et al., 2012). In an
experiment on electrodialysis Vermaas et al., (2012)
observed remnants of diatoms, clay minerals and
organic fouling at the anion exchange membrane and
scaling of calcium phosphate at the cation exchange
membrane. However, the membranes will show
distinct tendencies of fouling depending on the nature
of the foulant. Once the electrical charge of the foulant
is known this phenomenon can be easily used to
reduce fouling. On a neutral surface deposition of
foulants is less due to elimination of charge interaction
between the membrane and foulants. Also steric
repulsion due to presence of some surface-bound long-
chain hydrophilic molecule (like polyethylene glycol
etc.) prevents attachment of larger molecules of
foulants (Mcpherson et al., 1998; Nie et al., 2004;
Vermaas et al., 2012; Wang et al., 2002; Wang et al.,
2010).
Applying anionic or cationic surfactants at
suitable concentrations on hydrophobic surfaces not
only reduces bacterial attachment to the surface but
also removes previously attached bacteria on the
surface (Akuzov et al., 2013; Vermaas et al., 2012).
However, when tested on hydrophilic surfaces the
non-ionic surfactants showed no inhibition, showing
that the reduction of biofouling is not only due to the
toxicity of the surfactants but because of some
hydrophobic interaction between the membrane
surface and microbial cell. Vanillin (4-hydroxy-3-
methoxybenzaldehyde), a well-known food flavoring
agent, could be used as a potential compound that
reduces the biofilm formation on reverse osmosis
(RO) membranes (Ponnusamy et al., 2010). It is
reported that after adding Vanillin the biofilm
formation was inhibited up to 46.3% on polystyrene
surface. Vanillin was also found to be effective in
inhibition of biofouling in MFC as reported by
Chatterjee and Ghangrekar (2013). A mixture of PVA
and Vanillin was used as binder in a single chamber
air-cathode MFC, which demonstrated only 8%
decrease in power production as compared to 56%
decrease for the MFC using only PVA, after five
weeks of operation. Titanium dioxide (TiO2) being a
well-known photocatalytic agent and having a
hydrophilic behavior, use of nano TiO2 particles as
anti-fouling agent is reported in literature (Akuzov et
al., 2013). Other inorganic materials like silicon
dioxide (SiO2), Zeolite, silver nanoparticles and even
mesoporous materials improves membrane resistance
to fouling when coated on membrane surface (Kang
and Cao, 2012). Oxidative processes are important for
the cross-linking of the adhesive proteins secreted by
microorganisms to adhere to any surface and
inhibition of the oxidative processes could result in
low cohesion strength of the protein adhesive films of
microorganisms. Several antioxidants like gallic acid,
ascorbic acid, tannic acid can reduce fouling by
inhibiting their attachment to the surface (Akuzov et
al., 2013), but the feasibility of using all these agents
are yet to be tested on cathodes of MFC.
Today most researches on reducing membrane
fouling are based on introduction of a hydrophilic
layer, the reduction of surface roughness,
improvement of charge property, and utilization of
steric repulsion effect and use of biocides (Kang and
Cao, 2012). The easiest way to obtain all these
modifications of membrane surface is by surface
coating. The modifiers used might only be connected
to the membrane surface by van der Waals force of
attraction, hydrogen bonding or electrostatic
interaction, hence the antifouling properties of the
modified membrane may be gradually lost due to loss
or leaching of coating after long term use (Kang and
Cao, 2012). Certain chemical processes can also be
used to attain surface modification like using
hydrophilizing treatment by certain acids (Kulkarni et
al., 1996), radical grafting (Belfer et al., 2004),
chemical coupling (Van Wagner et al., 2011), plasma
polymerization (Zou et al., 2011), vapor deposition
(Yang et al., 2011) etc. Winfield et al. (2013) used
natural rubber as membrane and reported reduced
fouling even after operation for one year.
Researchers have obtained contradictive
results in terms of antifouling properties of
hydrophilic and hydrophobic surfaces. While some
claim hydrophilic surfaces are better fouling resistive,
others claim the opposite. Foulants can be inorganic or
organic, while inorganic fouling is resisted by
hydrophilic surfaces and it enhances biofouling,
providing a suitable surface for biotic organisms to
grow. The major research papers on fouling resistance
are dealing with water treatment and not with
wastewater treatment. The properties of foulants in
water and wastewater are different. While inorganic
materials have to be majorly dealt with while treating
drinking water, the major culprit in membrane based
1539
 Chatterjee et al./Environmental Engineering and Management Journal 17 (2018), 7, 1531-1544
wastewater treatment technologies is biological
growth. Hydrophobic surfaces by their water resisting
properties do not support biological growth; however,
they are not effective in reducing deposition of
inorganic foulants. Using a biocide on a hydrophilic
surface seems to be the solution to resist inorganic
fouling as well as biofouling.
6. Challenges
The main barrier to wide-scale implementation
of MFC technology is the issue with scaling it up for
real implementation in field; hence it is necessary to
develop cost-effective materials for fabrication of the
MFC cathode. With increase in the anodic chamber
volume requiring higher surface area of the electrode,
the conductive resistance of normal carbon electrode
increases resulting in higher ohmic losses. To avoid
this, metal current collectors on the surface of the
cathode are often used, as mentioned earlier. More
efficient design of current collector is necessary to
ensure proper contact between the cathode material
and the current collector so that no losses could occur.
Any loose contact or gaps between the collector and
the cathode will lead to current loss, which will affect
the power generated from the MFC.
The performance of gas diffusion layers also
needs further investigation. The materials generally
used for diffusion layer are hydrophobic, which
renders it non-conductive. This hampers the overall
performance of the MFC. Hydrophobic materials with
some polar functional groups could be a good choice
for diffusion layer material. Its hydrophobic nature
will serve the purpose of reducing electrolyte loss and
simultaneously the presence of polar groups will
ensure better conductivity of the cathode surface.
Rozendal et al., (2008) estimated the cost of
different materials used in laboratory scale MFCs and
reported that cathode materials in air-cathode MFCs
account for 47% of the total cost of the MFC. Pant et
al., (2011) did an economic analysis of several bio-
electrochemical systems and observed that the major
components contributing to high cost of these systems
are aeration equipment, costly catalysts, membranes
and electrolyte buffers in decreasing order. To reduce
the overall cost of MFC and make them commercially
feasible and successful it is thus necessary to reduce
the cost of materials used for cathode fabrication. This
can be done by using cheaper catalysts (like copper
pthalocyanin, iron pthalocyanin, co-tetramethyl
phenyl porphyrin, cobalt oxide, nanoparticles,
activated carbon powder), binders (PTFE, PVDF,
PDMS) and conductive materials (modified carbon
powder, polypyrroles).
As mentioned earlier the challenge lies in
scaling up of MFCs. The maximum voltage that can
be obtained in a MFC is 1.1 V using acetate as
substrate in anode and oxygen as the cathodic electron
acceptor (Logan, 2008). However, the voltage
obtained from a MFC is always less than this
maximum predicted voltage owing to several losses
1540
namely activation polarization losses, concentration
polarization losses and ohmic losses (Du et al., 2007;
Logan, 2008). Since the maximum voltage that can be
obtained from a MFC is limited, hence to maintain
similar Coulombic efficiency (around 30% say) as a
laboratory scale MFC, it is necessary to maximize
current produced in the scaled up MFC. To achieve
this, suitable selection of electrode materials and
configuration is necessary to optimize current density,
and to minimize the losses that occur at higher current
densities, that is concentration polarization and mass
transfer losses. Maximum current generation depends
on the maximum rate at which bacteria can oxidize a
substrate and transfer electrons to the anode, which in
turn depends on availability of substrate to bacteria
present in biofilm. Thus, there are two limitations to
current generation: bacterial kinetic limitation and
substrate limitation. Kinetic limitation comes into play
when the rate of electron uptake by the electrodes
surpasses the rate of generation of electrons by
bacteria. Electron generation by bacteria depends on
biofilm mass, substrate utilization rate of electrogenic
bacteria, and their growth rate and substrate
availability. Biofilm mass can be increased by
increasing anode surface area; however there is a
limitation to maximum surface area that can be used
for a particular substrate concentration otherwise
substrate availability will be the rate limiting step. The
maximum value of ratio of electrode surface area by
volume of reactor is yet to be determined and depends
on a number of factors like electrode material, reactor
configuration, etc. Thus by achieving the optimum
surface area by volume ratio substrate limitation and
kinetic limitations can be surpassed. Hence, by
reducing the mass transfer losses a scaled up MFC
with higher current generation capability can be aimed
for.
To scale-up a MFC instead of using a single
large unit several small units can be connected
together rather than increasing the size of a single
MFC (Ieropoulos et al., 2008) in order to avoid
increasing electrical resistance with size (Oliveira et
al., 2013b). When the MFC units are connected in
stacks then the total MFC size can be easily altered
just by adding or removing individual units, thus
adding up or reducing the power output respectively
of the MFC. Despite of the advantages associated with
stacking of MFCs, it results in problems like voltage
reversal etc. Also, it has been observed that the MFC
performs better when the stacks have their individual
feed and electrical connections. This can again be a
challenging task in the field. Several studies on MFC
scale up prove that it is feasible but it is important to
improve MFC design in order to achieve a marketable
cell. Aelterman et al., (2006) in a study investigated
the influence of the electrical circuit (series or parallel)
on the power, voltage and current output of MFCs and
observed that stacked MFCs did not deliver higher
power densities than the individual MFCs. Yet, they
created the possibility to produce an averaged power
at more practical voltages and currents.
 A brief review on recent advances in air-cathode microbial fuel cells
7. Conclusions
Air cathode MFC offers distinctive advantage
by reducing overall footprint required and eliminating
limitations of oxygen availability. A large number of
materials have been explored for its utility in air-
cathode, with optimum design for commercial
application yet to be achieved. Performance of air
cathode MFC varies with configuration and type of
materials used. Modification of cathode surface by
changing its water affinity, charge, surface roughness,
etc. can lead to a solution to the problem of fouling
currently being faced.
Further studies on current collectors,
conductive materials, diffusion layer materials,
catalysts, binders and fouling resistant economic
materials are required to improve cathode
performance with long term stability, and make MFC
competitive with the already established wastewater
treatment technologies.
Acknowledgements
The financial support received from the Department of
Science and Technology, Government of India (File No.
DST/TSG/NTS/2010/61) is duly acknowledged.
References
Aelterman P., Rabaey K., Pham H.T., Boon N., Verstraete
W., (2006), Continuous electricity generation at high
voltages and currents using stacked microbial fuel cells,
Environmental Science and Technology, 40, 3388-
3394.
Ahmed J., Yuan Y., Zhou L., Kim S., (2012), Carbon
supported cobalt oxide nanoparticles–iron
phthalocyanine as alternative cathode catalyst for
oxygen reduction in microbial fuel cells, Journal of
Power Sources, 208, 170-175.
Ajayi F.F., Weigele P.R., (2012), A terracotta bio-battery.
Bioresource Technology, 116, 86-91.
Akuzov D., Brümmer F., Vladkova T., (2013), Some
possibilities to reduce the biofilm formation on
transparent siloxane coatings, Colloids and Surfaces B:
Biointerfaces, 104, 303-310.
Alvarez-Gallego Y., Dominguez-Benetton X., Pant D.,
Diels L., Vanbroekhoven K., Genné I., Vermeiren P.,
(2012), Development of gas diffusion electrodes for
cogeneration of chemicals and electricity,
Electrochimica Acta, 82, 415-426.
Ang W.S., Lee S., Elimelech M., (2006), Chemical and
physical aspects of cleaning of organic-fouled reverse
osmosis membranes, Journal of Membrane Science,
272, 198-210.
Angenent L.T., Karim K., Al-Dahhan M.H., Wrenn B.A.,
Domíguez-Espinosa R., (2004), Production of
bioenergy and biochemicals from industrial and
agricultural wastewater, Trends in Biotechnology, 22,
477-485.
Behera M., Jana P.S., Ghangrekar M., (2010a), Performance
evaluation of low cost microbial fuel cell fabricated
using earthen pot with biotic and abiotic cathode,
Bioresource Technology, 101, 1183-1189.
Behera M., Jana P.S., More T.T., Ghangrekar M., (2010b),
Rice mill wastewater treatment in microbial fuel cells
fabricated using proton exchange membrane and
earthen pot at different pH, Bioelectrochemistry, 79,
228-233.
Belfer S., Fainshtain R., Purinson Y., Gilron J., Nyström M.,
Mänttäri M., (2004), Modification of NF membrane
properties by in situ redox initiated graft polymerization
with hydrophilic monomers, Journal of Membrane
Science, 239, 55-64.
Ben Liew K., Daud W.R.W., Ghasemi M., Leong J.X., Su
Lim S., Ismail M., (2014), Non-Pt catalyst as oxygen
reduction reaction in microbial fuel cells: A review,
International Journal of Hydrogen Energy, 39, 4870-
4883.
Bullen R.A., Arnot T.C., Lakeman J.B., Walsh F.C., (2006),
Biofuel cells and their development, Biosensors and
Bioelectronics, 21, 2015-2045.
Chatterjee P., Ghangrekar M., (2013), Preparation of a
fouling resistant sustainable cathode for a single
chambered microbial fuel cell, Water Science and
Technology, 69, 634-639.
Chatterjee P., Ghangrekar M.M., (2014), Design of clayware
separator-electrode assembly for treatment of
wastewater in Microbial Fuel Cells, Applied
Biochemistry and Biotechnology, 173, 378-390.
Chen G., Wei B., Luo Y., Logan B.E., Hickner M.A.,
(2012), Polymer separators for high-power, high-
efficiency Microbial Fuel Cells, ACS Applied Materials
and Interfaces, 4, 6454-6457.
Cheng S., Liu H., Logan B.E., (2006), Increased
performance of single-chamber microbial fuel cells
using an improved cathode structure, Electrochemistry
Communications, 8, 489-494.
Creber S., Vrouwenvelder J., Van Loosdrecht M., Johns M.,
(2010), Chemical cleaning of biofouling in reverse
osmosis membranes evaluated using magnetic
resonance imaging, Journal of Membrane Science, 362,
202-210.
Dong H., Yu H., Wang X., Zhou Q., Feng J., (2012), A novel
structure of scalable air-cathode without Nafion and Pt
by rolling activated carbon and PTFE as catalyst layer
in microbial fuel cells, Water Research, 46, 5777-5787.
Du Z., Li H., Gu T., (2007), A state of the art review on
microbial fuel cells: A promising technology for
wastewater treatment and bioenergy, Biotechnology
advances, 25, 464-482.
Duteanu N., Erable B., Senthil Kumar S., Ghangrekar M.
M., Scott K., (2010), Effect of chemically modified
Vulcan XC-72R on the performance of air-breathing
cathode in a single-chamber microbial fuel cell.
Bioresource Technology, 101, 5250-5255.
Elimelech M., Zhu X., Childress A.E., Hong S., (1997), Role
of membrane surface morphology in colloidal fouling
of cellulose acetate and composite aromatic polyamide
reverse osmosis membranes, Journal of Membrane
Science, 127, 101-109.
Elmekawy A., Hegab H. M., Dominguez-Benetton X., Pant
D., (2013), Internal resistance of microfluidic microbial
fuel cell: Challenges and potential opportunities,
Bioresource Technology, 142, 672-682.
Feng L., Yan Y., Chen Y., Wang L., (2011), Nitrogen-doped
carbon nanotubes as efficient and durable metal-free
cathodic catalysts for oxygen reduction in microbial
fuel cells, Energy and Environmental Science, 4, 1892-
1899.
Freguia S., Rabaey K., Yuan Z., Keller J., (2007), Non-
catalyzed cathodic oxygen reduction at graphite
granules in microbial fuel cells, Electrochimica Acta,
53, 598-603.
1541
 Chatterjee et al./Environmental Engineering and Management Journal 17 (2018), 7, 1531-1544
Ghasemi M., Daud W.R.W., Hassan S.H.A., Oh S.E., Ismail
M., Rahimnejad M., Jahim J. M., (2013a), Nano-
structured carbon as electrode material in microbial fuel
cells: A comprehensive review, Journal of Alloys and
Compounds, 580, 245-255.
Ghasemi M., Daud W.R.W., Rahimnejad M., Rezayi M.,
Fatemi A., Jafari Y., Somalu M.R., Manzour A.,
(2013b), Copper-phthalocyanine and nickel
nanoparticles as novel cathode catalysts in microbial
fuel cells, International Journal of Hydrogen Energy,
38, 9533-9540.
Gong X.-B., You S.-J., Wang X.-H., Zhang J.-N., Gan Y.,
Ren N.-Q., (2014), A novel stainless steel mesh/cobalt
oxide hybrid electrode for efficient catalysis of oxygen
reduction in a microbial fuel cell, Biosensors and
Bioelectronics, 55, 237-241.
Habermann W., Pommer E., (1991), Biological fuel cells
with sulphide storage capacity, Applied Microbiology
and Biotechnology, 35, 128-133.
Harnisch F., Wirth S., Schröder U., (2009), Effects of
substrate and metabolite crossover on the cathodic
oxygen reduction reaction in microbial fuel cells:
platinum vs. iron (II) phthalocyanine based electrodes,
Electrochemistry Communications, 11, 2253-2256.
Hays S., Zhang F., Logan B.E., (2011), Performance of two
different types of anodes in membrane electrode
assembly microbial fuel cells for power generation
from domestic wastewater, Journal of Power Sources,
196, 8293-8300.
Ieropoulos I., Greenman J., Melhuish C., (2008), Microbial
fuel cells based on carbon veil electrodes: stack
configuration and scalability, International Journal of
Energy Research, 32, 1228-1240.
Jana P.S., Behera M., Ghangrekar M., (2010), Performance
comparison of up-flow microbial fuel cells fabricated
using proton exchange membrane and earthen cylinder,
International Journal of Hydrogen Energy, 35, 5681-
5686.
Jin H., Zhang H., Zhong H., Zhang J., (2011), Nitrogen-
doped carbon xerogel: A novel carbon-based
electrocatalyst for oxygen reduction reaction in proton
exchange membrane (PEM) fuel cells, Energy and
Environmental Science, 4, 3389-3394.
Kang G.-D., Cao Y.-M., (2012), Development of antifouling
reverse osmosis membranes for water treatment: A
review, Water Research, 46, 584-600.
Khilari S., Pandit S., Das D., Pradhan D., (2014), Manganese
cobaltite/polypyrrole nanocomposite-based air-cathode
for sustainable power generation in the single-
chambered microbial fuel cells, Biosensors and
Bioelectronics, 54, 534-540.
Khilari S., Pandit S., Ghangrekar M., Das D., Pradhan D.,
(2013a), Graphene supported α-MnO2 nanotubes as a
cathode catalyst for improved power generation and
wastewater treatment in single-chambered microbial
fuel cells, RSC Advances, 3, 7902-7911.
Khilari S., Pandit S., Ghangrekar M.M., Pradhan D., Das D.,
(2013b), Graphene oxide-impregnated PVA–STA
composite polymer electrolyte membrane separator for
power generation in a single-chambered Microbial Fuel
Cell, Industrial and Engineering Chemistry Research,
52, 11597-11606.
Kim J.R., Premier G.C., Hawkes F.R., Dinsdale R.M., Guwy
A.J., (2009), Development of a tubular microbial fuel
cell (MFC) employing a membrane electrode assembly
cathode, Journal of Power Sources, 187, 393-399.
Kulkarni A., Mukherjee D., Gill W.N., (1996), Flux
enhancement by hydrophilization of thin film
1542
composite reverse osmosis membranes, Journal of
Membrane Science, 114, 39-50.
Kumar G.G., Sarathi V.G.S., Nahm K.S., (2013), Recent
advances and challenges in the anode architecture and
their modifications for the applications of microbial
fuel cells, Biosensors and Bioelectronics, 43, 461-475.
Kwon B., Lee S., Cho J., Ahn H., Lee D., Shin H.S., (2005),
Biodegradability, DBP formation, and membrane
fouling potential of natural organic matter:
Characterization and controllability, Environmental
Science & Technology, 39, 732-739.
Li F., Sharma Y., Lei Y., Li B., Zhou Q., (2010), Microbial
fuel cells: the effects of configurations, electrolyte
solutions, and electrode materials on power generation,
Applied Biochemistry and Biotechnology, 160, 168-
181.
Liu H., Logan B.E., (2004), Electricity generation using an
air-cathode single chamber microbial fuel cell in the
presence and absence of a proton exchange membrane,
Environmental Science & Technology, 38, 4040-4046.
Liu J., Feng Y., Wang X., Shi X., Yang Q., Lee H., Zhang
Z., Ren N., (2011), The use of double-sided cloth
without diffusion layers as air-cathode in microbial fuel
cells, Journal of Power Sources, 196, 8409-8412.
Liu L., Liu J., Bo G., Yang F., Crittenden J., Chen Y.,
(2012), Conductive and hydrophilic polypyrrole
modified membrane cathodes and fouling reduction in
MBR, Journal of Membrane Science, 429, 252-258.
Logan B.E., (2008), Microbial Fuel Cell, Hoboken, John
Wiley & Sons, Inc.
Logan B.E., Hamelers B., Rozendal R., Schröder U., Keller
J., Freguia S., Aelterman P., Verstraete W., Rabaey K.,
(2006a), Microbial fuel cells: Methodology and
technology, Environmental Science and Technology,
40, 5181-5192.
Logan B.E., Hamelers B., Rozendal R., Schröder U., Keller
J., Freguia S., Aelterman P., Verstraete W., Rabaey K.,
(2006b), Microbial fuel cells: methodology and
technology, Environmental Science and Technology,
40, 5181-5192.
Louie J.S., Pinnau I., Ciobanu I., Ishida K.P., Ng A.,
Reinhard M., (2006), Effects of polyether–polyamide
block copolymer coating on performance and fouling of
reverse osmosis membranes, Journal of Membrane
Science, 280, 762-770.
Martin E., Tartakovsky B., Savadogo O., (2011), Cathode
materials evaluation in microbial fuel cells: A
comparison of carbon, Mn2O3, Fe2O3 and platinum
materials, Electrochimica Acta, 58, 58-66.
Mcpherson T., Kidane A., Szleifer I., Park K., (1998),
Prevention of protein adsorption by tethered poly
(ethylene oxide) layers: experiments and single-chain
mean-field analysis, Langmuir, 14, 176-186.
Nie F.-Q., Xu Z.-K., Ye P., Wu J., Seta P., (2004),
Acrylonitrile-based copolymer membranes containing
reactive groups: effects of surface-immobilized poly
(ethylene glycol) s on anti-fouling properties and blood
compatibility, Polymer, 45, 399-407.
Oliveira V.B., Simões M., Melo L.F., Pinto A.M.F.R.,
(2013a), Overview on the developments of microbial
fuel cells, Biochemical Engineering Journal, 73, 53-64.
Oliveira V., Simões M., Melo L.F., Pinto A.M.F.R.,
(2013b), Overview on the developments of microbial
fuel cells, Biochemical Engineering Journal, 73, 53-64.
Oliveira V., Carvalho T., Melo L., Pinto A., Simoes M.,
(2016), Effects of hydrodynamic stress and feed rate on
the performance of a microbial fuel cell, Environmental
Engineering and Management Journal, 15, 2497-2504.
 A brief review on recent advances in air-cathode microbial fuel cells
Pant D., Arslan D., Van Bogaert G., Gallego Y.A., De
Wever H., Diels L., Vanbroekhoven K., (2013),
Integrated conversion of food waste diluted with
sewage into volatile fatty acids through fermentation
and electricity through a fuel cell, Environmental
Technology, 34, 1935-1945.
Pant D., Van Bogaert G., Diels L., Vanbroekhoven K,
(2010a), A review of the substrates used in microbial
fuel cells (MFCs) for sustainable energy production,
Bioresource Technology, 101, 1533-1543.
Pant D., Van Bogaert G., De Smet M., Diels L.,
Vanbroekhoven K., (2010b), Use of novel permeable
membrane and air cathodes in acetate microbial fuel
cells, Electrochimica Acta, 55, 7710-7716.
Pant D., Singh A., Van Bogaert G., Gallego Y.A., Diels L.,
Vanbroekhoven K., (2011), An introduction to the life
cycle assessment (LCA) of bioelectrochemical systems
(BES) for sustainable energy and product generation:
Relevance and key aspects, Renewable and Sustainable
Energy Reviews, 15, 1305-1313.
Pant D., Van Bogaert G., Alvarez-Gallego Y., Diels L.,
Vanbroekhoven K., (2016), Evaluation of
bioelectrogenic potential of four industrial effluents as
substrate for low cost Microbial Fuel Cells operation,
Environmental Engineering and Management Journal,
15, 1897-1904.
Ponnusamy K., Paul D., Kim Y.S., Kweon J.H., (2010), 2
(5H)-Furanone: a prospective strategy for biofouling-
control in membrane biofilm bacteria by quorum
sensing inhibition, Brazilian Journal of Microbiology,
41, 227-234.
Prakash G., Viva F.A., Bretschger O., Yang B., El-Naggar
M., Nealson K., (2010), Inoculation procedures and
characterization of membrane electrode assemblies for
microbial fuel cells, Journal of Power Sources, 195,
111-117.
Rabaey K., Verstraete W., (2005), Microbial fuel cells:
Novel biotechnology for energy generation, TRENDS
in Biotechnology, 23, 291-298.
Rana D., Matsuura T., (2010), Surface modifications for
antifouling membranes, Chemical Reviews, 110, 2448-
2471.
Rikame S.S., Mungray A.A., Mungray A.K., (2012),
Electricity generation from acidogenic food waste
leachate using dual chamber mediator less microbial
fuel cell, International Biodeterioration and
Biodegradation, 75, 131-137.
Rozendal R.A., Hamelers H.V., Rabaey K., Keller J.,
Buisman C.J., (2008), Towards practical
implementation of bioelectrochemical wastewater
treatment, Trends in Biotechnology, 26, 450-459.
Sagle A.C., Van Wagner E.M., Ju H., Mccloskey B.D.,
Freeman B.D., Sharma M.M., (2009), PEG-coated
reverse osmosis membranes: desalination properties
and fouling resistance, Journal of Membrane Science,
340, 92-108.
Saito T., Merrill M.D., Watson V.J., Logan B.E., Hickner
M.A., (2010), Investigation of ionic polymer cathode
binders for microbial fuel cells, Electrochimica Acta,
55, 3398-3403.
Sevda S., Dominguez-Benetton X., Vanbroekhoven K.,
Sreekrishnan T.R., Pant D., (2013), Characterization
and comparison of the performance of two different
separator types in air–cathode microbial fuel cell
treating synthetic wastewater, Chemical Engineering
Journal, 228, 1-11.
Sharma M., Bajracharya S., Gildemyn S., Patil S.A.,
Alvarez-Gallego Y., Pant D., Rabaey K., Dominguez-
Benetton X., (2014), A critical revisit of the key
parameters used to describe microbial electrochemical
systems, Electrochimica Acta, 140, 191-208.
Shi X., Feng Y., Wang X., Lee H., Liu J., Qu Y., He W.,
Kumar S., Ren N., (2012), Application of nitrogen-
doped carbon powders as low-cost and durable cathodic
catalyst to air–cathode microbial fuel cells, Bioresource
Technology, 108, 89-93.
Sun J., Hu Y., Bi Z., Cao Y., (2009), Improved performance
of air-cathode single-chamber microbial fuel cell for
wastewater treatment using microfiltration membranes
and multiple sludge inoculation, Journal of Power
Sources, 187, 471-479.
Tugtas A.E., Cavdar P., Calli B., (2011), Continuous flow
membrane-less air cathode microbial fuel cell with
spunbonded olefin diffusion layer, Bioresource
Technology, 102, 10425-10430.
Van Wagner E.M., Sagle A.C., Sharma M.M., La Y.-H.,
Freeman B.D., (2011), Surface modification of
commercial polyamide desalination membranes using
poly (ethylene glycol) diglycidyl ether to enhance
membrane fouling resistance, Journal of Membrane
Science, 367, 273-287.
Vermaas D.A., Kunteng D., Saakes M., Nijmeijer K.,
(2012), Fouling in reverse electrodialysis under natural
conditions, Water Research, 47, 1289-1298.
Vrijenhoek E.M., Hong S., Elimelech M., (2001), Influence
of membrane surface properties on initial rate of
colloidal fouling of reverse osmosis and nanofiltration
membranes, Journal of Membrane Science, 188, 115-
128.
Wang P., Tan K., Kang E., Neoh K., (2002), Plasma-induced
immobilization of poly (ethylene glycol) onto poly
(vinylidene fluoride) microporous membrane, Journal
of Membrane Science, 195, 103-114.
Wang Q., Huang L.P., Yu H.T., Quan X., Chen G.H.,
(2013), Recent developments of graphene electrodes in
bioelectrochemical systems, Wuli Huaxue Xuebao/
Acta Physico - Chimica Sinica, 29, 889-896.
Wang X., Feng Y., Liu J., Shi X., Lee H., Li N., Ren N.,
(2010), Power generation using adjustable
Nafion/PTFE mixed binders in air-cathode microbial
fuel cells, Biosensors and Bioelectronics, 26, 946-948.
Wei J., Liang P., Huang X., (2011), Recent progress in
electrodes for microbial fuel cells, Bioresource
Technology, 102, 9335-9344.
Wen Q., Wang S., Yan J., Cong L., Chen Y., Xi H., (2014),
Porous nitrogen-doped carbon nanosheet on graphene
as metal-free catalyst for oxygen reduction reaction in
air-cathode microbial fuel cells, Bioelectrochemistry,
95, 23-28.
Winfield J., Ieropoulos I., Rossiter J., Greenman J., Patton
D., (2013), Biodegradation and proton exchange using
natural rubber in microbial fuel cells, Biodegradation,
24, 733-739.
Xu J., Sheng G.-P., Luo H.-W., Li W.-W., Wang L.-F., Yu
H.-Q., (2012), Fouling of proton exchange membrane
(PEM) deteriorates the performance of microbial fuel
cell, Water Research, 46, 1817-1824.
Yang R., Xu J., Ozaydin-Ince G., Wong S.Y., Gleason K.K.,
(2011), Surface-tethered zwitterionic ultrathin
antifouling coatings on reverse osmosis membranes by
initiated chemical vapor deposition, Chemistry of
Materials, 23, 1263-1272.
Zhang F., Cheng S., Pant D., Bogaert G.V., Logan B.E.,
(2009a), Power generation using an activated carbon
and metal mesh cathode in a microbial fuel cell,
Electrochemistry Communications, 11, 2177-2179.
Zhang X., Cheng S., Wang X., Huang X., Logan B.E.,
(2009b), Separator characteristics for increasing
1543
 Chatterjee et al./Environmental Engineering and Management Journal 17 (2018), 7, 1531-1544
performance of microbial fuel cells, Environmental
Science and Technology, 43, 8456-8461.
Zhang F., Saito T., Cheng S., Hickner M.A., Logan B.E.,
(2010a), Microbial fuel cell cathodes with poly
(dimethylsiloxane) diffusion layers constructed around
stainless steel mesh current collectors, Environmental
Science and Technology, 44, 1490-1495.
Zhuang L., Feng C., Zhou S., Li Y., Wang Y., (2010b)
Comparison of membrane-and cloth-cathode assembly
for scalable microbial fuel cells: construction,
performance and cost, Process Biochemistry, 45, 929-
934.
Zhang F., Pant D., Logan B.E., (2011a), Long-term
performance of activated carbon air cathodes with
different diffusion layer porosities in microbial fuel
cells, Biosensors and Bioelectronics, 30, 49-55.
Zhang X., Cheng S., Liang P., Huang X., Logan B.E.,
(2011b), Scalable air cathode microbial fuel cells using
glass fiber separators, plastic mesh supporters, and
1544
graphite fiber brush anodes, Bioresource Technology,
102, 372-375.
Zhang F., Chen G., Hickner M.A., Logan B.E., (2012),
Novel anti-flooding poly (dimethylsiloxane)(PDMS)
catalyst binder for microbial fuel cell cathodes, Journal
of Power Sources, 218, 100-105.
Zou L., Vidalis I., Steele D., Michelmore A., Low S.,
Verberk J., (2011), Surface hydrophilic modification of
RO membranes by plasma polymerization for low
organic fouling, Journal of Membrane Science, 369,
420-428.
Zuo Y., Cheng S., Call D., Logan B.E., (2007), Tubular
membrane cathodes for scalable power generation in
microbial fuel cells, Environmental Science and
Technology, 41, 3347-3353.
Zuo Y., Cheng S., Logan B.E., (2008), Ion exchange
membrane cathodes for scalable microbial fuel cells,
Environmental Science and Technology, 42, 6967-
6972.
Environmental Engineering and Management Journal July 2018, Vol.17, No. 7, 1545-1554
http://www.eemj.icpm.tuiasi.ro/; http://www.eemj.eu

“Gheorghe Asachi” Technical University of Iasi, Romania

COMPARING MAMDANI AND SUGENO HIERARCHICAL FUZZY

SYSTEMS FOR ENVIRONMENTAL IMPACT ASSESSMENT:
A PIPELINE PROJECT CASE STUDY

Sharosin Givargis1∗, Azadeh Dabiri1, Negin Hakiminejad1,

Mahmood Mahmoodi2, Farzam Babaei Semiromi3
1Pouyandegan Mohit Zist (PMZ) Consulting Engineering Company Pvt. Ltd. No.1103, Valliasr Ave, Tehran, 113633865, Iran
2Department of Epidemiology and Biostatistics, School of Public Health, Tehran University of Medical Sciences, Iran
3Department of Environmental Management and Energy, Science and Research Branch, Islamic Azad University, Tehran, Iran

Abstract

Several studies have been conducted in various fields comparing Mamdani and Sugeno fuzzy inference systems (FISs). This study
contributes to the literature by comparing the performance of Mamdani and Sugeno hierarchical fuzzy systems (HFSs) in terms of
their technical performance within the context of environmental impact assessment (EIA), a multi-criteria decision analysis
(MCDA) method, with the aim of illuminating practical issues that need to be taken into account during application. The context
is a non-commercial, research-oriented EIA of a pipeline project in southern Iran. Mamdani and Sugeno HFSs were developed
with data collected for and expertise gained via a previous formal EIA of the same project. The two developed HFSs were of binary
structure, reducing complexity while also facilitating sensitivity analysis. A sensitivity analysis was carried out with the full range
of possible HFS input values. Excluding the final FISs, the FIS outputs did not differ significantly between the two examined HFSs.
However, the behavior of the Sugeno HFS was found to be more linear than that of its Mamdani counterpart with a higher degree
of sensitivity to input value changes. This study indicates that the Mamdami HFS is unreliable in some areas due to fluctuations in
the output surface, which necessitates smoothing before it can be applied.

Key words: environmental impact assessment, hierarchical fuzzy system, Mamdani inference system, Sugeno inference system

Received: February, 2014; Revised final: July, 2014; Accepted: September, 2014; Published in final edited form: July 2018

1. Introduction yields total scores for each alternative that reflect

Environmental impact assessment (EIA) is a
process of systematic identification and evaluation of
the potential impacts of projects, programs, or
legislative actions on the physical, chemical,
biological, cultural, and socioeconomic constituents of
the “total environment” (Canter, 1996). It is a multi-
criteria decision analysis method enables various
alternatives to be evaluated in relation to multifarious
criteria, such that each alternative is considered in
relation to criteria weighted on the basis of decision
makers’ or stakeholders’ judgments. Ultimately, EIA
decision makers’ preferences alongside the
alternatives’ performance in relation to the criteria
(FEI, 2013; Torretta and Capodaglio, 2017).
According to Rodriguez-Bachiller and Glasson,
(2004), EIA implementation has been improving, and
a standardized practice is close to being defined.
Nevertheless, most EIAs are far from satisfactory and
new paradigms for EIA are being proposed in the
literature. Rodriguez-Bachiller and Glasson (2004)
propose that expert systems be integrated with
geographical information systems (GISs), whereas
Shepard (2005) advocates the use of approximate

∗
Author to whom all correspondence should be addressed: e-mail: nirari6754@yahoo.com; Phone: +98-21- 66466682, +98-21- 66466733; Fax:
+98-21- 66480419
 Givargis et al./Environmental Engineering and Management Journal 17 (2018), 7, 1545-1554
reasoning and fuzzy modeling. Subjective judgment is
an inextricable component of EIA because it entails
social and cultural value judgments as well as
scientific data. Traditional multi-criteria decision-
making methods are insufficient for informing
decisions because they are not able to model
qualitative human thinking processes. Conversely,
fuzzy logic has the capacity to deal with such criteria
and problems. Fuzzy systems compute numerical
outputs based on a “qualitative human thinking
process” that uses words to analyze complex systems
(Chen et al., 2017; Liu et al., 2006). As a result, fuzzy
expert systems can help stakeholders understand the
rationale behind the tangible values computed by an
EIA (Aly and Vrana, 2007; Pislaru et al., 2010).
Likewise, these systems can provide a tool that
stakeholders can use to incorporate their views and
preferences into the assessment process and observe
the numerical results of their reasoning. A fuzzy
expert system is a knowledge-based, quantitative
expert system that functions using fuzzy sets,
membership functions, if-then rules, and logical
operators such as “AND”, “OR”, and “NOT”
(Horsekotte, 2005; Shepard, 2005). Mamdani and
Sugeno models are the most commonly employed
fuzzy expert systems (Zaheeruddin and Jane, 2006).
The main feature of the Mamdani fuzzy inference
system (FIS) is the fact that both the antecedents and
consequents of its rules are fuzzy. Building this type
of model requires four steps: fuzzification,
implication, aggregation, and defuzzification
(Mathworks Inc., 2006).
The Sugeno system (a.k.a. TSK based on its
originators Takagi, Sugeno and Kang) is similar to the
Mamdani system in the fuzzification step and its use
of fuzzy operators. However, it is distinguished by the
feature of having output membership functions that
are not fuzzy. The Sugeno system can employ a linear
(first-order) output membership function or a constant
(zero-order) membership function (Mathworks Inc.,
2006; Zaheeruddin and Jane, 2006), and its final
output is computed through the weighted average
method with the outputs of fired rules (Mathworks
Inc., 2006).
In standard fuzzy systems, the number of rules
and parameters increases exponentially in relation to
the number of input variables. This so-called curse of
dimensionality damages the transparency of systems
since humans are unable to understand and justify
“hundreds or thousands of fuzzy rules and parameters”
(Wang et al., 2006). This problem has been addressed
by hierarchical fuzzy systems (HFSs), which preclude
a large number of expert rules, prevent the expansion
of knowledge into a colossal system, and support the
organization of inter-input variable relationships into
a logical structure. Hierarchical systems accomplish
this restraint by combining or aggregating several
smaller fuzzy expert systems into the computation of
the final result.
In this context, combination means that the
outputs of logically-related fuzzy expert systems are
integrated into a single output using logical rules,
1546
whereas aggregation means that the outputs of
independent fuzzy expert systems are accumulated as
necessary for computing the final output of the entire
system (Aly and Vrana, 2007).
The purpose of the present work is to compare
the technical performance of Mamdani and Sugeno
HFSs in the realm of EIA from a multi-criteria
decision analysis perspective with the aim of
highlighting some crucial aspects of utilizing the
developed models. These HFS models have probable
applications in future studies of similar situations. The
present study is not a real EIA and does not analyze
and compare any actual alternatives, which would be
beyond the capacity of this paper. Rather, it scrutinizes
the way both systems can behave in supporting
environmental decisions.
Two EIA case studies utilizing HFSs are
summarized herein. These studies laid the foundation
upon which the current study builds and to which this
work adds new features and provides new practical
insights. This study utilizes a simplified structure
relative to the mentioned studies via developing binary
HFSs. In section 2, a brief discussion of some case
studies comparing Mamdani and Sugeno inference
systems is presented, mainly from the engineering
modeling perspective. These case studies provided the
impetus for the present study to investigate a different
facet of such comparisons to illuminate some practical
aspects of the systems.
2. Case studies
Some case studies of EIA schemes have made
use of HFSs as decision support tools. As a case in
point, Siqueira Campos Boclin and Mello, (2006)
developed an EIA fuzzy decision support system for a
highway project in Brazil. They examined four
alternative development scenarios, including the “pre-
existing situation” and “no action”. They created a
decision tree which integrates the outputs of fuzzy
sub-systems to compute the total environmental
situation. All of the subsystems were of the Mamdani
type and all of the variables in the model were
standardized to the universe of discourse [0, 1]. The
membership functions pertaining to the linguistic
terms “good”, “bad”, and “critical” were S-shaped.
The defuzzification method was the center of gravity,
which generated output values within an interval
smaller than in their study [0, 1]; put differently, the
final outputs were in the range [0.158, 0.842].
Therefore, the authors added a final normalization step
to each subsystem, as well as to the final output, to
rescale the outputs at each step to the interval [0, 1]. In
this example, the final outputs were converted by
subtracting the lower endpoint of the interval (0.158)
and dividing the result by the length of the interval.
Likewise, Liu et al., (2009) developed a fuzzy
decision support system for an EIA of a Taiwanese
high-speed rail project. They divided the project line
into three segments and analyzed three different
scenarios for each of the segments. They created a
hierarchy of Mamdani fuzzy expert systems to assess
 Application of landscape metrics for assessment of land use/ land cover (LULC) changes in Varjin protected area, Iran
the impact on environmental criteria and subcriteria.
Their model was developed using the Matlab Fuzzy
Logic Toolbox. The computed output of each fuzzy
expert system was the significance of the associated
environmental impact. The universe of discourse
relating to all output variables was the interval [0,
100]. Triangular fuzzy membership functions were
used for the output variables, with linguistic
interpretations ranging from “very insignificant” to
“very significant”. The defuzzification method was
the center of gravity. The authors reported that the real
output range was narrower than the interval [0, 100].
As with the Boclin and Mello (2006) study, Lui et al.,
(2009) rescaled all intermediate and final outputs so
that their lower and upper bounds corresponded to 0.0
and 100, respectively.
It should be mentioned that each higher-ranked
system of the hierarchy consisted of two to five sub-
systems in both of these models. Studies comparing
Mamdani and Sugeno FISs have been carried out in
fields other than EIA. For example, Jassbi et al.,
(2006) compared the performance of Mamdani and
Sugeno FISs in detecting generic system faults of the
ENVISAT satellite gyroscope. Both FISs were
developed to trigger an alarm. The authors observed
that, of the two, the Sugeno FIS was much faster,
requiring 14 times less processing time. The Sugeno
FIS was more robust than its Mamdani counterpart
when interpreting noisy data; however, it was also
more sensitive to very high noise levels. On this basis,
the authors concluded that the Sugeno FIS behaved
more realistically than did the Mamdani FIS.
Ultimately, they reported that transitions between
states were smoother in the Sugeno FIS and that the
Sugeno FIS was more sensitive to imprecision where
the input fuzzy sets overlapped. Guney and Sarikaya
(2009) compared Mamdani and Sugeno FISs for
resonant frequency calculation of rectangular
microstrip antennas. The designed parameters of both
FISs were adjusted using various training algorithms.
The absolute errors between the measured resonant
frequencies and the frequencies calculated by the FISs
were small. However, the best performance was
achieved by the Sugeno FIS trained by the least
squares algorithm. Özger (2009) compared Mamdani
and Sugeno FISs developed to predict stream flow of
the Euphrates River in Turkey. Both FISs were trained
on the basis of experimental data. The parameters of
the Mamdani FIS were adjusted using a genetic
algorithm, while the Sugeno FIS was trained using
artificial neural networks (ANNs).
In this case, the Mamdani FIS outperformed its
Sugeno counterpart in terms of accuracy. However,
the performances of both FISs were surpassed by that
of a conventional model. Kaur and Kaur (2012)
compared Mamdani and Sugeno FISs for the operation
of an air conditioning system. In both systems, the
input variables were temperature and humidity, and
the single output variable was compressor speed.
Despite the fact that both systems operated similarly,
the air conditioning system worked at full capacity
only when using the Sugeno FIS.
3. Material and methods
3.1. Background
This study is a research-oriented, non-
commercial EIA based on the data and expertise
utilized in a formal EIA study of the South Pars gas
pipeline project, Phases 15 and 16. The project area
extends from landfall in the vicinity of the coastal
village of Assaluyeh to the zone of the platforms of
South Pars Deck 15 (Phase 15) and South Pars Deck
16 (Phase 16), approximately 100 km off the southern
Iranian coast in the Persian Gulf. The landfall area
defined in the program stretches from Kilometer Point
(KP) 0.0 to KP 4.7. From KP 4.04 to KP 15.46, the
desired pipelines pass through Nayband national
Marine Park (Pouyandegan Mohit Zist, 2011).
Nayband National Marine Park is one of the most
valuable habitats of the Persian Gulf, sheltering coral
reefs and marine tortoises.
A prior formal EIA of this project was carried
out with a Leopold Matrix (Canter, 1996) with the aim
of assessing the environmental impact associated with
“action” and “no action” alternatives. The area
targeted by the EIA was a 5-km–wide pathway (2.5
km on either side of the pipeline) running through the
coastal regions and sea bed (Pouyandegan Mohit Zist,
2011).
3.2. Hierarchical fuzzy systems
In this study, we have developed two HFSs
using the same data and expertise that were involved
in creating the formal EIA. Our HFSs have been
developed with the cooperation of PMZ Company’s
experts through brainstorming sessions, using their
professional knowledge and the data that were
collected for the formal EIA. The conceptual models
(Fig. 1 and Fig. 2), fuzzy membership functions (Fig.
3), and all the rule bases (Table 1), are the resultant
outputs of such sessions. The present study does not
entail public participation or stakeholder views and
lies outside the scope of a formal EIA scheme.
We compare the results obtained with a
Mamdani FIS-based HFS and a Sugeno FIS-based
HFS with identical structures. The input variables are
all environmental factors scrutinized in the original
EIA. The HFSs are binary systems in which every
upper rank FIS consists of two sub-systems, thereby
reducing the complexity of the hierarchy and
facilitating sensitivity analysis.
3.2.1. Mamdani HFS
The Mamdani HFS (developed with the Matlab
R2006b Fuzzy Logic Toolbox) involves the same
membership functions for each fuzzy sub-system. All
input and output variables are rated in the overlapping
fuzzy categories “Poor”, “Mediocre”, and “Good”,
reflecting the performance quality of the project for
that environmental criterion.
1547
 Givargis et al./Environmental Engineering and Management Journal 17 (2018), 7, 1545-1554

Fig. 1. The conceptual model associated with the Biophysical Index (FL – Fuzzy logic)

Table 1. Fuzzy rules used to calculate the Sea-Bed, Biophysical and Total indices

If Then
Rule no.
Input1a is And Input 2b is Sea-Bed is Biophysical is Total is
1 Poor Poor Poor Poor Poor
2 Poor Mediocre Poor Poor Poor
3 Poor Good Mediocre Mediocre Poor
4 Mediocre Poor Mediocre Poor Poor
5 Mediocre Mediocre Mediocre Mediocre Mediocre
6 Mediocre Good Mediocre Good Good
7 Good Poor Mediocre Mediocre Poor
8 Good Mediocre Good Good Mediocre
9 Good Good Good Good Good
a: Sea-Bed Sediment for the Sea-Bed Index, Physical for the Biophysical Index, and Biophysical for the Total Index; b: Sea-Bed Morphology for
the Sea-Bed Index, Biological for the Biophysical Index, and Socio-Economic and Cultural for the Total Index.

The Mamdani and Sugeno HFSs compute manner to compute the total performance of the
project performance indices relating to the project in the order shown in Fig. 1 and 2, respectively.
environmental components within the interval [0, 1]. The conceptual models through which the Biophysical
The individual indices are integrated in a hierarchical and the Total Indices are computed are presented in

1548
 Application of landscape metrics for assessment of land use/ land cover (LULC) changes in Varjin protected area, Iran

Fig. 1. Table 1 shows the rules associated with the Sea- are computed are presented in Fig. 2, respectively.
Bed, Biophysical, and Total FISs that typify FISs The shapes of the corresponding membership
throughout the hierarchy. The conceptual models functions are triangular (Fig. 3), and the universe of
through which the Biophysical and the Total Indices discourse for all variables is [0, 1].

Fig. 2. The conceptual model associated with the Total Index (FL: Fuzzy logic)

Fig. 3. Fuzzy membership functions

1549
 Givargis et al./Environmental Engineering and Management Journal 17 (2018), 7, 1545-1554

It has to be mentioned that the Sea-Bed and
Biophysical FISs represent all FISs, except for the
Total FIS, which is a unique FIS throughout the
hierarchy. For this study, we use the Centroid
defuzzification method. The “AND”, implication, and
aggregation methods are Minimum, Minimum, and
Maximum, respectively.
3.2.2. Sugeno HFS
All the same input variables, input membership
functions, and rule bases used for the Mamdani HFSs
are used for the Sugeno HFS (developed with the
Matlab R2006b Fuzzy Logic Toolbox), in the systems
developed for the purpose of this study. In a zero order
Sugeno system, the output functions are constants.
Here, the “Poor”, “Mediocre” and “Good” categories
are mapped onto the constants 0.01, 0.5 and 1.0. The
selected “AND” method is the Product method.
3.3. Data generation
statistical differences between the outputs of both
HFSs.
Linear regression analysis is used to test the
linearity of the output function with respect to its input
variables. An HFS with linear behavior is equally
sensitive to all inputs, producing a balanced scale for
comparison. The variation between each pair of
adjacent output values is calculated as follows (Eq. 1):
∆i = i N − i N −1 , (1)
where i is the output value and N is the output number
4. Results and discussion
The results of diagonal simulation on the Sea-
Bed, Biophysical, and Total FISs are shown in Fig. 4.
The shapes of these curves are typical of all sub-
systems in the hierarchy.

To generate data, 41 input values ranging from

0.0 to 1.0, with an increment of 0.025, are generated
for each input variable. Then, the generated values of
both variables are paired in a 41×41 matrix.
Eventually, all input pairs generated in this fashion are
administered to all FISs of the developed Mamdani
and Sugeno HFSs. It is important to mention that, for
our purposes, the diagonal simulation means that the
corresponding pairs of input values are utilized to
generate output values.
Hence, the values of both input variables (a)
increase with a 0.025 increment. The row-oriented
simulation means that 41 unique values for the first
input variable are paired with the 41 unique values of
the second input variable in an incremental fashion.
Likewise, the column-oriented simulation means that
there are 41 unique values for the second input
variable, each of them paired with the 41 unique
values of the first input variable in an incremental
fashion.

3.4. Data analysis

Data analysis is carried out using sensitivity (b)

curves, the non-parametric two-related sample
Wilcoxon test (Siegel and Castellan, 1998; Statsoft
Inc., 2013), linear regression, and by characterizing
the variations between adjacent output values.
Sensitivity curves are used to observe the general
behavior of both HFSs, especially in terms of
operating at full capacity. Furthermore, the sensitivity
curves are utilized to compare the reliability of HFSs
over the full range of possible inputs.
The outputs of the Mamdani FISs are adjusted
using the method adopted by Siqueira Campos Boclin
and Mello (2006). In the next step, the outputs of the
Sugeno FISs are compared with both original and (c)
adjusted outputs of the developed Mamdani FISs. The
Wilcoxon test is used to detect any significant Fig. 4. Results of applying diagonal simulation to the: (a)
Sea-Bed FIS; (b) Biophysical FIS and (c) Total FIS.

1550
 Application of landscape metrics for assessment of land use/ land cover (LULC) changes in Varjin protected area, Iran

The Mamdani FISs fail to operate at full
capacity, meaning that their output range is narrower
than the input range [0, 1], as reported by other authors
(Kaur and Kaur, 2012; Liu et al., 2009; Siqueira
Campos Boclin and Mello, 2006). To put it in another
way, the Mamdani’s output range is [0.163, 0.837],
indicating the fact that the system is biased toward the
end of the interval. This means that the degree of bias
(overestimation or underestimation) increases as the
inputs move farther from the midpoint. This
phenomenon is especially pronounced in comparison
with the Sugeno FISs. On the other hand, the Sugeno
system operates without any bias toward the endpoints
of the interval. Hence, further adjustments have to be
made to the Mamdani outputs here, similar to the
method adopted by Siqueira Campos Boclin and
Mello (2006).
The results of the non-parametric two-related
sample Wilcoxon test (Table 2) demonstrate that for
the Sea-Bed and Biophysical FISs, there are no
statistically significant differences between the output
values of the Mamdani (with original or adjusted
output values) and Sugeno HFSs (|Z|< Z0.975= 1.96; P
> 0.05; PMontecarlo > 0.05). On the other hand, for Total
FIS, there are significant statistical differences
between the output values of the Mamdani (both with
original and adjusted output values) and Sugeno HFSs
(|Z| > Z0.975 = 1.96; P < 0.05; PMontecarlo < 0.05). This
means that, excepting the Total Index results, both
HFSs perform similarly in terms of evaluating the
performance of decision alternatives with respect to
each decision criterion or index. However, with regard
to the Total Index, the two HFSs function significantly
differently in terms of evaluating the aggregate
performance of decision alternatives. Hence, the
comparison among the final outputs of the decision
alternatives will have a different meaning for each
HFS in statistical terms.
Linear regression analysis (Table 3)
demonstrates that both the Mamdani (with original or
adjusted output values) and Sugeno HFSs are linear at
a confidence level of 99% (P < 0.01, F > F0.99 (2, 1678)
= 4.618), indicating that the transitions between output
values are fairly constant and continuous in both
HFSs. In other words, both systems provide a stable
measure for drawing comparisons between decision
alternatives. Nevertheless, the F and R2 values
associated with the Sugeno HFS are greater than those
belonging to the Mamdani HFS, with the original and
adjusted output values, indicating that the Sugeno
HFS produces more linear behavior, with a higher
percentage of points landing on the regression line.
Consequently, the Sugeno HFS produces a more
balanced scale for comparison.
In addition, the input variables’ coefficients for
the Sugeno system are greater than those for its
Mamdani counterpart. Therefore, any change in the
input values will bring about more changes in the
output values of the Sugeno system. Hence, besides
providing a more balanced scale for comparison, the
Sugeno system provides a more sensitive scale for
differentiating between decision alternatives, in terms
of both decision criteria and aggregate performance.
The percentages of negative variation (Eq. 1)
for the original and adjusted versions of the Mamdani
HFS are identical.

Table 2. Non-parametric two-related samples Wilcoxon test comparing Mamdani and Sugeno FIS outputs

Compared systems FISa Zb Pc Pd Montecarlo

Sea-Bed -0.698 0.485 0.482
Sugeno and Mamdani Biophysical -0.773 0.440 0.437
Total -16.330 0.000 0.000
Sea-Bed -0.874 0.382 0.379
Sugeno and Adjusted Mamdanie Biophysical -0.876 0.387 0.392
Total -15.380 0.000 0.000
a: Fuzzy inference system; b: Measure of the test; c: Asymptotic statistical P value (2-tailed); d: Montecarlo statistical P value (2-tailed) “based
on 10000 sampled tables; e: With adjusted output values

Table 3. Linear regression analysis results

Coefficients
HFSa FISb R2 Fe Pf
Input1c Input2d Constant
Sea-Bed 0.305 0.103 0.296 0.788 3.118exp3 <0.01
Original Mamdani g Biophysical 0.311 0.311 0.189 0.874 5.831exp3 <0.01
Total 0.216 0.319 0.184 0.682 1.802exp3 <0.01
Sea-Bed 0.453 0.153 0.197 0.788 3.118exp3 <0.01
Adjusted Mamdanih Biophysical 0.462 0.462 0.038 0.874 5.831exp3 <0.01
Total 0.320 0.474 0.030 0.682 1.802exp3 <0.01
Sea-Bed 0.736 0.254 0.007 0.934 1.184exp4 <0.01
Sugeno Biophysical 0.736 0.736 -0.233 0.962 2.118exp4 <0.01
Total 0.493 0.736 -0.238 0.824 3.920exp3 <0.01
a: Hierarchical fuzzy system; b: Fuzzy Inference System; c: Sea-Bed Sediment for the Sea-Bed FIS, Physical for the Biophysical FIS, and
Biophysical for the Total FIS; d: Sea-Bed Morphology for the Sea-Bed FIS, Biological for the Biophysical FIS, and Socio-Economic and Cultural
for the Total FIS; e: Fisher Test Critical Value; f: Statistical P value; g: With original output values; h: With adjusted output values

1551
 Givargis et al./Environmental Engineering and Management Journal 17 (2018), 7, 1545-1554

Negative variations are symptomatic of
fluctuations in the output surface. On the contrary, this
phenomenon is not observed with the Sugeno FISs.
Rather, in the corresponding areas of the Sugeno HFS,
the slope of change decreases (Fig. 5) or remains
constant (Fig. 6). Notice in Table 4 that, in most cases,
negative variations happen where the constant
variable values are within the extreme ranges of [0.0,
0.25] and [0.75, 1]. Moreover, it can be inferred that
when an input variable with a higher coefficient is held
constant (Table 3), the variation of the other variable’s
values spreads the mentioned error into other constant
variable value ranges, with the central value being
maintained (i.e. 0.5).
When the first input variable (i.e. Physical
Index) is held constant at 0.85 and paired with all
values related to the second input variable (i.e.
Biological Index) ranging from 0 to 1 with an
increment of 0.025 (Fig. 5), the slope of variation for
the Sugeno FIS increases from point 21 onward. On
the other hand, the output values for the Mamdani FIS
(original and adjusted output values) are constant from
point 21 to 27, followed by conspicuous fluctuating
pattern in the curve associated with the output values.
When the second input variable (i.e. Socio-
Economic and Cultural Index) is held constant at
0.950 and paired with all values of the first input
variable (i.e. Biophysical Index), ranging from 0 to 1
with an increment of 0.025, the Sugeno output values
increase at a constant rate of 0.047 between points 1
and 21 and then level out (Fig. 6 and Table 5). On the
other hand, in the Mamdani FIS, with both original
and adjusted output values, the output values increase
at a varying rate between points 1 and 21 and fluctuate
thereafter between two identical values. These
findings provide a practical explanation for the linear
regression results (Table 3) while underscoring the
fact that the Sugeno system provides a more balanced
scale for comparison than its Mamdani counterpart.
Between points 1 and 21, the outputs of the Sugeno
system react more sensitively to changes in the input
variable values (Table 5). This phenomenon provides
another practical explanation for the linear regression
results (Table 3), demonstrating that the Sugeno
system provides a more sensitive scale for comparison
than its Mamdani counterpart.
From the data shown in Fig. 6 and Table 5, it
can be inferred that in the fluctuating areas between
points 21 and 41, the Mamdani system fails to live up
to the expectation that when one input variable is held
constant, the response to increases in the other input
should be monotonic. In other words, this
phenomenon is construed as pseudo-sensitivity and
may distort differentiation between alternatives in
terms of decision criteria or aggregate performance.
For example, assume that we have two schemes
respectively named alternative 1 and 2 with an
identical Socio-Economic and Cultural input value of
0.95 using the Mamdani Total FIS with the original
output values for computation. The biophysical scores
of the two alternatives are 0.650 and 0.675,
respectively. According to the rule base (Table 1), one
could expect logically that alternative 2 would match
or surpass the Total Index score of alternative 1.
However, the computed Total scores associated with
alternative 1 and 2 are 0.73 and 0.72, respectively,
defying logic.

Table 4. Percentage of negative variation cases related to Mamdani FISs

Simulation Constant variable values

FISa
type [0.0, 0.25] [0.275, 0.475] 0.5 [0.525, 0.725] [0.750, 1.0]
Sea-Bed 12.5% 21.1% 0.0% 21.1% 12.5%
Row-Oriented
b Biophysical 12.5% 0.0% 0.0% 0.0% 12.5%
Total 35.2% 0.0% 0.0% 0.0% 0.0%
Sea-bed 8.6% 0.0% 0.0% 0.0% 8.6%
Column-
Biophysical 12.5% 0.0% 0.0% 0.0% 12.5%
Oriented c
Total 25.0% 10.6% 0.0% 10.6% 12.5%
a: Fuzzy inference system; b: Each unique value of the first variable, the constant variable, is paired with all unique values of the second variable;
first input variables are Sea-Bed Sediment for the Sea-Bed FIS, Physical for the Biophysical FIS, and Biophysical for the Total FIS; c: Each unique
value of the second variable, the constant variable, is paired with all unique values of the first variable; second input variables are Sea-Bed
Morphology for the Sea-Bed FIS, Biological for the Biophysical FIS, and Socio-Economic and Cultural for the Total FIS

Table 5. Variation values associated with results shown in Fig. 6

Input pair range

Type of fuzzy inference system 1−21 21−41
Mina Maxb Mean Min Max Mean
Original Mamdanic 0.0022 0.0542 0.0289 -0.0089 0.0089 0
Adjusted Mamdanid 0.0033 0.0805 0.0428 -0.0132 0.0132 0
Sugeno 0.0470 0.0470 0.0470 0 0 0
a: Minimum; b: Maximum; c: With original output values; d: With adjusted output values

1552
 Application of landscape metrics for assessment of land use/ land cover (LULC) changes in Varjin protected area, Iran

Fig. 5. Results for one row of row-oriented simulation for the Mamdani and Sugeno Biophysical FISs

Fig. 6. One column of the column-oriented simulation results for the Mamdani and Sugeno Total FISs

In contrast, the Sugeno HFS is reliable over the
entire curve. When using the Mamdani HFS, it is
necessary to take into account such areas of
fluctuation so as to prevent any misinterpretation. One
way to tackle such problems is to use smoothing
techniques. For example, the fluctuated area of the
total Mamdani FIS can be smoothed into a shape
similar to the corresponding area of the Sugeno Total
FIS by applying the linear interpolation method (Fig.
6).
5. Conclusions
This study compares Mamdani and Sugeno
HFSs in the context of a non-commercial research-
oriented EIA case study of a pipeline project in
southern Iran from a multi-criteria decision analysis
perspective. The developed HFSs are binary systems
that reduce complexity and facilitate sensitivity
analysis. The study aim is to provide insight into
practical aspects of the developed systems through
their comparison and thereby produce results that are
useful to consider during application. The developed
systems have the potential to be expanded, revised,
validated, and utilized in other EIA studies of similar
projects in the region.
Acknowledgements
The present work is a non-commercial research carried out
with the support of the Pouyandegan Mohit Zist Company.
The authors would like to express their special thanks to the
management and personnel of the Pouyandegan Mohit Zist
Company for their cooperation and support. Also, the first
author expresses his gratitude to Dr. Borhan Riazi, the
managing director of the company, for providing the
opportunity for him to initiate the study and cooperate as a
guest researcher. The authors thank Dr. Benjamin Mathiesen
from Write Science Right Co for editing of this article. The
authors express their thanks to the Editor-in-Chief and the
reviewers for their comments and suggestions.
References
Aly S., Vrana I., (2007), Multiple parallel fuzzy expert
systems utilizing a hierarchical fuzzy model,
Agricultural Economics − CZECH, 53, 89-93.

1553
 Givargis et al./Environmental Engineering and Management Journal 17 (2018), 7, 1545-1554
Canter L., (1996), Environmental Impact Assessment, 2nd
Edition, McGraw-Hill Inc., Singapore.
Chen Z., Zhang R.R., Han S.S., (2017), An enhanced
environmental multimedia modelling system
(FEMMS): Part I - development and model verification,
Environmental Engineering and Management Journal,
16, 317-328.
Durkin J., (1994), Expert Systems Design and Development,
Macmillan Publishing Company, USA.
Horsekotte E., (2005), Fuzzy expert systems, On line at:
http://www.austinlinks.com/Fuzzy/expert-
systems.html.
Guney K., Sarikaya N., (2009), Comparison of Mamdani
and Sugeno fuzzy inference system models for resonant
frequency calculation of rectangular microstrip
antennas, Progress in Electromagnetics Research B,
12, 81-104.
Jassbi J., Serra P.J.A., Ribeiro R.A., Donati A., (2006), A
Comparison of Mamdani and Sugeno fuzzy inference
systems for a fault detection application, Proc. the
World Automation Congress, Budapest. On line at:
https://ieeexplore.ieee.org/document/4259949/.
Kaur A., Kaur A., (2012), Comparison of Mamdani-Type
and Sugeno-type fuzzy inference systems for air
conditioning system, International Journal of Soft
Computing and Engineering, 2, 323-325.
Liu K.F.R., Liang H.H., Yeh K., Chen C.W., (2009), A
qualitative decision support for environmental impact
assessment using fuzzy logic, Journal of Environmental
Informatics, 13, 93-103.
FEI, (2013), Comparison of Multi-Criteria Decision
Analysis Software: Searching for Ideas for Developing
a New EIA-Specific Multi-Criteria Software,
IMPERIA Project Report, Mustajoki Jyri, Marttunen
Mika, Finnish Environment Institute, On line at:
http://imperia.jyu.fi/tuotokset/Annex7.5.13Compariso
nofMultiCriteriaDecisionAnalyticalSoftware.pdf.
Nath Pant S., Holbert K., (2004), Fuzzy logic in decision
making and signal processing, On line at:
http://enpub.fulton.asu.edu/PowerZone/FuzzyLogic/in
dex.html.
1554
Özger M., (2009), Comparison of fuzzy inference systems
for streamflow prediction, Hydrological Sciences, 54,
261-273.
Pislaru M., Trandabat A., Avasilcai S., (2010), Neuro-fuzzy
model for environmental impact assessment,
Environmental Engineering and Management Journal,
9, 381-386.
Pouyandegan Mohit Zist (2011), The Report of
Environmental Impact Assessment of South Pars
Pipeline Project: Phases 15 and 16, Pouyandegan
Mohit zist Consulting Engineering Company, Tehran,
Iran.
Rodriguez-Bachiler A., Glasson J., (2004), Expert Systems
and Geographic Information Systems for Impact
Assessment, Taylor& Francis, London and New York.
Shepard R.B., (2005), Quantifying Environmental Impact
Assessments Using Fuzzy Logic, Springer, USA.
Siegel S., Castellan J.R., (1998), Non Parametric Statistics
for the Bahavioural Sciences, Mc Graw Hill, Newyork.
Siqueira Campos Boclin A., Mello R., (2006), A decision
support method for environmental impact assessment
using a fuzzy logic approach, Ecological Economics,
58, 170-181.
Statsoft Inc., (2013), Electronic Statistics Textbook, On line
at: http:// www.statsoft.com/textbook/.
The Mathworks Inc., (2006), Matlab R2006b Help File:
Fuzzy Logic Tool Box. The Mathworks Inc., USA, On
line at: https://www.mathworks.com/products/fuzzy-
logic.html.
Torretta V., Capodaglio A.G., (2017), Strategic
environmental assessment: a critical review of
procedural soundness and reliability, Environmental
Engineering and Management Journal, 16, 105-112.
Wang D., Zeng X.J., Keane J.A., (2006), A survey of
hierarchical fuzzy systems, International Journal of
Computational Cognition, 4, 18-29.
Zaheeruddin E., Jain V.K., (2006), A fuzzy expert system
for noise-induced sleep disturbance, Expert Systems
with Applications, 30, 761-771.
Environmental Engineering and Management Journal July 2017, Vol.17, No. 7, 1555-1562
http://www.eemj.icpm.tuiasi.ro/; http://www.eemj.eu

“Gheorghe Asachi” Technical University of Iasi, Romania

DIMENSION AND STRUCTURAL TRAITS OF SOIL MICROPORES

IN CULTIVATIONS DIFFERING IN THE DURATION
OF ORGANIC MANAGEMENT

Nikolaos Monokrousos1,2∗, Efimia M. Papatheodorou2, George P. Stamou2

1Department of Biological Applications and Technology, University of Ioannina, 45 110 Ioannina, Greece
2Department of Ecology, Faculty of Sciences, School of Biology, Aristotle University, 54 124 Thessaloniki, Greece

Abstract

The objectives of this study were to investigate: a) whether the duration of organic farming influences the dimension, shape and
geometry (fractal dimension) of soil micropores (<50μm) and b) for relationships of morphometric traits with certain chemical and
biochemical soil variables (microbial biomass, N- and C-mineralization rates, NH4+, NO3-, organic C and N, extractable P, Mg+2,
K+, Ca+2). We compared soil micromorphometric traits (area, perimeter, compactness, solidity, eccentricity) among fields with
different duration of organic farming (2 (O2), 3 (O3), 5 (O5), and 6 years (O6)): these were planted with Asparagus officinalis and
one was conventional cultivation (CV). Νo significant differences were observed among the morphometric traits of all fields.
However, the soil of the older organic areas (O6, O5 and O3) was characterized by small-sized pores (<10 μm) while the newest
(O1) and the conventional field were characterized by medium-sized micropores (10-20 μm). The fractal dimension D2 of the larger
pores was found to be significantly higher in Ο2 and O3 fields, indicating larger outline irregularity for these particular pores.
Higher fractal dimension could be related to more heterogeneous distribution of the microbial community in space. All micropores
were correlated with the concentration of soil mineral nutrients (Mg+2, K+, Ca+2). In the small pore size category (≤10 μm), N-
microbial and NO3- concentrations, parameters involved in the nitrogen cycle, were found to be correlated to the structure
characteristics. Taking into account that the three older organic fields are characterized mainly by small sized pores (≤10 μm), it is
suggested that improved soil quality is mainly related with the N-cycle.

Key words: fractal dimension, image analysis, micromorphometric traits, soil thin sections

Received: February, 2014; Revised final: August, 2014; Accepted: September, 2014; Published in final edited form: July 2018

1. Introduction 1990; Youngs and Leeds-Harisson, 1990). Moreover,

The unfavorable effect of long lasting
conventional farming on soil structural characteristic
is well documented (e.g. Marks and Soane, 1987) and
many authors recommend the employment of
alternative farming practices to encounter harmful
effects (Pagliai et al., 1989; Shipitalo and Protz, 1987).
Soil porosity is generally considered an efficient
parameter of soil structure. Specifically, the size, the
shape and the connectivity of soil pores determine
important soil processes such as water and mineral
transportation, soil aeration etc. (Ringrose-Voase,
regarding pore size, the soil canal network constitutes
the habitat for a large array of soil fauna and microbes
(Elliott and Coleman, 1988).
To study soil micromorphometric traits,
techniques of image analysis are broadly used
especially for comparing the effects of alternative
farming practices on soil structure (Bui et al., 1989;
Kooistra et al., 1985; Livingston et al., 1985; Shipitalo
and Protz, 1987). Most relevant studies focus on soil
porosity (Pagliai et al., 1984, Ringrose-Voase et al.,
1990), as well as on void fractal dimension
(Mandelbrot, 1982). Hatano and Booltnik (1992) have

∗
Author to whom all correspondence should be addressed: e-mail: nmonokro@bio.auth.gr; Phone: +30 26510 07904; Fax: +30 26510 07064
 Monokrousos et al./Environmental Engineering and Management Journal 17 (2018), 7, 1555-1562
achieved to formulate a model adequately describing
the relationship between pore perimeter and surface.
The change in soil quality because of
management practice is a current topic of interest.
There is growing evidence that organically cultivated
systems exhibit improved soil quality than
conventional ones (Droogers and Bouma, 1996;
Monokrousos et al., 2006). However, very few studies
have been conducted regarding the changes that
organic farming might induce to the soil structure
during the conversion period from conventional to
organic farming and afterwards. Among them,
Gerhardt (1997) and Papadopoulos et al., (2006)
concluded that the management methods associated
with organic farming induce the formation of an
ameliorated soil structure, which is porous, better
developed and has increased organic matter content
and soil biota presence and activity. Based on
literature review, more studies have been conducted
on pores sized >50 μm in equivalent diameter,
whereas our research specifically focused our focused
on the study of soil micropores (<50 μm) on which
relevant information is limited (Menendez et al., 2005;
Pachepsky et al., 1996). These micropores play an
important role for agriculture as they are responsible
for water storage in the soil (Pagliai et al., 1981), form
the soil microaggregates, allowing the growth of root
hairs, and also form protective micro-habitats which
will allow the increase of soil microbial biomass
(Heijen and Van Veen, 1991). Thus, changes οf
morphological traits and numbers of soil micropores
in organically cultivated fields can be regarded as a
clear symptom of improved soil structure (Greenland,
1981).
In this paper we studied micromorphometric
traits of the void phase in four organic areas of
Asparagus officinalis (L), using image analysis of soil
thin sections. The areas differed regarding the age of
organic farming, so as to check for. We explore
questions relating as to whether the age of organic
farming influences the dimensions (surface,
equivalent diameter and perimeter), the shape
(eccentricity, solidity and compactness) and the fractal
dimension of soil micropores. Moreover, in an attempt
to associate micromorphometric traits with soil
biochemistry, we further sought for correlations
between the whole set of values relating to void phase
with the corresponding set of values of certain
biochemical variables.
2. Material and methods
2.1. Study area and sampling
The study area is located in Kria Vrisi 60 km
north-west of Thessaloniki, northern Greece. The
parent rock consists of Alluvial deposits (Holocene)
(IGMR, 1983). In order to evaluate the influence of
management practices on the soil profile, five
undisturbed vertical soil samples were collected with
cylindrical boxes (3 cm diameter, 4.5 cm high) from
each sampling area. The samples were taken in
1556
September 2003 from fields planted with the perennial
Asparagus officinalis (L). All fields were of similar
soil texture (Monokrousos et al., 2006). Five
organically cultivated fields and a conventional one
was sampled. The organic fields represented a
gradient in relation to the duration of organic
cultivation; 2 (O2), 3 (O3), 5 (O5) and 6 (O6) years.
Those cultivated organically for 2 and 3 years were
considered as fields in transition (Stanhill, 1990). All
organic fields were nearby to each other within a
continuous field area of about 30.000 m2. They were
conventionally cultivated for several years before
conversion to organic field. The conventional field
(CV) covered an area of 9.000 m2 and was located
almost 500 m away from the organic ones. It was
cultivated conventionally for more than 6 years.
2.2. Soil thin-section preparation
The samples were air-dried for a week and then
were dried in an oven at 350C for 48 h. The first step
included the placement of the boxes in a polyester
resin system applied under vacuum in the resin
impregnation process (Crystic17449, MEKP,
Accelerator ‘G’; Scot Bader) (Sole et al., 1992). The
resin droplets penetrated the soil pores by capillarity
until each soil sample was fully saturated with resin.
Impregnation time was 7 h and the hardening period 6
weeks. The resinous soil blocks were allowed to dry
over a period of 3 months, after which a thick section
was cut, mounted on a glass slide and polished to a
thickness of 20-25 μm. Then, the soil blocks were
bonded to polished glass slides (110 mm - 75 mm - 3
mm) with a standardized mass of epoxy (Epoxy 301;
Logitech), then cut and lapped. Thin sections were
then mechanically polished using 3 mm diamond
pasted.
2.3. Image acquisition and thin section porosity
analysis
To study the pore and soil particle distribution
the soil blocks were photographed using a digital
camera coupled with a polarizing microscope. Images
(752 x 582 pixels) were taken using a camera with an
objective magnification of 20x. Three replicates were
obtained from each block.
The use of polarized microscopy allowed the
discrimination of soil pores and matrix by color.
Minerals are represented by various colors, the
organic material by brown color, while the resin inside
the pores by black color. The segmentation and
classification of the digital image to patches was based
on these different colors as well as on differences on
the texture. Each soil section is considered a landscape
image and for image analysis the eCognition program
was used. The digital photos were transformed from
vector to raster and the results were processed using
Matlab. The following pore micromorphometric traits
were estimated: area, perimeter, equivalent diameter,
eccentricity, compactness and solidity. The first three
parameters stand for pore dimensions, while
 Dimension and structural traits of soil micropores in cultivations differing in the duration of organic management
eccentricity, compactness and solidity for the shape of
pores. Eccentricity is equal to the ratio of the distance
between the foci of the ellipse to the length of its major
axis. Eccentricity values range between 0 and 1 (zero
eccentricity corresponds to a circle while unity
corresponds to a line). To estimate compactness, we
used (Eq. 1)
( perimeter ) 2
C=
area
Finally, solidity is equal to the percentage of
the pore which is within a convex hull. The
relationship between area (A) and perimeter (P2) is
widely used in soil micromorphology to characterize
the shape of pores (Murphy et al., 1997). In this paper
we employed Swartz (1980) shape index (Eq. 2):
F=4πA/P2
This index is used to classify pore shapes into
elongated or planar (F<0.2), irregular (0.2<F<0.5) and
rounded (F>0.5) pores (Bouma et al., 1977, Mermut et
al., 1992). The pores were further subdivided into
three size classes (class 1: ≤10 μm, class 2: 10-20 μm,
class 3: 20-50 μm) according to either the equivalent
pore diameter for rounded and irregular pores and
their width for elongated pores (Pagliai et al., 1983).
The soil biochemical variables employed in
this paper are described in Table 1. These data relate
to the sampling period of September 2003 and were
presented in Monokrousos et al. (2008).
2.4. Statistical analyses
A correspondence analysis was used to
ordinate samples from various sampling areas with
regard of their relative frequencies of the three pore
size categories (≤10 μm, 10-20 μm, 20-50 μm). To
estimate the fractal dimension the linear model (Eq. 3)
was fitted on data:
Table 1. Mean values (± standard error) of soil chemical and biochemical variables
Biochemical variables
pH
Microbial C (μg g-1)
Microbial N (μg g-1)
NH4+ (μg g-1)
NO3- (μg g-1)
N-mineralization rate (μg g-1d-1)
C-mineralization rate (μg g-1d-1)
Organic C (mg g-1)
Organic N (mg g-1)
C/N ratio
P extractable (μg g-1)
Mg+2 (mg g-1)
K+ (mg g-1)
Ca+2 (mg g-1)
Log P= α + bLog S (3)
where: P stands for perimeter, S is surface, a is
elevation and b is slope.
Following Pachepsky et al. (1996) a piecewise
nonlinear regression model has been fitted. The
algorithm estimated parameters a1 and b1 for data up
to the breakpoint and a2 and b2 for data beyond the
breakpoint. To further seek for statistical differences
among slopes and elevations an Analysis of
(1) Covariance (ANCOVA) was performed. Finally, a
Canonical Analysis was performed to assess the
relationships between the pore micromorphological
traits and the soil biochemical variables.
3. Results and discussion
As it was revealed by estimating the Swartz
index, the large elongated pores contribute 64-72% to
(2) soil porosity, whereas the small rounded pores
participate with 12-17% (Fig. 1).
Pores were discriminated in three pore size
categories (≤10 μm, 10-20 μm, 20-50 μm) according
to their equivalent diameter. In Table 2 are presented
the average values of morphological traits in all
sampling sites for each pore size category. No
significant differences were revealed in relation to
management type (organic vs conventional farming)
in dimension and shape measurements, for any of the
three pore size categories. A One-Way ANOVA
accompanied by a Bonferroni test showed that
elongated pores occupy larger void surfaces in organic
plots of age 2 and 3 years.
The pores were further subdivided into three
classes according to the equivalent pore diameter for
the rounded and irregular pores and their width for the
elongated pores up to the size of 50 μm. The
percentage contribution of the three size categories is
shown in Fig. 2. The contribution of the small sized
pores in the total void of the soils is higher than the
other two categories, varying between 72-77%.
Mean values (±SE)
7.83 ± 0.09
170.79 ± 8.68
9.41 ± 0.32
20.43 ± 0.79
34.04 ± 1.17
0.26 ± 0.03
166.64 ± 13.08
12.53 ± 0.89
1.06 ± 0.05
12.02 ± 0.92
33.81 ± 3.73
0.31 ± 0.03
0.63 ± 0.09
0.22 ± 0.01
1557
 Monokrousos et al./Environmental Engineering and Management Journal 17 (2018), 7, 1555-1562

Fig. 1. Percentage contribution of the three pore shape categories in surface area in all sampling areas.
Different letters correspond to statistically significant differences, as indicated by ANOVA,
between the % contribution of the elongated pores of the sampling areas (p<0.001)

The results of the correspondence analysis
indicated that the first axis explain 84.32% of the
samples inertia. Samples from O6, O5 and O3 areas
were ordinated towards the left end point of the first
axis. Moreover, the smallest pore size category (≤10
μm) was ordinated to the left, while the middle pore
sized (10-20 μm) was ordinated in the middle and the
largest sized (20-50 μm) on the right end point of this
axis. These results indicate that the three oldest
organically farmed areas are characterized by small
sized (≤10 μm) pores while the newest organic area
and the conventional one are characterized mainly by
medium sized pores (10-20 μm).
The changes of microporosity due to different
cultivation management are a very important for
agriculture in general, because micropores determine
the amount of water available to plants and the soil
capacity for water storage (Pagliai et al., 1981). The
micropores whose diameter is about 10 μm form the
soil microaggregates, allowing the growth of root
hairs and providing the ground for a good soil
structure. Increase of micropores (0.2-50 μm) can be
regarded as a clear indication of improved soil
structure (Greenland, 1981). Soil structure is known to
have a large influence on the population dynamics,
ecology and activity of microorganisms (Stotzky,
1986). The small sized pores (≤10μm) form protective
micro-habitats which are large enough to allow
bacteria to enter but too small to be colonized by
predating protozoa (Heijen and Van Veen, 1991). As
shown the older organically cultivated fields in Krya
Vrisi are characterized by the smallest pore size
category and this is in accordance with the larger
amounts of microbial biomass recorded in these soils
(unpublished data). An increase of microbial biomass
is a clear indicator of improved soil quality and
therefore, it could be implied that the quality of soil
structure is improved after three years of continuous
biological cultivation (Fig. 3). Fractal dimension (D)
describes the geometry of a pore rather than its shape.
Dependencies of pore outline perimeter P on outlined
area A are shown in Fig. 4a-e. In Table 3 are given the
values of surface fractal dimension for voids as well
as the corresponding breakpoints.

Table 2. Mean values (± standard error) of morphological traits in all sampling sites of the three pore categories,
as these are shaped according to their size

Area Perimeter Eccentricity Solidity Compactness

≤10μm
O6 31.76 ± 0.64 21.91 ± 0.39 0.82 ± 0.01 0.80 ± 0.01 15.77 ± 0.47
O5 30.75 ± 0.33 21.78 ± 0.48 0.83 ± 0.01 0.79 ± 0.01 15.99 ± 0.55
O3 29.54 ± 0.60 20.83 ± 0.46 0.81 ± 0.01 0.81 ± 0.02 15.26 ± 0.58
O2 31.63 ± 1.24 21.73 ± 0.89 0.81 ± 0.01 0.81 ± 0.02 15.57 ± 0.83
CV 31.74 ± 0.92 21.35 ± 0.92 0.81 ± 0.01 0.83 ± 0.12 15.01 ± 0.53
10-20μm
O6 166.27 ± 5.32 79.13 ± 2.43 0.83 ± 0.01 0.64 ± 0.02 39.58 ± 1.97
O5 153.44 ± 3.77 76.65 ± 1.95 0.82 ± 0.01 0.62 ± 0.02 40.02 ± 1.88
O3 148.95 ± 3.69 74.83 ± 3.41 0.83 ± 0.01 0.65 ± 0.02 39.45 ± 3.23
O2 149.69 ± 3.91 68.53 ± 1.03 0.82 ± 0.01 0.68 ± 0.01 32.78 ± 1.18
CV 151.59 ± 3.37 69.57 ± 1.78 0.81 ± 0.01 0.68 ± 0.01 33.59 ± 2.12
20-50μm
O6 768.71 ± 16.53 281.09 ± 15.93 0.84 ± 0.01 0.54 ± 0.03 108.53 ± 10.67
O5 741.76 ± 31.84 267.18 ± 17.26 0.80 ± 0.01 0.54 ± 0.02 103.71 ± 11.44
O3 759.72 ± 45.63 283.77 ± 26.79 0.83 ± 0.01 0.55 ± 0.03 113.29 ± 16.38
O2 687.11 ± 25.02 235.96 ± 12.38 0.84 ± 0.01 0.56 ± 0.02 85.53 ± 6.88
CV 753.30 ± 20.05 240.11 ± 10.89 0.84 ± 0.01 0.59 ± 0.02 81.52 ± 6.56

1558
 Dimension and structural traits of soil micropores in cultivations differing in the duration of organic management

Fig. 2. Percentage contribution of the three pore size categories in the total void in all sampling areas

of water surplus and water retention under conditions

of water deficit (Yakovchenko et al., 1996).
Accordingly, it is concluded that more favorable
conditions for soil humidity and biological
transformations prevail in organic fields aged more
than 2 years.
Table 3. Fractal dimensions values and breakpoint for soil
samples from all sampling areas

D1 D2 Breakpoint
O6 1.216 1.474b 18.2 μm2
O5 1.272 1.540a 20.0 μm2
O3 1.242 1.538a 17.8 μm2
Fig. 3. Coordination of the pore size categories (1: ≤10 μm, O2 1.166 1.416b 19.1 μm2
2:10-20 μm, 3: 20-50 μm) and the sampling areas (O6: 6 CV 1.112 1.474b 18.6 μm2
yrs, O5: 5 yrs, O3: 3 yrs, O2: 2 yrs and CV: conventional
field) according to correspondence analysis Table 4. Results of ANCOVA on the lines from which
emerges the fractal dimensions D1 and D2 of the solid
phase of the sampling areas
Furthermore, fractal dimension values reported
for smaller voids (surface lower than that indicated by p-values
the corresponding breakpoint of the model) are lesser D1 D2
(1.112-1.272) than the fractal dimension values Elevation (a) 0.076 0.0001
estimated for their larger counterparts (1.416-1.54) Slope (b) 0.057 0.0001
(Table 3). The fact that the fractal dimension D1 for
the small pores is closer to 1 indicates their rounded To correlate the set of biochemical variables
geometry. with the set of micromorphometric variables of the
Analysis of covariance showed that the voids a Canonical Correlation Analysis was employed
elevations and the slopes of the relation for the for each pore size category separately. In all size
smallest pores did not differ statistically among the categories only the first root of the analysis was
areas (Table 4). Thus, no significant differences were statistically important, which accounts for the 87-96%
detected between fractal dimensions of pore outlines of the between sets correlation. Furthermore, the
among the sampling areas (Table 3). In contrast, analysis showed that the relationship between the
elevations and slopes for larger voids were found to be analyzed sets was indeed linear, while no point clouds
affected significantly by the age of farming (Table 4). are depicted on the relevant graphs (not provided).
Fractal dimension D2 is significantly larger in samples Accordingly, the linear correlation is homogeneous
from the O3 and O5 areas than in samples from the along the whole range of values. For all pore classes,
other areas (Table 3). According to Kampichler and significant are those correlations between
Hauser (1993) the variations in the fractal dimension morphometric variables, which are describing the
may result in variations in the available habitat for structure of the pores such as solidity and
different microbial species. The higher fractal compactness, and biochemical variables such as N
dimension could be related to more heterogeneous microbial, Ca+2, Mg+2, K+ and the rate of C-
distribution of the microbial community in space. mineralization. More specifically, in small sized pores
Highest D2 fractal dimension estimated for O3 and O5 (≤10 μm), a correlation exists between the pore
indicating more favorable biological transformation solidity and compactness and the N-microbial, NO3-,
processes as well as easy infiltration of water in case K+ and Ca+2. In the middle size category (10-20 μm)

1559
 Monokrousos et al./Environmental Engineering and Management Journal 17 (2018), 7, 1555-1562

the correlation is noticed between pore solidity and N-
microbial, N mineralization rate, K+, P extractable,
Ca+2 and Mg+2 concentrations, while in the larger size
category (20-50 μm) the correlation is between
solidity and compactness and the rate of C-
mineralization as well as the Mg+2 and K+
concentrations. Pores with equivalent diameter of <50
μm are considered to be the storage pores of soil. The
root hairs grow along these pore channels and
sometimes modify and enlarge them, at the same time
absorbing nutrients, water and oxygen via the adjoin
soil material (De Willigen and Van Noordwijk, 1987).
The correlation between the availability of soil
nutrients and the structure of voids could be explained
by the fact that the latter is driven to a large extent by
growing root hairs.
In the middle and larger (10-50 μm) pore size
categories the correlation between the N- and C-
mineralization rates with the structural morphometric
traits indicate high microbial activity. Strong et al.
(2004) also found that the decomposition rate of
organic C is faster in medium sized pores with neck
diameter 15-60 μm than in smaller ones. This is due to
greater microbial activity since this region supports
both microbial motility and best oxygen supply. Fungi
have been found to be more abundant in this pore class
(Strong et al., 2004) supporting larger numbers of
fungal-feeding nematodes (Jones and Thomason,
1976, van der Linden et al., 1989). Thus the fast rate
of decomposition of organic C in the 20-50 μm sized
could be enhanced by the nematode activity (Table 5).

Fig. 4. Dependencies of the pore outline perimeter (P) on the outlined area (A) for samples from all sampling areas:
(a) O2, (b) O3, (c) O5, (d) O6 and (e) CV

1560
 Dimension and structural traits of soil micropores in cultivations differing in the duration of organic management

Table 5. Factor structure of canonical analysis for chemical and biochemical and morphometric variables

Variables ≤10 μm 10-20 μm 20-50 μm

perimeter 0.272 -0.340 0.190
eccentricity -0.628 0.596 0.047
Morphometric
solidity -2.653 1.088 0.391
compactness -2.237 0.322 -0.801
pH 0.153 0.052 0.101
C microbial 0.171 0.299 -0.226
N microbial -0.486 -0.386 -0.109
NH4+ 0.247 -0.109 -0.005
NO3- -0.306 0.001 0.232
N-mineralization -0.190 0.379 -0.153
C-mineralization 0.050 0.180 -0.518
Biochemical C organic -0.183 -0.132 0.199
N organic -0.213 0.190 0.261
C/N ratio -0.153 -0.109 0.152
P extr -0.252 -0.516 -0.387
Mg+2 0.043 -0.457 -0.660
K+ 0.386 0.901 0.659
Ca+2 -0.380 -0.449 0.323

In the small pore size category (≤10 μm) N-
microbial and the concentrations of NO3- are also
correlated with the structure characteristics. It is
noticeable that these biochemical parameters are
involved in the nitrogen cycle. Consequently, it seems
plausible that in this region there are dominant
processes that transform nitrogen to absorbable by the
root hairs forms. Taking into account that the older
organic fields (O6, O5, O3) are characterized mainly
by the small sized pores (≤10 μm), we could suggest
that the improved soil quality in these field is rather
related with the N-cycle.
4. Conclusions
The duration of organic farming is reflected on
the dimension and structural characteristics of
micropores. Areas organically cultivated for more
than two years are mainly characterized by small sized
pores (≤10 μm), while the soil of organic cultivations
with duration of three and five years consist of pores
with higher fractal dimension supporting a more
heterogeneous distribution of microbes in space. In the
small pore size category (≤10 μm), N-microbial and
NO3- concentrations, parameters involved in the
nitrogen cycle, were found to be correlated to the
structure characteristics. All the changes related to the
properties of soil microporosity took place after the 2-
3 years’ time period that is necessary for the
conversion of a cultivation from a conventional to an
organic one. These results support the idea that
alternative cultivation practices, such as organic
farming result in improved soil structure and quality.
References
Bouma J., Jongerius A., Boersma O., Jager A.
Schoonderbeek, D., (1977), The function of different
types of macropores during saturated flow through four
swelling soil horizons, Soil Science Society of America
Journal, 41, 945-950.
Bui E.N., Mermut A.R., Santos M.C.D., (1989),
Microscopic and ultramicroscopic porosity of anoxisol
as determined by image analysis and water-retention,
Soil Science Society of American Journal, 53, 661-665.
De Willigen P., Van Noordwijk M., (1987), Roots, plant
production and nutrient use efficiency, PhD thesis,
Agricultural University, Wageningen, Holland.
Droogers P., Bouma J., (1996), Biodynamic versus
conventional farming effects on soil structure expressed
by simulated potential productivity, Soil Science
Society American Journal, 60, 1552-1558.
Elliott E.T., Coleman D.C., (1988), Let the soil work for us,
Ecological Bulletins, 39, 23-32.
Greenland D.J., (1981), Soil management and soil
degradation, Journal of Soil Science, 32, 301-322.
Hatano R., Booltink H.W.G., (1992), Using fractal
dimensions of stained flow patterns in a clay soil to
predict bypass flow, Journal of Hydrology, 135, 121-
131.
Heinjen C.E., van Veen J.A., (1991), A determination of
protective microhabitats for bacteria introduced into
soils, FEMS MicrobiologyEcology, 85, 73-80.
IGMR, (1983), Institute of Geology and Mining research,
Geological Map of Greece, scale 1:500000, Athens, On
line at
https://www.researchgate.net/institution/Institute_for_
Geological_and_Mining_Research.
Jones F.G.W., Thomasson A.J., (1976), Bulk density as an
indicator of pore space in soils useable by nematodes,
Nematologia, 22, 133-137.
Kampichler C., Hauser M., (1993), Roughness of soil pore
surface and its effect on available habitat space of
microarthropods, Geoderma, 56, 223-232.
Kooistra M.J., Bouma J., Boursma L., Jager A., (1985), Soil
structure differences and associated physical properties
of some loamy typic fluvaqvents in the Netherlands,
Geoderma, 36, 215-228.
Livingston S.J., Norton L.D., West L.T., (1985), Effect of
long-term cultivation on aggregate stability, organic
carbon distribution and porosity of two soil series,
Developments in Soil Science, 19, 89-95.
Mandelbrot B.B., (1982), The Fractal Geometry of Nature,
vol. 173, 3th Edition, New York.
Menendez I., Caniego J., Gallardo J.F., (2005), Use of
fractal scaling to discriminate between and macro- and

1561
 Monokrousos et al./Environmental Engineering and Management Journal 17 (2018), 7, 1555-1562
meso-pore sizes in forest soils, Ecological Modeling,
182, 323-335.
Mermut A.R., Grevers M.C.J., Dejong E., (1992),
Evaluation of pores under different management-
systems by image analysis of clay soils in
Saskatchewan, Canada, Geoderma, 53, 357-372.
Marks M.J., Soane G.C., (1987), Crop and soil response to
subsoil loosening, deep incorporation of phosphorus
and potassium fertilizer and subsequent soil
management on a range of soil types 1, Soil Use and
Management, 3, 115-123.
Monokrousos N., Papatheodorou E.M., Diamantopoulos J.
D., Stamou G.P., (2006), Soil quality variables in
organically and conventionally cultivated field sites,
Soil Biology and Biochemistry, 38, 1282–1289.
Monokrousos N., Papatheodorou E.M., Stamou G.P.,
(2008), The response of soil biochemical variables to
organic and conventional cultivation of Asparagus sp.,
Soil Biology and Biochemistry, 40, 198-206.
Murphy C.P., Bullock P., Turner R.H., (1997), The
measurement and characterization of voids in soil thin
sections by image analysis. Part I. Principles and
techniques, Journal of Soil Science, 28, 498-508.
Pachepsky Y., Yakovchenko V., Rabenhorst M.C., Pooley
C., Sikora L.J., (1996), Fractal parameters of pore
surfaces as derived from micromorphological data:
effect of longterm management practices, Geoderma,
74, 305-319.
Pagliai M., Guidi G., La Marca M., Giachetti M., Lucamante
G., (1981), Effects of sewage sludges and composts on
soil porosity and aggregation, Journal of
Environmental Quality, 10, 556-561.
Pagliai M., La Marca M., Lucamante G., (1983),
Micromorphometric and micro-morphological
investigations of a clay loam soil in viticulture under
zero and conventional tillage, Geoderma, 30, 35-53.
Pagliai M., Lamarca M., Lucamante G., (1984), Effects of
zero and conventional tillage on the length and
irregularity of elongated pores in a clay loam soil under
viticulture, Soil and Tillage Research, 4, 433-444.
Pagliai M., Pezzarossa B., Zerbi G., (1989), Soil porosity in
a peach orchand as influenced by water-table depth,
Agricultural Water Management, 16, 63-73.
1562
Papadopoulos A., Mooney S.J., Bird N.R.A., (2006),
Quantification of the effects of contrasting crops in the
development of soil structure: an organic conversion.
Soil Use and Management, 22, 172-179.
Ringrose-Voase A.J., (1990), One dimensional image
analysis of soil structure, Journal of Soil Science, 41,
499-512.
Shipitalo M.J., Protz R., (1987), Comparison of morphology
and porosity of a soil under conventional and zero
tillage, Canadian Journal of Soil Science, 67, 445-456.
Sole A., Josa R., Rardini G., Aringhieri R., Plana F., Gallart
F., (1992), How mudrock and soil physical properties
influence badland formation at Vallcebre (Pre-
Pyrenees, NE Spain), Catena, 19, 287-300.
Stanhill G., (1990), The comparative productivity of organic
agriculture, Agriculture, Ecosystems and Environment,
30, 1-26.
Stotzky G., (1986), Influence of Soil Mineral Colloids on
Metabolic Processes, Growth, Adhesion and Ecology of
Microbes and Viruses, In: Interactions of Soil Minerals
with Natural Organics and Microbes, Huang P.M.,
Schnitzer M. (Eds), Madison WI, 305-428.
Strong D.T., de Wever H., Merckx R., Recous S., (2004),
Spatial location of carbon decomposition in the soil
pore system, European Journal of Soil Science, 55,
739-750.
Swartz H., (1980), Two-dimensional feature shape indices,
Mikroskopie (Wien Supplement), 37, 64-67.
Van der Linden A.M.A., Jeurissen L.J.J., van Veen J.A.,
Schippers B., (1989), Turnover of Soil Microbial
Biomass as Influenced by Soil Compaction, In:
Nitrogen in Organic Wastes Applied to Soils, Hansen
J.A., Henriksen K. (Eds), Academic Press, London, 25-
36.
Yakovchenko Y., Sikora L.J., Kaufman D.D., (1996), A
biologically based indicator of soil quality, Biology and
Fertility of Soils, 21, 245-251.
Youngs E.G., Leeds-Harrison P.B., (1990), Aspects of
transport processes in aggregated soils, Journal of Soil
Science, 41, 665-675.
Environmental Engineering and Management Journal July 2018, Vol.17, No. 7, 1563-1573
http://www.eemj.icpm.tuiasi.ro/; http://www.eemj.eu

“Gheorghe Asachi” Technical University of Iasi, Romania

CHANGES OF ARSENIC SPECIATION DURING SWINE MANURE

WINDROW COMPOSTING AT FULL SCALE

Rui Wang1, Yuansong Wei1,3∗, Yulong Wu2, Jiaxi Zheng1, Bofu Zheng2
1State Key Joint Laboratory of Environmental Simulation and Pollution Control, Research Center for Eco-Environmental
Sciences, Chinese Academy of Sciences, Beijing 100085, P.R. China
2Nanchang University, Nanchang, 330031, P.R. China
3Ordos Institute of Solid Waste Technology, Research Center for Eco-Environmental Sciences,

Chinese Academy of Sciences, Ordos, 017000, P.R. China

Abstract

Organoarsenic compounds are used extensively as swine feed-additives to promote animal growth and control diseases, e.g.
roxarsone (4-hydroxy-3-nitrobenzenearsonic acid) and arsanilic acid (4-aminophenylarsonic acid) are usually overused in swine
industry in China. As a cost-effective technology for animal manure treatment and reuse, composting is widely used for swine
manure treatment. Therefore, the purpose of this study was to investigate changes of arsenic species and fractions during a full
scale swine manure windrow composting. Results showed that four arsenic species, arsenite (As (III)), arsenate (As (V)),
monomethylarsonate (MMA) and dimethylarsinate (DMA), were detected and varied differently which arsenate was on a slightly
increasing trend and arsenite presented a decreasing trend throughout the composting process. The percentages of arsenite, arsenate
and DMA were 24.10%, 29.68%, 4.87% of total arsenic in the beginning of the composting and 8.19%, 34.13%, 4.75% of total
arsenic respectively at the end of the composting. Arsenate had the similar trend to Fe/Mn oxides arsenic fraction, which decreased
dramatically in the first 10 days and then was on increasing trends during the rest of the composting. The percentages of Fe/Mn
oxides, residual, carbonates, organic arsenic fractions were 55.3%, 25.9%, 18.6%, 0% in the beginning and 46.6%, 22.7%, 19.1%,
11.5% at the end of the composting, respectively. These results demonstrated that the most toxic arsenic species, arsenite, was
declined dramatically during the composting process, and composting was a suitable method for detoxification of arsenic by
changing their forms.

Key words: arsenic fraction, arsenic species, composting, swine manure

Received: March, 2014; Revised final: August, 2014; Accepted: September, 2014; Published in final edited form: July 2018

1. Introduction et al., 2017). Roxarsone (4-hydroxy-3-

In the last several decades, along with the rapid
development of intensive livestock and poultry
industry in China, a great amount of animal manure
was produced, e.g., 243 million metric-tons in 2007
(The first national source census bulletin of China),
and may pose environmental problems, such as odor,
pollution of ground and surface waters due to leaching
and run-off of organics and nutrients, and soil
accumulation of heavy metals (Hsu and Lo, 2001; Yan
nitrophenylarsonic acid, C6H6NO6As, 28.5% As) and
arsanilic acid (p-aminophenylarsonic acid,
C6H8NO3As, 34.6% As) are aromatic organoarsenic
compounds widely used as feed additives in the
livestock and poultry industry for improving feed
efficiency, promoting growth, and controlling diseases
(Chapman and Johnson, 2002). Generally, roxarsone
is incorporated in the feed at levels of 25-50 mg·kg-1,
and arsanilic acid is at 50-100 mg·kg-1 (Sierra-Alvarez
et al., 2010). Several studies reported that arsenic

∗
Author to whom all correspondence should be addressed: e-mail: yswei@rcees.ac.cn, 15810948239@163.com; Phone: +86 10 62849690, Fax:
+86 10 62849109
 Wang et al./Environmental Engineering and Management Journal 17 (2018), 7, 1563-1573
concentrations in pig feeds ranged from 0.02 ~ 13.03
mg·kg-1 (Zhang et al., 2012), and average contents of
roxarsone and arsanilic acid were at 7.0 and 21.2 As
mg·kg-1, respectively (Yao et al., 2013). Additionally,
arsenite (As(III)) and monomethylarsonate (MMA)
were frequently detected due to arsenic impurities in
feeds bearing organoarsenic (Yao et al., 2013).
As a low bioaccumulation potential, roxarsone
and arsanilic acid are largely excreted unchanged in
animal manure (Garbarino et al., 2003). For example,
the distribution of arsenic species in fresh poultry litter
was 91% roxarsone, 1.5% dimethylarsinate (DMA),
1.1% arsenate (As(V)), 0.8% arsenite (As(III)) and
5.6% of other unknown arsenic compounds
(Garbarino et al., 2003). Based on broiler production
and roxarsone feed dosage, it is estimated that
approximately 900 metric tons of roxarsone are
released into environment in the U.S. annually by the
poultry industry alone, an amount that is equivalent to
250 metric tons of arsenic (Rutherford et al., 2003).
Several literatures reported that livestock and poultry
manure contained a great amount of arsenic, e.g.
arsenic concentration in swine manure in different
regions of China varied significantly from 0.004
mg·kg-1 DW (dry weight) to 194.50 mg·kg-1 DW
(Dong et al., 2008; Li and Chen, 2005; Long et al.,
2004; Zhang et al., 2012; Zhang et al., 2005).
Additionally, other investigation demonstrated that
the annual amount of arsenic input to the soil applied
with swine manures as organic fertilizers was 340.60
g· hm-2, and the increasing concentrations of arsenic
in arable soils was 0.15 mg·kg-1 (DW) annually (Dong
et al., 2008). With long-term application of swine
manure as organic fertilizers on agricultural lands,
organic arsenic could be converted into its inorganic
form, making it water-soluble and allowing it to seep
into the surface and groundwater (Garbarino et al.,
2003; Rutherford et al., 2003), which ultimately had a
great influence on plants, soil organisms and human
health (Carbonell-Barrachina et al., 1999).
As the bioavailability and toxicity of arsenic
compounds vary dramatically, the investigation of
arsenic degradation, transformation, migration,
environmental impact and human health risk should be
based on measurement of arsenic speciation as well as
total arsenic concentration (Al-Assaf et al., 2009).
Although the toxicity of organic arsenics, such as
arsanilic acid and roxarsone, is relatively low, their
degradation and transformation can happen and thus
form the more toxic metabolites, such as arsenic acid,
arsenous acid and monomethylarsonous acid during
their exposure to environment (Garbarino et al., 2003).
The toxicity of different arsenic species varies in the
order as following: As (III) > As (V) >> MMA > DMA
(Jain & Ali, 2000). Additionally, the bioavailability of
arsenic associated with different fractions have
dissimilar impacts on the environment and their
phytotoxicity connects to some forms rather the total
content (Hanc et al., 2011). The sequential chemical
extraction of elements can provide an understanding
of chemical fractions and a predicting mobility,
bioavailability and leaching rate (Tessier et al., 1979).
1564
Hanc et al., (2011) demonstrated that bioavailability
arsenic fractions, such as exchangeable arsenic,
decreased during the kitchen waste composting, and
the order of arsenic fractions in the final compost was
as follows: residual > oxidizable > exchangeable >
reducible. Although there were several researches
related to arsenic species (such as arsenite, arsenate
and so on) and arsenic fractions (exchangeable,
carbonates and so on) during the composting, the
relationship between them is not clear and needs
further study.
Composting, one of the most suitable
technologies for treating livestock manures can reduce
the mass, degrade the organic matter, provide
sufficient sanitation effect and produce valuable end
products (Jiang et al., 2011). In general, the chemical
speciation of As is particularly important in waste
management (Carbonell-Barrachina et al., 2000).
Garbarino et al., (2003) studied the degradation of
roxarsone in a lab-scale composting experiment,
which demonstrated that roxarsone degraded easily
and could enter into the environment in other form of
arsenic, such as arsenate. Arsenate is the
thermodynamically stable form under aerobic
conditions, of which one or two protons are
dissociated at natural soil/sediment pH value, and
arsenate is a chemical analogue of phosphate and may
interfere with oxidative phosphorylation (Ter Welle
and Slater, 1967). Additionally, arsenite is a neutral
species at natural pH value and inhibits the activity of
enzymes by binding to thiol groups. MMA and DMA
also form anions in soil/sediment (Carbonell-
Barrachina et al., 2000). Arsenic is subject to
chemically and/or microbiologically mediated
oxidation-reduction and methylation reactions in soils,
sediments or other medium (Masscheleyn et al.,
1991).
Once the residual arsenic in animal manure is
released into the environment with land application of
animal manure as fertilizer, arsenic may pose great
threat potential to food safety and environment
quality. There are growing interests about the arsenic
residues accumulation in animal manure and the soil
(Luo et al., 2009; Zeng et al., 2007; Zhang et al.,
2012), however, little work has been done to
investigate the different arsenic species and fractions
changed in the process of full scale swine manure
composting. The objective of this study was thus to
investigate different arsenic species and fractions
variation during the full scale swine manure windrow
composting to provide support for arsenic pollution
control in animal wastes treatment and disposal.
2. Material and methods
2.1. Full-scale swine manure composting
A full scale of swine manure composting was
carried out from May 22 to July 16, 2011 in a windrow
composting plant located in the suburb of Beijing (N
39°42', E 116°38'), to investigate degradation of
organic arsenic compounds and its metabolic
 Changes of arsenic speciation during swine manure windrow composting at full scale
products. Raw materials for composting were mixed
of swine manure and mushroom residues at the ratio
of 2: 3 (v/v). Fresh swine manure and mushroom
residues were sampled and stored at 4ºC before
composting for analysis. Parameters of all composting
materials, including pH, moisture content, organic
matter (OM), TC, TN, and TP, were determined
according to the methods (Bao, 2000) and the results
were shown in Table 1. The size of the windrow pile
was about 75 m3, at the length of 85 m, the height of
0.9 m and the width of 2 m at the bottom of the pile.
The full scale windrow composting was operated in
the open air and turned by a turner (BACKHUS 15.30,
Germany) once a week in the thermophilic stage, and
then the pile was moved and stored in a simple shed in
the mature stage in which there was no turning. During
the composting process, water was added irregularly
(almost once a week) to maintain stable moisture
content of the composting materials. The pile
temperature, including upper, central and lower part,
was recorded daily. The composting materials was
sampled by mixing the upper, central and lower sites
of the pile on day 1, 3, 7, 10, 23, 38 and 53,
respectively. All the samples collected were stored at
4oC before analysis.
Table. 1. Characteristics of raw materials in the full scale
swine manure windrow composting
Materials
Characteristics Swine
Mushroom
manure
pH 7.69 7.64
Moisture (%) 72.15 46
OMa (%) 68.9 87.0
TCb (g/kg) 331.42 395.28
TNc (g/kg) 25.98 22.55
TPd (g/kg) 12.01 4.69
C:N ratio 12.75 17.53
a. OM: organic matter; b. TC: total carbon; c. TN: total nitrogen;
d. TP: total phosphorus
2.2. Reagents and standards
Six high purity chemicals, including sodium
arsenite (NaAsO2 99.8%, Fluka, Switzerland), sodium
arsenate (Na3AsO4·7H2O 98.5%, Fluka, Switzerland),
disodium methyl arsenate (MMA,
(CH3)AsO3Na2·6H2O 99.5%, Sigma-aldrich, USA),
cacodylic acid (DMA, (CH3)2AsO(OH) 99.4%,
Sigma-aldrich, USA), aminophenyl arsenic acid (4-
arsanilic acid, C6H4AsO(OH)2NH2 99%, Dr.
Ehrenstorfer Gambh, Germany) and 4-Hydroxy-3-
nitrobenzenearsonic acid (roxarsone,
C6H3AsO(OH)2(OH)NO2 97.5%, Dr. Ehrenstorfer
Gambh, Germany) were purchased and used as
standard substances for determining 6 arsenic species
(arsenate, arsenite, DMA, MMA, roxarsone and
arsanilic acid) in this study. All of the following
chemicals, including phosphoric acid (H3PO4),
sodium hydroxide (NaOH) used as arsenic species
extraction buffers, ammonium hydroxide (NH4OH),
ammonium dihydrogen phosphate (NH4H2PO4),
ammonium nitrate (NH4NO3) used as mobile phases
of liquid chromatogram, magnesium chloride
(MgCl2), sodium acetate (NaOAc), acetic acid
(HOAc), nitric acid (HNO3), hydrogen peroxide
(H2O2), perchloric acid (HCIO4), hydrofluoric acid
(HF), hydroxylamine hydrochloride (NH2OH·HCl)
used as arsenic fraction extraction buffers were of
analytical pure grade and purchased from Sinopharm
Chemical Reagent Co. Ltd., China. All solutions were
prepared in ultra-pure water from a Millipore Milli-Q
system (Millipore, USA).
Phosphoric acid and ammonium hydroxide
were used for pH adjustment. Phosphate buffer
solution (PBS) was prepared from ammonium
dihydrogen phosphate and phosphoric acid. The daily
working standards for arsenic species were made from
standard stock solutions (1000 mg·L-1) prepared from
the six arsenic chemicals dissolved in ultra-pure water.
All these stock solutions of these 6 standards above
were kept in the dark at 4 ºC.
All mobile phases, standards and samples were
filtered through 0.45 µm polyethersulfone membrane
filters prior to analysis.
2.3. Analytical procedures
To determine different arsenic species, all of
composting samples stored at -20 ºC were firstly
freeze-dried (ALPHA1-2LD PLUS, Christ, Germany)
and then sieved by nylon screen with mesh size of 100,
Mixture and finally analyzed by using an improved method
7.39 described by Al-Assaf et al. (2009) and Garbarino et
47.28 al. (2006). Each sample at 1.0 g was extracted with 5
78.9 mL 0.10 mol·L-1 phosphoric acid solution (H3PO4)
223.62 and 5 mL 0.10 mol·L-1 sodium hydroxide solution
15.05 (NaOH) followed by shaking for 24 hours,
16.77 centrifugation at 7000 r·min-1 for 10 min and filtration
14.86 with 0.45µm filter. The extraction solution of arsenic
was used for speciation analysis by high-performance
liquid chromatography (HPLC) to inductively coupled
plasma mass spectrometry (ICP-MS) (Agilent 7700x,
Agilent Technologies, Japan).
To determine the total arsenic concentrations
of compost materials, 1.0 g sample was digested with
5mL HNO3 (69%) and 5mL HClO4 for 24 hours and
filtered through 0.45 µm polyether sulfone membrane
filters for ICP-MS analysis.
In order to determine the extraction efficiency
of the method mentioned above, samples were spiked
with the standard mixture solutions of arsenite,
arsenate, DMA, MMA, roxarsone and arsanilic acid at
three levels: 0.01 mg·kg-1 DW, 0.1mg·kg-1 DW and
5.0 mg·kg-1 DW, respectively. Samples were extracted
and analyzed, and the results showed that recoveries
of all target compounds varied from 89%±3.27% to
93.25%±2.38%, which suggested that it is an efficient
procedure for extracting arsenic species from compost
materials.
For better understanding the phytotoxicity,
bioavailability and mobility of arsenic, five different
arsenic fractions were analyzed during the swine
manure windrow composting. The sequential
1565
 Wang et al./Environmental Engineering and Management Journal 17 (2018), 7, 1563-1573
chemical extraction of arsenic was applied according
to Tessier et al. (1979) for the assessment of different
arsenic fractions including exchangeable (1M MgCl2,
pH=7), bound to carbonates (1M NaOAc, pH=5),
bound to Fe/Mn oxides (0.04 M NH2OH·HCl in 25%
(v/v) HOAc), bound to organic matter (0.02 M HNO3
+ 30% H2O2, pH=2), residual (digested with HF-
HClO4) in the swine manure windrow composting
samples. All of the arsenic extraction solutions were
stored at 4 ºC and arsenic concentrations were
determined by ICP-MS (Agilent 7700x, Agilent
Technologies, Japan).
2.4. Instrumentation
The arsenic speciation analysis was determined
by a coupling of HPLC with ICP-MS. In order to meet
the separation object, an ion exchange columns
(4.6×250 mm) manufactured by Agilent was selected
and isocratic elution using 10 mmol·L-1 ammonium
dihydrogen phosphate (pH=6.2) and 10 mmol·L-1
ammonium nitrate as mobile phase was applied. The
mobile phase was set at a flow rate of 1.0 mL·min-1
and the injection volume was 20 µL.
The ICP-MS has been used to determine
elements in natural-water, biota, sediment, soil
samples and other mediums since the early 1990s
(Garbarino et al., 2003). In this study, samples were
analyzed using an Agilent 7700x (Agilent
Technologies, Japan) at ambient temperature. The
Agilent 7700x is fitted with quartz, low-volume,
peltier-cooled, scott-type double-pass spray chamber
and a concentric MicroMist (borosilicate glass)
nebulizer.
The ICP-MS was tuned daily using a 1 µg·L-1
tuning solution containing 7Li, 89Y, 205Tl (Agilent
Technologies, Japan). The ICP-MS was used to
quantify the arsenic species at m/z 75. A helium gas
Fig. 1. Changes of pile temperature during the full scale swine manure windrow composting
1566
collision cell was utilized for the removal of
polyatomic interferences of 40Ar35Cl+, which usually
affect the determination of 75As when using ICP-MS.
The method separated the arsenic species within 13
min and the limits of detection for different arsenic
species ranged from 0.1 to 0.2 µg·L-1.
3. Results and discussion
3.1. Pile temperature
During 53 days composting, including 23 days
of the thermophilic stage and 30 days of the mature
stage, the pile temperature increased to approximately
55oC rapidly in the first 10 days (Fig. 1), and then
sustained above 55 ºC for more than 7 days, which met
with the requirements of sanitary standard for the non-
hazardous treatment of night soil in China (GB-7959-
87). Generally, the pile temperature in the central layer
was much higher than those in the bottom and the top.
The pile temperature increased successively up to over
60 ºC in the thermophilic stage then kept increasing to
70 ºC in the mature stage and decreased slightly but
still around 60 ºC at the end of the composting.
The pile temperature is an important factor of
affecting composting performance due to its influence
on the activity and diversity of microorganisms
(Finstein et al., 1986). In general, high pile
temperatures are consequences of biological activity
and have been considered a necessary condition for
good composting (de Bertoldi et al., 1983). In the
beginning of composting, mesophilic microorganisms
are initially present, and their activity raises the pile
temperature (Poincelot, 1975). As the pile temperature
kept climbing, thermophilic microorganisms
predominated gradually during the thermophilic stage
and parts of organic matter were decomposed in this
process.
 Changes of arsenic speciation during swine manure windrow composting at full scale
In the mature stage, the pile temperature
decreased, and mesophilic microorganisms kept
decomposing complex organic matter and the
composting was in an anaerobic condition because of
no turning (Miyatake and Iwabuchi, 2006). In this
study, due to low turning frequency of composting pile
and high ambient temperature in summer, the pile
temperature kept increasing and sustained highly
during the thermophilic stage and the mature stage
which was up to 70ºC. However, high pile
temperatures in fact inhibit growth of the majority of
microorganisms, thus slow down decomposition of
organic matter. Only a few species of thermophilic
sporigenous bacteria show a metabolic activity above
70ºC (de Bertoldi et al., 1983).
3.2. pH value, moisture content and organic matter
The pH values went up from 5.3 to 8.1,
increasing by more than two pH units during the
composting process (Fig. 2). This is attributed to
ammonia nitrogen generation from the organic matter
degradation. During the first 7 days, the pH values
fluctuated a little, e.g., it rose in the first 3 days and
then fell rapidly on the 7th day. This was a
consequence of the activity of acid-forming bacteria
which break down complex carbonaceous material
into organic acid intermediates (de Bertoldi et al.,
1983) in the early stage of swine manure windrow
composting. Organic matter and moisture contents
decreased 7.8% and 23.3%, respectively, as shown in
Fig. 2. In the process of composting, complex
physical, chemical, and biological reactions occurred,
i.e., organic matter was degraded while metals and
metalloids, such as arsenic, could potentially be
Fig. 2. Changes of moisture content, organic matter and pH value during the full scale swine manure windrow composting
transferred to other forms (Pinel-Raffaitin et al.,
2007).
3.3. Arsenic species and fractions
To determine changes of arsenic species during
the swine manure windrow composting, 6 forms of
arsenic compounds were investigated, including
arsenite, arsenate, MMA, DMA, arsanilic acid and
roxarsone. The results showed that neither arsanilic
acid nor roxarsone was detected in the composting
process, which demonstrated that these organoarsenic
compounds were degraded before the swine manure
composting in accordance with results of Garbarino et
al., (2003). Garbarino et al., (2003) studied that the
rate of roxarsone degradation increased with the
temperature. When the temperature was up to 40 ºC,
60% of roxarsone had degraded within 40 hours. In the
present study, raw material of composting was stored
in the open air for a few days before composting and
the pile temperature was up to 40 ºC. Therefore,
roxarsone and arsanilic acid had already degraded
before composting. The secondary metabolite was
arsenite that decreased from 24% (on the first day) to
8% (on the 53th day) of total arsenic. The organic
arsenic forms, including MMA and DMA, fluctuated
a little but maintained a stable low level at
approximately 5% of total arsenic. Furthermore,
unidentified arsenic compounds accounted for
approximately 27% to 60% in this process and need
further study in the future.
As shown in Fig. 3a, the main arsenic
metabolite was arsenate that varied drastically from
about 30% to 60% of total arsenic throughout the full
scale swine manure windrow composting.
1567
 Wang et al./Environmental Engineering and Management Journal 17 (2018), 7, 1563-1573

As shown in Fig. 3b, arsenate concentration
declined roughly from about 400 µg·kg-1 to almost
below the limit of detection during the first 7 days,
however, it went up dramatically to 727 µg·kg-1 from
the 7th day to the 10th day and then increased slowly
until the 40th day and finally dropped a little at the end
of the swine manure windrow composting. It was
interesting that the variation of arsenite was totally
opposite to that of arsenate, in other words,
concentration of arsenite went up when that of
arsenate went down obviously, vice versa. These
results showed that concentrations of arsenite reached
to the maximum (437 µg·kg-1) on the 7th day of the
swine manure windrow composting, while
concentration of arsenate dropped down to the
minimum (below the limit of detection).
Composting is an aerobic process which uses
naturally occurring microorganisms to convert
biodegradable organic matter into a humus-like
product (Georgacakis et al., 1996). Arsenic species
variation depended on many factors, including
temperature, pH value, moisture content, redox
potential and microorganism activity.
Additionally, the dominant form of arsenic in
oxygenated situation is arsenate (Byrd, 1990; Neff,
1997). This might explain why arsenate was the
primary arsenic species in the full scale swine manure
windrow composting process (Fig. 3a.).

Fig. 3. Changes of different arsenic species during the full-scale swine manure windrow composting:
(a) Percentage, (b) Concentration

1568
 Changes of arsenic speciation during swine manure windrow composting at full scale
Moreover, the other inorganic arsenic form,
arsenite, appeared in all of the compost samples. With
alteration of some composting factors, including
temperature, pH value and moisture content, both
inorganic arsenic forms, arsenate and arsenite,
transformed mutually in the composting process (Fig.
3b.). Garbarino studied that the moisture content and
the temperature influenced the rate of arsenic
transformation (Garbarino et al., 2003). In this study,
the full scale windrow composting was operated in the
open air and turned once a week during the
thermophilic stage. Pile turning increased the air flow
which could bring more oxygen through the pile and
its porosity. Jiang et al. reported that the pile
temperature risings were observed after turning and
these risings were caused by the degradation of
partially decomposed materials which were
transferred from the anaerobic areas to the aerobic
areas (Jiang et al., 2011).
In this study, the first turning happened on the
first day and the second turning happened on the 8th
day. Therefore, oxygen content might be deficiency
due to the activity of a great amount of aerobic
microorganisms and decomposition of organic matter
in the end of the first week, the pile might probably be
under anaerobic or anoxic condition. Additionally,
relatively high moisture content (approximately 50%)
in the beginning of composting might impede oxygen
internal diffusion, causing partial anaerobic condition.
The presence of microorganisms in the composting
pile combined to anaerobic condition might enhance
transformation of arsenic from As (V) to As (III).
These might well explain that the oxidation state
arsenic (arsenate) decreased to a low level and
reduction state arsenic (arsenite) increased up to a
relatively high level on the 7th day of composting (Fig.
3b). The mobility of five different arsenic fractions
changed in the order as following: exchangeable >
carbonates > Fe/Mn oxides > organic > residual. The
exchangeable fraction was the most toxic and highly
mobility arsenic form. However, the surprising result
(a)
Fig. 4. Changes of five arsenic fractions during the full scale swine manure windrow composting process:
(a) Percentage, (b) Concentration
of this study was that the exchangeable fraction of
arsenic kept in a really low level (0~84.9 µg·kg-1 DW)
during the composting (Fig. 4). The carbonates
fraction of arsenic varied from 18.6% to 45.3%, and
its average value was 30% in the composting process.
The mobility and bioavailability of Fe/Mn oxides
fraction of arsenic highly depended on Eh and pH
values, and changed dramatically during the swine
manure windrow composting process.
The organic and residual fractions of arsenic
were relatively low mobility arsenic forms which
accounted averagely for approximately 10% and 29%
respectively during the whole composting process.
Their concentrations at the end of composting were
not much difference (arsenate at 576.53 µg·kg-1DW
and arsenic bound to Fe/Mn oxides at 637.46 µg·kg-
1
DW). In general, arsenic presenting as arsenate under
oxic conditions is often adsorbed into organic matter,
iron and manganese oxides and oxyhydroxides
(Newton et al., 2006). According to this conclusion,
arsenic bound to Fe/Mn oxides to some extent ought
to be a part of arsenate during most of the time of
composting in this study. Furthermore, because
arsenate was the main species and arsenic bound to
Fe/Mn oxides was the main fraction, it seemed they
were linked. The low frequency of pile turning in the
thermophilic stage and a relatively low pH value
(around 5.0) at the start of composting might explain
both of arsenate and arsenic bound to Fe/Mn oxides
decreasing roughly in the first 10 days. As shown in
Fig. 5, it is interestingly found that variation of
arsenate had the similar trend to arsenic bound to
Fe/Mn oxides in the first 10 days of composting. In the
environment, As (V) and As (III) are the dominant
forms of inorganic arsenic, which As (III) is much
more toxic than As (V) (Jain and Ali, 2000). In this
study, the variation of the ratio of As (V) to As (III)
reflected their mutual transformation to some extent.
The point on the 7th day was extremely low and might
be due to low frequency pile turning in the first week
of composting.
(b)
1569
 Wang et al./Environmental Engineering and Management Journal 17 (2018), 7, 1563-1573
In this study, it seemed that the influence of the
pile turning on the ratio of As (V) to As (III) was
obvious. It was thrilling to notice that the ratio of As
(V) to As (III) kept increasing along with the
composting, which indicated that the proportion of the
more toxic As (III) decreased. In the thermophilic
stage, the pile was turned thoroughly once a week and
the composting was almost under aerobic condition.
This might be the main reason that the ratio of
As (V) to As (III) kept increasing in the first 38 days.
Nevertheless, there was no pile turning in the mature
stage and the composting was almost under partial
anaerobic condition, which might result in the ratio of
As (V) to As (III) dropping dramatically in the last
several days. Therefore, one pile turning should be
added in the mature stage to better control the
proportion of As (III) for improving detoxification on
arsenic. Moreover, some reports indicated that under
near-neutral conditions (pH=6.5) and different Eh
conditions (Eh=450, 325, 100 and 0mV), the major
inorganic arsenic species was always arsenate
(Carbonell-Barrachina et al., 2000).
There were some reports on the variation of the
ratio of As (V) to As (III) in other medium, like the
ocean, coastal waters and estuary (Neff, 1997),
however, little works have been done about the ratio
of As (V) to As (III) during the composting process.
For example, in open ocean seawater, the ratio of As
(V) to As (III) may exceed 300 (Li, 1991), and coastal
water was influenced by freshwater runoff and high
primary production, the ratio may be much lower
(Andreae and Andreae, 1989). Different from ocean,
coastal waters and estuary, high level activity of
microorganisms in the composting process might
contribute a lot to the variation of aerobic/anaerobic
condition, temperature, pH value and so on, therefore
Fig. 5. Correlation between arsenate and Fe/Mn oxides arsenic concentrations during the full scale
swine manure windrow composting process
1570
the actual true reason about the ratio of As (V) to As
(III) variation in the composting process needs further
studies. Compared with As (III) and As (V), the
concentrations of organic arsenics such as MMA and
DMA were relatively lower, in the range of 0~150
µg·kg-1 DW. Methylation theoretically will not occur
abiotically even if there is a supply of Carbon source,
such as protein and other organic compounds because
the C-C, C-N and C-O bonds are more stable than C-
As bonds (Williams, 1988). Biomethylation of metals
and metalloids occurs extensively in the environment
and methylated compounds show different properties
in mobility, bioavailability and toxicity in comparison
to their inorganic precursors (Sigel, 2010). Fig. 6
showed that the ratio of As (V) to As (III) kept
increasing in the first 38 days and then decreased in
the last several days of the composting.
It is interesting to point out that the ratio of As
(V) to As (III) in this study achieved maximum value
at pH of about 6.5, and stayed at a very low level at
pH of around 5.5 (Fig. 7). The process of
biomethylation was regarded as a detoxification
process for many years; however, more recent
research has indicated that the reverse is in fact the
case (Sigel, 2010). Fig. 3b showed that the tendency
of DMA was similar to that of arsenate, which both
declined in the beginning and then climbed smoothly
until the 40th day and decreased a little in the last days
of the composting. The level of MMA was always the
lowest, even lower than the limit of detection in the
full-scale swine manure windrow composting process.
The methylated arsenic accounted for 1% to
9% of total arsenic in the composting process.
Recently, Diaz-Bone et al. (2010) indicated that
metalloids, such as arsenic, could undergo intensive
biomethylation during composting.
 Changes of arsenic speciation during swine manure windrow composting at full scale

Fig. 6. The ratio of As (V) to As (III) variation with time during the full scale swine manure windrow composting process

Fig. 7. The relationship between pH value and the ratio of As (V) to As (III) during the full scale
swine manure windrow composting process

Extraordinary high concentrations of up to 150
mg·kg-1 methylated arsenic species as well as
conversion rates up to 50% were observed. In this
study, the percentages of methylated arsenic were not
so high because of a relatively low level of total
arsenic in the compost materials. The other reason
might be that DMA and MMA were only a small part
of methylated arsenic species and other forms of
methylated arsenic species were not detected in this
study. The total arsenic concentrations kept relatively
stable and showed a slightly increasing trend from
1.35 mg·kg-1 DW to 1.69 mg·kg-1 DW (Fig. 3b) due to
the decomposition of organic matter and the reduction
of bio-mass during the swine manure windrow
composting process. The primary arsenic species was
As (V), increasing from 29.6% of total arsenic on the
first day to 34.1% of total arsenic on the 53th day. A
very promising result was that the most toxic arsenic
species, arsenite, reduced a lot from 24.1% to 8.1% of
total arsenic (Fig. 3a). Additionally, organic arsenic

1571
 Wang et al./Environmental Engineering and Management Journal 17 (2018), 7, 1563-1573
(MMA and DMA) accounted for 4.7% at the end of
composting.
4. Conclusions
During 53 days composting, arsenite decreased
from 325.71 to 138.35 µg·kg-1 DW and arsenate
increased from 401.19 to 576.53 µg·kg-1 DW. The
percentages of arsenite, arsenate and DMA were
24.10%, 29.68%, 4.87% of total arsenic in the
beginning and 8.19%, 34.13%, 4.75% of total arsenic
respectively at the end of the composting. Arsenate
had the similar trend to Fe/Mn oxides arsenic fraction.
During this composting process, the most
toxic arsenic, arsenite, decreased about 16%. It was
proven that composting was a suitable method for
detoxification of arsenic by changing their forms.
Acknowledgements
This work is financially supported by Special Fund for
Agro-scientific Research in the Public Interest (No.
201303091), the National Natural Science Foundation of
China (No. 50578156 and No.21077122) and the National
Key Technology R&D Program (2012BAC25B00).
References
Al-Assaf K.H., Tyson J.F., Uden P.C., (2009),
Determination of four arsenic species in soil by
sequential extraction and high performance liquid
chromatography with post-column hydride generation
and inductively coupled plasma optical emission
spectrometry detection, Journal of Analytical Atomic
Spectrometry, 24, 376-384.
Andreae M., Andreae T., (1989), Dissolved arsenic species
in the Schelde estuary and watershed, Belgium,
Estuarine, Coastal and Shelf Science, 29, 421-433.
Bao S., (2000), Agricultural chemical analysis of soil, China
Agricultural, 56-58.
Byrd J.T., (1990), Comparative geochemistries of arsenic
and antimony in rivers and estuaries, Science of the
Total Environment, 97, 301-314.
Carbonell-Barrachina A., Jugsujinda A., Burlo F., Delaune
R., PatrickJr W., (2000), Arsenic chemistry in
municipal sewage sludge as affected by redox potential
and pH, Water Research, 34, 216-224.
Carbonell-Barrachina Á.A., Burló F., Valero D., López E.,
Martínez-Romero D., Martínez-Sánchez F., (1999),
Arsenic toxicity and accumulation in turnip as affected
by arsenic chemical speciation, Journal of Agricultural
and Food Chemistry, 47, 2288-2294.
Chapman H., Johnson Z., (2002), Use of antibiotics and
roxarsone in broiler chickens in the USA: analysis for
the years 1995 to 2000, Poultry Science, 81, 356-364.
de Bertoldi G.Vallini, A.Pera, (1983), The biology of
composting: a review, Waste Management & Research,
1, 157-176.
Diaz-Bone R.A., Raabe M., Keuter B., Menzel B., Kuppers
K., Widmann R., Hirner A.V., (2010), Investigation of
biomethylation of arsenic and tellurium during
composting, Journal of Hazardous Materials, 189, 653-
659.
Dong Z., Chen Yi-ding, Lin Xian-yong, Zhang Yong-song,
Dan-hua N., (2008), Investigation on the contents and
fractionation of heavy metals in swine manures from
1572
intensive livestock farms in the suburb of Hangzho,
Acta Agriculturae Zhejiangensis, 20, 35-39.
Finstein M., Miller F., Strom P., (1986), Waste treatment
composting as a controlled system, Biotechnology, 8,
396-443.
Garbarino J.R., Bednar A.J., Rutherford D.W., Beyer R.S.,
Wershaw R.L., (2003), Environmental fate of
roxarsone in poultry litter. I. Degradation of roxarsone
during composting, Environmental Science and
Technology, 37, 1509-1514.
Garbarino J.R., Kanagy L.K., Cree M.E, (2006),
Determination of elements in natural-water, biota,
sediment, and soil samples using collision/reaction cell
inductively coupled plasma-mass spectrometry, US
Department of the Interior, US Geological Survey.
Georgacakis D., Tsavdaris A., Bakouli J., Symeonidis S.,
(1996), Composting solid swine manure and lignite
mixtures with selected plant residues, Bioresource
Technology, 56, 195-200.
Hanc A., Szakova J., Svehla P., (2011), Effect of composting
on the mobility of arsenic, chromium and nickel
contained in kitchen and garden waste, Bioresource
Technology, 126, 444-452.
Hsu J.H., Lo S.L., (2001), Effect of composting on
characterization and leaching of copper, manganese,
and zinc from swine manure, Environmental Pollution,
114, 119-127.
Jain C., Ali I., (2000), Arsenic: occurrence, toxicity and
speciation techniques, Water Research, 34, 4304-4312.
Jiang T., Schuchardt F., Li G., Guo R., Zhao Y., (2011),
Effect of C/N ratio, aeration rate and moisture content
on ammonia and greenhouse gas emission during the
composting, Journal of Environmental Sciences, 23,
1754-1760.
Li Y.-H., (1991), Distribution patterns of the elements in the
ocean: A synthesis, Geochimica et Cosmochimica Acta,
55, 3223-3240.
Li Y., Chen T., (2005), Concentrations of additive arsenic in
Beijing pig feeds and the residues in pig manure,
Resources, Conservation and Recycling, 45, 356-367.
Long G., Yu-jun Z., Yuan-hua D., (2004), Heavy metals
pollution in poultry and livestock feeds and manures
under intensive farming in Jiangsu Province, China,
Journal of Environmental Sciences, 16, 371-374.
Luo L., Ma Y., Zhang S., Wei D., Zhu Y.G., (2009), An
inventory of trace element inputs to agricultural soils in
China, Journal of Environmental Management, 90,
2524-2530.
Masscheleyn P.H., Delaune R.D., Patrick Jr W.H., (1991),
Effect of redox potential and pH on arsenic speciation
and solubility in a contaminated soil, Environmental
Science & Technology, 25, 1414-1419.
Miyatake F., Iwabuchi K., (2006), Effect of compost
temperature on oxygen uptake rate, specific growth rate
and enzymatic activity of microorganisms in dairy
cattle manure, Bioresource Technology, 97, 961-965.
Neff J.M., (1997), Ecotoxicology of arsenic in the marine
environment, Environmental Toxicology and
Chemistry, 16, 917-927.
Newton K., Amarasiriwardena D., Xing B., (2006),
Distribution of soil arsenic species, lead and arsenic
bound to humic acid molar mass fractions in a
contaminated apple orchard, Environmental Pollution,
143, 197-205.
Pinel-Raffaitin P., Le Hecho I., Amouroux D., Potin-Gautier
M., (2007), Distribution and fate of inorganic and
organic arsenic species in landfill leachates and
biogases, Environmental Science & Technology, 41,
4536-4541.
 Changes of arsenic speciation during swine manure windrow composting at full scale
Poincelot R.P. (1975), The biochemistry and methodology
of composting, On line at: http://agris.fao.org/agris-
search/search.do?recordID=US201300543417.
Rutherford D., Bednar A., Garbarino J., Needham R., Staver
K., Wershaw R., (2003), Environmental fate of
roxarsone in poultry litter. Part II. Mobility of arsenic
in soils amended with poultry litter, Environmental
Science & Technology, 37, 1515-1520.
Sierra-Alvarez R., Cortinas I., Field J.A., (2010),
Methanogenic inhibition by roxarsone (4-hydroxy-3-
nitrophenylarsonic acid) and related aromatic arsenic
compounds, Journal of Hazardous Materials, 175, 352-
358.
Sigel A., (2010), Organometallics in environment and
toxicology, Royal Society of Chemistry, 86-87.
Ter Welle H., Slater E., (1967), Uncoupling of respiratory-
chain phosphorylation by arsenate, Biochimica et
biophysica acta, 143, 1-17.
Tessier A., Campbell P.G.C., Bisson M., (1979), Sequential
extraction procedure for the speciation of particulate
trace metals, Analytical Chemistry, 51, 844-851.
Williams R., (1988), The transfer of methyl groups: a
general introduction, Biological Alkylation of Heavy
Elements, 5-19.
Yan Z.Y., Li J., Liu X.F., Yuan Y.X., Liao Y.Z., Li X.D.,
(2017), Deodorization of swine manure using a
Lactobacillus strain, Environmental Engineering and
Management Journal, 16, 2191-2198.
Yao L., Huang L., He Z., Zhou C., Li G., (2013), Occurrence
of Arsenic Impurities in Organoarsenics and Animal
Feeds, Journal of Agricultural and Food Chemistry, 61,
320-324.
Zeng X., Li L., Bai L., Mei X., Yang J., Hu L., (2007),
Arsenic accumulation in different agricultural soils in
Shouguang of Shandong Province, The Journal of
Applied Ecology, 18, 310.
Zhang F., Li Y., Yang M., Li W., (2012), Content of Heavy
Metals in Animal Feeds and Manures from Farms of
Different Scales in Northeast China, International
Journal of Environmental Research and Public Health,
9, 2658-2668.
Zhang S., Zhang F., Liu X., Wang Y., Zou S., He X., (2005),
Determination and analysis on main harmful
composition in excrement of scale livestock and poultry
feedlots, Plant Nutrition and Fertilizer Science, 11,
822-829.
1573
1574
Environmental Engineering and Management Journal July 2018, Vol.17, No. 7, 1575-1581
http://www.eemj.icpm.tuiasi.ro/; http://www.eemj.eu

“Gheorghe Asachi” Technical University of Iasi, Romania

INTEGRATED USE OF TWO MICROALGAL SPECIES

FOR THE TREATMENT OF AQUACULTURE EFFLUENT
AND BIOMASS PRODUCTION

Katya Velichkova, Ivaylo Sirakov ∗, Yordan Staykov

Department of Biology and Aquaculture, Agricultural Faculty, Trakia University, 6000 Stara Zagora, Bulgaria

Abstract

One of the main challenges for the development of biomass algal system production is the high operational and capital costs for
these technologies. A great opportunity to overcome these challenges may however exist in the integration of wastewater treatment
with algal biomass production. The algae are used for treatment of industrial wastewater, for purification of effluent originating
from livestock production, while the studies concerning the use of algae species for the treatment of aquaculture effluent are very
few. The aim of our research was to compare the nitrogen and phosphate removal efficiency and the growth of two microalgae
species – Scenedesmus dimorphus and Botryococcus braunii, cultivated in wastewater, originating from freshwater aquaculture
production. A laboratory bioreactor was used for algae cultivation. It consisted of 500 mL Erlenmeyer flasks, containing wastewater
from semi closed recirculation aquaculture system. Light regime was adjusted at 15:9 h light:dark cycle, the air was enriched with
CO2 up to 1% and the water temperature was kept between 25 and 27oC. Samples for water chemical analysis and growth
measurement were taken at the beginning of the trial, at the 24th, 96th and the 168th hour after the start of the experiment. A better
removal efficiency of nitrogen compound from wastewater originate from aquaculture was determined for B. braunii and the
concentration of ammonia, nitrite, nitrate and total nitrogen decreased at the end of trial by 50%, 84.4%, 63.1% and 61.5%
respectively. A better phosphate removal efficiency (77.8%) and better growth rate were found for Sc. dimorphus when aquaculture
effluent was used as a growing media.

Key words: aquaculture, Botryococcus braunii, Scenedesmus dimorphus, wastewater treatment

Received: March, 2014; Revised final: February, 2014; Accepted: September, 2014; Published in final edited form: July 2018

1. Introduction Biofuels are among the major competitive alternatives

Protected areas have many benefits such as the
energy consumption of our planet is increasing
continuously, because of the rise in industrialization
and Earth’s population. Natural energy sources are
limited and cannot meet the high demand for power
consumption of our modern society. New energy
sources must be found and introduced to solve the so-
called “global energy crisis”. A possible solution of
this problem could be renewable energy sources,
whose importance has increased in these last years.
of fossil fuels.
Biodiesel is a biodegradable and non-toxic
alternative fuel obtained from renewable sources such
as waste cooking oil, palm, soy-bean, canola, rice
bran, sunflower, coconut, corn, fish oil, chicken fat
and algae (Pezzela et al., 2016; Sharif et al., 2007).
The microalgae are one of the most promising
resources, which could be used for biodiesel
production.One of the main advantages of algae’s
cultivation is connected with the amount of oils and
lipids per unit area. This difference can be as high as

∗
Author to whom all correspondence should be addressed: e-mail: ivailo_sir@abv.bg; Phone: +359896669829
 Velichkova et al./Environmental Engineering and Management Journal 17 (2018), 7, 1575-1581
30 times in comparison to oilseed crops which are
cultivated at land (Sheehan et al., 1998).
Botryococcus braunii is a green microalga that
produces hydrocarbons up to 75% from its dry
biomass and it has already been proven as a renewable
source of fuel (Brown et al., 1969; Banerjee et al.,
2002; Casadevall et al., 1983; Dayananda et al., 2005;
Velichkova et al., 2012). Many researches have
proven that Scenedesmus dimorphus is one of the most
promising algae species for biomass and biofuel
production as well (Becker, 1994; Goswami and
Kalita, 2011; Varsharani and Geeta, 2011; Velichkova
et al., 2013).
One of the main challenges for the
development of biomass algal system for biofuels
production is the high operational and capital costs for
these technologies (Alabi et al., 2009; Benemann,
2008). A great opportunity to overcome these
challenges may however exist in the integration of
wastewater treatment with algal biomass production.
The microalgae are used for treatment of
industrial wastewater: soya sauce effluent (Shirai et
al., 1998), brewery industries (Simate et al., 2011),
dairy wastewater (Danalewich et al., 1998; Woertz et
al., 2009), as well as for purification of effluent,
originating from livestock production – poultry
(Mahadevaswamy and Venkataraman, 1986;
Murugesan et al., 2010), pig farms (Aguirre et al.,
2011; Po Chung et al., 1978), cattle farms (Ayala and
Vargas, 1987; Lincoln et al., 1996). Until now few
studies concerning the exploitation of algae species for
the treatment of effluent in aquaculture have been
made (Borges et al., 2005; Dumas et al., 1998;
Hussenot et al., 1998; Laliberté et al., 1994; Lefebvre
et al., 1996). Nevertheless, these studies are ignored
(Wang, 2003) and the science publications concerning
them are highly limited. The aim of our research was
to compare the nitrogen and phosphate removal
efficiency as well as the growth of two microalgae
species - Scenedesmus dimorphus and Botryococcus
braunii cultivated in wastewater, originating from
freshwater aquaculture production to answer whether
it is possible to integrate the treatment of an
aquaculture effluent with algal biomass production.
2. Material and methods
2.1. Algae species and media
Two microalgae species were used in the study:
- Scenedesmus dimorphus (SKU: AC-1002)
- Botryococcus braunii (SKU: AC-1006)
Fig. 1. Semi-closed RAS and sampling point: 1. Fish farming tank; 2. Mechanical filter; 3. Biological filter; 4.
Pumping tank; 5. Inlet water; 6. Outlet water; 7. Pump; 8. Sampling point
1576
Both microalgae strains were supplied by
Algae depot (USA) (www.algaedepot.com). The
microalgae cultures were kept in glass flasks of 5 mL
volume with reseeding for their preservation in a good
physiological condition.
The wastewater originating from the semi
closed recirculation aquaculture system (semi – closed
RAS) was used as a growing medium for tested
microalgae. Effluent treatment in the semi – closed
RAS included a mechanical filter and a biofiltration
(nitrifying filter with bio balls) for water purification.
Daily, 40% of the water volume in semi - closed RAS
was replaced with fresh water. The wastewater, used
as medium for the tested microalgae was taken after
mechanical and biological filtration from the pumping
tank, before daily fresh water replacement (Fig. 1).
The total biomass of raised in the recirculation
aquaculture system common carps was 150 kg-1.
2.2. Description of the Bioreactor for microalgae
cultivation
A bioreactor consisting of 500 mL Erlenmeyer
flasks and filled with wastewater from the semi closed
RAS was used in the current trial. Three luminescent
lamps Sylvania Aqua Star – 18w, 10 000 K, were
placed at a distance of 30 mm from the flasks. Light
regime was adjusted at 15:9 h light:dark cycle in an
illumination incubator until the end of the experiment.
The air was enriched with CO2 up to 1%, regulated by
manometric control and injected to the bottom of the
flasks at a constant flow rate. The water temperature
was kept between 25 and 27oC, pH varied between 6.5
and 7.5 and for this reason was not adjusted.
Samples for water chemical analysis were taken at the
beginning of the trial and at the 24th, 96th and 168th
hour after the start of the experiment. They were
centrifuged at 300 rpm for 10 min, for precipitation
and removal of algal cells (Lee and Lee, 2002). Every
variant of the experiment was conducted in six
replications (Fig. 2). Other analyzed water chemical
parameters were measured spectrophotometrically
with spectrophotometer DR 2800 (Hach Lange). The
methods and range of tests conducted during the
experiment are shown in Table 1.
2.3. Growth measurement, chlorophyll and carotenoid
content of microalgae culture
Optical densities of microalgal cultures were
measured at 0, 24, 96 and 168 hours after the start of
the experiment, in three repetitions.
 Integrated use of two microalgal species for the treatment of aquaculture effluent and biomass production
Table 1. Methods and ranges of tests, used for monitoring
of hydrochemical parameters during experimental period
Determination Measuring
Parameters
method range (mg L-1)
Ammonia Indophenol blue 0.015 – 2
Nitrite –
Diazotization 0.015 – 0.6
nitrogen
Nitrate -
nitrogen
2.6 dimethylphenol 0.23 – 13.5
Total Koroleff digestion + braunii and Sc. dimorphus were statistically
5 - 40
nitrogen 2.6 dimethylphenol
Phosphorus
Phosphomolybdenum
0.05 – 1.5
(ortho +
blue
PO4-P
total) 0.15 – 4.5 PO4
The samples with a volume of 1 mL, were
appropriately diluted with deionized water and the
average absorbance at 550 nm was recorded with the
help of a spectrophotometer DR 2800 (Hach Lange).
The isolation of pigments from algae cells
included the following procedures: harvesting 2 mL of
microalgae cells by centrifugation at 10000 rpm, two
times for 3 min and discarding the supernatant,
suspending cells in 2 mL methanol/water 90:10 v/v
and Vortex mixing for 1 min, heating of the
suspension for half an hour, in a water bath at 60ºC,
cooling of the samples at room temperature,
centrifugating the suspension (10000 rpm for 3 min)
and discarding the supernatant with dissolved
pigments. The absorbance of the pigments extract
(665, 652 nm for chlorophyll content (a+b) and 470,
666 nm for carotenoids content) was recorded
spectrophotometrically. The chlorophyll content (mg
L-1) was computed according to Porra et al., (1989)
method and carotenoids - according to Lichtenthaler
(1987) method. Data analysеs were conducted by
using ANOVA (MS Office, 2010).
3. Results
3.1. Efficiency of nitrogen compounds and phosphate
removal in tested microalgae using aquaculture
effluent as a medium for their cultivation
The values of this parameter in the experiments
with B. braunii and Sc. dimorphus were similar at the
end of the trial and the differences were not
statistically significant (P≥0.05).
The measured nitrogen compounds (ammonia,
nitrite, nitrate and total nitrogen) in the wastewater
used for cultivation of both experimental microalgal
species decreased significantly during the experiment
(Table 2). The concentration of ammonia in the trial
with B. braunii decreased from 0.302 to 0.158 mg L-1
and for Sc. dimorphus from 0.301 to 0.20 mg L-1.
These results corresponded to 50% and 33.5%
reduction of ammonia respectively, but differences
were not statistically significant (P>0.05) (Table 2). B.
braunii reduced the concentration of nitrite and nitrate
in wastewater, originating from aquaculture
production by 84.4% and 63.1% respectively and for
Sc. dimorphus the according percentages were 75%
and 44.4%. The pH of wastewater increased in both
tested variants during the trial (Fig. 2).
The differences in nitrite values between the
tested variants were statistically significant for the
48th and 168th hour samples (P≤0.05). For the last
measurement differences were highly significant
(P≤0.01) (Table 2). The differences in the quantity of
nitrate in wastewater used for the cultivation of B.
significant for the 96th and 168th hour samples
(P≤0.05) (Table 2). The total nitrogen decreased in the
wastewater containing B. braunii and Sc. dimorphus
from the beginning until the end of the experiment by
61.5% and 54.4% respectively and the differences
were statistically significant at hours 96 and 168
(P≤0.05) (Table 2). The quantities of removed
phosphates from aquaculture wastewater were similar
for both tested strains (Table 2).
The results varied from 1.4 mgl-1 at the start to
0.38 mg L-1 at the end of experiment for Sc. dimorphus
and from 1.3 mg L-1 to 0.41 mg L-1 for B. braunii,
corresponding to 77.8% and 68.4% phosphate
removal efficiency respectively. The differences
concerning phosphate in the wastewater, used as a test
medium for both explored microalgae species, were
statistically significant for none of the taken
wastewater samples (P>0.05) (Table 2).
3.2. Growth parameters of tested microalgae, using
aquaculture effluent as a medium for their cultivation
The optical densities were higher at the 4th day
since the beginning of experiment (96 hour) for both
microalgae species cultivated in the wastewater and
reached 0.95 for Sc. dimorphus and 0.8 for B. braunii
(Fig. 3). After this day of maximum optical density,
the values of this parameter declinedin both tested
microalgae strains, as the nutritional compound in
wastewater decreased heavily.
The measured optical density in Sc. dimorphus
strains was higher by 10.5%, compared to the value of
B. braunii at the end of trial, but the differences were
not statistically significant (P≥0.05). At the end of the
experiment the results showed that the contents of
pigments - chlorophyll a+b and carotenoids were
higher by 8.6% and 5.15% respectively in Sc.
dimorphus, compared to their quantity in B. braunii,
when they are cultivated in wastewater originating
from freshwater aquaculture. However, the
differences were not statistically significant for any of
the measurements made concerning these growing
parameters (P≥0.05) (Fig. 4).
4. Discussions
Many authors stated that algal species
usually increase pH, due to the photosynthetic CO2
assimilation (Borowitzka, 1998; Chevalier et al.,
2000; Larsdotter, 2006). These points are confirmed in
recent research as well.
1577
 Velichkova et al./Environmental Engineering and Management Journal 17 (2018), 7, 1575-1581

Fig. 2. Bioreactor used in conducted trials:

1) Luminiscent lamps; 2). Erlenmeyer flasks; 3). CO2 bottle; 4). Manometer;

Table 2. Changes in nitrogen compounds of Sc. dimorphus and B. braunii in wastewater from aquaculture production during the
treatment period (mg L-1)

Hydrochemical parameters (mean ± SEM)

Microalgae
Hours
species Nitrite – Nitrate -
Ammonia Total nitrogen Phosphate
nitrogen nitrogen
Sc. dimorphus 0.301±0.01ns 0.08±0.02ns 1.8±0.2ns 18.7±0.2ns 1.4±0.02ns
0
B. braunii 0.302±0.02ns 0.09±0.01ns 1.9±0.1ns 17.7±0.9ns 1.3±0.01ns
Sc. dimorphus 0.241±0.01ns 0.0241±0.02ns 1.75±0.07ns 12.25±0.35ns 1.36±0.11ns
24
B. braunii 0.228±0.08ns 0.0218±0.007ns 1.6±0.1ns 11.75±0.35ns 1.17±0.28ns
Sc. dimorphus 0.202±0.09ns 0.0232±0.007ns 1.3±0.14* 11.1±0.28* 0.75±0.71ns
96
B. braunii 0.172±0.1ns 0.0162±0.007ns 0.95±0.07* 9.3±0.74* 0.95±0.03ns
Sc. dimorphus 0.20±0.05ns 0.0226±0.007* 1.00±0.014* 8.5±0.98* 0.31±0.75ns
168
B. braunii 0.158±0.12ns 0.0147±0.02* 0.7±0.04* 6.8±0.4* 0.41±0.02ns
Note: *-show statistically significant differences (p<0.05), ns=no significant difference

Fig. 3. Optical density (OD) of Sc. dimorphus and B. braunii, cultivated in wastewater from aquaculture production

One of the possible consequences of high pH
in media is the induction of flocculation in algal
cultures, which could be a reason for reduction of
nutrient uptake and growth.
If the pH values increase over 9, most of
inorganic carbon is converted to carbonate, which
cannot be assimilated by the algae.
In our opinion pH did not suppress microalgae
growth in the current trial, because pH values did not
reach very high levels and varied between 6.5 and
7.43. Because of the low values of ammonia in
wastewater originating from aquaculture, it could be
concluded that this parameter had no negative effect
on the microalgae growth.

1578
 Integrated use of two microalgal species for the treatment of aquaculture effluent and biomass production
Fig. 4. Chlorophyll (a+b) and carotenoids (mg L-1) concentrations in Sc. dimorphus and B. braunii,
cultivated in wastewater from aquaculture production
Concentrations higher than 20 mg NH4+N per
liter in wastewater are not recommended due to
ammonia toxicity (Borowitzka, 1998). The preferred
nitrogen compound for microalgae growth is ammonia
and when it is available, no alternative nitrogen
sources will be used (Bhaya et al., 2000). The low
ammonia concentration in wastewater used as a
medium for both tested microalgae species could
explain the advanced utilization of two other nitrogen
sources – nitrite and nitrate. The higher concentration
of nitrite makes it a less convenient nitrogen source for
algae, because of its toxicity (Becker, 1988). During
the experiment the values of nitrite were not so high,
which probably excluded the possibility of nitrite
toxicity for tested microalgae species.
The efficiency in nitrogen removal
(ammonium, nitrite, nitrate and total nitrogen) from
aquaculture effluent in B. braunii was higher
compared to the Sc. dimorphus strain in all
measurements. The results for nitrogen and
phosphorus removal by B. braunii in the current
research were found to be better than thosereported for
swine wastewater treatment, by using tubular
bioreactor (Lee et al., 1999). In their research removal
rates of swine wastewater nitrogen and phosphorus
byB. braunii were 43.9% and 41.7%, respectively,
after 14 days of incubation. The results for chlorophyll
density of B. braunii, cultivated in wastewater as a
growing medium from the current research were
similar with those of other authors (Tsukahara and
Sawayama, 2005) stating that B. braunii could
effectively remove nitrogen and phosphorus from
secondary sewage treatement (SST) in the continuous
bioreactor system.
The main factors affecting B. braunii’s growth
are availability of nitrogen and phosphate, light
intensity and pH (Qin, 2005). In the current study,
measured growth parameters indicated a better growth
of Sc. dimorphus, compared to B. braunii, when
wastewater from aquaculture was used as a growing
medium. Among the main disadvantages of B. braunii
is its slow growth: its doubling time is 7 hours (Qin,
2005). According to Banerjee et al., (2005) the growth
of B. braunii is 20 times slower, compared to fast-
growing algae so they recommended for its cultivation
a low-cost investment system, like raceway pond, as
economically effective. Our results for the hemical
parameters of effluent waters were very similar to
those of previous studies (Borges et al., 2005;
Hussenot et al., 1998; Tovar et al., 2000). Namely the
wastewaters from fish production are distinguished
mainly by dissolved inorganic nutrients like nitrogen
and phosphorus. This makes aquaculture effluent one
possible medium for low cost microalgae biomass
production (Laliberté et al., 1994). Ammonia removal
efficiency of Sc. dimorphus in the current research was
lower compared to other reports (Gonzalez et al.,
1997), but higher than the results of Chevalier and De
la Noüe (1985), which reported a removal efficiency
of 23.31%. Ammonia, total nitrogen and phosphate
removal efficiency were higher in other algal species
from genus Scenedesmus – Sc. quadricauda used for
swine wastewater treatment after appropriate dilution
with tap water (Gantar et al., 1991).
At the 8th day, the percent of ammonia, total
nitrogen and phosphate uptake varied between 99.5 -
100, 97.9% - 99.4% and 74.3 - 99.9% respectively at
different tested dilutions. The water samples from
locations around the Organized Industrial Zone were
found to stimulate the growth of green microalgae
Desmodesmus (=Scenedesmus) subspicatus (Katalay
et al., 2012). Charity et al., (2009) evaluated the
growth, removal of nutrients and organic matter of
microalgae Scenedesmus sp. in wastewater derived
from remains of fish for 15 days.
They established 2.24 mg/L chlorophyll and
0.89 mg/L carotenoids, and dry biomass 0.44 mg/mL.
In our results a high growth rate of the microalgae
culture and high chlorophyll and carotenoid contents
were also observed. Volotina et al. (2004) also
determined a high biomass after 24 hours cultivation
in artificial wastewater.
1579
 Velichkova et al./Environmental Engineering and Management Journal 17 (2018), 7, 1575-1581
The two tested microalgae are a good option
for the biological treatment of aquaculture effluent.
Both species showed a good nitrogen and phosphorus
removal efficiency and the current research provides a
good alternative technology for B. braunii and Sc.
dimorphus production - integration of effluent
treatment from aquaculture and biomass production,
which could decrease operational costs and will make
the technology for tested microalgae production
highly sustainable.
5. Conclusions
This research provided a proof of concept for
an integrated technology of aquaculture effluent
treatment process that combines nutrient removal and
microalgae production for potential use as a biofuel
feedstock.
A better removal efficiency of nitrogen
compound (ammonia, nitrite, nitrate and total
nitrogen) was determined for B. braunii and better
phosphate removal efficiency as well as better growth
rate were found for Sc. dimorphus when wastewater
from aquaculture effluent was used as a growing
medium.
Acknowledgements
This study was supported by the Faculty of Agriculture,
Trakia University (project 3E/2012). We also would like to
thank Mrs. Maia Andonova and Mr. Alexander Iordanov for
their great help inchecking and editing our manuscript.
References
Aguirre P., Álvarez E., Ferrer I., García J., (2011),
Treatment of piggery wastewater in experimental high
ratealgal ponds, Revista Latino-americana de
Biotecnologia Ambiental y Algal, 2, 57-66.
Alabi A.O., Tampier M., Bibeau E., (2009), Microalgae
Technologies and Processes for Biofuels/Bioenergy
Production in British-Columbia: Current Technology,
Suitability and Barriers to Implementation, Seed
Science Ltd, British Columbia Innovation Council,
Canada.
Ayala F., Vargas T., (1987), Experiments on Spirulina
culture on waste-effluent media and at the pilot plant,
Hydrobiologia, 151, 91-93.
Banerjee A., Sharma R., Chisti Y., Banerjee U. C., (2002),
Botryococcus braunii: A renewable source of
hydrocarbons and other chemicals, Critical Reviews in
Biotechnology, 22, 245-279.
Becker E.W., (1994), Microalgae: biotechnology and
microbiology, Cambridge University Press,
Cambridge, UK.
Becker E.W., (1988), Micro-algae for Human and Animal
Consumption, In: Micro-Algal Biotechnology,
Borowitzka M.A., Borowitzka L.J. (Eds.), Cambridge
University Press, Cambridge, 222-256.
Benemann J.R., (2008), Opportunities and challenges in
algae biofuels production, Algae World 2008, On line
at:
http://www.futureenergyevents.com/algae/whitepaper/
algae_positionpaper.pdf.
Bhaya D., Schwarz R., Grossman R., (2000), Molecular
Responses to Environmental Stress, In: The Ecology of
1580
Cyanobacteria - Their Diversity in Time and Space,
Whitton B.A., Potts M. (Eds.), Kluwer, Dordrecht, 397-
442.
Borges M.T., Silva P., Moreira L., Soares R., (2005),
Integration of consumer-targeted microalgal production
with marine fish effluent biofiltration–a strategy for
mariculture sustainability, Journal of Applied
Phycology, 17, 187-197.
Borowitzka M.A., (1998), Limits to Growth in Wastewater
Treatment with Algae, In: Inland Saline Aquaculture
Workshop, Wong Y.S., Tam N.F.Y. (Eds.), Springer
Verlag., 203-226.
Brown A.C., Knights B.A., Conway E., (1969),
Hydrocarbon content and its relationship to
physiological state in the green alga Botryococcus
braunii, Phytochemistry, 8, 543-547.
Casadevall E., Largeau C., Metzger P., Chirac C., Berkaloff
C., (1983), Hydrocarbon production by unicellular
microalga Botryococcus braunii, Biosciences, 2, 129.
Charity E., Andrade R., Alexandra L., Vera B., Carmen H.,
(2009), Biomass production of microalga Scenedesmus
sp. with wastewater from fishery, Revista Téchnica
Zulia, 32, 126-134.
Chevalier P., De la Noüe J., (1985), Wastewater nutrient
removal with microalgae immobilized in carrageenan,
Enzyme and Microbial Technology, 7, 621-624.
Chevalier P., Proulx D., Lessard P., Vincent W. F., De la
Noüe J., (2000), Nitrogen and phosphorus removal by
high latitude mat-forming cyanobacteria for potential
use in tertiary wastewater treatment, Journal of Applied
Phycology, 12, 105-112.
Danalewich J.R., Papagiannis T. G., Belyea R. L.,
Tumbleson M. E., Raskin L., (1998), Characterization
of dairy waste streams, current treatment practices, and
potential for biological nutrient removal, Water
Research, 32, 3555-3568.
Dayananda C., Sarada R., Bhattacharya S., Ravishankar G.,
(2005), Effect of media and culture conditions on
growth and hydrocarbon production by Botryococcus
braunii, Process Biochemistry, 40, 3125-3131.
Dumas A., Laliberté G., Lessard P., De la Noüe J., (1998),
Biotreatment of fish farm effluents using the
cyanobacterium Phomidium bohneri, Aquacultural
Engineering, 17, 57-68.
Gantar M., Obreht Z., Dalmacija B., (1991), Nutrient
removal and algal succession during the growth of
Spirulina platensis and Secenedesmus quadricauda on
swine wastewater, Bioresource Technology, 36, 167-
171.
Gonzalez L. E., Cañizares R. O., Baena S., (1997),
Efficiency of ammonia and phosphorus removal from
Colombian agro industrial wastewater by the
microalgae Chlorella vulgaris and Scenedesmus
dimorphus, Bioresourse Technology, 60, 259-262.
Goswami R., Kalita M., (2011), Scenedesmus dimorphus
and Scenedesmus quadricauda: two potent indigenous
microalgae strains for biomass production and CO2
mitigation – A study on their growth behavior and lipid
productivity under different concentration of urea as
nitrogen source, Journal of Algal Biomass Utilization,
2, 42-49.
Hussenot J., Lefebvre S., Brossard N., (1998), Open-air
treatment of wastewater from land- based marine fish
farms in extensive and intensive systems: Current
technology and future perspectives, Aquatic Living
Resources, 11, 297-304.
Katalay S., Boyacioglu M., Cakalarslan O., Parlak H.,
Karaaslan M., (2012), Phytotoxicity of water and
sediment from Nif Brook (Izmir, Turkey) on green
 Integrated use of two microalgal species for the treatment of aquaculture effluent and biomass production
algae Desmodesmus (= Scenedesmus) subspicatus,
Ekoloji, 21, 25-31.
Laliberté G., Proulx D., De Pauw N., De la Noüe J., (1994),
Algal technology in wastewater treatment, Archiv fur
Hydrobiologie- Beiheft Ergebnisse der Limnologie, 42,
283-302.
Larsdotter K., (2006), Wastewater treatment with
microalgae - a literature review, Vatten, 62, 31-38.
Lee S.J., Kim S.B., Kim H.S., Kwon G.S., Yoon B.D.,
(1999), Mass cultivation of Botryococcus braunii for
advanced treatment of swine waste water and lipid
production in photobioreactor, Sanop Misaengmul
Hakhoechi, 27, 166.
Lee K., Lee C.G., (2002), Nitrogen removal from
wastewaters by microalgae without consuming organic
carbon sources, Journal of Microbiology and
Biotechnology, 12, 979-985.
Lefebvre S., Hussenot J., Brossard N., (1996), Water
treatment of land-based fish farm effluents by outdoor
culture of marine diatoms, Journal of Applied
Phycology, 8, 193-200.
Lichtenthaler H. K., (1987), Chlorophylls and carotenoids:
Pigments of photosynthetic biomembranes, Methods in
Enzymology, 148, 350-382.
Lincoln E.P., Wilkie A.C., French B.T., (1996),
Cyanobacterial process for renovating dairy
wastewater, Biomass Bioenergy, 10, 63-68.
Mahadevaswamy M., Venkataraman L.V., (1986),
Bioconversion of poultry droppings for biogas and algal
production, Agricultural Waste, 18, 93-101.
Murugesan S., Venkatesh P., Dhamotharan R., (2010),
Phycoremediation of poultry wastewater by microalga,
Bioscience Biotechnology Research Communications,
3, 142-147.
Pezzella C., Vastano M., Casillo A., Corsaro M.M., Sannia
G., (2016), Production of bioplastic from waste oils by
recombinant Escherichia coli: a pit-stop in waste frying
oil to bio-diesel conversion race, Environmental
Engineering and Management Journal, 15, 2003-2010.
Po Chung W.G., Pond J.M., Kingsbury E.F., Walker J.,
Krook L., (1978), Production and nutritive value of
Arthrospira platensis, a spiral blue-green alga grown on
swine wastes, Journal of Animal Science, 47, 319-330.
Porra R.J., Thomson W.A., Kriedemann P.E., (1989),
Determination of accurate extinction coefficients and
simultaneous equations for assaying chlorophylls a and
b extracted with four different solvents: verification of
the concentration of chlorophyll standards by atomic
absorption spectroscopy, Biochimica et Biophysica
Acta, 975, 384-394.
Qin J., (2005), Bio-hydrocarbons from algae: impacts of
temperature, light and salinity on algae growth, Rural
Industries Research and Development Corporation, On
line at: http://www.rirdc.gov.au/reports/EFM/05-
025.pdf.
Sharif A., Nasrulhaq A., Majid H., Chandran S., Zuliana R.,
(2007), Biodiesel Production from Waste Cooking Oil
as Environmental Benefits and Recycling Process: A
Review, Asia Biofuel Conference Book, Singapore, 28.
Sheehan J., Dunahay T., Benemann J., Roessler P., (1998),
A look back at the US Department of energy’s aquatic
species program-biodiesel from algae, National
Renewable Energy Laboratory (NREL) Report:
NREL/TP-580-24190, Golden, Colorado, On line at:
https://www.nrel.gov/docs/legosti/fy98/24190.pdf.
Shirai F., Kunii K., Sato C., Teramoto Y., Mizuki E., (1998),
Cultivation of microalgae in the solution from the
desalting process of soy sauce waste treatment and
utilization of the algal biomass for ethanol
fermentation, World Journal of Microbiology and
Biotechnology, 14, 839-842.
Simate G. J., Cluett J., Iyuke S., Musapatika E. T., Ndlovu
S., (2011), The treatment of brewery wastewater for
reuse: state of the art, Desalination, 273, 235-247.
Tovar A., Moreno C., Manuel-Vez M. P., Garcia-Vargas M.,
(2000), Environmental impacts of intensive aquaculture
in marine waters, Water Research, 34, 334-342.
Tsukahara K., Sawayama S., (2005), Liquid fuel production
using microalgae, Journal of the Japan Petroleum
Institute, 48, 251-259.
Varsharani H., Geeta G., (2011), Isolation of microalgae
with biodiesel productivity prospects, Karnataka
Journal of Agricultural Sciences, 24, 585-588.
Velichkova K., Sirakov I., Georgiev G., (2012), Cultivation
of Botryococcus braunii strain in relation of its use for
biodiesel production, Journal of Bio Science and
Biotechnology, 157-162.
Velichkova K., Sirakov I., Georgiev G., (2013), Cultivation
of Scenedesmus dimorphus strain for biofuel
production, Agricultural Science and Technology, 5,
181-185.
Wang J.K., (2003), Conceptual design of a microalgae-
based reticulating oyster and shrimp system,
Aquacultural Engineering, 28, 37-46.
Woertz I., Feffer A., Lundquist T., Nelson Y., (2009), Algae
grown on dairy and municipal wastewater for
simultaneous nutrient removal and lipid production for
biofuel feedstock, Journal of Environmental
Engineering, 135, 1115-1122.
1581
1582
Environmental Engineering and Management Journal July 2018, Vol.17, No. 7, 1583-1590
http://www.eemj.icpm.tuiasi.ro/; http://www.eemj.eu

“Gheorghe Asachi” Technical University of Iasi, Romania

ENVIRONMENTAL PERFORMANCE OF SAJOR-CAJU MUSHROOM

PRODUCTION BASED ON FARM SIZES IN THAILAND

Siriprapa Ueawiwatsakul1, Thumrongrut Mungcharoen2, Rungnapa Tongpool3∗

1TAIST-Tokyo Tech, Faculty of Engineering, Kesetsart University, Bangkok 10900, Thailand
2Faculty
of Engineering, Kasetsart University, Bangkok 10900, Thailand
3LCA Lab, National Metal and Materials Technology Center, Pathumthani 12120, Thailand

Abstract

Unlike other vegetables, mushroom production requires several substrate ingredients, energy to sterilize the substrate, water to
humidify the fruiting house and waste management of the spent substrate and used materials. This work studied the environmental
impacts and eco-efficiency of sajor-caju mushroom (Pleurotus sajor-caju (Fr.) Sing.) production in terms of climate change
potential, acidification potential, water depletion potential and fossil fuel depletion potential, using life cycle assessment (LCA)
method. The results showed that the mushroom production in small and large farms had nearly the same environmental
performance. Their environmental impacts were lower and their eco-efficiencies were higher than those for medium-sized farms.
This means the medium-sized farms would cause more environmental impacts in order to obtain the same profit as the other two
farms. It was found that there was inefficient use of both substrate and energy in the medium-sized farms. This is probably because
the medium-sized farm had relatively low financial limitations, compared to the small farms, and relatively less concern over
efficient use of resources, compared to the large farms. The reduction of sawdust and rice bran, used as substrate ingredients, as
well as wood, used as energy source for sterilization, to the same amounts as those used in the small farms could reduce
environmental impacts (5-25%) and improve eco-efficiencies (10-40%) of the medium-sized farms to be close to those for large
and small farms.

Key words: climate change, environment, farm size, life cycle assessment, mushroom

Received: May, 2014; Revised final: August, 2014; Accepted: September, 2014; Published in final edited form: July 2018

1. Introduction
Sajor-caju mushroom (Sajor-caju (Fr.) Sing.)
is in the oyster mushroom family. It is a common
mushroom with the second largest production volume
in Thailand (DOAE, 2012a). In 2012, the total
production was 5,679 ton (DOAE, 2012b). Unlike
other vegetables, mushroom production requires
preparation and sterilization of substrate which
comprises several ingredients. After mixing and
putting substrate ingredients in a plastic bag, the
substrate is sterilized by steam. Then spawn is put
inside the substrate bag and incubated in a humidified
house. After harvest, the spent plastic bags and
substrate are burnt or left unmanaged. It can be seen
∗
Author to whom all correspondence should be addressed: e-mail: rungnapatongpool@gmail.com; Phone: +66 25646500; Fax: +66 25646404
that energy and water are required and air emissions
are released from these processes. Emissions of CO2,
CH4 and N2O contribute to climate change potential
while SO2 and NOx can cause terrestrial acidification.
The production of substrate ingredients, energy and
other raw materials used in the farms also consumes
fossil and water resources as well as releases
emissions to air. Although these environmental
burdens take place outside the farms, they arise from
the sajor-caju production and thus their environmental
impacts should be included.
Life Cycle Assessment (LCA) methodology is
an efficient tool to evaluate environmental impacts
quantitatively. It considers the use of resources and the
emissions from a product’s life cycle, i.e., from the
 Ueawiwatsakul et al./Environmental Engineering and Management Journal 17 (2018), 7, 1583-1590
stages of raw material provision, production, use, to
disposal and recycling. Data on resources used and
emission substances under the scope of the study are
collected and then classified according to their
environmental impact potential. For example, CO2,
CH4 and N2O are classified as climate change
substances. Each substance is given a characterization
factor according to its impact potential. For example,
characterization factors of 1 kg CO2, 1 kg CH4 and 1
kg N2O are 1, 25 and 298 kg CO2 eq, respectively
(BSI, 2008). Then the characterization factor of each
substance is multiplied by the amount of the
substance. The sum of the multiplication is the impact
of the concerned category. The methodology of LCA
has been explained in detail elsewhere (Dong et al.,
2016; ISO, 2006a, 2006b; Finnveden et al., 2009;
Tongpool et al., 2010). Using LCA, Gunady et al.
(2012) studied greenhouse gas emissions from
production of button Agaricus bisporus mushroom in
Western Australia. They found that life cycle
greenhouse gas emissions of the button mushroom
was 2.8 kg CO2eq/kg and that the main impact
contributor was the transport of raw materials such as
peat, compost and spawn.
While the farm activities affect the
environment, the environment influences farm
productivity and economy. Kanellopoulos et al.
(2014) illustrated that future farming systems were
challenged to adapt to changes in climate, market and
environment in order to remain competitive and to
meet increasing demand for food. To remain viable,
smaller farms were said to require technological
development to increase yields and compensate for
substantial increase in price of inputs. Technology and
resource management in different sizes of farms were
shown to have impacts on performance and economic
viability of the farms. Pishgar-Komleh et al. (2012)
reported that, in Esfahan province of Iran, the energy
input for potato production of large farms (larger than
5 ha) was relatively low while the total energy output
was relatively high, compared to small (up to 1 ha) and
medium-sized (1-5 ha) farms. They also found that the
amounts of seeds and chemical fertilizers applied in
Fig. 1. Scope of data collection under the study
1584
the large farms were relatively low due to better
management. However, Manjunatha et al. (2013)
found that, in the southern part of India, small farms
(up to 1.01 ha) showed better resource use efficiency
than large farms (> 1.01 ha). In Slovenia, small farms
(up to 5 ha) were relatively more profitable and the
large farms (> 20 ha) were more technically efficient
(Bojnec and Latruffe, 2013). It was reported that the
number of medium-sized farms (5-20 ha) in Slovania
has been decreasing because they were too small to be
economically efficient but they were too large to be
profitable.
Environment, economic and social factors are
concerned in sustainable development. The indicator
measuring progress toward economic and
environmental sustainability is eco-efficiency. It was
defined as the value of the product or service, divided
by environmental influence (Blengini et al., 2017;
Verfaillie and Bidwell, 2000). The environmental
impact, economic benefit and eco-efficiency of
different sizes of mushroom farms have not been
reported elsewhere.
Therefore this work compared the
environmental performance of sajor-caju production
of small, medium and large farms. The concerned
environmental impact categories were climate change,
terrestrial acidification, water depletion and fossil fuel
depletion. The key impact contributors were also
investigated and measures to reduce environmental
impacts as well as increasing eco-efficiency proposed.
2. Case studies presentation
2.1. Environmental impact assessment
As shown in Fig. 1, the production processes at
sajor-caju farms include mixing and sterilizing
substrates, inoculating with spawn, maintaining
environmental controls for spawn run and fruiting, as
well as harvesting and managing wastes. Wastes are
burnt in the open air or used as fuel for steam
generation or left unmanaged or sent to landfill,
depending on individual farm practice.
 Environmental performance of sajor-caju mushroom production based on farm sizes in Thailand
The data of inputs (such as substrate
ingredients, plastic bags, and fuel) and outputs (such
as air emissions, wastes) of each process were
collected by interviewing the farmers in Ratchaburi
province, Thailand. Sizes of mushroom farms were
divided into small, medium and large farms, having an
annual production of less than 20-ton, 20-40 ton and
more than 40 ton, respectively (DOA, 2012). The
obtained data came from fifteen large farms, eight
medium farms and eight small farms that together
contributed 22 % of total sajor-caju production of the
country in 2012 (DOAE, 2012b). Then the inputs and
outputs (inventory) for each farm size were weight
averaged according to the annual production to obtain
the averaged inventory of 1 kg sajor-caju production.
Amounts of air emissions from urea fertilizer and
combustion of wood, LPG, gasoline and burning of
wastes were calculated using emission factors of IPCC
2006 (Eggleston et al., 2006) and EMEP/EEA 2013
(EEA, 2013). Carbon dioxide from wood combustion
was not counted due to carbon neutral rule. The inputs
applied to sajor-caju production processes are
produced outside the farms and then transported to the
farms. Therefore the data of resources consumption
and emissions due to transportation and production of
the inputs were included in the environmental impact
assessment. These data were obtained from the Thai
National LCI database (MTEC, 2013) and the
Ecoinvent database (Ecoinvent Centre, 2007).
Then the total resources and emissions were
classified and multiplied by characterization factors
according to their environmental impact potential,
using the methodology of ReCiPe midpoint (H)
(Goedkoop et al., 2009) in order to quantify impacts in
the categories of climate change (unit: kg CO2
equivalent (eq)), terrestrial acidification (unit: kg SO2
eq), water depletion (unit: m3) and fossil fuel depletion
(unit: kg oil eq). SimaPro 7.3.3 software (Pre
Consultants, 2013) was used to facilitate the
calculation and to reduce any possible mistakes.
2.2. Eco-efficiency
The cost of raw materials, energy,
transportation and labor used for the production of
sajor-caju in each farm was obtained by interviewing
the farmers and suppliers. The sale price at sajor-caju
farm was 30 Baht/kg (about 1 US$/kg) in 2013. The
profit at the farm was equal to the difference between
the sale price and the total cost. Eco-efficiency of
sajor-caju production was derived from the profit of 1
kg sajor-caju divided by the environmental impact of
1 kg sajor-caju production.
3. Results and discussion
3.1. Environmental impact assessment
The inventory for production of 1 kg sajor-caju
is shown in Table 1. The medium-sized farms used
higher amount of substrate while the large farms used
more electricity than the others. It can be seen that, due
to the use of machinery, the amount of electricity used
in the large farms was about two and eleven times of
that used in the medium and small farms, respectively.
In the sterilization process, the medium and small
farms used only wood as an energy source. The large
farms used both wood and LPG. It was reported that
energy content of wood and LPG was 15.99 and 49.3
MJ/kg, respectively (DEDE, 2011). Therefore the
energy applied for the sterilization process in the
medium-sized farms was the highest (94.5 MJ),
compared to the small (71.3 MJ) and large (76.6 MJ)
farms. Water consumption for steam generation in all
farms was nearly the same but the water applied to
humidify the fruiting houses of the large farms was
only half of that used in the small and medium farms.
This is because every large farm used sprinkler
system, providing a better control of water usage,
which is not the case for small and medium farms.
The environmental impacts of 1 kg sajor-caju
are shown in Fig. 2. The impacts from production,
transportation and mixing of the substrates were
denoted as “substrate”. The impacts from the
production and transportation of fuel and water used
for sterilization, as well as air emissions from fuel
combustion during sterilization were denoted as
“sterilizing”. The environmental burdens from
inoculating spawn and environmental controls for
spawn run and fruiting were shown as “inoculating”
and “fruiting”, respectively. The contribution of
“harvest” came from the production and transportation
of plastic bags used during harvesting. The
contribution of “waste” came from management of
wastes after the harvest.
It is shown that the sajor-caju production of the
medium-sized farms caused the highest environmental
impacts. The main impact contributors in all
categories were “sterilizing” and “substrate”. The
climate change potentials of the large, medium and
small farms were 1.1, 1.4 and 1.2 kg CO2 eq/kg sajor-
caju, respectively. This was much lower than the level
obtained for button mushroom (Agaricus bisporus)
production in Australia (2.34 kg CO2eq/kg) reported
by Gunady, et al. (2012). In the category of fossil fuel
depletion, about 0.2 kg oil eq was consumed by the
production of 1 kg sajor-caju which was one third of
shiitake production (0.6 kg oil eq/kg) (Tongpool and
Pongpat, 2013). However, the impacts of sajor-caju in
these two categories were higher than those for normal
produce, such as tomato, onion, bush bean, baby corn
and potato (0.1-0.5 kg CO2 eq/kg, 0.03-0.07 kg oil eq)
(Tongpool et al., 2011).
“Sterilizing” was a major impact contributor in
the categories of climate change and terrestrial
acidification, while the “substrate” was the main
contributor in the categories of water depletion and
fossil fuel depletion. Both “sterilizing” and “substrate”
recorded for the medium-sized farms were higher than
those of small and large farms.
1585
 Ueawiwatsakul et al./Environmental Engineering and Management Journal 17 (2018), 7, 1583-1590

Table 1. Inventory of 1 kg sajor-caju produced at large, medium and small farms

Farm size
List Unit
Large Medium Small
Input Mixing of substrates
Sawdust kg 2.0725 2.4779 2.3343
Rice bran kg 0.1084 0.1495 0.0969
Urea fertilizer kg 0.0015 0.003 0.0015
Other ingredients kg 0.0583 0.04845 0.0392
Pakaging e.g. bag, neck ring kg 0.0256 0.0274 0.0269
Total weight of substrate kg 2.2663 2.7063 2.4988
Electricity kWh 0.0042 0.0030 0.0014
Transportation ton-km 1.4711 1.7131 1.5100
Sterilizing of substrates
Wood kg 4.7650 5.9070 4.4527
Liquefied petroleum gas (LPG) kg 0.0084 0 0
Transportation ton-km 0.1236 0.1673 0.1533
Tap water m3 0.0008 0.0010 0.0020
Water from well or river m3 0.0012 0.0013 0.0001
Inoculation with spawn
Transportation ton-km 0.0002 0.0002 0.0002
Spawn kg 0.0091 0.0102 0.0101
Environmental control for spawn run and fruiting
Pesticides kg 2 x 10-5 4 x 10-5 2 x 10-5
Water from well or river m3 0.0031 0.0065 0.0016
Tap water m3 0.0018 0.0060 0.0094
Gasoline m3 2 x 10-7 5 x 10-7 5 x 10-7
Electricity kWh 0.0226 0.0117 0.0011
Transportation ton-km 0.0001 0.0002 0.0002
Harvesting
Transportation ton-km 0.0042 0.0041 0.0041
Plastic bag kg 0.015 0.015 0.015
Output Sajor-caju in plastic bag kg 1 1 1
Unmanaged solid wastes kg 1.2832 1.8662 1.6520
NOx kg 0.0070 0.0087 0.0065
SO2 kg 0.0008 0.0010 0.0008
NH3 kg 0.0030 0.0038 0.0028
CO2 kg 0.0466 0.0298 0.0194
CH4 kg 0.0230 0.0284 0.0215
N2O kg 0.0003 0.0004 0.0003

(a) (b)

(c) (d)
Fig. 2. Environmental impacts of 1 kg sajor-caju production of the large, medium and small farms, in the categories of:
(a) climate change, (b) terrestrial acidification, (c) water depletion and (d) fossil fuel depletion

1586
 Environmental performance of sajor-caju mushroom production based on farm sizes in Thailand

Fig. 3 shows components contributing to the
impact of “sterilizing” in the medium-sized farms. It
can be seen that the emissions from “sterilizing”
contributed almost 100% in the categories of climate
change and terrestrial acidification. These emissions
came from wood combustion which releases relatively
high amounts of CH4, N2O and SOx, compared to LPG
combustion, considering the same given energy
(Eggleston, et al., 2006; EEA, 2013).
Fig. 4 shows components that caused
“substrate” of the medium-sized farms to have high
environmental impacts. It can be seen that the
production of sawdust, rice bran, packaging and
transportation of sawdust were the main sources of the
impacts. This corresponded to the data shown in Table
1 that the medium-sized farms used relatively large
amounts of sawdust, rice bran and packaging.
Although rice bran, broken rice and rice husk are by-
products of milled rice production, they have
economic values. As a result, the environmental
burdens from the milled rice production were
allocated to all products according to their economic
values; 71% to milled rice, 23% to broken rice, 4% to
rice bran and 2% to rice husk. This ratio was derived
from the multiplication of the market price and the
amount of each product obtained from rice milling.
The amounts of rice products from the rice milling
were obtained from Nadsathaporn (2007). Sawdust
and slab are also by-products of lumber production.
Since all products can be sold, the environmental
burdens from the lumber production were allocated to
all of them according to the mass ratio; 50% to lumber,
10% to sawdust and 40% to slab (MTEC, 2013).
Although Table 1 shows that water used for
humidifying fruiting houses of the small and medium
farms was 2-3 times of that used in the large farms,
Fig. 2e (water depletion category) reveals that its
contribution to life-cycle water depletion of sajor-caju
was insignificant compared to the water consumption
of “substrate” dominated by rice bran production (Fig.
4).

Fig. 3. Impact components of “sterilizing” in the medium-sized farms

Fig. 4. Impact components of “substrate” for the medium-sized farms

1587
 Ueawiwatsakul et al./Environmental Engineering and Management Journal 17 (2018), 7, 1583-1590

Fig. 5. Components of wastes from the production of 1 kg sajor-caju in the large, medium and small farms

Moreover, the impact of the “fruiting” (Fig. 2)

was relatively small although a large amount of
electricity was used in the environmental control for
spawn run and fruiting (Table 1). Components of
wastes from the small, medium and large farms are
shown in Fig. 5. The medium-sized farms produced
the highest amount of wastes, most of which were
spent substrates. Hence, in medium-sized farms, more
attention should be paid to efficient use of substrate.
(a) (b)
3.2. Eco-efficiency

Table 2 shows the cost of raw materials, energy

and labor as well as profit for the production of 1 kg
sajor-caju. The major cost came from sawdust, rice
bran, packaging, wood and labor. The total cost in
small farms was relatively low and thus the profit was
relatively high. This leads to high eco-efficiencies of
sajor-caju production in small farms as shown in Fig.
6. Medium-sized farms showed the lowest eco- (c) (d)
efficiencies as a result of having the lowest profit and
the highest environmental impacts (Table. 2). This Fig. 6. Eco-efficiencies of sajor-caju production for the
means that medium-sized farms would cause more large (L), medium (M) and small (S) farms, in the
environmental impacts in order to obtain the same categories of (a) climate change, (b) terrestrial
acidification, (c) water depletion and
profit as the small and large farms. (d) fossil fuel depletion
Table 2. Cost of raw materials, energy, labor and profit for
the production of 1 kg sajor-caju in the large,
The reason that the small and large farms had
medium and small farms similar environmental performance and the medium-
sized farms performed worst can be related to the
Cost (Bath) resource management in each type of farm. In the
Details
Large size Medium size Small size small farms, the amounts of wood and substrates used
Cost for sajor-caju production were relatively low,
Sawdust 2.8925 3.4776 3.2972 implying that raw materials usage was more carefully
Rice bran 0.8405 1.1955 0.7646
Urea fertilizer 0.0533 0.0891 0.0511
controlled, probably because of financial limitation. In
Spawn 0.6243 0.7032 0.6944 the large farms, the amounts of substrates, wastes and
Packaging 2.0989 2.1794 2.1541 water used for humidifying the fruiting house were the
Other materials 0.6616 0.6090 0.5735 lowest, implying that attention was given to efficient
Electricity 0.0739 0.0406 0.0069 use of resource.
Gasoline 0.0064 0.0209 0.0203
Wood 1.0417 1.1918 1.2737
In the medium-sized farms, the amounts of
LPG 0.1550 0 0 wood, substrates and wastes were relatively high,
Labor 2.0393 1.4097 1.5109 implying that efficient use of resources was not a
Total 10.4874 10.9168 10.3467 concern, due to relatively low financial limitation,
Profit 19.5126 19.0832 19.6533 compared to the small farms.

1588
 Environmental performance of sajor-caju mushroom production based on farm sizes in Thailand
3.3. Improvement of environmental performance
It was shown that, in the medium-sized farms,
a relatively large amount of substrate was used and
subsequently left after the harvest. This implies that it
is possible for these farms to reduce the substrate
amount while still achieving the same production
yield. As sawdust and rice bran largely contributed to
environmental impacts (Fig. 4) and total cost (Table 2)
of the sajor-caju production, it is proposed that the
medium-sized farms use the same amounts of sawdust
and rice bran as those of the small farms. It was
revealed in section 3.1 that the medium-sized farms
used more energy in the sterilizing process than the
other two sizes of farm. Moreover, the emissions from
the sterilizing process were the main impact
contributors (Fig. 3). Therefore it is proposed that the
amount of wood was also reduced to the same amount
of that used in the small farms.
The results from this option, shown in Table 3,
demonstrated that 5-25% of environmental impacts of
sajor-caju could be reduced. The climate change
potential of the medium-sized farms decreased from
1.40 to 1.14 kg CO2eq/kg which was closed to that of
the small farms (1.13 kg CO2eq/kg) and slightly lower
than that of the large farms (1.18 kg CO2eq/kg). Water
depletion potential was reduced from 0.28 to 0.21
m3/kg, which was close to those of the small and large
farms. The profit of the medium-sized farms was
consequently increased from 19.08 to 20.00 Baht/kg.
The change in both profit and environmental impacts
resulted in 10-40% increase in eco-efficiencies (Table
3). These new eco-efficiencies were slightly higher
than those recorded for the large farms.
4. Conclusions
The environmental impacts of sajor-caju
production are derived mainly from the production,
transportation and sterilization of substrates. Medium-
sized farms showed relatively high environmental
impacts and relatively low profit and eco-efficiency,
compared to small and large farms. It was found that
there was inefficiency use of substrate and energy in
the medium-sized farms. After the sawdust, rice bran
and wood were reduced to the same amounts as those
used in the small farms, the environmental
performance of medium-sized farms was improved to
be close to that of the large and small farms.
Acknowledgements
This study was accomplished through the financial support
Table 3. Environmental impacts and eco-efficiencies of 1 kg sajor-caju produced at medium-sized farms after the amounts of
sawdust, rice bran and wood were reduced
Eco-efficiency
% Increase Unit New value % Reduction
29.1 Bt/kg CO2eq 17.6 18.8
32.9 Bt/kg CO2eq 1420 21.1
39.5 Bt/m3 93 24.9
10.6 Bt/kg oil eq 107 5.3
of National Science and Technology Development Agency
(NSTDA) Thailand, under the Thailand National LCI
Database Project, Thailand Advanced Institute of Science
and Technology and Tokyo Institute of Technology
(TAIST-Tokyo Tech), and Faculty of Engineering,
Kesetsart University. We acknowledge the department of
agriculture and farmers in Ratchaburi province, Thailand,
for data provision. We thank Dr. John Thomas Harry Pearce
for proof reading the article.
References
Blengini G.A., Garbarino E., Bevilacqua P., (2017),
Sustainability and integration between mineral
resources and C&DW management: Overview of key
issues towards a resource-efficient Europe,
Environmental Engineering and Management Journal,
16, 493-502.
Bojnec S., Latruffe L., (2013), Farm size, agricultural
subsidies and farm performance in Slovenia, Land use
policy, 32, 207-217.
BSI, (2008), PAS 2050:2008 Specification for the
assessment of the life cycle greenhouse gas emissions
of goods and services, British Standards Institution, On
line at: http://shop.bsigroup.com/Browse-By-
Subject/Environmental-Management-and-
Sustainability/PAS-2050/.
DEDE, (2011), Annual report: Oil of Thailand 2011,
Department of Alternative Energy Development and
Efficiency, Ministry of Energy, Thailand, On line at:
http://www.dede.go.th/dede/index.php?option=com_c
ontent&view=article&id=1841&Itemid=318&lang=en
DOAE, (2012a), Statistic of mushroom cultivation in
Thailand, Department of Agricultural Extension,
Thailand.
DOAE, (2012b), Statistic of sajor-caju mushroom
cultivation in Thailand, Department of Agricultural
Extension, Thailand.
DOA, (2012), Statistics of sajor-caju mushroom cultivation
in Ratchaburi province, Department of Agriculture,
Ratchaburi, Thailand, On line at:
http://www.ratchaburi.doae.go.th/webpage/main.html.
Dong Y.H., Ng S.T., Kumaraswamy M.M., (2016), Critical
analysis of the life cycle impact assessment methods,
Environmental Engineering and Management Journal,
15, 879-890.
Ecoinvent Centre, (2007), The Ecoinvent database, Swiss
Centre for life Cycle Inventories, On line at:
http://www.ecoinvent.org/database.
Eggleston H.S., Buendia L., Miwa K., Ngara T., Tanabe K.,
(2006), 2006 IPCC Guidelines for National Greenhouse
Gas Inventories, Intergovernmental Panel on Climate
Change, On line at: http://www.ipcc-nggip.iges.or.jp.
EEA, (2013), EMEP/EEA Air Pollutant Inventory
Guidebook, European Environment Agency, On line at:
http://www.eea.europa.eu/publications/emep-eea-
guidebook-2013.
Environmental impact Impact category
Unit New value
kg CO2 eq 1.135 Climate change
kg SO2 eq 0.014 Terrestrial acidification
m3 0.214 Water depletion
kg oil eq 0.187 Fossil fuel depletion
1589
 Ueawiwatsakul et al./Environmental Engineering and Management Journal 17 (2018), 7, 1583-1590
Finnveden G., Hauschild M.Z., Ekvall T., Guine J., Heijungs
R., Hellweg S., (2009), Recent developments in life
cycle assessment, Journal of Environmental
Management, 91, 1-21.
Ghinea C., Petraru M., Simion I.M., Sobariu D., Bressers
H.T.A., Gavrilescu M., (2014), Life cycle assessment
of waste management and recycled paper systems,
Environmental Engineering and Management Journal,
13, 2073-2085.
Goedkoop M.J., Heijungs R., Huijbregts M., De Schryver
A., Struijs J., Van Zelm R., (2009), ReCiPe 2008, A life
cycle impact assessment method which comprises
harmonized category indicators at the midpoint and the
endpoint level, Ministry of Housing, Spatial Planning
and Environment, The Netherlands, On line at:
http://www.lcia-recipe.net.
Gunady M.G.A., Biswas W., Solah V. A., James A. P.,
(2012), Evaluating the global warming potential of the
fresh produce supply chain for strawberries,
romaine/cos lettuces (Lactuca sativa), and button
mushrooms (Agaricus bisporus) in Western Australia
using life cycle assessment (LCA), Journal of Cleaner
Production, 28, 81-87.
ISO, (2006a), ISO 14040 Environmental management-life
cycle assessment-principles and framework,
International Organization for Standardization,
Geneva, Switzerland, On line at:
https://www.iso.org/standard/37456.html.
ISO, (2006b), ISO 14044 Environmental management-life
cycle assessment-requirements and guidelines,
International Organization for Standardization,
Geneva, Switzerland, On line at:
https://www.iso.org/standard/38498.html.
Kanellopoulos A., Reidsma P., Wolf J., van Ittersum M. K.,
(2014), Assessing climate change and associated socio-
economic scenarios for arable farming in the
Netherlands: An application of benchmarking and bio-
economic farm modeling, European Journal of
Agronomy, 52, 69-80.
1590
Manjunatha A.V., Anik A. R., Speelman S., Nuppenau E.
A., (2013), Impact of land fragmentation, farm size,
land ownership and crop diversity on profit and
efficiency of irrigated farms in India, Land Use Policy,
31, 397-405.
MTEC, (2013), Thai national life cycle inventory database,
National Metal and Materials Technology Center, On
line at: http://www.thailcidatabase.net.
Nadsathaporn H., (2007), Environmental life cycle
assessment of rice products, MEng Thesis, Faculty of
Environmental Engineering, Suranaree University of
Technology, Nakorn Ratchasima, Thailand.
Pishgar-Komleh S.H., Ghahderijani M., Sefeedpari P.,
(2012), Energy consumption and CO2 emissions
analysis of potato production based on different farm
size levels in Iran, Journal of Cleaner Production, 33,
183-191.
Pre Consultants, (2013), SimaPro software, the Netherlands,
On line at: http://www.pre-
sustainability.com/simapro?gclid=CNegroK29b0CFdc
Vjgodp04AVg.
Tongpool R., Jirajariyavech A., Yuvaniyama C.,
Mungcharoen T., (2010), Analysis of steel production
in Thailand: Environmental impacts and solutions,
Energy, 35, 4192-4200.
Tongpool R., Yuvaniyama C., Mungcharoen T., (2011), The
Sustainable Production of Food Crops via Eco-
efficiency, Proc. 6th Dubrovnik Conference on
Sustainable Development of Energy, Water and
Environment Systems, Dubrovnik, Croatia.
Tongpool R., Pongpat P., (2013), Analysis of Shiitake
environment performance via life cycle assessment,
International Journal of Environmental Science and
Development, 4, 552-557.
Verfaillie, H.A., Bidwell, R., (2000), Measuring eco-
efficiency-a guide to reporting company performance,
World Business Council for Sustainable Development,
Archive, On line at:
https://www.gdrc.org/sustbiz/measuring.pdf.
Environmental Engineering and Management Journal July 2018, Vol.17, No. 7, 1591-1597
http://www.eemj.icpm.tuiasi.ro/; http://www.eemj.eu

“Gheorghe Asachi” Technical University of Iasi, Romania

INDIRECT ENVIRONMENT-RELATED EFFECTS

OF ELECTRIC CAR VEHICLES USE

Bogdan Ovidiu Varga, Florin Mariasiu ∗

Technical University of Cluj-Napoca, Automotive Engineering & Transports Department,

103-105 Muncii Bd., Cluj-Napoca, Romania

Abstract

The article presents a study on indirect environment-related effects of electric car vehicles use. Although electric vehicles are
considered zero emission vehicles, their use may be linked to indirect pollution of environment caused by energy sources used for
electricity production and also other sources of pollution such as poor recycling procedures. A computer simulation was performed
to determine the energy consumption of electric vehicles available on the Romanian car market. The values obtained were
correlated with emissions of CO2, NOx and SOx pollutants. Furthermore, a supplementary analysis was performed to determine the
necessary combination and related management of energy sources to minimize the environmental impact of electric vehicles, in
order to increase their use for a future sustainable transportation domain. The study shows that SOx emissions in particular will
increase in Romania alone, as a result of the implementation of electric cars, unless alternative and renewable energy sources are
implemented in the society. This study reports crucial statistics and information which outline that legislative policies must be in
place before a full implementation of electric vehicles is carried out, in order to avoid increase in emissions countries without the
direct access to renewable energy resources.

Key words: electric vehicles, energy sources, environment, management, pollutant emissions, modeling

Received: June, 2014; Revised final: August, 2014; Accepted: September, 2014; Published in final edited form: July, 2018

1. Introduction hybrid (internal combustion engine + electric motor)

According to recent scientific articles and
studies, the road transport sector is responsible for
23.1% of the total emissions in the atmosphere in the
European region (OECD/ITF, 2010). As an
immediate response for reducing the amount of GHG
(Green House Gases) emissions, the use of renewable
energy resources in internal combustion engines
fueled using blends of biofuels and fossil fuels was
intensified (Mariasiu, 2013). Ng et al., (2010) and
Moldovanu and Burnete (2013) showed that although
it resulted a significant reduction in vehicle’s related
pollutant emissions using renewable sources, it is still
necessary to continue research and development
activities in the field of road vehicles to increase the
efficiency of these results. At present, the use of
and electric vehicles (EV) technologies in vehicle’s
powertrain appears to be a feasible solution for
passenger cars. For example, the CO2 emission of the
Toyota Prius hybrid vehicle is 20% lower than
emission of a similar car that using only an internal
combustion engine powered with fossil fuels (Fuhs,
2009; Thompson et al., 2011). The continuous
development of technology employed in electric
vehicles is a promising and viable solution to the
emission problems and a potential approach for
significant reduction of GHG emissions (Holdway et
al., 2010; Huo et al., 2010; Huang and Ren, 2013).
Even so, Grunig et al. (2011) reported that the
market penetration of electric vehicles will remain low
compared to conventional vehicles in the near future
(the main market is represented only by passenger

∗
Author to whom all correspondence should be addressed: e-mail: florin.mariasiu@auto.utcluj.ro; Phone +040730011747; Fax +040264415950.
 Varga and Mariasiu/Environmental Engineering and Management Journal 17 (2018), 7, 1591-1597
cars). Mariasiu (2012) and Liu et al. (2008) consider
that the main technologically barriers for increasing
the car market share for EVs are considered to be the
following factors:
- the average EV purchase price is 100% greater than
for a new IC engine car;
- there are serious limitations regarding the EVs
autonomy and charging time (average value for
autonomy is considered to be 130 km and mostly EVs
need 6 hours for a complete charging cycle);
- the low capacity of electric battery;
- the usable energy depending directly on many factors
(type of battery, type of powertrain, driving style,
traffic conditions, ambient temperature and charging
technology);
- proper recycling schemes for car batteries
throughout the cradle-to-grave life cycle.
Given the above-presented barriers, at present,
the EU countries and authorities are taking numerous
political, economic and social measures, to promote
and increase the penetration of electric vehicles in
transportation domain. To achieve a target of
4,752,100 EVs for the end of 2015 at European level
measures as: free parking, right to use bus-lanes, free
entry to city-centers, financial subsidies and
incentives, exempts from registration tax, VAT,
annual car tax and congestion charges are some of
adopted measures (Reiner et al., 2010; Nemry and
Brous, 2010).
It is also interesting to highlight that measures
aimed at increase number of EVs include many
subsidy programs in EU, most of which refer to
taxation of CO2 emission level (“CO2-based car
taxes”). It can be easily to see that the indirect CO2
emissions from energy production are not perceived to
be a major barrier to EVs use and are not taken into
consideration (Brinkman et al., 2010; Jansen et al.,
2010). Unfortunately, on closer analysis, it is evident
that EVs must be powered with electricity (to charge
battery), which is not obtained only from
environmentally friendly sources (or at least from
renewable sources) (Varga, 2013). Europe-wide
primary sources for obtaining electricity rely heavily
on coal, oil, gas and nuclear energy. In Fig. 1 is shown
Fig. 1. Structure of primary sources for electricity production (Eurostat, 2013)
1592
the structure of primary energy sources for electrically
energy production comparatively for Europe and
Romania. The coal remains an important primary
source of electrical energy production with direct
implications on the level and structure of pollutant
emissions in atmosphere, both for EU 28 and
Romania. In Romania, the use of coal for energy
production is with 14.3% higher than the EU average,
indicating that Romania must adopt energy policies
that lead to achieving at least EUs average value. As
major issue of energetic strategy, the Romanian
government makes efforts to complete units 3 and 4 of
Cernavoda nuclear power plant, together with the
massive development of renewable energy production
units. The reducing CO2 (and also GHG) emission
policy is also highlighted by government policies to
support the increase of EVs market. Thus, for each EV
purchased, government offers a subsidy from purchase
price for a new EV acquisition.
Currently, in Romania, there are only four EV
manufacturers that offer to customers a total of six
models (Table 1). There are important differences in
EVs models, between values pertaining to the
autonomy for a full battery charge cycle. For these
EVs, models available on Romania’s car market, a
computer simulation process was used to identify the
EVs with minimum and maximum electric energy
consumption during a standard NEDC test. Further,
the electric energy consumption was converted to
pollutant emissions, considering the particular
structure of primary energy sources (POST, 2006).
The scenario-based analyses were conducted in order
to establish whether the indirect pollutant emissions
(CO2, NOx and SOx) could achieve the limits imposed
by contemporary pollution norm (Euro 6), and
evaluation in terms of their environmental impact.
Besides major influence on air pollution of CO2
emissions (by increasing the greenhouse effect), and
other pollutant emissions from road transport have
direct influence on the environment, and their
magnitude when using electric vehicles is a novelty
presented in this paper. SOx emissions were
considered to be dangerous due to their contribution to
the direct creation of smog and acid rain.
 Indirect environment-related effects of electric car vehicles use
Those negative effects on environment and
human health have been drastically eliminated by
international norms that involving the use of fuels with
low sulfur content. To present date (pending the
introduction of Euro 6 pollution norm), NOx emissions
were not a major concern in terms of limiting by law
to certain emission limit values.
The advent of well-documented studies
regarding the effects of direct and indirect NOx
emissions on the environment and human health, were
determined their influence. From an environmental
perspective, the negative effects associated with NOx
emissions are soil nitrification with immediate
detrimental effects on plant growth and soil
microorganisms’ life. Also, the toxic effect of NOx
emissions on human health manifests (especially in
big cities) with respiratory and cardiovascular
problems (Gustavsson et al., 2000; Weiss et al., 2012).
For these reasons the present work shows a broad
approach on to the quantity and composition of
pollutant emissions emitted indirectly, in case of use
of an electric vehicle.
2. Material and methods
Analysis and computer simulation are research
tools widely used in many scientific fields. An
immediate advantage of using computer simulation is
the opportunity to obtain, at a relatively low cost, a
multitude of experimental data and possible cases of
studied phenomena. In the present work, to obtain the
EVs CO2 emission level for the standardized cycle
(NEDC - testing procedure recommended by the
European Union) conditions, CRUISE simulation
environment were used. CRUISE software package
was developed and implemented successfully in the
automotive industry by AVL List GmbH (2006)
(Varga, 2012).
Major benefit of CRUISE simulation software
is its modularity and interactive process of simulation
and design of model for a various type of vehicles. For
brevity, specific details of using CRUISE, how to
build a vehicle model and how each function is used
are given in AVL User Manual (2009) and Varga
(2012). Primary data used as input for the simulation
model, are technical data of considered electric
vehicles presented in Table 1. The method used to
Table 1. Technical data of EVs models available on Romanian car market (Fuhs, 2009; Brooke, 2011)
Vehicle Battery
Urban
EV Electric vehicle mass mass
range (km)
(kg) (kg)
Mitshubishi
1 1,080 200 160
iMiEV
2 Chevrolet Volt 1,588 170 64
3 Renault Zoe ZE 1,100 200 210
4 Renault Kangoo ZE 1,520 250 160
5 Renault Fluence ZE 1,600 250 160
6 Toyota Prius Plug-in 1,380 155 23
calculate the amount of the each considered pollutant
emission is presented below, in the case of CO2 (but
applicable also for NOx and SOx emissions). After
obtaining data related to energy consumption for an
electric vehicle, the transformation into the amount of
pollutant emitted into the atmosphere has been
achieved, taking into account the characteristics of
Romania’s primary energy sources mix used for
electricity production in power stations. In case of CO2
pollutant, the following relation was used (Mariasiu,
2012) (Eq. 1):
(1)
were EVCO2/km is the amount of CO2 emitted by energy
consumption per km, LNEDC is the length of NEDC
cycle (11.008 km), EVCO2/NEDC is the amount of CO2
emitted by energy consumption during a NEDC cycle,
CO2energy_source is the amount of CO2 emission function
of energy source and the Plosses = 1.15-1.22 are the
losses caused by electric distribution grid and internal
EVs electrical circuits (Tesla Motors, 2010).
3. Results and discussion
The input data required to perform computer
simulation are presented in previous sections. To
obtain accurate results, the technical and operational
particularities of each model of EV were taken into
account (total weight of the vehicle, wheel diameter,
transmission type, aerodynamic coefficient, battery
type, energy efficiency and losses).
Further, the pollutant emission levels obtained
for EVs operation, considered in the study, was
correlated with energy sources pollutant emissions
used to produce the electricity. The results analysis
indicates that the lowest energy consumption is
obtained for EV2 (Chevrolet Volt) and the highest for
EV3 (Renault Zoe). Fig. 2 depicts battery energy
consumption for Chevrolet Volt during the NEDC
cycle and related cumulative results for considered
EVs and obtained from computer simulation are
presented in the Table 2. The amount of pollutant
emissions corresponding to each source of energy for
electricity production and distribution of these sources
for EU28 and Romania are presented in the Table 3.
Charging
Battery capacity Purchase
time
(kWh) cost (Euro)
(hrs:min)
16 07:00 37,767
16 08:00 29,347
22 06:00 20,700
24 06:00 21,340
24 06:00 26,300
5.2 01:40 34,750
1593
 Varga and Mariasiu/Environmental Engineering and Management Journal 17 (2018), 7, 1591-1597

The EVs emission of CO2 varies between a

minimum for EV2 (98.42 gCO2/km) and maximum
for EV3 (100.84 gCO2/km). Compared with a compact
car that is powered by fossil fuels (with average of 130
gCO2/km emission), these values are 24.3% and
22.4% lower, respectively. However, EU directives
stipulate that by 2020, the value of CO2 emissions in
transport must not exceed 95 g/km (Eurostat, 2011),
and values presented above are 3.47 – 6.14% higher.
In Romania, if some scenarios (Tables 6 to 8)
are applied, we can analyze the environmental effects
of electric vehicles use (currently available in the
Romania’s car market.) Different amount of coal, oil
and natural gas as energy sources were replace by
nuclear and renewable energy. The scenario-based
analyses considers that: nuclear unit no.3 is functional
in Scenario A, scenario B considers that nuclear units Fig. 2. Energy consumption during the NEDC cycle
no.3 and 4 are functional and scenario C considers that
nuclear units no.3 and 4 are functional and 14% to The pollutant emissions data related to the use
20% more electrical generation from wind and solar of considered EVs are presented in the Tables 3 to 5
(values were calculated using Eq. 1).
sources (renewable sources).

Table 2. EVs energy consumption vs. CO2 emissions (for NEDC test)

Electric Vehicle (EV)

Parameters
1 2 3 4 5 6
Energy consumption (kWh) 1.711 1.672 1.713 1.678 1.689 1.688
CO2 emission (g/km) 100.72 98.42 100.84 98.78 99.43 99.37

Table 3. CO2 emissions depending on the nature and distribution of energy sources

EU 28 Romania
Electrical Generation Grams of CO2
Share of energy CO2 emission Share of energy CO2 emission
Sources emissivity (g/kWh)
sources (g/kWh) sources (g/kWh)
Coal 1000 18% 180 22% 220
Oil 650 35% 227.5 25% 162.5
Natural Gas 500 23% 115 31% 155
Renewable 15 10% 1.5 14% 2.1
Nuclear 5 14% 0.7 8% 0.4
Total - 100% 524.7 100% 540

Table 4. NOX emissions depending on the nature and distribution of energy sources

EU 28 Romania
Electrical Grams of NOX
Share of energy NOX emission Share of energy NOX emission
Generation Sources emissivity (g/kWh)
sources (g/kWh) sources (g/kWh)
Coal 6 18% 1.08 22% 1.32
Oil 4 35% 1.4 25% 1
Natural Gas 1.7 23% 0.391 31% 0.527
Renewable 0.006 10% 0.006 14% 0.008
Nuclear 0 14% 0.00 8% 0.00
Total - 100% 2.877 100% 2.855

Table 5. SOX emissions depending on the nature and distribution of energy sources

EU 28 Romania
Electrical Generation Grams of SOX
Share of energy SOX emission Share of energy SOX emission
Sources emissivity (g/kWh)
sources (g/kWh) sources (g/kWh)
Coal 13 18% 2.34 22% 2.86
Oil 12 35% 4.2 25% 3
Natural Gas 0.1 23% 0.023 31% 0.031
Renewable 1.22 10% 0.122 14% 0.17
Nuclear 0 14% 0.00 8% 0.00
Total - 100% 6.685 100% 6.061

1594
 Indirect environment-related effects of electric car vehicles use

Considering Romania’s future policy and If we consider the most favorable case of EV2
strategy in energy production and following the above type use, the CO2 emissions (from energetic source
considerations and presented algorithm for mix) are shown in the Fig. 3, in line with considered
calculations, the results and related discussions are scenarios (reducing the share of energy from coal and
presented below. CO2 emission decreases 5.5% in increase in same time the energy share from renewable
scenario A, 12.9% in scenario B and 19.1% in scenario sources and nuclear).
C, compared to the present situation. The target of
below 95 g/km of CO2 emissions (future Euro 6
pollution standard) is achieved in all cases of proposed
scenarios (Fig. 3).
Comparative analysis of NOx emissions
(compounded emissions of N2O and NO2) depending
on the energy source used for electricity generation
and selected scenarios are presented in the Table 7. It
is worth mentioning that with the introduction of Euro
6 emission norms for road vehicles (EC, 2007), NOx
emissions are taken into account in evaluating the
performance of a car (pollutant emissions). The
maximum accepted by Euro 6 norm is considered to
be 0.06 g/km for passenger cars. By using electric
vehicles considered in the study, a maximum level of
emissions of SOx (1.1 g/km) results, considering the
current structure of energy sources used to produce
electricity. Fig. 3. EV2’s CO2 emission function of scenarios

Table 6. Scenarios on the share of primary energy resources - CO2 case

Scenario A Scenario B Scenario C

Electrical CO2 CO2 CO2
Share of energy Share of energy Share of energy
Generation Sources emission emission emission
sources (%) sources (%) sources (%)
(g/kWh) (g/kWh) (g/kWh)
Coal 19 190 15 150 15 150
Oil 25 162.5 25 162.5 22 143
Natural Gas 31 155 31 155 28 140
Renewable 14 2.1 14 2.1 20 3
Nuclear 11 0.55 15 0.75 15 0.75
Total 100 510.55 100 470.35 100 436.75

Table 7. Scenarios on the share of primary energy resources - NOX case

Scenario A Scenario B Scenario C

Electrical NOX NOX NOX
Share of energy Share of energy Share of energy
Generation Sources emission emission emission
sources (%) sources (%) sources (%)
(g/kWh) (g/kWh) (g/kWh)
Coal 19 1.14 15 0.9 15 0.9
Oil 25 1 25 1 22 0.88
Natural Gas 31 0.527 31 0.527 28 0.476
Renewable 14 0.008 14 0.008 20 0.012
Nuclear 11 0.00 15 0.00 15 0.00
Total 100 2.675 100 2.435 100 2.268

Table 8. Scenarios on the share of primary energy resources - SOX case

Scenario A Scenario B Scenario C

Electrical
Share of Share of Share of
Generation SOX emission SOX emission SOX emission
energy sources energy sources energy sources
Sources (g/kWh) (g/kWh) (g/kWh)
(%) (%) (%)
Coal 19 2.47 15 1.95 15 1.95
Oil 25 3 25 3 22 2.64
Natural Gas 31 0.031 31 0.031 28 0.028
Renewable 14 0.17 14 0.17 20 0.244
Nuclear 11 0.00 15 0.00 15 0.00
Total 100 5.671 100 5.151 100 4.862

1595
 Varga and Mariasiu/Environmental Engineering and Management Journal 17 (2018), 7, 1591-1597
Note that using electric vehicles, the accepted
level is exceeded more than 8 times (Fig. 4), with
values between 0.49 g/km (scenario A) and 0.41 g/km
(scenario C).
Fig. 4. EV2’s NOx emission function of scenarios
SOx emissions are directly responsible for the
existence of acid rain, rain that present an immediate
danger potential on the environment and human
health. SOx emissions are high due to the high
proportion of use of coal in electricity production, both
for Romania (22% of total energy production), as well
as in the case of Europe (18% of total energy
production).
In case of programs and / or policies to improve
the structure of energy sources through intensive use
of renewable and/or non-fossil (Table 8, scenario C),
emissions of SOx emissions may reach a minimum
estimated value of 0.89 g/km, but still remain a great
value considering that actually the emission of a
passenger car is limited to 0.02 g/km (Fig. 5).
Fig. 5. EV2’s SOx emission function of scenarios
The proposed scenarios offer the possibilities
to reduce (but not to eliminate) the negative effects on
environment of acid rain.
4. Conclusions
This study evaluates the importance of indirect
emissions as environmental barrier for EVs rapid and
consistent car market penetration. EVs are considered
1596
zero-emission vehicles; however, electricity (used to
recharge the EV’s battery) produced in power plants
is still predominantly based on the fossil fuel usage
and thus directly affects the quantity (amount) of
pollutant and greenhouse gases emissions.
Given the structure of energy sources mix (in
particularly, for Romania’ case) used for electricity
generation, the major pollutant emissions originating
from the combination of these energy sources (related
to energy consumption of available EVs on Romanian
market) was determined.
In terms of compliance with future European
standards of pollution emissions (95 gCO2/km –
2020), the calculated CO2 indirect emissions of EVs
use, failed to achieve the imposed limits, considering
the actual mix of Romania’s energy sources for
electricity production (scenarios A and B). If a part of
coal share (only 4%) will be replaced with renewable
and non-fossil (nuclear) sources, the EVs use indirect
CO2 emission is with 2% lower than the future EURO
6 pollution standards. Worryingly are the high values
of indirect NOx and SOx emissions that occur when
using electric vehicles, which can lead to increasing
the frequency of occurrence of acid rain, with
immediate adverse effects on the environment and
human health.
As a major conclusion, it can be said that,
before introducing EVs on car market, economic
policies and management of energy sources mix used
for electricity production must take into account the
indirect environmental effects that may occur as a
consequence of this initiative. Prior to make major
investments in recharging infrastructure of EVs
batteries and related policies to encourage the EVs
market penetration, it is necessary to invest in
modernization of actual power plants (increasing
production efficiency and reducing pollutant
emissions) and the structure of energy sources used
(renewable and/or non-fossil sources).
However, even if electric vehicles use poses
certain problems of indirect pollutants emissions high
levels, there is the possibility of better management
regarding technologies applied to reduce pollution.
Reduction may be made directly to the source (power
plants) as a whole, with much better results than
implementing clean technologies (catalytic converter)
for each vehicle that uses an internal combustion
engine. It can thus achieve an economic cost
optimization in the reduction process of pollutant
emissions; differences can be used and/or directed to
other programs to increase environmental quality.
Acknowledgements
This work was done with support of “Parteneriat
interuniversitar pentru excelenta in inginerie - PARTING”,
POSDRU/159/1.5/S/137516 project.
References
AVL LIST GmbH, (2009), AVL Cruise User manual, On
line at: www.avl.com.
 Indirect environment-related effects of electric car vehicles use
Brinkman G.L., Denholm P., Hannigan M.P., Milford J.B.,
(2010), Effects of plug-in hybrid electric vehicles on
ozone concentrations in Colorado, Environmental
Science and Technology, 44, 6256-6262.
Brooke L., (2011), Chevrolet Volt - AEI best engineered
vehicle 2011, Automotive Engineering Magazine, 5, 22-
25.
EC, (2007), Regulation (EC) No. 715/2007 of the European
Parliament and of the Council of 20 June 2007 on type
approval of motor vehicles with respect to emissions
from light passenger and commercial vehicles (Euro 5
and Euro 6) and on access to vehicle repair and
maintenance information, Official Journal of the
European Union, L 171, 1-16.
Eurostat, (2011), Energy, transport and environmental
indicators, European Commission Publication no. KS-
DK-13-001-EN-C, On line at:
http://ec.europa.eu/eurostat/documents/3930297/5966
062/KS-DK-11-001-EN.PDF/a1caaacc-1f22-42fc-
8bab-93d0f808ea75.
Eurostat, (2013), Energy, transport and environmental
indicators, Pocket edition, European Commission
Publication no. KS-DK-11-001-EN-C, On line at:
http://ec.europa.eu
/eurostat/documents/3930297/5968878/KS-DK-13-
001-EN.PDF
Fuhs A.E., (2009), Hybrid Vehicles and the Future of
Personal Transportation, CRC Press, Boca Raton.
Grunig M., Witte M., Marcellino D., Selig J., Van Essen H.,
(2011), Impact of Electric Vehicles on the Market and
in Development, Publication no. 11.4058.03, CE Delft,
Delft, On line at: https://ec.europa.eu
/clima/sites/clima/files/transport/vehicles/docs/d1_en.p
df.
Gustavsson P., Jakobsson R., Nyberg, F., Pershagen G.,
Jarup L., Scheele P., (2000), Occupational exposure
and lung cancer risk: a population-based case-referent
study in Sweden, American Journal of Epidemiology,
152, 32-40.
Holdway R.A., Williams A.R., Inderwildi O.R., King D.A.,
(2010), Indirect emissions from electric vehicles:
emissions from electricity generation, Energy
Environment Science, 3, 1825-1832.
Huang S., Ren Y., (2013), Indirect carbon emission
measurement of electric vehicle in China, Applied
Mechanic and Materials, 476, 772-776.
Huo H., Zhang Q., Wang M.Q., Streets D.G., He K., (2010),
Environmental implication of electric vehicles in
China, Environmental Science and Technology, 44,
4856-4861.
Jansen K.H., Brown T.M., Samuelsen G.S., (2010),
Emisssions impacts of plug-in hybrid electric vehicle
deployment on the US Western grid, Journal of Power
Sources, 195, 5409-5416.
Liu X., Hildebrandt D., Glasser D., (2012), Environmental
impacts of electric vehicles in South Africa, South
African Journal of Science, 108, 1-6.
Mariasiu F., (2012), Energy sources management and future
automotive technologies: environmental impact,
International Journal of Energy Economics and Policy,
2, 342-347.
Mariasiu F., (2013), Consumers’ attitude related to biofuel
use in transportation, International Review of
Management and Marketing, 1, 1-9.
Moldovanu, D., Burnete, N., (2013), Computational fluid
dynamics simulation of a single cylinder research
engine working with biodiesel, Thermal Science, 17,
195-203.
Nemry F., Brous M., (2010), Plug-in Hybrid and Battery
Electric Vehicles. Market penetration scenarios of
electric drive vehicles, Technical note JRC-IPTS
58748, On line at: http: //ftp.jrc.es/ EURdoc /JRC
58748_TN.pdf.
Ng J.H., Ng H.K., Gan S., (2010), Advances in biodiesel fuel
for application in compression ignition engines, Clean
Technology and Environmental Policy, 12, 459-493.
OECD/ITF, (2010), Reducing Transport Greenhouse Gas
Emissions: Trends & Data 2010, On line at:
http://www.internationaltransportforum.org/Pub/pdf/1
0GHGTrends.pdf.
POST, (2006), Parliamentary Office of Science and
Technology, (Carbon footprint of electricity generation,
Report Number 268, On line at:
http://www.parliament.uk/documents/post/postpn268.
pdf.
Reiner R., Cartalos S., Evrigenis A., Viljamaa K., (2010),
Challenges for a European Market for Electric
Vehicles, Report ITRE-PE 440.276, On line at:
http://www.europarl.europa.eu/document/activities/co
nt/201106/20110629ATT22885/20110629ATT22885
EN.pdf.
Tesla Motors, (2010), Using energy efficiently, On line at:
http://www.teslamotors.com / go electric / efficiency.
Thompson T.M., King C.W., Allen D.T., Webber M.E.,
(2011) Air quality impacts of plug-in hybrid electric
vehicles in Texas: evaluating three battery charging
scenarios, Environment Research Letters, 6, 1-11.
Varga B.O., (2012), Energy management of electric and
hybrid vehicles dependent on powertrain configuration,
Central European Journal of Engineering, 2, 253-263.
Varga B.O., (2013), Electric vehicles, primary energy
sources and CO2 emission: applied scenario, Energy
Journal, 49, 61-70.
Weiss M., Bonnel P., Kuhlwein J., Provenza A., Lambrecht
U., Alessandrini S., Carriero M., Colombo R., Forni F.,
Lanappe G., Le Lijour P., Manfredi U., Montigny F.,
Sculati M., (2012), Will Euro 6 reduce the NOx
emissions of new diesel cars? – Insights from on-road
tests with Portable Emissions Measurement Systems
(PEMS), Atmospheric Environment, 62, 657-665.
1597
1598
Environmental Engineering and Management Journal July 2018, Vol.17, No. 7, 1599-1611
http://www.eemj.icpm.tuiasi.ro/; http://www.eemj.eu

“Gheorghe Asachi” Technical University of Iasi, Romania

THE “INSURANCE EFFECT”: HOW TO INCREASE THE INVESTMENT

AMOUNT IN GREEN BUILDINGS –A MODEL-BASED APPROACH TO
REDUCE THE ENERGY EFFICIENCY GAP

Hans Ulrich Buhl1, Tobias Gaugler2, Philipp Mette1∗

1FIM Research Center, University of Augsburg, Universitaetsstrasse 12, 86159 Augsburg, Germany
2Institute of Materials Resource Management (MRM), University of Augsburg, Alter Postweg 101, 86159 Augsburg, Germany

Abstract

Real estate is an industry sector with high potential to increase energy efficiency. However, many of the existing green building
investment opportunities (GBIO) are not utilized because economic valuation tools are complex and oftentimes difficult to
understand. For this reason, we develop a formal, but comprehensible bottom-up model to determine the optimal investment amount
from an economic perspective, placing particular emphasis on the descriptive valuation of risk, and point out the applicability of
GBIO as insurance against energy price volatility. We also give examples of the model’s potential application. Our work shows
that considering the insurance effect will increase the optimal investment amount and that certain investment amounts lead to both
economic and ecological benefits in properties and property portfolios. Our findings can be used for a comprehensible enhancement
of existing valuation methods and tools to reduce the energy efficiency gap (EEG). They constitute a quantitative basis for the
adaption of laws to counteract the current underinvestment in the real estate sector.

Key words: energy efficiency gap, energy price risk, green building, investment valuation

Received: September, 2012; Revised final: July, 2014; Accepted: September, 2014; Published in final edited form: July 2018

1. Introduction to increase energy efficiency is emphasized both from

To protect against the consequences associated
with global warming, European Union – similar to
other economic regions – aims to significantly reduce
the current energy consumption. This goal is related to
a reduction of greenhouse gas emissions by at least
20% until 2020 and by at least 80% until 2050
compared to 1990 levels (European Union, 2011). To
achieve this, a variety of measures to increase energy
efficiency is required. From a scientific perspective –
and based on Hirst and Brown (1990) – a large number
of publications refers to energy efficiency (Gillingham
et al., 2009; Patterson, 1996) and especially the energy
efficiency gap (e.g. Allcott and Greenstone (2012);
Jaffe and Stavins (1994); Koopmans and te Velde
(2001)). The development and application of measures
an economic perspective as well as for reasons of
climate change. Furthermore, the existence of an
energy efficiency gap (EEG) – defined as “the
difference between the amount of energy that
households and business currently consume and the
amount they ‘should‘ consume, relatively to (…) the
optimal level” (Klemick and Wolverton, 2013) – is
unanimously confirmed. This gap (resulting in the
energy paradox) is attributable to the existence of
various barriers, which – based on classifications
schemes of Gillingham et al. (2009) and Levine et al.
(1994) – can be categorized as follows:
Existing price distortions: The majority of
environmental and (long-term) social costs –
summarized as “external costs” – are not included in
the observable market price of energy. Furthermore,

∗
Author to whom all correspondence should be addressed: e-mail: mette@mail.de; Phone: + 49177 599 7011
 Buhl et al./Environmental Engineering and Management Journal 17 (2018), 7, 1599-1611
also state intervention (e.g. taxes and subsidies)
influence the market price of energy. These distortions
complicate conclusions about the ‘real’ height of the
EEG, which is also dependent on energy prices
(Levine et al., 1994).
Information problems: A variety of
information problems also influences the EEG. In
addition to the lack of information (e.g. on current
energy consumption patterns and ways to reduce this
consumption) (Levine, 1994), the principle-agent
problem is particularly important: In the context of
investments in green buildings this problem occurs, if
one party (the landlord) chooses the level of
investments in energy efficiency measures, but the
energy bills are paid by the other party (the tenant)
(Brown, 2001; Jaffe et al., 2005). In general,
information problems lead to an underinvestment in
energy efficiency (Gillingham et al., 2009; Jaffe and
Stavins, 1994).
Transaction costs: In our context, transaction
costs can be defined as costs of gathering, assessing
and applying information on the characteristics and
performance, making decisions as well as enforcing
contracts relating to the purchase and installation of
technologies, which reduce the energy demand of a
building (Hein and Blok, 1995; Levine et al., 1994).
The existence of transactions costs (which increase the
total cost of energy efficiency measures) results in
lower – and/or later – investments in energy efficiency
measures than it would be assumed from a purely neo-
classical perspective (Golove and Eto, 1996).
Capital market imperfections: Especially in
capital intensive industries like the real estate
business, liquidity constraints complicate the
quantification of the ‘real’ EEG. Limited access to
financing for energy-efficient investments is resulting
in underinvestment and therefore in an increasing
EEG (Gillingham et al., 2009).
In current literature, the existence of these
barriers (and the difficulties resulting from their
quantification) (Golove and Eto, 1996; Hein and Blok,
1995) lead to disagreements over the actual extent of
the EEG. One of the main reasons for this it that
different scales are used: To demonstrate the problems
related to the quantification of the EEG, Jaffe and
Stavins distinguish (already in 1994) between the
“economist’s economic potential”, the “technologist’s
economical potential” (which, in comparison, is
higher than the first scale), the even higher
“hypothetical potential” and the “narrow social
optimum” and the “true social optimum”. To illustrate
these differences, we refer to Cullen and Allwood
(2010), who argue that – from a purely technical
perspective – the global demand for energy could be
reduce by almost 90%, if energy conservation devices
were entirely employed. In contrast, other authors like
Allcott and Greenstone argue that „the actual
magnitude of the EEG is small relative to the
assessment from engineering analyses“ (2012). No
matter how big the EEG actually is – it plays an
important role in scientific discussions on real estates
(Jakob, 2006; Myers, 2014; Shove, 1999; Wilson and
1600
Dowlatabadi, 2007). The real estate sector is one of
the largest consumers of energy and is responsible for
approximately 40% of global energy use (Pinkse and
Dommisse, 2009) and 40% of all carbon emissions
(Apgar, 2009). Energy consumption of properties is
responsible for more than 30% of global final energy
demand and one-third of energy-related CO2
emissions and is feared to increase by further 33%
until 2050 (compared to 2005) in a sub-optimal
scenario (Global Energy Assessment, 2012). A study
by the McKinsey Global Institute (2007) revealed that
the largest potential for reducing energy demand and
the negative effects on the environment (commonly
quantified in the emission of CO2 equivalents, CO2-eq)
lies in the real estate sector. Existing technologies
could be integrated into so-called “green buildings”,
thereby increasing energy efficiency by 35%
(WBCSD, 2007). Although the ecological advantages
of green buildings to significantly improve their CO2-
eq footprint are evident (Edwards and Naboni, 2013;
Teng and Wu, 2014) and even net zero energy (NZE)
buildings are possible from a technical perspective
(Kibert, 2014), the investment rate in higher energy
efficiency is only 1–1.5 % in non-residential buildings
and only 0.07% in residential buildings (Rottke,
2009). These figures demonstrate the existence and
high relevance of EEG especially in the real estate
sector.
Various strands of research try to reduce the
EEG in the building sector. So, for example, a macro-
economically oriented strand examines the impact of
government programs to increase energy efficiency
(Abadie et al. 2012; Clinch and Healy, 2001; Dixon et
al., 2010; Hull et al., 2009). Another line of research
evaluates and ranks the variety of technical devices,
which are available for energy efficiency measures in
the real estate sector, in terms of their economical
appropriate use (Jakob, 2006; Kobos et al., 2006). This
includes tools such as MAKRAL and RETScreen: The
International Energy Agency (2001) uses a detailed
and technology-specific linear programming approach
to connect individual energy use to macroeconomic
effects. In this way, an optimal combination of
technologies can be determined, which holds energy
service demands stable while minimizing energy
system total costs. Software-based solutions like
RETScreen provide the opportunity to compare
different technologies with regard to their energy
generation, savings, life cycle costs, emission
reduction, financial figures and technology-specific
risks.
Against this background, it is not our goal, to
correctly determine the actual amount of the EEG in
the real estate sector, nor to quantify or integrate the
effects of market failures (like the above-mentioned
energy price distortions, information problems
including principle-agent problems between landlord
and tenant, transaction costs, capital market
imperfections) or behavioral failures (Gillingham et
al., 2009), nor to develop a selection algorithm to
classify or rank specific technical devices in terms of
their ability to reduce the EEG. Rather, we first
 The “insurance effect”: How to increase the investment amount in green buildings
develop a rudimentary optimization model, by which
we quantified an “optimal investment amount in green
buildings” in the setting of rented buildings and
building portfolios. In the context of focused
abstraction – and without loss of generality – only a
potential trade-off between the two factors “ecologic
potential savings” and “economic investment amount”
(necessary to attain the first factor) is modelled. On the
basis of this basic model, we want to formally derive
the impact of the so-called „insurance effect” in the
second part of this paper. The insurance effect
describes the effect that reduced energy demand leads
to reduced volatility of energy costs – an added value
to risk-averse tenants, who can use GBIO as insurance
against the impact of a possible damage event (energy
price volatility) by paying a premium (rent increase).
Correctly integrating this effect an increase of the
initial optimal investment amount in green buildings
can be derived. We would like to point out that factors
such as the location of the property, its age, the
original level of its technical equipment, etc. are
consciously disregarded in our model-theoretic
considerations. These factors, in reality, play an
important role in determining the optimal level of
investment in EEG. However, our paper primarily
focuses on the “insurance effects” and its
consequences on the investment level.
Since the “insurance effect” can be worked out
even without additional factors (such as location, etc.),
we have decided to abstract from these effects. In the
main part of the paper, we will show that it is possible
to reduce the EEG if this insurance effect is taken into
account, which in fact is a novelty in the scientific
discussion of the EEG. It should be noted that the
further use of the term “optimal” only refers to the
model-theoretic considerations outlined in this paper,
as potential market and behavioral failures (already
discussed in the first part of the introduction) remain
unattended. Against this background, the ‘model-
theoretic optimum’ differs from the ‘real’ optimal
level of investments in green buildings. Since the
existence of an EEG can be regarded as proved and
since the gap is likely to be even greater if the already
discussed failures are included, the model-theoretic
optimality considerations lead in the right direction
(i.e. towards the ‘real live optimum’).
Existing literature, which refers to EEG in the
context of the real estate sector, only sporadically
addressed uncertainty and the associated risks to the
EEG. If uncertainty/risk is focused, it is for example
related to uncertainty about future government
intervention (such as support programs and subsidies
or stricter regulations to internalize external effects)
(Hirst and Brown, 1990), uncertainty about the future
energy price (Bristow and Kennedy, 2010), weather-
related uncertainty effecting different amounts of
future energy use (Wang et al., 2012), uncertainty on
the energy usage patterns of tenants (Silva and Ghisi,
2014) or uncertainty about the actual long-term impact
of new (and unproven) devices (Mills et al., 2006).
In existing publications, often one of these
risks is considered in detail and quantified, but without
model-based referring to its effective implication.
Furthermore, most existing literature does not address,
that investments in the real estate area are – caused by
the capital intensity for the landlord – characterized by
a specific type of risks, which prohibits conventional
beta pricing. Thus, an isolated consideration of the
investment’s idiosyncratic risk – on the basis of
established capital market-theoretical methods – is not
possible.
We will focus in detail on this issue and its
impact on our modelling approach in section 3.1.1. If
risk is discussed in existing publications, it is often
restricted to the statement, that due to risks, less
measures to increase energy efficiency are useful or
can be performed (Hirst and Brown, 1990; Szabó et al.
2010; Thompson, 1997); an increase of risk leads – so
it is concluded – (analogous to other market
imperfections) to an increase of the EEG in the real
estate sector.
Extending our above-mentioned basic model,
we will be showing in the second part of the paper,
that, as a result of uncertainty about the future energy
price and unlike the above-mentioned common
integration of risks, more measures to increase energy
efficiency are meaningful. By working out this
insurance effect, which is new and so far unconsidered
both in scientific discourse and business, a
contribution to the future reduction of the EEG in real
estate sector can thus be brought off.
We believe that the capability of energy
efficiency investments to work as an insurance against
energy price volatility is a comprehensible benefit,
which should be used by means of awareness-raising
and informative actions, which are needed to
accomplish behavior / social change (Vedung, 1998).
As a consequence, our approach may contribute to
counteract the energy efficiency gap by overcoming
associated barriers like uncertainty and, first and
foremost, information gaps (Hirst and Brown, 1990).
To illustrate our approach, the paper is structured as
follows: In section 2, we develop our formal valuation
model with regard to specific characteristics of GBIO
and determine the optimal economic investment
amount under certainty in section 3. In section 4, we
show the impact of the insurance effect when the
landlord has a real estate portfolio with several
tenants. We use an example to illustrate the model’s
practical applicability in section 5 and conclude with
a summary of results as well as directions for future
research in section 6.
We would like to point out that we already used
some of the assumptions and the modeling approach
of section 2 and 3 in an earlier version of the paper
(Buhl et al., 2011). This paper builds upon our
preliminary groundwork and extends it significantly
with regard to the modeling of property portfolios
(section 4), the demonstration of the model’s practical
applicability (section 5) and inferable results (section
6). Furthermore, this significantly extended version of
the paper contains a comprehensive review of existing
literature regarding the energy paradox, EEG
investments and uncertainty (section 1).
1601
 Buhl et al./Environmental Engineering and Management Journal 17 (2018), 7, 1599-1611
2. Model development
As already outlined in the previous section, it
is our main goal to derive the impact of the so called
“insurance effect” on the investment level in rented
green buildings. To achieve this, we will only derive a
rudimentary optimization model in the following
section. By implication, we do not focus on the
potential market or behavioral failures already
outlined in the introduction. To set up our optimization
model and to work out the insurance effect in the
second part, only the following characteristics of
green buildings are required:
Characteristic 1: Achievable energy cost
savings depend on a property’s energy demand (that
can be reduced by green building investment
opportunities [GBIO]) on one hand, and on the
prevailing price for energy on the other hand. Thus,
the benefits of GBIO depend on future energy prices
(Atkinson et al., 2009).
Characteristic 2: GBIO reduce energy costs
and, at the same time, volatility, which originates from
the volatile energy prices. As we will point out, GBIO
reduce risk and operate like insurance. Thus, the
profitability of GBIO depends on the volatility of
energy prices (Thompson, 1997) and the insurance
effect.
Characteristic 3: If a portfolio of properties is
considered, the advantage of GBIO also depends on
energy price correlations, which affect the overall
portfolio risk.
Although more special characteristics of GBIO
might take effect as investment drivers (cp. the
discussion of market failures like energy price
distortions, information problems including principle-
agent problems between landlord and tenant,
transaction costs, capital market imperfections or
behavioral failures in the introduction), the focus of
this paper is on these three characteristics only.
According to the model setting of the present paper,
an investment in green building technology is planned
to be performed in a first step. Depending on the
investment amount, the question arises, which single
measure (or which set of measures) is most suitable to
increase energy performance of the building from a
purely technical perspective.
As it is elaborated in the main part of the paper,
the application of these technical measures also
increases the resale value of the building. In turn, the
improved energy performance leads to a reduction of
the future energy consumption of the building. From
an ecological perspective, this results in a reduction of
damage induced by the consumption of energy.
Although the negative impact of energy consumption
relates to many different dimensions (e.g.
consumption of finite energy sources, environmental
pollution, climate change), it is commonly normalized
and aggregated to CO2 equivalents (CO2-eq).
Therefore, it stands for reason to measure ecological
effects in terms of CO2-eq emission reductions in the
context of our paper.
1602
We do not consider other possible ecological
impacts of GBIO in the following. In addition to these
environmental aspects, reduced power consumption
leads to a reduction in energy costs for the users of the
property. In addition, the insurance effect can be
observed. Its implications for landlord and tenant and
their willingness to conduct the respective investment
in green buildings is the main focus of our work and
will be discussed in detail in chapter 5. Since it is the
aim of the present article to examine ecological and
economic impacts of investments in green building
technologies, we will consciously not consider or
evaluate technical performance criteria of single
technical measures, which are available to improve the
energy performance of buildings.
Thus, it is not the aim of the paper to examine
the (dis)advantages of possible individual measures
from a technical perspective, nor to make an optimal
decision selection from various available
technologies. Fig. 1 may help to summarize and
clarify our model setting: As we will point out,
available approaches in literature are appropriate to
address the listed characteristics, but do not focus on
the risk reducing insurance effect of GBIO. We
believe that this insurance effect can be a
comprehensible figure, which can be used to inform
decision makers about the (thereby increasing)
benefits of GBIO. Vyas and Cannon (2008) state,
“The current information stream has become polluted
with advocacy and lobbying rather than useful
metrics.” In this vein, the “major challenge to make
home buyers aware of the advantages of clean
technologies and to inform them about the exact
consequences of adoption” is difficult to master
(Pinkse and Dommisse, 2009).
In order to meet this challenge, this paper will
develop a bottom-up model putting special emphasis
on the insurance effect of GBIO. Based on our model
setting (see Fig. 1) we will therefore derive purposeful
assumptions to answer the following research
question: How can GBIO in rented properties and
property portfolios be correctly evaluated and how
does this affect the valuation of the ecological
contribution of the real estate sector and contribute to
overcome the energy efficiency gap?
Thus, we aim at providing an awareness-
raising and applicable instrument for consumers in
order to accomplish social change leading to higher
investment rates in GBIO. Since this paper focuses
only on investment valuation, we do not consider the
resolution of the landlord-tenant dilemma (for this
purpose, see, e.g., Schleich and Gruber, 2008). In
order to incentivize technology adoption, bottom-up
models evaluate the benefits of energy efficiency
technologies. Bottom-up models are disaggregated
representations of the energy-economy system
considered of “prime importance to support the most
suitable design of policies by assessing whether they
are capable of achieving the impacts that would justify
their implementation” (Mundaca et al., 2010).
 The “insurance effect”: How to increase the investment amount in green buildings
Fig. 1. Investment in green buildings and its economic and ecological results
However, bottom-up models are also criticized
to lack the ability to portray both broader
macroeconomic effects (Rivers and Jaccard, 2006)
and microeconomic decision-making by businesses
and consumers when selecting technologies. In this
vein, it has long been argued that bottom-up models
provide only “an unrealistic portrait of microeconomic
decision-making frameworks for technology choice”
(Mundaca et al., 2010). Despite the criticism, bottom-
up models can be used to describe current and
prospective competition of energy technologies in
detail on the supply and on the demand side. In fact,
also the aforementioned McKinsey study (McKinsey
Global Institute, 2007) is based on a bottom-up
approach arguing that macroeconomic phenomena are
“really nothing more than the sum of demand from
hundreds of microeconomic sectors”.
Thus, bottom-up models can provide valuable
insights in radically different technology futures and
enable decision making which may help to take full
advantage of the benefits of energy efficiency
technologies in general, and GBIO in particular
(Hourcade et al., 2006). Gallinelli (2008) uses a
discounted-cash-flow approach, which discounts an
investment’s expected future cash flows using a
discount rate to reach a present value. In this
connection, Keown et al. (1998) recognized the
particular importance of risk for energy-efficiency
investments and suggested a basic approach to
integrate risk into a discounted-cash-flow valuation.
They suggested the adaption of the discount rate
depending on the height of the risk. However, Johnson
(1994) realizes that the use of the adapted discount rate
is broadly discussed in literature, but is not a satisfying
way to take risk into consideration adequately.
2.1. Key rationale
The potential of green buildings to lower
carbon emissions and energy consumption is
estimated to be 35% (WBCSD, 2007). In order to
achieve this amount, all possible GBIO have to be
implemented. This might be advantageous from an
ecological perspective, but not necessarily from an
economic perspective. For a meaningful combination
of ecological and economic perspectives (cp. Figure
1), we develop a quantitative optimization model to
identify the economical optimal investment amount.
Investing this amount of money will lead to
economical optimal and at least ecologically
advantageous results. In the framework of our model
the optimal (private) investment amount in green
buildings is derived. This single, private investment
simultaneously contributes to reach the overall
economic optimum while being ecologically
advantageous.
2.2 Main definitions and assumptions
We consider a setting with a property being
rented and used by a tenant from time T0 until T1. The
tenant pays a constant periodic basic rental charge
(excluding energy costs) at the specific amount RC to
the landlord. Furthermore, the tenant has to pay energy
costs EC to a gas/electricity supply company, which is
necessary for the property’s operation. These energy
costs are the product of the property’s energy demand
d and the effective energy price at time t, P(t).
Additional expenses like, for example, expenses for
water supply are irrelevant for this analysis and are not
considered. The landlord receives the periodic basic
rental charge RC from the tenant. At the end of the
letting in T1, the landlord sells the building and
receives the resale return RR. The amount of RR can
be considered the net present value of all future
achievable rental charges (as already motivated in the
introduction, the value of the land and other factors
like location, condition or technical equipment of the
building shall be disregarded at this point shall be
disregarded at this point). Hence, it is irrelevant for our
consideration whether the property is actually sold. In
the following account, we assume the resale of the
property in T1 for the sake of simplicity. The
property’s energy demand can be reduced with the
help of a green building investment, which is
determined by the amount of its necessary investment
payout I, with . However, many energy-
saving measures of green buildings require energy
themselves. The net present value of these costs, as
well as further costs (e.g., for possible breakdown) is
integrated in I. The property’s energy costs can be
lowered permanently through GBIO to a level of
ECafter < ECbefore by reducing the energy demand from
dbefore to dafter. Moreover, the resale return RR can be
1603
 Buhl et al./Environmental Engineering and Management Journal 17 (2018), 7, 1599-1611
raised to RRafter > RRbefore. This relationship is verified
by an empirical survey, assuming that the demand for
energy-efficient properties is on the rise because of the
expected long-term increase in energy prices.
Consequently, increased resale returns can be realized
(Bienert, 2009). For our model, we assume a
diminishing marginal resale return.
The necessary investment payout occurs in T0
and has to be paid completely by the landlord initially,
although the landlord has the possibility of turning
over a certain portion of the investment payout to the
tenant (see next section). On account of this, the basic
rental charge rises to RCafter > RCbefore.
Thus, GBIO generate economic benefits for
both tenant (energy cost reduction) and landlord
(resale return increase). However, these benefits are
not necessarily equally high. Often, energy-efficiency
investments are not made because of the above-
mentioned landlord-tenant dilemma (The Climate
Group, 2008), that is, the landlord decides on the
investment amount, but only the tenant benefits from
the resulting energy cost savings. In contrast, tenants
do not have the right to claim GBIO from the landlord.
Since this paper focuses on the derivation of the
“insurance effect”, we can consider the landlord and
tenant as one and do not analyze the division of the
investment amount. At this point, game theory can
provide some important directions for research, for
example, as done by Bengtsson (1998) and also the
WBCSD (2009) drill on this in their simulation study.
To proceed, the following assumptions are considered
necessary for further analysis:
A.1: Landlord and tenant decide on the
investment amount together and evaluate GBIO
according to their willingness to pay.
A.2:
Landlord and tenant calculate with
the identical risk-free discount rate i.
A.3: The energy price increases in the
long run.
Fig. 2. Impact of GBIO on the considered cash flows
1604
2.3 Objective function
Energy prices have increased sharply during
the last few decades. The price for light fuel oil
increased from ~8€ ct/l in 1970 to 80€ ct/l in 2008 (i.e.,
~6.2% p.a.). Furthermore, the world population will
continue to rise (Tucker, 2007), and the consumption
level of many nations will approach Western
standards. Hence, we can forecast a rising energy
demand. Because of the inevitable excess demand, we
can assume exponentially rising energy costs in the
future. This forecast is supported by Buhl and Jetter
(2009), who stated that the price of each non-
renewable resource, – depending on the specific
availability and demand, rises exponentially. One
possibility to formalize the exponential rise of the
considered energy price P(t) is P(t ) = P0 ⋅ (1 + r ) t .
Here, P0 is the energy price in its initial state in T0. The
parameter r is the periodical growth rate of the energy
price compared to the previous period. The time
dependence of the energy price is implied by the
exponent t. We obtain the objective function through
pretest-posttest study, that is, by comparing all cash
flows of the landlord-tenant unit after GBIO, CFafter,
to all cash flows before any investment, CFbefore, and
maximizing the (positive) difference between the two.
This maximal difference is obtained through
optimization of the objective function Eq. (1).
∆CF ( I ) = CFafter ( I ) − CFbefore =
T1
∆RR( I )
= − I + ∑ ∆EC ( I ) +
(1)
=
t =1 (1 + i )T1
T1
P0 ⋅ (1 + r ) t ∆RR( I )
= − I + ∆d ( I ) ⋅ ∑
+
t =1 (1 + i ) t (1 + i )T1
 The “insurance effect”: How to increase the investment amount in green buildings
As mentioned above, the resale return rises
with the investment amount. One cannot assume that
GBIO can raise the resale return of a property linearly.
Hence, we can assume that investments’ effects
decline. Thus, we can use a strictly monotone
dRRafter d 2 RRafter
increasing ( > 0 ), concave (
dI d 2I
course of the function for RRafter (starting from the
resale return without any GBIO RRbefore). This can be
formalized exemplarily for I ≥ 1 as
RRafter ( I ) = RRbefore + s ⋅ ln I . The parameter
determines the inclination of the resale return curve.
The higher s we choose, the more GBIO increase the
building’s resale return. Hence, the achievable
increase of the resale return with GBIO is
∆RR( I ) = RRafter − RRbefore = ( RRbefore + s ⋅ ln I ) −
RRbefore = s ⋅ ln I . The second element of the objective
function describes the development of the property’s
energy demand dafter(I) depending on the investment
amount. As mentioned above, the energy demand
decreases permanently when the investment amount
dd
rises ( after < 0 ). At this point, we use a linear
dI
relation between energy demand and investment
amount in the relevant region. One possible function
for this is d after ( I ) =d before−v ⋅ I . dbefore is the property’s
energy demand in the initial state, that is, before any
investment. v determines the curve’s inclination and
equates to the marginal energy demand of the
property: If the investment amount is raised by one
monetary unit, the energy demand of the property falls
permanently by approximately v units. This
observation is in line with Christen et al. (2002), who
found that the energy-efficiency gain of GBIO has a
linear connection with the hence emerging costs. Note
that our model also allows for negative energy
demand, which can be realized with sufficiently high
investment amounts and energy generation devices,
such as solar panels. It is important that these
relationships are technology-dependent. It is self-
Fig. 2. Course of the objective function (exemplary)
evident that an investment for the integration of a
modern roof insulation material has a different impact
on a property’s energy demand and its resale return
than an investment of the same amount to integrate an
information system-based heating and air-
conditioning management system. Furthermore,
< 0) properties have individual cost functions, which are
determined by specific prevailing conditions
(Atkinson et al., 2009). We approach this problem by
using only generic functions that encompass a general
case to illustrate basic interdependencies. Our model
s
can be tailored to specific GBIO by simply adapting
the course of the functions. To sum up, we can
formalize the induced effects of GBIO in properties to
raise energy efficiency using Eq. (2).
T1
P0 ⋅ (1 + r ) t s ⋅ ln I
∆CF ( I ) = − I + v ⋅ I ⋅ ∑ + (2)
t =1 (1 + i ) t (1 + i )T1
Setting the objective function’s first derivative
to 0 and verifying the second-order condition shows
that the objective function is strictly concave in the
domain and reaches its maximum at the investment
amount.
T1 P ⋅ (1 + r ) t
s
I RiskNeutra
*
l = + v ⋅ ∑ 0 (3)
(1 + i )T1 t =1 (1 + i ) t
Consequently, it is reasonable to raise the
I*
investment amount up to RiskNeutral Eq. (3). Below this
investment amount, an elevation of the investment
sum leads to a higher resale return increase and energy
cost reduction than the necessary payout. In contrast,
the positive effects above the investment amount
I RiskNeutra
*
l , in fact, exceed the incidental payouts, but
disproportionally high capital expenditure is
necessary. These relationships are depicted in Fig. 2
(like before, we assume the potential of green
buildings to lower carbon emissions and energy
consumption to be 35% (WBCSD, 2007)).
1605
 Buhl et al./Environmental Engineering and Management Journal 17 (2018), 7, 1599-1611
3. Optimal investment amount under uncertainty
Resources and commodities are ever more
subject to speculative transactions. Investment in
commodity indexes increased by a factor of 20 from
US $13 billion in 2003 to US $260 billion in 2008
(Masters, 2008). As shown by Shiller (1981),
increased speculation and trading of commodities and
energy sources cause an increase in price volatility
(Duffie et al., 1999). Other sources of uncertainty, for
example, a volatile resale return due to market price
changes, are irrelevant for our consideration and are
thus excluded. In order to account for energy price
volatility, the following assumptions and definitions
are necessary.
3.1.1. Assumptions and definitions
For this consideration, we assume normally
distributed energy prices. As mentioned above, the
energy cost savings are composed of the product of the
energy prices at time t with the demand reduction ∆d
and are thus normally distributed, too. Then, the time-
dependent, volatile energy savings can be discounted
to an expected present value of the savings, which is
also normally distributed. Because of the volatility of
energy prices, the calculated present value of the
energy cost savings is volatile, too. Here, we interpret
the volatility of the present value as the possible
positive or negative deviation of the present value
from the expected present value of the savings. We
measure this deviation with the variance σ2 (σ2 > 0).
σE2 is the variance of the energy price. In order to
integrate measures for risk (σ2) and return (μ), we use
a preference function:
A.4: The risk-adjusted value of the
investment is determined by both parties with
Bernoulli’s theory of expected utility (Bernoulli,
1954) and the preference function Eq. (4):
α mentioned in A.4, we assume risk-averse (and
Φ(µ , σ ) = µ − ⋅σ 2
2 consequently insurance-affine) decision makers
We assume risk-averse decision makers, that
is, the present value of the investment’s cash flow is
valued lower if its variance is higher (assuming a fixed
expected value μ).
The risk-adjusted value corresponds to a
preference function, which is developed according to
established methods of decision theory and integrates
an expected value, its deviation, and the decision
maker’s risk aversion. This preference function is
based on the utility function U ( x) = −e 2αx and is
compatible to the Bernoulli principle (Bernoulli,
1954). The Arrow-Pratt characterization of absolute
risk aversion (Arrow, 1971) is with , = µ after −
modeling a risk-averse decision maker. It seems
reasonable for us to choose this approach to
adequately include energy price risks in our specific
context: While risks of other investments can usefully
be quantified by established approaches derived from
1606
capital market theory, which help to differentiate
between market and project specific risks, this is not
the case for the investor/landlord in the context of real
estate. Investments in real estate use up a very large
part of a private investor’s capital. Since it is not
possible in this case to diversify the market risk,
individual pricing of risk by means of a preference
functional appears to be appropriate. This approach is
consistent with Johnson (1994), who draws attention
to boundaries of established investment theories – like
the capital asset pricing model (CAPM) and arbitrage
pricing theory (APT) – in the context of energy
technology choices and rather proposes variations of
the net present value (NPV).
Considering energy price volatility in our
valuation, we discover a particular effect: Taking into
account the rules of linear transformation of random
variables (in our consideration, the present value of the
energy savings), the reduced energy demand results in
reduced volatility of the energy costs (Greene, 2008).
Thus, green buildings operate like insurance: By
paying a premium (rent increase), the insurance holder
(tenant) can insure himself against the impact of a
possible damage event (energy price volatility). The
tenant’s willingness to negotiate such insurance and
the amount of premium he is willing to pay depends
on his individual risk attitude:
Tenants who negotiate insurance want to avoid
(or lower) risk. They prefer to pay a certain amount of
money (here, the rent increase) rather than accept an
uncertain, more volatile payout. In this manner, we
can mentally divide the rental charge RCafter into three
parts: The first part is the basic rental charge RCbefore.
The second part is the counter value for the achievable
energy cost savings. The third part is the insurance
premium, which is the value of the achievable risk
reduction (Buhl et al., 2011). This (over all periods of
the letting) accumulated value equates to the
difference of the second part of Bernoulli’s preference
function before and after the investment. As
(4)
(Bamberg and Spremann, 1981). For them, the risk-
reducing effect of green buildings creates added value.
3.1.2. Optimization model
Valuating all cash flows and risks of the
landlord-tenant unit before and after GBIO with
Bernoulli’s preference function, we obtain the new
objective function to be optimized Eq. (5):
∆Φ ( µ , σ ) = Φ after ( I ) − Φ before =
Φ ( µ after ( I ), σ after ( I )) − Φ ( µ before , σ before ) =
α 2 α 2
⋅ σ after − ( µ before − ⋅ σ before ) = ∆Φ ( I ) =
2 2
T
P ⋅ (1 + r ) t ∆RR( I ) α
= − I + ∆d ( I ) ⋅ ∑ 0
1
+ + ⋅ ∆σ 2 ( I )
t =1 (1 + i ) t (1 + i ) T 2 1
(5)
 The “insurance effect”: How to increase the investment amount in green buildings
The first three terms are similar to Eq. (2) in
our consideration under certainty. The last term
corresponds to the volatility reduction of the energy
costs weighted with the risk-aversion parameter α and
that can be increased with the investment amount.
To sum up, we can formalize the induced
effects of GBIO in properties under consideration of
risk as shown in Eq. (6).
P0 ⋅ (1 + r ) t s ⋅ ln I α
T1
∆Φ( I ) = − I + v ⋅ I ⋅ ∑ + + ⋅
t =1 (1 + i) t (1 + i)T1 2
⋅(d 02 − (d 0 − v ⋅ I ) 2 ) ⋅ σ E2
The objective function is strictly concave in the
domain again and reaches its maximum at the
investment amount in Eq. (7).
T1
P0 ⋅ (1 + r ) t
−1 + v ⋅ ∑ + α ⋅ v ⋅ d 0 ⋅ σ E2
t =1 (1 + i ) t
I*
2 ⋅ α ⋅ v 2 ⋅ σ E2
T1
P0 ⋅ (1 + r ) t
(1 − v ⋅ ∑ + α ⋅ v ⋅ d 0 ⋅ σ E2 ) 2
t =1 (1 + i ) t
s dependencies affect the overall volatility of a property
+4 ⋅ ⋅ α ⋅ v 2 ⋅ σ E2
(1 + i )T1
+ overall variance of a portfolio is determined by
2 ⋅ α ⋅ v 2 ⋅ σ E2
By comparing the computed optimal investment
amount considering the insurance effect to the optimal
investment amount of a risk-neutral decision maker,
we can see that the optimal investment amount
considering energy price volatility is always higher
than assuming a non-volatile energy price.
Considering the energy price and the insurance effect
accurately will consequently lead to an increase of the
actual optimal investment amount. Consequently,
valuating GBIO correctly will lead to economic
decisions that will generate more benefits also from an
ecological perspective.
4. Addressing the ‘insurance-effect’ in multiple
properties
In this section, we transfer our presented
analysis to an approach with multiple properties and
analyze how the insurance effect takes effect in
building portfolios.
4.1.1. Assumptions and definitions
Now assume a landlord with n ≥ 2 properties i
and multiple tenants. The landlord and his tenants
together decide on the investment amount to be spent
on GBIO. For this, they need to determine the overall
optimal investment amount I overall
*
for the entire
property portfolio as well as the optimal allocation of
this investment amount to each property. If we
consider the properties to be supplied by different
energy sources (e.g., gas, oil etc.) and compare their
prices, we discover a special characteristic: There are
similarities in the prices, for example, for domestic
fuel oil and domestic gas in the past. Many drivers for
energy prices, such as global demand or inflation, are
present across all energy sources. As theoretical and
empirical research affirms (Da Silva, 2007), energy
prices are not independent of each other, but are
correlated, that is, they are systematically associated
with each other (Eydeland and Wolyniec, 2003). In
order to accurately determine the optimal investment
amount for a property portfolio, we have to account
for the impact of energy price dependencies on the
(6) overall risk and the manner in which it is affected by
GBIO. If energy price dependencies were disregarded,
the computed optimal investment amounts would not
account for possible diversification effects and would
consequently be false. For this reason, we want to
adapt our model and add the following assumption:
A.5: All n properties are supplied by
energy sources whose prices depend on each other.
The measurement for the energy price dependencies is
+ the correlation coefficient kij. We consider these
correlations coefficients to be given. All other possible
(7) dependencies among the properties are disregarded.
As mentioned above, energy price
portfolio. According to Keown et al. (2008), the
summing up the stand-alone risks of each property σi
and considering the dependencies between the energy
prices with the help of the correlation coefficient.
Hence, the overall energy costs’ variance of a property
portfolio (without any GBIO) is shown in Eq. (8)
n n
σ overall
2
= ∑∑ σ iσ j kij (8)
i =1 j =1
If we consider the properties being supplied by
energy sources whose prices do not show perfect
positive correlation (i.e., kij = [−1;+1[ ), we can see
that the overall property portfolio’s variance is smaller
than its variance with perfect positive correlation
( kij = +1 ). This risk-reducing effect is called the
diversification effect and can be interpreted in the
following manner: If the considered properties are
supplied by the same energy source, kij equals +1 since
an energy price is always fully positively correlated
with itself. If this energy price increases, the overall
payouts for energy will then increase in equal
proportion since energy costs only depend on one
stochastic factor. However, if the landlord and tenants
decide on properties supplied by different energy
sources, kij is likely to take on a value of k ij = [−1;+1[
, that is, the energy prices will not show perfect
positive correlation with each other. If one of the
energy prices increases, the payouts for energy will
then increase in a smaller proportion since the risk of
a possible price increase is diversified on multiple
stochastic factors. Consequently, the resulting overall
portfolio risk is lower than the aggregate stand-alone
risks of the properties.
1607
 Buhl et al./Environmental Engineering and Management Journal 17 (2018), 7, 1599-1611
Since the stand-alone risks of each property of
a portfolio codetermine the value of the portfolio’s
diversification effect, GBIO also indirectly affect the
value of the diversification effect. As a well-grounded
investment valuation has to account for all investment-
caused consequences on financial figures, we need to
consider the investment’s effect – as well as the
insurance effect – on the value of the diversification
effect in our valuation, too.
4.1.2. Optimization model
Eq. (9) is the objective function for multiple
properties accounting for all aforementioned GBIO-
caused effects:
∆Φ overall ( µ , σ ) = Φ overall , after ( I i ) − Φ overall , before =
= Φ ( µ overall , after ( I i ), σ overall , after ( I i )) −
Φ ( µ overall , before , σ overall , before ) = ∆Φ overall ( I i ) =
n n T1, i
P ⋅ (1 + ri ) t ∆RRi ( I i ) 
= −∑ I i + ∑  ∆d i ( I i ) ⋅ ∑ 0, i + +
(1 + i ) 1, i 
T
i =1 i =1  t =1 (1 + i ) t
α  n n 2  The duration of the letting is 30 years (T1 = 30). The
⋅ ∆ ∑∑ σ i ( I i )σ 2j ( I j )kij 
2  i =1 j =1  property after 30 years and calculates (like their
(9)
This function consists of the following terms:
The negative sum of the investment amounts in each
property (first term) and the sum of the achievable
energy cost savings and resale return increases for
each property (second term). Additionally, the third
term considers the individual risk reductions for each
property as well as the investments’ effect on the value
of the diversification effect.
In order to obtain the overall optimal
investment amount as well as the individual optimal
investment amounts for each property, this objective
function has to be optimized. At its maximum, the
necessary (but not sufficient) conditions shown in Eq.
(10) are fulfilled.
(10)
For the verification of the second-order
condition, we test the Hessian matrix to be negatively
definite in the domain. Since the multiple unknown
decision variables cannot always be separated, an
explicit solution for the objective function cannot
always be determined. However, a numerical solution
of the optimization model at hand with a limited
number of properties can be computed in most cases.
The authors made several model calculations
and compared the calculation results of the presented
model to results of calculations disregarding the
impact of diversification. In this manner, we
ascertained that the overall optimal investment
amount of a portfolio under consideration for the
diversification effect can be higher or lower than the
optimal investment amount disregarding price
dependencies. If the diversification effect is raised by
1608
GBIO, another potential of green buildings becomes
visible: In certain cases, GBIO positively affect the
value of the diversification effect and are consequently
able to reduce the overall risk in a property portfolio
in two ways: reducing the stand-alone risks of each
property due to energy demand reduction and reducing
the portfolio risk due to a positively affected
diversification effect. Consequently, a correctly
conducted economic valuation will lead again to an
increased overall optimal investment amount and to a
higher ecological contribution to energy efficiency.
Despite not computing higher overall optimal
investment amounts in all cases, it is evident that the
high economic and resulting ecological potential of
GBIO only becomes visible, if risk is considered
correctly.
5. Example of potential application and results
A building society wants to design a block of
offices in an energy-efficient manner. Therefore, the
company wants to exploit the high potential of GBIO.
For this example, we consider the following situation:
company receives a one-time resale return for the
tenants) with a discount rate of i = 3%. The resale
return can be raised by an energy-efficiency
investment, beginning from an amount of 1,000,000
monetary units (MU) in a form that can be described
by the following function:
RRafter ( I ) = 1000000 + 10000 ⋅ ln I . The property’s
demand for domestic fuel oil can be lowered by
approximately 0.06 l p.a. with each invested MU
starting from a basic demand of 3000 l p.a. Hence, we
use the following relationship for the energy demand:
d after ( I ) = 3000 − 0.06 ⋅ I . We assume an energy
price increase of 7% p.a. with an initial price of P0 =
0.85 MU/l. We determined the optimal investment
amount to be 4,604 MU. Hence, the company should
invest exactly this amount to maximize its economic
profit and thus contribute to energy efficiency.
Investing more than the computed optimal amount is
advantageous from an ecological perspective, but not
from an economic perspective. In order to account for
risk and determine the optimal investment amount
from an economic viewpoint under uncertainty, we
simply assume the energy price to be volatile (σE2 =
0,006). For the parameter of risk aversion, we assume
α = 1. All input values are summarized in Table 1.
Considering risk and the insurance impact of
green buildings, we determine an optimal investment
amount of I* = 18,748 MU. Considering the insurance
impact clearly raises the optimal investment amount
and energy demand, and thus the carbon footprint can
be reduced much more. The company in this example
will save 37.5% of its periodic energy demand if it
chooses the optimal investment alternative, whereas
the company would only save 9.2% of energy if it
disregarded the insurance impact and conducted a
valuation under certainty.
 The “insurance effect”: How to increase the investment amount in green buildings
Table 1. Input values for the example with one property
T1 =30 i=3 % r=7 %
P0=0,85 MU/l σE 2=0,006 α=1
Table 2. Input values for the example with n=2 properties
T1,1=30 α=1 n=2
T1,2=30 k12=0,5 i=3 %
P0,E=0,85 MU/l σE2=0,006 rE=7 %
P0,F=0,6 MU/u σF2=0,009 rF=5%
Imagine the building society together with their
tenants planning to invest in GBIO for n = 2
properties. Both properties are under possession of the
building society and used by their tenants. Property 1
is the familiar block of offices. All given input values
remain unmodified. Property 2 is a production site for
commercial uses. All input values given are
summarized in Table 2.
We can determine an optimal investment
amount for property 1 of I1* = 41,064 MU and for
property 2 of I 2* = 74,529 MU. By summing up the
two optimal investment amounts, we obtain the
overall optimal investment amount I overall
*
= 115,593
MU. Compared to a consideration without price
dependencies, we can observe significantly different
results: In the current example, the overall optimal
investment amount is only 65,121 MU if energy price
dependencies are disregarded. Here, the consideration
of energy price dependencies leads to a further
increase in the optimal overall investment amount.
Hence, we could model-based show that an accurate
economic valuation raises the ecological contribution
again.
As already stated in the introducing section, it
is not the aim of this fictional example, to determine
the optimum level of investment in devices to reduce
energy consumption in green buildings (for which we
once again refer to the application of established
tools). Instead, we focus on deriving the relevant
cause-effect relationships particularly resulting from
the effect insurance: If this – as shown in our
numerical example – is taken into consideration, there
is an increase in the optimal level of investment.
6. Conclusions
In the first part of this paper, a rudimentary
optimization model is developed to determine the
optimal investment amount in green building
investment opportunities (GBIO). We also show that
investments in green buildings are appropriate up to a
certain level (which can be identified, if the presented
model is applied) from an economic as well as from
an ecological perspective. Without focusing on an
exact quantification of the EEG and potential market
or behavioral failures, it is also shown that the (model
RRafter ( I ) = 1000000 + 10000 ⋅ ln I
d after ( I ) = 3000 − 0,06 ⋅ I
RRafter ,1 ( I1 ) = 1000000 + 10000 ⋅ ln I1
RRafter , 2 ( I 2 ) = 3000000 + 30000 ⋅ ln I 2
d after ,1 ( I1 ) = 3000 − 0,06 ⋅ I1
d after , 2 ( I 2 ) = 5000 − 0,03 ⋅ I 2
theoretic) optimal investment amount can be increased
by considering the so called “insurance effect” of
green buildings. The model-theoretic conception of
this effect (and the impact on the EEG) is a novelty in
the scientific literature, which is examining the energy
efficiency gap in the real estate sector. Furthermore,
we pointed out that a diversification effect is a further
possible benefit of GBIO in property portfolios. Of
course, a theoretical model gives only limited
information about real world impacts. Thus, it cannot
be concluded with certainty that planning tools and
businesses considering the insurance effect lead to a
practical reduction of the EEG in every case. The
practical effectiveness of the insurance effect must be
proven in real world settings, for example in pilot test
versions of planning tools which may give insight
about user acceptance and the actual EEG impact of
GBIO investments considering the insurance effect.
Moreover, several assumptions and conditions
of this paper have to be examined critically, which
opens up possibilities for further research. First, the
landlord-tenant dilemma might be an investment
barrier and identified economical and ecologically
meaningful investments are not being made because
landlord and tenant are not willing to share the
investment amount.
Furthermore, existing legislation can impede
reasonable investments if the investment amount
cannot be divided equitably (e.g., German landlords
can only pass on 11% p.a. of the costs of certain
energy-efficient refurbishment measures to the
tenant). In that case, it is possible that the investment
amount to be spent is limited. Second, it is obvious that
the value (and thus the resale return) of properties
depends on many more factors than the investment
amount of GBIO and energy price dependencies (e.g.
location, condition, technical equipment). A model
integrating these factors should be the subject of
further research.
References
Abadie L.M., Ortiza R.A., Galarraga I., (2012),
Determinants of energy efficiency investments in the
US, Energy Policy, 45, 551-566.
Allcott H., Greenstone M., (2012), Is There an Energy
Efficiency Gap?, Journal of Economic Perspectives,
26, 3-28.
1609
 Buhl et al./Environmental Engineering and Management Journal 17 (2018), 7, 1599-1611
Apgar M., (2009), What every leader should know about real
estate, Harvard Business Review, 10, 100-107.
Arrow K.J., (1971), Essays in the Theory of Risk Bearing,
Mark-ham Publishing, Chicago.
Atkinson J.G.B., Jackson T., Mullings-Smith E., (2009),
Market influence on the low carbon energy
refurbishment of existing multi-residential buildings,
Energy Policy, 37, 2582-2593.
Bamberg G., Spremann K., (1981), Implications of constant
risk aversion, Mathematical Methods of Operations
Research, 25, 205-224.
Bengtsson B., (1998), Tenants’ dilemma – On collective
action in housing, Housing Studies, 13, 99-120.
Bernoulli D., (1954), Exposition of a new theory on the
measurement of risk, Econometrica, 22, 23-36.
Bristow D., Kennedy C.A., (2010), Potential of building-
scale alternative energy to alleviate risk from the future
price of energy, Energy Policy, 38, 1885-1894.
Brown, M.A., (2001), Market failures and barriers as a basis
for clean energy policies, Energy policy, 29, 1197-
1207.
Buhl H.U., Gaugler T., Mette P., (2011), Determining the
Optimal Investment Amount of an Intel-ligent House-
Potentials of Information and Technology to Combine
Ecology and Economy, Proc. of the 19th European
Conference on Information Systems, Helsinki, Finland.
Buhl H.U., Jetter M., (2009), BISE’s responsibility for our
planet, Business and Information Systems Engineering,
1, 273-276.
Christen K., Jakob M., Jochem E., (2002), Marginal cost in
residential buildings with enforced energy efficiency
measures (Grenzkosten bei forcierten Energie-
Effizienzmassnahmen in Wohngebäuden), On line at:
http://www.cepe.ch/research/projects/grenzkosten/Gre
nzkosten_Schlussbericht_Teil1.pdf.
Clinch J.P., Healy J.D., (2001), Cost-benefit analysis of
domestic energy efficiency, Energy Policy, 29, 113-
124.
Cullen J.M., Allwood J.M., (2010), Theoretical efficiency
limits for energy conversion devices, Energy, 35, 2059-
2069.
Da Silva P.P., (2007), EU spot prices and industry structure:
assessing electricity market integration, International
Journal of Energy Sector Management, 2, 340-350.
Dixon R.K., McGowan E., Onysko G., Scheer, R.M.,
(2010), US energy conservation and efficiency policies:
Challenges and opportunities, Energy Policy, 38, 6398-
6408.
Duffie D., Gray S.F., Hoang P., (1999), Volatility in Energy
Prices, In: Managing Energy Price Risk, Risk
Publications, London, UK.
Edwards B.W., Naboni E., (2013), Green Buildings Pay.
Design, Productivity and Ecology, Routledge Chapman
& Hall, London.
European Union, (2011), A Roadmap for moving to a
competitive low carbon economy in 2050, On line at:
http://eur-lex.europa.eu/legal-
content/EN/TXT/?uri=CELEX:52011DC0112.
Eydeland A., Wolyniec K., (2003), Energy and power risk
management – New developments in modeling, pricing,
and hedging, John Wiley & Sons, Inc., New Jersey.
Gallinelli F., (2008), Mastering real estate investment:
Examples, metrics and case studies, Real-data Inc.,
Southport, UK.
Gillingham K., Newell R., Palmer K., (2009), Energy
efficiency policies: A Retrospective Examination,
Annual Review of Environment and Resources, 31, 161-
92.
1610
Golove W.H., Eto J.H., (1996), Market Barriers to Energy
Efficiency: A Critical Reappraisal of the Rationale for
Public Policies to Promote Energy Efficiency,
Lawrence Berkeley National Labora-tory, CA, United
States.
Greene W.H., (2008), Econometric Analysis, Prentice Hall,
New Jersey.
Hein, L.G., Blok, K., (1995), Transaction costs of energy
efficiency improvement, The energy efficiency
challenge for Europe. ECEEE summer study,
Mandelieu, France.
Hirst E., Brown M., (1990), Closing the efficiency gap:
barriers to the efficient use of energy, Resources,
Conservation and Recycling, 3, 267-281.
Hourcade J.C., Jaccard M., Bataille C., Ghersi F., (2006),
Hybrid modeling: new answers to old challenges, The
Energy Journal, 2, 1-12.
Hull D., Ó Gallachóir B.P., Walkerd N., (2009),
Development of a modelling framework in re-sponse to
new European energy-efficiency regulatory
obligations: The Irish experience, Energy Policy, 37,
5363-5375.
International Energy Agency, (2001), Energy Technology
Systems Analysis Program, On line at: http://www.iea-
etsap.org/web/Markal.asp.
Jaffe A. B., Richard G.N., Stavins R.N., (2005), A tale of
two market failures: Technology and environmental
policy, Ecological Economics, 54, 164-174.
Jaffe A.B., Stavins R.N., (1994), The energy-efficiency gap
– What does it mean?, Energy Policy, 22, 804-810.
Jakob M., (2006), Marginal costs and co-benefits of energy
efficiency investments: The case of the Swiss
residential sector, Energy Policy, 34, 172-187.
Johnson B.E., (1994), Modeling energy technology choices:
Which investment tools are appropriate?, Energy
Policy, 22, 877-883.
Keown A.J., Scott D.F., Martin J.D., Petty J.W., (2008),
Foundations of Finance: The Logic and Practice of
Financial Management, Prentice Hall, New Jersey.
Kibert C.J., (2014), Sustainable Construction: Green
Building Design and Delivery, John Wiley & Sons,
Hoboken, NJ.
Klemick H., Wolverton A., (2013), Energy Efficiency Gap,
Encyclopedia of Energy, Natural Resource, and
Environmental Economics, Elsevier, Oxford, 74-81.
Kobos P.H., Erickson J.D., Drennen T.E., (2006),
Technological learning and renewable energy costs:
implications for US renewable energy policy, Energy
Policy, 34, 1645-1658.
Koopmans C.C., te Velde D.W., (2001), Bridging the energy
efficiency gap: using bottom-up information in a top-
down energy demand model, Energy Economics, 23,
57-75.
Levine M.D., Hirst E., Koomey J.G., McMahon J.E.,
Sanstad A.H., (1994), Energy Efficiency, Market
Failures, and Government Policy, Lawrence Berkeley
National Laboratory, CA, United States.
Masters M.W., (2008), Testimony before the committee on
homeland security and governmental affairs United
States senate, On line at:
http://hsgac.senate.gov/public/_files/052008Masters.p
df.
McKinsey Global Institute, (2007), Curbing Global Energy
Demand Growth: The Energy Productivity
Opportunity, MGI, San Francisco.
Mills E., Kromer S., Weiss G., Mathew, P.A., (2006), From
volatility to value: analyzing and managing financial
 The “insurance effect”: How to increase the investment amount in green buildings
and performance risk in energy savings projects,
Energy Policy, 34, 188-199.
Mundaca L., Neij L., Worrell E., McNeil, M., (2010),
Evaluating energy efficiency policies with energy-
economy models, Annual Review of Environment and
Resources, 35, 305-344.
Myers E., (2014), Asymmetric Information in Residential
Rental Markets: Implications for the Energy Efficiency
Gap (working paper), On line at:
http://areweb.berkeley.edu/candidate/sites/areweb.berk
eley.edu/files/Asymmetric%20Information%20in%20
Residential%20Rental%20Markets%20Implications%
20for%20the%20Energy%20Efficiency%20Gap_0.pdf
Patterson M.G., (1996), What is energy efficiency? –
Concepts, indicators and methodological issues, Energy
Policy, 24, 377-390.
Pinkse J., Dommisse M., (2009), Overcoming barriers to
sustainability: an explanation of residen-tial builders'
reluctance to adopt clean technologies, Business
Strategy and the Environment, 18, 515-527.
Rottke N.B., (2009), Economy vs. Ecology – Sustainability
in the Real Estate Industry, Immobilien Manager,
Cologne, Germany.
Schleich J., Gruber E., (2008), Beyond case studies: Barriers
to energy efficiency in commerce and the service
sector, Energy Economics, 30, 449-464.
Shiller R.J., (1981), Do stock prices move too much to be
justified by subsequent changes in dividends?,
American Economic Review, 71, 421-436.
Shove E., (1999), Gaps, barriers and conceptual chasms:
theories of technology transfer and energy in buildings,
Energy Policy, 26, 1105-1112.
Silva A.S., Ghisi, E., (2014), Uncertainty analysis of the
computer model in building performance simulation,
Energy and Buildings, 76, 258-269.
Szabó S., Jäger-Waldau A., Szabó L., (2010), Risk adjusted
financial costs of photovoltaics, Energy Policy, 38,
3807-3819.
Teng J., Wu X., (2014), Eco-footprint-based life-cycle eco-
efficiency assessment of building pro-jects, Ecological
Indicators, 39, 160-168.
The Climate Group, (2008), Smart 2020: Enabling the Low
Carbon Economy in the Information Age, On line at:
http://www.smart2020.org/_assets/files/02_Smart2020
Report.pdf.
Thompson P.B., (1997), Evaluating energy efficiency
investments: accounting for risk in the dis-counting
process, Energy Policy, 25, 989-996.
Tucker P., (2007), World population forecast rises, Futurist,
41, 1-12.
Vedung E., (1998), Policy instruments: typologies and
theories. Carrots, Sticks and Sermons, Transaction,
New Brunswick.
Vyas U.K., Cannon S.E., (2008), Shifting the sustainability
paradigm: From advocacy to good business, Real
Estate Issues, 33, 1-7.
Wang L., Mathew P., Pang X., (2012), Uncertainties in
energy consumption introduced by building operations
and weather for a medium-size office building, Energy
and Buildings, 53, 152-158.
Wilson C., Dowlatabadi H., (2007), Models of decision
making and residential energy use, Annual Review of
Environment and Resources, 32, 169-203.
WBCSD, (2007), World Business Council for Sustainable
Developmen, Mortgage Banking, On line at:
https://www.wbcsd.org.
WBCSD, (2009), World Business Council for Sustainable
Development, Energy efficiency in buildings -
Transforming the market, On line at:
http://www.wbcsd.org/transformingthemarketeeb.aspx
.
1611
1612
 Environmental Engineering and Management Journal July 2018, Vol.17, No. 7, 1613-1620
http://www.eemj.icpm.tuiasi.ro/; http://www.eemj.eu

“Gheorghe Asachi” Technical University of Iasi, Romania

ADSORPTION BEHAVIOR OF GADOLINIUM(III)

FROM AQUEOUS SOLUTION BY D418 RESIN

Leilei Pi1, Chunhua Xiong1∗, Jianxiong Jiang2, Xuming Zheng3, Rongrong Zhao1,
Caiping Yao1, Qing Chen1, Yaqin Fu4, Yaofeng Zhu4
1Department of Applied Chemistry, Zhejiang Gongshang University, No.149 Jiaogong Road,
Hangzhou City, Zhejiang Province, 310012, China
2Key Laboratory of Organosilicon Chemistry and Material Technology of Ministry of Education, Hangzhou Normal University,

Hangzhou City, Zhejiang Province, 310012, China

3Engineering Research Center for Eco-Dyeing & Finishing of Textiles, Ministry of Education,

Zhejiang Sci-Tech University, Hangzhou City, Zhejiang Province, 310018, China

4
School of Materials and Textile, Zhejiang Sci-Tech University, Hangzhou City, Zhejiang Province, 310018, China

Abstract

Adsorption, desorption, and kinetics behavior of Gd(III) ion on chelating resin were investigated in this paper. Batch adsorption
studies were carried out with different pH, contact time, and temperature. The adsorption of Gd(III) follows the Langmuir model
better than the Freundlich model. The apparent activation energy Ea and adsorption rate constant are k308 4.16 kJ/mol and 6.01×10-
5 s-1, respectively. Thermodynamic parameters such as ∆H, ∆S and ∆G indicated that Gd(III) ion adsorption by D418 resin was an

endothermic and spontaneous process in nature. Column adsorption experiments indicated the maximum dynamically adsorption
capacity of 281mg/g for Gd(III). The desorption rate of Gd(III) was 98.3% when the elution agent is 2.0 mol/L HCl solution and
the chelating resin can be regenerated and reused. Analysis of IR spectroscopy for D418 resin before and after adsorption of Gd(III)
ions and after desorption indicated the attendance of coordination and good regenerative ability of resin. Thermo gravimetric
analysis for D418 resin before and after adsorption of Gd(III) ions proved the adsorption of Gd(III) ions.

Keywords: adsorption, desorption, D418 resin, gadolinium

Received: July, 2013; Revised final: September, 2014; Accepted: September, 2014; Published in final edited form: July 2018

1. Introduction Gadolinium oxide, one of its largest uses in nuclear

The development of new technology makes
higher demands on various functional materials. Rare
earth elements and their chemical compounds have
their peculiar optical, electric, magnetic, acoustic,
thermal, mechanical properties and mutual
transformations among them. Therefore, they play
special significant role and occupy very important
position in the science and technology of modern
materials (Alexandratos, 2000; Awual et al., 2013a;
Chen and Zhao, 2003). Gadolinium, a soft silvery
metallic element, is found in the lanthanide series of
inner transition metals of the periodic table.
power reactors is as shielding and fluxing devices.
Gd(III) ion, known as fission product, can leak to the
environment and cause risks. Consequently, it is
considered as hazardous material (El-Sofany, 2008;
Junk et al., 1974).
For years, many methods have been developed
to remove rare earth metals from the effluents. Several
techniques, such as ion exchange, chemical
precipitation, membrane separation, and extraction
chromatography have been reported for the adsorption
and separation of rare earth elements from aqueous
solution (Awual et al., 2013b; Kondo and Kamio,
2002; Liang et al., 2008; Matlock et al., 2002). Various

∗
Author to whom all correspondence should be addressed: e-mail: xiongch@163.com, Phone: +86 13357196703
 Pi et al./Environmental Engineering and Management Journal 17 (2018), 7, 1613-1620
materials such as bamboo charcoal, zeolite,
polystyrene beads and resins have been used in these
techniques (Awual et al., 2013c; Khalil et al., 2016;
Xiong et al., 2014). Amongst those, resins were
studied as available material in industrial wastewater
because of their high removal efficiency, feasibility in
application, as well as the potential to extract toxic or
valuable metals from the waste water (Li et al., 2016;
Lin et al., 2018; Shen et al., 2016). D418 resin, as a
typical ion exchange resin, possesses the advantages
of polymeric adsorbents, such as the relatively low
cost and effectiveness in removing metal ions. D418
chelating resin is a polymeric microporous material
which contains aminophosphonic function group of [-
NHCH2PO3Na2]. It could be adopted to the adsorption
of metal ions from aqueous solutions.
In this work, study of the adsorption of Gd(III)
on D418 resin is carried out in the form of batch and
column methods. Some factors to affect the
adsorption, such as contact time, initial pH of solution,
and temperature were examined. Kinetics and
isotherm adsorption experiments were carried out, and
thermodynamic parameters of adsorption for Gd(III)
were calculated. The experimental results present a
path for the removal of Gd(III) in the environmental
protection.
2. Experimental
2.1. Materials
The chelating resin (D418) was provided by
Nankai University, China and the properties are
shown in Table1. The standard solutions of Gd(III) ion
was prepared from Gd2O3 (AR). HOAc-NaAc buffer
solution with pH 4.50-6.50 and C6H15NO3-HNO3
buffer solutions with pH 7.20 were prepared from
acetic acid (HOAc), sodium acetate (NaAc),
triethanolamine (C6H15NO3) and Nitric acid (HNO3)
solutions. The chromophoric reagent of 0.1%
arsenazo-I solution was obtained by dissolving 0.1000
g arsenazo-I powder into 100 mL purified water. All
other reagents were of AR grade.
Table 1. General description and properties of D418 resin
Items Properties
Functional group -NHCH2PO3Na2
Specific surface area (m2/g) 400-440
Pore diameter (Å) 8.0-9.0
Pore volume (mL/g) 1.0-1.1
Mass capacity (mmol/g) ≥14
Moisture (%) 50-60
Wet apparent density (g/mL) 0.70-0.80
True wet density (g/mL) 1.10-1.18
Particle size (≥95%) 0.4-1.25mm
Whole bead after osmotic attrition ≥95
2.2. Apparatus
The Gd(III) ion was determined with Shimadzu
UV-2550 UV-visiable spectrophotometer. Mettler
Toledo Delta 320 pH meter was used for measuring
1614
pH of solutions. The sample was shaken in the DSHZ-
300A and the THZ-C-1 temperature constant shaking
machine. The water used in the present work was
purified using Molresearch analysis-type ultrapure
water machine. The IR spectra were recorded on a
Nicolet 380 FT-IR spectrometer. Thermo gravimetric
analysis was performed on Mettler Toledo
TGA/DSC1 STARe.
2.3. Adsorption experiments
Experiments were run in a certain range of pH,
temperature, contact time and adsorption isotherms.
The operations for the adsorption and desorption of
Gd(III) were carried out in batch vessels and glass
columns.
Batch experiments were performed under
kinetics and equilibrium conditions. A desired amount
of treated D418 resin was weighed and added into a
conical flask, followed by the addition of a desired
volume of buffer solution with pH 5.5. After 24h, a
required amount of standard solution of Gd(III) was
put in. The flasks were shaken in a shaker at constant
temperature. The upper layer of clear solution was
taken for analysis until adsorption equilibrium
reached. The procedure of kinetic tests was identical
to that of the equilibrium tests. The aqueous samples
were taken at preset time intervals and the
concentrations of Gd(III) were similarly measured.
In the column experiments, continuous packed
bed studies were performed in a fixed bed mini glass
column of 0.3 cm inner diameter with resin and filled
with Gd((III) solution. The Gd(III) ion solution was
fed from the top at a fixed flow rate. The Gd(III)
solutions at the outlet of the column were collected at
regular time intervals and the concentrations of Gd(III)
were measured till the outlet and inlet concentrations
became equal.
2.4. Desorption experiments
Desorption experiments were performed
immediately following the completion of the
adsorption experiments.
Batch desorption: at the end of adsorption
experiments, the exhausted resins were separated from
the aqueous solution by filtration, washed by
deionized water and were stirred with eluent solutions
of various concentrations, oscillating to balance at
constant temperature. After equilibrium reached, the
concentration of Gd(III) in aqueous phase was
determined and then the percentage of elution for
Gd(III) was calculated.
Column desorption: after equilibrium was
reached, D418 resins which were separated from the
aqueous solution was washed five times with optimal
pH HOAc-NaAc buffer solution. Then column
desorption experiments carried on by use the preferred
desorption conditions obtained in the batch desorption
experiments. The samples in the outlet were taken at
the preset volume intervals and the concentrations of
Gd(III) were determined, until the concentration of
Gd(III) in the effluent is zero.
 Adsorption behavior of gadolinium(III) from aqueous solution by D418 resin
2.5. Analytical method
A solution containing lower than 75µg/mg of
Gd(III) was added into a 25 mL colorimetric tube, and
then 1.0 mL of 0.1% arsenazo-I solution and 10 mL
pH 7.20 C6H15NO3-HNO3 buffer solution were added,
after the addition of purified water to the mark of the
colorimetric tube, the absorbance was determined in
an 1 cm colorimetric vessel at wavelength of 575 nm
and compared with blank test. The adsorption capacity
(Q, mg/g) and distribution coefficient (D, mL/g) were
calculated with the following formulas Eq. (1) and Eq.
(2) (Chen et al., 2014):
Q = ( C0 C e )V / m (1)
D = Q / Ce (2)
where: C0 is initial concentration in solution (mg/mL);
Ce is equilibrium concentration in solution (mg/mL);
V is solution volume of solution(mL); m is resin dry
mass (g).
3. Results and discussion
3.1. Influence of pH on the distribution coefficient of
Gd(III)
The pH of solution has been identified as the
most important variable governing the adsorption
capacity. The influence of pH on the sorption
behaviors of D418 resin for Gd(III) is shown in Fig. 1.
It can be easily seen that the distribution coefficient
for Gd(III) was the highest when the pH was 5.5 in the
HOAc-NaAc medium and decreased by either raising
or lowing pH under the experimental conditions.
Fig. 1. Influence of pH on the distribution coefficient D
(resin 15.0 mg, 298 K, 100rpm, C0=10.0 mg/30.0 mL)
At lower pH values, the Gd(III) ion uptake was
inhibited in the acidic medium and it can be attributed
to the presence of H+ ion competing with the Gd(III)
ion for the adsorption sites. Due to the Gd(III) ion was
prone to deposit beyond pH 5.5, the adsorption of
Gd(III) ions is reduced. Therefore, we chose the
HOAc-NaAc system at pH 5.5 for experiments.
3.2. Determination of adsorption rate constant and
apparent activation energy
The influence of contact time on the adsorption
of Gd(III) onto D418 resins (Fig. 2(a)) was
investigated at 288, 298 and 308 K. It is easily seen
that the amount of adsorption increased with
increasing contact time and the maximum adsorption
was observed after 45 h, beyond which there was
almost no further increase in the adsorption. In
addition, the maximum adsorption capacities
increased with increase in the temperatures.
Adsorption kinetics curves were obtained for
Gd(III) on D418 resins. The kinetics of adsorption can
be described by the liquid film diffusion model
(Xiong, 2015), using the Brykina method (Eq. 3)
(Brykina et al., 1980):
ln( 1 - F ) = kt (3)
where: F is the fractional attainment of equilibrium (F
= Qt/Qe); Qe and Qt are the amounts of Gd(III)
adsorbed on the adsorbent at equilibrium (mg/mL) and
at any time t, respectively; k is the adsorption rate
constant.
The experimental results accorded with the
equation and a straight line was obtained by plotting -
ln(1-F) vs. t. The results are listed in Table 2.
According to the Boyd equation, it can be deduced
from the linear relationship of -ln(1-F) vs. t that the
liquid film spreading might be the predominating step
of the adsorption process (Boyd et al., 1947). Also the
results show that the sorption rate constant k increases
with temperature, from which we may infer that the
sorption reaction is endothermic (Xiong et al., 2013).
Table 2. Adsorption rate constants
Linearity relation
T(K) k × 10–5 (s–1) R2
of-ln (1-F) and t
288 y=0.1934x-0.1761 5.37 0.993
298 y=0.2093x-0.1457 5.66 0.992
308 y=0.2165x-0.0857 6.01 0.996
According to the Arrhenius equation (Eq. 4):
Ea (4)
lg k = + lg A
2.303 RT
where: Ea is the Arrhenius activation energy for the
adsorption process indicating the minimum energy
that reactants must have for the reaction to proceed
(kJ/mol), A is the Arrhenius factor, R is the gas
constant (8.314 J/(mol K)), k is the adsorption rate
constant and T is the temperature.
As can be observed in Fig. 2(b), the coefficient
of determination of the straight line R2 = 0.997 was
achieved. The activation energy Ea was 4.16kJ/mol,
which could be considered as a low energy barrier in
this study (Balan et al., 2017).
1615
 Pi et al./Environmental Engineering and Management Journal 17 (2018), 7, 1613-1620

coverage. The Freundlich isotherm equation is used in

the general form, as Eq. (6) (Freundlich, 1906):

1
lg Qe = lg C e + lg K F (6)
Q(mg/g)

n
288K
298K where KF and n are the Freundlich isotherm constants
308K related to adsorption capacity and intensity of
adsorption, respectively.
The isotherm constants of all the previous
equations are very useful parameters for predicting
t(h)
adsorption capacities and also for incorporating into
(a)
mass transfer relationships to predict the design of
batch reactors. According to the results that show in
Fig. 3, higher R2 values (R2288K = 0.9993, R2298K =
0.9998, R2308K = 0.9998) were obtained from
Langmuir model for Gd(III) than from the Freundlich
model (R2288 K = 0.9679, R2298 K = 0.9575, R2308 K =
0.9513). It is evident that the adsorption of Gd(III) ion
onto D418 resin is fitted better to the Langmuir
isotherm model than that of the Freundlich isotherm
models. This indicates that the adsorption of Gd(III)
ions by D418 resin is monolayer-type.

(b)

Fig. 2. (a) Adsorption kinetics and capacity Q at different

times and different temperatures and (b) relationship
between lgk and 1/T (resin 15.0 mg, pH=5.5, 100 rpm,
C0= 10.0 mg/30.0 mL)

3.3. Adsorption isotherms

In order to perform the further characterization

of the adsorption behavior of Gd(III) on D418, the
equilibrium adsorption isotherms on D418 at different (a)
temperatures were obtained for Gd(III). Therefore, the
correlation of equilibrium data by either theoretical or
empirical equations was essential to the practical
design and operation of adsorption systems. The
Langmuir and Freundlich isotherm models
(Jaafarzadeh et al., 2014) are used to reveal the
linearity fitting and to describe how solutes interact
with the resins.
The Langmuir isotherm is governed by the
following relationship (Eq. 5) (Langmuir, 1918):

Ce = 1 + Ce
Qe Q0 K L Q0
where Qe is the equilibrium Gd(III) ions adsorption
capacity (mg/g), Ce is the equilibrium Gd(III) ions
concentration in solution (mg/mL), Q0 is the
maximum adsorption capacity corresponding to
complete monolayer coverage (mg/g), KL is the
Langmuir constant and related to the free energy of
adsorption.
The Freundlich isotherm model assumes
heterogeneous surface energies in which energy term
in Langmuir equation varies as a function of surface
(5)
(b)
Fig. 3. (a) Langmuir isotherm and (b) Freundlich isotherm
curves of Gd(III)
3.4. Thermodynamic parameters
In any adsorption procedure, both energy and
entropy considerations should be taken into account in
order to conclude whether the process is spontaneous
or not.

1616
 Adsorption behavior of gadolinium(III) from aqueous solution by D418 resin
The amounts of Gd(III) adsorbed at
equilibrium at 288, 298 and 308K have been examined
to obtain thermodynamic parameters for the
adsorption system. Thermodynamic parameters of
adsorption have been determined by using classical
thermodynamic equations under the conditions of
15.0mg D418 resin and C0=10.0mg/30.0 mL at pH
5.5, 100 rpm.
Thermodynamic parameters such as Gibbs free
energy change (ΔG), enthalpy change (ΔH) and
entropy change (ΔS) can be estimated using
equilibrium constants changing with temperature. The
slope and intercept of lgD vs. 1/T was used for the
calculation of ΔH and ΔS using the Eq. (7) (Ünlü and
Ersoz, 2006):
∆H ∆S
lg D = − + (7)
2.303 RT 2.303 R
Gibbs free energy ΔG calculated by using
equation given below (Eq. 8):
∆G = ∆H − T∆S
where R is the gas constant and T is the absolute
temperature.
The linear equation was y = -0.978x + 6.934,
the correlation coefficient of the straight line R2=
0.9765 was achieved. ∆H and ∆S values can be
estimated from slope and intercept value of this plot
lgD vs. 1/T and the ∆G values at different temperatures
were calculated using the Eq. (8), respectively. The
values of ΔS, ΔH and ΔG are ΔH 18.7 kJ/mol; ΔS
132.8 J/(K·mol); and ΔG (–)19.5 kJ/mol for 288K, (–
)20.9 kJ/mol for 298K, and (–) 22.2 kJ/mol for 308K.
The values of ∆G were negative, and confirm
that the spontaneity of the adsorption process with
increasing temperature. The positive value of ∆H
suggests that the adsorption is endothermic in nature.
In addition, the values of ∆S were found to be positive
due to the exchange of Gd(III) ions with more mobile
ions (-NHCH2PO3Na2) present on the exchanger,
which would cause increase in the entropy, during the
adsorption process.
3.5. Elution and regeneration tests
Efficient elution is essential to ensure of
adsorbed solute from resin is the reuse of resin for
repeated adsorption-desorption cycles.
After adsorption equilibrium reached, the
D418 resin separated from aqueous phase was washed
three times with pH 6.5 buffer solution. 30.0mL HCl
solution with various concentrations were added into
the flasks respectively as eluents followed by shaking
the flasks. After equilibrium was reached, the
concentration of Gd(III) in aqueous phase was
determined and then the percentage of elution for
Gd(III) was obtained. The percentages of elution are
88.7%, 93.9%, 98.3% and 96.5% for 1.0, 1.5, 2.0 and
3.0mol/L HCl concentration respectively.
The experimental results of elution show that
2.0 mol/L HCl is the most efficient. In order to show
the reusability of the adsorbent, adsorption-desorption
cycle of metal ion was repeated six times by using the
same beads. The adsorption capacities for the D418
resin did not noticeably change during the repeated
adsorption-desorption operations. These results show
that the D418 resin has good regeneration ability.
3.6. Dynamic adsorption and desorption
3.6.1. Dynamic adsorption curve
The performance of packed beds is described
through the concept of the breakthrough curve. The
breakthrough curve shows the dynamic adsorption
capacity of Gd(III) adsorbed on D418 resin and is
usually expressed in terms of adsorbed Gd(III)
concentration [Cad = inlet Gd(III) concentration (C0) -
outlet Gd(III) concentration (Ce)] or normalized
concentration defined as the ratio of effluent Gd(III)
concentration to inlet Gd(III) concentration (Ce/C0) as
a function of time or volume of effluent for a given
(8) bed height (Xiao et al., 2009). The area under the
breakthrough curve obtained by integrating the
adsorbed concentration (Cad; mg/mL) vs the
throughput volume (V; mL) plot can be used to find
the total adsorbed Gd(III) quantity (maximum column
capacity). Total adsorbed Gd(III) quantity (Q; mg/g)
in the column for a given feed concentration and flow
rate is calculated using Eq. (9):
v C0 − C e
Q=∫ dV (9)
0 m
where m (g) is the mass of the adsorbent. The capacity
value Q was obtained by graphical integration as 281
mg/g. Traditionally, the Thomas model is used to
fulfill the purpose. The model has the following form
(Eq. 10)(Min Min et al., 2017):
1 (10)
Ce / C0 =
1 + exp[K T ( Qm - C 0 V ) / θ ]
where KT [mL/(min·mg)] is the Thomas rate constant;
and θ (mL/min), the volumetric flow rate. The
linearized form of the Thomas model is (Eq. 11):
ln( C 0 / C e − 1 ) = K T Qm / θ − K T C 0 V / θ (11)
The dynamic coefficient KT and the adsorption
capacity of the bed Q can be determined from the
curve of ln(Co/Ce-1) vs. t at a certain flow rate as
shown in Fig. 4. (a) The outlet time t is obtained from
V/θ. Thomas equation coefficients for Gd(III)
adsorption were KT = 7.8×10-3 mL/(min·mg) and
Q=273mg/g. The theoretical predictions based on the
model parameters were compared with the observed
data as shown in Fig. 4. (b) Therefore, it can be
concluded that the experimental data fitted well to the
Thomas model, which shows that the external and
internal diffusion isn't the limiting step.
1617
 Pi et al./Environmental Engineering and Management Journal 17 (2018), 7, 1613-1620
ln(Co/Ce-1)
y = -0.0471x + 2.2304
R² = 0.9642
T(h)
(a)
Ce/Co
V(mL)
(b)
Fig. 4. (a) Plot of ln(C0/Ce-1) vs t and (b) Breakthrough
curve for adsorption of Gd(III) (resin 150 mg, pH = 5.5,
C0= 0.20 mg/mL, flow rate = 0.139 mL/min)
3.6.2. Dynamic desorption curve
Efficient elution of adsorbed solute from D418
in column was essential to ensure the Recycling
performance of D418. With respect to the separate
Gd(III) from D418, using 2mol/L HCL as eluent.
Desorption curve was plotted as the effluent
concentration (Ce) vs elution volume from the column
at 0.278mL/min flow rate. It can be seen from Fig. 5
that the adsorption flow rate (0.278mL/min) was low
so that the volume of elution was small which helped
in easy handling, and high in concentration for
economic recovery of Gd(III). It was observed that the
total volume of eluent was 175 mL and after which
further desorption was negligible. Therefore, the
Weight(%)
(a)
Fig. 6. (a) FTIR spectra and (b) TGA of D418 resin before and after being loaded Gd(III)
1618
2.0mol/L HCl eluant could help in easy handling and
removing of Gd(III).
Ce(mg/mL)
V(mL)
Fig. 5. Dynamic desorption curve
(flow rate = 0.278mL/min)
3.7. Infrared spectra analysis
The FTIR spectral analysis is important to
identify some characteristic functional groups which
are responsible of adsorbing metal ion. From the
above results, it can be deduced that the adsorption of
Gd(III) ion by D418 resin belongs to a chemical
adsorption. Therefore, in the functional groups of
D418 resin, N-H, P=O, P-OH and Gd(III) are
supposed to form chemical bonds. Fig. 6 shows typical
FTIR spectra for D418 resin before and after the
adsorption of Gd(III) ions and after desorption. It
should be noticed that the characteristic adsorption
peaks of N-H shift from 3420cm-1 to 3411cm-1. P=O
characteristic adsorption peaks shift from 1175cm-1 to
1168cm-1.
These results suggest that the formation of the
coordination of the oxygen and Gd(III) weakens the
stretch contraction vibration and causes the peak to
shift to the lower frequency. The peak shifted from
928cm-1 to 918cm-1 indicates that H of P-OH has been
replaced. The infrared spectra of resin that before
adsorption and after desorption are basically the same,
indicating that D418 resin can regeneration effectively
after eluted with HCl.
D41
8
Temperature (℃)
(b)
 Adsorption behavior of gadolinium(III) from aqueous solution by D418 resin
3.8. Thermo gravimetric analysis
TGA were carried out at a heating rates
20oC/min under nitrogen (flow rate, 20 mL/min). In
the present work, according to the TGA curves, the
residue of D418 resin before and after Gd(III) ion
adsorption is about 30.7wt.% and 48.1wt.%,
respectively. The difference of residue (17.4 wt.%) is
the weight of adsorptive Gd(III) ion, and it approaches
the experimental value (16.9 wt.%), which can be
calculated from the maximum adsorption capacity
(281 mg/g). These results prove the adsorption of
Gd(III) ion on D418 resin.
4. Conclusions
Gd(III) can be optimally adsorbed on D418
resin in pH 5.5 HOAc-NaAc system. At 298K, the
dynamic saturated adsorption capacity of Gd(III) is
281mg/g. The dynamic adsorption curve of D418
resin for Gd(III) obeys the Langmuir isotherm.
The analysis of the thermodynamic parameters
(ΔS, ΔH and ΔS) of Gd(III) adsorption showed that the
adsorption process was spontaneous and endothermic.
The Gd(III) adsorbed on D418 resin can be effectively
eluted by using 2.0mol/L HCl solution. FTIR and
TGA of D418 resin adsorbed Gd(III) show that the
functional group of resin is coordinated with Gd(III)
to form coordination compound. D418 resin has very
good potential for utilization as an adsorbent for
Gd(III) from aqueous medium.
Acknowledgements
The work is supported by the National Natural Science
Foundation of China (No.21276235), The Key Research and
Development Program of Zhejiang Province
(No. 2017C03049), and The Program of Science and
Technology of Zhejiang Province, China (No. 2017C311,
LGN18C200013)
References
Alexandratos S.D., (2000), Ion-complexing polymer-
supported reagents: synthesis and application to
separations science, Chinese Journal of Reactive
Polymers, 9, 4-4.
Awual M.R., Kobayashi T., Miyazaki Y., Motokawa R.,
Shiwaku H., Suzuki S., Okamoto Y., Yaita T., (2013a),
Selective lanthanide sorption and mechanism using
novel hybrid Lewis base (N-methylN-phenyl-1,10-
phenanthroline-2-carboxamide) ligand modified
adsorbent, Journal of Hazardous Materials, 252-253,
313-320.
Awual M.R., Kobayashi T., Shiwaku H., Miyazaki Y.,
Motokawa R., Suzuki S., Okamoto Y., Yaita T.,
(2013b), Evaluation of lanthanide sorption and their
coordination mechanism by EXAFS measurement
using novel hybrid adsorbent, Chemical Engineering
Journal, 225, 558-566.
Awual M.R., Yaita T., Shiwaku H., (2013c), Design a novel
optical adsorbent for simultaneous ultra-trace
cerium(III) detection, sorption and recovery, Chemical
Engineering Journal, 228, 327-335.
Balan C., Robu B., Bulai P., Bilba D., Balan C., Robu B,
(2017), Modeling equilibrium data for Cd (Ⅱ) adsorption
by peat using non-linear regression analysis,
Environmental Engineering and Management Journal, 16,
569-580.
Boyd G.E., Adamson A.W., Myers L.S., (1947), The
exchange adsorption of ions from aqueous solutions by
organic zeolite II kinetics, Journal of American
Chemical Society, 69, 2836-2848.
Brykina G.D., Marchak T.V., Krysina L.S., Belyavskaya T.A.,
(1980), Ion sorption of copper and nickel by AV-17X8
anion-exchanger, modified with 1-(2-thiazolylazo)-2-
naphthol-3, 6-disulfonic acid disodium salt, Journal of
Analytical Chemistry, 35, 2294-2299.
Chen Y.Y., Zhao Y., (2003), Synthesis and characterization
of polyacrylonitrile-2-amino-2-thiazoline resin and its
sorption behaviors for noble metal ions, Reactive and
Functional Polymers, 55, 89-98.
Chen X.Y., Xiong C.H., Yao C.P., (2014), Evaluation of the
adsorption of Copper(II) from aqueous solution by D151
resin, Environmental Engineering and Management
Journal, 13, 783-790.
El-Sofany E.A., (2008), Removal of lanthanum and
gadolinium from nitrate medium using Aliquat-336
impregnated onto Amberlite XAD-4, Journal of
Hazardous Materials, 153, 948-954.
Freundlich H.M.F., (1906), Over the adsorption in solutions,
Journal of Physical Chemistry, 57, 385-470.
Jaafarzadeh N., Teymouri P., Babaei A.A., Alavi N.,
Ahmadi M., (2014), Biosorption of cadmium(II) from
aqueous solution by NaCl-treated ceratophyllum
demersum, Environmental Engineering and Management
Journal, 13, 763-773.
Junk G.A., Richard J.J., Grieser M.D., Witiak D., Witiak
J.L., Arguello M.D., Vick R., Svec H.J., Fritz J.S.,
Calder G.V., (1974), Use of marcoreticular resins in the
analysis of water for trace organic contaminants,
Journal of Chromatography A, 99, 745-752.
Kondo K., Kamio E., (2002), Separation of rare earth metals
with a polymeric microcapsule membrane,
Desalination, 144, 249-254.
Langmuir I., (1918), The adsorption of gases on plane surfaces
of glass, mica, and platinum, Journal of the America
Chemistry Society, 40, 1361-1403.
Liang T., Ding S.M., Song W.C., Chong Z.Y., Zhang C.S.,
Li H.T., (2008), A review of fractionations of rare earth
elements in plants, Journal of Rare Earth, 26, 7-15.
Li J., Miao X.X., Chen X.Y., Lu L., Yang Y., Fu Y.Q., Xiong
C.H (2016). Application and characterization of grafted
polytetrafluoroethylene fiber for enhanced adsorption
of Cu(II) in aqueous solutions, Journal of Central South
University, 23, 2513-2519.
Lin B., Li S., Fang L., Zhu M., Xiong, C. (2018). Synthesis
and characterization of amino-terminated chloration
modified peanut shell and its application to
preconcentrate and detect the concentration of sunset
yellow in drink and jelly samples, Food Analytical
Methods, 1-14.
Matlock M.M., Howerton B.S., Atwood D.A., (2002),
Chemical precipitation of heavy metals from acid mine
drainage, Water Research, 36, 4757-4764.
Min M., Shen C., Fang L., Zhu B., Li J., Yao L., (2017), Design
of a selective regenerable cellulose microcolumn for
selenium efficient recovery and economic determination.
Chemical Engineering Research & Design, 117, 773-783.
Shen C., Chang Y., Fang L., Min M., Xiong C.H., (2016),
Selective removal of copper with polystyrene–1,3-
1619
 Pi et al./Environmental Engineering and Management Journal 17 (2018), 7, 1613-1620
diaminourea chelating resin: synthesis and adsorption
studies, New Journal of Chemistry, 40, 3588-3596.
Tofan L., Wenkert R., Paduraru C., (2016), Natural and
waste materials as green sorbents for Cd(Ⅱ) removal
from aqueous effluents, Environmental Engineering
and Management Journal, 15, 1049-1058.
Ünlü N., Ersoz M., (2006), Adsorption characteristics of heavy
metal ions onto a low-cost biopolymeric sorbent from
aqueous solutions, Journal of Hazardous Materials, 136,
272-280.
Xiao K., Wang X.M., Huang X., Waite T.D., Wen X.H., (2009),
Analysis of polysaccharide, protein and humic acid
retention by microfiltration membranes using Thomas’
dynamic adsorption model, Journal of Membrane
Science, 342, 22-34.
Xiong C.H., Li Y., Wang G., Fang L., Zhou S., Yao C.P.,
(2015). Selective removal of Hg(Ⅱ) with polyacrylonitrile-
1620
2-amino-1,3,4-thiadiazole chelating resin: batch and
column study, Chemical Engineering Journal, 259, 257-
265.
Xiong C.H., Jia Q., Chen X.Y., Wang G.T., Yao C.P., (2013),
Optimization of polyacrylonitrile-2-aminothiazole resin
synthesis, characterization, and its adsorption performance
and mechanism for removal of Hg(II) from aqueous
solutions, Industrial and Engineering Chemistry Research,
52, 4978-4986.
Xiong C.H., Zheng Y.Q., Feng Y.J., Yao C.P., Ma C.A., Zheng
X.M., Jiang J.X., (2014), Preparation of a novel
chloromethylated polystyrene-2-Amino- 1,3,4-thiadiazole
chelating resin and its adsorption properties and
mechanism for separation and recovery of Pt(IV) from
aqueous solutions, Journal of Materials Chemistry A, 2,
5379-5386.
Environmental Engineering and Management Journal July 2018, Vol.17, No. 7, 1621-1634
http://www.eemj.icpm.tuiasi.ro/; http://www.eemj.eu

“Gheorghe Asachi” Technical University of Iasi, Romania

REMOVAL OF HEXAVALENT CHROMIUM USING TWO

INNOVATIVE ADSORBENTS

Reda Abubeah1∗, Hossam Altaher2, Tarek E. Khalil3

1Department of Chemical Engineering, Faculty of Engineering, El-Minia University P.O. Box 6151, El-Minia, Egypt
2Departmentof Chemical Engineering Technology, Yanbu Industrial College P.O. Box 30436,
Yanbu Industrial City, Saudi Arabia
3Chemistry Department, Faculty of Science, Alexandria University P.O. Box 426 Ibrahimia, Alexandria 21321, Egypt

Abstract

Hexavalent chromium Cr(VI), a constituent of wastewater from many industries, is regulated by the Environmental Protection
Agency (EPA) as one of the priority pollutants. Adsorption is an important technique that can be used for Cr(VI) removal from
wastewater. The challenge in adsorption techniques is to find a cheap, widely available adsorbent that has a high adsorption
capacity to Cr(VI). Dates stones (DS) and palm fiber (PF) are two agricultural wastes that are produced from palm trees. These
adsorbent materials were tested in batch systems to investigate the different factors that may affect the adsorption process, e.g.
adsorbent dose, initial adsorbate concentration, pH of aqueous solution and agitation time. Equilibrium models tested include
Langmuir, Freundlich, Temkin and Dubinin-Radushkevich. Freundlich adsorption isotherm was found to best fit the
experimental data. The maximum adsorption capacities obtained for DS and PF were 16 and 6 mg/g, respectively. Kinetic studies
indicated that the adsorption of Cr(VI) on both DS and PF was fast in the first stage of the adsorption process. The equilibrium
was reached in less than 2 h for both adsorption systems. The adsorption process was found to be second order with both
adsorbent and adsorbate concentration to affect the adsorption process. Film diffusion and intraparticle diffusion models were
found to simultaneously influence the adsorption. The adsorption process was found to be favorable and of physical nature.

Key words: adsorption, chromium (VI), dates stones, kinetics, palm fiber, thermodynamic

Received: August, 2013; Revised final: August, 2014; Accepted: September, 2014; Published in final edited form: July 2018

1. Introduction Chromium has many adverse effects on human

The excessive industrial disposal of heavy
metals represents a major source of pollution
problems (Štrkalj et al., 2013). Chromium is heavily
used in many industries like tanning of leather,
cement manufacturing, dyeing of textile,
electroplating, chromate preparation, pulp and paper,
fertilizer, alloy and steel manufacturing, and
colorants production. The wastewater from these
industries contains chromium and its compounds in
concentrations higher than the permissible limit.
Chromium has two oxidation states;
hexavalent and trivalent. Chromium (VI) is 100-1000
times more toxic to organisms than Cr (III).
∗
Author to whom all correspondence should be addressed: e-mail: redaabubeah@yahoo.com; Phone: + 966597454109; Fax: +96643920213
beings like damage to liver, kidney, circulatory and
nerve tissues (Blázquez et al., 2009; Hlihor et al.,
2009; Natarajan and Nagarajan, 2010).
The treatment of wastewater containing
chromium includes many physical and chemical
methods like solvent extraction, reverse osmosis,
cementation onto iron, electrocoagulation (Sadeghi et
al., 2017), bioremoval (Fathima et al., 2017, Ghorbani
and Younesi, 2016), floating electrochemical
precipitation, ultra filtration, ion exchange, electro
dialysis, chemical oxidation and reduction (Gheju
and Balcu, 2011; Štrkalj et al., 2013). These
conventional methods were found to be expensive
and inefficient, especially when treating wastewater
 Abubeah et al./Environmental Engineering and Management Journal 17 (2018), 7, 1621-1634
with low concentration of heavy metals. Another
problem that must be taken into consideration is that
some of these conventional methods produce
chemical or biological toxic sludge. For example, the
chemical precipitation produces large amounts of
sludge that must be disposed of. Another drawback
of most of these methods is the difficulty of recovery
of the metals from these sludges (Attia et al., 2010).
Adsorption process represents an excellent
alternative for treatment of wastewater containing
heavy metal. Some important advantages of the
adsorption technique are low initial cost, low
operating cost, flexibility and simplicity of design,
and ease of operation compared to other techniques
(Chen et al., 2012).
Adsorption has the ability of complete
removal of pollutants even from dilute solutions. It
also does not result in the formation of harmful
substances (Rafatullah et al., 2010). Activated carbon
has high adsorption capacity for many pollutants
(Altaher and Dietrich, 2014). This is the reason why
many adsorption studies have focused on the
applications of such material for wastewater
treatment (Štrkalj et al., 2013). The major drawbacks
of using activated carbon include the high cost of
production and problems of regeneration. This has
led many researchers to investigate the use of low
cost and efficient adsorbents to remove heavy metals
and other pollutants from wastewater.
Many materials have been reported as
adsorbent for the removal of heavy metals, some of
them are, grape stalk (Fiol et al., 2008), coconut
(Gonzalez et al., 2008), trunk palm fibers (Altaher et
al., 2015), rice shells (Bhattacharya et al., 2008),
carbonized Eupatorium adenophorum Spreng (Jin-fa
et al., 2018), treated oil palm fibre (Isa et al., 2008),
green tea waste (Park et al., 2008), olive stone
(Blázquez et al., 2009) seaweeds (Vijayaraghavan et
al., 2005), coconut husks (Gupta et al., 2010b), waste
rubber tire (Gupta et al., 2012), deolied mustard
(Gupta et al., 2010a), Ficus carica bast (Gupta et al.,
2013), polypyrrole (Hasaniet et al., 2015), treated
alga (Gupta et al., 2002), rice husk and its ash
(Ahmaruzzaman and Gupta, 2011), fertilizer industry
waste material (Gupta et al., 2010c), nanoparticles
(Predescu et al., 2016), orange peel (Gupta and
Nayak, 2012), and carbon nanotubes (Gupta and
Saleh, 2013).
In the current study, dates stones (DS) and
dates palm fibers (PF); two agricultural wastes from
palm trees, were investigated as low-cost adsorbents
for the removal of hexavalent chromium ions from
aqueous solutions. The experimental parameters that
may affect the adsorption process such as initial
concentration, temperature, pH, contact time, stirring
rate and the adsorbent dose were studied. Some
adsorption and kinetic models were applied to
investigate the validity of the produced data.
1622
2. Materials and methods
2.1. Adsorbate
Potassium dichromate stock solution was
prepared by dissolving 1 g of the solid pure reagent
(weighed to the fourth decimal point) in 1000 mL of
distilled water. The working solution was prepared
by diluting the stock solution with distilled water to
the required concentration. The initial and final
concentration of Cr(VI) was determined following
ASTM method number (3500 Cr B) based on the
reaction of the hexavalent chromium ions with
diphenyl carbazide to form a red-violet color that can
be measured at a wavelength of 540 nm. Shimadzu
UV-Spectrophotometer was utilized for the
determination.
2.2. Adsorbent
Date stones and palm fibers were obtained
from the local market at Yanbu, Saudi Arabia. The
samples were washed several times with tap water to
remove foreign materials and dust, followed by
washing three times with distilled water. The samples
were then dried in an oven at 105°C overnight.
The samples were ground into fine powder by
a kitchen mixer. The powder was sieved to the
required size and kept in sealed glass bottles. No
chemical treatment was applied to the two
adsorbents.
2.2.1. Characterization of adsorbents
2.2.1.1. Fourier Transform Infrared (FT-IR) study
The FTIR spectra of the two adsorbents were
investigated using Infra red Spectrophotometer
Varian, USA, FT IR 800 Scimitar Series. The spectra
were recorded in the range between 4000 and 400
cm-1 as KBr pellets. The pressed KBr pellets were
prepared by grinding 200 mg of adsorbent samples
with 0.5 g of KBr.
2.2.1.2. Scanning Electron Microscope (SEM) study
The surface morphology of the two adsorbents
was examined using scanning electron microscope
[SEM (JEOL-JSM 5300)] at different magnification.
2.2.1.3. Zero point of charge
The zero point of charge (pHzpc) was
determined by preparing 2 series of flasks containing
0.1N NaCl. One series was intended for DS and the
other was for PF. The pHs of the two series were
adjusted using NaOH and HCl. A fixed amount of
the adsorbent (0.15 g) was added to each flask. The
mixtures were shaken for 48 hours at room
temperature. The final pH of each solution was
determined (Kun-yu et al., 2008).
 Removal of hexavalent chromium using two innovative adsorbents
2.2.1.4. Specific surface area, apparent density,
moisture and ash content
The specific surface area was determined by
using Sears method (Sears, 1956). According to this
method, 1.5 g of the adsorbent was added to 100 mL
dilute HCl to reach a pH value in the range 3 - 3.5.
Sodium chloride (30 g) was added to the acidic
solution with continuous stirring. The mixture
volume was made up to 150 mL using distilled water.
The suspension was titrated with standard 0.1 M
NaOH to reach a pH 4 and then 9. The added volume
(V) required to raise the pH from 4 to 9 was
recorded. The test was made in triplicate and the
value of surface area (S) was determined depending
on the average of the three obtained value.
The apparent density of each adsorbent was
determined by filling a 10 mL measuring cylinder
with the adsorbent. Gentle tapping was applied on the
cylinder wall to fill the gaps between adsorbents
particles. The apparent density was determined as the
mass of the adsorbent required to fill the cylinder
divided by the volume of the cylinder.
To determine the moisture content of the
adsorbent, 10 grams of the adsorbent (weighed to the
fourth decimal point) were placed into weighing
watch glass. The watch glass was then placed inside
an oven at 105 oC. After 5 hours of drying, the watch
glass was cooled in desiccator. The mass of the dried
adsorbent was then measured. The average moisture
content of the samples was assumed to be the mass
loss from the adsorbent sample. The experiment was
repeated five times and the average was calculated.
For determination of ash content, three clean
empty crucibles were ignited at 550oC for 30
minutes, cooled in a desiccator and then weighed.
Accurately weighed 10g of each adsorbent were
placed in each crucible. The samples were ignited at
550oC for 2 hours. The difference in weight of
crucibles before and after ignition represented the ash
content. The experiment was repeated three times
and the average was considered.
2.2.1.5. Determination of the functional groups
(Boehm’s titration)
The surface functional groups of the two
adsorbents were determined using Boehm’s titration
procedure (Boehm, 2002). This method includes
acid-base titration of the adsorbent to identify the
nature of functional groups on the surface of
adsorbent. Twenty-five milli-grams of date stones or
palm fibers were placed in 25 ml of 0.1 M solution of
each of sodium bicarbonate, sodium carbonate or
sodium hydroxide. The mixtures were shaken for 24
h then centrifuged at 6000 rpm. A given amount of
each supernatant (5 ml) was titrated with 0.1 M
hydrochloric acid solution. Similar procedures were
performed with blank samples. Triplicate of this
experiment was conducted and the average was
calculated. The number of acidic sites was
determined under the assumptions that sodium
bicarbonate neutralizes carboxylic groups, sodium
carbonate neutralizes both carboxylic and lactonic
groups and sodium hydroxide neutralizes carboxylic,
lactonic and phenolic groups.
2.3. Adsorption equilibrium studies
The adsorbate uptake at equilibrium, qe
(mg/g), was determined by (Eq. 1).
(Co − Ce )V
qe = (1)
W
where Co and Ce (mg/L) are the initial and final
concentration of the dye, respectively, W (g) is the
mass of the adsorbent used, and V (L) is the volume
of the solution. Percentage dye removal (R) is
determined according to (Eq. 2).
Co − Ce
R= × 100 (2)
Co
The adsorbent capacity towards chromium
removal (effect of adsorbent mass) was investigated
by adding fixed volumes (0.05 L) of the working
solution of chromium of known concentration (50
mg/L) to 125 mL conical flasks containing different
masses of adsorbent. The flasks were stoppered and
shaken for 300 min at a rate of 300 rpm. The
solutions were filtered using 0.45 µm membrane
filters and the residual hexavalent chromium in the
filtrate was determined using the dipheny carbazide
method as described before. To investigate the other
factors that may affect the adsorption process (pH,
initial concentration and time of agitation), the
previous procedure was conducted several times. In
every experiment all factors were kept constant and
the studied factor was changed. All experiments were
conducted at room temperature (20 ± 2oC). Other
experimental conditions are summarized in Table 1.
2.4. Mathematical models
2.4.1. Adsorption isotherms
The linear form of Langmuir isotherm is given
by (Eq. 3).
Ce 1 C
= + e (3)
qe q m . K L q m
where Ce (mg/L) is the equilibrium concentration, qm
(mg/g) is the maximum adsorption capacity, qe
(mg/g) is the adsorption capacity at equilibrium, and
KL (L/mg) is a constant related to adsorption rate.
The favorability of the adsorption is determined by
the dimensionless separation factor RL which is given
by (Eq. 4).
1
RL = (4)
1 + K L Co
where Co (mg/L) is the initial dye concentration and
KL is Langmuir constant.
The value of separation factor indicates the
favorability of the adsorption process as follows:
1623
 Abubeah et al./Environmental Engineering and Management Journal 17 (2018), 7, 1621-1634

Table 1. Range of variables for batch experiments

Initial concentration, Contact Adsorbent Stirring Particle Volume,

Experiment pH
mg/L time, min mass, g rate, rpm size, µm mL
Effect of initial concentration 15-450 300 0.15 300 < 250 ≈4.5 50
Effect of adsorbent mass 50 300 0.05-0.3 300 < 250 ≈4.5 50
Effect of time 50 5-360 0.15 300 < 250 ≈4.5 50
Effect of pH 50 300 0.15 300 < 250 2-7 50
Effect of temperature
50 300 0.15 300 < 250 ≈4.5 50
10, 20, 30, 40oC

Unfavorable (RL > 1), linear (RL = 1), R is the universal gas constant in J/mol. K and
favorable (0 < RL < 1), and irreversible (RL = 0). The T is the temperature (K).
linear form of Freundlich model is given by (Eq. 5).
2.4.2. Kinetic models
1 Many kinetic models have been introduced to
ln qe = ln k f + ln Ce
n (5) examine the mechanism of adsorption processes.
Some of the commonly used kinetic models were
where qe and Ce have the same definition as applied. These kinetic models are pseudo first order,
Langmuir equation, kf (mg/g(mg/L)1/n) is constant pseudo second order, Boyd model, intraparticle
related to the bonding energy. Kf is defined as the diffusion model and Elovich equation.
adsorption coefficient, and 1/n identifies the The linear form of pseudo first order model is
adsorption intensity of adsorbate onto adsorbent. If given by Eq. (11):
the value of exponent n is greater than 1, the
adsorption represents favorable adsorption ln(qe − qt ) = ln qe − k1t (11)
conditions.
The linear form of Temkin isotherm is given where qe (mg/g) and qt (mg/g) are the amounts of dye
by Eqs. (6-7). adsorbed on the surface of the adsorbate at
equilibrium and at any time t (min), respectively; and
qe = BT (ln KT ) + BT ln(Ce ) (6) k1 (min-1) is the rate constant of the pseudo first order
equation.
RT The linear form of the pseudo second order
BT = (7)
b model is given by Eq. (12):

where qe (mg/g) and Ce (mg/L) have the same t 1 t (12)

= +
definition as in Langmuir equation, BT (J/mol) is qt k 2 qm qm
2

Temkin isotherm constant related to the heat of

adsorption, b is Temkin isotherm constant, R is the
universal gas constant (8.314 J/mol K), and T is the
absolute temperature (K). The constant KT (L/mg) is
the equilibrium binding constant, corresponding to
maximum binding energy.
Dubinin-Radushkevich model is represented
by Eq. (8):
where k2 (g/mg.min) is the rate constant of the
pseudo second order equation, qm (mg/g) is the
maximum adsorption capacity, qt (mg/g) is the
amount adsorbed at time t (min).
Boyd model (Hu et al., 2011) is applied to
distinguish between the pore and film diffusion. This
model is expressed in Eq. (13):

ln qe = ln qm − Bε 2 (8) Bt = − 0.4977 − ln(1 − F ) (13)

where qe is the amount of adsorbate adsorbed onto where Bt is the mathematical function of F and F
adsorbent surface (mg/g), qm represents the represents the fraction of solute adsorbed at time t (h)
maximum adsorption capacity of adsorbent (mg/g), and F = qt / qe
and B is a constant related to the mean free energy of According to the model, if the plot of Bt
adsorption per mole of adsorbate. versus t passes through the origin, pore diffusion is
The Polanyi potential (ε) is given by Eqs. (9- the rate determining step. Otherwise, the adsorption
10): process is film diffusion controlled.
Intraparticle diffusion model considers the
1
ε = RT ln (1 + ) (9) intraparticle diffusion as the rate controlling step for
Ce adsorption. It is given by Eq. (14):
1
E= (10) qt = k id t 0.5 + C (14)
(2 B) 0.5

1624
 Removal of hexavalent chromium using two innovative adsorbents
where qt (mg/g) is the amount adsorbed at time t
(min) and kid (mg/g.min0.5) is the rate constant for the
intraparticle diffusion model.
A simplified linearized form of Elovich
equation (Malkoc and Nuhoglu, 2007) can be
introduced by Eq. (15):
1 1
qt = ln(αB) + ln t (15)
B B
where α (mg/g min) is the initial adsorption rate and
B (g/mg) is the adsorption constant related to the
extent of the surface coverage and activation energy
for chemisorption.
2.5. Thermodynamic study
The thermodynamics data obtained from the
adsorption process can be used to introduce the
adsorption mechanism. The Gibbs free energy
change ∆G can be calculated by Eq. (16):
∆G = − RT ln K L (16)
where KL is the Langmuir adsorption constant, R is
the universal gas constant and T is the absolute
temperature. The enthalpy change and the entropy
can be obtained from Van’t Hoff equation (Eq. 17):
∆S ∆H
ln K L = − (17)
R RT
The enthalpy change and entropy change can
be obtained from the slope and intercept of the line
obtained by plotting ln KL against 1/T, respectively.
3. Results and discussion
3.1. Infrared (FTIR) Spectral Analysis
FTIR spectra for both adsorbents are
illustrated by Figs. 1a-b. The figures indicate a
number of absorption peaks suggesting complex
properties of the adsorbents. Spectra are similar for
both samples. The peaks in the region 1200–1000
cm-1 may represent C–O stretching vibrations (Sain
and Panthapulakkal, 2006). The band at 1261 cm-1 in
the spectrum of palm fibers is ascribable to the
bending modes of O–C–H, C–C–H and C–O–H
(Bouchelta et al., 2008). The peaks at 1377 and 1379
cm-1 are due to C–H bending vibrations. The peaks at
1516, 1523 and 1426, 1441 cm-1 may be ascribed to
C=C stretching of aromatic skeletal mode (Pavan et
al., 2008; Sain and Panthapulakkal, 2006). The bands
at 1641 and 1631 cm-1 may represent C=C stretching
vibration of alkenes or C=N vibrations in aromatic
region. The peak at 1743 cm-1 in the spectrum of date
stones is assigned to C=O stretching vibration of
carboxylic groups of lignin and/or hemicelluloses
(Pavan et al., 2008; Sun et al., 2005). The bands
observed at 2924 and 2860 cm-1 are assigned to
asymmetric and symmetric C–H and symmetric C–H
bands, respectively, present in alkyl groups. The
dominant peaks at 3426 and 3430 cm-1 respectively
for palm fibers and date stones, are attributed to O–H
stretching vibrations in hydroxyl groups and appears
in this lower region where these groups are involved
in hydrogen bonds.
The FTIR spectra of date stones and palm
fibers waste are in good agreement with the results of
Yang (Yang et al., 2007) who studied the
characteristics of cellulose, hemicellulose and lignin.
This result is in accordance with the composition of
lignocellulosic material such as date stones and palm-
trees waste, which are essentially composed of
cellulose, hemicellulose and lignin (Bouchelta et al.,
2008). Consequently, the FTIR results indicate that
the biosorbents presented different functional groups
such as hydroxyl, carboxyl and carbonyl, which may
be potential biosorption sites for the removal of
chromium.
3.2. Scanning Electron Microscope (SEM)
The results of the scanning electron
microscopy are shown in Figs. 2-3. It is quite clear
that the morphology of the two adsorbents is quite
different except for one feature which is the
macropores position. The macropores seem to be on
the surface of the two adsorbents.
Also, these macropores are connected to meso
and micropores through channels. The surface of DS
is very porous and the pores are irregular and have
different dimension. Layers of adsorbent can be
clearly seen. On the other hand, the surface of PF
showed relatively smooth regions. The pores on the
surface of PF seem to be more regular and
homogeneously distributed all over the surface.
3.3. Specific surface area, moisture, ash content and
apparent density
The proximate analyses are shown in Table 2.
The percentage moisture content (%MC) was
computed as follows (Eq. 18):
loss in weight on drying
% MC = × 100 (18)
initial mass of sample before drying
The moisture content is relatively high and
indicates the necessity of drying the adsorbent before
use or otherwise, taking the moisture content into
consideration during calculation. The ash content was
calculated using (Eq. 19)
ash weight
Ash content = × 100 (19)
oven dry weight
Ash content can affect the adsorbent
efficiency, i.e. it reduces the overall affinity of
adsorbent towards adsorbate. The ash content is low
and it indicates that the adsorbent could be used in
further studies for production of activated carbon.
The specific surface area was calculated using Eq.
(20), where S and V signification is explained in the
section 2.2.1.4.
1625
 Abubeah et al./Environmental Engineering and Management Journal 17 (2018), 7, 1621-1634

(a)

(b)

Fig. 1. (a). FTIR spectrum for DS, b. FTIR spectrum for PF

Fig. 2. Scanning electron micrograph of DS Fig. 3. Scanning electron micrograph of PF

Table 2. Physical properties of adsorbents

Specific surface area, m2/g Ash content, % Moisture content, % Apparent density, g/cm3
DS PF DS PF DS PF DS PF
399 237.4 3.19 3.62 3.03 3.3 0.47 0.34

1626
 Removal of hexavalent chromium using two innovative adsorbents

S (m 2 g −1 ) = 32 V − 25 (20) Another factor that must be taken into

consideration is the aggregation of adsorbent
The specific area is comparable to many of particles that may take place due to their high
the agricultural wastes that are commonly used concentration in the solution (high dose of adsorbent
(Ioannidou and Zabaniotou, 2007). in the solution). This aggregation will mask some of
the pores on the surface of the adsorbent and also
3.4. Adsorbent surface functional groups increase the diffusion path length (Ahmaruzzaman,
2008).
The results of functional groups distribution
on the adsorbent surface are summarized in Table 3.
These results are based on Boehm’s titration
illustrated in previouis section. It is well known that
surface chemical groups are more complex than
shown by Boehm's titration; however, this method
gives a semi-quantitative indication about the surface
functionalities. It is clear from the results that the
acidic sites in both date stones and palm fibers are
mainly due to the presence of phenol and carboxyl
groups. Few lactones were detected on date stones
surface, whereas the palm fiber sample does not
contain any lactones.
From the results obtained by the Boehm
titration method, it can be observed that the amount Fig. 4. Effect of DS dose on chromium removal
of basic groups was significantly higher than total
amount of acidic groups. These results suggest that
both date stone and palm fibers materials had a basic
character.

Table 3. Distribution of functional groups on adsorbent

surface

Basic Acidic Sites, mmol/ g

sites, Total
Adsorbent
mmol/ Carboxyl Lactone Phenol acidic
g sites
Date
1.3 0.25 0.10 0.75 1.1
stones
Palm
1.2 0.30 0.00 0.75 1.05
fibers Fig. 5. Effect of PF dose on chromium removal

3.5. Effect of adsorbent dose on adsorption 3.6. Comparison of adsorption capacity of DP and
performance PF with other adsorbents

One very important factor when designing an
adsorption system is the optimum adsorbent dose
required for this system. Fig. 4 and 5 indicate that
increasing the dose of adsorbent increased the
adsorption efficiency of adsorbate. Increasing the
dose from 1 to 6 g/L resulted in an increase of
adsorption removal efficiency from 23 to 69%.
However, the adsorption capacity decreased from
11.4 to 5.7 mg/g. On the other hand, for PF, the same
increase in adsorbent dose (from 1 – 6 g/L) had a
lower effect. The corresponding increase in
adsorption removal efficiency was from 15 – 35 %.
The adsorption capacity for this system increased
from 3 – 7.5 mg/g.
These results can be attributed to the effect of
the mass of adsorbent on the number of adsorption
sites. Increasing the mass of adsorbent increases the
number of active sites available for adsorption, thus
resulting in increasing removal efficiency.
The adsorption capacity of Cr (VI) on both
dates stones and palm fibers compared with other
agricultural wastes adsorbents reported in literature is
illustrated in Table 4. It can be noticed that the
adsorption capacity of DS (16 mg/g) is higher than
some of these adsorbents and similar to others. The
adsorption capacity of PF (6 mg/g) is lower than DS
and most of the other adsorbents. Many reasons can
explain the difference in adsorption capacity among
these adsorbents including the difference in specific
surface area, the surface functional groups, and the
pore size distribution of these adsorbents.
3.7. Effect of pH and point of zero charge on
adsorption performance
The effect of aqueous solution pH on the
adsorption capacity of chromium ion is depicted by
Fig. 6.

1627
 Abubeah et al./Environmental Engineering and Management Journal 17 (2018), 7, 1621-1634

Table 4. Comparison of adsorption capacity of DP and PF with other adsorbents

Operation conditions Adsorption

Adsorbent Adsorbate initial Adsorption Mass of Temp. capacity, Reference
pH
concentration, mg/L time, min adsorbent, g/L oC mg/g
Gupta and
Saw dust 216 250 5.4 30 1.0 41.5
Babu, 2009
Thamilarasu
Ricinus seed 10 60 50 50 2.0 8.8
et al., 2012
Pistia stratiotes Das et al.,
10 15 5 40 2.0 7.24
biomass 2013
Gao et al.,
Rice straw 40 1440 10 47 2.0 3.15
2008
Carnation
flowers waste Vargas et
10 180 10 18 ± 2 2.0 6.25
(carnation al., 2012
compost)
Date stones 50 300 0.15 20 ± 2 2 16 Present
Palm fiber 50 300 0.15 20 ± 2 2 6 study

As shown in Fig. 6, the adsorption capacity of positive charge, while the adsorbate will be present
both DS and PF increased with decreasing the pH of mostly in its anionic form.
the solution. The highest adsorption capacity was
obtained at pH 2. This result is supported by the work
of other researchers (Karthikeyan et al., 2005;
Malkoc and Nuhoglu, 2007; Sarin and Pant, 2006).
The effect of pH is due to the surface
properties of the adsorbent and properties of
chromium ions in solution (Jain et al., 2009).
Hexavalent chromium may be present in various
species in aqueous solutions, e.g.
− 2− 2−
H 2 CrO4 , HCrO4 , CrO4 and Cr2 O7 according to the
solution pH.

Fig. 6. Effect of initial solution pH on chromium removal

The point of zero charge gives a good
indication about the surface charge of adsorbent.
When the pH of the aqueous solution is below the
pHzpc, the surface charge of the adsorbent is
positive. Meanwhile, the surface of the adsorbent is
negatively charged when the pH of the solution is
greater than the pHzpc. The values of point of zero
charge on the surface of DS and PF are 5.5 and 6.5,
respectively as illustrated in Fig. 7. These values
represent the points at which the initial and final pH
values of the system are equal. Below these pH
values the surface of the adsorbents will carry
Fig. 7. Zero point of charge; A: DS, B: PF
Attraction of these opposite charged species
will take place. So, it is expected to have higher
adsorption capacity at lower pH. In the pH range 2-6,
both HCrO4− and Cr2 O72− species are present in
2−
equilibrium. At lower pH ranges, Cr3O10 and
2−
Cr4O13 are produced. The adsorption of Cr(VI) may
take place through different mechanisms. The
electrostatic attraction between the dominant

1628
 Removal of hexavalent chromium using two innovative adsorbents
chromium species ( HCrO4− ) present in solution with
positively charged sites on the adsorbent is an
important mechanism. Chelation with acidic groups
on the adsorbent surface and exchange with anions
loaded on the residues are two other common
mechanisms (Gandhia et al., 2010).
3.8. Effect of initial concentration on adsorption
The initial concentration of the metal ion is
one of the important factors that may affect the
adsorption of metal ions in aqueous solution on the
surface of a certain adsorbent. The initial
concentration provides the driving force required to
overcome the mass transfer resistance of the metal
between aqueous solution and solid phase (Aksu and
Akpinar, 2000). Fig. 8 illustrates the effect of initial
Cr(VI) concentration on the adsorption into both
adsorbents. For DS, the adsorption capacity increased
from 1.99 mg/g to 74.5 mg/g when increasing the
initial concentration from 15 to 450 mg/L. For PF,
the increase was from 0.93 mg/g to 62.6 mg/g for the
same concentration range.
Fig. 8. Effect of initial concentration on chromium
removal, contact time 300 min, adsorbent mass 0.15 g,
stirring rate 300 rpm, particle size < 250 μm
For a fixed adsorbent dose, the number of
active adsorption sites to accommodate adsorbate
ions remains unchanged but with increasing
adsorbate concentration, the adsorbate ions to be
accommodated increases and hence many of the sites
available for adsorption will be occupied. It can be
noticed from the figure that adsorption saturation was
not reached due to high availability of free adsorption
sites on both adsorbents compared to the number of
chromium species to be adsorbed. The increase in
adsorption capacity accompanied by the increase in
initial ion concentration indicates increasing
occupation of available binding sites. As was stated
before, the initial concentration provides an
important driving force to overcome all mass transfer
resistances of all molecules between the aqueous and
adsorbents. Another effect of high initial
concentration of the adsorbate molecules is the
probable increase of contact between chromium
molecules and adsorbent particles. Some factors must
be taken into account when studying the effect of
initial ion concentration. First, some of the active
adsorption sites are not occupied by adsorbate
molecules which lead to less adsorption capacity.
Second, aggregation of the adsorbent particles may
take place which lead masking of some of the
adsorption sites, thus decreasing the total surface area
of the adsorbent in the solution (Yasemin and Tez,
2007).
3.9. Adsorption equilibrium
The experimental data were tested with
adsorption models namely Langmuir, Freundlich,
Temkin and D-R. The characteristic parameters from
the slope and intercept of the plots are determined
using regression analysis. The parameters are
reported in Table 5. The dimensionless separation
factor RL values (calculated according to Langmuir
model) for both DS and PF systems, being less than
unity, indicate that adsorption of Cr(VI) on both
adsorbents to be favorable. The maximum adsorption
capacity of DS is much higher than PF. However,
their values are not consistent with the experimental
values as indicated by the relatively low R2 for both
adsorbents.
Freundlich is an empirical expression. It
assumes that adsorption takes place on multilayers on
the surface of adsorbent. As illustrated by the results
in Table 5, the Freundlich model fits the data for both
DS and PF systems better than the Langmuir model,
this is demonstrated by high correlation coefficient
(R2) for the data obtained from Freundlich model.
The value of 1/n for DS is 0.772 which indicates high
favorability of the adsorption of Cr(VI) on DS. On
the other hand, 1/n value in case of PF system is
0.362. Comparing these results indicates that the DS
adsorption system is more favorable than PF
adsorption system. Analysis of the adsorption data
using D-R model has a significant Importance. It
gives an indication about the nature of the adsorption
process; physical or chemical.
The mean adsorption energy of the D-R model
gives information about the nature of the adsorption
process; physical or chemical. If E<8 kJ/mol, the
adsorption process is physical, and if 8<E<16, the
adsorption process is ion exchange, and if E > 16
kJ/mol the adsorption process is chemical. As
indicated in Table 5, the correlation coefficients of
adsorption system for DS and PF are high (0.989 and
0.957 for DS and PF, respectively). The maximum
adsorption capacity for DS is 15.441 mg/g and that
for PF is 21.22 mg/g. The first value agrees well with
experimental data, while the other one is not
consistent. The E values being less than 8 kJ/mol
suggest the physical adsorption to be the mechanism
of adsorption.
Temkin isotherm fits well to the experimental
data as indicated by correlation coefficients. Its
constants are given in Table 5. The Temkin
adsorption potential, KT, of DS is 0.159 compared to
0.034 for PF indicating a lower PF-metal ion
potential than DS.
1629
 Abubeah et al./Environmental Engineering and Management Journal 17 (2018), 7, 1621-1634
Table 5. Adsorption parameters for the applied adsorption models
Isotherm
Langmuir 20 oC
Freundlich 20 oC
Kf (mg/g)(mg/L)1/n
Temkin 20 oC
Dubinin-Radushkevich 20 oC
This is probably due to the difference in pore
diameter and pore shape of both adsorbents. As it is
depicted by Figs. 2-3, the shape of the DS pores is
irregular in contrast to the regular shape of PF pores.
The pore radius must be close to the ionic radius for
adsorption to take place.The Temkin constants, b
related to heat of sorption for DS and PF are 0.381
and 0.22 kJ/mol, respectively. It has been reported
that the typical range of bonding energy for ion-
exchange mechanism is 8-16 kJ/mol (Kiran and
Kaushik, 2008). The low values in this study
indicates a weak interaction between sorbate and
sorbent, supporting a physical adsorption mechanism.
This result is in agreement with that obtained from
D-R isotherm.
3.10. Kinetic study
Adsorption kinetics describes the solute
uptake rate, which governs the residence time of
sorption reaction. It also describes the reaction
pathway along times until reaching equilibrium. Fig.
9 depicts the adsorption rate of hexavalent chromium
ions on both DS and PF. The adsorption of chromium
ions on both DS and PF displays a fast adsorption
rate.
Fig. 9. Effect of contact time on the adsorption of
chromium on DS and PF
1630
Adsorbent
Parameters
DS PF
R2 0.916 0.929
qm, mg/g 35.7 2.24
KL, L/mg 0.013 0.017
RL 0.61 0.547
R2 0.986 0.955
n 1.295 0.362
0.71 1.7x10-4
R2 0.997 0.923
BT, mg/g 6.385 11.07
KT, L/mg 0.159 0.034
R2 0.989 0.957
qm, mg/g 15.441 21.22
E, kJ/mol 0.1 0.016
For both adsorbents equilibrium is reached
within the first two hours. The adsorption took place
in two different stages. The first one was very fast.
This is illustrated by the fact that for DS, 86 % of the
chromium ions was removed within the first 10 min
(maximum removal efficiency was 94%). The next
stage was a slow stage ending with equilibrium. A
similar behavior could be observed for PF. Within
the first 10 min, 21 of the chromium ions were
removed compared to 34% removal efficiency at
equilibrium. The rapid uptake of the metal ions at the
early stage of the adsorption process may be
attributed to the availability of active adsorption sites
on the adsorbent surface. With time, these sites got
occupied and the remaining adsorbate ions would
compete for the remaining unoccupied sites which
results in slower adsorption rate (Ozdes et al., 2011).
Another reason is the concentration gradient. As was
mentioned before the high initial concentrations
provide a driving force for chromium ions
adsorption. At early stages, a great concentration
gradient is present between the adsorbate in the bulk
of solution and on the surface of adsorbent which
easily overcomes the mass transfer resistance. With
time, the concentration gradient decreases which
decrease the driving force. The kinetic data were
evaluated using pseudo-first and second order
models. The results of fitting the experimental data to
the model are illustrated in Table 6. The data in the
table shows that the experimental data had its best fit
with the pseudo second order model with higher R2
value (0.999 and 0.995 for DS and PF, respectively)
compared with pseudo first order.
Furthermore, the values of qt predicted by the
model agree closely with those determined
experimentally. Therefore, it can be concluded that
the rate of hexavalent chromium ions adsorption is
influenced by both the metal ion concentration and
adsorbent dose (Abramian and El-Rassy, 2009). The
prediction of the rate controlling mechanism is
important in adsorption processes since it affects the
design and scale-up of the adsorption system. The
adsorption process in an aqueous solution is
 Removal of hexavalent chromium using two innovative adsorbents

characterized by either external mass or intraparticle
diffusion or both.
Table 6. Kinetic parameters for adsorption of chromium on
DS and PF
layer for the former. These results support the finding
obtained before from fitting the data to film diffusion
model.

Adsorbate
Kinetic model Parameters
DS PF
R2 0.619 0.503
Pseudo first order K1 0.029 0.026
qm 5.474 7.374
R2 0.999 0.995
Pseudo second order K2 0.032 0.018
qm 15.873 5.714
R2 0.984 0.947
Elovich B 2.513 1.835
α 2.31 46.64
R2 0.958 0.91
Intraparticle diffusion Kid 0.104 0.131
C 14.14 3.387
Boyd R2 0.956 0.88

Since the solution containing the adsorbent is

vigorously agitated during the adsorption period, it
can be assumed that the rate of adsorption is not
affected by mass transfer of ions from the bulk liquid
to the particle external surface. In this case, the rate
limiting step may be either film or intraparticle
diffusion (Moussavi and Khosravi, 2011). As they act
in series, the slower of the two will be the rate-
determining step. Film diffusion was analyzed using Fig. 10. Liquid film diffusion (Boyd model) representation
Eq. 13). The plot of liquid film diffusion on both DS of adsorption of chromium; A: DS, B: PF
and PF is illustrated in Fig. 10. Examining this Figure
indicates that two distinct regions can be identified
for the two adsorbents. This suggests that two
different mechanisms are involved at different
intervals. The results show that the film diffusion is
not the sole limiting step. The intraparticle diffusion
may also be involved in controlling the adsorption
process.
To gain further insight into the adsorption
behavior the kinetic results were analyzed by
intraparticle diffusion model. The fitting of the
experimental data to this model was examined by
applying Eq. (14). If the intraparticle diffusion is
involved in the adsorption process, the plot of the
Fig. 11. Intraparticle diffusion representation of adsorption
adsorption capacity at any time (qt, mg/g) against t of chromium on both DS and PF
(min) should give a straight line. If the intraparticle
diffusion is the only rate controlling mechanism, the The Elovich equation is useful in describing
straight line should pass through the origin. If the line adsorption on heterogeneous surfaces. However, it
does not pass through the origin other mechanism does not predict any mechanism. If the experimental
may be involved in the adsorption process. The he data fits well to this model that indicates the
intercepts (c) of the produced straight line gives an heterogeneity of adsorbent surface. The data for both
indication about the thickness of the boundary layer. DS and PF fit the model very well according to the
The larger the intercept, the greater is the high correlation coefficient (Table 6 and Fig. 12).
boundary layer effect. As illustrated in Fig. 11, the This proves the heterogeneity of surfaces of the two
adsorption on both DS and PF can be represented by adsorbents.
straight line according to this mode. Experimental
data followed a linear distribution with high 3.11. Effect of temperature and thermodynamics of
correlation coefficient (R2 values are 0.958 and 0.91 adsorption
for DS and PF, respectively). The value of the
intercept in case of DS system is much greater than Temperature has an important effect on the
that of PF, indicating the great effect of boundary adsorption process. The increase of solution

1631
 Abubeah et al./Environmental Engineering and Management Journal 17 (2018), 7, 1621-1634

temperature will decrease the viscosity of the
aqueous solution containing chromium ions.
The positive values of ∆S may suggest that
Cr(VI) ions replace some water molecules from the
solution previously adsorbed on the surface of
adsorbent. These displaced molecules gain more
translation entropy than is lost by the adsorbate ions,
thus allowing the prevalence of randomness in the
system. The positive values of ∆S may also show the
increased disorder and randomness at the solid
solution interface during the adsorption process for
both systems. In addition, the positive value of ∆S
reflects the affinity of the adsorbents for the Cr(VI)
ions (Najm and Yassin, 2009).

3.12. Feasibility study

Fig. 12. Elovich model for adsorption The two agricultural wastes used in this study
of chromium on both DS and PF are widely available with low cost. These two
substances have no significant industrial and
This will resilt in increasing the rate of commercial uses but become an issue and contribute
adsorbate diffusion across the external boundary to environmental problems. Hence, the utilization of
layer and in the internal pores of the adsorbent. such agriculture solid waste for wastewater treatment
Temperature may also affect the equilibrium of the is beneficial. The alternative of using agricultural
adsorbate depending on the nature of the adsorption waste as adsorbent is activated carbon. Activated
process; whether it is endothermic or exothermic. carbon has better adsorpitivity but its cost of
Finally, it has an effect on the stability of the metal production is very high.
ion species initially placed in solution. Dates stones collected from dates processing
The adsorption capacity increased from 6.803 plant are given away for free. Its use in other
mg/g to 28.6 mg/g for DS and from 0.062 mg/g to purposes is negligible. The cost of processing
5.16 mg/g for PF (Figures not shown) when the including transportation, washing, drying, crushing,
temperature was increased from 10 to 40oC, sieving, and labor is SR 200/ton. No chemical
indicating that the adsorption process is endothermic treatment is required for its processing. PF is
in nature for both adsorbents. collected during the trimming the palm tree. It is less
The thermodynamic parameters ∆H and ∆S available. However, its cost of processing is less; SR
are introduced in Table 7. These data were calculated 100/ton. Compared with the cheapest commercial
from the slope and intercept of the plot between ln Kc activated carbon available in the local market (SR
and 1/T. The values of ∆G indicate the favorability of 4000/ton), the two adsorbents studied are very good
physical adsorption. The positive values of ∆H replacement.
indicate the endothermic nature of adsorption and
this governs the possibility of physical adsorption 4. Conclusions
which was also supported by the increase in
adsorption capacity with increasing temperature. The Tis study indicates that DS has a better
negative values of ∆G in case of adsorption on DS adsorption capacity to hexavalent chromium than PF.
suggest the high favorability of Cr(VI) adsorption on The adsorption of Cr(VI) on both DS and PF was
DS. They also indicate that the metal ion adsorption found to be pH-dependent. The highest removal
was spontaneous. On the other hand, the positive efficiency was obtained at pH 2. The initial
values of ∆G in case of PF system indicate that this concentration of the metal ion in solution was found
adsorption system is less favorable. Furthermore, the to have a pronounced effect on the adsorption
low values of ∆G suggest physical adsorption process. Increasing the initial concentration resulted
(Kumar and Kirthika, 2009). in higher adsorption capacity. Freundlich model was
found to be the best adsorption equilibrium model
Table 7. Thermodynamic parameters of Cr(VI) adsorption
onto DS and PF
that can interpret the experimental data.
The mechanism of adsorption involves both
∆G ∆H ∆S liquid film diffusion and intraparticle diffusion as the
T (K)
(kJ/mol) (kJ/mol) (J/mol.K) rate determining mechanisms. It can be concluded
283 0.71 from the results demonstrated here that DS is a
Dates 293 -2.54
31.06 108
potential effective adsorbent for removal of Cr(VI)
stones 303 -2.03 from aqueous solutions.
313 -2.33
283 4.67 References
Palm 293 0.4
33.61 102.6
fiber 303 2.1 Abramian L., El-Rassy H., (2009), Adsorption kinetics and
313 1.6 thermodynamics of azo-dyes orange II onto highly

1632
 Removal of hexavalent chromium using two innovative adsorbents
porous titania aerogel, Chemical Engineering Journal,
150, 403-410.
Ahmaruzzaman M.D., (2008), Adsorption of phenolic
compounds on low cost adsorbents: A review,
Advances in Colloid and Interface Science, 143, 48-
76.
Ahmaruzzaman M., Gupta V.K., (2011), Rice husk and its
ash as low-cost adsorbents in water and wastewater
treatment, Industrial & Engineering Chemistry
Research, 50, 13589-13613.
Aksu Z., Akpinar D., (2000), Modeling of simultaneous
biosorption of phenol and nickel (II) onto dried
aerobic activated sludge, Separation and Purification
Technology, 21, 87-99.
Altaher H., Dietrich A.M., (2014), Characterizing o-and p-
nitrophenols adsorption onto innovative activated
carbon prepared from date pits, Water Science and
Technology, 69, 31-37.
Altaher H., Alghamdi A., Omar W., (2015), Innovative
biosorbent for the removal of cadmium Ions from
wastewater, Environmental Engineering and
Management Journal, 14, 793-800.
Attia A.A., Khedr S.A., Elkholy S.A., (2010), Adsorption
of chromium ion (Vl) by acid activated carbon,
Brazilian Journal of Chemical Engineering, 27, 183-
193.
Bhattacharya A.K., Naiya T.K., Mandal S.N., Das S.K.,
(2008), Adsorption, kinetics and equilibrium studies
on removal of Cr(VI) from aqueous solutions using
different low-cost adsorbents, Chemical Engineering
Journal, 137, 529-541.
Blázquez G., Hernáinz F., Calero M., Martín-Lara M.A.,
Tenorio G., (2009), The effect of pH on the
biosorption of Cr (III) and Cr (VI) with olive stone,
Chemical Engineering Journal, 148, 473-479.
Boehm H., (2002), Surface oxides on carbon and their
analysis: a critical assessment, Carbon, 40, 145-149.
Bouchelta C., Medjram M.S., Bertrand O., Bellat J.P.,
(2008), Preparation and characterization of activated
carbon from date stones by physical activation with
steam, Journal of Analytical and Applied Pyrolysis,
82, 70-77.
Chen C., Zhang M., Guan Q., Li W., (2012), Kinetic and
thermodynamic studies on the adsorption of xylenol
orange onto MIL-101(Cr), Chemical Engineering
Journal, 183, 60-67.
Das B., Mondal N.K., Roy P., Chattaraj S., (2013),
Equilibrium, kinetic and thermodynamic study on
chromium(VI) removal from aqueous solution using
Pistia Stratiotes biomass, Chemical Science
Transaction, 2, 85-104.
Fathima A., Sadiq M., Rao J.R., Nair B.U., (2016),
Bioremoval of trivalent chromium using indigenous
bacillus species– a biofilm based approach,
Environmental Engineering and Management Journal,
15, 1603-160.
Fiol N., Escudero C., Villaescusa I., (2008), Chromium
sorption and Cr(VI) reduction to Cr(III) by grape
stalks and yohimbe bark, Bioresource Technology, 99,
5030-5036.
Gandhia M.R., Viswanathanb N., Meenakshi S., (2010),
Adsorption mechanism of hexavalent chromium
removal using Amberlite IRA 743 resin, Ion Exchange
Letters, 3, 25-35.
Gao H., Liu Y., Zeng G., Xu W., Li T., Xia W., (2008),
Characterization of Cr(VI) removal from aqueous
solutions by a surplus agricultural waste-rice straw,
Journal of Hazardous Materials, 150, 446-452.
Gheju M., Balcu I., (2011), Removal of chromium from
Cr(VI) polluted wastewaters by reduction with scrap
iron and subsequent precipitation of resulted cations,
Journal of Hazardous Materials, 196, 131-138.
Ghorbani F., Younesi H., (2016), Bioremediation of
chromium (VI) from aqueous solution by Aspergillus
niger: optimization of process parameters,
Environmental Engineering and Management Journal,
15, 2551-2558.
Gonzalez M.H., Araujo G.C., Pelizaro C.B., Menezes E.A.,
Lemos S.G., de Sousa G.B., Nogueira A.R., (2008),
Coconut coir as biosorbent for Cr(VI) removal from
laboratory wastewater, Journal of Hazardous
Materials, 159, 252-256.
Gupta S., Babu V., (2009), Removal of toxic metal Cr(VI)
from aqueous solutions using sawdust as adsorbent:
Equilibrium, kinetics and regeneration studies,
Chemical Engineering Journal, 150, 352-365.
Gupta V.K., Ganjali M.R., Nayak A., Bhushan B., Agarwal
S., (2012), Enhanced heavy metals removal and
recovery by mesoporous adsorbent prepared from
waste rubber tire, Chemical Engineering Journal, 197,
330-342.
Gupta V.K., Jain R., Malathi S., Nayak A., (2010a),
Adsorption–desorption studies of indigocarmine from
industrial effluents by using deoiled mustard and its
comparison with charcoal, Journal of colloid and
interface science, 348, 628-633.
Gupta V.K., Jain R., Shrivastava M., Nayak A., (2010b),
Equilibrium and thermodynamic studies on the
adsorption of the dye Tartrazine onto waste “Coconut
Husks” carbon and activated Carbon, Journal of
Chemical & Engineering Data, 55, 5083-5090.
Gupta V.K., Nayak A., (2012), Cadmium removal and
recovery from aqueous solutions by novel adsorbents
prepared from orange peel and Fe2O3 nanoparticles,
Chemical Engineering Journal, 180, 81-90.
Gupta V.K., Pathania D., Agarwal S., Sharma S., (2013),
Removal of Cr (VI) onto Ficus carica biosorbent from
water, Environmental Science and Pollution Research,
20, 2632-2644.
Gupta V.K., Rastogi A., Nayak A., (2002), Biosorption of
nikel onto treated alga (Oedogonium hatei):
Application of isotherm and kinetic models, Journal of
Colloid and Interface Science, 342, 533-539.
Gupta V.K., Rastogi A., Nayak A., (2010c), Adsorption
studies on the removal of hexavalent chromium from
aqueous solution using a low-cost fertilizer industry
waste material, Journal of Colloid and Interface
Science, 342, 135-141.
Gupta V.K., Saleh T.A., (2013), Sorption of pollutants by
porous carbon, carbon nanotubes and fullerene-An
overview, Environmental Science and Pollution
Research, 20, 2828-2843.
Hasani E,, Farnam M., Asl S.M.H., Reza Katal , Rastegar
S.O., (2015), Batch and column removal of chromium
(VI) from aqueous solution using polypyrrole,
Environmental Engineering and Management Journal,
14, 17-28.
Hlihor R.M., Apostol L.C., Smaranda C., Pavel L.V.,
Caliman F.A., Robu B.M., Gavrilescu M., (2009),
Bioavailability processes for contaminants in soils and
their use in risk assessment, Environmental
Engineering and Management Journal, 8, 1199-1206.
Hu X.J., Wang J.S., Liu Y.G., Li X., Zeng G.M., Bao Z.L.,
Zeng X.X., Chen A.W., Long F., (2011), Adsorption
of chromium (VI) by ethylenediamine-modified cross-
linked magnetic chitosan resin: Isotherms, kinetics and
1633
 Abubeah et al./Environmental Engineering and Management Journal 17 (2018), 7, 1621-1634
thermodynamics, Journal of Hazardous Materials,
185, 306-314.
Ioannidou O., Zabaniotou A., (2007), Agricultural residues
as precursors for activated carbon production-A
review, Renewable and Sustainable Energy Reviews,
11, 1966-2005.
Isa M.H., Ibrahim N., Aziz H.A., Adlan M.N., Sabiani
N.H.M., Zinatizadeh A.A.L., Kutty S.R.M., (2008),
Removal of chromium (VI) from aqueous solution
using treated oil palm fibre, Journal of Hazardous
Materials, 152, 662-668.
Jin-fa C., Yong G., Da-gang S., Sha Y., (2018), Adsorption
characteristics of Cr (VI) on a carbonized Eupatorium
adenophorum SPRENG, Environmental Engineering
and Management Journal, 17, 115-122.
Jain M., Garga V.K., Kadirvelu K., (2009), Chromium (VI)
removal from aqueous system using Helianthus
annuus (sunflower) stem waste, Journal of Hazardous
Materials, 162, 365-372.
Karthikeyan T., Rajgopal S., Miranda L.R., (2005),
Chromium (VI) adsorption from aqueous solution by
Hevea brrasilinesis sawdust activated carbon, Journal
of Hazardous Materials, B124, 192-199.
Kiran B., Kaushik A., (2008), Chromium binding capacity
of Lyngbya putealis exopolysaccharides, Biochemical
Engineering Journal, 38, 47-54.
Kumar P.S., Kirthika K., (2009), Equilibrium and kinetic
study of adsorption of nickel from aqueous solution
onto bael tree leaf powder, Journal of Engineering
Science and Technology, 4, 351-363.
Kun-yu Z., Hui-ping H., Li-juan Z., AQi-yuan C., (2008),
Surface charge properties of red mud particles
generated from Chinese diaspore bauxite, Transaction
of Nonferrous Metals Society of China, 18, 1285-1289.
Malkoc E., Nuhoglu Y., (2007), Determination of kinetic
and equilibrium parameters of the batch adsorption of
Cr(VI) onto waste acorn of Quercus ithaburensis,
Chemical Engineering and Processing, 46, 1020-
1029.
Moussavi G., Khosravi R., (2011), The removal of cationic
dyes from aqueous solutions by adsorption onto
pistachio hull waste, Chemical Engineering Research
and Design, 89, 2182-2189.
Najm T., Yassin S., (2009), Removal of chromium from
aqueous solution using modified pomegranate peel:
mechanistic and thermodynamic studies, E-journal of
Chemistry, 6, S153-S158.
Natarajan B., Nagarajan S., (2010), Utilisation of
agriculture weed for the removal of Cr(VI) from
aqueous solution, Acta Chimica Slovenica, 57, 693-
699.
Ozdes D., Duran C., Senturk H.B., (2011), Adsorptive
removal of Cd(II) and Pb(II) ions from aqueous
solutions bu using Turkish illitic clay, Journal of
Environmental Management, 92, 3082-3090.
1634
Park D., Lim S.L., Yun Y.S., Park J.M., (2008),
Development of a new Cr (VI)-biosorbent from
agricultural biowaste, Bioresource Technolology, 99,
8810-8818.
Pavan F.A., Lima E.C., Dias S.L.P., Mazzocato A.C.,
(2008), Methylene blue biosorption from aqueous
solutions by yellow passion fruit waste, Journal of
Hazardous Materials, 150, 703-712.
Predescu A. Matei E., Predescu A., Berbecaru A., Sohaciu
M., Predescu C., (2016), Removal of hexavalent
chromium from waters by means of a TiO2-Fe3O4
nanocomposite, Environmental Engineering and
Management Journal, 15, 989-994.
Rafatullah M., Sulaiman O., Hashim R., Ahmad A., (2010),
Adsorption of methylene blue on low-cost adsorbents:
a review, Journal of Hazardous Materials, 177, 70-80.
Sain M., Panthapulakkal S., (2006), Bioprocess preparation
of wheat straw fibers and their characterization,
Industrial Crops and Products, 23, 1-8.
Sarin V., Pant K.K., (2006), Removal of chromium from
industrial wastewater by using eucalyptus bark,
Bioresource Technology, 97, 15-20.
Sears G.W., (1956), Determination of specific surface area
of colloidal silica by titration with sodium hydroxide,
Analytical Chemistry, 28, 1981-1983.
Sadeghi S., Moghaddam M.R., Arami M., (2017), Techno-
economical evaluation of hexavalent chromium
removal by electrocoagulation process with the aid of
polyaluminum chloride as coagulant: optimization
through response surface methodology, Environmental
Engineering and Management Journal, 16, 93-104.
Štrkalj A., Glavaš Z., Brnardić I., (2013), Application of
foundry waste for adsorption of hexavalent chromium,
Chemical and Biochemical Engineering, 27, 15-19.
Sun X.F., Xu F., Sun R.C., Fowler P., Baird M.S., (2005),
Characteristics of degraded cellulose obtained from
steam-exploded wheat straw, Carbohydrate Research,
340, 97-106.
Thamilarasu P., Sharmila R., Karunakaran K., (2012),
Kinetic, equilibrium and thermodynamic studies on
removal of Cr(VI) by modified activated carbon
prepared from Ricinus communis seed shell,
Coromandal Journal of Science, 1, 30-39.
Vargas C., Brandᾶ P.F.B., Ágreda J., Castillo E., (2012),
Bioadsorption using compost: An alternative for
removal of chromium (VI) from aqueous solutions,
BioResources, 7, 2711-2727.
Vijayaraghavan K., Jegan J., Palanivelu K., Velan M.,
(2005), Biosorption of cobalt (II) and nickel (II) by
seaweeds: batch and column studies, Separation and
PurificationTechnology, 44, 53-59.
Yang H., Yan R., Chen H., Lee D.H., Zheng C., (2007),
Characteristics of hemicellulose, cellulose and lignin
pyrolysis, Fuel, 86, 1781-1788.
Yasemin B., Tez Z., (2007), Adsorption studies on ground
shells of hazelnut and almond, Journal of Hazardous
Materials, 149, 35-41.
 Environmental Engineering and Management Journal June 2018, Vol.17, No.7, 1635-1644
http://www.eemj.icpm.tuiasi.ro/; http://www.eemj.eu

“Gheorghe Asachi” Technical University of Iasi, Romania

VERTICAL DISTRIBUTION AND POTENTIAL MOBILITY

OF HEAVY METALS IN NEW AND OLD TAILINGS
OF A LEAD/ZINC SULFIDE MINE

Malik Anju

Department of Energy and Environmental Sciences, Chaudhary Devi Lal University, Sirsa-125055, Haryana, India
E-mail: anjumalik27@yahoo.com; Phone: +91-9466733030

Abstract

Sulfide bearing mine tailings generated through mining and processing of mineral ores is one of the serious problems around the
world due to their potential environmental hazards. Current study focuses on vertical distribution and potential mobility of metals
in New and Old tailings of a Lead and Zinc mining area of India. Mine tailing profiles sampled from New (0-500mm) and Old
tailing (0-400mm) dams were chemically (pH, EC, Carbonate content, Cd, Pb, Zn and Mn levels) and mineralogically (using X-
ray Diffraction) characterized. Both the new and old tailings are alkaline having high carbonate content. However, the pH of old
tailings is less than the new tailings. The vertical profile distribution of metals in new tailings indicate a decrease in Pb levels with
depth up to 200 mm and a substantial increase in Zn and Cd levels were noted below 200 mm depth. However, no differences were
observed for Mn levels with depth. In case of Old Tailings, no discernible pattern for Cd, Pb, Zn and Mn levels with depth was
observed. These chemical and mineralogical observations confirm that the new tailings are undergoing oxidation and weathering
in the surface layers, however, the old tailings are sort of stabilized at least up to sampling depth. To estimate the potential mobility
of Cd, Pb and Zn in New and Old tailings, single extraction (EDTA 0.05 mol/L) was applied to surface tailings. The correlation
coefficients between total and EDTA extractable Cd and Zn (r values for Cd and Zn being 0.965 and 0.976, respectively, and P
value being 0.001) indicate statistically significant relation. The results of EDTA extraction show that out of the three metals, Pb
has maximum extractability in new tailings, whilst Cd has the maximum mobility in old tailings, and Zn has the least potential
mobility in both types of tailings.

Key words: correlation coefficient, heavy metal, mine tailing, single extraction, vertical distribution

Received: March, 2014; Revised final: August, 2014; Accepted: September, 2014; Published in final edited form: June 2018

1. Introduction water and oxygen, the resulting oxidative weathering

Mining and milling operations are the
important point source of trace elements in the
environment (Adriano, 1986). Mining and
beneficiation processes generate four categories of
large-volume waste: mine waste (overburden, barren
rocks), tailings, dump heap leach, and mine water
(Dudka and Adriano, 1997). Extraction and
processing of base and precious metals from the
sulfide ore deposits generate large volumes of sulfide
tailings (Moncur et al., 2015). Sulfide minerals are
thermodynamically unstable in presence of oxygen
and water. When sulfidic mine tailings are exposed to
of sulfides can generate H+ and release sulfate and
metal(oid)s to tailing pore waters (Blowes et al., 2013;
Lindsay et al., 2015). Depending on the nature of ore
deposits, metals such as Cu, Zn, Pb, Ni, Cd, Co, Hg,
Al, Mn and U and metalloids including As, Sb and Se
are released in mine drainage (Nordstrom, 2011).
Sulfide-bearing mine tailings are a serious
environmental problem around the world (Schuwirth
et al., 2007). The extent of environmental impacts
associated with sulfide tailings deposits depends upon
their mineralogical and geochemical composition, and
upon in situ chemical, biological, and physical
processes (Lindsay et al., 2015). Mine tailings may
 Anju/Environmental Engineering and Management Journal 17 (2018), 7, 1635-1644
weather and leach metals into tailing profile and
ground water for several hundreds of years after
mining activities have ceased. So, field studies of
metals in vertical profile of tailings can be useful for
predicting the pollution risk to groundwater from the
sulfide tailings.
The shift in paradigm from mere determination
of total elemental concentration to a more advanced
fractionation-based approach points to the general
recognition that information about the physiochemical
forms of trace elements is required for understanding
their environmental fate and behavior, including
mobility, pathways and bioavailability. A variety of
standardized leaching tests with variable reagents, and
experimental conditions are used to mimic differing
environmental conditions. Strong acids are used to
determine total concentrations of potentially harmful
elements, whereas weaker solutions are used to
determine compounds that are relatively soluble and
bioavailable (Salminen and Sipila, 1996). Various
extraction schemes have been developed for
discerning individual geochemical phases, ranging
from simplistic methods for distinguishing the “labile”
and “residual” fractions (single extractions), to more
sophisticated approaches enabling the sequential
removal of water soluble, exchangeable, adsorbed,
carbonate bound, phosphate, Fe/Mn oxide bound,
sulphide and organic bound and silicate fractions
(sequential extraction procedures). In single
extractions a large spectrum of extractants have been
used to extract the “mobile” or “bioavailable” forms
of metals (He and Singh, 1995; Karczewska et al.,
1998), viz.
 Very strong acids such as aqua regia (Vidal et
al., 1999), nitric acid (Sutherland et al., 2001), HCl
(Filipek and Pawlowski, 1990; Rieuwerts et al., 2000;
Sutherland and Tack, 2000; Sutherland et al., 2001;
Taylor et al., 1993).
 Chelating agents like EDTA (Anju and
Banerjee, 2011; Diaz-Barrientos et al., 1999; Filipek
and Pawlowski, 1990; McGrath, 1996; Sanders et al.,
1987; Sutherland et al., 2001; Tokalioglu and Kartal,
2005; Vidal et al., 1999), DTPA (Taylor et al., 1993),
CH3COOH (Filipek and Pawlowski, 1990; Vidal et
al., 1999).
 Buffered salt solutions e.g. ammonium acetate
(pH 7) (Ure et al., 1993).
 Unbuffered salt solutions e.g. CaCl2 (Diaz-
Barrientos et al., 1999; Sanders and Adams, 1987;
Vidal et al., 1999), NaNO3 (Gupta and Aten, 1993;
Hani and Gupta, 1982), BaCl2 (Juste and Solda, 1988)
etc.
These extractants are hypothesized to extract
exchangeable or weakly bound “available” metals in
the solid matrices viz. soils, sediments etc., which are
believed to be bioavailable. The content of extractable
forms of heavy metals is a more precise index of their
ability to migrate (Anderson, 1976; Haynes and Swift,
1986; Wiklander and Wahtras, 1977).
The present study is centered on Zawar, a
historical Zn-Pb mining community in India. The
relics of the historical mining can be seen as hillocks
1636
surrounding old surface and subsurface mine
workings, earthen retorts and slag heaps. Modern
mining, however, commenced in 1942. Zawar lies
within a belt of Aravalli system. The ore is confined to
the main dolomite horizon (Chakrabarti, 1967; Smith,
1964). Sphalerite (ZnS) is the predominant sulphide
mineral in the ore, second most abundant being pyrite
in a gangue of dolomite and quartz. Galena (PbS) is
more locally concentrated. Native silver, chalcopyrite,
arsenopyrite, pyrolite, and magnetite are also known
to occur (Mookherjee, 1964). The conventional froth-
flotation technology is adopted to produce separate
lead and zinc concentrates. On an average the ore
contains 5% metal (both Pb and Zn) and 95% of the
total mined ore is pumped out as tailings. Fine-grained
tailings generated during the beneficiation of sulphide
ores are directed to the New Tailing Dam. Prior to
1980's, tailings were dumped in an area called Old
Tailing Dam, which has been stabilized by plantation.
The volume of tailings generated is normally in far
excess of the liberated resource and the tailings often
contain potentially hazardous contaminants (Kossoff
et al., 2014). Thus, both tailing dams represent by far
the biggest accumulations of environmentally
dangerous waste materials in Zawar area.
The purpose of this study was to investigate the
vertical distribution of metals (viz. Cd, Pb, Zn and Mn)
in tailing profiles from New and Old Tailing Dams.
Additionally, the potential mobility of Cd, Pb and Zn
in New as well as Old tailings was estimated by using
single extraction (EDTA 0.05 mol/L).
2. Material and methods
2.1. Sample collection and preparation
Tailing profiles were sampled as undisturbed
core, by using a specially designed cylindrical
stainless-steel corer from New tailing dam (n = 5,
NTD1 to NTD5, 0-500mm) and Old tailing dam (n =
4, OTD1 to OTD4, 0-400mm). Soon after collection,
the core samples were cut into 100mm sections and
carefully transferred to clean and dry self-sealing
polyethylene bags for transport to laboratory. The
tailings were air-dried at room temperature,
disaggregated with a wooden roller and then sieved
through 2 mm sieve. The samples were homogenized
by coning and quartering and stored in tightly sealed
polyethylene bags for further analysis.
2.2. Analytical techniques
The pH and electrical conductivity (EC) of
the tailings were measured in solid/liquid ratio of 1:
2.5 with distilled water (Okalebo et al., 1993). The
carbonate content analysis involved the dissolution of
carbonates in an excess of standard hydrochloric acid
and back titration of excess acid with standard sodium
hydroxide (Rowell, 1994). Total metal content of the
tailings was determined after tri-acid digestion (using
a combination of HClO4, HNO3 and HF) as described
by Agemian and Chau (1975).
 Vertical distribution and potential mobility of heavy metals in new and old tailings of a lead/zinc sulfide mine
Only surface (0-100mm) samples from both
the locations were considered for single extraction
analysis. Measurement and Testing Programme
(formerly BCR) of the European Commission had
launched a project aiming at harmonizing
measurements for extractable trace metal contents in
soil. EDTA, acetic acid and ammonium acetate with
DTPA were the three extractants selected for
harmonization and validation of single extractions
(Quevauviller et al., 1997). EDTA 0.05 mol/L (pH 7)
has been used for the certification of soil reference
materials by BCR.
This test is assumed to extract both carbonate
bound and organically bound fractions of metals and
was hence considered to be suitable for calcareous soil
analysis (Quevauviller, 1998). To assess the potential
mobility of Cd, Pb and Zn in tailings, a single
extraction scheme using EDTA as described by
Quevauviller (1998) was preferred for the present
work as it has been extensively tested in previous
studies also. All the vessels in contact with samples or
reagents were cleaned by soaking in HNO3 4 Mol/L
(overnight) and rinsed repeatedly with distilled water.
EDTA extractions were conducted in triplicate in 100
mL centrifugation tubes with tight lids by using a
mechanical end-over-end shaker. All chemicals used
were of analytical grade, and double distilled or Milli.
Q. water was used throughout.
Cadmium, lead, zinc and manganese in
digests as well as extracts were determined by flame
atomic absorption spectrometry (FAAS) using a
Shimadzu AA-6800 system with air/acetylene flame.
Calibrants were prepared from 1000 µg/mL
Spectrosol standard solutions from Merck as well as
from standard solutions prepared in laboratory
following APHA (1995) standard methods. Multi-
level standards were prepared in the same matrix as
the extracting reagent to minimize matrix effects.
Blanks were run simultaneously for background
correction and other sources of error. Mineralogical
analysis of New and Old tailings was done by using
Philips X’Pert model XRD instrument. Samples were
scanned from 50 to 900 2θ. The peak analysis was done
by using X’Pert data collector.
3. Results and discussion
3.1. Physicochemical and mineralogical analysis
pH, EC, Carbonate content, Cd, Pb, Zn and
Mn levels in surface tailings (0-100mm) have already
been discussed in previous publication (Anju and
Banerjee, 2010). Briefly, the pH of both type of
tailings was alkaline, which contrasts with the other
studies in similar type of metal mining areas, where
the pH is usually acidic (Grimalt et al., 1999; Morrell
et al., 1996; Shu et al., 2001; Simon et al., 1999). In
general, sulfide rich and carbonate-poor materials
produce acidic conditions in areas having metal
sulphide as the ores. In contrast, alkaline-rich
materials, even with significant sulfide concentrations,
often produce alkaline conditions (Skousen et al.,
2000). The pH of old tailings was less than the new,
which could be because of the ability of sulphide
tailings to generate acid upon weathering. The
carbonate content of both the New and Old tailings
was quite high, indicating calcareous nature of these
tailings and hence the alkaline pH.
The color of the new tailings was gray,
sometimes brownish silty material, but the colour of
old tailings was always yellowish brown. The
formation of efflorescent salts (ocherous and white
deposition) on the surfaces of old tailings was also
observed. Mineralogical analysis indicated the
presence of Dolomite, Quartz, Ankerite and other
sulphide minerals in New and Old tailings.
3.2. Profile distribution of metals in tailings
The flotation tailings profiles were
investigated for Cd, Pb, Zn and Mn levels at two
dumping sites called New tailing dam and Old tailing
dam, which represent by far the biggest accumulations
of environmentally dangerous waste materials in the
area. The mean concentrations of Pb, Zn, Mn and Cd
in New tailings (0-100mm, n=5) were 1099, 1908,
2932 and 13.96 mg kg-1 and in Old tailings (0-100mm,
n=4) were 1903, 5098, 2271 and 22.93 mg kg-1,
respectively. The results indicate high levels of Cd,
Pb, Zn, and Mn in both types of tailings. The levels of
metals present in the tailings are comparable with the
levels observed in many previous studies on tailings in
similar kind of mining areas (Table 1).
The vertical profile distribution of Cd, Pb, Zn
and Mn in New Tailings (0-100, 100-200, 200-300,
300-400 and 400-500 mm) and Old Tailings (0-100,
100-200, 200-300 and 300-400 mm) is shown in Fig.
1(a-d) and Fig. 2(a-d), respectively. In case of new
tailings, the depth-wise distributions indicate a
decrease in Pb levels with depth up to 200 mm and a
substantial increase in Zn and Cd levels were noted
below 200 mm depth. However, no differences were
observed for Mn levels with depth.
The contrasting trend noted for Pb is
explained by the formation of highly insoluble
anglesite (PbSO4) in the surface layer as confirmed by
XRD analysis of surface samples. Shu et al. (2001)
have also observed significantly lower concentrations
of Zn and Cd in surface (0-10cm) and subsurface layer
(10-20cm), than those of deep layers (20-50cm and
50-100cm) in their study on Pb/Zn mine tailings at
Lechang, China. Morrell et al. (1996) in their study on
mine tailings from New Zealand and Hofmann and
Schuwirth (2008) in a study on sphelarite bearing
mine waste tailings from Adolf-Helene mine,
Germany have also observed increase in Zn and
decrease in Pb concentration with depth.
The occurrence of anglesite in surface layer
and a substantial increase in concentration of Zn and
Cd below 200mm depth are two chemical and
mineralogical observations confirming that oxidation
and weathering have been largely confined to surface
tailings in new tailing dam.
1637
 Anju/Environmental Engineering and Management Journal 17 (2018), 7, 1635-1644

Table 1. Heavy metal levels in surface tailings (µg/g) of mining areas as reported in literature

Pb Zn Cd Cu As
Location Mean Mean Mean Mean Mean Reference
Range Range Range Range Range
(Median) (Median) (Median) (Median) (Median)
Adolf-Helena Schuwirth
mine, Germany 27 10 0.035 3.5 0.5 et al., 2007
(former tailing mmol/kg mmol/kg mmol/kg mmol/kg mmol/kg
pond)
Aznalcollar Lacal et al.,
300mg/k 2.05 5.33
spill, Spain 7.05mg/g 14 mg/g 2003
g mg/g mg/g
Boleslaw, Krzaklewsk
Cracow, Poland 735- 1578- i&
4039 1.5% 9-371 1-390
19555 36000 Pietrzykow
ski, 2002
Wiesloch, Stuben et
188mg/k
Germany 0.72% 3.13% - 2.1% al., 2001
g
Lechang Pb/Zn Shu et al.,
2784 2029 12 132
mine, China 2001
Aznalcollar Alastuey et
spill, pyrite al., 1999
7141 9448 31 1968
mud, Agrio
river, Spain
Pyrite mine Simon et
Aznalcollar, (9507.3) (7183.5) (28.6) (3749) al., 1999
Spain
Baxter Springs Zhu et al.,
in Cherokee 1999
County, 351 17850 125
southeast
Kansas
Aznalcollar, Vidal et al.,
Spain 7957 10707 40 2126 5321 1999

Aznalcollar, Lopez-
Spain 8091 8832 33 1323 4692 Pamo et al.,
1999
Guadiamar, Cabrera et
Spain 7888 7096 25.1 1552 2878 al., 1999

Montevecchio(S Fanfani et
ardinia) Coarse 17200 16400 150 700 - al., 1997
material (C)
Montevecchio(S Fanfani et
ardinia) fine 20900 21300 140 570 - al., 1997
material (F)
Montevecchio(S Fanfani et
ardinia) al., 1997
19000 16300 120 520 -
Very fine
material (VF)
Tui base metal Morrell et
mine, Te Aroha, 10568 486 113 al., 1996
New Zealand
Southeast Abdel-
Kansas 1370 13700 90 Saheb et
al., 1994
Old Lead Belt, Clevenger,
Missouri (400 1990
3867 1329 33.9 88.6
mesh size)

Gold mine, Andrade

Serra de Santa Lima et al.,
21 55 0.7 87 71
Cruz(Bahia, 2008
Brazil)
Goldmine, Zaranyika
Zimbabwe 1750±10 and
- 225±20 17.6±0.1
0 Chirinda,
2011
*Units are µg/g unless otherwise mentioned

1638
 Vertical distribution and potential mobility of heavy metals in new and old tailings of a lead/zinc sulfide mine

However, no discernible pattern for Cd, Pb, Zn
and Mn levels with depth (Fig. 2a-d) was observed for
Old Tailing Profile (except profile 4), indicating that
these tailings are sort of stabilized for weathering at
least upto a depth of 400 mm. For profile 4, an increase
in Cd, Pb and Zn concentration was observed at 30-
40cm depth. This indicated that acidification
accelerated metal solubility, which resulted in
depletion of Cd, Pb and Zn in surface tailings and
relative enrichment in deep layers (30-40cm).

(a)

(b)

(c)

(d)

Fig. 1. (a) Depthwise distribution of Cd in New Tailings; (b) Depthwise distribution of Pb in New Tailings;
(c) Depthwise distribution of Zn in New Tailings; (d) Depthwise distribution of Mn in New Tailings

1639
 Anju/Environmental Engineering and Management Journal 17 (2018), 7, 1635-1644

(a)

(b)

(c)

(d)

Fig. 2.(a) Depthwise distribution of Cd in Old Tailings; (b) Depthwise distribution of Pb in Old Tailings;
(c) Depthwise distribution of Zn in Old Tailings; (d) Depthwise distribution of Mn in Old Tailings

3.3. Single extraction analysis nutritional deficiency effects) and environmentally

3.3.1. EDTA extraction
Single extraction tests are commonly used to
study the eco-toxicity and mobility of metals in solid
matrices, e.g. to assess the bioavailable metal fraction
(and thus to estimate the related phyto-toxic and
accessible trace metals upon disposal of sediment or
wastes on a soil (thus assessing contamination of
ground waters). EDTA has been extensively used for
determination of extractable metal content of polluted
and unpolluted soils (Anju and Banerjee, 2011;
McGrath, 1996; Schramel et al., 2000; Tokalioglu and

1640
 Vertical distribution and potential mobility of heavy metals in new and old tailings of a lead/zinc sulfide mine

Kartal, 2005; Viro, 1955), road sediments (Sutherland
et al., 2001), municipal and industrial sludges (Lo and
Chen, 1990) and is increasingly being used in the
reclamation of contaminated soils and sediments
(Cline and Reed, 1995; Mench et al., 1994; Sun et al.,
2001). Table 2 shows the EDTA extractable
concentrations and percentage of total metal extracted
by EDTA in both type of tailings. Relative EDTA
extractability of Cd, Pb and Zn in New tailings (Table
3) followed the order: Pb (96.53 ± 4.43) >>Cd (55.98
± 4.81) >>Zn (33.16 ± 7.57). The EDTA extractability
was maximum for Pb followed by Cd in all the
profiles. Out of the three metals, Zn was least
mobilizable. Vidal et al. (1999) in their study on
Aznalcollar (Spain) toxic spill have also observed
similar order i.e. Pb>Cd>Zn, of EDTA extractability
of metals from mine sludge (tailings). Many other
investigations have demonstrated the ability of EDTA
to extract a large proportion of total soil Pb (Abdel-
Saheb et al., 1994; McGrath and Cegarra, 1992;
Pichtel et al., 2001; Sposito et al., 1982; Tsadilas et al.,
1995). Clayton and Tiller (1979) observed that EDTA
extracted 74-93% of total Pb from soils contaminated
by smelter emissions, but only 20-60%
uncontaminated soils.
Relative EDTA extractability of Cd, Pb and Zn
in Old tailings (Table 3) followed the order: Cd (61.94
± 25.09) ≥ Pb (55.42 ± 10.54) > Zn (36.23 ± 6.14). The
relative extractability of these metals is consistent with
the findings of the study of Anju and Banerjee (2011)
on soils of a lead and zinc mining area, India and
McGrath (1996)’s on Irish soils, where EDTA
extractability followed the sequence
Cd>Cu=Pb>Ni=Zn>Cr. EDTA extracts of soils tend
to correlate well with plant contents, in particular with
the plant-available fraction for Cd, Cu, Ni, Pb and Zn
(Sanders et al., 1986), Cd (Lag and Elsokkary, 1978)
and Cu (Sanders et al., 1987). Therefore, complexants
like EDTA are frequently used to extract mobilizable
(potentially bioavailable or potentially leachable)
metal fraction in soil (Gupta et al., 1996).
3.3.2. Correlation
Correlation coefficients between the EDTA
extractable Cd, Pb and Zn vs their ‘total’ elemental
contents in Tailings are given in Table 4. The amount
of Cd, Pb and Zn extracted by EDTA and their total
concentrations showed a linear positive correlation,
which are statistically significant (r values for Cd, Pb
and Zn being 0.965, 0.638 and 0.976, respectively and
P values being <0.001). The correlation coefficients
indicate a strong relation between EDTA extractable
and total Cd and Zn. Anju and Banerjee (2011) has
also observed statistically significant linear positive
correlations between EDTA extractable and total
concentrations of Cd, Pb and Zn (r values for Cd, Pb
and Zn being 0.901, 0.971 and 0.795, respectively, and
P values being <0.001), in soils of a lead and zinc
mining area, India. These findings are in agreement
with other studies such as McGrath’s (1996) work on
Cd, Cu and Ni; Boon and Soltanpour’s (1991) work
on Cd, Zn and Pb and Mehra et al.’s (1999) work on
Cd and Cu for contaminated soils. These results can
appear to justify the use of ‘total’ metal contents as a
useful preliminary indicator of areas where the risks of
metal excess or deficiency are high.

Table 2. EDTA extractable concentrations (mg kg-1) and percentages of total Cd, Pb and Zn in New and Old surface tailings

Cd Pb Zn
Sample ID
EDTA-Extractable Total % Extracted EDTA-Extractable Total % Extracted EDTA-Extractable Total % Extracted
NTD1 3.85 6.98 55.16 735 768 95.70 355 1227 28.93
NTD2 7.06 12.96 54.48 979 987 99.19 597 1990 30.00
NTD3 9.81 19.94 49.20 1379 1525 90.43 754 3069 24.57
NTD4 10.41 16.95 61.42 1186 1246 95.18 727 1758 41.35
NTD5 7.74 12.96 59.72 992 971 102.16 613 1498 40.92
OTD1 3.33 13.46 24.74 747 1584 47.16 956 3538 27.02
OTD2 13.81 18.44 74.89 799 1691 47.25 1686 4276 39.43
OTD3 14.19 20.44 69.42 1070 1845 57.99 1785 4588 38.91
OTD4 30.99 39.38 78.69 1727 2493 69.27 3161 7991 39.56

Table 3. Relative mobility of Cd, Pb and Zn in New and Old tailings as estimated by single extraction

Matrix Type Relative EDTA extractability and mobility of Cd, Pb and Zn

New Tailings Pb (96.53 ± 4.43) >>Cd (55.98 ± 4.81) >>Zn (33.16 ± 7.57)
Old Tailings Cd (61.94 ± 25.09) ≥ Pb (55.42 ± 10.54) > Zn (36.23 ± 6.14)

Table 4. Correlation coefficients between amount of Cd, Pb and Zn extracted by EDTA and their “Total” content

EDTA-extractable vs ‘total’
Element
( n =9)
Cd r =0.965 **
Pb r = 0.638
Zn r = 0.976**
** All correlations are significant at the 0.01 level (2-tailed)

1641
 Anju/Environmental Engineering and Management Journal 17 (2018), 7, 1635-1644
4. Conclusions
From this study, it is concluded that new and
old tailings are alkaline and have high carbonate
content. Also, the pH of old tailings is less than the
New tailings, which could be because of H+ ions
produced during mine tailing weathering.
Chemical and mineralogical observations
confirm that the new tailings are undergoing oxidation
and weathering in the surface layers. However, the old
tailings are sort of stabilized at least up to the sampling
depth. The response of the elements to single
extraction (EDTA 0.5mol/L) enabled us to estimate
their potential mobility in the tailings.
Based on single extraction, Zn has the least
potential mobility in both types of tailings, whilst Cd
in old tailings and Pb in new tailings have the highest
potential mobility. Present study represents a first
approach for a further environmental assessment of
these environmentally hazardous tailings and possible
reclamation efforts.
Acknowledgements
The grants of scholarship awarded by the University Grants
Commission (UGC) and Council of Scientific and Industrial
Research (CSIR) at different stages of this work are
gratefully acknowledged.
References
Abdel-Saheb I., Schwab A.P., Banks M.K., Hetrick B.A.,
(1994), Chemical characterisation of heavy metal
contaminated soil in southeast Kansas, Water, Air, &
Soil Pollution, 78, 73-82.
Adriano D.C., (1986), Trace Elements in the Terrestrial
Environment, Springer-Verlag, New York.
Agemian H., Chau A.S.Y., (1975), An atomic absorption
method for the determination of 20 elements in. lake
sediments after acid digestion, Analytica Chimica Acta,
80, 61-66.
Alastuey A., Garcia-Sanchez Lopez F., Querol X., (1999),
Evolution of pyrite mud weathering and mobility of
heavy metals in the Guadiamar valley after the
Aznalcollar spill, south-west Spain, Science of the Total
Environment, 242, 42-55.
Anderson A., (1976), On the determination of the
ecologically significant fraction of some heavy metals
in soils, Swedish Journal of Agricultural Research, 6,
19-25.
Andrade Lima L.R.P., Bernardez L.A., Barbosa L.A.D.,
(2008), Characterization and treatment of artisanal gold
mine tailings, Journal of Hazardous Materials, 150,
747-753.
Anju M., Banerjee D.K., (2010), Comparison of two
sequential extraction procedures for heavy metal
partitioning in mine tailings, Chemosphere, 78, 1393-
1402.
Anju M., Banerjee D.K., (2011), Associations of cadmium,
zinc, and lead in soils from a lead and zinc mining area
as studied by single and sequential extractions,
Environmental Monitoring and Assessment, 176, 67-
85.
APHA, (1995), Standard Method for the Examination of
Water and Wastewater, 19th Edition, Washington DC.
1642
Blowes D.W., Ptacek C.J., Jambor J.L., Weisener C.G.,
Paktunc D., Gould W.D., Johnson D.B., (2013), The
Geochemistry of Acid Mine Drainage, In:
Environmental Geochemistry. Treatise on
Geochemistry, Lollar, B.S., Holland, H.D., Turekian,
K.K. (Eds.), vol. IX, 2nd Edition, Elsevier-Pergamon,
Oxford, 131-190.
Boon D.Y., Soltanpour P.N., (1991), Estimating total Pb, Cd
and Zn in contaminated soils from NH4HCO3-DTPA-
extractable levels, Communications in Soil Science and
Plant Analysis, 22, 369-378.
Cabrera F., Clemente L., Diaz Barrientos E., Lopez, R.,
Murillo, J.M., (1999), Heavy metal pollution of soils
affected by the Guadiamar toxic flood, Science of the
Total Environment, 242, 117-129.
Chakrabarti A.K., (1967), Genesis of the lead-zinc deposits
at Zawar, Rajasthan, India, Economic Geology, 62,
554-556.
Clayton P.M., Tiller K.G., (1979), A Chemical Method for
the Determination of Heavy Metal Content of Soils in
Environmental Studies, Commonwealth Scientific and
Industrial Research Organization, CSIRO, Australia.
Clevenger T.E. (1990). Use of sequential extraction to
evaluate the heavy metals in mining wastes, Water, Air,
& Soil Pollution, 50, 241-254.
Cline S.R., Reed B.E., (1995), Lead removal from soils via
bench-scale soil washing techniques, Journal of
Environmental Engineering, 121, 700-705.
Diaz-Berrientos E., Madrid L., Cardo I., (1999), Effect of
flood with mine wastes on metal extractability of some
soils of Guadiamar River basin (SW Spain), Science of
the Total Environment, 242, 149-165.
Dudka S., Adriano D., (1997), Environmental impacts of
metal ore mining and processing: a review, Journal of
Environmental Quality, 26, 590-602.
Fanfani L., Zuddas P., Chessa A., (1997), Heavy metals
speciation analysis as a tool for studying mine tailings
weathering, Journal of Geochemical Exploration, 58,
241-248.
Filipek T., Pawlowski L., (1990), Total and extractable
heavy metal content of some soils of the Lublin coal
mining region, Science of the Total Environment, 96,
131-137.
Grimalt J.O., Ferrer M., Macpherson E., (1999), The mine
tailing accident in Aznalcollar, Science of the Total
Environment, 242, 3-11.
Gupta S.K., Aten C., (1993), Comparison and evaluation of
extraction media and their suitability in a sample model
to predict the biological relevance of heavy metal
concentrations in contaminated soils, International
Journal of Environmental Analytical Chemistry, 51, 25-
46.
Gupta S.K., Vollmer M.K., Krebs R., (1996), The
importance of mobile, mobilisable and pseudo total
heavy metal fractions in soil for three-level risk
assessment and risk management, Science of the Total
Environment, 178, 11-20.
Hani H., Gupta S., (1982), Environmental Effects of Organic
and Inorganic Contaminants in Sewage Sludges, R.D.
Davis, G. Hucker, P.L. Hermite, D. Ridel, (Eds.),
Dordrecht, 121.
Haynes R.J., Swift R.S., (1986), Effects of soil acidification
and subsequent leaching on levels of extractable
nutrients in a soil, Plant and Soil, 95, 327-336.
He Q.B., Singh, B.R., (1995), Cadmium availability to
plants as affected by repeated applications of
phosphorus fertilizers, Acta Agriculturae Scandinavica,
Section B, Soil and Plant Science, 45, 22-31.
 Vertical distribution and potential mobility of heavy metals in new and old tailings of a lead/zinc sulfide mine
Hofmann T., Schuwirth N., (2008). Zn and Pb release of
sphalerite (ZnS)-bearing mine waste tailings, Journal of
Soils & Sediments, 8, 433-441.
Juste C., Solda P., (1988), Changes in the Cadmium,
Manganese, Nickel and Zinc bioavailability of a
sewage sludge treated sandy soil as a result of
ammonium-sulphate, acid-peat, lime and iron
compound addition, Agronomie, 8, 897-904.
Karczewska A., Szerszen L., Kabala C., (1998), Forms of
selected heavy metals and their transformation in soils
polluted by the emissions from copper smelters. Adv.
GeoEcology, 31, 705-712.
Kossoff D., Dubbin W.E., Alfredsson M., Edwards S.J.,
Macklin M.G., Hudson- Edwards K.A., (2014), Mine
tailings dams: characteristics, failure, environmental
impacts, and remediation, Applied Geochemistry, 51,
229–245.
Krzaklewski W., Pietrzykowski M., (2002), Selected
physico-chemical properties of zinc and lead ore
tailings and their biological stabilization, Water, Air, &
Soil Pollution, 141, 125-142.
Lacal J., Pilar de Silva M., Garcia R., Teresa Sevilla M.,
Procopio J.R., Hernandez L., (2003), Study of
fractionation and potential mobility of metal in sludge
from pyrite mining and river sediment: changes in
mobility over time and use of artificial ageing as a tool
in environmental impact assessment, Environmental
Pollution, 124, 291-305.
Lag J., Elsokkary I.M., (1978), A comparison of chemical
methods for estimating cadmium, lead and zinc
availability to six food crops grown in industrially
polluted soils of Odda, Norway, Acta Agricultuare
Scandinavica, 28, 76-80.
Lindsay M.B.J., Moncur M.C., Bain J.G., Jambor J.L.,
Ptacek C.J., (2015), Geochemical and mineralogical
aspects of sulfide mine tailings, Applied Geochemistry,
57, 157-177.
Lo K.S.L., Chen Y.H., (1990), Extracting heavy metals from
municipal and industrial sludges, Science of the Total
Environment, 90, 99-116.
Lopez-Pamo E., Barettino D., Anton-Pacheco C., Ortiz G.,
Arranz J.C., Gumiel J.C., Martinez-Pledel B., Aparicio
M., Montouto O., (1999), The extent of the Aznalcollar
pyritic sludge spill and its effects on soils, Science of
the Total Environment, 242, 57-88.
McGrath D., (1996), Application of single and sequential
extraction procedures to polluted and unpolluted soils,
Science of the Total Environment, 178, 37-44.
McGrath S.P., Cegarra J., (1992), Chemical extractability of
heavy metals during and after long – term applications
of sewage sludge to soil, Journal of Soil Science, 43,
313-321.
Mehra A., Cordes K.B., Chopra S., Fountain D., (1999),
Distribution and bioavailability of metals in soils in the
vicinity of a copper works in Staffordshire, UK,
Chemical Speciation & Bioavailability, 11, 57-66.
Mench M.I., Didier V.L., Loffler V.L., Gomez A., Masson
P., (1994), A mimicked in situ remediation study of
metal-contaminated soils with emphasis on cadmium
and lead, Journal of Environmental Quality, 23, 58-63.
Moncur M.C.V., Ptacek C.V.J., Lindsay M.B.J., Blowes
D.W., Jambor J.L., (2015), Long term mineralogical
and geochemical evolution of sulfide mine tailings
under a shallow water cover, Applied Geochemistry, 57,
178-193.
Mookherjee A., (1964), The geology of the Zawar lead–zinc
mine, Rajasthan, India, Economic Geology, 59, 656–
677.
Morrell W.J., Stewart R.B., Gregg P.E.H., Bolan N.S.,
Horne D., (1996), An assessment of sulphide oxidation
in abandoned base-metal tailings, Te Aroha, New
Zealand, Environmental Pollution, 94, 217-225.
Nordstrom D.K., (2011), Mine waters: acidic to
circumneutral, Elements, 7, 393–398.
Okalebo J.R., Kennth W.G., Woomer P.L., (1993),
Laboratory manual of soil and plant analysis: a working
manual, Soil Chemistry Laboratory, The Kenya
Agriculture Research Institute, National Agriculture
Research Centre, Muguga Kenya.
Pichtel J., Vine B., Kuula-Vaisanen P., Niskanen P., (2001),
Lead extraction from soils as affected by lead chemical
and mineral forms, Environmental Engineering
Science, 18, 91-98.
Quevauviller Ph., Rauret G., Lopez-Sanchez J.F., Rubio R.,
Ure A., Muntau H., (1997), Certification of trace metal
extractable contents in a sediment reference material
(CRM 601) following a three-step sequential extraction
procedure, Science of the Total Environment, 205, 223-
234.
Quevauviller Ph., (1998), Operationally defined extraction
procedures for soil and sediment analysis 1.
Standardization, Trends in Analytical Chemistry, 17,
289-298.
Rieuwerts J.S., Farago M., Cikrt M., Bencko V. (2000),
Differences in lead bioavailability between a smelting
and a mining area, Water, Air and Soil Pollution, 122,
203-229.
Rowell D.L., (1994), Soil Science: Methods and
Applications, Longman Group UK Ltd, Harlow, Essex,
UK, 32-34.
Salminen R., Sipila P., (1996), The environmental impact of
sulphide mines measured with organogenic sampling
media, Applied Geochemistry, 11, 277-283.
Sanders J.R., Trevor Mc. M.A., Christensen B.T., (1986),
Extractability and bioavailability of Zn, Ni, Cd and Cu
in three Danish soils sampled 5 years after application
of sewage sludge, Journal of the Science of Food and
Agriculture, 37, 1155-1164.
Sanders J. R., McGrath S.P., Adams T.McM., (1987), Zinc,
copper and nickel concentrations in soil extracts and
crops grown on four soil treated with metal loaded
sewage sludges, Environmental Pollution, 44, 193-210.
Sanders J.R., Adams T.McM., (1987), The effects of pH and
soil type on concentrations of zinc, copper and nickel
extracted by calcium chloride from sewage sludge-
treated soils. Environmental Pollution, 43, 219-228.
Schramel O., Michalke B., Kettrup A., (2000), Study of
copper distribution in contaminated soils of hop fields
by single and sequential extraction procedures, Science
of the Total Environment, 263, 11-22.
Schuwirth N., Voegelin A., Kretzschmar R., Hofmann T.,
(2007), Vertical distribution and speciation of trace
metals in weathering flotation residues of a zinc/lead
sulfide mine, Journal of Environmental Quality, 36, 61-
69.
Shu W.S., Ye Z.H., Lan C.Y., Zhang Z.Q., Wong M.H.,
(2001), Acidification of lead/zinc mine tailings and its
effect on heavy metal mobility, Environment
International, 26, 389-394.
Simon M., Ortiz I., Garcia I., Fernandez E., Fernandez J.,
Dorronsoro C., Aguilar J., (1999), Pollution of soils by
the toxic spill of a pyrite mine (Aznalcollar, Spain),
Science of the Total Environment, 242, 105-115.
Skousen J.G., Sexstone A., Ziemkiewicz P.F., (2000), Acid
Mine Drainage Control and Treatment, In:
Reclamation of Drastically Disturbed Lands,
Agronomy Monograph No. 9, American Society of
1643
 Anju/Environmental Engineering and Management Journal 17 (2018), 7, 1635-1644
Agronomy, Crop Science Society of America, Soil
Science Society of America, Madison, WI 53711, USA,
131-168.
Smith A.W., (1964), Remobilisation of sulfide ore bodies,
Economic Geology, 59, 930-941.
Sposito G., Lund L.J., Chang A.C., (1982), Trace metal
chemistry in arid-zone field soils amended with sewage
sludge: I. Fractionation of Ni, Cu, Zn, Cd and Pb in
solid phases, Soil Science Society of America Journal,
46, 260-264.
Stuben D., Berner Z., Kappes B., Puchlet H., (2001),
Environmental monitoring of heavy metals and arsenic
from Ag-Pb-Zn mining, Environmental Monitoring and
Assessment, 70, 181-200.
Sun B., Zhao F.J., Lombi E., McGrath S.P. (2001), Leaching
of heavy metals from contaminated soils using EDTA,
Environmental Pollution, 113, 111-120.
Sutherland S.A., Tack F.M.G., (2000), Metal phase
associations in soils from an urban watershed,
Honolulu, Hawaii, Science of the Total Environment,
256, 103-113.
Sutherland R.A., Tack F.M.G., Tolosa C.A., Verloo M.G.,
(2001), Metal extraction from road sediment using
different strength reagents: Impact on anthropogenic
contaminant signals, Environmental Monitoring and
Assessment, 71, 221-242.
Taylor R.W., Ibeabuchi I.O., Sistani K.R., Shuford J.W.
(1993), Heavy metal concentration in forage grasses
and extractability from acid mine spoils, Water, Air and
Soil Pollution, 68, 363-372.
Tokalioglu S., Kartal S., (2005), Comparison of metal
fractionation results obtained from single and
sequential extractions, Bulletin of Environmental
Contamination and Toxicology, 75, 180-188.
1644
Tsadilas C.D., Matsi T., Barbayiannis N., Dimoyiannis D.,
(1995), Influence of sewage sludge application on soil
properties and on the distribution and availability of
heavy metal fractions, Communications in Soil Science
and Plant Analysis, 26, 2603-2619.
Ure M., Thomas R., Litlejohn D., (1993), Ammonium
acetate extracts and their analysis for the speciation of
metal ions in soils and sediments, International Journal
of Environmental Analytical Chemistry, 51, 65-84.
Vidal M., Lopez-Sanchez J.F., Sastre J., Jimenez G., Dagnac
T., Rubio R., Rauret G., (1999), Prediction of the
impact of Aznalcollar toxic spill on the trace element
contamination of agricultural soils, Science of the Total
Environment, 242, 131-148.
Viro P.J., (1955), Use of ethylene diamine tetra acetic acid
in soil analysis: I. Experimental, Soil Science, 79, 459-
465.
Wiklander L., Wahtras K., (1977), Solubility and uptake of
heavy metals from Swedish soil, Geoderma, 19, 123-
129.
Zaranyika M.F., Chirinda T., (2011), Heavy metal
speciation trends in mine slime dams: A case study of
slime dams at a goldmine in Zimbabwe, Journal of
Environmental Chemistry and Ecotoxicology, 3, 103-
115.
Zhu D., Schwab A.P., Banks M.K., (1999), Heavy metal
leaching from mine tailings as affected by plants,
Journal of Environmental Quality, 28, 1727-1732
 Environmental Engineering and Management Journal July 2018, Vol.17, No. 7, 1645-1656
http://www.eemj.icpm.tuiasi.ro/; http://www.eemj.eu

“Gheorghe Asachi” Technical University of Iasi, Romania

SUPPLY CHAIN EMISSION REDUCTION OPTIMIZATION UNDER

CONSUMER CARBON SENSITIVITY AND CARBON TAX POLICY

Lan Bai1,2, Xianliang Shi1, Honghu Gao1, Qiwen Du1∗

1School of Economics and Management, Beijing Jiaotong University, Shangyuan 3#, Haidian District,Beijing, 100044,China
2Management School, Hebei Agricultural University, Jianshenan Road, Baoding, 071000, China

Abstract

This paper presents decentralized and centralized decision-making conditions based on consumer carbon sensitivity and carbon tax
policy in a supplier-driven two echelons supply chain. The analysis shows that: 1) the optimal emission reduction amount under
centralized decision-making is more than that under decentralized decision-making; 2) the carbon tax should be levied properly
because the carbon emission reduction will decrease as the carbon tax is higher than a critical value; 3) the consumer carbon
sensitivity is positively correlated with carbon emission reduction, but it may cause total mission rised in certain condition. Then
an extensive numerical analysis is conducted to enrich the discussion and to draw some managerial insights on how to reduce
carbon emission in the supply chain.

Key words: carbon emission reduction, consumer carbon sensitivity, carbon tax policy, Stackelberg game, supply chain

Received: January, 2018; Revised final: May, 2018; Accepted: June, 2018; Published in final edited form: July 2018

1. Introduction Scholars have discovered that when the products are

Since global warming has been widely
concerned by the governments and companies, people
concern more and more about carbon emission. Many
governments apply relevant policies to address the
issue of global warming. Amongst them, the carbon
tax is the government's specific tax on companies to
constrain carbon emissions. Owing to the obvious
effect and low cost of carbon tax collection, the United
States, Canada and Australia and other countries have
adopted it (Lin and Li, 2011). In addition, due to
global concern about environmental protection,
consumer demand for products is not only concerned
with the quality of products (Du et al., 2013). More
and more, consumers are paying more attention to the
social value and environmental value of products. In
response to the different preferences of consumers
with low carbon, the level of carbon emissions per unit
of product has gradually become one of the most
important attributes for consumers to pursue products.
sold within the form of carbon labels (Vanclay et al.,
2011). Environmentally friendly consumers are more
willing to make such purchases (Androniceanu and
Dragulanescu, 2016; Michaud et al., 2012). At the
same time, those consumers are more willing to pay
higher prices for low-carbon product, which means
that if companies reduce carbon emission in
production process, then they can get their products
better competitiveness in the market (Lu et al., 2007;
Zhou et al., 2017). The low-carbon production
decisions of companies in the production process have
become serious business decision-making problems.
On one hand, the government levying carbon taxes on
companies will increase the operating costs of
companies, while consumers' demand for low-carbon
products will prompt companies to produce
environmentally friendly products. On the other hand,
low-carbon investments by companies can reduce
carbon tax costs and meet the needs of
environmentally friendly consumers but increase costs

∗
Author to whom all correspondence should be addressed: e-mail: tobailan@foxmail.com; Phone: +8617713288335; Fax: +86312 7528673
 Bai et al./Environmental Engineering and Management Journal 17 (2018), 7, 1645-1656
in the manufacturing process. Therefore, how the
government levies the carbon tax and how they
produce a reasonable preventive solution is a problem
influenced both government and companies. Thus,
questions about how the carbon tax and the consumer
carbon sensitivity affect the carbon emission reduction
decisions of companies, how the government
formulates the optimal carbon tax considering
consumer carbon sensitivity have become urgent
problems to be solved.
In this paper, research focuses on the carbon
tax levied on companies and consumer preferences for
low-carbon products from the perspective of supply
chain; the relevant literature review is hence done
from two perspectives as following:
1) Carbon tax:
Carbon tax can induce companies to reduce
their carbon emissions (Bruvoll and Larsen, 2002;
Meng et al., 2013; Bjørner and Jensen, 2002), and
different carbon tax rates result in different carbon
emission reductions. In the supply chain, revenues of
companies change under variable decision-making
strategy. Tseng and Hung (2014) have established a
decision-making model for the sustainable
development of garment processing supply chain
network under the carbon tax policy, and the strategic
decision-making of companies under different carbon
tax policies. Comas Martí et al. (2015) take demand
uncertainty into account and includes decisions on
supply chain responsiveness under different carbon
policies for both functional and innovative products.
Yue and Shuaihua (2016) have analyzed the
manufacturers and supplier decisions under the carbon
tax policy when suppliers involved and not involved
in carbon emission and the impact of carbon tax on the
two kinds of decision-making. Yang and Luo (2016)
produce a model expressing both the emission
reduction decisions of manufacturers within carbon
tax constraints and the stochastic demand of the
wholesale price of companies and their revenue
sharing contract and announce that there only existed
an emission reduction rate within two kinds of
decision-making to maximize the supply chain
revenue.
Besides the decision making, revenue
distribution and are also analyzed. Yang et al. (2017)
analyzed the revenue sharing and first-run advantage
in manufacturers' carbon emission reduction efforts
and firms' profitability in linear demand settings of a
manufacture-driven supply chain.
These papers show that carbon tax play a
significant role in reducing emissions and useful
methods are also given, but joint influence of carbon
tax and other related items such as are not mentioned.
2) Consumer carbon sensitivity:
The impact of consumer carbon sensitivity
means that environmentally friendly consumers who
cares about carbon emission are willing to pay high
prices for low-carbon products (Chitra, 2007).
Considering of the impact of consumers’ preference to
1646
low carbon in an emission sensitive supply chain or
market, both the channel profit and the emission
reduction increase in the consumers’ preference to
low-carbon consumption simultaneously in particular
cases (Du et al., 2015). Liu et al. (2012) analyzed the
decision-making and profit-making of retailers and
manufacturers in the supply chain, putting the
emphasis on consumer environmental preferences. Ji
et al. (2017) compared two emission reduction
strategies, including single manufacturer's emission
reduction in production strategy and joint emission
reduction strategy. They indicated that when
considering consumers' low-carbon preference, the
retailer always has motivation to implement low
carbon promotion though without the manufacturer's
incentives. Yilei et al. (2017) focused their research on
three contract models under low carbon preferences of
consumers to analyze the cooperation and profit
reduction in the supply chain. They concluded here
that companies constraining the wholesale price can
make the supply chain reduce emissions and optimize
supply chain coordination. Wang et al. (2016) analyze
by contracting emission reduction for supply chains
considering market low-carbon preference. Therefore,
we can get that consumer carbon sensitivity affects the
revenue of the companies and the entire supply chain.
Under different strategy, the effects are not the same.
To sum up, the papers reviewed make
contribution to the carbon emission reduction to
protect the environment, but seldom paper takes both
of carbon tax and consumer carbon sensitivity. In this
paper, we present a game theory model to assist
companies facing these joint affect items, and analyze
the revenue distribution to help them get proper
decision to make the profits maximized for the entire
supply chain. Suggestions of the carbon tax for the
government are given as well. More precisely, we
contribute to the literature with a two-echelon supply
chain game model that simultaneously considers
carbon tax and consumer sensitivity of low-carbon
products. It also provide managerial implications for
the companies to maximize revenue and the
government to levy carbon at proper amount.
The structure of this paper is as follows. In
Section 2, we present the game modal in both
decentralized and centralized mode for the carbon
emission reduction and the balance analysis. In
Section 3, we present detailed numerical analyse to
illustrate the influence of the carbon tax and consumer
carbon sensitivity captured by the model and the type
of managerial insights that it allows to derive. In
Section 4, we conclude and discuss future research
opportunities.
2. Material and methods
2.1. Problem description and model assumptions
This article aims at a supplier-driven two-
echelon supply chain consisting of a supplier and a
manufacturer. Both the supplier and the manufacturer
are risk-neutral and the information between them is
 Supply chain emission reduction optimization under consumer carbon sensitivity and carbon tax policy

symmetrical. The manufacturer can only order generic
or low-carbon raw materials from upstream, so the
supplier is in the leading position in the supply chain.
Since the supply chain is driven by the supplier,
which means the supplier is in the dominant position,
the government charges carbon tax on the supplier is
more feasible to get the influence of the entire supply
chain. Consumers who are sensitive to low-carbon
environmental protection also grow in number.
Therefore, the supplier will implement carbon
reduction technologies to reduce the carbon emissions
per unit of product. Hence, several situations of the
consumer carbon sensitivity are considered under the
carbon tax: decentralized decision-making when
demand is unrelated to emission reduction; centralized
decision-making when the demand is related or
unrelated to emission reduction; and how the supplier
and the manufacturer will make the decision. The
following assumptions have hence been made for the
simplification and research convenience of the model:
Hypothesis 1: Assuming that upon the occasion
the supplier does not invest in carbon emission
reductions, the initial carbon emission reduction per
unit of production is e0. Then the government imposes
a carbon tax on carbon dioxide emissions from the
supplier at a rate of t yuan/ton CO2.
Hypothesis 2: The wholesale price and
production cost of the supplier's unit product are w and
c respectively, as the wholesale price is the supplier's
decision variable and the manufacturer sell their
products with the price p per unit.
Hypothesis 3: Carbon emission reductions for
the supplier after carbon reduction investment is e,
0≤ e ≤ e0 and e is the supplier's decision variable. The
emission cost of the supplier after carbon emission
reduction satisfies c(e ) = ke 2 / 2 , k > 0 (Ma et al., 2013;
Peng et al., 2013), where k is the carbon emission
reduction cost coefficient of the supplier; that is, c(e )
satisfies c' (e ) < 0 , c'' (e ) < 0 . In the actual production
process, the reduction of carbon emission e will lead
to increasing marginal costs, indicating that the early
stage emission reduction will be easy to carry on, but
become more and more difficult later.
Hypothesis 4: When there is some demand
for low-carbon products in the market, the supplier
adopts the order-oriented production method with the
same market demand without regard to stock-out and
stock-extrusion problems. This article uses the
symbols and their meanings as shown in Table 1.
2.2. Methods and models
2.2.1. Decentralized decision when demand unrelated
to carbon emission reduction
In this situation, the product demand is
expressed as Q = a - p. When the demand is unrelated
with the consumers’ sensitivities of carbon emission
reduction, the game is normally a two-echelon game.
In the game, the supplier first determines the
wholesale price w and the emission reduction rate e ,
while the manufacturer determines the final retail
price p according to the wholesale price and the
emission reduction rate provided by the supplier (w,e )
and the observed market information.
In the model, the government levies carbon tax
t of the CO2 , which is generated by the supplier in the
production process. At this point, we can model the
profit maximization decision function of the supplier
and the manufacturer as Eqs. (1-2):
π sN = (w − c − te0 )(a − p) − ke 2 / 2 (1)
π mN = ( p − w)(a − p) (2)

Table 1. Symbol meaning summary

Variable Meaning
w whole sale price of a unit product
p retail price of a unit of product
Q market demand for products
πs supplier's revenue
πm manufacturer's revenue
πc supply chain’s revenue
* optimal decision-making
t a unit of carbon tax levied by he government
c supplier production costs
N decentralized decision when demand isn’t related to carbon emission reduction
D decentralized decision when demand is related to carbon emission reduction
C centralized decision-making when demand is related to carbon emission reduction
e the amount of carbon emission reduction of the supplier
k cost coefficient for carbon emission reduction
e0 initial discharge of per unit product
E discharge of supply chain
η consumer sensitivity to low carbon products

1647
 Bai et al./Environmental Engineering and Management Journal 17 (2018), 7, 1645-1656
Calculating the first-order partial derivative of
π rN with respect to the retail price p in Eq.(1) and
make ∂π mN / ∂p = 0 , we can obtain the manufacturer's
retail price relative to the wholesale price of the
supplier as Eq. (1):
p = (a + w) / 2 (3)
Substituting Eq. (3) into Eq. (1) and calculating
the first-order partial derivative of the wholesale price
w and letting ∂π sN / ∂w = 0 , we get the optimal
wholesale price w of the supplier, as Eq. (4):
a + c + te0
wN =
*
2 (4)
Substituting Eq. (4) into Eq. (3), we could yield
*
the optimal retail price p N as Eq. (5):
3a + c + te0
pN =
*
(5)
4 the supplier and manufacturer are constituted by Eqs.
Substitute the optimal decision of the supplier
and the manufacturer into the demand function and
revenue Eqs. (1-2), and then the market demand, the
optimal revenue for the supplier and the manufacturer
will be as Eq. (6):
a − c − te0
QN =
*
4 Stackelberg game model between supplier and
π sN =
* (a − c − te0 )2 − 4ke 2
8 (6)
π mN =
* (a − c − t 0 )2
16
According to Eq. (1), we can get that when
demand is unrelated to the initial emissions of the
products, the optimal emission reduction of the
supplier is e = 0 . It means that the supplier will not be
willing to reduce the carbon emission because
emissions reduction has no effect on consumers’
purchase of the product. However, if consumers'
carbon sensitivity is related to the carbon emission of
the product, which means when the manufacturer
applies emission reductions, the demand of consumers
will increase, the manufacturer will actively increase
the carbon emission reduction investment.
The following study will thus focus on the
impact of carbon tax and consumer carbon sensitivity
on corporate emissions reduction, pricing and supply
chain performance under both decentralized and
centralized decision-making models when consumer
demand is related to carbon emission reduction.
1648
2.2.2. Decentralized decision when demand related to
carbon emission reduction
Since consumers' sensitivity for low-carbon
products will result in lower sales of products of
carbon-laden companies, if companies want to sell
more products, they will have to take active measures
to deal with carbon emission. Energy-saving and
emission reduction technology allows companies to
reduce energy consumption in the process of
production, and the manufacturer will invest in
applying certain technology to achieve the reduction
of carbon emissions rather than traditional production,
which means the carbon emission of per unit product
will decrease. The demand of consumers will thus
increase because of the carbon sensitivity.
As long as the carbon emission varies, the
carbon tax cost of companies, the increase of market
demand and the revenue of the supply chain
companies change. When companies reduce carbon
emissions, the demand will be expressed as
Q = a − p + ηe within the emission reduction amount
e as the decision variable. The revenue functions for
(7-8):
π sD = [w − c − t (e0 − e )](a − p + ηe ) − ke 2 / 2 (7)
π mD = ( p − w)(a − p + ηe ) (8)
We thus get the formation of the two-echelon
manufacturer. In the first stage, the supplier
determines its own wholesale price and emission
reduction amount according to the maximization of
revenue. In the second stage, the manufacturer
determines the retail price according to the emission
reduction amount of the supplier and the wholesale
price. Then we can use the inverse inductive method
to solve the problem, while the supplier would offer
the optimal wholesale price as Eq. (9):
a + c + te0 + (η − t )e
wD =
*
2 (9)
The optimal retail price is computed as Eq.
(10):
3a + c + te0 + (3η − t )e
pD =
*
4 (10)
Hence, we can obtain the demand of the
products, the benefit of supplier and manufacturer
respectively as given by Eq. (11):
a − c − te0 + (t + η )e
QD =
*
4
π sD
*
=
[a − c − te0 + (t + η )e] − 4ke 2
2
(11)
8
 Supply chain emission reduction optimization under consumer carbon sensitivity and carbon tax policy

π mD =
* [a − c − t 0 + (t + η )e]2  ∂ 2π C / ∂e 2 ∂ 2π C / ∂p∂e 2ηt − k t + η 
H = 2 C =
2 
16 ∂ π / ∂p∂e ∂ π / ∂p   t + η
2 C
− 2 

Considering that both the supplier and the According to ∂ 2π C / ∂p 2 = −2 and

manufacturer are rational, the emission reduction of
H = 2k − (t + η )2 > 0 , π
C
* * is a joint concave function
the supplier needs to meet π sD > π sN , which means
with respect to the retail price p and carbon emission
two possibilities should be considered as given by Eqs.
(12-13): reduction e , so there will be an optimal solution. By
calculating the first-order partial derivative of Eq. (1)
2(t + η )(a − c − te0 ) of the retail price p and carbon emission reduction
e>
4k − (t + η )
2 e respectively and making the results zero, we get
4k − (t + η ) > 0 ,
2
(12) Eqs. (17-18):

2(t + η )(a − c − te0 ) ∂π C

0<e< = a − 2 p + c + te0 − (t + η )e = 0
4k − (t + η ) < 0 , 4k − (t + η )
2 2
(13) ∂p (17)

Then we substitute the supplier's wholesale ∂π C

price (Eq. 9) and the retail price (Eq. 10) into Eq. (11), = ta + ηc − (t + η ) p + ηte0 − 2ηte − ke = 0
calculate the first-order partial derivative and let it be ∂e (18)
*
zero, as ∂π mN / ∂e = 0 , entailing that we can then obtain
* According to the calculating of Eqs. (17-18),
the optimal emission reduction eD expressed as Eq. *

(14): the retail price P C and carbon emission reduction

*
e C are computed as Eq. (19):
eD =
* (t + η )(a − c − te0 )
4k − (t + η )
2 k (a + c + te0 ) − (t + η )(at + ηte0 + ηc )
PC =
*
(14)
2k − (t + η )
2

*
and e N satisfies Eq. (12).
*
eC =
(t + η )(a − c − te0 )
(19)
Substitute the optimal emission reductions into 2k − (t + η )2
Eq. (11), then the demand, the revenue of supplier and
the manufacturer are computed as Eq. (15): * *
Taking P C and e C into demand, we get that
k (a − c − te0 ) the total demand and revenue of the whole supply
QD =
*

chain follow Eq. (20):

4k − (t + η )
2

k ( a − c − te 0 )
QC =
*

k (a − c − te0 )
2

π sD = 2 k − (t + η )
*
2

[
2 4k − (t + η )
2
] (15)
k ( a − c − te 0 )
2

πC =
*

k 2 (a − c − te0 )
[ ]
2

π mD = 2 2 k − (t + η )
*
2

[4k − (t + η ) ] 2 2
(20)

Since the supplier and the manufacturer in the

2.2.3. Centralized decision-making when demand is supply chain are rational, they would accept the profits
related to carbon emission reduction only if they are more than that under decentralized
In the centralized decision-making mode, the decision-making as Eq. (21):
supplier and the manufacturer are an integrated
system. The supplier and the manufacturer jointly π C * > π D*
determine the retail price p and the carbon emission s s
 C* (21)
reduction e in order to maximize the revenue of the
*
π m > π mD
entire supply chain system. We get the revenue
function of the entire supply chain as Eq. (16): Then, the optimal wholesale price under
*
centralized decision satisfies wmin < wC < wmax , and
π C = [ p − c − t (e0 − e )](a − p + ηe ) − ke 2 / 2
(16) the result of calculating is Eq. (22):
The Hessian matrix of π C on retail price p
and carbon emission reduction e is:

1649
 Bai et al./Environmental Engineering and Management Journal 17 (2018), 7, 1645-1656

wmin =
*
π sD + ke2 / 2 + cQC + t e0 − eC QC
*

*
( *
) *
reduction amount is more than decentralized. It shows
that under the precondition that the government levies
C
Q carbon tax, the emission reduction effect will be more
(22) effective within the cooperation between the supplier
p C Q C − π mD
* * *
and the manufacturer in the supply chain. Thus, when
wmax = the carbon tax policy is levied, the government should
QC
*

promote the benefits of the corporation of the entire

supply chain to reduce carbon emissions.
The retail price is obtained by bargaining Proposition 2: Under the decentralized
between the supplier and the manufacturer. If the decision-making mode and the centralized decision-
manufacturer's bargaining power is named as making mode, the optimal emission reduction will
α ,( 0 < α < 1 ), the supplier's bargaining power is 1-α decrease as the carbon emission reduction coefficient
correspondingly. According to the Nash bargaining *
of the supplier increases; that is ∂eC / ∂k < 0 ,
model, the problem of seeking the best w then boils *

down to N is Eq. (23): ∂e D / ∂k < 0 .

Therefore, no matter whether the supplier
N : Max
w
*
= π mC − π mD
*
( ) (π
α
C*
s > π sD
*
)(
1−α )
(23)
cooperates with the manufacturer to reduce emissions
or not, the optimal emission reductions will decrease
as the supplier's emission reduction coefficient
increases. This movement demonstrates that there is
which means (Eq. 24): scale diseconomy in the process of investment in
emission reduction.
N : Max  C
( )
α
= p − w Q C − π mD  , Proposition 3:
* * *

w   (24) When

{ (
 w − c − t e − eC  Q c − k eC
) ( ) (1−α )
2
/ 2 −π D*
}
* * *

 0  s 2
4e0  4k − ( t + η )  −
2

 
Then the best solution to the retail price follows  4ke − e ( t + η )2 − ( t + η ) ( a − c − te )  > 0
Eq. (25):  0 0 0


we can know that under decentralized decision-

p C − π mD − α∆π
* *

w= making mode, carbon emission of the whole supply

QC
*

(25) chain will decrease; and when:

According to Eq. (25), the wholesale price is

negatively related to the manufacturer's bargaining
[ ] [ ]
2e0 2k − (t + η )2 − 2ke0 − e0 (t + η )2 − (t + η )(a − c − te0 ) > 0
2

power, which means the stronger the bargaining

power of the manufacturer is, the lower the wholesale
price given by the supplier will be. On the contrary, it
will be higher. By using the Shapley method to
distribute the supply chain revenue under the
centralized decision, the profitability of the
manufacturer and the supplier should be expressed as
Eqs. (26-27):
π sD π C − π mD
* *
carbon emission of the whole supply chain will
decrease under the centralized decision-making mode.
When under the centralized and decentralized
decision, if the values of t and η are satisfied, then
the carbon emissions in the supply chain under the two
decision-making modes both decrease. Due to the
complexity of the inequality, the numerical analysis
will be verified in the next chapter.
*

π C*
s = + 3. Results and discussion
2 2 (26)
When the government imposes carbon tax on
π mD π C − π sD
* * *
the supplier, the supplier will consider the cost and
π mC = +
*

revenue of emission reductions to make a comparative

2 2 (27) decision. Because when the increasing consumer's
demand for low-carbon products brings more revenue
2.3. Balance results analysis than the cost, the supplier will actively invest in
emission reduction. Conversely, it will lack of power
Proposition 1: If 2k − (t + η )2 > 0 and to do this. Therefore, the carbon tax levied on the
a − c − te0 > 0 , when the consumer demand is related to supplier needs to be considered with consumer carbon
carbon emission reduction, then the emission sensitivity. Only if a proper carbon tax is levied, it will
reduction efficiency of the supplier is increased, and it play an incentive role in reducing emissions.
the carbon reduction in production is carried out. At the meantime, the companies have to face
Moreover, when the supplier and the manufacturer carbon tax and consumers are carbon-sensitive at the
take a centralized decision-making, the emission same time, so that they need make the carbon emission

1650
 Supply chain emission reduction optimization under consumer carbon sensitivity and carbon tax policy
decision to meet the low-carbon needs of both
consumers and government, as well as try to get
maximum revenue.
Owing to the complexity of solving the model,
to verify the model conclusion and simplify the model
calculation process, we use Matlab (R2016a) as a
calculation tool to solve each equation in the model.
To illustrate the effect of carbon tax and carbon-
sensitive coefficient on emissions reductions, it is
assumed that a = 100 , c = 60 , k = 2 , η = 0.4 ,
e0 = 40 , t = 0.1 , then compare companies revenue
and total revenue on the supply chain under demand
related/unrelated to carbon emission or not under
centralized decision-making and decentralized
decision-making, as shown in Table 2. This Table
shows that, when consumers' demand is related to
carbon emission reduction, companies have increased
the revenue of both the supplier and the manufacturer
by investment in emission reduction. At the same
time, Nash bargaining constrains the wholesale price
and the Shapley method to allocate companies revenue
under centralized decision-making, so that the profits
of all parties under decentralized decision-making
have been improved and the supply chain participates
are coordinated. We then analyze the impact of carbon
tax changing demand, supply chain revenue, emission
reductions and total emissions, as shown in Figs. (1-
4). We set the consumer preference as η = 0.4 ,
while the range of variation of t is 0~1. In Fig. 1, we
can get that no matter consumers are carbon-sensitive
Table 2. Revenues of companies and entire supply chain under different decision-making mode
Demand unrelated to carbon
Reve Demand related to carbon emission,
emission, decentralized
nue
decision-making
π *s 162
π *m 81
π *c 243
Fig. 1. Impact of carbon tax on demand
or not, the demand is negatively correlated with the
carbon tax. In addition, consumers have the maximum
demand when they are carbon-sensitive under
centralized decision-making. Meanwhile, demands
under decentralized decision-making only experience
minor differences in demand both related and
unrelated to carbon emission reduction. In Fig. 2, the
supply chain revenue are all negatively related to the
carbon tax in the three game models. The revenue
under centralized decision making within carbon-
intensive is the highest, and when the demand is
unrelated to the emission reduction, the supply chain
revenue is the lowest. Therefore, the emission
reduction effect within supply chain cooperation is
higher than that within individual emission reduction.
When the government has levied a carbon tax, we
should encourage upstream and downstream
companies in the supply chain to cooperate in
emission reduction.
According to Figs 1-2, when the government
wants to levy carbon tax to protect the environment, it
should carefully decide the tax amount. That’s
because if the tax is too heavy, the market will
decrease badly and even the supplier and manufacturer
may quit the market leading to industry decline in the
long term. It illustrates in Fig. 3 that when the tax rate
is low, while the level of emission reduction of supply
chain rises with the increase of carbon tax. When the
tax rate exceeds a certain threshold, the optimal
emission reduction will decrease with the increase in
carbon tax.
Demand related to carbon
emission, centralized
decentralized decision-making
decision-making
167.23 213.47
86.31 132.13
253.54 345.60
1651
 Bai et al./Environmental Engineering and Management Journal 17 (2018), 7, 1645-1656
450
400
350
300
250
200
150
100
50
0
0 0.2
Fig. 2. Impact of carbon tax on supply chain revenue
4 D*
e
3
e
0
0 0.2
Fig. 3. Impact of carbon tax on emission reductions
In Fig. 4, it shows that the total carbon
emissions under the three decision-making modes
have a negative correlation to the carbon tax. When
0 < t < 0.33 , carbon emissions under decentralized
decision-making within carbon-sensitive demand is
larger than that within non-carbon-sensitive demand,
which means that when 0 < t < 0.33 , the carbon tax
rate is low and so does the carbon emission. When
0.33 < t < 1 , since the carbon tax rate rise,
companies should sell more products to make up for
carbon emission reduction costs and the carbon tax,
while the total carbon emission is largest under
centralized decision making.
Therefore, the government should not make the
tax amount too low or too high when levying it
because a high tax rate will decrease the enthusiasm of
companies in emission reduction and a low tax rate
will not achieve the purpose of emission reduction.
The government needs to balance between industry
and environment protection. Then we analyze the
impact of the consumer changes on demand, supply
chain revenue, emission reduction and total emissions,
1652
C*
D*
N*
0.4 0.6 0.8 1
t
*
eC
*
eN
0.4 0.6 0.8 1
t
as shown in Fig. (5-8). We set carbon tax as t = 0.5
, η range of 0 ~ 1. In Fig. 5 it is shown that, when
consumers are carbon-sensitive, if carbon-sensitivity
increases, the demand increases too, and will reach the
top under centralized decision-making. Meanwhile,
there is only a little difference of the demand under
decentralized decision-making no matter the demand
is related or unrelated to emission reduction.
Fig. 6 presents that when the demand is related
to consumer carbon sensitivity, the supply chain
revenue has a positive correlation to consumers’
carbon sensitivity under both decentralized and
centralized decision-making, and the revenue of the
supply chain is higher under centralized rather than
decentralized decision-making.
Therefore, we use Nash bargaining to restrict
the wholesale price and then distribute the revenue
under the centralized decision-making within the
Shapley method. Hence, under decentralized decision-
making, both of the supplier and the manufacturer can
get more profit.
 Supply chain emission reduction optimization under consumer carbon sensitivity and carbon tax policy

800 24
N*
E
D*
E
22
700 C*
E

20
600

18

500
16

400
E

14

Q
300 12

10
200
C* D*
Q Q
8 N*
Q
100
6

0
0 0.2 0.4 0.6 0.8 1 4
0 0.2 0.4 0.6 0.8 1
t

Fig. 4. Impact of carbon tax on total emissions Fig. 5. Impact of consumers’ carbon sensitivity differences
on demand

240

220

200

180

160

140

D*
C*

120

N*

100

80

60
0 0.2 0.4 0.6 0.8 1

Fig. 6. Impact of consumers’ carbon sensitivity differences on supply chain revenue

According to Figs. 5-6, the companies in the
supply chain should improve the promotion of low-
carbon products to increase consumer carbon
sensitivity, then they can work together within
centralized decision-making in order to sell more
products and get more revenue.
In Fig. 7, we find out that with the increasing
of consumer carbon sensitivity, the carbon emission
reduction increases as well under both centralized
decision-making mode and decentralized decision-
making mode, which can be concluded that
consumers' carbon sensitivity can push companies
forward to reduce carbon emissions. In Fig. 8, the
total carbon emissions under the three decision-
making modes are positively correlated with the
consumers’ carbon sensitivities. Under decentralized
decision-making, when 0 < η < 1 , the total carbon
emissions amount within demand related to consumer
carbon sensitivity is lower than demand unrelated to
the sensitivity. When η = 1 , the level of emissions
almost achieves zero. Under centralized decision-
making, when 0 < η < 0.61 , the total carbon
emissions amount within demand related to consumer
carbon sensitivity is lower than demand unrelated to
the sensitivity. When 0.61 < η < 1 , the emission is
higher than the demand unrelated to carbon
sensitivity, and when η = 1 , the carbon emission
reaches the maximum value. It means that although
the emission reduction increases, the increase of the
consumers’ carbon preference leads to the increase in
sales and finally causes more emissions. Therefore,
the consumer carbon sensitivity plays a vital role in
carbon emission reduction of the supply chain.

1653
 Bai et al./Environmental Engineering and Management Journal 17 (2018), 7, 1645-1656
12
10
8
6
e
*
eN
2
* *
eC eD
0
0 0.2 0.4 0.6 0.8
Fig. 7. Impact of consumers’ carbon sensitivity differences on
carbon emission reduction
On the basis of encouraging the supplier and
the manufacturer to cooperate on emissions reduction,
the government should try to keep the consumer
carbon sensitivity at 0 < η < 0.61 , in order to get
better environment protection.
4. Conclusions
This paper s presents the game model between
the supplier and the manufacturer under the carbon tax
policy and consumers' carbon-sensitive differences.
By comparing the demand related or unrelated with
the emission reduction under both decentralized and
centralized modes of decision-making, we have
provided optimal emission reduction decisions, profit
decisions and optimal emissions in the supply chain
within different decision-making situations.
Comparing with reviewed paper’s conclusion
as manufacturer’s optimal carbon emissions per unit
product the centralized channel, no matter whether the
supply chain is coordinated or not, or when
considering consumers' low-carbon preference, the
retailer always has motivation to implement low-
carbon promotion though without the manufacturer's
incentives, we focus on more comprehensive effect by
considering both carbon tax and consumer carbon
sensitivity and finally, we get different conclusions by
calculating the equilibrium solution and numerical
analysis as below:
(1) Demand and supply chain revenue are
negatively related to carbon tax, but positively related
to consumers’ carbon-sensitivity preferences. Besides,
the demand and supply chain benefits are highest
under centralized decision-making under both demand
and carbon sensitivities;
1654
350
300
D*
C* E
E N*
250 E
200
E
150
100
50
0
1
0 0.2 0.4 0.6 0.8 1
Fig. 8. Impact of consumers’ carbon sensitivity differences on
carbon emissions
(2) When demand is unrelated to carbon
sensitivities, carbon tax and consumers’ carbon
sensitivities will not affect the emission reductions of
companies. When demand is related to carbon
sensitivities, emission reductions will first increase
and then decrease with the carbon tax increases,
meaning that emission reductions and consumer
carbon sensitivity are positively correlated;
(3) When the carbon tax is considered, the total
emission reductions in the supply chain decrease when
0 < t < 0.33 under decentralized decision-making.
Considering the carbon sensitivities of consumers,
when 0 < η < 0.61 , the total emission reductions of the
supply chain will increase under centralized decision-
making. Therefore, the comprehensive consideration
of the carbon tax and consumer carbon sensitivity is
particularly important for the formulation of emission
reduction strategies.
This paper is based on recognition of the reality
that both the supplier and the manufacturer are risk-
neutral and information symmetric. In fact, since the
cost of the supplier's emission reduction investment is
uncertain and affected by risk aversion, it is very
difficult to realize the complete sharing of information
on the supply chain, while uncertainty factors will also
affect the cost and profit distribution between the
supplier and the manufacturer. Therefore, considering
the information uncertainty of both parties is one of
the principal research directions to be taken in the
future.
Acknowledgements
We thank the support provided by the Project 691249 RUC-
APS (EU H2020-MSCA-RISE-2015), National Natural
Science Foundation of China: Urban Logistics Management
(71390334).
 Supply chain emission reduction optimization under consumer carbon sensitivity and carbon tax policy

Appendix we can conclude that the emission under the

centralized decision is reduced.
Certify of Proposition 1: If emission reduction
is greater than zero, then the following inequalities References
would be satisfied:
4k − (t + η )2 > 0 , a − c − te0 > 0 , Androniceanu A., Dragulanescu I.V., (2016), A survey on
the buyers' eco-responsibility and the urban white
which means, when pollution, Environmental Engineering and
2k − (t + η )2 > 0 , a − c − te0 > 0 Management Journal, 15, 481-487.
the emission reduction is positive. From Table (1) we Bjørner T.B., Jensen H.H., (2002), Energy taxes, voluntary
can get: agreements and investment subsidies-a micro-panel
analysis of the effect on Danish industrial companies’
energy demand, Resource & Energy Economics, 24,
e N = 0 , e D = (t + η )(a − c − te0 ) , eC = (t + η )(a − c − te0 )
* * *
229-249.
4k − (t + η )2 2k − (t + η )2 Bruvoll A., Larsen B.M., (2002), Greenhouse gas emissions
in Norway: do carbon taxes work?, Energy Policy, 32,
* * * * 493-505.
Then e C > e N , e D > e N . Chitra K., (2007), In search of the green consumer: a
perceptual study, Journal of Services Research, 7, 173-
Since * * 2k , we can get
eC − e D =
[4k − (t + η ) ][2k − (t + η ) ] > 0
191.
2 2
Comas Martí J.M., Tancrez J.-S., Seifert R.W., (2015),
* * Carbon footprint and responsiveness trade-offs in
eC > e D .
supply chain network design, International Journal of
Certify of Proposition 2: Under centralized Production Economics, 166, 129-142.
decision-making, the first-order partial derivative of Du S., Zhu L., Liang L., Ma F., (2013), Emission-dependent
*
optimal emission reduction e C with respect to supply chain and environment policy-making in the
‘cap-and-trade’ system, Energy Policy, 57, 61-67.
emission reduction coefficient k satisfies the Du S., Zhu J., Jiao H., Ye W., (2015), Game-theoretical
following: analysis for supply chain with consumer preference to
∂e C / ∂k = −2(t + η )(a − c − te0 ) / 2k − (t + η )
* 2
[ ] low carbon, International Journal of Production
Research, 3, 3753–3768.
and we can conclude from Proposition 1 that Ji J., Zhang Z., Yang L., (2017), Carbon emission reduction
2k − (t + η ) > 0 , then obtain ∂e D / ∂k < 0 ,
2 *
decisions in the retail-/dual-channel supply chain with

∂e C / ∂k = −2(t + η )(a − c − te0 ) / 2k − (t + η ) < 0 .

*
[ 2
] consumers' preference, Journal of Cleaner Production,
141, 852-867.
Certify of Proposition 3: When the demand is Lin B., Li X., (2011), The effect of carbon tax on per capital
independent of carbon emissions, the carbon CO2 emissions, Energy Policy, 39, 5137-5146.
Liu Z., Anderson T.D. Cruz J.M., (2012), Consumer
emissions per unit product is e0 , then the carbon environmental awareness and competition in two-stage
emission of the whole supply of the supply chain is supply chains, European Journal of Operational
E N = e0Q N ; when the demand is related to carbon Research, 218, 602-613.
Lu Y.Y.L., Wu C.H., Kuo T.-C., (2007), Environmental
emissions in the decentralized decision-making mode, principles applicable to green supplier evaluation by
the total carbon emissions is E D = (e0 − e )Q D , and the
*
using multi-objective decision analysis, International
emission reduction is: Journal of Production Research, 45, 4317-4331.
Ma P., Wang H., Shang J., (2013), Contract design for stage
supply chain coordination: Integrating manufacturer-
∆E D = E N − E D = e0 Q N − ( e0 − e ) Q D =
*

quality and retailer-marketing efforts, International

Journal of Production Economics, 146, 745-755.
e0 ( a − c − te0 ) / 4 − ( e0 − e ) k ( a − c − te0 )  /  4k − ( t + η ) 
2

  Meng S., Siriwardana M., McNeill J., (2013), The

Then, if environmental and economic impact of the carbon tax
2 in Australia, Environmental & Resource Economics,
4e0  4k − ( t + η )  −
2

  54, 313-332.
Michaud C., Llerena D., Joly I., (2012), Willingness to pay
 4ke − e ( t + η )2 − ( t + η ) ( a − c − te )  > 0 for environmental attributes of non-food agricultural
 0 0 0

products: a real choice experiment, European Review of
we can get ∆E D = E N − E D > 0 . Agricultural Economics, 40, 313-329.
Also, in centralized decision-making mode, we can Tseng S.-C., Hung S.-W., (2014), A strategic decision-
obtain: making model considering the social costs of carbon
∆E C = E N − E C = e0 Q N − ( e0 − e ) Q C =
*
dioxide emissions for sustainable supply chain
management, Journal of Environmental Management,
e0 ( a − c − te0 ) / 4 − ( e0 − e ) k ( a − c − te0 )  /  2k − ( t + η ) 
2

  133, 315-322.
Vanclay J.K., Shortiss J., Aulsebrook S., Gillespie A.M.,
If
2
Howell B.C., (2011), Customer response to carbon
2e0  2k − ( t + η )  − labelling of groceries, Journal of Consumer Policy, 34,
2

 
153-160.
 2ke − e ( t + η )2 − ( t + η ) ( a − c − te )  > 0 Wang Q., Zhao D., He L., (2016), Contracting emission
 0 0 0

reduction for supply chains considering market low-

1655
 Bai et al./Environmental Engineering and Management Journal 17 (2018), 7, 1645-1656

carbon preference, Journal of Cleaner Production, 120, Yang H., Luo J., Wang H., (2017), The role of revenue
72-84. sharing and first-mover advantage in emission
Yalabik B., Fairchild R.J., (2011), Customer, regulatory, and abatement with carbon tax and consumer
competitive pressure as drivers of environmental environmental awareness, International Journal of
innovation, International Journal of Production Production Economics, 193, 691-702.
Economics, 131, 519-527. Yilei W., Qinghua Z., Xiqiang X., (2017), Coordinated
Zhou Z.F., Xiao Y.X., Ziqi D., Liu L.M., Chen X.H., Zeng contract game model for joint emission reduction at
H.X., (2017), Determinants of corporate carbon upstream and downstream of supply chain based on
efficiency: evidence from CDP 2011-2014 consumption preference (in Chinese), Journal of
questionnaire for Standard & Poor's 500 index Systems Engineering, 32, 188-198.
companies, Environmental Engineering and Yue J., Shuaihua H., (2016), Effect of supplier participation
Management Journal, 16, 1595-1608. on enterprise emission reduction decision under the
Yang H., Luo J., (2016), Study on decision of supply chain carbon tax regulation, (in Chinese), Soft Science, 30,
emission reduction under carbon tax policy (in 43-48.
Chinese), Systems Engineering -Theory & Practice, 36,
3092-3102.

1656
Environmental Engineering and Management Journal July 2018, Vol.17, No. 7, 1657-1666
http://www.eemj.icpm.tuiasi.ro/; http://www.eemj.eu

“Gheorghe Asachi” Technical University of Iasi, Romania

FACTORS AFFECTING THE PATTERN OF VEGETATION CARBON

DENSITY IN A KARST REGION IN NORTHWEST GUANGXI, CHINA

Mingyang Zhang1,4, Kelin Wang1,4∗, Huiyu Liu2, Weijian Luo3, Jing Wang1,4, Yuemin Yue1,4
1KeyLaboratory of Agro-ecological Processes in Subtropical Region, Institute of Subtropical Agriculture,
Chinese Academy of Sciences, Hunan 410125, China
2College of Geography Science, Nanjing Normal University, Nanjing 210046, China
3Central South Forest Inventory and Planning Institute of State Forestry Administration, Changsha 410014, China
4Huanjiang Observation and Research Station for Karst Ecosystems, Chinese Academy of Sciences, Huanjiang, 547100, China

Abstract

The changing characteristics of vegetation carbon patterns in the karst region of northwest Guangxi, China and their impact factors
were analyzed on the basis of vegetation inventory data from 2005 to 2010. A radial basis function network model (RBFN) was
constructed using data from 1377 samples and 13 environmental factors. The results for the 5-year study period were as follows:
(1) The total carbon storage of vegetation had increased with an annual growth rate of 1.84% and the carbon density of vegetation
increased from 29.04 t hm-2 to 29.57 t hm-2. The carbon density in the west (>40 t hm-2) was greater than that in the Middle East
(<25 t hm-2). Hot spot analysis revealed a random distribution of vegetation carbon density in 2005, but a highly aggregated
distribution in 2010. (2) The four most important impact factors on spatial distribution of vegetation carbon density in this area
were land type, forest type, forest category, and vegetation type (significance <50%). The least important factors were location,
slope, aspect, and elevation (significance of 2–11%). Vegetation carbon density increased significantly with the implementation of
rocky desertification control measures. Factors changed by human activities had much greater impacts than topographic factors on
the spatial distribution of vegetation carbon density. Therefore, the Ecological Immigration Program, returning farmland to forests,
and rocky desertification control measures had an important effect on the pattern of vegetation carbon density.

Key words: Guangxi China, impact factors, karst, northwest, pattern characteristic, radial basis function network (RBFN),
vegetation carbon

Received: February, 2014; Revised final: August, 2014; Accepted: September, 2014; Published in final edited form: July 2018

1. Introduction carbon to soil carbon is much higher in karst

Estimation of vegetation carbon is one of the
key steps in examining greenhouse gas CO2 exchange
between terrestrial ecosystems and the atmosphere
(Dixon et al., 1994; Piao et al., 2009). Karst ecological
systems are an important component of the terrestrial
ecosystem (Cao et al., 2005) and the CO2-H2O-CaCO3
system with biological and chemical active processes.
Thus, karst systems play an important role in the
global carbon cycle. There is relatively little carbon
stored in karst soil because it is shallow and
discontinuous. Consequently, the ratio of vegetation
ecosystems than in other types of ecosystems (Cai et
al., 2012). China has invested more than CNY 5
billion in rocky desertification control, and had
implemented the following programs, beginning in the
1990's: the National Poverty Alleviation Plan, the
Western Development Strategy, the Green for Grain
Program, the Mountain Closure Program (to facilitate
afforestation), and the Ecological Immigration
Program. The focus of the Ecological Immigration
Program was to help farmers in karst areas to move to
non-karst areas where ecosystem conditions were
deemed to be considerably better. During this

∗
Author to whom all correspondence should be addressed: e-mail: my_1223@163.com; Phone: +86 731 84615208; Fax: +86 731 84612685
 Zhang et al./Environmental Engineering and Management Journal 17 (2018), 7, 1657-1666
campaign, 49,133 families (232,705 persons) were
relocated in northwest Guangxi, China. The total area
of the Green for Grain Program was up to 1,278.67
km2 in Baise during 2001–2004 (Zhang et al., 2011).
This included 605.33 km2 of farmland that was
restored to woodland and 673.33 km2 of reforestation
activities of barren hills. For example, in Duan
County, the land area of rocky desertification had been
reduced to 333,356.9 hm2, and 45,333.33 hm2 of land
were converted from farmland to forest as part of the
Green for Grain program by 2009. The comprehensive
treatment of rocky desertification was expected to
have caused changes in the vegetation carbon pattern.
Impact factor analysis could be used to evaluate the
treatment of rocky desertification.
Vegetation carbon in the karst area has been
studied by many researchers (Liu et al., 2008; Wang
et al., 2012; Ye et al., 2010; Luo et al., 2010; Tian, et
al., 2011). However, most of the studies focused on
the micro or local ecological system scale.
Investigation of the spatial pattern of vegetation
carbon is needed at regional scale. Furthermore,
factors that affect the formation of vegetation carbon
spatial patterns have been scarcely studied.
Investigation of the changes in vegetation carbon
patterns in a typical karst area, and further analysis of
the main impact factors can provide a scientific
reference for the evaluation of karst rocky
desertification control and basic data for
understanding the driving mechanisms of carbon
Fig. 1. Location and land cover of the study area
1658
characteristics in karst regions. This will permit an
accurate estimation of the carbon storage in karst
regions. We used forest resources inventory data
collected in 2005 and 2010 to study vegetation carbon
density in a typical karst area in northwestern
Guangxi, China. In addition, a radial basis function
network (RBFN) model was applied to test the
influence of environmental factors on carbon spatial
distribution.
2. Study area and methods
2.1. Study area
The study area was located in northwest
Guangxi, China (104°29´–109°09´E, 23°41´–
25°37´N) (Fig. 1), which includes 23 counties
covering an area of 69,400 km2 and sustaining a
population of 7.97 million people. The region has a
subtropical wet monsoon climate with an annual
temperature 19.5°C and an annual precipitation of
1,000–1,600 mm. The elevation in this hilly region
ranges from 100 m to 2,000 m above sea level and the
dominant vegetation communities are mixed
subtropical evergreen and deciduous forests. The
landforms in this region are typical karst landforms,
including poljes, cockpits, towers, and dolines. The
region supports a mountainous agricultural region in
which the cropland areas are generally not very fertile.
 Factors affecting the pattern of vegetation carbon density in a karst region in northwest Guangxi, China
2.2. Data acquisition and preprocessing
2.2.1. Data sources and software
The main data used consisted of 2005 and 2010
forest inventory data from the Central South China
Forestry Survey and Design Institute (a design
qualification survey, planning, and design institute
that is part of the State Forestry Administration,
China). The data was obtained using systematic, fixed
sampling of 25.82 × 25.82 m (0.0667 ha) sampling
plots. More than 60 variables were measured at each
sampling location, including altitude, slope position,
slope, vegetation type, forest type, land type,
vegetation coverage, soil thickness, humus, rocky
desertification. ArcGIS 9.3 (Environment Systems
Research Institute) was used to prepare data and for
spatial analysis. Weather and radiation datasets were
interpolated (i.e., Kriging) in ARCGIS 9.3. All data
were projected or re-projected to an Albers Conical
Equal Area, Krasovsky Spheroid projection and then
re-sampled to a 100 × 100-m pixel spacing.
2.2.2. Biomass and vegetation carbon
We used biomass expansion factors (BEFs) to
estimate the total biomass in the study area from the
forest inventory data. BEFs are comprehensive
reflections of habitat, climate, age, and other factors
that influence vegetation (Jiao et al., 2005). BEFs are
generally based on regression equations. We used the
following equation to calculate BEF (fBEF) (Jiao et al.,
2005) (Eq. 1):
fBEF= a + b / V (1)
where a and b are conversion factors (constants) and
V is forest volume.
The density of a biological stand (B, t hm-2) can
be estimated by multiplying V (m3 hm-2), i.e.(Jiao et
al., 2005) (Eq. 2):
B aV + b
= (2)
Consequently, stand biomass (w) can be
expressed by the following equation (Jiao et al., 2005)
(Eq. 3):
Fig. 2. The structure of a Radial Basis Function Network (RBFN)
k
w= ∑A ×F
i =1
i BEFi × Vi
(3)
where i is the dominant species, Ai is the tree stand
area, Vi is the mean volume of the forest, and FBEFi is
the corresponding conversion factor.
Carbon storage can be estimated using a
constant (Fang et al., 2002), derived from molecular
formulas (Li et al., 1996), or measured directly (Mo et
al., 2003). We used 0.05 as the constant for estimating
carbon storage (Fang et al., 2002).
Vegetation carbon storage (C) was calculated
by multiplying the carbon in vegetation biomass (W)
by the carbon content (Cc), as follows (Fang et al.,
2002) (Eq. 4):
C= W × Cc (4)
Vegetation carbon density was defined as the
vegetation carbon of per unit area. The carbon in this
paper only refers to the vegetation carbon and does not
include carbon in the litter layer.
2.2.3. RBFN principle and training algorithm
RBFNs are artificial neural network models
that are widely used in the fields of taxonomy and
biodiversity (Schwenker et al., 2001; Huang et al.,
2005). These models are composed of an input layer,
a hidden layer, and an output layer that consists of
three-layer feed-forward networks. Each layer has a
plurality of neurons and a one-way connection
between layers (Fig. 2), where Xk is the input layer (k
= 1, 2, ..., N), Vj is the hidden layer input (j = 1, 2, …
L), and Oi is the output layer (i = 1, 2, ..., M). Cjk =
(Cj1, Cj2, ..., Cjn) is center and width of the hidden
layer unit function. N, M, and K are the number of
input, hidden, and output units, respectively. We used
RBFN to analyze factors that affect the pattern of
vegetation carbon in the karst area of northwest
Guangxi, China by approximating the dataset
distribution using a linear combination of Green’s
functions (Haykin, 1994; Moody and Darken, 1989;
Park and Sandberg, 1991) (Eq. 5).
1659
 Zhang et al./Environmental Engineering and Management Journal 17 (2018), 7, 1657-1666

N
(2) Identify the winner yw = max (yi), i < n.
F ( x)
= ∑ w G( x − x )
i =1
i i (3) If yw > αT, then change the position of the center
(5)
of the winner using (Eq. 13):
where G (x-xi ) is a Green’s function and x is its center.
The Green’s functions were characterized by a cw( k +1) =cx( k ) + r ( x j − cw( k ) )
(13)
mean vector xj and common variance σ2, as (Eqs. 6-7):
x − xi where α is an arbitrary ratio (α = 2.0 in this case) and
−
G ( x − xi ) =
e
2δ 2j T is the threshold value (T = 10-4).
(6) (4) If yw < T/α, then construct a new hidden node with
center at the point xj and variance equal to δ and
and
increase the number of total hidden nodes n by 1.
−
x − xi The algorithm stops when there are no new
N
F ( x) = ∑ wi e
2δ 2j hidden nodes after testing all points of data set D.
i =1 (7)
2.2.4. Impact factors of vegetation carbon density
where: each equation consists of a linear superposition The following 13 factors were selected for
of multivariate Gaussian basis functions (probability analysis:
bells) with centers xj, located at the data points and (1) Altitude: the range of altitude in the study area
δ j2 was 92–1876 m;
widths . (2) Aspect: divided into 9 aspects; North (338–22°),
The initial number of nodes was set at two. The the Northeast (23–67°), East (68–112°), Southeast
first node was assigned a Gaussian function centered (113–157°), South (158–202°), Southwest (203–
on the center of data set, determined by Eq. (8): 247°), West (248–292°), Northwest (293–337°), and
no slope (<5° slope). Aspects were coded 1, 2, 3, 4, 5,
N
xi
xav = ∑
6, 7, 8 and 9, respectively
i =1 N (8)
(3) Slope position: 6 positions, ridge (mountain
watersheds), above, middle, lower (the downhill), vale
(collection waterline on both sides of the valley), flat
and the second node was assigned a Gaussian function (plain and tableland). Slope positions were coded 1, 2,
centered on the point determined by the vector xp that 3, 4, 5, and 6, respectively;
maximized the function (Eqs. 9-10) (4) Slope angle: the slope of samples in the study
area were between 0 and 80°;
P( x) = ∫ S Q( x)dS (9) (5) Soil thickness: sample thickness was 1–180 cm;
(6) Humic layer: the thickness of humus layer was
where: 0–35 cm;

Q( x) = {1,0, if x∈D
(7) Vegetation coverage: total vegetation coverage
if x∉D in the study area was 10–100%;
(10)
(8) Land type: according to the characteristics of
The variance of Gaussian functions is assumed land cover and utilization, including arbor, shrub,
to be constant and equal to (Eq. 11): farmland, grass, and construction land;
(9) Vegetation type: including coniferous forest,
N m
( j) evergreen broad-leaved forest, deciduous broadleaf
∑∏ ( x
i =1 j =1
( j)
i −x ) forest, bamboo forest, deciduous shrubs, and
δ =r m
evergreen shrub;
N
( j)
∏ ∑ (x
j =1 i =1
( j)
i − x )2 (10) Grade of rocky desertification: the degree of
rocky desertification was defined as in the Forest
(11)
Inventory Check Operating Rules of Guangxi. The
where m is the number of dimensions (number of study area included the following grades: no rocky
nodes in the hidden layer), x-(j) is the mean value for desertification, potential, mild, moderate, and severe
dimension j and r is an arbitrary ratio defining the total rocky desertification;
number of Gaussian functions. (11) Forest species: 23 sub forest species were
The second layer was carried out using a self- categorized as in the forest inventory check operating
organizing algorithm with the following steps: rules of Guangxi. Species were categorized according
(1) Calculate the output of the Gaussian functions for to their potential for such uses as soil and water
all hidden nodes (Eq. 12) conservation, national protection, scenic value,
timber, et al.;
x j − ci
− (12) Forest type: the main two forest types were
yi = e δ2
(12) ecological forest and commercial forest. The
ecological forest was subdivided into national
where i < n (number of nodes). ecological forest and local ecological forest.

1660
 Factors affecting the pattern of vegetation carbon density in a karst region in northwest Guangxi, China

Commercial forest was divided into timber forest,
firewood forest, and economic forest;
(13) Area level: there were 7 categories, according
to the contiguous area of each vegetation type.
Categories were <1 hm-2, 1–4.9 hm-2, 5–9.9 hm-2, 10–
19 hm-2, 20–49 hm-2, 50–99 hm-2, and >100 hm-2 and
were coded 1, 2, 3, 4, 5, 6, and 7, respectively.
3. Results and discussion
3.1. Pattern characteristics of vegetation carbon
3.1.1. Temporal change in vegetation carbon
Vegetation carbon density in the study area
increased from 29.04 t hm-2 in 2005 to 29.57 t hm-2 in
2010 (Table 1). Accordingly, the storage of vegetation
carbon had increased from 4.19 × 104 t to 4.27 × 104 t
with an increase of 1.84%. The vegetation carbon
density in the study area was lower than the overall
vegetation carbon density of China (38.05 t hm-2),
Fujian Province (32.85 t hm-2) and Hainan Province
(32.59 t hm-2).
However, the vegetation carbon density in the
study area was higher than in northern Hunan
Province (18.53 t hm-2) and Sichuan Province (18.47 t
hm-2) (Fang et al., 1996; Wang, 2004; Cao et al., 2002;
Jiao et al., 2005; Huang et al., 2008). The vegetation
carbon density at the study site was close to that of
Jiangxi (25.38 t hm-2) (Wang and Wei, 2007). The
cover type with the highest vegetation carbon density
was arbor (34.82 t hm-2 and 35.12 t hm-2 in 2005 and
2010, respectively), followed by bamboo, which
accounted for about 50.82% and 54.62% of total
vegetation carbon storage in 2005 and 2010,
respectively.
Because the samples were fixed, changes in the
number of samples in each land type classification
may indicate regional vegetation changes to some
degree. In the period from 2005 to 2010, the number
of samples classified as arbor and shrub increased,
whereas samples classified as no stumpage forest,
unused land, and farmland decreased dramatically
(Table 1). These changes indicate that the vegetation
coverage in the study area has been improved by rocky
desertification control policies, such as ecological
migration program and the Green for Grain program.
Although the greatest vegetation carbon
density was in provincial and national nature
protection areas, the other projects also resulted in
dramatic increases in vegetation carbon (Fig. 3).

Table 1. Vegetation carbon density and their changes in Guangxi, China between 2005 and 2010

Vegetation Vegetation
Sample number
Land type carbon density (t.hm-2) carbon storage (* 104 t)
2005 2010 05-10 2005 2010 05-10 2005 2010 05-10
Arbor 612 664 52 34.82 35.12 0.30 2.13 2.33 0.20
Bamboo 10 11 1 29.35 28.52 -0.83 0.03 0.03 0.00
Shrub 231 244 13 24.63 24.53 -0.11 0.57 0.60 0.03
No stumpage forest 134 108 -26 24.66 25.20 0.53 0.33 0.27 -0.06
Farmland 324 308 -16 24.86 24.80 -0.06 0.81 0.76 -0.04
Grassland 18 10 -8 24.90 25.31 0.40 0.04 0.03 -0.02
Water area 14 19 5 24.55 24.41 -0.15 0.03 0.05 0.01
Unused land 70 45 -25 24.41 24.55 0.14 0.17 0.11 -0.06
Construction land 31 35 4 25.11 25.61 0.50 0.08 0.09 0.01
Total 1444 1444 - 29.04 29.57 0.53 4.19 4.27 0.08

Pearl River
Water
Protection Forest conservation
50 forest
40
40 35
30
No forestry 30 Land conversion 25
engineering projects 20
20 15 Protected
Fruit trees 10
10 national forests
5
0 0

Other forestry National Nature

engineering Reserve
Firewood forest Timber forest

2005 2005
Provincial Nature
Reserve 2010
2010

(a) (b)

Fig. 3. Vegetation carbon density of areas influenced by (a) different conservation projects
and (b) different vegetation species (t.hm-2)

1661
 Zhang et al./Environmental Engineering and Management Journal 17 (2018), 7, 1657-1666
The vegetation carbon density of water
conservation forest and special use forest was
relatively high (more than 30 t hm-2) (Fig. 3). This
could be attributed to the implementation of rocky
desertification control measures.
3.1.2. Spatial distribution of vegetation carbon
Vegetation carbon density tended to be greater
in the western part of the study area than in the eastern
part of the study area (Fig. 4). Whereas vegetation
carbon density was 28–35 t hm-2 in the western part, it
was 25–30 t hm-2 in the eastern part of the study area.
The distribution of vegetation carbon density was
similar to that of vegetation net primary productivity
(NPP) and vegetation coverage. The Previous research
showed that mean vegetation coverage was more than
60% and mean NPP was more than 1000 g m-2 in the
west, and vegetation coverage was less than 30% and
NPP was 100 g m-2 in the east in 2005 (Zhang et al.,
2011). The rocky desertification land in Guangxi
China was the third largest region of rocky
desertification in China and was mainly distributed in
the counties of Pingguo, Dahua, Duan, Masan,
Donglan, Bama, and Fengshan (Yang, 2003). These
counties were mainly located in the eastern or middle
part of the study area.
The spatial variation of vegetation carbon
density differed from the spatial distribution of
vegetation carbon density. The vegetation carbon
density generally increased in the study area (Fig. 5).
Between 2005 and 2010, the vegetation carbon density
of 16 counties remained at the same level and that of
7 counties increased, including the counties of Napo,
Pingguo, Longlin, Donglan, Leye, Huanjiang, and
Tiane. According to results of rocky desertification
monitoring in 2005, there was 83,366.67 hm2 of rocky
desertification in Pingguo county. In this county, the
Green for Grain program was carried out in areas
totaling 45,333.33 hm2.
Fig. 4. Distribution of vegetation carbon density of samples in Guangxi, China in 2005 and 2010
1662
Vegetation carbon density had improved in this
study region as the result of recent implementations of
control measures for dealing with rocky
desertification. Results of previous researches showed
that the ecosystem service has improved since 2005 in
typical karst regions (Zhang et al., 2011).
3.1.3. Distribution pattern of vegetation carbon
In order to evaluate the pattern of vegetation
carbon distribution in the study area, we used hot spot
analysis (spatial statistics, Getis-Ord General G). The
resultant Z scores indicated whether features were
tightly clustered or widely scattered. This method
works by examining each feature within the context of
neighboring features. A feature with high vegetation
carbon density was interesting but may not be a
statistically significant hot spot.
To be a statistically significant hot spot, a
feature must have a high value and be surrounded by
other features with high values as well. The local sum
for vegetation carbon density and its neighbors is
compared proportionally to the sum of all features;
when the local sum is much different from the
expected local sum, and that difference is too large to
be the result of random chance, a statistically
significant Z score was generated. The G statistic
returned in the dataset is a Z score.
Hot spot analysis of vegetation carbon density
indicated a random pattern of distribution in 2005, but
a highly clustered pattern in 2010 (Fig. 6). Large,
positive Z scores indicate intense clustering of high
values (hot spot). Small, negative Z scores indicate
intense clustering of low values (cold spot). The Z
score was -0.273287 in 2005, but was 2.780397 in
2010. This indicates that the distribution pattern of
vegetation carbon density may have been impacted by
rocky desertification control measures.
.
 Factors affecting the pattern of vegetation carbon density in a karst region in northwest Guangxi, China

Fig. 5. Vegetation carbon density in each county in Guangxi, China

(a) (b)

Fig. 6. The general G for distribution of vegetation carbon in the study area in 2005 (a)and in 2010 (b)

3.2. Impact factor analysis of vegetation carbon
pattern
3.2.1. The model
The total number of samples was 1377,
excluding non-vegetated samples. The effective
number of samples was 1297 (80 test samples were
excluded from analysis automatically). We used
74.3% of the effective sample (964) for training and
25.7% of the effective sample (333) to test the model.
There were 13 impact factors in the input layer of the
FRBN model in this study. The output layers were
vegetation carbon density, significance of impact
factors, and an error function (Table 2). There were
516 hidden layer units in the trained model. The
activation function was softmax and the error function
was the sum of squares.
3.2.2. Results of impact factors on vegetation carbon
The rankings of the significance of the 13
factors that affect the spatial distribution of vegetation
carbon density were (Fig. 7): land type > forest type >
vegetation species > vegetation type > degree of rocky
desertification > humus layer thickness > area level >
vegetation coverage > soil thickness > slope position
> slope > aspect > altitude. The top 4 impact factors
were land type, forest type, vegetation species, and
vegetation type, and their significances were all above
50%. The next 5 factors, rocky desertification degree,
humus layer thickness, area level, vegetation
coverage, and soil thickness, had standardization
significances of 15–30%. The least significant impact
factors for vegetation carbon density were the 4
topography factors, and their standardization
significances were only 2–11%. This indicated that the
impacts of topography factors on the distribution of
vegetation carbon density were limited, and factors
associated with human activities, such as land type,
forest type, vegetation species, and vegetation type
had relatively larger impacts on vegetation carbon
density. Therefore, the ecological immigration
program and the Green for Grain program had
important impacts on the spatial distribution of
vegetation carbon density.

1663
 Zhang et al./Environmental Engineering and Management Journal 17 (2018), 7, 1657-1666

Table 2. The network information of model processing

Levels Information
Input layer Factors 1 Altitude
2 Aspect
3 Slope position
4 Slope
5 Soil thickness
6 Humus layer
7 Vegetation coverage
8 Land type
9 Vegetation type
10 Degree of rocky desertification
11 Vegetation species
12 Forest type
13 Area grade
Units 516
Hidden layer Units number 5a
Activation function Softmax
Output layer Dependent variable 1 Total_c (Vegetation carbon density)
Units number 1
Scale dependent variable rescaling method Standardization
Activation function Identical
The error function Quadratic sum

Fig. 7. The significance of impact factors for the distribution of vegetation carbon density

Table 3. The correlations between environmental factors and vegetation carbon density

AL AS SP SL ST HL VC LT VT DR VS FT AG
VC 202** -072** -160** 065* 323** 348** 246** .374** -.089** -205** 305** 236** 155**
VC: Vegetation carbon density, AL: Altitude, AS: Aspect, SP: Slope position, SL: Slope, ST: Soil thickness, HL: Humus layer, VC: Vegetation
coverage, LT: Land type, VT: Vegetation type, DR: Degree of rocky desertification, VS: Vegetation species, FT: Forest type, AG: Area grade. **
Significant at 0.01 level * Significant at 0.05 level

1664
 Factors affecting the pattern of vegetation carbon density in a karst region in northwest Guangxi, China
Of the 13 impact factors investigated, 12 were
significantly correlated with vegetation carbon density
at the 0.01 level (Table 3). The remaining impact
factor (slope) was significant at the 0.05 level. Soil
thickness, humus layer, vegetation coverage, forest
type, vegetation type, altitude, and area level were
positively related to vegetation carbon density. The
pattern of vegetation carbon density was positively
affected by these environmental factors. Slope aspect,
slope position, and rocky desertification were
negatively correlated to vegetation carbon density. As
the degree of rocky desertification increased, the
vegetation carbon density decreased (Hu et al., 2008).
4. Conclusions
In this study, we used forest inventory data in
2005 and 2010 and geospatial technology to analyze
the spatial pattern of vegetation carbon and factors that
impact it in a typical karst area of northwest Guangxi,
China. The results showed that vegetation carbon
density increased from 29.04 t hm2 to 29.57 t hm2
between 2005 and 2010. The storage of vegetation
carbon increased 1.84% during the same period.
Spatial variation in vegetation carbon density
generally increased in the study area between 2005
and 2010. Hot spot analysis indicated a random
distribution in 2005, but a highly clustered distribution
in 2010. The impacts of factors associated with human
activities had relatively high impacts on the
distribution of vegetation carbon density. The results
indicated that the ecological immigration program and
Green for Grain program had important impacts on the
spatial distribution of vegetation carbon density.
The distribution trend of vegetation carbon
density was similar to the distribution of vegetation.
The spatial pattern of vegetation changes in the study
region indicated that vegetation carbon density had
increased as a result of recent rocky desertification
control measures. For example, planting and
protection of forest areas was the main measures for
desertification control in Hechi City. Beginning in
2008, the counties of Duan, Dahua, Fengshan, and
Huanjiang started to implement rocky desertification
control projects, in which 9,000 hm2 per county were
comprehensively managed as part of a pilot program.
In summary, the vegetation carbon density has
increased in this study region due to rocky
desertification control. The results of the present study
showed that factors affected by humans have
important effects on the pattern of vegetation carbon
density in a typical karst region. Our study also
showed that rocky desertification control measures
had an important impact on the pattern of vegetation
carbon density.
The conflict between economic development
and environmental protection was a common issue
worldwide and was also noted in rocky desertification
in karst areas (Brown et al., 2011). Poor environmental
quality was known to be an important restricted
condition in karst areas. Therefore, conservation of
karst areas should take priority over restricting the
uncontrolled reclamation of these areas for economic
purposes in future land use practices. Controlling karst
rocky desertification requires the optimization of land
use structure. Although discontinuing all reclamation
activities in karst areas might not be possible, future
land reclamation projects need to be controlled and
should be implemented after rigorous environmental
impact assessment. More detailed studies of the
impacts of karst reclamation projects on ecosystem
services provided in the southwest China were
necessary.
The present study has some limitations that
should be addressed in future studies. First, we did not
consider the soil carbon pool, so the carbon capacity
of the whole forest system was not estimated. Second,
the degree of uncertainty associated with the BEF
method of vegetation biomass estimation requires
further research. Finally, although the main impact
factors on the pattern of vegetation carbon density
were investigated using RBFN, more comprehensive
and accurate investigation into the driving
mechanisms of the distribution of vegetation carbon
density in karst regions are needed.
Acknowledgements
This study was supported by National Natural Science
Foundation of China (31370484, 31470519, 41471445), he
National Key Basic Research Program of China
(2015CB452703), the National Key Re- search and
Development Program of China (2016YFC0502400), the
Youth Innovation Team Project of ISA, CAS
(2017QNCX_ZJ), and the Public Service Technology
Center, Institute of Subtropical Agriculture, Chinese
Academy of Sciences.
References
Brown D.R., Dettmann P., Rinaudo T., Tefera H., Tofu A.,
(2011), Poverty alleviation and environmental
restoration using the clean development mechanism: a
case study from Humbo, Ethiopia, Environmental
Management, 48, 322-333.
Cai G.P., Han H.Q., Zhang F.T., Gao H.J., Zhu J., (2012),
Impacts of land use change on the vegetation carbon
storage in Suiyang, Guizhou province in the karst area,
Research of Soil and Water Conservation, 9, 122-125.
Cao J.H., Yuan D.X., Pei J.G., (2005), Karst Ecosystem
Constrained by Geological Conditions in Southwest
China, Geological Publishing House, Beijing.
Cao J., Zhang Y.L, Liu Y.H., (2002), Changes in forest
biomass carbon storage in Hainan Island over the last
20 years, Geographical Research, 21, 551-560.
Dixon R.K., Solomon A.M., Brown S., Houghton R.A.,
Trexier M.C., Wisniewski J., (1994), Carbon pools and
flux of global forest ecosystems, Science, 263, 185-190.
Fang J.Y., Chen A.P., Zhao S.Q., Ci L.J., (2002), Estimating
biomass carbon of China forests: supplementary notes
on report published in science (291:2320-2322) by
Fang et al.(2001), Acta Phytoecologica Sinica, 26, 243-
249.
Fang J.Y., Liu G.H., Xu H.L., (1996), Biomass and net
production of forest vegetation in China, Acta
Ecologica Sinica, 16, 497-508.
1665
 Zhang et al./Environmental Engineering and Management Journal 17 (2018), 7, 1657-1666
Haykin S., (1994), Neural Networks: A Comprehensive
Foundation, Macmillan, London.
Hu B.Q., Li L., Jiang S.F., (2005), Spatial pattern analysis
on Karst rocky desertification of Guangxi based on
landscape spatial method, Earth and Environment, 33,
581-587.
Hu Y.C., Liu Y.S., Wu P.L., Zou X.P., (2008), Rocky
desertification in Guangxi Karst mountainous area: its
tendency, formation causes and rehabilitation,
Transactions of the CSAE, 24, 96-101.
Huang C.D., Zhang J., Yang W.Q., Tang X., Zhao A.J.,
(2008), Dynamics on forest carbon stock in Sichuan
province and Chongqing city, Acta Ecologica Sinica,
28, 966-975.
Huang Q.H., Cai Y.L., (2005), Hazardous assessment of
karst rocky desertification in Guizhou Province: An
application of the RBFN Model, Acta Geographica
Sinica, 60, 771-778.
Jiao X.M., Xiang W.H., Tian D.L., (2005), Carbon storage
of forest vegetation and its geographical distribution in
Hunan province, Journal of Central South Forestry
University, 25, 4-8.
Li M.H.,Yu M.J., Chen Q.C., Chang J., Pan X.D., (1996),
Dynamics of carbon in the evergreen broadleaved forest
dominated by Cyclobalanopsis glauca in south-east
China, Acta Ecologica Sinica,16, 645-651.
Liu C.Q., Jiang Y.K., Tao F.X., Lang Y.C, Li S.L., (2008),
Chemical weathering of carbonate rocks by sulfuric
acid and the carbon cycling in Southwest China,
Geochimica, 37, 404-410.
Luo D.H., Xia J., Yuan J.W., Zhang Z.H., Zhu J.D., Ni J.,
(2010), Root biomass of karst vegetation in a
mountainous area of southwestern China, Chinese
Journal of Plant Ecology, 34, 611-618.
Mo J.M., Fang Y.T., Peng S.L., Brown S., Zhou G.Y.,
(2003), Carbon accumulation and allocation of lower
subtropical evergreen broad-leaved forests in a MAB
reserve of China, Acta Ecologica Sinica, 23, 1970-
1976.
Moody J., Darken C., (1989), Fast learning in networks of
locally-tuned processing units, Neural Computation, 4,
740-747.
1666
Park J., Sandberg I., (1991), Universal approximation using
radial basis function network, Neural Computation, 3,
246-257.
Piao S.L, Fang J.Y, Ciais P., Peylin P., Huang Y., Sitch S.,
Wang T., (2009) The Carbon balance of terrestrial
ecosystems in China, Nature, 458, 1009-1013.
Schwenker F., Kestler H.A., Palm G., (2001), Three learning
phases for radial-basis-function networks, Neural
Networks, 14, 439-458.
Tian D.L., Wang X.K., Fang X., Yan W.D., Ning X.B.,
Wang G.J., (2011), Carbon storage and spatial
distribution in different vegetation restoration patterns
in karsts area, Guizhou Province, Scientia silvae
sinicae, 47, 7-14.
Wang B., Wei W.J., (2007), Carbon storage and density of
forests in Jiangxi province, Jiangxi Science, 25, 681-
687.
Wang S.L., Yeager K.M., Wan G.J., Liu C.Q., Wang Y.C.,
Lv Y.C., (2012), Carbon export and HCO3- fate in
carbonate catchments: A case study in the karst plateau
of southwestern China, Applied Geochemistry, 27, 64-
72.
Wang Y.X., (2004), Carbon stock of main forest types in
Fujian province and carbon sequestration of
Cunninghamis Lanceolata Plantatieon, Agriculture
and Forestry University, Fujian (In Chinese.
http://cdmd.cnki.com.cn/Article/CDMD-10389-
2004109155.htm).
Yang C.M., (2003), A discussion on the remote sensing
analysis of karst stone desertification in Guangxi,
Remote Sensing for Land & Resources, 56, 34-36,63.
Ye J.S., She G.H., (2010), Forest carbon dynamics in
Guangdong province, Journal of Nanjing Forestry
University, 34, 7-12.
Zhang M.Y., Zhang C.H., Wang K.L., Yue Y.M., Qi X.K.,
Fan F.D., (2011), Spatiotemporal variation of karst
ecosystem service values and its correlation with
environmental factors in northwest Guangxi, China,
Environmental Management, 48, 933-944.
Environmental Engineering and Management Journal July 2018, Vol.17, No. 7, 1667-1677
http://www.eemj.icpm.tuiasi.ro/; http://www.eemj.eu

“Gheorghe Asachi” Technical University of Iasi, Romania

ANALYSIS OF COMMERCIALLY AVAILABLE

BOTTLED WATER IN POLAND

Rajmund Michalski ∗, Magdalena Jabłońska-Czapla,

Sebastian Szopa, Aleksandra Łyko

Institute of Environmental Engineering of the Polish Academy of Sciences

Skłodowska-Curie 34 Street, 41-819 Zabrze, Poland

Abstract

Bottled water (commonly called mineral) makes an important source of minerals for humans. Its consumption has been
systematically rising in the highly industrialized countries, even though not every water type has appropriate health properties. As
a result, drinking bottled water ought to be supported with understanding its qualities and physicochemical content.
The following article presents the results of the research into pH levels, electrical conductivity, selected inorganic anions (F-, Cl-,
NO3-, PO43-, SO42-), cations (Na+, NH4+, K+, Mg2+, Ca2+), metals and metalloids (Co, Ni, Cu, Zn, Cd, Pb, As, Cr, Mn, Ba, Rb, Sr,
Ag, Tl, V, U, Bi, Mo, Sb) of determined for the 48 bottled waters commercially available in Polish supermarkets. The methodology
optimization applied for the selected inorganic anions and cations as well as metals and metalloids helped to determine them on
the ultra-trace levels (<µg L-1). Inorganic ions were analyzed with the ion chromatography method. On the other hand, the ICP-MS
method was applied to analyze metals and metalloids. The validation parameters (limits of detection and quantification, recovery,
regression coefficients and coefficient of variation) were sufficient for the ultra-trace analysis. Due to the comparison of the
obtained results with the data provided by producers, numerous significant differences were observed.

Key words: bottled water, ion chromatography, inorganic ions, metals, trace elements

Received: May, 2013; Revised final: September, 2014; Accepted: September, 2014; Published in final edited form: July 2018

1. Introduction the world is expensive tap water. The problem is that

Water has vital importance for the human
organism as it participates in the transportation of
nutrients into cells and tissues. It also removes
unwanted metabolic products. It is delivered not only
by consuming various beverages, such as coffee or tea,
or eating different dishes, but also by drinking bottled
water, commonly called “mineral”. There is a great
number of the commercially available water types. To
choose the one that is beneficial to human health, it is
important to read the labels carefully. There are also
many types of table water, which are often confused
with mineral water. These are frequently distilled
waters which producers supply with mineral salts. It is
estimated that nearly 50% of the bottled water sold in
a liter of tap water is 250-600 times cheaper than a liter
of bottled water, even though its quality is often as
high as the one of the so-called mineral water. Bottled
water makes an important source of minerals for the
human organism. Its consumption has been
systematically rising in the highly industrialized
countries, even though not every water type has
appropriate health properties. The World Bank
estimates that the global mineral water market is worth
$ 800 billion per year. In 2011, nearly $ 4.5 billion was
spent on non-alcoholic beverages in Poland.
Therefore, the quality of drinking bottled water has to
be controlled systematically. Most tested mineral
waters available in the Polish market are compliant
with the required hygiene standards. The

∗
Author to whom all correspondence should be addressed: e-mail: rajmund.michalski@ipis.zabrze.pl; Phone: +48 32 2716481; Fax: +48 322717470
 Michalski et al./Environmental Engineering and Management Journal 17 (2018), 7, 1667-1677
nomenclature concerning the so-called “mineral
water” is equivocal (Diduch et al., 2011). The water is
usually classified as lowly, medium and highly
mineralized water. Curative, spring and table waters
are incorrectly included within this category. It is also
possible to classify water either as carbonated or non-
carbonated.
Researchers believe that nearly 20 properly
concentrated minerals should be present in the human
body to ensure proper functions of the human
organism and to maintain its homeostasis. Selected
ions and metals/metalloids constitute these minerals
(Birke, et al., 2010). Some of them can be found in
drinking and mineral water. Mineral waters usually
contain sodium, potassium, magnesium, calcium,
chlorides, sulfates, and several trace metals and
metalloids. If the mineral contents in the organism are
not appropriate, several imbalances may occur. Some
of them may lead to serious negative human effects
(WHO, 2005). When water is saturated with CO2, its
taste becomes slightly tart, bacteria development is
stopped and its expiry date is longer. CO2 improves
digestion and whets the appetite. As it enhances urine
production and helps to detoxify the organism, it is
advisable to drink after the physical effort. It should
be drunk in small portions as it is easily absorbed. On
the other hand, it is not recommended to people on a
diet and or those who suffer from the peptic ulcer
disease.
Non-carbonated water is appropriate for people
suffering from heartburn, flatulence and
cardiovascular disorders. Depending on its source,
mineral water contains various amounts of macro and
microelements (e.g. Mg, Na, K, Li, Ca, Sr, Zn, Mn, Fe,
Se, Cu, chlorides, sulfates, fluorides, iodides,
bromides, and nitrates) that have a significant impact
on living organisms. Magnesium regulates heart work
and participates in the thermoregulation and nerve
conduction. Potassium helps to control the water-
mineral balance, decreases blood pressure and
regulates nerve and muscle activities. Calcium has
impact on the muscle contractility, nerve impulse
Table 1. Regulations and standards for bottled water intended for human consumption
International
European Economic
Parameter Bottled Water
Community (2003)
Association (2009)
Physical parameters
Color - 5
pH - 6.5-8.5
Total dissolved
- 500
substances
Turbidity - 0.5
Disinfection by-products [µg/L)
Bromate - 10
Chlorine - 100
Chlorite - 1,000
Haloacetic acids - 60
Trihalomethanes
- 10
(total)
Inorganic chemicals [mg/L)
Aluminum - 0.2
1668
transmission, blood coagulation and it is contained in
bones and teeth. Iron is essential for the transportation
and storage of oxygen in the organism. Zinc, copper
and manganese take part in the metabolic processes
while chlorides regulate blood pH. On the other hand,
people suffering from the elevated blood pressure and
kidney problems should not drink sodium-rich water
as this element increases blood pressure and
contributes to water retention and edemas. The
occurrence of those substances in water has usually a
positive impact on human health. Unfortunately,
consumers are not informed on the presence and
concentrations of harmful substances that may also
occur in water.
The maximum concentrations of the potentially
harmful substances, i.e. those whose presence in water
must be controlled in the Polish bottled waters, are
listed in Annex 1 of the appropriate Directive
(Regulation of the Minister of Health of 31 March
2011). The requirements concerning determination
methods of the potentially toxic substances are given
in Annex 2. Spring water undergoes the same standard
analyses as the natural mineral water. Spring water at
the intake and in bottles must meet specific
microbiological, chemical, chemicophysical and
organoleptic requirements defined for drinking water.
Bottled table water must meet microbiological
requirements.
It must also be compliant with the regulations
on the permissible levels of chemical components,
which are specifically defined for drinking water and
whose excess could pose a threat for the public health.
Mineral components that can be used in the table water
production are: calcium chloride, magnesium chloride
or sulfate, chlorides, sulfates, or potassium and
sodium bicarbonates. Their chemical purity must
comply with the regulations for foodstuff. Bottled
water can be admitted for consumption only when the
concentrations of its constituents do not exceed the
standards specified in various legal regulations. Table
1 provides a list of such standards recommended both
by the international and Polish institutions.
Polish legalization
U.S. Food and Drug
Dz. U. Nr 85, poz.466
Administration (2010)
(May 7th, 2011)
15 5
6.5-8.5 6.5-8.5
500 -
5 -
10 3
4000 -
1000 -
60 -
80 -
0.2 -
 Analysis of commercially available bottled water in Poland

Ammonium - - - -
Antimony 0.005 0.006 0.006 0.005
Arsenic 0.01 0.01 0.01 0.01
Barium - 1 2 1
Beryllium - 0.004 0.004 -
Boron 1 - - 5
Cadmium 0.003 0.005 0.005 0.003
Chloride - 250 250 -
Chromium 0.05 0.05 0.1 0.05
Copper 1 1 1 1
Cyanide 0.07 0.1 0.1 0.07
Fluoride 5 - - 5
Iron - 0.3 0.3 -
Lead 0.01 0.005 0.005 0.01
Manganese 0.5 0.05 0.05 0.5
Mercury 0.001 0.001 0.002 0.001
Molybdenum - - - -
Nickel 0.02 0.1 0.1 0.2
Nitrate 50 50 50 50
Nitrite 0.1 1 1 0.1
Selenium 0.01 0.01 0.05 0.01
Silver - 0.025 0.1 -
Sodium - - - -
Sulfate - 250 250 -
Thallium - 0.002 0.002 -
Uranium - 0.03 0.03 -
Zinc - 5 5 -
Organic chemicals [µg/L)
Acrylamide - - - -
Benzene - 0.001 0.001 -
Pesticides (the sum
- - - -
of)
Phenols - 0.001 0.001 -

The analytical chemists must develop reliable
and conclusive methods for determinations a wide
range of substances present in bottled waters in the
qualitative and quantitative ways. Determining
contents of inorganic and organic compounds is
difficult, due to: low levels of individual compounds
present in samples; complex compositions (high
mineralization degree); interactions between the
constituents present in samples; and possible changes
in the composition during transport and storage
resulting from reactions and potential desorption of
constituents from the packaging materials (Birke at al.,
2010; Muntean and Mihaiescu, 2016). Therefore, it is
necessary to use appropriate, sensitive and selective
methods, such as gas and liquid chromatography
coupled with the mass spectrometry detectors (Jin et
al., 2010). The aim of the study was to test 48 mineral
waters available in the Polish market to determine the
contents of selected inorganic anions and cations,
several dozens of trace elements, and the pH and
electrical conductivity values. Additionally, the data
obtained from water producers and the research results
was compared whenever it was possible.
2. Experimental
2.1. Research subject and range
A number of 48 types of bottled water (mineral,
table, spring and curative ones) were bought in a few
Polish supermarkets. The samples were determined in
terms of the electrical conductivity, pH and contents
of the main inorganic anions (F-, Cl-, NO3-, PO43- and
SO42-), cations (Li+, Na+, NH4+, Mg2+ and Ca2+),
selected metals and metalloids (Co, Ni, Cu, Zn, Cd,
Pb, As, Cr, Mn, Ba, Rb, Sr, Ag, Tl, V, U, Bi, Mo, Sb).
2.2. Research methods
Electrical conductivity and pH were measured
with the multifunction CX-401 meter (Elmetron,
Poland) with the ERH 111 electrode (Hydromet,
Poland) and CD-2 conductivity detector (Hydromet,
Poland). Inorganic ions were determined with an ion
chromatograph (Metrohm, Switzerland) equipped
with: IC 818 pump, IC 837 eluent degasser, IC 830
interface, IC 820 separation center, Valco injection
valve, Metrodata 2.3 software, IC 838 automatic
sample feeder and IC 819 conductivity detector. The
Institute of Environmental Engineering of the Polish
Academy of Sciences in Zabrze, where analyses were
carried out, has the laboratory that possesses the ISO
17025 ISO (ISO standard, 2005) accreditation of the
Polish Accreditation Centre (PCA), number 950.
The analytical conditions for the determination
of inorganic ions were:
− Anions (F-, Cl-, NO3-, PO43-, SO42-);
− Column-Metrohm Metrosep A Supp 3;
− Eluent-1.7 mM Na2CO3 + 1.6 mM NaHCO3;

1669
 Michalski et al./Environmental Engineering and Management Journal 17 (2018), 7, 1667-1677

− Eluent flow rate-0.85 mL/min;
− Injection volume-20 µL;
− Detection-suppressed conductivity;
− Cations (Li+, Na+, NH4+, K+, Mg2+, Ca2+);
− Column-Metrohm Metrosep C3;
− Eluent-5 mM HNO3;
− Eluent flow rate-1.0 mL/min;
− Injection volume-100 µL;
− Detection-non-suppressed conductivity.
The contents of selected metals and metalloids
were determined with the ICP-MS technique (ICP-MS
Elan 6100 DRC-e Perkin Elmer spectrometer). The
ICP-MS spectrometer analytical parameters are given
in Table 2.
Table 2. ICP-MS spectrometer parameters
Parameter Settings
RF generator power [W] 1,125
Argon - plasma gas flow [L/min] 15
Argon - nebulizer gas flow [L/min] 0.76-0.82
Argon - auxiliary gas flow [L/min] 1.15-1.16
Nebulizer cross-flow
Torch quartz
Scanning mode peak hopping
Dwell time [ms] 250
Sweeps/Reading 1 2.4. Sample preparation
Number of replicates 830
2.3. Reagents
NaHCO3 and Na2CO3 of analytical purity were
used to prepare the eluent for the inorganic anion
determinations. Ultrapure HNO3 (POCh, Gliwice,
Poland) was employed to prepare the cation
determination eluent. The properly composed and
concentrated eluents were prepared by weighing an
appropriate sample mass (or collecting appropriate
volume) of a given substance. It was then solved in a
2-L laboratory flask. Standard solutions of specific
anions (F-, Cl-, NO3-, PO43- and SO42-) and cations
(Li+, Na+, NH4+, K+, Mg2+, Ca2+) were made with the
Fluka reference materials (Steinheim, Switzerland).
Their concentrations amounted to 1000 ± 2 mg/L.
Water used for standards and eluents preparation came
from the Millipore deionizer (Merck, Germany). Its
electrical conductivity was < 0.05 μS/cm. Calibration
solutions were made by diluting appropriate standard
solutions right before their application. All solutions
were kept in glass or high-density polyethylene
(HDPE) containers at room temperature.
The ICP-MS spectrometer was optimized daily
with a 10-µg/L solution (Mg, Cu, Rh, Cd, In, Ba, Ce,
Pb, U) in 1% HNO3 Elan 6100 Setup/Stab./Masscal.
Solution (Perkin-Elmer). An application enabling
measurements of 53Cr, 55Mn, 59Co, 60Ni, 65Cu,
66Zn, 69Ga, 75As, 85Rb, 88Sr, 98Sr, 107Ag, 114Cd,
121Sb, 130Te, 138Ba, 205Tl, 208Pb, 209Bi, 238U
isotopes was prepared. To determine Pb in a
quantitative way, a correction equation
(Pb208=Pb208+Pb206+Pb207) was employed. The
analyses were carried out with the internal standard
method, which was constituted by 10 µg/L Rh solution
introduced on-line with a peristaltic pump. All
solutions and standards were made with the high-
quality deionized Milli-Q-Gradient Millipore water
(Merck, Germany). 1 g/L single-element standard
solutions of Cd, Pb, Zn, Sb (Merck, Germany) and a
certified multi-element standard solution were used
for calibration. It was obtained by diluting 10 mg/L of
the VI calibration solution (Merck, Germany).
Calibration was verified with another certified multi-
element standard, i.e. XXI (Merck, Germany). All
standards were prepared daily with weight dilution.
The method validation was performed on the basis of
the certified reference material - NIST 1643e Trace
Elements In Water.
The carbonated water samples were degassed
in a sonic bath before the analysis. The examined
water samples were filtered through a 0.2-µm
nitrocellulose filter. Subsequently, they underwent
chromatographic and ICP-MS analyses. Water
samples were acidified with the ultra-pure
concentrated nitric acid (Merck, Germany) before the
quantitative analysis with ICP-MS spectrometer.
2.5. Validation of the applied analytical methods
Taking into consideration the inorganic ion
determination methodologies, the calibration was
performed in compliance with the ISO 8466-1 (ISO
Standard, 1990). 7 standard solutions which contained
various concentrations of given ions were prepared to
define validation parameters for the methodology of
the inorganic anion determination. Each solution went
through the chromatographic analysis three times. The
obtained peak areas of the specific ions served to
calculate: standard deviation, variation coefficient,
limits of detection and quantification, and correlation
coefficient. The results are given in Table 3.

Table 3. Validation parameters for the ion chromatographic method (inorganic ions)

Anions Cations
Parameter
F- Cl- NO3- PO43- SO42- Na+ NH4+ K+ Mg2+ Ca2+
Concentration range [mg/L] 0.2-2.0 1-100 0.2-20 0.2-20 1-100 1-100 0.2-2.0 0.5-50 1-100 1-100
Standard deviation [mg/L] 0.019 0.232 0.079 0.029 0.132 0.098 0.159 0.102 0.081 0.115
Coefficient of variation [%] 2.68 4.57 2.76 4.82 3.17 3.82 4.94 1.84 3.67 3.43
Limit of detection [mg/L] 0.012 0.064 0.048 0.019 0.131 0.091 0.027 0.095 0.17 0.211
Limit of quantification [mg/L] 0.036 0.192 0.144 0.057 0.353 0.273 0.081 0.285 0.51 0.633
Regression coefficient (r2) 0.999 0.999 0.999 0.999 0.999 0.999 0.998 0.999 0.999 0.998

1670
 Analysis of commercially available bottled water in Poland

The ICP-MS method validation was performed
on the basis of the certified reference material - NIST
1643e Trace Elements in Water. The validation
parameters are given in Table 4.
2. Results and discussion
25 non-carbonated, 12 lightly carbonated and
11 carbonated waters constituted the analyzed group
of the 48 water brands. The bottle volume was usually
0.5 L or 1.5 L, however there were also 0.33-L and
2.5-L bottles. As it is apparent from the data in Table
1, the requirements of various organizations for the
quality of bottled waters are highly diverse.
The qualitative and quantitative ranges of
substances regulated by the International Bottled
Water Association (IBWA) and U.S. Food and Drug
Administration (U.S. FDA) are much greater than the
ranges defined by the European Economic
Community and Polish regulations. Fig. 1 presents
conductivity and pH values for the analyzed water
samples. The results of the selected inorganic anion (F-
, Cl-, NO3-, PO43- and SO42-) and cation (Li+, Na+,
NH4+, Mg2+ and Ca2+) determinations are shown in
Figs. 2 and 3, respectively. The water producer values
given in Figs. 2 and 3 are marked in red, and the black
color concerns the analysis data.

Table 4. Validation parameters for metals and metalloids determined with ICP-MS

Limit of Limit of Relative standard deviation Relative standard deviation of

Element Recovery [%]
detection [µg/L] quantification [µg/L] of reproducibility [%] repeatability [%]
Co 0.002 0.007 6.4 3.3 116.43
Ni 0.024 0.072 3.8 3.2 82.09
Cu 0.064 0.191 10.4 6.3 96.07
Zn 0.181 0.543 4.6 4.6 80.27
Cd 0.04 0.12 4.6 3.9 87.99
Pb 0.036 0.107 5.8 3.5 108.08
As 0.096 0.289 11.4 7.6 106.95
Cr 0.013 0.039 5 4.8 109.63
Se 0.041 0.123 3.1 3.9 96.81
Mn 0.033 0.098 5.1 5.9 107.31
Ba 0.01 0.03 6.9 3.2 111.14
Rb 0.003 0.008 5.3 13.9 104.62
Sr 0.008 0.025 8.2 6.5 95.39
Ag 0.002 0.007 7.1 3.9 107.02
Tl 0.002 0.006 13.9 11.3 105.67
V 0.09 0.27 5.5 5.9 103.25
U 0.011 0.033 4.7 3.7 -
Te 0.008 0.024 6.3 5.1 101.44
Bi 0.018 0.054 5.8 5.5 98.36
Ga 0.042 0.126 5.2 4 -
Mo 0.027 0.081 9.7 8.1 99.87

(a)

(b)

Fig. 1. Conductivity and pH values of the analyzed samples: (a) pH, (b) conductivity

1671
 Michalski et al./Environmental Engineering and Management Journal 17 (2018), 7, 1667-1677

(a) (b)

(c) (d)

Fig. 2. Concentrations of common inorganic anions in the analyzed samples: (a) F-, (b) Cl-, (c) NO3- (d) SO42-

(a) (b)

(c) (d)

Fig. 3. Concentrations of common inorganic cations in the analyzed samples: (a) Na+, (b) K+, (c) Mg2+, (d) Ca2+

Figs. 4-8 present the determination results for
the selected metal and metalloid concentrations.
Before deliberating over the trace elements, some
major elements should be discussed as they are usually
reported on the label. Manufacturers do not give the
information on the metal and metalloid
concentrations, which was the subject of this research.
Usually, only the content of the main inorganic anions
and cations is given. The graphical comparison of the
study results for inorganic ions with the values
reported by producers on the bottled water labels is
shown in Fig. 9.
In some cases, the same water type was
analyzed but either the bottles in which it was sold

1672
 Analysis of commercially available bottled water in Poland

differed (samples 11 and 29, 28 and 33, 30 and 31) or
the water was saturated with carbon dioxide to a
different extent (samples 4 and 15, 16 and 40, 34 and
48). Most of the examined water types came from
Poland but some were produced in the Czech Republic
(sample 24) or Lithuania (samples 17, 22, 30 and 31).
Multinational corporations, such as PepsiCo (samples
4, 15 and 43), Coca-Cola (samples 10 and 26) and
Nestle (samples 8, 11, 29 and 46), produce several
popular water types in Poland. The obtained pH values
ranged between 4.69 (sample 43) and 8.28 (sample 9).
Higher pH values were determined for non-carbonated
water, while lower ones were observed for carbonated
water (Fig. 1). A classic example is sample 4. The non-
carbonated water version had pH of 7.91, but when
saturated with carbon dioxide (sample 43), it was only
4.69. There were slighter differences in the case of the
lightly carbonated sample 34 (pH= 5.29) and
carbonated sample 48 (pH=5.59). The average value
of electrical conductivity for non-carbonated water
was < 1,500 µS/cm. What seems interesting is the fact
that this value was the same for a few carbonated water
types (Fig. 1). Samples 30 and 31 had definitely higher
values, which were > 5,000 µS/cm.

(a) (b)

(c) (d)

Fig. 4. Concentrations of V, Cr, Mn and Co in the analyzed samples: (a) V, (b) Cr, (c) Mn, (d) Co

(a) (b)

(c) (d)

Fig. 5. Concentrations of Ni, Cu, Zn and Ga in the analyzed samples: (a) Ni, (b) Cu, (c) Zn, (d) Ga

1673
 Michalski et al./Environmental Engineering and Management Journal 17 (2018), 7, 1667-1677

(a) (b)

(c) (d)

Fig. 6. Concentrations of Rb, Sr, As and Mo in the analyzed samples: (a) Ru, (b) Sr, (c) As, (d) Mo

(a) (b)

(c) (d)

Fig. 7. Concentrations of Ag, Sb, Te and Ba in the analyzed samples: (a) Ag, (b) Sb, (c) Te, (d) Ba

(a) (b)

(c) (d)

Fig. 8. Concentrations of Tl, Bi, Pb and U in the analyzed samples: (a) Tl, (b) Bi, (c) Pb, (d) U

1674
 Analysis of commercially available bottled water in Poland
10000
1000
100
Concentration
10
1
F- Cl-
0,1
0,01
Fig. 9. Graphical comparison of the study results with the values reported on the bottled water labels. The first line applies
to the results obtained in the laboratory; the second (parallel) line concerns the value declared by the manufacturer
for a given parameter, if it was stated on the bottle label
The average contents of F- ions for most water
types did not exceed 0.5 mg/L (Fig. 2). Higher values,
close to the limit, were detected for samples 23 and 31.
When comparing the obtained results with the data
provided by producers, numerous significant
differences were observed. In many cases, the
information given by producers (if specified in the
label) indicated lower values than those determined in
the research process. There may be multiple
explanations of this fact. Nevertheless, the most
important one seems to be the fact that producers
present information (analyses are performed only
when necessary) which is very convenient as they
possibly assume that the consumers are not be able to
verify the actual data. The water composition at
intakes may change. It happens not only in emergency
situations. Moreover, the cost of such analyses is high.
The Directive allows for deviations from the declared
characteristic component contents that are not higher
than 20%. The average content of chloride ions did not
exceed 20 mg/L (Fig. 2). However, an unusually high
value (over 1.0 g/L) was observed for samples 30 and
31. The analyzed water types did not contain NO2-
ions, whereas the NO3- content ranged between 0.51
mg/L (sample 28) and 11.41 mg/L (sample 39).
Unfortunately, only one manufacturer presented the
declared nitrate ion content in the produced water (Fig.
2). The PO43- content was not indicated in any label.
Their contents varied between 1.28 and 1.56 mg/L and
only three examined water types had PO43- levels
below the detection limit.
The determined sulfate ion values did not differ
as much as the Cl- ion contents and ranged between
1.51 mg/L (sample 1) and 80.85 mg/L (sample 31). In
most cases, the declared values of sulfate ions were
consistent with the calculated ones. Fig. 3 shows
results of the selected cation determinations. Lithium
and ammonium ions (over the detection limit) were
not present in any sample. The medium concentration
of Na+ ions was lower than 20 mg/L for non-
carbonated water. In the remaining cases, it varied
between 50 and 270 mg/L (sample 36).
max
min
mediana
NO3- SO42- Na+ K+ Mg2+ Ca2+ pH
Ions
The highest content of K+ ions (31.68 mg/L)
was observed for sample 25, even though the producer
had declared the value of 37.5 mg/L. In most cases,
the K content did not exceed 10 mg/L and was usually
the same as the one given in the label (Fig. 2). Calcium
and magnesium make important constituents of the so-
called mineral water. Most producers indicated their
contents but they were often higher than the
determined ones (Fig. 3). The highest Mg value
(131.14 mg/L) was determined in sample 24 but it was
still much lower than the declared one (179 mg/L). A
similar situation was observed for Ca2+ ions. The
contents did not exceed 100 mg/L in most water types.
Interestingly, these values were usually lower than the
declared ones. The highest Ca2+ level was found in
samples 33 (238.13 mg/L; producer data - 369 mg/L)
and 34 (138.57 mg/L; producer data-129 mg/L).
Sample 5 is interesting in terms of the Ag
content. Its value reached 0.7 µg/L. The average value
for the remaining 45 samples was < 0.1 µg/L (Fig. 7).
Similar deviations were observed for Te (sample 36,
Fig. 7), Ba (sample 36, Fig. 7), Tl (sample 21, Fig. 8)
and Ga (sample 36, Fig. 5). The Pb content did not
exceed 0.2 µg/L in most cases. However, it was higher
for samples 20 and 44 (Fig. 8). Bismuth was detected
in 9 water types (Fig. 8). The highest Mo content was
determined in sample 39, while the median value was
< 0.5 µg/L (Fig. 6). Interestingly, U was detected in
almost all water types (Fig. 8). The highest U value
(approx. 1.2 µg/L) was observed for samples 3, 7 and
39. It did not exceed 0.2 µg/L in most remaining
samples.
Most water types contained Cr (< 1 µg/L).
However, its content for samples 44 and 46 was
approx. 20 µg/L (Fig. 4). A slighter diversification was
noticed for Sb. Its average concentrations ranged
between 0.3 and 0.8 µg/L (Fig. 7). Its highest value
was observed for sample 45. Taking into account the
quality requirements for bottled water, Table 5
summarizes the number of times when the limit values
for all the analyzed waters was exceeded.
1675
 Michalski et al./Environmental Engineering and Management Journal 17 (2018), 7, 1667-1677

Table 5. Number of times when water standards were exceeded

International Polish legalization

European Economic U.S. Food and Drug
Parameter Bottled Water Dz. U. Nr 85, poz.466
Community (2003) Administration (2010)
Association (2009) (May 7th, 2011)
Antimony 13 8 8 13
Arsenic 3 3 3 3
Barium - 8 4 8
Chloride - 2 2 -
Chromium 0 0 0 0
Copper 0 0 0 0
Fluoride 0 0 0 0
Lead 0 0 0 0
Manganese 3 9 9 3
Nickel 3 0 0 0
Nitrate 1 1 1 1
Silver - 0 0 -
Sulfate - 0 0 -
Thallium - 0 0 -
Uranium - 0 0 -
Zinc - 0 0 -

In the literature, there are various examples of
modern analytical techniques (including ion
chromatography) applied to examine mineral water
compositions. The studies concerned Slovenia (Gros
and Gorenc, 1994), South Korea (Jong-Hun, 1997),
Turkey (Guler, 2007; Guler and Alpaslan, 2009), Sri
Lanka (Herath et al., 2012), Italy (Dinelli et al., 2012),
the USA (Saleh, 2008), or Arabic countries (Shomar,
2012). The Polish mineral water was also investigated
(Astel et al., 2014; Ciężkowski et al., 2007; Michalski,
2003), but the research range was different from that
described in the following paper. In conclusion, there
is a steady tendency related to the bottled water
consumption. Hopefully, the decision whether to drink
it and which one to drink could be facilitated by the
findings provided in this article.
4. Conclusions
In the following study, 48 various carbonated,
lightly carbonated and non-carbonated mineral waters
available in the Polish market were examined. The
researchers determined the pH and electrical
conductivity values, and the concentrations of the
main inorganic ions, metals and metalloids.
Most of the measured parameter values were
below the permissible limits defined by the WHO, EU
and Polish institutions. For the metal and metalloid
contents, some values were exceeded in a few cases
(antimony – 13 waters; barium – 8 waters; manganese
– 3 waters; and arsenic – 3 waters). No value exceeded
the permissible limits for the inorganic anions and
cations. Nonetheless, it must be stated that the
qualitative research range embraced a much larger
number of analytes than that to which the regulations
on mineral waters apply.
The comparison of the analysis results with the
water label data showed a significant discrepancy
between the declared and measured values (for ions,
metals and metalloids). Importantly, it should be
noticed that the mineral water composition may
change due to the long-term storage, which could be
investigated in the further research.
Abbreviations
FDA-United States Food and Drug Administration
IBA-International Bottled Water Association
IC-Ion Chromatography
ICP-MS-Inductively Coupled Plasma Mass Spectrometry
ISO-International Standardization Organization
NIST-National Institute of Standards and Technology
PCA-Polish Accreditation Centre
WHO-World Health Organization
References
Astel A., Michalski R., Łyko A., Jabłońska-Czapla
M., Bigus K., Szopa S., Sarapata A., Kwiecińska
A., (2014), Characterization of bottled mineral
waters marketed in Poland Using hierarchical
cluster analysis, Journal of Geochemical
Exploration, 143, 136-145.
Birke M., Reimann C., Demetriades A., Rauch U., Lorenz
H., Harazim B., Glatte W., (2010), Determination of
major and trace elements in European bottled mineral
water - analytical methods, Journal of Geochemical
Exploration, 107, 217-226.
Ciężkowski W., Chowaniec J., Górecki W., Krawiec A.,
Rajchel L., Zuber A., (2010), Mineral and thermal
waters of Poland, Przegląd Geologiczny, 58, 762-773.
Diduch M., Polkowska Ż., Namieśnik J., (2011), Chemical
quality of bottled waters: A Review, Journal of Food
Science, 76, 178-196.
Dinelli E., Lima A., Albanese S., Birke M., Cicchella D.,
Giaccio L., Valera P., De Vivo B., (2012), Comparative
study between bottled mineral and tap water in Italy,
Journal of Geochemical Exploration, 112, 368-389.
Gros N., Gorenc B., (1994), Ion chromatographic
determination of alkali and alkaline earth metals in
mineral waters, Chromatographia, 39, 448-452.
Guler C., (2007), Evaluation of maximum contaminant
levels in Turkish bottled drinking waters utilizing

1676
 Analysis of commercially available bottled water in Poland

parameters reported on manufacturer’s labeling and column ion chromatography, Journal of

government-issued production licenses, Journal of
Food Composition and Analysis, 20, 262-272.
Guler C., Alpaslan M., (2009) Mineral content of 70 bottled
water brands sold on the Turkish market: Assessment
of their compliance with current regulations, Journal of
Food Composition and Analysis, 22, 728-737.
Herath A. T., Abayasekara C. L., Chandrajith R., Adikaram
N. K. B., (2012), Temporal variation of microbiological
and chemical quality of noncarbonated bottled drinking
water sold in Sri Lanka, Journal of Food Science, 77,
160-164.
ISO Standard (1990), Water quality - Calibration and
evaluation of analytical methods and estimation of
performance characteristics Part 1: Statistical
evaluation of the linear calibration function, ISO 8466-
1:1990, On line at:
https://www.iso.org/standard/15664.html
ISO Standard (2005), General requirements for the
competence of testing and calibration laboratories,
ISO/IEC 17025:2005, On line at:
https://www.iso.org/standard/39883.html.
Jin B.-H., Xiao F., Chen B., Chen P.-J., Xie L., (2010),
Simultaneous determination of 42 organic chemicals in
bottled water by combining C18 extraction disk with
GC-MS and LC/MS/MS technique, Journal of Water
and Health, 8, 116-125.
Jong-Hun K., Joung-Hae L., (1997), Simultaneous
determination of six cations in mineral water by single-
Chromatography A, 782, 140-146.
Michalski R., (2003), Selected anions and cations in mineral
waters, Archives of Environmental Protection, 29, 8-18.
Muntean E., Mihaiescu T., (2016), Groundwater quality
studies in two Transylvanian rural communities using
parallel ion chromatography, Environmental
Engineering and Management Journal, 15, 2703-2708.
RMH, (2011), Regulation of the Minister of Health on 31
March 2011, the natural mineral waters, natural spring
waters and table waters, Official Journal, No. 85 pos.
466th On line at: http://www.infor.pl/akt-
prawny/DZU.2011.085.0000466,metryka,rozporzadze
nie-ministra-zdrowia-w-sprawie-naturalnych-wod-
mineralnych-wod-zrodlanych-i-wod-stolowych.html.
Saleh M.A., Abdel-Rahman F.H., Woodard B.B., Clark S.,
Wallace C., Aboaba A., Zhang W., Nance J.H., (2008),
Chemical, microbial and physical evaluation of
commercial bottled waters in greater Houston area of
Texas, Journal of Environmental Science and Health,
Part A, 43, 335-347.
Shomar B., (2012), Zamzam water: Concentration of trace
elements and other characteristics, Chemosphere, 86,
600-605.
WHO, (2005), Nutrients in Drinking Water, Water,
Sanitation and Health Protection and the Human
Environment, On line at:
http://www.who.int/water_sanitation_health/dwq/nutri
entsindw.pdf.

1677
1678
Environmental Engineering and Management Journal July 2018, Vol.17, No. 7, 1679-1683
http://www.eemj.icpm.tuiasi.ro/; http://www.eemj.eu

“Gheorghe Asachi” Technical University of Iasi, Romania

SURVEY OF INDUSTRIAL WASTEWATER TREATMENT

PERFORMANCE WITH BIOLOGICAL MONITORING OF PROTOZOA
IN THE INDUSTRIAL PARK OF ABADAN

Abotaleb Bay1, Mahdi Sadeghi2∗, Hadi Rahimzade2, Nasser Behnampour3

1Department of Environmental Health Engineering, Faculty of Health,
Shahid Beheshti University of Medical Sciences, Tehran, Iran
2Environmental Health Research Center, Golstan University of Medical Sciences, Gorgan, Iran
3Department of Biostatistics, School of Health, Golstan University of Medical Sciences, Gorgan, Iran

Abstract

Industrial wastewater is of particular importance. In most wastewater treatments, physicochemical parameters are used to determine
the performances which are costly while protozoa can be used as less expensive biological indicators to determine the efficiency
of the treatment plant. The purpose of this study was to investigate the relationship between protozoa and physic-chemical
characteristics of the Abadan wastewater treatment industry.
Samples were collected from wastewater stabilization ponds in the Abadan industrial park. A total of 66 samples for biochemical
and chemical tests and protozoa counts were taken. The sites of sampling were influent of sewage and effluent from maturation
pond. For the parameters from influent, first 300 mL of samples was collected in depths of 0.1, 0.35 and 1 m and then mixed to
provide a composite sample. 300 mL sample was directly taken from the effluent.
The results of this study indicate that there is a positive relationship between the number and population percentage of amoeba and
mastigophora (flagellate) with BOD and COD, whereas is the negative relationship between sticking ciliates with BOD and COD.
Furthermore there is a positive relation between DO, the percentage, number of ciliates and amoeba population and there is a
correlation, with flagellate population. Increase of flagellate numbers is indicative of high effluent Kjeldahl nitrogen. There was
also an increase of input TSS and amoeba population percentage and a reduction of free ciliates numbers.
The type and abundance of protozoa in industrial wastewater treatment following biological methods is an indicator of the system
performance and effluent quality. This study shows the use of biological indicators can be used for monitoring as a much cheaper
alternative resulting in a better economy. It is also a good indicator for the detection of toxic materials entering the plant or refinery
and also its organic load shocks.

Key words: Abadan, amoeba, ciliate, industrial wastewater, mastighophora, monitoring, protozoa,

Received: August, 2013; Revised final: September, 2014; Accepted: September, 2014; Published in final edited form: July, 2018

1. Introduction physicochemical parameters. It has also been well

The nature of industrial waste disposal for
elimination in the environment is of special interest
and nowadays, due to rapid industry and population
growth, the importance of wastewater treatment is
even more significant than before. In most refineries
the correct treatment filtration is determined by the
established that industrial wastewater contains high
concentration of dissolved organic and various toxic
compounds (Luna-Pabello et al., 1990; Popescu et al.,
2009). For the aerobic treatment of wastewater,
protozoa play a significant role in removal of colloidal
and suspended organic matter (Madoni, 1994; Luna-
Pabello et al., 1990). It is demonstrated that the

∗
Author to whom all correspondence should be addressed: e-mail:mahdikargar1@gmail.com; Phone:+9817- 32431602; Fax: +981732431607
 Bay et al./Environmental Engineering and Management Journal 17 (2018), 7, 1679-1683
presence of organisms such as protozoa, their
abundance and species diversity are associated with
quality, type of waste and emissions which can be used
to evaluate the performance of refineries or plants
(Puigagut et al., 2005). In recent decades, regulatory
authorities in industrialized regions have also
endeavored to improve water quality by advanced
wastewater treatment (Eslami et al., 2016; Foley et al.,
2010). Therefore, evaluation and monitoring of
industrial wastewater treatment is necessary. The
reason that the use of stabilization ponds in
wastewater treatment are growing rapidly around the
globe and especially in tropical areas is due to their
methods of function.
Since the type and frequency of biological
agents represent the performance of wastewater
treatment plants, therefore the possibility of
monitoring refineries performance by measuring
protozoan is possible. So due to its low cost and high
speed, this technique for monitoring of wastewater
treatment plants have been recommended
(Papadimitriou et al., 2010; Zhou et al., 2006). Variety
of studies such as the following research has reported
the existence of protozoa in sewage in many countries.
Papadimitriou et al. (2010) studied protozoa as
an indicator of wastewater treatment efficiency in
wetlands. They also showed there is a significant
relationship between the total coliform and
phosphorus removal rate and by increasing the
removal of loaded organics, the number and variety of
protozoa also increase. Dubber et al. (2011) in Ireland
during the study of SBR reactors evaluated the effect
of anoxic and anaerobic process on protozoa
population. They measured chemical and biochemical
parameters such as BOD, COD, TKN in four different
times of the reactor use and also determined the
abundance and diversity of protozoan. Their studies
showed that prolonged anoxic and anaerobic
processes were reducing the entire population of
protozoa in the reactor. Chen et al. (2005) conducted
a survey on wastewater sturgeon fish in south of
China. They used protozoa as bio-indicators to deposit
effluent quality assessment. They found that the
protozoa populations are dependent on distribution
and pollution rotation period that this property can be
used as a treatment bio-indicator. They also reported
that more than 80 percent of the 216 ciliates species
are omnivorous (Chen et al., 2005). Perez-Uzet al.
(2010) studied the likely assessment of bio-indicators
to determine the performance of treatment in advanced
wastewater treatment systems.
Their results showed that the reduction of
nitrogen removal efficiency increased ciliates
population or the high existence of Ciliates resulted in
low efficiency in system nitrogen removal. Poole
compared the species of parasites where they operated
activated sludge systems in which nitrification and
operating systems were not performed and indicator
organisms were introduced to the system (Poole,
1984). The organic load entering relation to plant and
protozoa population were studied by Salvador and
Gracia (1993) in Spain.
1680
Since the use of stabilization pond treatment
system is rapidly increasing specially in the tropical
areas and due to the difficulties and the high cost of
wastewater quality control testing, in this study, we
have tried to propose appropriate solutions for
monitoring and evaluation of system's performance by
investigating the relationship between protozoan and
physicochemical indicators.
2. Material and methods
In this study, samples were collected from
wastewater stabilization ponds in the Abadan
industrial park. Abadan is a region with a warm
climate in the Iran. The pool was made of concrete and
the average of raw sewage was 800 cubic meters per
day.
The major wastewater productions were from
soft drink industry, tuna fish and dairy and food
industry. The pond design specifications are given in
Table 1.
2.1. Sampling
In this study the samples were collected within
four months (during spring and summer). Eight
samples were collected per month through duplicate
sampling. A total of 66 samples for BOD, COD, TSS,
TKN tests and protozoa counts were taken. DO and
pH parameters were measured daily in place by a Hach
portable. The sites of sampling were influent of
sewage and effluent from maturation pound. For the
parameters from influent, 300 mL of samples were
collected in a depth of 0.1, 0.35 and 1 m and mixed
that to provide a composite sample and 300 mL
sample was taken from the effluent directly.
2.2. Determination of physicochemical parameters
For measuring BOD, COD and TKN, the
samples were filtered by fiber glass filters. In order to
determine the amount of TSS the effluent was passed
through Whatman filter and then dried at 103 to 105°C
and the weight difference filter was calculated. All
physic-chemical parameters were measured according
to standard methods (APHA, 2005).
2.3. Protozoan population monitoring
For protozoa counting after stirring the
sample with distilled water, the sample was prepared
diluted in a ratio of 1 to 3. Then 1 mL of the diluted
sample was placed on counting slide of Sedgwick-
Rafter and observed with optical microscope with
100X enlarging and protozoa were counted.
Field method was used to count protozoa and
10 randomly selected fields and protozoa counts were
performed in four major categories of moving ciliates
(free), sticking ciliates, amoeba and mastigophore
(flagellate). This count for each sample was repeated
twice and the results were calculated based on the
number of mL from Eq. (1) (APHA, 2005).
 Survey of industrial wastewater treatment performance with biological monitoring of protozoa

Table 1. Details of wastewater stabilization ponds design of Abadan industrial park

Depth Water Influent flow

Pond Width (m) Length (m) Area (m2) Width/Length
(m) depth(m) rate (m3/day)
Anaerobic 4 3.4 32 48 1536 1:1.5 400
Facultative 2.3 1.9 100 153 15300 1:1.53 800
Maturation 1.8 1.4 100 252 25200 1:2.52 800

C × 1000 3. Results and discussion

NO = (1)
A× D × F × f
Table 2 provides an important list of genus or
C: Number of organisms counted species of protozoan were identified during the study.
A: The area of each field view Moreover, some of the achievements and results of
D: Depth of field view this study are summarized in Tables 3 to 6. These
F: The number of field counted Tables (3 to 6) shows the general characteristics of
f: Dilution of the sample quality and statistical indicators during the review.

2.4. Data analysis
Data were analyzed by using SPSS v.18
software. For data interpretation, the mean and
standard deviation (mean ± SD) were used. Data
normality was evaluated by Shapiro-Wilk test and for
normal data the average. Comparing of input and
output variables paired T-test was used. The Pearson
correlation was used to determine the correlation
severity and linear regression models to determine the
relationship between variables studied in input and
output places. P-value of 0.05 was considered as
statistical significance for all tests.
3.1. Relationship between BOD and COD with
protozoa
The results of this study indicate that there is a
positive relationship between the number and
population percentage of amoeba and mastigophore
(flagellate) with BOD and COD whereas is negative
the relationship between sticking ciliates with BOD
and COD. Furthermore the results demonstrated that
there is a positive relation between DO, the
percentage, number of ciliates and amoeba population
and there is a correlation with flagellate population.

Table 2. The important genus or species of protozoa that were found in wastewater treatment plants

Category Genus or species

Ciliates Paramecium sp (180-300um)
Trachelophyllum pusillum (40-50um)
Tetrahymena pyriformis (40-80um)
Colpidium sp. (90-150um)
Euplotes sp.
Aspidisca sp.
Vorticella sp.
Opercularia
Carchesium polypinum
Epistylis sp.
Didinium (80-200um)
Mastigophore Bodo sp.
(flagellate) Euglena sp.
Peranema sp.
Amoeba Acanthamoeba sp.
Amoeba proteus
Amoeba sp.

Table 3. The quality and statistical indicators of the raw wastewater entering the anaerobic treatment pond in industrial park

Parameter Unit Mean SD Maximum Minimum

BOD mg/L 905 204.7 1266 541
COD mg/L 1711 300.5 2406 1090
DO mg/L 0.4 0.41 1.2 0
TKN mg/L 151.4 25.7 198.1 98.1
TSS mg/L 893 173.7 1299 628
pH - 8.3 0.32 8.8 7.7
oC
Temperature 26.2 3.8 32.5 19.4

1681
 Bay et al./Environmental Engineering and Management Journal 17 (2018), 7, 1679-1683

Table 4. The quality and statistical indicators of the final effluent in established pond of the industrial park of Abadan

Parameter Unit Mean SD Maximum Minimum Mean Efficiency* (%)

BOD mg/L 107 36.3 211 68 88
COD mg/L 232 58.3 370 128 86.4
DO mg/L 3.3 0.49 4.4 2.5 -
TKN mg/L 22.9 7.4 44 13.8 84.8
TSS mg/L 219 50.7 298 102 75.4
pH - 8.9 0.26 9.2 8.2 -
oC
Temperature 27.8 4.4 35.4 21.6 -
*Mean Efficiency=Mean of Influent (Table 3) – Mean of Final effluent (Table 4) / Mean of Influent ×100

Table 5. The quality and biological indicators of the effluent entering the maturation pond of industrial park of Abadan

Parameter Unit Mean SD Maximum Minimum

Mastigophora(flagellate) No/mL 18692 6250 32146 10846
Free Ciliates No/mL 5911 5904 18972 864
Attached Ciliates No/mL 1879 595 2916 808
Amoeba No/mL 728 647 2536 93

Table 6. The quality and biological indicators in final effluent of industrial park of Abadan

Parameter Unit Mean SD Maximum Minimum

Mastigophora(flagellate) No/mL 12082 4082 24752 6471
Free Ciliates No/mL 12597 7727 29714 1379
Attached Ciliates No/mL 3127 828 4136 1211
Amoeba No/mL 561 524.9 1946 64

In other words, with the increase in flagellate
population it can be expected that the output effluent
quality will not be satisfactory and with the increase in
number of attached ciliates it can be expected to
achieve satisfactory effluent quality. These findings
are consistent with Dubber reports on the terms of
ciliates populations.
According to Dubber’s study there is a direct
relationship between the reduced amount of BOD and
organic matter with the total population of protozoa
(Dubber et al.2011).
3.2. Relationship between DO concentration and
Protozoa
The results of this study show that there is a
direct relation between DO and the number and
percentage of Ciliates and Amoeba population and
also a correlation with flagellate population (P<0.05).
But the highest correlation is related to the association
of DO and attached (sticking) ciliates. The existence
of high dissolved oxygen can cause spreading in
attached ciliates population.
3.3. Relationship between Kjeldahl nitrogen and
protozoa
Since Kjeldahl nitrogen amounts in sequential
wastewater ponds is decreasing, comparing Kjeldahl
nitrogen amounts with the number and percentage of
ciliates and also the number of amoeba results in a
direct relationship. Whilst, total Kjldal nitrogen in
pond and the final effluent has a significant direct
correlation with the number of flagellate (P<0.05). In
other words, increasing of flagellate numbers is
indicative of high effluent Kjeldahl nitrogen. The
research findings were consistent with Puigagut and
colleagues on nitrogen amount and total population of
ciliates (Puigagut et al. 2005).
3.4. Relationship between TSS and Protozoa
The Relationship between incoming suspended
solids and protozoa population showed that there is a
significant relation between this parameter and the
percentage of amoebas population and the numbers of
ciliates (P<0.05) which both have direct and inverse
relationship. This means that by increasing input TSS,
the amoeba population percentage and free ciliates
numbers were reduced. Moreover the results of input
TSS measurements have a direct correlation with
flagellate population.
4. Conclusions
Changes in the abundance of biologic indicator
species of the microbial community can give advance
warning of impending problems in wastewater
treatment plant. Changes in operating conditions of
the wastewater treatment plant may cause rapid
changes in the structure of the communities of
microscopic organisms, and it is hoped to work
towards an understanding of how to regulate the
operating conditions to maintain an ideal balanced
community of protozoa.
The type and abundance of protozoa in
industrial wastewater treatment working on biological
methods are the indicator of system performance and
effluent quality. Among protozoa, the presence of
certain variety of ciliates is considered highly
important compared to other species. This study
shows increase in the ciliates population reduced

1682
 Survey of industrial wastewater treatment performance with biological monitoring of protozoa
concentrations of BOD, COD, TSS and nitrogen in
output wastewater effluent and leads to satisfactory
quality of wastewater. Whereas the increasing in the
number of amoeba and flagellates indicates the
sewage treatment is incomplete. It is also
demonstrated that there is susceptibility between DO
concentration and flagellate population in which by
increasing dissolved oxygen, the number of flagellate
decreases.
Thus, it can be deduced from the findings of
this study that low cost biological indicators can be
used for monitoring and hence result in a better
economy especially in tropical area. Furthermore it is
also a good indicator for the detection of toxic
materials entering the plant or refinery and also its
organic load shocks. Especially in industrial
wastewater because of the cost of testing is expensive,
also in developing countries and technical problems of
the analysis, the use of bio-indicators for operating
plant can be very helpful.
Acknowledgments
Special thanks to the head of environmental health research
center and company officials of industrial park in Khuzestan
province as well as Dr. Bijan Bina and Dr. Jafarzadeh for
their efforts in providing research facilities and their
guidance.
References
APHA., AWWA., WEF., (2005), Standard Methods for the
Examination of Water and Waste Waters, 21st Edition,
USA, American Public Health Association.
Chen Q., Xu R., Tam N., Cheung S., Shin P., (2008), Use of
ciliates (protozoa: Ciliophora) as bioindicator to assess
sediment quality of two constructed mangrove sewage
treatment belts in Southern China, Marine Pollution
Bulletin, 57, 689–694.
Dubber D., Gray N.F., (2011), The effect of anoxia and
anaerobia on ciliate community in biological nutrient
removal systems using laboratory-scale sequencing
batch reactors (SBRs), Water Research, 45, 2213-2226.
Eslami A., Moradi M., Ghanbari F., Vatanpour V., (2016),
Electrochemical catalytic degradation and
biodegradability enhancement of real textile
wastewater by anodic and cathodic reactions,
Environmental Engineering and Management Journal,
15, 2319-2326.
Foley J., Haas D., Hartley K., Lant P., (2010),
Comprehensive life cycle inventories of alternative
wastewater treatment systems, Water Research,44,
1654-166.
Luna-Pabello V.M., Mayen R., Olevera-Viascan V.,
Saavedra J., Duran de Bazua C., (1990), Ciliated
Protozoa as Indicators of a Wastewater Treatment
System Performance, Biological Wastes,32 , 81-90.
Madoni P., (1994), Estimated of ciliated protozoa biomass
in activated sludge and biofilm, Bioresource
Technology,48, 245-249.
Papadimitriou C.A., Papatheodoulou A., Takavakoglou V.,
Zdragas A., Samaras P., Sakellaropoulos G.P.,
Lazaridou M., Zalidis G., (2010), Investigation of
protozoa as indicators of wastewater treatment
efficiency in constructed wetlands, Desalination,250,
378-382.
Perez-Uz B., Arregui L., Calvo P., Salvado H., Fernandez
N., Rodrıguez E., Zornoza A., Serrano S., (2010),
Assessment of plausible bio indicators for plant
performance in advanced wastewater treatment
systems, Water Research,44, 5059-5069.
Popescu M., Koncsag C.I., Birladeanu A.I., (2009),
Monitoring of PAHs in refinery effluents,
Environmental Engineering and Management Journal,
8, 67-71.
Poole J.E.P., (1984). A study of the relationship between the
mixed liquor fauna and plant performance for a variety
of activated sludge sewage treatment works, Water
Research,18, 281-287.
Puigagut J., Salvado H., Garcia J., (2005), Short-term
harmful effects of ammonia nitrogen on activated
sludge microfauna, Water Research,39, 4397-4404.
Salvado H., Gracia M.P., (1993), Determination of organic
loading rate of activated sludge plants based on
protozoan analysis, Water Research, 27, 891-895.
Zhou K., Xu M., Dai J., Cao H., (2006), The microfauna
communities and operational monitoring of an
activated sludge plant in China, European Journal of
Protistology, 42, 291-295.
1683
1684
Environmental Engineering and Management Journal July 2018, Vol.17, No. 7, 1685-1692
http://www.eemj.icpm.tuiasi.ro/; http://www.eemj.eu

“Gheorghe Asachi” Technical University of Iasi, Romania

APPLICATION OF RESPONSE SURFACE METHODOLOGY (RSM)

FOR OPTIMIZATION OF ZINC EXTRACTION
FROM ANAEROBIC SEWAGE SLUDGE

Vladimír Frišták1∗, Martin Pipíška1,2, Michaela Valovčiaková2, Juraj Lesný2

1Trnava University in Trnava, Chemistry, Department of Chemistry, Trnava, SK-918 43, Slovak Republic
2University of SS. Cyril and Methodius in Trnava, Faculty of Natural Sciences, Department of Ecochemistry and Radioecology,
Nám. J. Herdu 2, Trnava, SK-917 01, Slovak Republic

Abstract

The reuse of sewage sludge will be increasing in the future, with expansion of sewerage works and advanced sewage treatments.
During anaerobic digestion of sewage sludge from wastewater treatment plants the humification of organic matter will have an
important effect on the physico-chemical forms of heavy metals. The aim of our paper was to investigate mobility of zinc from
dried anaerobic sludge (DANS) from municipal wastewater treatment plant (WWTP) via single step (simultaneous) extraction
protocols. We tried to optimize extraction process as a function of amount of sludge biomass, concentration of extracting agent and
reaction time. The three parameters namely 0.55 g of anaerobic sludge, 50.5 mmol/dm3 HCl as extracting agent and 3 hour of
reaction time were chosen as center points. The experimental data on extractable Zn concentration (cZn) from DANS were obtained
by stripping chronopotentiometry and fitted into a quadratic polynomial model using multiple regression analysis. The optimal
extraction parameters were studied using experimental Box-Behnken design under response surface methodology (RSM). Results
showed that interaction effect of reaction time and amount of biomass to Zn extraction is minimal, but in combination with the
concentration of HCl significant increase of extractable zinc from DANS was detected. The maximum metal solubilization was
obtained at 0.55 g of DANS, 100 mmol/dm3 HCl and 5 hour of reaction time. Correlation among to predicted and experimental
data showed ability to reducing the number of total experiments from 39 to 17.

Key words: DANS, extraction, RSM, stripping chronopotentiometry, zinc

Received: August, 2013; Revised final: August, 2014; Accepted: September, 2014; Published in final edited form: July 2018

1. Introduction utilization is mainly due to pathogens, persistent

The production of sewage sludge from
wastewater treatment plants have been increased due
to rapid industrialization and urbanization.
Unsettleable form and high water content can be one
of the problems of sewage sludge disposal. Sewage
sludge can be stabilized by anaerobic digestion and
dewateration before its deposition in landfills. The
feasibility of adequate sites and the risk of
contamination of superficial and subterranean waters
are some of the drawbacks of sludge disposal on land
(Marchioretto et al., 2002). The risk of the sludge
organic pollutants and heavy metals content. Metal
contaminants are particularly problematic because,
unlike most organic contaminants, they are non-
biodegradable and can be bioaccumulated in living
organisms, thus becoming concentrated throughout
the food chain (Shrivastava et al., 2003). Metals
concentrations in sewage sludge vary from one site to
another, depending on municipal and industrial input
into system. Binding of metals into organic matrix,
organic mineral aggregates and inorganic particles and
to carbonates or manganese oxides affect the
reversibility of bound and leaching of toxic

∗
Author to whom all correspondence should be addressed: e-mail: fristak.vladimir.jr@gmail.com; Phone: +421335921459; Fax: +421335921403
 Frišták et al./Environmental Engineering and Management Journal 17 (2018), 7, 1685-1692
component from sewage sludge into environment.
Sludge as an organic matter rich material with low
toxic metals concentration can be utilized as fertilizer
and soil conditioner in agriculture (Frišták et al.,
2013). The benefit of biosolids application to land is
to supply and recycle organic matter and nutrients
(Martinez et al., 2003). The ecological and
toxicological effect of potentially toxic metals from
sewage sludge on agro-ecosystem is directly related to
their partitioning between mobile and complexed
species in the highly heterogeneous soil system
(Merrington et al., 1997). Many factors influence this
partitioning of toxic metals in soil amended with
sludge including the soil and sludge characteristics.
The extraction of toxic metals from the sewage
sludge, before land application, is therefore a
necessary step for reducing the adverse effects of
biosolids application (McBride, 2003). For extraction
of heavy metals from sewage sludge a variety of single
and sequential extraction techniques can be used. In
recent years there has been a great interest for the
methods of toxic metals removing from sewage sludge
(Babel and Dacera, 2006; Frišták et al., 2012a; 2012b;
Marchioretto et al., 2002). These methods can be
clasified as a: chemical extraction, using inorganic
acids (Marchioretto et al., 2002), organic acids
(Veeken and Hamelers, 1999), alkaline reagents
(Hsiau and Lo, 1997), chelators (Lo and Chen, 1990),
bioloeaching (Pathak et al., 2009), electroreclamation
(Wang et al., 2005) and advanced biosolids treatment
methods (Dewil et al., 2006). Chemical extractions as
a tool for Zn removal from sewage sludge were
extensively descripted in work Shiro and Tahei
(2000). Zhang and Zhang (2012) showed the
application of [S, S]-EDDS as extracting agent for Zn
extraction from sewage sludge. Zn extraction by
organic acids provides some disadvantages (Babel and
Dacera, 2006). Our previous paper (Frišták et al.,
2012b) found out, that inorganic extractants such as
hydrochloric or nitric acid solutions remove the Zn
from sludge structures with highest efficiency. For
differences in biosolids structures, extracting
parameters and using only limited number of chemical
extractants, it is difficult to perform comparison
among experimental data and find out the optimal
conditions of extraction protocols.
In the paper, we tried to optimize chemical
extraction efficiency of zinc from dried anaerobic
sludge (DANS) of municipal wastewater treatment
plant (WWTP) via single step (simultaneous)
extraction protocols using stripping
chronopotenciometry. The optimal extraction
parameters were studied using experimental Box-
Behnken design under response surface methodology
(RSM). RSM is a collection of mathematical and
statistical techniques useful for analyzing the effects
of several independent variables on the response (Box
and Draper, 1987). The methodology is more practical
as it arises from experimental methodology which
includes interactive effects of the variables and,
eventually, it depicts the overall effects of the
parameters on the process (Bas and Boyaci, 2007).
1686
The main aim of paper was to describe the
effect of amount of sludge biomass, concentration of
extracting agent and reaction time on zinc extraction
from DANS. Therefore, three level three factor (33)
quadratic polynomial model under Box-Behnken
design (BBD) was developed to predict the
relationship between the experimental variables on the
total of zinc extracted (response variable) from
DANS. Obtained results are useful for application of
anaerobic sludge as a soil amendment and conditioner.
2. Material and methods
2.1. Preparation of sludge sample
Municipal anaerobic sludge was obtained from
mechanical-biological WWTP in Zeleneč (Slovak
Republic). After repeated washing in deionised water,
sludge biomass was oven dried at 105°C for 48 h,
ground, sieved and stored in closed bottles at 22°C.
For extraction experiments fraction < 2 mm was used.
2.2. Extraction experiments
Leachable and bioavailable fractions of zinc
from dried anaerobic sludge (DANS) were obtained
after one-step extractions with HCl of variable
concentrations. HCl as extraction agent with highest
extraction efficiency for Zn and anaerobic sludge was
used (Frišták et al. 2012b).
Extraction experiments were carried out by
suspending of DANS (10 g/dm3 d.w.) in 10 cm3 of
extraction agent. Flasks were agitated on a reciprocal
shaker (150 rpm) at 22°C. After centrifugation (5 min,
5000 rpm) the supernatant was obtained and
concentration of zinc was determined by
electrochemical analysator. All experiments were
carried out in triplicates. Amounts of DANS,
concentrations of HCl and reaction time were set
according to experimental design methodology.
2.3. Experimental design and optimization of zinc
extraction
Variable number of factors such as reaction
time, amount of DANS and concentration of HCl as
extracting agent can significantly affect the total zinc
extraction efficiency. Therefore, Box-Behnken design
(BBD) under RSM was used to identify the
relationship between the response variable (total
extracted zinc concentration) and the experimental
variable (amount of sludge biomass, concentration of
extracting agent and reaction time).
The statistical software Statgraphics Centurion
XV (StatPoint Inc., USA) was used for the
experimental design, data analysis, quadratic model
building, graph plotting (three-dimensional response
surface and contour plots) and to estimate the
responses of amount of sludge biomass, concentration
of extracting agent and reaction time. The range of
experimental variables and their units and notation are
indicated in Table 1.
 Application of response surface methodology (RSM) for optimization of zinc extraction from anaerobic sewage sludge

Table 1. Experimental range and levels of independent variables for Zn extraction

Coded levels
Independent variables Symbols Unit
-1 0 +1
mDANS A mg 100 50 1000
cHCl B mmol/dm3 1 50.5 100
T C h 1 3 5

The response variable cZn (total extractable Zn 3. Results and discussion

concentration) can be expressed as a function of the
independent experimental variables according to the 3.1 Characterization of soluble and solid fraction of
following RSM guadratic model (Eq. 1): DANS

k k k Physico-chemical characteristics of DANS are

c Zn = β 0 + ∑ β i xi + ∑ β ij xi + ∑β
2
ij
i =1 i =1 i ≤i ≤ j
where: cZn is predicted response (total extractable Zn
concentration); βi, βii, βij - linear, quadratic and
interaction effects; β0 - the constant coefficient; xi, xj,
…xk - the input variables (mDANS; cHCl; t) which affect
the cZn value; ε - the error.
In the optimization procedure we studied the
response of the statistically designed combinations,
estimated the coefficients by fitting the experimental
data to the response functions and predicted the
response of the fitted model. The adequacy of the
model and statistical significance of the regression
coefficients were checked using the analysis of
variance (ANOVA).
2.4. Galvanostatic stripping chronopotentiometry and
pretreatment of samples
Extractable zinc was measured by
galvanostatic stripping chronopotentiometry on
macroporous electrode E104 L using electrochemical
analyzer EcaFlow model GLP 150 (Istran, Ltd.,
Bratislava, Slovak Republic). Aliquots (5 ml of
extracts) were added to flasks with 20 ml 0.1 mol/dm3
HCl and 0.5 ml 0.01 mol/dm3 KMnO4. Solution was
heated to 80-95°C for 5-10 min. For reducing the
excess of permanganate oxalic acid was added. The
volume of samples was adjusted to 50 ml with 0.1
mol/dm3 HCl.
2.5. Determination of pseudo-total zinc concentration
Aqua regia extractable zinc fraction as an
adequate fraction for analyzing total-recoverable zinc
was determined by extraction in strong acid conditions
and analyzed by atomic absorption spectrometry with
electrothermal atomization (ETAAS) device
(Shimadzu AA-6300, USA) with electrothermal
atomizer (Shimadzu GFA-EX7i) using an automatic
dispenser (Shimadzu ASC 6100) and background
correction method of Smith-Hieftje. Samples were
microwave digested by the Multiwave system MW
3000 (Anton Paar GmbH, Austria)
xi x j + ε
presented in Table 2. On the basis of experimental data
(1) we determined cation exchange capacity (CEC) of 25
meq/100g. CEC is commonly used to obtain the soil
sorption capacity for heavy metals. Content of the
soluble fraction is important for utilization of sludge
as soil conditioner. We found, that after one step
extraction by 0.9 % NaCl at 22°C, DANS had well
sedimentation properties.
Table 2. Chemical and physico-chemical characteristics
of DANS studied
DANS characterization
Soluble fractiona Solid fractionb
Proteins1 Cobalt5
9.6 ± 0.3 86 ± 3
[mg/g] [µg/kg]
Carbohydrates2 Manganese5
5.1 ± 0.2 163 ± 5
[mg/g] [µg/kg]
Zinc5
Tiols3 [µmol/g] 6.1 ± 0.1
[mg/kg]
649 ± 12
Soluble
CEC6
inorganic salts4 <LOD 25
3 [meq/100g]
[mg/dm ]
a
extracted by 0.9 % NaCl at 22 °C; 1determination of soluble
proteins (Lowry et al., 1951); 2determination of carbohydrates by
anthrone reagent (Dreywood, 1946); 3 determination of thiols by
Ellman’s reagent (5,5-dithiobis-(2-nitrobenzoic acid) (Ellman,
1959), 4 determination of soluble inorganic salts by IC (Values of
LOD for anions used (µg/dm3) : fluorides 6, chlorides 7, nitrites 20,
bromides 29, sulphates 25, nitrates 24, phosphates 7);
5
concentration of cobalt, manganese and zinc after aqua regia
digestion (ETAAS), 6cation exchange capacity
Concentration of mobile metals in the sewage
sludge is important for assessment its hazardous
character or using as a soil conditioner and fertilizer
(Frišták et al, 2012a). Pseudototal zinc concentration
in DANS as an aqua regia extractable zinc was 649 ±
12 mg/kg (d.w.). The comparison of obtained
extractable fractions of other heavy metals is shown in
Table 2. Meers et al (2007) showed that used digestion
process can be considered adequate for analysing
total-recoverable heavy metals in soils. We suppose
that residual zinc that is not released by aqua regia
digestion is bound to silicate minerals and is
considered unimportant for estimating the mobility of
metals. Similar results were obtained by Niskavaara et
al. (1997).

1687
 Frišták et al./Environmental Engineering and Management Journal 17 (2018), 7, 1685-1692

3.2. Optimization of zinc extraction protocol – RSM
application
Generally, sewage sludge as an organic matter-
rich material contains larger concentration of heavy
metals than the background values found in soil.
Application of sludge as soil conditioner can gradually
raise the total toxic metal concentration of soil (Smith,
et al., 2009). Although, metals present in compost
exist in relatively resistant forms, these are more labile
in the environment of soil matrix. Zinc is the element
which the largest increase in labile forms is reported
when sewage sludge is applied to soil, particularly
under acidic soil conditions (Planquart et al., 1999).
Eneji et al. (2003) found that Zn availability increased
during anaerobic digestion of organic wastes. One of
the main approaches to study the behaviour of toxic
metals in sewage sludge has been to use chemical
extraction protocol to remove metals bound in
different defined phases (He et al., 1992). Parameters
of extraction protocols are variable and it is difficult to
find out the optimal conditions. For that reason, a Box-
Behnken design under RSM was used to analyze
interactive effect of the amount of sludge biomass,
concentration of extracting agent and reaction time on
total extractable zinc concentration from DANS. The
advantage of the RSM method is the minimization of
the number of experiments and time needed. A total of
17 runs were used to optimize the extraction process
parameters (Table 2). Each of experiment was
performed in triplicate.
Using multiple regression analysis, the second
order polynomial model which characterizes the
relationship between total extractable zinc
concentration cZn from DANS and actual studied
variables can be written as given by (Eq. 2). The
analysis of variance (Table 3), which is a statistical
technique that subdivides the total variation in a set of
data into components associated with specific sources
of variation to test hypotheses on the parameters of the
model, suggests the statistical significance of the
model equation, as it is evident from the Fisher`s F-
value (86.19) with a low probability value (p-value
<0.0001).
c Zn = +5.49265 + 4.06566.10 −3 × m DANS − 0.13619 ×
c HCl − 2.42866 × t + 1.01010.10 − 4 ×
(2)
m DANS × 4 + 8.33333.10 − 4 × m DANS × t + 0.032828 ×
c HCl × t − 5.55556.10 −6 × m DANS 2 +
1.17335.10 −3 × c HCl 2 + 0.21875 × t 2
A large F-value indicates that most of the
variation can be explained by a regression equation
whereas a low p-value indicates that the model is
considered to be statistically significant (Myers and
Montgomery, 2002). The goodness of fit was also
evaluated by the correlation coefficient (R2) between
the model predicted and experimental values.
Coefficient R2 indicates the statistically signification
of model used. Both values R2 and adjusted R2 are
closed to 1 and indicate a high correlation between the
observed and the predicted values (Garg et al., 2008).
On the basis of this investigation, the
relationship between the independent variables
(amount of sludge biomass, concentration of
extracting agent and reaction time) and the response
(total extractable zinc concentration from DANS) can
be explained according to the regression model. cZn is
attributed to the independent variables and only about
1.90 % of the total variation cannot be explained by
the model. Regression analysis (Table 3) shows that
the square as well as the interaction effects of the
independent experimental variables were significant
(p<0.0001). The adequacy of quadratic model was
tested also through the correlation between calculated
and experimental values for total extractable zinc
concentration from DANS, which is shown in
scattered plot (Fig. 1).

Table 3. BBD matrix, the experimental and predicted values of total extractable Zn concentration cZn from DANS for three
independent variables (A: mDANS (mg); B: cHCl (mmol/dm3); C: t (h))

Coded levels Y (cZn)

Run
A B C cZn (exp.) (mg/kg) cZn (pred.) (mg/kg)
1 -1 0 +1 5.05 3.25
2 +1 +1 0 19.10 19.37
3 0 +1 -1 12.00 12.00
4 0 0 0 6.00 6.00
5 0 0 0 6.00 6.00
6 +1 0 +1 12.05 12.00
7 0 0 0 6.00 6.00
8 +1 -1 0 1.00 1.12
9 0 -1 -1 5.02 3.12
10 0 +1 +1 21.00 21.87
11 -1 0 -1 1.05 1.00
12 +1 0 -1 5.00 6.75
13 -1 +1 0 10.13 9.87
14 -1 -1 0 1.00 2.87
15 0 0 0 6.08 6.00
16 0 0 0 6.08 6.00
17 0 -1 +1 1.00 0.87

1688
 Application of response surface methodology (RSM) for optimization of zinc extraction from anaerobic sewage sludge
Table 4. Analysis of variance (ANOVA) for Zn extraction from DANS
Source Sum of squares
Model 554.06
A 50.00
B 364.50
C 32.00
AB 20.25
AC 2.25
BC 42.25
A2 5.33
B2 34.80
C2 3.22
Lack of fit 5.00
Pure Error 0.000
Total 559.06
R2 0.9754
From correlation between predicted cZn pred and
experimental cZn exp values of zinc extraction from
DANS is evident that the regression model can
represent the experimental extraction data well. On the
basis of the evaluation of ANOVA outputs, the
statistical significance of a quadratic model for the
response was confirmed and it can be concluded that
the model can be used for further analysis of effect of
process variables (Frišták et al., 2012a; Remenárová
et al., 2012).
Fig. 1. Correlation between predicted cpred and
experimental cexp values of zinc extraction from DANS
As was already mentioned above, the aim of
this paper was to optimize the parameters of zinc
extraction protocol, namely: the amount of sludge
biomass, concentration of extracting agent and
reaction time. The statistical significance of each
coefficient was determined by p-values shown in
Table 3. Lower p-value indicates a higher significance
of the parameter.
Our results demonstrate the statistical
significance of linear quadratic coefficients: A (F =
70.00, p <0.0001), B (F = 510.30, p <0.0001) and C
(F = 44.80, p = 0.0003), and quadratic coefficients A2
(F = 7.46, p = 0.0293), B2 (F = 48.72, p = 0.0002).
Other quadratic form C2 (F = 0.37, p = 0.0713), as well
as interaction coefficient AC (F = 3.15, p = 0.1192)
p-value
DF F value
Prob>F
9 86.19 <0.0001
1 70.00 <0.0001
1 510.30 <0.0001
1 44.80 0.0003
1 28.35 0.0011
1 3.15 0.1192
1 59.15 0.0001
1 7.46 0.0293
1 48.72 0.0002
1 4.51 0.0713
3 0.00
4 0.00
16
R2adj 0.98
not confirmed the statistical significance. For
interpretation of the interaction coefficients can be
used three-dimensional plots of regression model.
Such three-dimensional surfaces can provide useful
information about the behaviour of the extraction
system within the experimental design, facilitate an
examination of the effects of the experimental factors
on the responses and contour plots between the factors
(Ahmad and Hameed, 2010).
Fig. 2 shows that the extraction process of zinc
is strongly influenced by the amount of biomass and
concentration of exracting agent. Effect of sludge
amount in combination with reaction time was less
significant. According to identified relations and
predicted values we found out that the extraction
process of zinc from DANS has highest efficiency at
following parameters: the amount of sludge biomass
550 mg, the concentration of extracting agent (HCl)
100 mmol/dm3 and contact time 5 h. Correlation
among to predicted and experimental data showed
ability to reducing the number of total experiments
from 39 to 17. Mannan et al. (2007) showed the
usefulness of RSM methods and Box-Behnken design
for the optimization of process parameters of
bioconversion of activated sludge by Penicillium
corylophilum.
Response surface method (RSM) appears to be
a better tool for the evaluation of difficult studies of
multi-component systems than the simple
extrapolation from single-component systems in
sorption process of heavy metals by dried activated
sludge (Remenárová et al., 2012). Zhang and Zhang
(2012) showed the application of [S,S]-EDDS as
extraction agent of Zn from mixed sludge with
effectivity more than 65%. Authors described equally
the possibility of Zn extraction from this type of
sewage sludge by EDTA with high efficiency.
However extraction of Zn from sewage sludge
requires optimization of reaction conditions (Babel
and Dacera, 2006). Our paper showed the RSM as
useful statistical tool for optimization of Zn extraction
process by HCl with comparable removing effect such
as extraction by H3PO4 published in paper Zhang and
Zhang (2012).
1689
 Frišták et al./Environmental Engineering and Management Journal 17 (2018), 7, 1685-1692

Fig. 2 Three-dimensional surface plots of the combined effect of (A) reaction time (t) and amount of DANS (m); (B)
concentration of extracting agent (cZn) and amount of DANS (m); (C) reaction time (t) and concentration of extracting agent (cHCl)
on total extractable zinc concentration (cZn) from DANS (VHCl = 10 cm3, 22°C, 150 rpm)

1690
 Application of response surface methodology (RSM) for optimization of zinc extraction from anaerobic sewage sludge
4. Conclusions
The present investigation was carried out to
optimize extraction process as a function of amount of
DANS biomass, concentration of extracting agent and
reaction time using a Box-Behnken design under the
Response surface methodology (RSM).
The results demonstrate that interaction effect
of reaction time and amount of biomass to Zn
extraction is minimal, but in combination with the
concentration of HCl significant increase of
extractable zinc from DANS was detected. ANOVA
analysis as well as 3D surface plots revealed that the
maximum metal solubilization was obtained at 0.55 g
of DANS, 100 mmol/dm3 HCl and 5 hour of reaction
time. On the basis of results it can be concluded that
RSM present an excellent tool in prediction of optimal
parameters of extraction protocols and reduces the
number of needed experiments from 39 to 17.
References
Ahmad A.A, Hameed, B.H., (2010), Effect of preparation
condition of activated carbon from bamboo waste for
real textile wastewater, Journal of Hazardous
Materials, 17, 487-493.
Babel S., Dacera D.D.M., (2006), Heavy metal removal
from contaminated sludge for land application, Waste
Management, 26, 988-1004.
Bas D., Boyaci I.H, (2007), Modeling and optimization I:
usability of response surface methodology, Journal of
Food Engineering, 78, 836-845.
Box G.E.P., Draper N.R., (1987), Empirical Model-Building
and Response Surfaces, John Wiley and Sons, New
York.
Dewil R., Baeyens J., Neyens E., (2006), Reducing the
heavy metal content of sewage sludge by advanced
sludge treatment methods, Environmental Engineering
Science, 23, 994-999.
Dreywood R., (1946), Qualitative test for carbohydrate
material, Industrial and Engineering Chemistry
Analytical Edition, 18, 499-504.
Ellman G.L., (1959), Tissue sulfhydryl groups, Archives of
Biochemistry and Biophysics, 82, 70-77.
Eneji A.E., Honna T., Yamamoto S., Masuda T., Endo T.,
Irshad M., (2003), The relationship between total and
available heavy metals in composted manure, Journal
of Sustainable Agriculture, 23,125-134.
Frišták V., Remenárová L., Lesný J., (2012a), Response
surface methodology as optimization tool of
competitive effect of Ca2+ and Mg2+ ions in sorption
process of Co2+ ions by dried activated sludge, Journal
of Microbiology, Biotechnology and Food Sciences, 5,
1235-1249.
Frišták V., Valovčiaková M., Pipíška M., Augustín J.,
(2012b), Simultaneous and sequential extraction
protocols as tools for determination of zinc
bioavailability in dried anaerobic sludge, Nova
Biotechnologica et Chimica, 11, 167-175.
Frišták V., Pipíška M., Horník M., Augustín J., Lesný J.,
(2013), Sludge of wastewater treatment plants as Co2+
ions sorbent, Chemical Papers, 67, 265-273.
Garg U.K., Kaur M.P., Gark V.K., Sud D., (2008), Removal
of nickel (II) from aqueous solution by adsorption on
agricultural waste biomass using a response surface
methodological approach, Bioresource Technology, 99,
1325-1331.
He X.T, Traina S.J., Logan T.J., (1992), Chemical properties
of municipal solid waste composts, Journal of
Environmental Quality, 21, 318-329.
Hsiau P.C., Lo S.L., (1997), Characteristic of four alkaline
biosolids produced from sewage sludge, Resources
Conservation and Recycling, 21, 185-197.
Lo K.S.L., Chen Y.H., (1990), Extracting heavy metals from
municipal and industrial sludges, Science of the Total
Environment, 90, 99-116.
Lowry O.H, Rosebrough N.J., Farr A.L., Randall R.J.,
(1951), Protein measurement with the folin phenol
reagent, Journal of Biological Chemistry, 193, 266-
275.
Mannan S., Razi F.A., Alam M.Z., (2007), Optimization of
process parameters for bioconversion of activated
sludge by Penicillium corylophilum, using response
surface methodology, Journal of Environmental
Sciences, 19, 23-28.
Marchioretto M.M., Bruning H., Loan N.T.P., Rulkens
W.H., (2002), Heavy metals extraction from
anaerobically digested sludge, Water Science and
Technology, 46, 1-8.
Martinez F., Cuevas G., Calvo R., Walter I., (2003),
Ecosystem restoration. Biowaste effects on soil and
native plants in a semiarid ecosystem, Journal of
Environmental Quality, 32, 472-479.
McBride M.B., (2003), Toxic metals in sewage sludge-
amended soils: has promotion of beneficial use
discounted the risks? Advances in Environmental
Research, 8, 5-19.
Meers E., Laing D.D., Unamuno V., Ruttens A.,
Vangronsveld J., Tack F.M.G., Verloo M.G., (2007),
Comparison of cadmium extractability from soils by
commonly used single extraction protocols, Geoderma,
141, 247-259.
Merrington G., Winder L., Green I., (1997), The
bioavailability of Cd and Zn from soils amended with
sewage sludge to winter wheat and subsequently to the
grain aphid Sitobion avenae, Science of Total
Environment, 205, 245-254.
Myers R.H, Montgomery, D.C., (2002), Response Surface
Methodology: Process and Product Optimization Using
Designed Experiments, John Wiley and Sons, New
York.
Niskavaara H., Reimann C., Chekushin V., Kashulina G.,
(1997), Seasonal variability of total and easily
leachable element contents in topsoils (0-5 cm) from
eight catchments in the European Arctic (Finland,
Norway and Russia), Environmental Pollution, 96, 261-
274.
Pathak A., Dastidar M.G., Sreekrishnan T.R., (2009),
Bioleaching of heavy metals from sewage sludge by
indigenous iron-oxidizing microorganisms using
ammonium ferrous sulfate and ferrous sulfate as energy
sources: a comparative study, Journal of Hazardous
Materials, 171, 273-278.
Planquart P., Bonin G., Prone A., Massiani C., (1999),
Distribution, movement and plant availability of trace
metals in soils amended with sewage sludge composts:
application to low metal loadings, Science of The Total
Environment, 241, 161-179.
Remenárová L., Pipíška M., Horník M., Rozložník M.,
Augustín J., Lesný J., (2012), Biosorption of cadmium
and zinc by activated sludge from single and binary
solutions: Mechanisms, equilibrium and experimental
1691
 Frišták et al./Environmental Engineering and Management Journal 17 (2018), 7, 1685-1692
design study, Journal of the Taiwan Institute of
Chemical Engineers, 43, 433-443.
Shiro Y., Tahei T., (2000), Principle and process of heavy
metal removal from sewage sludge, Environmental
Science &Technology, 25, 113-118.
Shrivastava R., Upreti R.K., Chaturvedi U.C., (2003),
Various cells of the immune system and intestine differ
in their capacity to reduce hexavalent chromium, FEMS
Immunology &Medical Microbiology, 38, 65-70.
Smith K.M., Fowler D.G., Pullket S., Graham N.J.D.,
(2009), Sewage sludge-based adsorbents: A review of
their production, properties and use in water treatment
applications, Water Research, 43, 2569-2594.
1692
Veeken A.H.M., Hamelers H.V.M., (1999), Removal of
heavy metals from sewage sludge by extraction with
organic acids, Water Science & Technology, 40, 129-
136.
Wang C., Hu X., Chen M.L., Wu Y.H., (2005), Total
concentration and fractions of Cd, Cr, Pb, Cu, Ni and
Zn in sewage sludge from municipal and industrial
wastewater treatment plants, Journal of Hazardous
Materials, 119, 245-249.
Zhang L., Zhang H., (2012), Zinc extraction from sewage
sludge with [S,S]-EDDS, Journal of Central South
University, 19, 3546-3550.
Environmental Engineering and Management Journal July 2018, Vol.17, No. 7, 1693-1700
http://www.eemj.icpm.tuiasi.ro/; http://www.eemj.eu

“Gheorghe Asachi” Technical University of Iasi, Romania

ANALYSIS OF THE CONTINUOUS MEASUREMENTS OF PM10

AND PM2.5 CONCENTRATIONS IN BEIRUT, LEBANON

Wehbeh Farah1∗, Myriam Mrad Nakhlé2,5, Maher Abboud3, Nelly Ziade4,

Isabella Annesi-Maesano5, Rita Zaarour6, Nada Saliba6, Georges Germanos1,
Najat Aoun Saliba7, Alan L. Shihadeh8, Jocelyne Gerard6
1Physics Department, Saint Joseph University of Beirut, Beirut, Lebanon
2Biology Department, Saint Joseph University of Beirut, Beirut, Lebanon
3Chemistry Department, Saint Joseph University of Beirut, Beirut, Lebanon
4Rheumatology Department, Saint Joseph University of Beirut, Beirut, Lebanon
5EPAR, Institute Pierre Louis of Epidemiology and Public Health, UMR-S 1136 INSERM & UPMC Paris 6, Sorbonnes

Universities, Medical School Saint-Antoine, Paris, France

6Geography Department, Saint Joseph University of Beirut, Beirut, Lebanon
7Mechanical Engineering Department, American University of Beirut, Beirut, Lebanon
8
Chemistry Department, American University of Beirut, Beirut, Lebanon

Abstract

Atmospheric concentrations of PM2.5 and PM10 were measured in Beirut, Lebanon, for a period of 12 months. The daily average
concentrations of PM10 and PM2.5 were found to be 51.3 ± 33.1 and 30.3 ± 9.4 µg.m−3, respectively, with corresponding maximum
values of 359.7 and 208.6 µg.m−3. The annual average concentrations of PM10 and PM2.5 exceeded the World Health Organization’s
standards by 150% and 200%, respectively. The mean concentration of coarse particles (PM10–2.5) was found to be 41% of the
average PM10, suggesting that the site was also influenced by re-suspended surface dust and soil. The mean PM2.5/PM10 ratio for
the entire study period was 0.61 ± 0.12. This indicates that in Beirut, PM2.5 accounts for about 61% of PM10. Such a large fraction
of fine particles could have considerable effect on health; thus, it is necessary to quantify its impact. Daily concentrations of PM10
and PM2.5 exceeded the upper threshold limit on 133 and 129 days, respectively, representing 39% and 38% of the entire sample,
respectively. These findings indicate the important role dust events play within this area. Concentrations of PM2.5 were highly
correlated with NO2, whereas concentrations of PM10 and PM10–2.5 were not associated with any gaseous pollutant. Regression
analysis showed that 93% of PM2.5 and 43% of PM10 particle mass concentrations were derived from road traffic exhaust in Beirut.

Key words: air quality, dust event, health effect, particulate matter, regression analysis

Received: October, 2013; Revised final: September, 2014; Accepted: September, 2014; Published in final edited form: July 2018

1. Introduction associated with traffic, wood fires used for heating,

Fine and ultrafine particles are the inhalable
fraction of atmospheric particulate matter (PM) that
can become embedded within the lungs and lower
respiratory system (Oberdorster et al., 2005).
Generally, ultrafine particles are related to soot
derived from the incomplete combustion of fuels
and incineration. The chemical compositions of these
particles define their biological reactivity. Various
organic and biological compounds can adhere to these
particles and it has been proven that these compounds,
which can be found in the lungs of humans, have
potentially harmful effects if they are not eliminated
by the clearance mechanism of the respiratory system.

∗
Author to whom all correspondence should be addressed: e-mail: wehbeh.farah@usj.edu.lb; Phone: +96 1142 1374; Fax: +96 1453 2657
 Farah et al./Environmental Engineering and Management Journal 17 (2018), 7, 1693-1700
Since the early 1990s, thousands of studies
have reported associations between airborne particles
and a range of respiratory and cardiovascular diseases
and even death. For example, Schwartz (1994)
reported an association between concentrations of
daily total suspended particles and the risk of mortality
in Ohio, whereas Dockery et al. (1993) associated
premature death in six eastern US cities with fine
atmospheric particulates (inhalable particles).
Similar results have been reported in
association with fine particles, but this time with
regard to hospital admissions. In Birmingham (UK), a
study in 1997 showed that following every 10 μg.m−3
increase in concentration of PM10, respiratory and
cerebrovascular hospital admissions increased by
2.4% and 2.1%, respectively (Wordley et al., 1997).
Another study revealed an association between the
short-term increase in hospital admission rates and
PM2.5 for all health problems except injuries
(Dominici et al., 2006). The largest association
regarded heart failure, which had a 1.28% (95%
confidence interval; 0.78–1.78) increase in risk per 10
μg.m−3 increase in same-day PM2.5 concentration.
Dominici et al. (2006) concluded that short-term
exposure to PM2.5 increased the risk of hospital
admission of people with cardiovascular and
respiratory diseases.
Other studies have also found relationships
between low levels of fine particles and hospital
admissions of people suffering respiratory symptoms.
In Vancouver, Canada, where pollutant levels are
relatively low, Chen et al. (2004) confirmed a positive
relationship between PM10, PM2.5, and PM10–2.5 and
emergency hospitalizations for Chronic Obstructive
Pulmonary Disease in people aged over 65 years.
Schwartz (1995) constructed daily counts of
Fig. 1. Locations of Lebanon on the Mediterranean Sea and of the HUV and CPF sites
1694
admissions to all hospitals in New Haven
(Connecticut) and Tacoma (Washington) for persons
aged 65 years and older with respiratory diseases
(ICD-9 460-519). The results showed significant
association between the three studied pollutants (SO2,
PM10, and O3) and admissions to hospitals of the older
persons with respiratory problems; however, the
strongest evidence of an independent association was
for PM10. These significant relationships are also
found for children, another group vulnerable to the
effects of air pollution and fine particles in particular
(Brauer et al., 2002; Braun-Fahrlander et al., 1997;
Penard-Morand et al., 2005).
These findings underline the importance of
ambient PM and for need to monitor PM10 and the
associated fraction of PM2.5. In this paper, we explore
the annual and daily average concentrations of PM10
and PM2.5 and compare them with the corresponding
WHO standards to estimate their effects on health. We
also determine the relative contributions of fine,
PM2.5, and PM10–2.5 to total inhalable particles (PM10),
and discuss the relationship between levels of PM and
concentrations of other gaseous pollutants.
2. Material and methods
2.1. Study area
Beirut, Lebanon’s capital, is located on the
eastern coast of the Mediterranean Sea (Fig. 1),
covering an area of 20.8 km2 and it has a population
of about 361,366 inhabitants (CAS, 2007). Beirut is
characterized by two hills that lie to the north and west
of the city, separated by a thalweg. Each hill rises
about 100 m above sea level and overlooks the
Mediterranean Sea.
 Analysis of the continuous measurements of PM10 and PM2.5 concentrations in Beirut, Lebanon
The typical unplanned urban morphology, high
population density, and high concentration of human
activities of the developing world, together with the
Mediterranean climate and topography of the capital,
are major factors that contribute to the degradation of
air quality. The main sources of air pollution are
vehicular traffic and dust storms, because industrial
activity does not exist within the study area. The
number of registered vehicles in Lebanon exceeds one
million, and the average age of the vehicle fleet is over
10 years (El-Zein et al., 2007). The government
exercises little effective control over the quality of
fuel, which is imported (El-Zein et al., 2007).
However, in 2002, a law was introduced banning the
use of diesel fuel for light- and medium-duty engines.
The main thermal power station of the country is
located 13 km northeast of Beirut, and the airport is
located 7 km to the southwest. Pollution data were
collected from only one station even though there are
several stations within Beirut for the measurement of
air quality. However, we chose the most suitable site,
which we monitored consistently throughout the study
period. Wordley et al. (1997) confirmed that
significant variation in pollutant levels could exist
between different stations, but that data obtained from
one station could be used as proxies for the levels of
air pollution required for time series studies.
2.2. Instrumentation and sampling procedure
The concentrations of PM10 and PM2.5 were
measured using an MP101M (Environnement SA,
2013) atmospheric monitoring system. The MP101M
is an ambient air-suspended particulate analyzer that
uses a beta gauge. A beta gauge consists of a Carbon-
14 radioactive source with very low activity, and a
Geiger–Müller tube receives the emitted beta
radioactive radiations. The detector is fixed below the
glass fiber filter tape and it collects air-suspended
particulates. Sampling is performed using a vacuum
pump attached to a sampling head that is connected to
the top of the analyzer. The measurement cycle can be
described as follows:
• Stage 1: The tape is blank. The radioactive
source is placed in front of the Geiger–Müller
detector. At this stage, a specified quantity of beta
radiation is absorbed by both the filter and the
compressed air located between the radioactive source
and the detector.
• Stage 2: The sample is absorbed by the
sampling head and collected on the filter tape. Beta
radiation emitted by the source is absorbed by the
filter, the compressed air, and the fixed particulates.
The particulate concentration is calculated
based on the difference between the radiation count at
the beginning of the cycle when the tape is blank, and
at the end of the cycle when particulates have adhered
to the tape. This difference in count is directly
proportional to the mass of particulates collected on
the tape. The duration of each cycle is 24 h, which is
divided into 12 equal periods. Once a cycle has
completed, the supply reels rotate and a new cycle
begins.
The MP101M has a built-in reference gauge for
calibration and the procedure of calibration is repeated
every 6 months. This equipment is manufactured
according to ISO 10473 standards, certified equivalent
to the US EPA-approved reference methods Nos.
EQPM-0404-151 and EN 12341.
Measurements of the trace gases were taken by
an atmospheric monitoring system for O3 and
NO/NO2 (Environnement SA, 2013), the instruments
of which were installed within a shelter in Beirut,
Lebanon. Ambient air was sampled through ¼-in
Teflon sampling tubing and all the recording
instruments were connected to an XR workstation data
logger, which was used to store their outputs.
The AC32M monitor for NO2 is based on
chemiluminescence (EN 14211) and the O342M
monitor is based on the absorption of ultraviolet
radiation (EN 14625). A gas dilutor (LNI Schmidlin
SA, Model SONIMIX 3012) was used to calibrate the
instruments using zero and standard gases. The zero
air generator (LNI Shmidlin SA, Model SONIMIX
3052) provided high-purity zero air, while the
standard gas consisted of 29 ppm NO balanced with
high-purity N2. The ozone analyzer calibration was
conducted using a standard ozone generator regulated
by ultraviolet light and integrated in the SONIMIX
dilutor.
The daily average concentrations of
atmospheric PM10, PM2.5, and other pollutants were
obtained between January and December 2012. In
addition to the continuous measurements of the
pollutant levels, measurements of wind direction and
speed, and temperature and humidity were routinely
monitored at this site to understand the influence of
meteorological parameters on pollutant levels. A
quality control procedure was established such that the
average calculations were accepted if 75% of the
quarter-hourly data were valid for any given hour and
75% of the hourly data valid for any given day.
Relationships between PM10, PM2.5, and PM10–
2.5 and other pollutant concentrations were
investigated. This was accomplished either by
estimating correlations between the different variables
(Pearson correlation coefficients) or by conducting
regression analysis.
3. Results and discussions
3.1. PM10, PM2.5, and PM10–2.5 concentrations
The concentrations of PM10 and PM2.5 were
measured on 343 days, out of the 365 days, from
January through December 2012. These
measurements produced 8232 data points for
assessment of the concentration of each pollutant.
Statistics for the daily concentrations of PM10, PM2.5,
and PM10–2.5 are shown in Table 1. The coarse particle
1695
 Farah et al./Environmental Engineering and Management Journal 17 (2018), 7, 1693-1700
mass concentration (PM10–2.5) was calculated based on
the difference between PM10 and PM2.5. The daily
average concentrations of PM10 and PM2.5 were 51.3 ±
33.1 and 30.3 ± 19.4 µg.m−3, respectively, with
corresponding maximum values of 359.7 and 208.6
µg.m−3. The mean concentration of coarse particles
(PM10–2.5) was found to be 41% of the average PM10,
suggesting that the site was influenced by re-
suspended surface dust and soil.
The annual average concentrations of PM10 and
PM2.5 exceeded the annual average WHO standards
(i.e., 20 µg.m−3 for PM10 and 10 µg.m−3 for PM2.5) by
150% and 200%, respectively, suggesting that urgent
intervention is required to control particulate pollution
in Beirut.
Table 1. Statistics for daily concentrations of PM10, PM2.5,
and PM10–2.5 from January to December 2012
N
Min Max
total
(µg/m3) (µg/m3)
(days)
PM10 343 16.0 359.7
PM2.5 343 4.5 208.6
PM10-2.5 343 1.0 157.7
Similar results to the present study have been
reported following comparable work performed in
other urban and coastal–urban areas throughout
Europe and Asia (Chaloulakou et al., 2003;
Gerasopoulos et al., 2007; Ho et al., 2003; Marcazzan
et al., 2003; Massoud et al., 2011; Pey et al., 2008;
Rodriguez et al., 2004). However, the mean
concentration of PM2.5 observed in December by
Meng and Lu (2007) is at least five to six times higher
than the values presented in this study.
This large difference is because the
concentrations of PM2.5 were measured during autumn
and winter, when the level of combustion of fossil fuel
is very high. In addition, adverse meteorological
conditions during these seasons can prevent
particulate dispersion, producing high concentrations
near the ground, which was not the case in Lebanon in
2012.
400
300
HUV-PM10 (µg.m-3)
200
100
0
0 100
Fig. 2. Comparison between PM10 concentration measurements at HUV and CPF sites
1696
3.2. Site representativeness
To assess the health effects more accurately, it
is important to examine whether data from a single
monitoring site are representative of community
exposure. Various studies on the spatial variation of
PM have shown that the concentration of particulates
is higher at locations affected by traffic emissions
compared with traffic-free areas. For example,
Buzorius et al. (1999) studied the spatial variations of
aerosol concentrations in Helsinki and found that
concentrations were higher at locations affected by
traffic emissions. Similar results have shown that
PM10 and PM2.5 concentrations measured near a busy
road were on average 1.3 times higher than at a
background site (Jansen et al., 1997).
Therefore, it is possible that the high
concentration levels reported here could be attributed,
at least in part, to the location of the monitoring site.
Mean To test this hypothesis, a limited number of PM10
(µg/m3) samples were collected at Beirut Saint Joseph
51.3 ± 33.1 University’s Huvelin area (HUV site). This site is
30.3 ± 19.4 located approximately 4 km from the CPF air-quality
21.0 ± 17.3 monitoring station. The CPF site is less affected by
traffic emissions and it is more representative of
exposure encountered within the Beirut Metropolitan
area. Fig. 2 compares the concentrations of PM10
observed at the HUV and CPF sites. Although this
comparison is based on a limited number of sampling
days (170 days), these results strongly support the
presumption that PM10 concentrations measured at
HUV were higher than at the CPF site; concentrations
of PM10 measured at the HUV site were on average
approximately 16% higher. Furthermore, PM10
concentrations measured at the HUV and CPF sites
exhibited correlation with an r2 value of 0.82.
3.3. Time series of PM concentrations
Fig. 3 shows the time series plot of daily
concentrations of PM10 and PM2.5 from January 2012
through December 2012.
HUV-PM10 vs CPF-PM10
Regression line
200 300 400
CPF-PM10 (µg.m-3)
 Analysis of the continuous measurements of PM10 and PM2.5 concentrations in Beirut, Lebanon

PM10
350
PM2.5

300

PM10 and PM2.5 (µg/m3) 250

200

150

100

50

0
1/1/2012

2/1/2012

3/1/2012

4/1/2012

5/1/2012

6/1/2012

7/1/2012

8/1/2012

9/1/2012

10/1/2012

11/1/2012

12/1/2012

1/1/2013
Fig. 3. Temporal trends of daily average PM10 and PM2.5 concentrations over the study period in Beirut.
Inset shows the correlation between PM10 and PM2.5 concentrations

This Figure shows that the mean values of both study periods. However, the average PM2.5/PM10 ratios
PM10 and PM2.5 reached peaks in February–July, when observed by Khodeir et al. (2012), Shahsavania et al.
wind speeds were also the highest, which is consistent (2012), and Saliba et al. (2010) are much lower than
with Draxler et al. (2001) and also identified by other those found in the present study.
investigators (Badarinatha et al., 2010; Querol et al., The primary reason for this difference is that,
2009). unlike our study, the pollutants in these other studies
Being an enclosed area, the Middle East is were not measured continuously. Thus, the
easily affected by dust storms from two sources: the PM2.5/PM10 ratios calculated for the entire year
Sahara desert in Africa and the Arabian Desert in Asia. include spring and summer. During these seasons, the
In Lebanon, dust events that carry great quantities of fine fraction of PM (i.e., PM2.5) is highest because
sand and dust from the deserts, usually occur in March various sources release fine particulates primarily
and April, but occasionally, they can occur into July. during these seasons; therefore, this produces an
The relationship between PM10 and PM2.5 increase in the PM2.5/PM10 ratio.
concentrations was analyzed in terms of linear
regression and the result revealed that the coefficient
of determination (r2) was relatively high (r2=0.83), as
shown in the inset of Fig. 3. This is expected, because
both fine and coarse particles are associated with local
traffic.

3.4. PM ratios

Fig. 4 shows the trend in the monthly average

of the PM2.5/PM10 ratio. This ratio shows the
variability of the data, which has a range of 0.45–0.7.
The monthly variability in the data suggests that the
contributions of fine and coarse particles are derived
from different sources. A gradual decrease can be
observed in the value of the PM2.5/PM10 ratio from
September to December, but the mean for the entire Fig. 4. Monthly PM2.5/PM10 ratios during the study period
study period is 0.61 ± 0.12. This average indicates that in Beirut
in Beirut, PM2.5 comprises about 61% of PM10,
revealing that the fine particles, which have greater 3.5. Statistical analysis of exceedance of the WHO
effect of human health, constitute a large fraction of air-quality standard
PM10. This finding is similar to that of many other
studies. For example, Aldabe et al. (2011), The daily mean PM10 and PM2.5 WHO air-
Gerasopoulos et al. (2007), He et al. (2001), and Ye et quality standards are 50 and 25 µg.m−3, respectively,
al. (2003) have reported mean PM2.5/PM10 ratios of and are not to be exceeded on more than 3 days per
0.64, 0.55, 0.63, and 0.63, respectively, for their entire year. Fig. 5 illustrates the percentage of days

1697
 Farah et al./Environmental Engineering and Management Journal 17 (2018), 7, 1693-1700
exceeding these standards for PM10 and PM2.5 during
the different months of the study period. For PM10, the
greatest number of exceedance days (25 days)
occurred in July (25 days) followed by August (19
days) and April (16 days). There were 133 days during
the observation period on which daily concentrations
of PM10 exceeded the upper threshold limit, which
represents 39% of the sampling period (133 out of
343). For PM2.5, the greatest number of exceedance
days occurred in April (19 days) followed by May,
March, and June (all 16 days). There were 129 days
during the observation period on which daily
concentrations of PM2.5 exceeded the upper threshold
limit, which represents 38% of the sampling period
(129 out of 343). The highest daily averages of the
PM10 and PM2.5 concentration were observed during
spring and summer (March to July). These findings
indicate the role played by dust events within this area
(Kutiel and Furman, 2003).
3.6. Relationships between PM and other pollutants
Studies have shown that PM causes greater
adverse effects when associated with SO2, NO2, CO,
and O3 (e.g., Cheng et al., 1998; Dockery et al., 1993).
Gaseous pollutants (except ozone) are associated
mostly with local sources, including exhaust
(a)
Fig. 5. Percentage of exceedance days for PM10 and PM2.5 concentrations during the different months of the study period: (a); (b)
Fig. 6. Monthly average concentrations of O3 and NO2 during the study period at the HUV site
1698
emissions, domestic heating during the cold season,
and industrial activities. To establish the association
between PM and traffic-related gaseous pollutants,
data from the HUV site were considered.
Simultaneous measurements of NO2 and O3 were
conducted at the same monitoring site to achieve this
objective, and Fig. 6 summarizes the monthly average
concentrations of all the gaseous species obtained
during the study period. The main source of the high
concentration of NO2 is largely the high volumes of
vehicular traffic in Beirut (Afif et al., 2009).
Concentrations of PM10, PM2.5, and PM10–2.5
were correlated with NO2 and O3 measured at the same
monitoring site. The most significant correlation was
found for PM2.5 with NO2 (r = 0.7), whereas PM10 and
PM10–2.5 were found not to be associated with any
gaseous pollutant. Strong associations between fine
particle mass concentrations and primary gaseous
pollutants have been reported previously reported
following air-quality surveys in Birmingham, UK,
suggesting the importance of road traffic-related
emissions (Harrison et al., 1997). The contribution to
levels of PM10 and PM2.5 from vehicular exhausts was
estimated using regression analysis (Table 2). NO2
was used as a vehicular emission tracer because it is
assumed that NO2 is almost entirely attributable to
emissions from road traffic.
Dec. Jan.
PM2.5 6%
8%
Feb.
Nov.
12%
12%
Oct.
1%
Sep.
2% Mar.
July 12%
7%
June
12% Apr.
15%
May
14%
(b)
 Analysis of the continuous measurements of PM10 and PM2.5 concentrations in Beirut, Lebanon
Table 2. Regression analysis between PM and NO2
Regression relationship
PM10 [PM10] = 29 + 0.28 [NO2]
PM2.5 [PM2.5] = 2.2 + 0.23 [NO2]
PM10-2.5 [PM10-2.5] = 11.5 + 0.16 [NO2]
For PM2.5, a very small value of the intercept
was found (2.2 µg.m−3) for the background together
with 28.1 µg.m−3 (mean–intercept) of vehicular-
derived material, whereas for PM10, the background
concentration was 29 µg.m−3 with a value of 22.3
µg.m−3 representing the traffic-derived material.
These results suggest that, on average, 93% of PM2.5
and 43% of PM10 particle mass concentrations are
contributed by the transported pollution of road traffic
exhaust from the city.
5. Conclusions
Sampling of PM10 and PM2.5 was conducted
during the period January through December 2012 in
Beirut. To the best of our knowledge, this represents
the longest record of simultaneous PM10 and PM2.5
measurements in Beirut. The daily average
concentrations of PM10 and PM2.5 were 51.3 ± 33.1
and 30.3 ± 19.4 µg.m−3, respectively, with
corresponding maximum values of 359.7 and 208.6
µg.m−3. The mean concentration of coarse particles
(PM10–2.5) was found to be 41% of the average PM10,
suggesting that the site was influenced by re-
suspended surface dust and soil. The annual average
concentrations of PM10 and PM2.5 exceeded the WHO
standards (i.e., 20 µg.m-3 for PM10 and 10 µg.m−3 for
PM2.5) by 150% and 200%, respectively. The mean
PM2.5/PM10 ratio for the entire study period was 0.61
± 0.12, indicating that in Beirut, about 61% of PM10
comprises PM2.5. Such a large fraction of fine particles
could have considerable effect on health.
The highest daily averages of PM10 and PM2.5
were observed in spring and summer (March to July),
indicating the important role dust events play within
this area. The correlation between PM and NO2
indicated that vehicular exhaust emissions contribute
an average of 93% of PM2.5 and 43% of PM10 with the
remainder derived from other sources. The present
study demonstrated the scale and importance of high
concentrations of PM.
Therefore, we have initiated the collection of
data on health outcomes from seven hospitals in Beirut
(80% of hospital beds) to correlate them with air-
pollution indicators. This study comprises the
framework of the Beirut Air Pollution and Health
Effects (BAPHE) project. The outcome of this study
is expected to provide compelling evidence for policy
makers and local authorities regarding the necessity
for the design and implementation of appropriate air-
quality control measures in Lebanon.
Acknowledgments
The authors thank the Lebanese National Council
for Scientific Research (CNRS), Regional Council of the
Isle de France, University of Saint Joseph, Lebanese non-
r
0.39
0.70
0.24
governmental organization arcenciel, and the CEDRE
project (Ministry of Foreign Affairs in France and the
Ministry of Higher Education and Research in France and
Lebanon) for supporting this research financially.
References
Afif C., Dutot A., Jambert C., Abboud M., Adjizian-Gérard
J., Farah, W., Perros P.E., Rizk T, (2009), Statistical
approach for the characterization of NO2 concentrations
in Beirut, Air Quality Atmosphere and Health, 2, 57-67.
Aldabe J., Elustondo D., Santamaria C., Lasheras E.,
Pandolfi M., Alastuey A., Querol X., Santamaria J.M.,
(2011), Chemical characterisation and source
apportionment of PM2.5 and PM10 at rural, urban and
traffic sites in Navarra (north of Spain), Atmospheric
Research, 102, 191-205.
Badarinatha K.V.S., Kharola S.K., Kaskaoutisb D.G.,
Sharmaa A.R., Ramaswamy V., Kambezidis H.D.,
(2010), Long-range transport of dust aerosols over the
Arabian Sea and Indian region - A case study using
satellite data and ground-based measurements, Global
and Planet Change, 72, 164-181.
Brauer M., Hoek G., Van Vliet P., Meliesfste K., Fisher P.
H., Wijga A., Koopman L. P., Neijens, H. J., Gerritsen
J., Kerkhof M., Heinrich J., Bellander T., Brunekreef
B., (2002), Air pollution from traffic and the
development of respiratory infections and asthmatic
and allergic symptoms in children, American Journal of
Critical Care and Respiratory Medicine, 166, 1092-
1098.
Braun-Fahrlander C., Vuille J. C., Sennhauser F.H., (1997),
Respiratory health and long-term exposure to air
pollutants in Swiss schoolchildren. SCARPOL Team.
Swiss Study on Childhood Allergy and Respiratory
Symptoms with Respect to Air Pollution, Climate and
Pollen, American Journal of Critical Care and
Respiratory Medicine, 155, 1042-1049.
Buzorius G., Hameri K., Pekkanen J., Kulmala M., (1999),
Spatial variation of aerosol number concentration in
Helsinki city, Atmospheric Environment, 33, 553-565.
CAS, (2007), Central Administration of Statistics. Living
Conditions Survey 2007, On line at:
http://www.cas.gov.lb/index.php/en/demographic-and-
social-en/population-en. Accessed 15 April 2013
Chaloulakou A., Kassomenos P., Spyrellis N., Demokritou
P., Koutrakis P., (2003), Measurements of PM10 and
PM2.5 particle concentrations in Athens, Greece,
Atmospheric Environment, 37, 649-660.
Chen Y., Yang Q., Krewski D., Shi Y., Burnett R.T., Mc
Grail K., (2004), Influence of relatively low level of
particulate air pollution on hospitalization for COPD in
elderly people, Inhalation Toxicology, 16, 21-25.
Cheng L., Sandhu H.S., Angle R.P., Myrick R.H., (1998),
Characteristics of inhalable particulates matter in
Alberta cities, Atmospheric Environment (UK), 32,
3836-3844.
Dockery D.W., Pope C.A., Xu X., Spengler J.D., Ware J.H.,
Fay M.E., Ferris B.G., Speizer F.E., (1993), An
association between air pollution and mortality in six
cities, Journal of Medicine (UK), 329, 1753-1759.
1699
 Farah et al./Environmental Engineering and Management Journal 17 (2018), 7, 1693-1700
Dominici F., Peng R.D., Bell M.L., Pham L., Mc Dermott
A., Zeger S.L., Samet J.M., (2006) Fine Particulate Air
Pollution and Hospital Admission for Cardiovascular
and Respiratory Diseases, Journal of the American
Medical Association, 295, 1127-34.
Draxler R.R., Gillette D.A., Kirkpatrick J.S., Heller J.,
(2001), Estimating PM10 air concentrations from dust
storms in Iraq, Kuwait and Saudi Arabia, Atmospheric
Environment, 35, 4315-4330.
Environnement SA, (2012), French company offering
advanced environmental monitoring equipment, On
line at: http://www.environnement-sa.com/produits-
2/surveillance-de-la-qualite-de-lair
El-Zein A., Nuwayhid I., El-Fadel M., Mroueh S., (2007),
Did a ban on diesel fuel reduce emergency respiratory
admissions for children?, Science of the Total
Environment, 384, 134-140.
Gerasopoulos E., Koulouri E., Kalivitis N., Kouvarakis G.,
Saarikoski S., Makela T., Hillamo R., Mihalopoulos N.,
(2007), Size-segregated mass distributions of aerosols
over eastern Mediterranean: seasonal variability and
comparison with AERONET columnar size-
distributions, Atmosphere Chemistry and Physics, 7,
2551-2561.
Harrison R. M., Deacon A.R., Jones M.R., Appleby R.S.,
(1997), Sources and processes affecting concentrations
of PM10 and PM2.5 particulate matter in Birmingham,
(UK), Atmospheric Environment, 31, 4103-4117.
He K., Yang F., Ma Y., Zhang Q., Yao X., Chan C. K., Cadle
S., Chan T., Mulawa P., (2001), The characteristics of
PM2.5 in Beijing, China, Atmospheric Environment, 35,
4959-4970.
Ho K.F., Lee S.C., Chan C.K., Yu J.C., Chow J.C., Yao X.
H., (2003), Characterization of chemical species in
PM2.5 and PM10 aerosols in Hong Kong, Atmospheric
Environment, 37, 31-39.
Jansen N.A., Van Manson D.F.M., Vander Jagt K.,
Hasseman H., Hoek G., (1997), Mass concentration and
elemental composition of airborne particulate matter at
street and background locations, Atmospheric
Environment, 31, 1185-1193.
Kutiel H., Furman H., (2003), Dust Storms in the Middle
East: Sources of Origin and their Temporal
Characteristics, Indoor Built Environment, 12, 419-426.
Khodeir M., Shamy M., Alghamdi M., Zhong M., Sun H.,
Costa M., Chen L. C., Maciejczyk P., (2012), Source
apportionment and elemental composition of PM2.5 and
PM10 in Jeddah City, Saudi Arabia, Atmospheric
Pollution Research, 3, 331-340.
Marcazzan G.M., Ceriani M., Valli G., Vecchi R., (2003),
Source apportionment of PM10 and PM2.5 in Milan
(Italy) using receptor modeling, Science of the Total
Environment, 317, 137-147.
Massoud R., Shihadeh A., Roumié M., Youness M., Gerard
J., Saliba N., Zaarour R., Abboud M., Farah W., Aoun
Saliba N., (2011), Interurban variability of PM10 and
PM2.5 in an Eastern Mediterranean city, Atmospheric
Research, 101, 893-901.
1700
Meng Z., Lu B., (2007), Dust events as a risk factor for daily
hospitalization for respiratory and cardiovascular
diseases in Minqin, China, Atmospheric Environment,
41, 7048-7058.
Pey J., Rodriguez S., Querol X., Alastuey A., Moreno T.,
Putaud J. P., Van Dingenen R., (2008), Variations of
urban aerosols in the western Mediterranean,
Atmospheric Environment, 42, 9052-9062.
Oberdörster G., Oberdörster E., Oberdörster J., (2005),
Nanotoxicology: An emerging discipline evolving from
studies of ultrafine particles, Environmental Health
Perspective, 113, 823-39.
Penard-Morand C., Charpin D., Raherison C.,
Kopferschmitt C., Caillaud D., Lavaud F., Annesi-
Maesano I., (2005), Long-term exposure to background
air pollution related to respiratory and allergic health in
schoolchildren ISAAC, Clinical and Experimental
Allergy, 35, 1279-1287.
Querol X., Peya J., Pandolfi M., Alastuey A., Cusack M.,
Pérez N., Moreno T., Viana M., Mihalopoulos N.,
Kallos G., Kleanthous S., (2009), African dust
contributions to mean ambient PM10 mass-levels
across the Mediterranean Basin, Atmospheric
Environment, 43, 4266-4277.
Rodriguez S., Querol X., Alastuey A., Viana M.M., Alarcon
M., Mantilla E., Ruiz C. R., (2004), Comparative PM10-
PM2.5 source contribution study at rural, urban and
industrial sites during PM episodes in eastern Spain,
Science of the Total Environment, 328, 95-113.
Saliba N.A., El Jam F., El Tayar G., Obeid W., Roumie M.,
(2010), Origin and variability of particulate matter
(PM10 and PM2.5) mass concentrations over an eastern
Mediterranean city, Atmospheric Research, 97, 106-
114.
Shahsavania A., Naddafia K., Jafarzade Haghighifard N.,
Mesdaghiniaa A., Yunesiana M., Nabizadeha R.,
Arahamid M., Sowlata M. H., Yarahmadia M., Sakib
H., Alimohamadia M., Nazmaraa S., Motevaliane S.A.,
Goudarzib G., (2012), The evaluation of PM10, PM2.5,
and PM1 concentrations during the Middle Eastern Dust
(MED) events in Ahvaz, Iran, from April through
September 2010, Journal of Arid Environment, 77, 72-
83.
Schwartz J., (1994), Total Suspended Particulate Matter and
Daily Mortality in Cincinnati, Ohio, Environmental
Health Perspective, 102, 186-189
Schwartz J., (1995), Short-term fluctuations in air pollution
and hospital admissions of the elderly for respiratory
disease, Thorax, 50, 531-538.
Wordley J., Walters S., Ayres, J.G., (1997), Short-term
variations in hospital admissions and mortality and
particulate air pollution, Journal of Occupational and
Environmental Medicine, 54, 108-116.
Ye B., Ji X., Yang H., Yao, X., Chan C. K., Cadle S. H.,
Chan T., Mulawa P. A., 2003, Concentration and
chemical composition of PM2.5 in Shanghai for a 1-year
period, Atmospheric Environment, 37, 499-510.
Environmental Engineering and Management Journal July 2018, Vol.17, No. 7, 1701-1709
http://www.eemj.icpm.tuiasi.ro/; http://www.eemj.eu

“Gheorghe Asachi” Technical University of Iasi, Romania

CONTRASTIVE SOIL PROPERTIES, MICROBIAL STRUCTURE

AND SOIL ENZYMES IN THE RHIZOSPHERE OF Scirpus triqueter
AND BULK SOIL IN PETROLEUM-CONTAMINATED WETLAND

Jing Wei, Xiaoyan Liu ∗, Chuanhua Wang, Xueping Chen, Xia Liang, Qian Wang

Laboratory of Environmental Remediation, College of Environmental and Chemical Engineering,

Shanghai University, 99 Shangda Road, Shanghai, 200444, China

Abstract

Though Scirpus triqueter plays an important role in protecting the ecological functions of Yangtze River estuary wetland which is
frequently contaminated by petroleum, knowledge is limited about the mechanisms involved. In this study, comparative analysis
of soil properties, microbial community structure and soil enzymes in the rhizosphere of S. triqueter and bulk soil was conducted
throughout one year to provide an insight into the phytoremediation mechanism in the field. The total petroleum hydrocarbon
content and soil pH in the rhizosphere were averagely 0.06 g kg-1 and 0.1 lower, respectively, than in the bulk soil. Analysis of
phospholipid fatty acids shew the microbial community structure was quite different from that in the bulk soil with much higher
activity and diversity. The relative abundance of Gram-positive to Gram-negative bacteria was 0.12 lower in the rhizosphere, while
the relative abundance of aerobes to anaerobes was 0.74 higher, than in the bulk soil. Positively correlated with the relative
abundance of fungi to bacteria and microbial diversity, the activities of dehydrogenase, catalase, polyphenol oxidase were
significantly lower (P < 0.05) in the bulk soil than in the rhizosphere. Redundancy analysis and Pearson correlation shew that soil
organic matter, total nitrogen, total phosphorus, as well as temperature were crucial factors influencing soil microbial communities
and enzymes which were responsible for the degradation of petroleum hydrocarbons. These results would improve our
understanding of the phytoremediation processes in fragile contaminated wetland and supply valuable information for its
application in the restoration of damaged ecosystems.

Key words: microbial community structure, petroleum pollution, phytoremediation, Scirpus triqueter, soil enzymes

Received: February, 2013; Revised final: August, 2014; Accepted: September, 2014; Published in final edited form: July 2018

1. Introduction HYRE wetland, helps greatly to protect, maintain and

The Huangpu-Yangtze River estuary (HYRE)
wetland is the most valuable natural estuarine wetland
in Shanghai, China, which plays an important role in
maintaining the ecological environment. It is also an
ideal model to study the impact of human activities on
estuarine wetland. However, HYRE wetland is
suffering from frequent petroleum contamination in
recent years. It is estimated that more than 760 tons of
petroleum from oil leakage has intruded into the
HYRE wetland in the last few years (Liu et al., 2010b).
Fortunately, Scirpus triqueter, the dominant plant in
restore the wetland ecosystem through
phytoremediation (Liu et al., 2011; Zhang et al.,
2011). Though the petroleum-degrading ability of S.
triqueter has been studied in greenhouse experiments,
no field study is conducted till now. The situation is
much more complicated in the field, the extremely
intricate interactions among plants, environmental
factors (such as weather, soil properties and tidewater)
and pollution may lead to discriminating results.
Therefore it is necessary to carry out field study in
HYRE wetland to make up for the current inadequacy
of scientific data.

∗
Author to whom all correspondence should be addressed: e-mail: lxy999@shu.edu.cn; Phone: +86 21 66137767; Fax: +86 21 66137761
 Wei et al./Environmental Engineering and Management Journal 17 (2018), 7, 1701-1709
Soil microorganisms play an important role in
the degradation of organic contaminants (Ciancio et
al., 2016; Gerhardt et al., 2009). Liu et al. (2010a)
found that 58.2% of the petroleum hydrocarbons were
degraded by indigenous microbes after one year of
bioremediation in a pot experiment. It is thought that
plants can enhance the pollutant-removal ability of
soil microbes through the so called rhizosphere effect
(Liu et al., 2010a; Mihalache et al., 2016). Sun et al.
(2010) proved that root-exudates enhanced the
biodegradation of phenanthrene and pyrene by 15.5%
and 21.3%, respectively, in a low organic matter soil.
Meng and Zhu (2010) found that components from
celery roots increased the removal of pyrene by
around 22% in an industry soil, and lipophilic
components enhanced the relative abundances of
pyrene-degraders. In addition, the influences of plants
on soil microorganisms are plant-dependent (Yao and
Wu, 2010), which may lead to discrepant
phytoremediation efficiencies (Wei et al., 2013). As a
culture-independent approach to explore the
fingerprints of soil microbes, phospholipid fatty acid
(PLFA) model is a preferable method to monitor the
complex soil microbial community structure (Cowie
et al., 2010; Frostegård et al., 2011). Therefore, PLFA
model was used to analyze the relative abundance of
certain microflora, including bacteria (B), fungi (F),
Gram-positive bacteria (GP) and Gram-negative
bacteria (GN), aerobes (AE) and anaerobes (AN), in
this study.
Soil enzymes, deriving primarily from soil
microorganisms, plant roots, plant and animal
residuals, are sensitive indicators for the soil function
and environmental strain in affected ecosystems
(Zhou et al., 2011). Their activities vary widely
depending on the enzyme species, soil types,
environmental factors, plants, as well as pollutants
(Wang et al., 2009). Wyszkowska and Wyszkowski
(2010) found that petroleum contamination inhibited
the activities of soil dehydrogenases and urease to a
large degree in spring rape and oat soils. Considering
that soil enzymes can serve to measure the health and
sustainability of disturbed ecosystems (Boerner et al.,
2005), activities of important soil oxidoreductases and
hydrolases, as well as the relationship between soil
microbial communities and enzymes, were measured
to probe into the influences of S. triqueter on the soil
function in this study.
This study was carried out in HYRE wetland
with the objectives: (1) to explore the differences of
soil properties, microbial community structure and
soil enzymes in the rhizosphere and bulk soil; (2) to
monitor seasonal fluctuation of microbial
communities and soil enzymes throughout one year;
(3) to investigate the interactions between soil
properties, soil microbial communities and soil
enzymes during phytoremediation of petroleum
polluted wetland.
2. Materials and methods
2.1. Field sampling
1702
Sampling sites were set in the HYRE wetland,
the north of Shanghai, China. Characterized by the
north semi-tropical monsoon climate, the annual
precipitation is 1022 mm and approximately 70
percent falls between May and September. The annual
mean temperature is around 15.3ºC according to
Shanghai Meteorological Center. Almost 80 percent
of the mesolittoral zone in HYRE wetland is colonized
by S. triqueter, a kind of perennial herb, whose
growing period is from April to November.
In this study, rhizosphere and bulk soil samples
were collected every month from March, 2011 to
February, 2012. Rhizosphere soil samples (SS) were
randomly collected from the root surface of S.
triqueter at the depth of 0-20 cm and bulk soil samples
(BS) were collected at the same depth, but at least 1 m
away from the nearest plants (Toyama et al., 2011).
All soil samples were processed within 6 h according
to previous studies with some modifications (Nie et
al., 2009). Plant roots and stones were hand-picked
out. Soils were freeze-dried, then passed through 1
mm mesh to get homogenized. All the samples were
divided into two parts: one was stored at 4ºC for the
analysis of soil properties and soil enzymes as
mentioned previously (Wei et al., 2013); the other part
was stored at -80ºC to avoid the degradation of PLFAs
and used for soil microorganism analysis.
2.2. Soil properties
Total petroleum hydrocarbon (TPH) was
ultrasonically extracted from 5 g of treated soil
samples with dichloromethane, then the content was
determined through gravimetric analysis as described
in previous studies (Liu et al., 2011; Mishra et al.,
2001; Zhang et al., 2013). Values of pH and total
dissolved solid (TDS) were determined by HI255 pH-
TDS-conductivity-salinity meter after mixing the soil
with deionized water (1:10, w/w); total nitrogen (TN)
content was analyzed by Kjeldahl apparatus (Buchi
Distillation Unit B-324); soil temperature-moisture
meter (LZB-I) was used for the measurement of the
soil temperature (T); total phosphorus (TP) and
organic matter (OM) were determined by colorimetric
and volumetry method, respectively (Rukun, 1999).
2.3. Soil microbial communities
The activity of fluorescein diacetate hydrolase
(FDA), expressed in A490 g-1 dry soil 2 h-1, was used to
reflect the total activity of soil microorganisms (Green
et al., 2006; Wei et al., 2013). To measure the FDA
activity, 1 g of treated soil was mixed with 30 mL of
phosphate buffer (pH 7.6) and 1 mL of fluorescein
diacetate solution (2 g mL-1, dissolved in acetone),
then the mixture was incubated in the dark at 30ºC for
2 h. After centrifugal separation, the absorbance at 490
nm was detected.
Microbial community structure was analyzed
using PLFA method. In general, crude fatty acids were
extracted from 3 g of soil by a mixture of citrate buffer,
methanol, and chloroform (Frostegård et al., 1993; Li
 Contrastive soil properties, microbial structure and soil enzymes in the rhizosphere
et al., 2010). Then the crude fatty acids were passed
through silica gel columns with methanol as the eluent
to obtain phospholipids. After that, phospholipids
were methylated and quantified using GC-MS (Li et
al., 2010). PLFAs were named according to the
nomenclature described in previous study (Cao et al.,
2012). Shannon index H was used to evaluate the
diversity of microbial communities and calculated by
Eq. (1):
R
H = −∑ pi ln pi
i =1
where H represented the value of Shannon index, pi
was the relative abundance of each detected PLFA, R
was the total number of detected PLFAs (Yao and Wu,
2010; Zornoza et al., 2009)
2.4. Soil enzymes
Activities of three oxidoreductases
[dehydrogenase (DHA, EC 1.1), catalase (CAT, EC
1.11.1.6) and polyphenoloxidase (PPO, EC 1.10.3.1)]
and three hydrolases [sucrase (SUC, EC 3.2.1.26),
urease (URE, EC 3.5.1.5) and alkaline phosphatase
(PHOS, EC 3.1.3.1)] were measured.
URE activity, expressed as mg NH4-N g-1 soil
-1
24 h , was measured according to the hydrolysis rate
of carbamide. SUC activity, expressed as mg glucose
g-1 soil 24 h-1, was determined by incubating treated
soil with sucrose solution (Schinner and Mersi, 1990).
PHOS (mg phenol g-1 soil 3 h-1), DHA (mg TF g -1 soil
48 h-1) and PPO (mg purpurogallin g-1 soil 2 h-1) were
measured following the methods previously described
(Guan, 1986; Tang et al., 2010; Wyszkowska and
Wyszkowski, 2010). Hydrogen peroxide was used as
the substrate for CAT determination, and the unit of
its activity was mL 0.1 M KMnO4 g-1 soil 30 min-1
(Johnson and Temple, 1964).
2.5. Statistical analysis
Each data was the mean value (±SD) of three
replicates. SPSS 17.0 was used for paired t test to
examine the differences of rhizosphere and bulk soils
in data sets of soil properties, soil microorganisms and
soil enzymes. Principal component analysis (PCA)
was used to compare microbial community structure
in the rhizosphere and bulk soil. Redundancy analysis
(RDA) performed by Canoco for windows 4.5 was
used to analyze the microbial characteristics across
soil property gradients, as well as soil enzymes across
microbial population gradients. Pearson correlation
matrix was used to assess the covariance between soil
properties, microbial communities and enzyme
activities.
3. Results and discussion
3.1. Soil properties
Paired t-test was conducted to compare the soil
characteristics in the rhizosphere and bulk soil (Table
1). TPH content in the rhizosphere was significantly
lower (P < 0.05) than in the bulk soil, indicating that
S. triqueter can greatly enhance the degradation of
petroleum hydrocarbons. In addition, the pH value
(7.3 ± 0.1) in the bulk soil was significantly higher
than that (7.2 ± 0.0) in the rhizosphere (P < 0.001),
while the OM content of 75.03 ± 18.21 g kg-1 in the
rhizosphere of S. triqueter was slightly higher than
that of 61.28 ± 17.07 g kg-1 in unplanted soil. By
contrast, differences of soil temperature, TN, TP and
(1) TDS contents, between the two types of soil were less
remarkable.
Organic acids released by plants, including
citric acid, malic acid, oxalic acid and so on, may
contribute to the decrease of soil pH. It is thought that
petroleum hydrocarbons can affect increase soil pH,
deteriorate soil function and decrease soil
permeability.
On the other hand, plants can improve soil
properties through releasing ample carbon sources and
mineral nutrients into the soil in the form of root
exudate, mucilage, colloidal and so on (Bertin et al.,
2003; Hinsinger, 2001; Richardson et al., 2009).
Therefore different soil properties between the
rhizosphere and bulk soil were the combined results
by petroleum hydrocarbons and plants.
3.2. Soil microbial communities
A total of 23 PLFAs were detected, most of
which had 15-18 carbon in their molecules. PLFAs
i12:0, a13:0, 13:0, i14:0, i15:0, 15:0, a15:0, i16:0,
16:1ω7c, 16:1ω7t, i17:0, a17:0, 17:0, 18:0, cy17:0,
18:1ω7c and cy19:0 were used to indicate bacterial
biomass; 16:1ω5t, 18:1ω9t, 18:1ω9c, 18:2ω6,9c were
used as fungal biomarkers; 16:1ω7c, 16:1ω7t,
18:1ω7c, cy17:0, cy19:0 were chosen to represent GN;
branched saturated fatty acids i14:0, i15:0, i16:0,
i17:0, a15:0 and a17:0 indicated GP; 14:0 and 16:0
were common PLFAs in most microorganisms
(Frostegård et al., 1993; Velasco et al., 2010; Zhang et
al., 2010). The ratio of monounsaturated fatty acids to
branched saturated fatty acids was used to indicate the
relative abundance of aerobic to anaerobic microbes
(AE/AN) (Li et al., 2010; Su and Yang, 2009).
PCA plot was used to compare the microbial
community structure in the rhizosphere and bulk soil
(Fig.1). The first principal component accounted for
75.4% of the total variance and the second explained
17.3%. Samples in the rhizosphere were completely
separated with those in the bulk soil as illustrated in
the PCA plot, therefore their microbial structures were
quite different from each other. Microbial
communities in the rhizosphere were characterized by
monounsaturated fatty acids, such as 18:1ω7c,
16:1ω5t, 16:1ω7t and 16:1ω7c, with higher diversity
and activity. By contrast, the relative abundances of
saturated fatty acids were much higher in the bulk soil,
resulting in the higher GP/GN.
Paired t-test analysis was used to compare the
microbial community structure in the bulk and
rhizosphere soil (Table 2).
1703
 Wei et al./Environmental Engineering and Management Journal 17 (2018), 7, 1701-1709

.
Table 1. Comparison of soil properties in the bulk and rhizosphere soil through paired t-test

Bulk Rhizosphere Difference t P

T（ºC） 8.3 ± 7.0 8.2 ± 7.0 0.1 ± 0.2 1.129 0.283
TN（g kg-1） 1.01 ± 0.18 1.07 ± 0.19 0.06 ± 0.13 -1.642 0.129
TP（g kg-1） 0.11 ± 0.02 0.11 ± 0.01 0.00 ± 0.01 1.687 0.120
OM（g kg-1） 61.3 ± 17.1 75.0 ± 18.2 13.8 ± 24.7 -1.931 0.080
pH 7.3 ± 0.1 7.2 ± 0.0 0.1 ± 0.1 6.934 0.000
TPH（mg kg-1） 712 ± 178 470 ± 244 241 ± 221 3.781 0.003
TDS（mg L-1） 113 ± 55 91 ± 33 21 ± 45 1.601 0.138
T, soil temperature; TN, total nitrogen; TP, total phosphorus; OM, organic matter; TPH, total petroleum hydrocarbon; TDS, total dissolved solid

Fig. 1. PCA plot of microbial communities in the rhizosphere and bulk soil (FDA, fluorescein diacetate hydrolase; AE/AN, the
relative abundance of aerobes to anaerobes; GP/GN, the relative abundance of Gram-positive bacteria to Gram-negative bacteria;
F/B, the relative abundance of fungi to bacteria; H, Shannon index. BS, bulk soil; SS, soil in the rhizosphere of S. triqueter,
digitals reflect the sampling month)

Table 2. Comparison of soil microbial structure in the bulk and rhizosphere soil by paired t-test
Bulk Rhizosphere Difference t P
FDA (abs g-1) 0.22 ± 0.10 0.29 ± 0.14 0.45 ± 0.32 -2.352 0.038
AE/AN 1.89 ± 0.43 2.63 ± 0.33 0.07 ± 0.10 -4.996 0.000
GP/GN 0.62 ± 0.18 0.50 ± 0.14 0.74 ± 0.51 4.017 0.002
F/B 0.03 ± 0.03 0.12 ± 0.09 0.12 ± 0.10 -4.267 0.001
H 1.94 ± 0.20 2.10 ± 0.25 0.16 ± 0.13 -4.203 0.001
FDA, fluorescein diacetate hydrolase; AE/ AN, the relative abundance of aerobes to anaerobes; GP/GN, the relative abundance of Gram-positive
bacteria to Gram-negative bacteria; F/B, the relative abundance of fungi to bacteria; H, Shannon index

FDA activity was significantly higher (P <
0.05) in the rhizosphere (0.29 ± 0.14 abs g-1) than that
in bulk soil (0.22 ± 0.10 abs g-1), which indicated the
soil microbes were more active in the rhizosphere than
in the bulk soil. The activity of soil microorganisms
was thought to be largely related with petroleum
biodegradation (Kirk et al., 2005; Liu et al., 2011).
The relative abundance of AE to AN was
distinctly higher (P < 0.001) in the rhizosphere than in
bulk soil (Table 2), and this disparity reached to the
summit in August (Fig.2A). Similarly, the relative
abundance of fungi to bacteria in the rhizosphere was
3-4 times higher than in the bulk soil (Table 2), and
the biggest difference appeared in October (Fig.2B).
However, the relative abundance of GP to GN
decreased by about 20% in the rhizosphere compared
with that in bulk soil (Table 2) during the whole
vegetative period of S. Triqueter (Fig.2C). In addition,
the average value of H was 2.10 ± 0.25 in the
rhizosphere, while 1.94 ± 0.20 in the bulk soil,
reflecting that soil microbes in the rhizosphere were
more diverse during the whole year (Table 2, Fig.2D).
The lower TPH content and moderate pH
(Table 1) in the rhizosphere, as well as various organic
carbons released by plant roots, provided a better
habitat for soil microbes and accounted for the higher
activity and diversity of microbial communities in the
rhizosphere (Table 2). It was believed that the
microbial community structure was plant-dependent
in the rhizosphere (Smalla et al., 2001). For example,
soil microbial populations were dominated by cy17:0,
16:1ω5c and 18:1ω5c in the rhizosphere of V. tricolor,
while by 18:1ω7c, 18:0 and 10Me18:0 in the
rhizosphere of C. comosum (Wei et al., 2013).

1704
 Contrastive soil properties, microbial structure and soil enzymes in the rhizosphere

Fig. 2. Soil microbial community structure in the rhizosphere and bulk soil throughout one year
AE/AN, the relative abundance of aerobes to anaerobes; GP/GN, the relative abundance of Gram-positive bacteria to Gram-
negative bacteria; F/B, the relative abundance of fungi to bacteria; H, Shannon index. BS, bulk soil; SS,
soil in the rhizosphere of S. triqueter

The relative abundances of effective
petroleum-degrading microorganisms (AE and GN)
were much higher in the rhizosphere of S. triqueter,
which can explain the reason why S. triqueter had a
good ability to enhance the degradation of petroleum
hydrocarbons in the HYRE wetland (Wei et al., 2014;
Zhang et al., 2011). RDA was used to explore the
effect of soil properties on microbial communities.
The first two axes explained 69.4% and 25.4% of the
total variance respectively (Fig.3).
Fig. 3. RDA plot based on microbial communities and soil
properties in the rhizosphere and bulk soils throughout one
year (FDA, fluorescein diacetate hydrolase; AE/AN, the
relative abundance of aerobes to anaerobes; GP/GN, the
relative abundance of Gram-positive bacteria to Gram-
negative bacteria; F/B, the relative abundance of fungi to
bacteria; H, Shannon index. T, soil temperature; TN, total
nitrogen; TP, total phosphorus; OM, organic matter; TPH,
total petroleum hydrocarbon; TDS, total dissolved solid.
BS, bulk soil; SS, soil in the rhizosphere of S. triqueter,
digitals reflect the sampling month)
Rhizosphere samples were located on the left
of Axis1, while bulk samples on the right. With the
greatest arrow lengths among all the environmental
factors, pH and TPH concentration were two main
determinants of the soil microbial community
structure. By contrast, TDS affected the soil microbial
communities slightly. The completely contrary
tendencies of TPH and AE/AN revealed that TPH
concentration made a negative influence on the
relative abundance of AE/AN. However, OM, TP and
TN influenced soil microbial structure by increasing
microbial diversity and activity. In addition, soil
microbial structure appeared to be similar in the same
season, for example, rhizosphere samples in March,
April and May located adjacently in the ordination
diagram of RDA.
As an important indicator for the soil microbial
ecology, the ratio of F/B has been used to indicate the
shift of soil microbial community structure under
environmental stress (Kaur et al., 2005; Yao and Wu,
2010). It has been reported that F/B ratio was
positively correlated with soil fertility quality, but
negatively with pollution level (Liu and Herbert,
2002; Zhang et al., 2012). Therefore, the higher
relative abundance of fungi in the rhizosphere might
result from the lighter pollution level to some extent.
pH deviation could impose great stress on soil
microorganisms by changing the integrity and
function of cellular membranes or altering the
bioavailability of nutrients and contaminants (Baath
and Anderson, 2003; Fierer and Jackson, 2006; Wick
et al., 2010), therefore it was one of the main factors
determining microbial community structure (Fig.3).
Pearson correlation coefficients between soil
properties and microbial characteristics were listed in

1705
 Wei et al./Environmental Engineering and Management Journal 17 (2018), 7, 1701-1709

Table 3. In the rhizosphere, microbial activity The obvious higher activities of

(indicated by FDA activity) and diversity (indicated
by H) were both positively correlated with OM
content (P < 0.05). In addition, OM was positively
associated with the relative abundances of F/B and
GP/GN. Plant residues are one of the main sources of
OM in the soil (Rukun, 1999).
Therefore one of the pathways that plants act
on soil microbial communities is by controlling the
soil organic matters. Microbial diversity and the ratio
of F/B were positively correlated with soil temperature
both in the bulk soil and in the rhizosphere (P < 0.001),
therefore, microbiota was more diverse in Summer
with higher relative abundance of F/B than in Winter
(Fig. 2). The positive correlation (P < 0.001) between
microbial diversity and TN was also observed.
Nitrogen is one of the most important nutrients that
regulate the growth and metabolism of microbes. In
most polluted sites, nitrogen is the limiting factor that
controls the activity and function of soil microbiota
(Nannipieri et al., 2003).
3.3. Soil enzymes
Paired t-test analysis revealed that activities of
CAT, PPO, SUC, PHOS and DHA were significantly
higher (P < 0.001) in the rhizosphere than in bulk soil
(Table 4). Soil enzymes participate in various
biogeochemical processes, such as the degradation of
organic pollutants, the decomposition of humus and
the redox of inorganic substances (Zhou et al., 2011).
oxidoreductases, related with the degradation of
organic pollutants, in the rhizosphere should be the
results of the more active soil microorganisms, and
made a good contribution to the degradation of
petroleum hydrocarbons. The higher activities of SUC
and PHOS, responsible for the transformation of
organic nitrogen and organic phosphorus,
respectively, indicated better soil fertility in the
rhizosphere. Seasonal shifts of soil enzyme activities
were observed in both types of soil (Fig. 4). The
highest CAT activity (3.73-4.33 mL 0.1M KMnO4 g-1
soil 30min-1) appeared in June, while the submit of
PPO (0.39-0.47 mg purpurogallin g-1 soil 2 h-1) was in
August. Activities of URE, DHA and PHOS were 2-4
times higher in summer (June, July and August) than
in winter (December, January and February). SUC
was 0.78 ± 0.06 and 0.26 ± 0.05 mg glucose g-1 soil 24
h-1 in May and February, respectively.
RDA was also performed to explore the
relationships between microbial communities and soil
enzymes (Fig. 5). The first and second axes explained
86.7% and 12.0% of the total variance respectively.
The similar trends of H and F/B in the RDA plot
demonstrated that they affected soil enzymes alikely.
The acute angles between F/B, CAT, DHA and PPO
indicated that the relative abundance of F/B was
positively correlated with activities of CAT, DHA and
PPO. Similarly, GP/GN exerted a positive effect on
DHA activity but a negative influence on PHOS
activity.

Table 3. Coefficients of Pearson correlation between soil properties and microbial characteristics

T TN TP OM pH TPH TDS
Bulk
FDA 0.374 0.444 0.394 -0.358 -0.279 0.317 -0.280
AE/AN -0.423 -0.632* -0.091 0.007 -0.154 0.204 0.667*
GP/GN 0.284 0.656* 0.604* -0.298 -0.568 0.223 -0.455
F/B 0.711** 0.546 0.320 0.447 0.215 0.133 -0.178
H 0.781** 0.799** 0.495 0.278 -0.055 -0.182 -0.296
Rhizosphere
FDA 0.080 0.341 0.709** 0.690* 0.501 0.688* 0.578*
AE/AN 0.271 0.117 0.112 -0.252 0.456 -0.465 -0.168
GP/GN 0.020 0.156 0.276 0.617* -0.082 0.626* 0.078
F/B 0.825** 0.663* 0.225 0.637* 0.391 0.137 -0.121
H 0.823** 0.801** 0.549 0.812** 0.563 0.387 -0.108
Asterisks mark different levels of significance: *P < 0.05, **P < 0.01. FDA, fluorescein diacetate hydrolase; AE/AN, the relative abundance of
aerobes to anaerobes; GP/GN, the relative abundance of Gram-positive bacteria to Gram-negative bacteria; F/B, the relative abundance of fungi
to bacteria; H, Shannon index, represents the microbial diversity. T, soil temperature; TN, total nitrogen; TP, total phosphorus; OM, organic matter;
TPH, total petroleum hydrocarbon; TDS, total dissolved solid

Table 4. Comparison of soil enzymes in bulk and rhizosphere soil by paired t-test

Bulk Rhizosphere Difference t P

CAT (mL 0.1M KMnO4 g-1 soil 30min-1) 2.41 ± 0.76 2.93 ± 0.72 0.52 ± 0.40 -4.576 0.001
PPO (mg purpurogallin g-1 soil 2 h-1) 0.17 ± 0.10 0.23 ± 0.12 0.06 ± 0.04 -4.345 0.001
SUC (mg glucose g-1 soil 24 h-1) 0.43 ± 0.17 0.53 ± 0.18 0.01 ± 0.04 -8.285 0.000
URE (mg NH4-N g-1 soil 24 h-1) 0.17 ± 0.08 0.14 ± 0.85 0.02 ± 0.10 -0.852 0.412
PHOS (mg phenol g-1 soil 3 h-1) 0.04 ± 0.01 0.08 ± 0.03 0.03 ± 0.02 -6.717 0.000
DHA (mg TF g -1 soil 48 h-1) 1.02 ± 0.43 1.47 ± 0.44 0.45 ± 0.32 -4.852 0.001
DHA, dehydrogenase; CAT, catalase; PPO, polyphenoloxidase; SUC, sucrase; URE, urease; PHOS, alkaline phosphatase

1706
 Contrastive soil properties, microbial structure and soil enzymes in the rhizosphere

5
A BS 2.8 D BS
4 SS SS

CAT
2.1

DHA
3 1.4
2 0.7
1 0.0

0.4 B BS
SS 1.0 E
BS
SS
URE

SUC
0.2 0.5

0.0 0.0
C 0.18
0.4 BS F BS
SS SS
PPO

0.12

PHOS
0.2
0.06

0.0 0.00

Apr. 2011
Mar. 2011
Mar. 2011

May 2011

Aug. 2011

May 2011
Apr. 2011

Jun. 2011

Jul. 2011

Sep. 2011

Dec. 2011
Jul. 2011

Sep. 2011

Jun. 2011

Aug. 2011

Oct. 2011

Jan. 2012
Nov. 2011
Dec. 2011

Nov. 2011
Oct. 2011

Feb. 2012

Feb. 2012
Jan. 2012

Date

Fig. 4. Variations of soil enzyme activities in the rhizosphere and bulk soil throughout one year
(DHA, dehydrogenase, mg TF g -1 soil 48 h-1; CAT, catalase, mL 0.1M KMnO4 g-1 soil 30min-1; PPO, polyphenoloxidase, mg
purpurogallin g-1 soil 2 h-1; SUC, sucrase, mg glucose g-1 soil 24 h-1; URE, urease, mg NH4-N g-1 soil 24 h-1; PHOS, alkaline
phosphatase, mg phenol g-1 soil 3 h-1. BS, bulk soil; SS, soil in the rhizosphere of S. triqueter)

temperature and TN, but negatively with TDS in the

Fig. 5. RDA plot based on microbial communities and soil
enzymes in rhizosphere and bulk soil (FDA, fluorescein
diacetate hydrolase; AE/AN, the relative abundance of
aerobes to anaerobes; GP/GN, the relative abundance of
Gram-positive bacteria to Gram-negative bacteria; F/B, the
relative abundance of fungi to bacteria; H, Shannon index.
DHA, dehydrogenase; CAT, catalase; PPO,
polyphenoloxidase; SUC, sucrase; URE, urease; PHOS,
alkaline phosphatase. BS, bulk soil; SS, soil in the
rhizosphere of S. triqueter, digitals reflect the sampling
month)
bulk soil. However, PHOS was only positively
correlated with OM content in the rhizosphere.
Moreover, CAT, PPO and SUC were positively
correlated with pH in the rhizosphere rather than in
bulk soil.
DHA, PPO and CAT are important
oxidoreductases and responsible for vital metabolic
processes including the decomposition and
detoxification of xenobiotics in the soil (Tang et al.,
2010; Zhang et al., 2009). As an important participant
in the natural recycle of phosphorus, phosphatase is
responsible for the hydrolysis of organically bound
phosphate into free ions; urease and sucrase are
responsible for the hydrolysis of urea and organic
carbohydrate, respectively (Wang et al., 2009).
Previous studies proved that plants can alleviate the
adverse effects of petroleum hydrocarbons on soil
enzymes, however, the relationships of soil enzymes
and microbial community structure were firstly
explored in this paper (Wang et al., 2009;
Wyszkowska and Wyszkowski, 2010; Zhang et al.,
2011).
4. Conclusions

Therefore, soil microbes can indirectly
influence the degradation of petroleum by regulating
activities of soil enzymes. Coefficients of Pearson's
correlation between soil characteristics and enzymes
were listed in Table 5. In both types of soil, CAT, PPO
and SUC were positively related with soil
temperature. PHOS was positively correlated with soil
In this study, rhizosphere effects on soil
properties, microbial community structure and soil
enzymes, as well as their interactions, were explored
at the HYRE wetland for the first time. Characterized
with higher microbial activity and diversity, the
microbial community structure in the rhizosphere was
quite different from that in the bulk soil.

1707
 Wei et al./Environmental Engineering and Management Journal 17 (2018), 7, 1701-1709

Table 5. Coefficients of Pearson's correlation between soil characteristics and enzymes

T TN TP OM pH TPH TDS
Bulk
CAT 0.782** 0.369 0.249 0.366 0.378 -0.378 -0.093
PPO 0.855** 0.486 0.407 0.674* 0.382 -0.409 0.097
SUC 0.589* -0.193 0.166 0.675* 0.531 -0.313 0.563
URE -0.023 0.133 0.035 -0.074 0.270 0.247 -0.218
PHOS 0.643* 0.843** 0.404 0.031 -0.264 0.135 -0.585*
DHA 0.638* 0.590* 0.480 -0.157 -0.137 -0.030 -0.293
Rhizosphere
CAT 0.743** 0.636* 0.244 0.290 0.772** -0.240 0.033
PPO 0.893** 0.894** 0.446 0.618* 0.652* -0.013 0.084
SUC 0.662* 0.416 0.255 0.206 0.677* -0.425 0.174
URE 0.486 0.365 0.351 0.230 0.502 0.103 -0.041
PHOS 0.452 0.462 0.266 0.696* 0.044 0.479 -0.180
DHA 0.499 0.435 0.205 0.419 0.551 0.142 0.199
Asterisks mark different levels of significance: *P<0.05, **P<0.01. DHA, dehydrogenase; CAT, catalase; PPO, polyphenoloxidase; SUC, sucrase;
URE, urease; PHOS, alkaline phosphatase. T, soil temprature; TN, total nitrogen; TP, total phosphorus; OM, organic matter; TPH, total petroleum
hydrocarbon; TDS, total dissolved solid

Positively correlated with H and the relative
abundance of F/B, activities of DHA, CAT, PPO in the
bulk soil were significantly lower (P < 0.05) than in
the rhizosphere. The microbial structure and activities
of soil enzymes were greatly affected by OM, TN, TP
and pH.
Acknowledgements
The work was funded by the National Natural Science
Foundation of China (Nos.41373097, 41073072, 41101230,
41203051), China Postdoctoral Science Foundation funded
project (No.2013M541506), Program for Innovative
Research Team in University (No.IRT13078).
References
Baath E., Anderson T.H., (2003), Comparison of soil
fungal/bacterial ratios in a pH gradient using
physiological and PLFA-based techniques, Soil Biology
& Biochemistry, 35, 955-963.
Bertin C., Yang X.H., Weston L.A., (2003), The role of root
exudates and allelochemicals in the rhizosphere, Plant
and Soil, 256, 67-83.
Boerner R.E.J., Brinkman J.A., Smith A., (2005), Seasonal
variations in enzyme activity and organic carbon in soil
of a burned and unburned hardwood forest, Soil Biology
and Biochemistry, 37, 1419-1426.
Cao Z., Liu X., Zhang X., Chen L., Liu S., Hu Y., (2012),
Short-term effects of diesel fuel on rhizosphere
microbial community structure of native plants in
Yangtze estuarine wetland, Environmental Science and
Pollution Research International, 19, 2179-2185.
Ciancio A., Colagiero M., Pentimone I., Rosso L., (2016),
Soil microbial communities and their potential for root-
knot nematodes management: a review, Environmental
Engineering and Management Journal, 15, 1833-1839.
Cowie B.R., Greenberg B.M., Slater G.F., (2010),
Determination of microbial carbon sources and cycling
during remediation of petroleum hydrocarbon impacted
soil using natural abundance C-14 analysis of PLFA,
Environmental Science & Technology, 44, 2322-2327.
Fierer N., Jackson R.B., (2006), The diversity and
biogeography of soil bacterial communities,
Proceedings of the National Academy of Sciences of the
United States of America, 103, 626-631.
Frostegård Å., Tunlid A., Bååth E., (1993), Phospholipid
fatty acid composition, biomass, and activity of
microbial communities from two soil types
experimentally exposed to different heavy metals,
Applied and Environmental Microbiology, 59, 3605-
3617.
Frostegård Å., Tunlid A., Bååth E., (2011), Use and misuse
of PLFA measurements in soils, Soil Biology and
Biochemistry, 43, 1621-1625.
Gerhardt, K.E., Huang, X.D., Glick, B.R., et al., (2009),
Phytoremediation and rhizoremediation of organic soil
contaminants: Potential and challenges, Plant Science,
176, 20-30.
Green V.S., Stott D.E., Diack M., (2006), Assay for
fluorescein diacetate hydrolytic activity: Optimization
for soil samples, Soil Biology & Biochemistry, 38, 693-
701.
Guan S., (1986), Soil Enzymes and Research Methods (in
Chinese), China Agricultural Science Press, Beijing,
China.
Hinsinger P., (2001), Bioavailability of soil inorganic p in
the rhizosphere as affected by root-induced chemical
changes: A review, Plant and Soil, 237, 173-195.
Johnson, J.L., Temple, K.L., (1964), Some variables
affecting the measurement of “catalase activity” in soil,
Soil Science Society of America Journal, 28, 207-209.
Kaur A., Chaudhary A., Choudhary R., Kaushik R., (2005),
Phospholipid fatty acid-A bioindicator of environment
monitoring and assessment in soil ecosystem, Current
Science, 89, 1103-1112.
Kirk J.L., Klironomos J.N., Lee H., Trevors J.T., (2005),
The effects of perennial ryegrass and alfalfa on
microbial abundance and diversity in petroleum
contaminated soil, Environmental Pollution, 133, 455-
465.
Li M., Zhou Q., Tao M., Wang Y., Jiang L., Wu Z., (2010),
Comparative study of microbial community structure in
different filter media of constructed wetland, Journal of
Environmental Sciences, 22, 127-133.
Liu X., Herbert S.J., (2002), Fifteen years of research
examining cultivation of continuous soybean in
northeast china: A review, Field Crops Research, 79, 1-
7.
Liu W., Luo Y., Teng Y., Li Z., Ma L.Q., (2010a),
Bioremediation of oily sludge-contaminated soil by

1708
 Contrastive soil properties, microbial structure and soil enzymes in the rhizosphere
stimulating indigenous microbes, Environmental
Geochemistry and Health, 32, 23-29.
Liu X., Zhong C., Wang Z., Zhang X., Wang J., (2010b),
Distribution characteristics of oily contaminant in
Wusong estuary sediments, Journal of Shanghai
University, 16, 621-625.
Liu X., Wang Z., Zhang X., Wang J., Xu G., Cao Z., Zhong
C., Su P., (2011), Degradation of diesel-originated
pollutants in wetlands by Scirpus triqueter and
microorganisms, Ecotoxicology and Environmental
Safety, 74, 1967-1972.
Meng L., Zhu Y.G., (2010), Pyrene biodegradation in an
industrial soil exposed to simulated rhizodeposition:
How does it affect functional microbial abundance?,
Environmental Science & Technology, 45, 1579-1585.
Mihalache G., Zamfirache M.M., Hamburda S., Stoleru V.,
Munteanu N., Stefan M., (2016), Synergistic effect of
Pseudomonas lini and Bacillus pumilus on runner bean
growth enhancement, Environmental Engineering and
Management Journal, 15, 1823-1831.
Mishra S., Jyot J., Kuhad R.C., Lal B., (2001), In situ
bioremediation potential of an oily sludge-degrading
bacterial consortium, Current Microbiology, 43, 328-
335.
Nannipieri P., Ascher J., Ceccherini M.T., Landi L.,
Pietramellara G., Renella G., (2003), Microbial
diversity and soil functions, European Journal of Soil
Science, 54, 655-670.
Nie M., Zhang X.D., Wang J.Q., Jiang L.F., Yang J., Quan
Z.X., Cui X.H., Fang C.M., Li B., (2009), Rhizosphere
effects on soil bacterial abundance and diversity in the
Yellow river deltaic ecosystem as influenced by
petroleum contamination and soil salinization, Soil
Biology & Biochemistry, 41, 2535-2542.
Richardson A.E., Barea J.M., Mcneill A.M., Prigent-
Combaret C., (2009), Acquisition of phosphorus and
nitrogen in the rhizosphere and plant growth promotion
by microorganisms, Plant and Soil, 321, 305-339.
Rukun L., (1999), Analysis Methods of Soil Agricultural
Chemistry (in Chinese), Agricultural Science and
Technology Press, Beijing, China.
Schinner, F., Mersi von W., (1990), Xylanase-, CM-
cellulase- and invertase activity in soil: An improved
method, Soil Biology and Biochemistry, 22, 511-515.
Smalla K., Wieland G., Buchner A., Zock A., Parzy J.,
Kaiser S., Roskot N., Heuer H., Berg G., (2001), Bulk
and rhizosphere soil bacterial communities studied by
denaturing gradient gel electrophoresis: Plant-
dependent enrichment and seasonal shifts revealed,
Applied and Environmental Microbiology, 67, 4742-
4751.
Sun T.R., Cang L., Wang Q.Y., Zhou D.M., Cheng J.M., Xu
H., (2010), Roles of abiotic losses, microbes, plant
roots, and root exudates on phytoremediation of PAHs
in a barren soil, Journal of Hazardous Materials, 176,
919-925.
Tang J.C., Wang R.G., Niu X.W., Zhou Q.X., (2010),
Enhancement of soil petroleum remediation by using a
combination of ryegrass (lolium perenne) and different
microorganisms, Soil & Tillage Research, 110, 87-93.
Toyama T., Furukawa T., Maeda N., Inoue D., Sei K., Mori
K., Kikuchi S., Ike M., (2011), Accelerated
biodegradation of pyrene and benzo[a]pyrene in the
Phragmites australis rhizosphere by bacteria-root
exudate interactions, Water Research, 45, 1629-1638.
Velasco A.V., Probanza A., Mañero F.J., Solano B., Lucas
J., (2010), Characterization of the rhizosphere
microbial community from different Arabidopsis
thaliana genotypes using phospholipid fatty acids
(PLFA) analysis, Plant and Soil, 329, 315-325.
Wang Q.Y., Zhou D.M., Cang L., (2009), Microbial and
enzyme properties of apple orchard soil as affected by
long-term application of copper fungicide, Soil Biology
and Biochemistry, 41, 1504-1509.
Wei J., Liu X., Wang Q., Chen X., Li H., Li X., Deng K.,
(2014), Different biomass allocation, soil enzyme
activities and microbial characteristics between diesel-
degrading plants, Clean-Soil, Air, Water, 42, 1765-
1770.
Wei J., Liu X., Wang Q., Wang C., Chen X., Li H., (2014),
Effect of rhizodeposition on pyrene bioaccessibility and
microbial structure in pyrene and pyrene-lead polluted
soil, Chemosphere, 97, 92-97.
Wick L.Y., Buchholz F., Fetzer I., Kleinsteuber S., Härtig
C., Shi L., Miltner A., Harms H., Pucci G.N., (2010),
Responses of soil microbial communities to weak
electric fields, Science of the Total Environment, 408,
4886-4893.
Wyszkowska J., Wyszkowski M., (2010), Activity of soil
dehydrogenases, urease, and acid and alkaline
phosphatases in soil polluted with petroleum, Journal
of Toxicology and Environmental Health-Part a-
Current Issues, 73, 1202-1210.
Yao H., Wu F., (2010), Soil microbial community structure
in cucumber rhizosphere of different resistance
cultivars to Fusarium wilt, FEMS microbiology
ecology, 72, 456-463.
Zhang C., Xu J., Liu X., Dong F., Kong Z., Sheng Y., Zheng
Y., (2010), Impact of imazethapyr on the microbial
community structure in agricultural soils,
Chemosphere, 81, 800-806.
Zhang X., Liu X., Liu S., Liu F., Chen L., Xu G., Zhong C.,
Su P., Cao Z., (2011), Responses of Scirpus triqueter,
soil enzymes and microbial community during
phytoremediation of pyrene contaminated soil in
simulated wetland, Journal of Hazardous Materials,
193, 45-51.
Zhang X., Liu X., Zhong C., Cao Z., Liu F., Chen L., Liu S.,
Hu Y., (2012), Soil microbial community response to
pyrene at the presence of Scirpus triqueter, European
Journal of Soil Biology, 50, 44-50.
Zhang X., Wang Z., Liu X., Hu X., Liang X., Hu Y., (2013),
Degradation of diesel pollutants in Huangpu-Yangtze
river estuary wetland using plant-microbe systems,
International Biodeterioration & Biodegradation, 76,
71-75.
Zhang Z.Z., Su S.M., Luo Y.J., Lu M., (2009), Improvement
of natural microbial remediation of petroleum-polluted
soil using graminaceous plants, Water Science and
Technology, 59, 1025-1035.
Zhou X., Yu G., Wu F., (2011), Effects of intercropping
cucumber with onion or garlic on soil enzyme activities,
microbial communities and cucumber yield, European
Journal of Soil Biology, 47, 279-287.
Zornoza R., Guerrero C., Mataix-Solera J., Scow K.M.,
Arcenegui V., Mataix-Beneyto J., (2009), Changes in
soil microbial community structure following the
abandonment of agricultural terraces in mountainous
areas of eastern spain, Applied Soil Ecology, 42, 315-
323.
1709
1710
Environmental Engineering and Management Journal July 2018, Vol.17, No. 7, 1711-1719
http://www.eemj.icpm.tuiasi.ro/; http://www.eemj.eu

“Gheorghe Asachi” Technical University of Iasi, Romania

STUDY OF NITROGEN FORMS IN SEASONAL DYNAMICS

AND KINETICS OF NITRIFICATION AND DENITRIFICATION
IN PRUT AND NISTRU RIVER WATERS

Petru Spataru1, Igor Povar1, Tudor Lupascu1∗, Alfredo Carlos Alder2, Elena Mosanu1
1Institute of Chemistry, Academy of Sciences of Moldova, 3 Academiei str., Chisinau, MD2028, Republic of Moldova
2Eawag, Environmental Chemistry, Eawag, Environmental Chemistry, 133 Uberlandstrasse str., P.O. Box 611,
Dubendorf, CH-8600, Switzerland

Abstract

The presence of nitrogen in its reduced forms serves as an indicator of natural aquatic system degradation. The excessive
concentrations of nitrogen in reduced forms (NH4+, NO2-) found in the sections of the the Shireutsi (of the Prut River) and the
Olaneshti (of the Nistru River) prove that these sites are areas to focus efforts on improving water quality. Laboratory simulations
indicate a good environmental buffer action of natural waters to potential pollution in Sculeni and Ungheni sections of the Prut
River and downstream of Naslavcea and Tiraspol (Nistru River), as well as in sites most affected by pollution with wastewater
from Frasineshti (of the Prut River) along with Lencautsi, Cosautsi and downstream of Bender (the Nistru River). Overcoming the
maximum allowable concentration (MAC) level for the nitrate ion in water from wells is an index of pollution of the land areas
where pollutants can leach into and pollute groundwater. Of the 12 wells from which water samples were collected over three years
(2010-2012), only two wells, in the Cuhneshti and Cobani villages showed MAC level compatible with drinking water quality.

Key words: Moldovan rivers, nitrogen forms, organic compounds, oxidation, seasonal dynamics

Received: January, 2013; Revised final: October, 2014; Accepted: October, 2014; Published in final edited form: July, 2018

1. Introduction The high pollution of wells is another huge

The issue of water as a whole, and specifically
for drinking water in the Republic of Moldova is
extremely important due to the phenomenon of aridity,
which has become more and more pronounced over
the last 5-7 years, from which one or two are usually
dried. In the most part of the country the average
return period of drought varies from 3 to 5 years. For
the period spanning from 2000 to 2012 the Republic
of Moldova has already experienced 4 years (2000,
2003, 2007, and 2012) with the devastating droughts
(Cazac, et al., 2014). In the southern part of the
country a significant number of small rivers in the
summer period remains without water and only a few
reservoirs are used.
concern, taking into account that wells are used as a
source of water supply for about 75% of the rural
population. Due to the increase in human activity,
intensive agriculture, especially in the livestock
sector, and the lack of ecological management, the
water quality of wells does not correspond to the
standards of "Drinking Water". This also regards the
water from the distribution systems throughout the
country. The ratio of the samples that exceeded the
sanitary-chemical parameters from the underground
centralized sources was 70.8% in 2009 and 67.8% in
2010. The most severe situation was recorded in the
districts of Anenii Noi, Faleshti, Glodeni, Hinceshti,
Causheni, Ciadyr-Lunga, Cimishlia, Comrat, Drochia,
Sangerey and the Baltsy municipality where this index

∗
Author to whom all correspondence should be addressed: e-mail: lupascut@gmail.com; Phone: +37322739736; Fax: +37322739736
 Spataru et al./Environmental Engineering and Management Journal 17 (2018), 7, 1711-1719
was between 85-100%. The absence of waste disposal
systems as well as basic measures of water sources
protection are the main causes of water quality
worsening in wells. The ratio of wells that exceeded
the sanitary-chemical parameters constituted 84.2% in
2010 (80.7% in 2009) (NRSERM, 2011).
The rivers Nistru and Prut are the largest in the
Republic of Moldova and their basins comprise the
entire territory of the country, form and store the most
significant amounts of water. The waters of Nistru and
Prut rivers meet the regulatory requirements for
"Drinking Water" main mineral parameters and
represent the greatest source of water for water
treatment plants. The quality of water in these rivers is
guaranteed mainly by self-purification processes that
take place in water ecosystems.
Nitrogen, being one of the basic elements in the
formation of organic substances, is among the most
sensitive elements to change in the natural water due
to eutrophication and pollution. The drinking water
MAC level is for NH4+ 0.39 mg/L and NO2- - 0.02
mg/L (CD, 1998; State Register, 2007). The study of
nitrogen mineral forms highlights the sustainability, as
well as the vulnerability of the aquatic environment.
Studies of the hydrochemistry and quality of the water
of these rivers have been published in national
specialized journals (Gladchi et al., 2008; Zubcov et
al., 2011). Findings show that, for a wide range of
bacteria, in contrast to the situation in animal
(Martinelle et al., 1993) and plant cells (Britto et al.,
2001, Britto et al., 2002), ammonia is not toxic, even
in molar concentrations. This happens because most
bacteria prefer ammonia as a nitrogen source; some
species even produce ammonium ion – for example,
N2-fixing Rhizobia Cyanobacteria and Proteolytic
Clostridia create ammonia through the fermentation
of amino acids. Resistance to ammonia is a common
property in terms of bacteria (Müller et al., 2006).
However, nitrites are highly toxic to humans, flora and
fauna, being an important concern regarding water
quality. They implicate also in the pathology of gastric
cancer (Brunning-Fann et al., 1993, Weng et al., 1992)
and are a possible cause of migraines (Bradberry et al.,
1994); they as well compromise the binding capacity
of oxygen in the blood and may result in respiratory
deficiencies of aquatic animals and humans (Blackall,
2000, Van Leeuwen, 2000). Understanding the
toxicity of reduced forms of nitrogen requires the
identification of methods of enhancing the oxidation
of nitrification processes. Furthermore, the pollutants
are frequently associated into ionogenic groups,
becoming less active, and their toxicity decreases and
the rate of oxidation/assimilation processes increase,
respectively. Otherwise, if pollutants are not
associated, their toxicity is higher and the processes of
self-purification are blocked (Lewis, 1992).
It is well-known that in the case of nitrification
and denitrification the nitrite ion is an intermediate
product. Both processes, oxidation of ammonia (first
stage of nitrification) and denitrification are generated
by specific enzymes, which cause the production of
1712
the same intermediate product - nitrite ion (Philips et
al., 2002). The nitrite ion concentration, as an
indicator of oxidation processes generated by the
enzymatic catalysis, is higher than the MAC value in
the abovementioned samples, as well as in such
localities near the Prut River, as the town of Leova and
the city of Cahul. In the composed sample from the
Valea Mare section, this concentration is higher than
the average value from the same section and period of
time, which denotes its authenticity due to
biochemical processes.
Consequently, the main goal of this paper was
to study the nitrogen-containing species in function of
seasons as well as kinetics of nitrification and de-
nitrification in the Prut and Nistru waters.
2. Methods
Within the Scientific Co-oPeration between
Eastern Europe and Switzerland (SCOPES) project
(see Acknowledgments), during the 2010-2012 years,
a large amount of aquatic samples were taken from
different sections of the Prut and Nistru rivers (see the
map, Fig.1), and analytical tests were carried out.
Sampling on the Prut River was started from the
border between the Republic of Moldova and Ukraine,
above the lake dam (Criva and Shireuti, placed after
the administrative centre Lipcani) and almost
immediately after the lake (Branishte). The next
sections were Sculeni and upstream Ungheni (32.5),
before wastewater treatment plant (WWTP), which
are less affected by human activity. Valea Mare is a
section after the WWTP Ungheni. The Frasinesti
section is downstream of the confluence with Jijia
crossing the urban centres Dorohoi (24.3) and Vladeni
and bringing through the Bahlui River polluted waters
from Iassy (321.6). Cotul Morii is a less affected area,
with reduced population. The Leova and Cahul cities
have respectively 11.4 and 40.9 thousands inhabitants.
The section Giurgiulesti is influenced by the Danube
waters due to frequent changes of its water level. In
this segment the Prut water is mixed with that of
Danube. The sampling sites of the Nistru River were
mostly chosen close to the towns and cities, as Soroca
(28.6), Camenca (23.2), Ribnita (73.6) and Rezina
(10.2) on its middle and Dubasari (35.2), Bender
(99.3), Tiraspol (146.6) and Chisinau (804) on its
bottom part.
The oxygen fixing and laboratory simulation
initiating were started along with sampling processes.
Under real conditions many investigations are
impossible or extremely difficult to accomplish. At the
same time, laboratory simulations offer the
opportunity to study the process at the desired stage
and analyze the degree of influence of a series of
chemical and physicochemical parameters, as the
concentration of different components, redox
potential, temperature etc. Preliminary tests were
made for the main mineral parameters: (Ca2+, Mg2+,
NH4+, NO2-, NO3-, SO42+, HCO3-, Cl-), pH and
dissolved oxygen, BOD5, COO-Cr.
 Study of nitrogen forms in seasonal dynamics and kinetics of nitrification and denitrification in Prut and Nistru river waters
Fig.1. The map of the investigated sites, denoted by read squares
The natural water tests were carried out
according to ISO methods (ISO 7150-1, 2001; SR ISO
7890-3, 1988; ISO 8466-1, 1990; SR ISO 7890-3,
2000; SR EN 26777, 2006; Sandu et al., 1989).
Laboratory simultations were carried out respecting
the minimum recommended of water amount for the
tested sample in glass bottles (Matveeva et al., (1988).
In each sample a solution of NH4Cl was added in
order to obtain an initial concentration of 2 mg/L of
ammonium. The tested water samples were kept in
natural light conditions and spared from direct rays.
Composed samples over a period of 20 days were also
taken and studied. Test methods were carried out using
HACH Company spectrophotometer (2004).
3. Results and discussion
3.1. Analysis of experimental data for the Prut River
In the first year of our studies (2010)
ammonium ion concentrations that exceeded the MAC
("Drinking Water") were detected in some sampling
sections. Ammonium produced by the decomposition
of organic substances affects mostly small rivers. It
also has an evident impact on the larger rivers.
For instance, the ammonium ion concentration
in the spring water samples of the Camenca River
(near the Balatina village) exceeded 2 MAC; for the
Lapusna River (in the summer period) exceeded 1.1
MAC; for the Prut River (the Sculeni village) in
summer, the Valea Mare village in summer and in the
Leova town in fall overcome the MAC by 1.5, 1.4 and
1.2 times respectively.
The obtained results were compared with data
obtained by the State Hydrometeorological Service,
HydroMeteo (MB, 2011). The samples were collected
monthly, so in total for each section in two years 24
samples were collected. The comparative analysis of
data obtained by HydroMeteo and our laboratory
simulations reveals the strength of natural waters to
return to the required standard levels for drinking
water and provides a deeper understanding about
changes in the aquatic environment. The data on
exceeded MAC values for ammonium and nitrite in
the period of 2011-2012 yrs denote the changes that
take place in considered cases in the course of the
middle and bottom of the Prut River (Fig.2), that were
reported by the State Hydrometeorological Service
(MB, 2011).
1713
 Spataru et al./Environmental Engineering and Management Journal 17 (2018), 7, 1711-1719

[NO2-]
Giurgiulesti [NH4+]

Cahul

Leova
Cotul Morii
Valea Mare
Ungheni
Branesti
Siriuti
Criva
0 1 2 3 4 5 6
Number of exceeded MAC

Fig. 2. The number of samples in which the concentrations of NH4+ and NO2- overcome the MAC values
in the middle and bottom sections of the Prut River

Table 1. The concentrations of NH4+, NO2- and NO3- in samples taken from the middle and lower sections
of the Prut River in 2011

Sampling site/month May November May November May November

Sections of the Prut River [NH4+], mgN/L MAC=0.39 [NO2-], mgN/L MAC=0.02 [NO3-], mg/L
village Sculeni 0.16 0.006 0.46
city Ungheni 0.38 0.2 0.018 0.009 1.5 0.48
village Macaresti 0.59 0.026 1.9
village Valea Mare after the Jijia 0.19 0.003 0.97
River
after the Leova town 0.48 0.026 1.95
village Stoianovca 0.28 0.017 1.91
village Branza after the Cahul city 0.55 0.13 0.063 0.007 1.88 0.86

Table 2. The concentrations of NH4+, NO2-, NO3-, O2 and BOD5 in samples taken from the middle and lower sections
of the Prut River in 2012

Sampling site/concentration [NH4+] [NO2] BOD5 [O2]

[NO3-] mg/L
Sections of the Prut River mgN/L mgN/L mgO/L (NO2-) mg/L
village Suleni 0.14 0.038 2.1 0.97(0.126) 9.53
city Ungeni after WWTP 0.21 0.038 3.2 0.82 (0.086) 9.41
village Valea Mare after Jijia 0.16 0.038 4.1 0.80(0.016) 9.09
After the Leova town 0.22 0.023 3.2 0.79(0.257) 9.11
village Branza after the Cahul city 0.3 0.046 2.9 0.86(0.211) 9.07

The dynamics of the accident cases (that
exceeded the MAC values for NH4+ and NO2-)
demonstrates an evident influence of anthropic impact
on the aquatic environment, at the border with Ukraine
in the North (up to Lake Costesti-Stanca) and the
Branishte section, which is downstream of this
reservoir. In the South (the Giurgiuleshti section), the
water quality of the River Prut is impacted by the
Danube River (Teodoru et al., 2007). The processes of
assimilation and oxidation of organic matter are
significantly more intense upstream of Lake Costesti -
Stanca (the Criva and Shireuti sections) as well as in
the Leova section; while in the Giurgiuleshti section
they are less obvious. In other words, the more
polluted is area, the less intense are the processes of
assimilation and oxidation of organic matter.
The study of the nitrogen-containing species in
waters of the Prut River in the years 2011-2012
highlights the values that exceed the MAC ("drinking
water"), especially in the Branza village, which
overran the MAC for the ammonium ion concentration
in 2011, as well as for the nitrite ion in 2011 and 2012
yrs (Tables 1 and 2).
The most significant decompositions and
oxidation of organic material in waters are more
evident in the Prut River, especially in the Branza
section downstream of the Cahul city.
3.2. Laboratory kinetic simulation for the Prut waters
Using the laboratory simulations, the water
samples were studied from the following Prut River
sections: the Sculeni village, the Ungheni city
(downstream of WWTP), the Valea Mare village
(downstream of the Jijia River), downstream of the
Leova town and the Branza village. During a period of
8 to 13 days, the ammonia oxidation reached a level of
about 2 mg/L (Fig. 3).

1714
 Study of nitrogen forms in seasonal dynamics and kinetics of nitrification and denitrification in Prut and Nistru river waters

a
3,0
Prut, Sculeni
Prut, Sculeni b
[NH4+],mg/L

4.5 Prut, Ungheni (after WWTP)

2,5 Prut, Ungheni (after WWTP) Prut, Frasinesti
Prut, Frasinesti Prut, after Leova

[NO2-], mg/L
4.0
Prut, after Leova Prut, Branza
2,0 Prut, Branza 3.5

3.0

1,5 2.5

2.0
1,0
1.5

1.0
0,5
0.5

0,0 0.0
0 2 4 6 8 10 12 14 0 5 10 15 20 25 30
Days Days

Fig. 3. The kinetics of the NH4+ (a) and NO2- (b) oxidation for the samples from the following Prut River sections: 1 – the Sculeni
village, 2 – the Ungheni city (downstream of WWTP), 3 –the Valea Mare village (downstream of the Jijia River), 4 - downstream
of the Leova town and 5 –the Branza village

The obtained data of inorganic forms of (ammonium, nitrites) were recorded in the North,
nitrogen are comparable to those of the Yangtze River which was caused by the Novodnestrovsc Dam Lake,
in which there is a noticeable human impact (Muller where the processes of sedimentation and sewage
et al., 2008) and are overcome by the eutrophic water causing an "inertia" of indicators took place, at the
of the Haihe River, from a semi-arid zone of China same time as the Dam Lake Stanca-Costesti gave a
(Pernet-Coudrier et al., 2012). In sections of water stability of the nitrogen forms in the water of the Prut
downstream of Leova and Branza, the ammonia River. The water quality of the Nistru River is affected
oxidation, assimilation of organic compounds, as well especially at the Olaneshti section, and to a lesser
as undergoing transformations under the influence of extent in the sections of Soroca, Dubasari, Vadul-lui-
natural enzymes produced by micro-organisms, Voda and Bender. In the Bik River (tributary of
proceeded very fast. Dniester), downstream of Chisinau, the ammonium
It is known that the numeric rising of concentration is found to be about 10-40 mg/L while
heterotrophic populations in natural water (Daum et at the confluence with the River Nistru ammonia
al., 1998) is coupled with the increase of nitrite exceeds 50mg/L. Formation of ammonium in
concentration (Spataru, 2011; Lin et al., 2010). The wastewaters is mainly supposed to be due to the
samples least prepared for pollution are from the decomposition of organic matter from foodborne/
Sculeni and Ungheni sections. The full process of digestion products containing proteins and urea.
assimilation /oxidation of the reduced nitrogen and
carbon forms for Frasineshti, however, is the longest,
that according to the BOD5 index and the
concentration of nitrites in the destructible sample in
the dark, demonstrates that water sample from the Prut
- Valea Mare, along with the highest content of
organic substance, containts xenobionts, that break the
oxidation of carbon and reduced nitrogen forms. The
complex compounds, which are the ammonia
associated with a wide range of organic substances
(humic complexes), can reduce the toxic effect of
ammonia (for example, in the Leova and Branza
samples). In the case of the water sample from the Prut
– (Valea Mare section), the toxic effect of ammonia is
diminished. Instead, a large amount of organic matter
(in comparison with the rest of the samples) along with Fig. 4. The number of samples in which the concentrations
the presence of toxic pollutants causes a delay of of NH4+ and NO2- exceeded the MAC values in waters of
purification processes (Philips et al., 2002). the Nistru River

3.3. Analysis of experimental data for the Nistru River As one can see from Fig. 4, the anthropogenic
accumulations lead to the increase of the exceeded
Fig. 4 illustrates that for the Nistru River the MAC cases, that is a curve with peaks, which
lowest values of these parameters for accidental cases amplitude is growing to inferior sections, culminating

1715
 Spataru et al./Environmental Engineering and Management Journal 17 (2018), 7, 1711-1719
in the Olaneshti one, and then returns to the normal
shape up to the Palanca section.
A special case is the pollution of the Dubasari
section. Upstream of this is the Rezina section (Fig. 4),
where a minimum frequency of prevailing MAC was
found for both ammonia and nitrites. Up to the
Dubasari section, where many cases of the MAC
overruns for nitrites have been found, there are not
urban centers which would so strongly influence the
aquatic environment. It appears that the impact on this
segment is from the secondary pollution of lake
sediments. The secondary pollution is called the
pollution that takes place at the interface “water-
sediment”, caused by underwater deposits, as
pesticides and heavy metals (Linnik and Zubenko,
2000; Zubcov et al., 2010). The Lake Stanca-Costeshti
is younger than Lake Dubasari, resulting in a lack of
evident impact of the secondary pollution in the
accumulation lake of the Prut River. Therefore, in this
case the secondary pollution is a major factor that
leads to the decrease of sustainability of the water
treatment system on a segment of the Nistru River.
3.4. Laboratory kinetic simulations for the Prut waters
The laboratory kinetic simulation from 2011
demonstrates that in the sample, taken from the
Naslavcea section, the oxidation process occurred
most quickly (8 days). In the previous publication
(Spataru, 2011) there were presented data on solved
oxygen in this section that was lower than in the
downstream sections, while the value of BOD5 was
higher in comparison with the same sections.
So, it is found that increasing the oxygen
concentration has an opposite effect. This result can be
explained by the following evidence. The value of
BOD5 in the Naslavcea village is higher in comparison
with the values of this parameter in the downstream
sections, since enzyme activity is greater. The
oxidation of ammonia, nitrates and organic substances
is related to the fact that the water spilled out of the
lake dam comes from the depth zone of thermocline,
where the activity of micro-organisms and
biochemical activity is maximum, compared to the
epilimnion and hypolimnion layers of bulk water
a b
2.5
[NO2-], mg/L
2.0
1.5
1.0
0.5
0.0
0 2 4 6 8 10
Days
Fig. 5. The kinetics of the NH4+ ion oxidation in the samples from the Nistru River: (a) represents the sections upstream of the
Bac River; (b) represents the sections downstream of the Tiraspol city
1716
(Prokopkin, 2010). The change of the hydrochemical
regime is also a phenomenon with an important impact
(Mengis et al., 1997).
In the Lencauti and Cosauti sections the
oxidation takes place more slowly, which shows
considerable biochemical diminishes compared to the
Naslavcea village. It can be assumed that this area is
one of the more vulnerable in the case of the pollution
treatment. In the upstream section of the Bac,
downstream of the Bender city, the BOD5 values are
larger and growing compared to the sample from the
Cosauti village, however the loads of organic matter
are much higher. This fact leads to the equalization of
the oxidation time of ammonia in these samples which
are different as regards the composition of organic
matter and the nitrogen mineral forms from water.
As one can see from Fig. 4, the anthropogenic
accumulations lead to the increase of the exceeded
MAC cases, that is a curve with peaks, which
amplitude is growing to inferior sections, culminating
in the Olaneshti one, and then returns to the normal
shape up to the Palanca section.
A special case is the pollution of the Dubasari
section. Upstream of this is the Rezina section (Fig. 4)
where a minimum frequency of prevailing MAC was
found for both ammonia and nitrites. Up to the
Dubasari section, where many cases of the MAC
overruns for nitrites have been found, there are not
urban centers which would so strongly influence the
aquatic environment. It appears that the impact on this
segment is from the secondary pollution of lake
sediments. The Lake Stanca-Costeshti is younger than
Lake Dubasari, resulting in a lack of evident impact of
the secondary pollution in the accumulation lake of the
Prut River. This fact demonstrates the decrease of
sustainability of the water treatment system on a
segment of the the Nistru River.
The laboratory kinetic simulation from 2011
proves that in the sample taken from the Naslavcea
section, the oxidation process occurred most quickly
(8 days). In the previous publication (Spataru, 2011)
there were presented data on solved oxygen in this
section that was lower than in the downstream
sections, while the value of BOD5 is higher in
comparison with the same sections.
Nistru, Naslavcea 2.5
Nistru upstream of debus. Bac r.
[NO2-], mg/L
Nistru, Lencauti Nistru downstream Bender sit.
Nistru, Cosauti
2.0 Nistru downstream Tiraspol sit.
1.5
1.0
0.5
0.0
12 0 2 4 6 8 10 12 14 16 18
Days
 Study of nitrogen forms in seasonal dynamics and kinetics of nitrification and denitrification in Prut and Nistru river waters

a b
7
Nistru upstream of debus. Bac r.
[NO2-], mg/L 8 Nistru downstream Bender sit.
6

[NO2-], mg/L
Nistru, Naslavcea Nistru downstream Tiraspol sit.
7
Nistru, Lencauti
5 Nistru, Cosauti 6

4 5
4
3
3
2 2

1 1
0
0
0 2 4 6 8 10 12 14 16 18 20 22 0 2 4 6 8 10 12 14 16 18 20
Days Days

Fig. 6. The kinetics of oxidation of NO2- -ions in water samples taken from the Nistru River: (a) represents the sections upstream
of the Bac River; (b) represents the sections downstream of the Tiraspol city

As one can see from Fig. 5a, 5b and 6a, 6b the
oxidation of nitrites is more rapidly, although the
ammonia oxidation is almost the same as in the
samples from the Nistru River in the sections upstream
of the Bac River, downstream of the Tiraspol city,
compared with the Lencauti and Cosauti sections. This
process keeps its high speed, even if in the lower
sections of the Nistru River the organic load is greater.
This fact confirms our earlier finding that the time of
oxidation of all reduced nitrogen species depends
proportionally on the organic matter amounts in
natural waters (Spataru, 2011).
The diagram on Fig. 7, with all the accidental
cases, demonstrates that the impact of the pollution is
more evident in the winter and summer periods. The
climate change mitigation the impact occurred in the
spring, compared with the winter and summer. Due to
the pollution of both wastewater and washing of land
by the pluvial waters the River Prut becomes more
affected. These situations can cause low dissolved
oxygen concentrations due to the intensification of
processes of oxidation and destruction, especially in
the sections with an increased concentration of
nitrites. The study of the samples of water from wells
on the left side of the Prut River showed an obvious
pollution with the nitrate ion (Table 3). Of 12 wells
from which the water samples were collected within
three years (2010-2012), in only two wells (in the
Cuhnesti and Cobani villages) the water corresponds
to the MAC level (for drinking water). Three of these
contain nitrates over MAC, two contain nitrates over
2 MAC, and 5 of these contain over 3 and up to 13
MAC of nitrites. In dry seasons many wells remain
without water, leaving the population of villages
without any water supply sources.
4. Conclusions
Most of the main mineral parameters of the
rivers Prut and Nistru waters, except ammonia and
nitrates, correspond to the European statutory norms
of "Drinking Water".

Number of excedences of MAC of [NO2-]

Summary number of NO2- of MAC
Sum of concentration NO2-, mg/L
9

0
Winter Spring Summer Autumn
Seasons

Fig. 7. The dynamics of the number of accidental cases (the exceeded MAC values) in the 2010-2012 yrs
in the Prut and Nistru rivers (recorded by the State Hydrometeorological Service (MB, 2011))

1717
 Spataru et al./Environmental Engineering and Management Journal 17 (2018), 7, 1711-1719

Table 3. Nitrogen-containing species concentrations in groundwater

[N- [N-NO2-] [N-NO3]

[N- [N-NO3-
NH4+] mg/L [N-NO2-] mg/L
No Sampling site/concentration NH4+] ]
mg/L mg/L
mg/L mg/L
MAC MAC MAC
1 Cobani village, dug well 0.08 0.09 0.7
Cobani well at the piquet frontier
2 0.00 0.00 12.4
guard, dug well
3 Cuhneshti, dug well, grab sample 0.00 0.00 4.6
4 Bisericani, dug well, grab sample 0.00 0.00 9.2
5 Sirma, dug well, grab sample 0.00 0.00 9.7
6 Macareshti, well tube 0.00 0.02 0.00 55.8
0.39 9
7 Macareshti, dug well (1) 0.00 0.00 62.2
8 Macareshti, dug well (2) 0.00 0.00 32.5
9 Frasineshti, dug well 0.00 0.00 19.8
10 Aqueduct Branza 0.00 0.00 1.1
11 Colibashi, Dug well 0.00 0.00 83.7
12 Branza, dug well 0.00 0.00 116.5
13 Semeni, dug well 0.00 0.00 19.6

The dynamics of nitrogen mineral forms shows
the impact of urban activities, where the stations for
wastewater treatment do not work satisfactorily. It is
necessary to improve the environmental management
upstream of the border with the Ukraine for the Prut
River and upstream of the Olaneshti section in the case
of the Nistru River. Also the secondary pollution
caused by sediments of the Lake dam Dubasari should
be taken into account.
The water in the wells on the east side of the
Prut River is polluted with nitrates and disappears in
the most wells in dry periods. It is compulsory to run
a series of projects, in order to fill the lack of drinking
water in the abovementioned villages.
Acknowledgments
This work has been supported by the international program
SCOPES 2009–2012 Swiss National Science Foundation
Joint Research Project “Xenobiotic Input to the Prut River
(XENOPRUT)”.
References
Blackall L.L., (2000), A Summary of Recent Microbial
Discoveries in Biological Nutrient Removal from
Wastewater, Australasian Biotechnology, 10, 29-32.
Bradberry S.M., Gazzard B., Vale J.A., (1994),
Methemoglobinemia caused by the accidental
contamination of drinking water with sodium nitrite,
Clinical Toxicology, 32, 173-178.
Britto D.T., Siddiqi M.Y., Glass A.D.M., Kronzucker H.J.,
(2001), Futile Transmembrane NH4 + Cycling: A
Cellular Hypothesis to Explain Ammonium Toxicity in
Plants, Proc. of the National Academy of Sciences,
USA, vol. 98, 4255-4258.
Britto D.T.; Konzucker H.J., (2002), NH4+ toxicity in higher
plants: a critical review, Journal of Plant Physiology,
159, 567-584.
Brunning-Fann C.S., Kaneene J.B., (1993), The effects of
nitrate, nitrite, and N-nitroso compounds on human
health, Veterinary and Human Toxicology, 35, 521-
538.
Cazac V., Ceres V., Gherghelegiu V., (2014), Water
Resources in the Territory of the Republic of Moldavia,
L.P. Estetini, Kishinev.
CD, (1998), Council Directive 98/83/EC on the quality of
water intended for human consumption Official Journal
of the European Communities, L330, 32-43.
Daum M., Zimmer W., Papen H., Kloos K., Nawrath K,
Bothe H., (1998), Physiological and molecular
biological characterization of ammonia oxidation of the
heterotrophic nitrifier Pseudomonas putida, Current
Microbiology, 37, 281-288.
Gladchi V., Goreaceva N., Duca G., Bunduchi E., Borodaev
R., Mardari I., Romanciuc L., (2008), Chemical
composition and redox state of water from the medial
Nistru River, Environment (Mediul Ambiant), 39, 20-
28.
ISO 8466-1, (1990), Water quality - Calibration and
evaluation of analytical methods and estimation of
performance characteristics. Part 1: Statistical
evaluation of the linear calibration function, On line at:
https://www.iso.org/standard/15664.html.
Lewis M.A., (1992), The effects of mixtures and other
environmental modifying factors on the toicities of
surfactants to freshwater and marine life, Water
Reserarch, 26, 1013-1023.
Lin Y., Kong H., Wu D., Li C., Wang R., Tanaka S., (2010),
Physiological and molecular biological characteristics
of heterotrophic ammonia oxidation by Bacillus sp. LY.
World Journal of Microbiology and Biotechnology, 26,
1605-1612.
Linnik P., Zubenko I., (2000), Role of bottom sediments in
the secondary pollution of aquatic environments by
heavy-metal compounds, Lakes & Reservoirs:
Research & Management, 5, 11-21.
Martinelle K., Häggström L., (1993), Mechanisms of
ammonia and ammonium ion toxicity in animal cells:
Transport across cell membranes, Journal of
Biotechnology, 30, 339-350.
Matveeva N.P., Klimenko O.A., Trunov N.M., (1988),
Simulation of Self-Purification of Natural Treatment of
Organic Pollutants in the Laboratory,
Gidrometeoizdat, Leningrad, Russia.
Mengis M., Gachter R., Wehrli B., (1997), Nitrogen
elimination in two deep eutrophic lakes, Limnology and
Oceanography, 42, 1530-1543.

1718
 Study of nitrogen forms in seasonal dynamics and kinetics of nitrification and denitrification in Prut and Nistru river waters
MB, (2011), Monthly Bulletin, Ambient Quality in Republic
of Moldova, Chisinau, On line at: www. meteo.md.
Muller B., Berg M., Yao Z.P., Zhang X.F., Wang D., Pfluger
A., (2008), How polluted is the Yangtze river? Water
quality downstream from the Three Gorges Dam,
Science of the total environment, 402, 232-247.
Müller T., Walter B., Wirtz A., Burkovski A., (2006),
Ammonium Toxicity in Bacteria, Current
Microbiology, 52, 400-406.
NRSERM, (2011), National Report on the State of the
Environment in the Republic of Moldova 2007–2010,
Chisinau, 44, On line at: www.mediu.gov.md.
Pernet-Coudrier B., Qi W., Liu H., Muller B., Berg M.,
(2012), Sources and pathways of nutrients in the semi-
arid region of Beijing−Tianjin, China, Environmental
Science and Technology, 46, 5294-5301.
Philips S., Hendrikus J., Laanbroek, Verstraete W., (2002),
Origin, causes and effects of increased nitrite
concentrations in aquatic environments, Reviews in
Environmental Science and Bio-Technology, 1, 115-
141.
Prokopkin G., Mooij W.M., Janse J.H., Degermendzhy
A.G., (2010), A general one-dimensional vertical
ecosystem model of Lake Shira: description,
parameterization and analysis, Aquatic Ecology, 44,
585-618.
Sandu M., Lozan R., Ropot V., (1989), Determination of
nitrate and nitrite in natural waters, Chemistry and
Technology of Water, 3, 120-125.
SR EN 26777:2002/C91:2006, (2006), Water quality -
determination of the content of nitrites. The method of
the spectrometry of molecular absorption.
Spataru P., (2011), Transformation of organic matter in the
surface waters of the Republic of Moldova, PhD thesis,
Institute of Chemistry of the Academy of Sciences of
Moldova, Chisinau, Republic of Moldova.
SR ISO 7890-3, (2000), Water quality - The determination
of the content of nitrates. The part 3: The spectrometric
method with sulfosalicylic acid, On line at:
http://agris.fao.org/agris-
search/search.do?recordID=XF8981389.
State Register (2007), State of natural mineral water, soft
drinks and bottled drinking publshed: 24.08.2007,
Official Monitor No.131-135 art. No 970 (in
Romanian), Kishinev, Republic of Moldavia.
Teodoru C.R., Friedl G., Friedrich J., Roehl U., Sturm M.,
Wehrli B., (2007), Spatial distribution and recent
changes in carbon, nitrogen and phosphorus
accumulation in sediments of the Black Sea, Marine
Chemistry, 105, S52-S58.
Van Leeuwen, F.X.R., (2000), Safe drinking water: the
toxicologist’s approach, Food and Chemical
Toxicology, 38, 51-58.
Weng Y.M., Hotchkiss J.H., Babish J.G., (1992), N-
nitrosamine and mutagenicity formation in Chinese
salted fish after digestion, Food Additives and
Contaminants, 9, 29-37.
Zubcov E., Bagrin N., Ungureanu L., Biletchi L., Borodin
N., Bogonin Z., (2011), Dynamics of the
hydrochemical parameters and the quality of the Prut
River, Bulletin of the Academy of Sciences of Moldavia.
Life Sciences (Buletinul Academiei de Științe a
Moldovei Științele vieții), 313, 103-110 .
Zubcov E., Ungureanu L., Ene A., Zubcov N., Bagrin N.,
Borodin N., Lebedenco L., Biletchi L., (2010),
Assessment of chemical compositions of water and
ecological situation in Dniester river, Journal of
Science and Arts, 10, 47-52.
1719
1720
Environmental Engineering and Management Journal July 2018, Vol.17, No. 7, 1721-1728
http://www.eemj.icpm.tuiasi.ro/; http://www.eemj.eu

“Gheorghe Asachi” Technical University of Iasi, Romania

PHYSICAL NETWORKS BASED ON GELATIN

AND AZO-POLYSILOXANES

Elena- Luiza Epure1, Elena Resmeriţă1, Anca Daniela Rusu1,2,

Bogdana Simionescu3, Constanţa Ibănescu1,4*
1“Gheorghe Asachi” Technical University of Iasi, Department of Natural and Synthetic Polymers, 73,
Prof. Dimitrie Mangeron Street, 700050 - Iasi, Romania; Fax number +40232271311;
2CEA, LIST Saclay, Laboratoire Capteurs et Architectures Électroniques, 91191 Gif-sur-Yvette, Cedex, France
3“Costin D. Nenitescu” Centre of Organic Chemistry, 202B Splaiul Independentei, 71141 Bucharest, Roumania
4“Petru Poni” Institute of Macromolecular Chemistry, 41A, Grigore Ghica Voda Alley, 700487 - Iasi, Romania

Abstract

Novel systems based on modified azo-polysiloxanes and gelatin with potential biological applications were prepared and
characterized. Simultaneous rheological and UV irradiation studies allowed evidencing structural modification inside the gelatin
matrix. For all analyzed samples a shift in the temperature corresponding to the physical network destructuration, from 30˚C
(corresponding to gelatin) to 40˚C (for the composite) was noticed. The experimental results proved the existence of interactions
between gelatin and polysiloxanes intensified after UV irradiation. The increase in the values of G’ and G” is a consequence of
system restructuration leading to more arranged architectures able to release the included active principle. As a function of the azo-
polysiloxane structure, the destructuration temperature of the composite can be tuned in the domain 30- 40°C.

Key words: biopolymers, dynamic moduli, gelatin, nucleobases, rheology

Received: January, 2014; Revised final: October, 2014; Accepted: October, 2014; Published in final edited form: July, 2018

1. Introduction multiple applications in this field. A lot of synthetic

Drug delivery systems have been intensively
studied in the last period, representing a rapidly
growing area of interest for researchers all over the
world. The possibility of controlling the time
dependent release of the active principle and to target
it to the right place represents a challenge accepted by
scientists for many years and different classes of
materials have been produced, analyzed and tested to
be used as drugs carriers (Gupta et al., 2012; Kim et
al., 2009; Kost and Langer, 2012; Machín et al., 2012;
Malafaya et al., 2007; Oh et al., 2009; Qiu and Park,
2012; Vashist et al., 2012; Wang and Grayson, 2012;
Zhang et al., 2013). Considering their low toxicity,
biodegradability and chemical stability in the
physiological environment, polymers have found
polymers have been used as drug carriers and among
the most frequently studied we can mention
polylactide/polyglycolide (Panyam and Labhasetwar,
2003), poly(ε-caprolactone) (Potineni et al., 2003),
poly(ester-anhydride) (Pfeifer et al., 2005),
polyaspartamides (Caldwell et al., 1997; Machado et
al., 1992), poly (amidoamines) (Ferruti et al., 2002),
poly(hydroxypropyl-methacrylate) (Hovorka et al,
2006) etc. In addition to synthetic polymers, naturally
derived components such as chitosan (Janes et al.,
2001), collagen (Lee et al., 2001) and gelatin (Young
et al., 2005) are very attractive materials used in drug
delivery systems. Due to its biocompatibility,
biodegradability without toxic degradation products
and high physiological tolerance, gelatin presents
multiple advantages as drug carrier. Gelatin is derived

*
Author to whom all correspondence should be addressed: email: ibanescu@ch.tuiasi.ro
 Epure et al./Environmental Engineering and Management Journal 17 (2018), 7, 1721-1728
from collagen contained in animal tissues by hot water
extraction following an acid or alkaline pre-treatment
and has distinctive sol-gel behavior forming 100%
thermo-reversible gels. From more than 25 collagen
types, only three can be used to prepare gelatin
(collagen type I, II and III) (Schrieber and Gareis,
2007). Function of the preparation method, gelatin is
characterized by different physical properties and
isoelectric points. The ability of tuning the electric
properties of gelatin has contributed to the
understanding of the concept of sustained release of
proteins from polymer matrices (Young et al., 2005).
Because the physical network of gelatin hydrogel is
destroyed around 30°C, crosslinking agents are
necessary for 3D-system stabilization. There are three
important methods to prepare gelatin carriers: gelatin
microspheres, gelatin block hydrogels and porous
gelatin block hydrogels (Hosaka et al., 2004; Kang et
al., 1999; Yamamoto et al., 2000). Two types of
crosslinking agents can be used: non-zero length and
zero-length (Kuijpers et al., 2000). A major problem
arising when chemical crosslinking agents are used is
the total elimination of the unreacted products.
Here in we propose a new system based on
gelatin and different polysiloxanes containing azo-
benzene groups in the side-chains, with potential
applications as drug release systems. Due to the
presence of azo-polysiloxanes sequences the
modification of the diffusion rate inside the gelatin
matrix under UV light irradiation is expected. The
azo-benzene groups undergo a specific behavior as a
consequence of the azo-benzene groups’ trans-cis
isomerization capacity under UV/VIS irradiation
(Moleavin et al., 2011; Rau, 2003). The geometrical
modifications of the azo-groups are accomplished by
strong alteration of the dipole-moment (in the case of
unsubstituted azobenzene from 0.1 D to 3.5 D) that
will also influence the diffusion process inside the
material.
Two different azo-polymers structures mixed
with gelatin were investigated: azo-polysiloxane
modified with adenine or thymine and azo-
polysiloxanes grafted with poly(dimethyl acrylamide)
and poly(butyl acrylate) segments. A rheological
method was used to evaluate the composites behavior
under UV light irradiation. Following the UV
irradiation a shift in the temperature corresponding to
the gel network destructuration, from 30°C
(corresponding to gelatin) to 40 °C (for gelatin-
polymer composite) was evidenced.
2. Experimental
2.1. Materials
Tetrahydrofurane, ethyl acetate,
dichloromethane, nucleobases (adenine and thymine),
dimethyl sulfoxide, methanol, chloroform, acryloyl
chloride, dicyclohexylcarbodiimide, 4-N,N
dimethylaminopyrydine and 4-phenylazophenol were
purchased from Aldrich, Steinheim, Germany and
used without supplementary purification.
1722
The ((chloromethyl) phenylethyl)
methyldichlorosilane was purchased from ABCR
GmbH & Co. KG, Karlsruhe, Germany and also used
without supplementary purification. The azo-
methacrylate monomer was obtained through Sterlich
esterification of acryloyl chloride with 4-
phenylazophenol in the presence of 4-N,N-
dimethylaminopyrydine and
dicyclohexylcarbodiimide (Neises and Sterlich,
1978). The compound was purified by thin layer
chromatography technique on silica gel using a
mixture of 20/1 dichloromethane and ethyl acetate as
eluent. Type A gelatin (G2500) was acquired from
Aldrich, Steinheim, Germany and used as received.
The linear polysiloxane was prepared in
agreement with ref. (Kazmierski et al., 2004). A two-
step reaction was employed for the modification of
azo-polysiloxanes with nucleobases, starting from a
polysiloxane containing chlorobenzyl groups in the
side-chain (Fig. 1). In the first reaction step, the
polysiloxane was modified with sodium salt of 4-
hydroxyazobenzene (50–60% substitution degree)
and, in the second step, the unreacted chlorobenzyl
groups were substituted with adenine (Sample 1) or
thymine (Sample 2). The polymers were characterized
by 1H-NMR, details concerning polymer synthesis and
characterization being previously reported (Hurduc et
al., 2007). The 1H-NMR spectra were recorded on a
Bruker 400 MHz apparatus. SET-LRP (single electron
transfer - living radical polymerization) technique
(Fig. 1) was used to obtain the graft azo-polysiloxanes
(Rusu, 2012a; 2012b; 2013). In a typical reaction 0.09
g bipyridyl were added to 0.1 g cyclic polysiloxane
macroinitiator previously dissolved in 1.5 mL DMSO.
The reaction mixture was introduced in a 25 mL flask
where monomers (0.1 g of azo-methacrylate
(AZOMA), 0.5 g of N,N-dimethyl acrylamide
(DAM), 0.5 g of butyl acrylate (BA)), and Cu wire
(0.25 g) were added. The system was degassed by five
freeze–pump–thaw cycles. The flask was heated at the
reaction temperature (80 °C) and the synthesis was
performed under nitrogen atmosphere and stirring, for
8 h. At the end of the reaction, the mixture was diluted
with CHCl3 and the catalyst was removed by passing
the solution through a basic alumina column. The
purified solution was then precipitated in diethyl ether,
washed three times with the same solvent, and dried
under vacuum (Sample 3).
2.2. Methods
The polymer composites were prepared by
mixing a 1% azo-polysiloxane solution in THF with a
5 % gelatin solution in water.
The gelatin solution in water was prepared at
90°C in agreement with the literature data (Benton et
al., 2009). Following the complete dissolution of the
gelatin the solution was cooled at 60 °C and the azo-
polysiloxane solution was added drop by drop under
intense agitation. After the complete addition of the
azo-polysiloxane, the mixture was heated to 66 °C for
another 30 minutes to eliminate the THF.
 Physical networks based on gelatin and azo-polysiloxanes
Fig. 1. Synthesis of azo-polysiloxanes modified with nucleobases (A) and of graft polysiloxanes (B)
Afterwards, the mixture was cooled at 35 °C
and introduced between the plates of the rheometer
measuring system. The rheological properties were
investigated using an Anton Paar Physica MCR 501
rheometer equipped with a Peltier system for the
temperature control. The temperature can be adjusted
between -40 and 200 °C. Measurements were carried
out using parallel-plate geometry (Danu et al., 2012).
Firstly, a strain sweep was performed to establish the
limits of the linear viscoelastic range (LVE) and then
temperature tests were carried out at constant
oscillation amplitude (γ = 0.5% situated in linear
viscoelastic range) and frequency (f = 1Hz). The
temperature had been varied between 10 and 100 °C
with a heating rate of 0.5 °C /min. Simultaneous UV
irradiation and rheological studies could be performed
using an advanced UV system, under a Peltier
controlled P-PTD-UV chamber on Physica MCR 501
rheometer. As measuring system a 43 mm parallel
glass plate fixture was used. The UV source was the
Omnicure Series 1000 manufactured by EXFO
(Vanier, QC, Canada), using a 100 W lamp able to
deliver UV light of 365 nm using a Mercury bulb.
The UV/Vis kinetic studies in THF solution
were carried out using a Shimadzu spectrophotometer
(Shimadzu UV 1700 Pharma Spy, Kyoto, Japan). The
azo-polymer solutions were irradiated using a UV
lamp (50 W) equipped with 365 nm filter. The
isomerization kinetics of azobenzene can be
investigated due to the strong absorption peaking at
350 nm, attributed to the π-π* transition of the
azobenzene chromophore. During UV irradiation at
365 nm the absorbance corresponding to the π-π*
transition strongly decreases, permitting to calculate
the trans (respectively cis) content of the azo-groups.
3. Results and discussion
The main characteristics of the synthesized
modified azo-polysiloxanes are presented in Table 1.
The polymers’ molecular weights (Mn) were
calculated using 1H-NMR spectra.
1723
 Epure et al./Environmental Engineering and Management Journal 17 (2018), 7, 1721-1728

The chemical structure of the obtained
polymers was confirmed by NMR spectroscopy using
CDCl3 as a solvent, at room temperature. If we refer
to Sample 3 in a typical 1H-NMR spectrum we can
identify: the signal corresponding to the methyl groups
(CH3-Si-) from polysiloxane (reference signal for
polysiloxane structural unit) at 0.1 ppm, the
characteristic signal for the methyl groups (-CH3)
corresponding to the poly (butyl acrylate, BA)
segment at 0.9 ppm, the signal for (-CH2-CH2-) groups
at 1.5 ppm and (-CH2-O-CO-) signal present at 4 ppm
. The signal corresponding to the methyl groups of
dimethyl acrylamide (DAM) is present at 2.9 ppm.
The signals corresponding to the azo-groups are found
in the domain 7.5 - 8.0 ppm.
Firstly, the photochromic behavior of the
azo-polymers under UV irradiation was investigated.
Fig. 2 presents the kinetic curves of the irradiation
process. Evidently, the polymers response in solution
is very fast, less than 2 minutes of UV irradiation
being necessary to obtain the maximum value of cis-
group. In these circumstances 5 minutes of UV
exposure were selected for the rheological
experiments.
It was also noticed that the trans-cis
conversion rate of the azo-groups is influenced by the
substituted molecule. Likely due to the smaller size of
thymine comparing with the adenine, the structures of
Sample 2 are possibly more compact than Sample 1.
The 60% conversion rate of sample containing
thymine may be due to the steric hindrance caused by
the network density (Resmerita et al., 2010). The
evaluation of the rheological properties as a function
of temperature evidenced the existence of interactions
between gelatin and azo-polysiloxanes, the addition of
modified polymers having significant influence on the
network formation and stability. A 5% gelatin solution
in water was used as reference for the temperature
sweep tests. From Fig. 3a the destructuration process
of the physical network corresponding to gelatin (5%
in water) can be evidenced at 30°C, in agreement with
literature data (Benton et al., 2009). For temperatures
under 30°C, a stable gel structure is developed with
the storage modulus G’ higher than the loss modulus,
G”. Increasing temperature, the physical bonds
between chains are broken and the system turns liquid
with G’<G”.

Table 1. Characteristics of the modified azo-polysiloxanes

Azo groups Nucleobases or other Nucleobase

Spl. Azo-units groups/
content monomers Mn or monomers-units
no. polymeric chain
(%) Content (%) /polymeric chaina
1 50 15 7,650 12a 4
2 50 11 7,530 12a 3
3 12 22(DAM+BA) 4,200 3b 12DAM + 10BA
a
Average polymerization degree of the polysiloxanic chains = 25; b Average polymerization degree of the polysiloxanic chains (cyclic) = 4; DAM
= dimethyl acrylamide; BA = butyl acrylate

Fig. 2. Polymer response after UV irradiation: (a) Sample 1; (b) Sample 2; (c) Sample 3

1724
 Physical networks based on gelatin and azo-polysiloxanes

Fig. 3. Temperature sweep for a 5% gelatin solution in water:

(a) and mixtures of gelatin with Sample 1 (b), Sample 2 (c) and Sample 3 (d)

This particular behavior of gelatin networks
turning from solid-like to liquid-like behavior by
changing temperature can be exploited for the
controlled release of active principles, but the
characteristic temperature of phase transition is too
low. The idea is to bring this temperature closer to the
physiological temperature.
The polymer addition into gelatin network does
not significantly modifies the destructuration
temperature, but in the case of Sample 1 considerable
differences in curves profile were evidenced (Fig. 3b):
an increase in both dynamic moduli can be observed
starting with 40°C, followed by a stabilization of the
values, two orders of magnitude lower as compared
with the initial one. Over 60°C some kind of stable
week gel structure is formed. The purine structure of
adenine may be responsible for the complex
rheological behavior of Sample 1.
In the case of Sample 2 added into gelatin
network (Fig. 3c) more stable and flexible structure at
higher temperatures can be observed. The rheological
behavior of the gelatin network incorporating a
polysiloxane grafted with thermo-sensitive segments
(Sample 3) is very similar to pure gelatin, but the
flexibility and stability of the resulted system is
increased (Fig. 3d).
The non-covalent interactions type π-π staking
and hydrogen bonding (Mignon et al., 2005) are
responsible for the different behaviors of the samples
containing nuceobases moieties when geometrical
conditions are accomplished. Different factors
(nucleobases type, main chain conformation, network
density) lead to the development of physical
interactions makeing this mechanism more
complicated than in case of solitary nucleobases. The
possibility of thymine dimers formation (Li et al.,
2008; Goodsell, 2001) explains the higher stability of
Sample 2 as compared with Sample 1.
The next step was to investigate the influence
of the trans-cis isomerisation of included modified
azo-polysiloxanes on the rheological properties of the
gelatin networks. The samples were irradiated for 5
minutes before running the temperature sweep tests,
enough time for the complete isomerization of the azo-
groups, according to the kinetic curves of the
irradiation process. As expected a shift towards higher
values in the specific phase transition temperatures
was noticed for all samples (Fig. 4).

1725
 Epure et al./Environmental Engineering and Management Journal 17 (2018), 7, 1721-1728

Fig. 4. The influence of UV irradiation on rheological properties of gelatin network incorporating modified azo-polysiloxanes;
(a) 5% gelatin in water with Sample 1, (b) 5% gelatin in water with Sample 2, (c) 5% gelatin in water with Sample 3

The destructuration temperature increases at corresponding to the azo-group in the cis

34°C for gelatin with Sample 1, 39°C for gelatin with
Sample 2 and 40°C when gelatin is mixed with Sample
3. If below the phase transition the rheological
behavior is not modified by the incorporation of azo-
polymers into gelatin network, things are changing
after the irradiation process. The trans-cis
isomerization is increasing the gel stability, G’ being
increased with one order of magnitude. When
covalently bound to the polymer chains, the azo-
molecule motion can have a strong impact on the
conformation of the entire polymer chain and on its
organization within a matrix. A remarkable example
is given in the case of azo-polymer films subjected to
a light interference pattern. In this case, the molecular
isomerization process leads to large topographic
modifications reproducing the intensity modulations
(Rochon et al., 1995, Damian et al., 2014).
But more important modification of the
rheological behavior takes place in the liquid state,
even for the samples for which no polymer/polymer
interactions were identified before the irradiation (Fig.
4a, b, c). Probably the higher mobility of both
polymers (gelatin and azo-polysiloxane) in the liquid
state and the strong modification of the dipole moment
configuration (cis azophenol dipole-moment = 3.5D)
are responsible for the intensification of the
polymer/polymer interactions, leading to a system
restructuration at high temperature.
Moreover, the temperature is inducing the azo-
molecules relaxation (Garcia-Amorós et al., 2010),
which combined with the dipole-dipole interactions
and shear rate leads to an increase of viscoelastic
moduli with six or seven orders of magnitude,
becoming even higher than the moduli values at 15 °C.
A pseudogel like system is formed. Comparing the
three graphs (Fig.4) a delay between the
destructuration temperature and the sharp increase of
G’ and G’’ was noticed. The star shape polymer of
Sample 3 (Fig. 4c) is letting the azo-groups more
exposed and around 55°C both moduli are increasing.
In the case of linear azo-polysiloxanes
substituted with nucleobases the sharp increase of the
two moduli starts at around 65°C; because of the coil
conformation, and due to the thermal relaxation of
azo-molecules, some steric hindrance takes place.
All the results are an indication that the
addition of the modified azo-polysiloxane in gelatin
has noticeable influence on the overall rheological

1726
 Physical networks based on gelatin and azo-polysiloxanes
temperature dependent behavior. Moreover, the UV
irradiation of the modified systems allows the increase
of the phase transition temperature bringing it closer
to the physiological temperature. In this way the
systems become suited for the use as drug release
systems and the entities included in the azo-polymers
can determine specific behaviors.
4. Conclusions
New systems based on gelatin and three
different polysiloxanes containing azo-benzene
groups in the side-chain with potential application as
drug release systems were investigated by rheological
methods combined with UV irradiation.
The experimental results proved the existence
of interactions between gelatin and polysiloxanes
intensified after UV irradiation. Due to its structure,
gelatin is able to form multiple hydrogen bonds
between the protein chains, thus generating a network
which can be broken by chemical or physical stress
and reformed after the stress is removed (Kozlov,
1983). By embedding azo-polysiloxanes substituted
with nucleobases into this network new hydrogen
bonds with significant effect onto cohesive forces
between molecules are generated. Moreover, after the
UV irradiation of the azo-molecules, dipol-dipol
interactions are changing the physical proprieties of
gelatin network. The increase in the values of G’ and
G” is a consequence of system restructuration leading
to more organized architectures able to release the
included active principle. As a function of the azo-
polysiloxane structure, the destructuration
temperature of the composite can be tuned in the
domain 30-40 °C.
Acknowledgements
The authors want to thank to ANCS for financial support of
this research (Grant CEA - C1- 01/2010).
References
Benton J.A., DeForest C.A., Vivekanandan V., Anseth K.S.,
(2009), Photocrosslinking of gelatin macromers to
synthesize porous hydrogels that promote valvular
interstitial cell function, Tissue Engineering: Part A.,
15, 3221-30.
Caldwell G., Neuse E.W., Perlwitz A.G., (1997), Water
soluble polyamides as potential drug carriers. IX.
Polyaspartamides grafted with amine-terminated poly
(ethylene oxide) chains, Journal of Applied Polymer
Science, 66, 911-9.
Damian V., Resmerita E., Stoica I., Ibanescu C., Sacarescu
L., Rocha, L., Hurduc N., (2014), Surface relief gratings
induced by pulsed laser irradiation in low glass-
transition temperature azopolysiloxanes, Journal of
Applied Polymer Science, 41015, doi:
10.1002/app.41015
Danu M, Rotaru I, Ibănescu C, Hurduc N., Ibănescu S.A.,
Simionescu B.C., (2012), Gluten content influence on
rheological behaviour of starch-gluten networks,
Environmental Engineering and Management Journal,
11, 1883-8.
Ferruti P., Marchisio A., Duncan R., (2002), Poly(amido-
amine)s: Biomedical Applications, Macromolecular
Rapid Communications, 23, 332-55.
Garcia-Amorós J., Sanchez-Ferrer A., Massad W.A, Nonell
S., Velasco D., (2010), Kinetic study of the fast thermal
cis-to-trans isomerization of para -, ortho - and
polyhydroxyazobenzenes, Physical Chemistry
Chemical Physics, 12, 13238–42
Goodsell D.S., (2001), The Molecular Perspective:
Ultraviolet Light and Pyrimidine Dimers, The
Oncologist, 6, 298-9.
Gupta S., Tyagi R., Parmar V.S., Sharma S.K., Haag R.,
(2012), Polyether based amphiphiles for delivery of
active components, Polymer, 53, 3053-78.
Hosaka A., Koyama H., Kushibiki T., Tabata Y., Nishiyama
N., Miyata T., Shigematsu H., Takato T., Nagawa H.,
(2004), Gelatin hydrogel microspheres enable pinpoint
delivery of basic fibroblast growth factor for the
development of functional collateral vessels,
Circulation, 110, 3322-8.
Hovorka O., Etrych T., Subr V., Strohalm J., Ulbrich K.,
Ríhová B., (2006), HPMA based macromolecular
therapeutics: internalization, intracellular pathway and
cell death depend on the character of covalent bond
between the drug and the peptidic spacer and also on
spacer composition, Journal of Drug Targeting, 14,
391-403.
Hurduc N., Enea R., Scutaru D., Sacarescu L., Donose B.C.,
Nguyen A.V., (2007), Nucleobases modified azo-
polysiloxanes, materials with potential application in
biomolecules nanomanipulation, Journal of Polymer
Science Part A: Polymer Chemistry, 45, 4240-8.
Janes K.A., Calvo P., Alonso M.J., (2001), Polysaccharide
colloidal particles as delivery systems for
macromolecules, Advanced Drug Delivery Reviews, 47,
83-97.
Kang H.W., Tabata Y., Ikada Y., (1999), Fabrication of
porous gelatin scaffolds for tissue engineering,
Biomaterials, 20, 1339-44.
Kazmierski K., Hurduc N., Sauvet G., Chojnowski J.,
(2004), Polysiloxanes with chlorobenzyl groups as
precursors of new organic-silicone materials, Journal of
Polymer Science Part A: Polymer Chemistry, 42, 1682-
92.
Kim S., Kim J.H., Jeon O., Kwon I.C., Park K., (2009),
Engineered polymers for advanced drug delivery.,
European Journal of Pharmaceutics and
Biopharmaceutics, 71, 420-30.
Kost J., Langer R., (2012), Responsive polymeric delivery
systems, Advanced Drug Delivery Reviews, 64, 327-41.
Kozlov P.V., (1983), The structure and properties of solid
gelatin and the principles of their modification,
Polymer, 24, 651-66.
Kuijpers A.J., Engbers G.H., Krijgsveld J., Zaat S.A.,
Dankert J., Feijen J., (2000), Cross-linking and
characterisation of gelatin matrices for biomedical
applications, Journal of Biomaterials Science, Polymer
Edition, 11, 225-43.
Lee C.H., Singla A., Lee Y., (2001), Biomedical
applications of collagen, International Journal of
Pharmaceutics, 221, 1-22.
Li G., Sichula V., Glusac K.D., (2008), Role of Adenine in
Thymine-Dimer Repair by Reduced Flavin-Adenine
Dinucleotide, The Journal of Physical Chemistry B,
112, 10758-64.
Machado M.L., Neuse E.W., Perlwitz A.G., (1992), Water-
soluble polyamides as potential drug carriers, V.
Carboxy-functionalized polyaspartamides and
1727
 Epure et al./Environmental Engineering and Management Journal 17 (2018), 7, 1721-1728
copolyaspartamides, Angewandte Makromolekulare
Chemie, 195, 35-56.
Machín R., Isasi J.R., Vélaz I., (2012), β-cyclodextrin
hydrogels as potential drug delivery systems,
Carbohydrate Polymers, 87, 2024-30.
Malafaya P.B., Silva G.A., Reis R.L., (2007), Natural–origin
polymers as carriers and scaffolds for biomolecules and
cell delivery in tissue engineering applications,
Advanced Drug Delivery Reviews, 59, 207-33.
Mignon P., Loverix S., Steyaert J., Geerlings P.,(2005),
Influence of the π–π interaction on the hydrogen
bonding capacity of stacked DNA/RNA bases, Nucleic
Acids Research, 33, 1779-89.
Moleavin I., Ibanescu C., Hodorog-Rusu A., Peptu E.,
Doroftei F., Hurduc N., (2011), Amphiphilic
azopolymers capable to generate photo-sensitive
micelles, Central European Journal of Chemistry, 9,
1117-25.
Neises B., Sterlich W., (1978), Simple Method for the
Esterification of Carboxylic Acids, Angewandte
Chemie International Edition, 17, 522-4.
Oh J.K., Lee D.I., Park J.M., (2009), Biopolymer-based
microgels/nanogels for drug delivery applications,
Progress in Polymer Science, 34, 1261-82.
Panyam J., Labhasetwar V., (2003), Biodegradable
nanoparticles for drug and gene delivery to cells and
tissue, Advanced Drug Delivery Reviews, 55, 329-47.
Pfeifer B.A., Burdick J.A., Langer R., (2005), Formulation
and surface modification of poly(ester-anhydride)
micro- and nanospheres, Biomaterials, 26, 117-24.
Potineni A., Lynn D.M., Langer R., Amiji M., (2003),
Poly(ethylene oxide)-modified poly(β-amino ester)
nanoparticles as a pH-sensitive biodegradable system
for paclitaxel delivery, Journal of Controlled Release,
86, 223-34.
Qiu Y., Park K., (2012), Environment-sensitive hydrogels
for drug delivery, Advanced Drug Delivery Reviews,
64, 49-60.
Rau H., (2003), Azo Compounds, In: Photochromism:
Molecules and Systems, Dürr H, Bouas-Laurent H
(Eds.), 2nd Edition, Elsevier, Amsterdam, 165-88.
Resmerita A.-M., Epure L., Hurduc N., Adès D., Siove A.,
(2010), Surface properties, thermal behavior, and
1728
molecular simulations of azo-polysiloxanes under light
stimuli. Insight into the relaxation, Macromolecular
Research, 18, 721–9.
Rochon P., Batalla E., Natansohn A. L., Kumar J., (1995)
Optically induced surface gratings on azoaromatic
polymer films, Applied Physics Letters, 66, 136-8.
Rusu-Hodorog A.D., Ibanescu C., Danu M., Simionescu
B.C., Rocha L., Hurduc N., (2012a), Thermo-sensitive
polymers based on graft polysiloxanes, Polymer
Bulletin, 69, 579-95.
Rusu (Hodorog) A.D., Ibanescu C., Moleavin I., Hurduc N.,
(2012b), Thermo-responsiveness of polysiloxanes
grafted with poly(dimethyl acrylamide) segments,
Central European Journal of Chemistry, 10, 1338-48.
Rusu A.D., Ibanescu C., Resmerita E., Simionescu B., Danu
M., Hurduc V., Ibanescu S.A., (2013), Stimuli
Responsive Graft Polysiloxanes, Revista de Chimie -
Bucharest, 64, 667-71.
Schrieber R., Gareis H., (2007), From Collagen to Gelatine,
Gelatine Handbook: Theory and Industrial Practice,
Wiley-VCH, Weinheim, 45-117.
Vashist A., Gupta Y.K., Ahmada S., (2012), Interpenetrating
biopolymer network based hydrogels for an effective
drug delivery system, Carbohydrate Polymers, 87,
1433-9.
Wang Y., Grayson S.M., (2012), Approaches for the
preparation of non-linear amphiphilic polymers and
their applications to drug delivery, Advanced Drug
Delivery Reviews, 64, 852-65.
Yamamoto M., Tabata Y., Hong L., Miyamoto S.,
Hashimoto N., Ikada Y., (2000), Bone regeneration by
transforming growth factor β1 released from a
biodegradable hydrogel, Journal of Controlled Release,
64, 133-42.
Young S., Wong M., Tabata Y., Mikos A.G. (2005), Gelatin
as a delivery vehicle for the controlled release of
bioactive molecules, Journal of Controlled Release,
109, 256-74.
Zhang Y., Chan H.F., Leong K.W., (2013), Advanced
materials and processing for drug delivery: The past and
the future, Advanced Drug Delivery Reviews, 65, 104-
20.
Environmental Engineering and Management Journal July 2018, Vol.17, No. 7, 1729-1741
http://www.eemj.icpm.tuiasi.ro/; http://www.eemj.eu

“Gheorghe Asachi” Technical University of Iasi, Romania

ENERGY SOURCES ANALYSIS FROM THE PERSPECTIVE

OF SUSTAINABLE DEVELOPMENT

Tiberiu Rusu1∗, Ciprian Cristea2, Tudor Rusu1, Ileana-Codruţa Groze3, Laurenţiu Stoica4
1TechnicalUniversity of Cluj-Napoca, Department of Environmental Engineering and Sustainable Development
Entrepreneurship, 103-105 Muncii Ave, Cluj-Napoca, Romania
2Technical University of Cluj-Napoca, Department of Electrical Machines and Drives, 26-28 G. Bariţiu Street,

Cluj-Napoca-, Romania
3Babeş-Bolyai University, Faculty of Environmental Science and Engineering, 30 Fântânele Street, Cluj-Napoca, Romania
4Technical University of Cluj-Napoca, Department of Manufacturing Engineering, 103-105 Muncii Ave, Cluj-Napoca, Romania

Abstract

A judicious use of energy is a fundamental factor to achieve sustainable development. The limited character of energetic fossil
resources and the pollution generated by burning fossil fuels for electricity production generate the need to replace them with other
sources of energy. Despite the fact that fossil fuels would continue to play a prevailing role in the energy supply for decades to
come, renewable energy resources have the potential of contributing to the increasing global energy demands, while simultaneously
emerge the most efficient solutions for clean and sustainable energy development in the world. In this framework, the main scope
of the present study is to provide an analysis of the current state of world natural resources used to produce energy and energy
consumption degree across different regions of the world. At the same time, this paper aims to compare the environmental impacts
in water, air, soil and ecosystem produced by a range of conventional and renewable energy sources, which is necessary to be
reduced for building a genuine low-carbon society.

Key words: energy sources, natural resources, sustainable development, urbanization

Received: August, 2014; Revised final: September, 2014; Accepted: November, 2014; Published in final edited form: July 2018

1. Introduction century has been associated with urban sprawl and,

Energy is a vital factor in economic growth and
prosperity, improving the quality of life and satisfying
human needs (Michalena and Angeon, 2009; Tantau
et al., 2017). The availability of energy is a
requirement for the functioning of modern societies,
and the demand for energy resources has an effect on
the politics of countries in all stages of development
(Chalvatzis and Hooper, 2009; Sheau-Ting et al.,
2016). The urbanization and economic development
are strictly connected phenomena that have numerous
implications on global environmental change. The
intensive urbanization since the mid-half of the last
with urban land area doubling in the OECD countries
and growing by a factor of five in the rest of the world
(OECD, 2010). The human impact on the environment
manifests through many and various ways, but the
most important factor, in terms of extending and
consequences is represented by natural resources
degradation and abatement (IEA, 2010a, b; Tosovic et
al., 2016).
Frequently, specific literature mentioned that
the most common natural resources classification
divides into non-renewable resources (e.g. oil, natural
gas, coal, uranium, minerals) and renewable
resources, which may be power source for renewable

∗
Author to whom all correspondence should be addressed: e-mail: tiberiu.rusu@sim.utcluj.ro
 Rusu et al./Environmental Engineering and Management Journal 17 (2018), 7, 1729-1741
energy (wind energy, solar energy, geothermal energy,
hydroelectric energy) (Axinte et al., 2015; Boyle et al.,
2003; Casper, 2010; Grübler and Fisk, 2012; Tong,
2010; Zaharia, 2004). There are many definitions of
sustainable development, including the following
common one: “development that meets the needs of
the present without compromising the ability of future
generations to meet their own needs” (Bruntland,
1987).
Due to the rapid development of the global
economy, energy demand has grown quickly,
particularly in emergent countries, which tend to have
a greater part of their economies in manufacturing
industries. Fossil fuels supply the majority of energy
needs worldwide, with 86.7 percent of global energy
consumption in 2013 (BP, 2014). A recent World
Energy Council study found that without any change
in our current practice, the world energy demand in
2050 would be 50 percent higher than 2013 level, at
least 80 percent of this increase is anticipated to come
from developing countries (WEC, 2013). Because the
fossil fuel resources required for the generation of
energy are expected to dominate the global
consumption of energy resources in the coming years,
as well as because these are becoming scarce,
alternatives that would be viable and regenerative to
attain sustainability have to be explored (Shamsuddin
et al., 2014). In this regard, renewable energy
resources represent one of the most efficient solutions
to environmental problems that are faced today by
humanity.
The renewable energy technologies are
recording significant worldwide growth due to
declining costs of generation, increasing electricity
demand and the environmental concerns, especially
those concerning climate change. According to
Bloomberg New Energy Finance (BNEF, 2014),
global investment in clean energy was $254 billion in
2013, almost five times the total of $54 billion in 2004.
Natural resources exploitation represents a key factor
of economic growth and development, but is
important to consider the consequences of potential
irreversible changes that may occur to the
environment (Bond et al., 2003; Zaharia, 2004).
Sustainable development and natural resources are
indissoluble related with the social, economic,
technological development, and environmental
protection. The economic competitiveness and
ecological sustainability are complementary aspects of
current society’s common goal (ASCE, 2010). In this
context, the main scope of this paper is to analyze and
compare the current state of natural resources used
worldwide to produce energy, in terms of their
potential of contributing to climate change mitigation
and sustainable development, as well as to review the
environmental impact assessment for water, air, soil
and ecosystem for a range of conventional and
renewable energy sources.
This paper is based on data sources including
peer-reviewed papers, state and agency reports, non-
governmental organizations’ reports, and recent
statistical data and reviews. A unitary manner of data
1730
synthesis clearly and meticulously gathered from the
sources mentioned above is proposed by the authors,
along with a pertinent analysis of results.
2. Urban-industrial systems
Urbanization and economic growth are
inextricably linked. Globally, there is no country that
has reached middle-income status without a
substantial population shift into cities (Hughes and
Cain, 2003). Studies have revealed that the simple
correlation coefficient across countries based on the
percentage of urban residents in a country and gross
domestic product (GDP) per capita is of about 0.85
(Henderson, 2003), fact that leads to the conclusion
that urbanization is an inevitable part of a modern
society. Cities generate about 80 percent of world’s
GDP (Grubler and Fisk, 2012). The world’s GDP is
expected to rise by an average of 3.6 percent per year
from 2010 to 2040, particularly in emerging, non-
OECD regions, where it increases by 4.7 percent per
year. This increase drives the sustained growth in
future energy consumption, based mostly on fossil
fuels, which in turn plays a key role in the upward
trend in CO2 emissions (EIA, 2013). However,
relatively high levels of urbanization do not always
assure high levels of economic welfare in developing
countries. This is most evident in Latin America where
the urbanization levels are above 80%, yet with real
per capita incomes and GDP about a third that of
developed nations (UNEP, 2012).
In the present era of rapidly growing
development, urban sustainability is threatened by
complex global uncertainty and change which have
broad range, and threaten multiple sectors – such as
water, energy, transport and waste – that are critical
for urban sustainability (MGI, 2011). The urban
population growth rate is superior to that of the total
population. As can be noticed in Table 1 not all
regions around the world have reached the same level
of urbanization (United Nations, 2014).
Rosenthal and Strange (2004) show that
doubling city size increases productivity across
industries in the United States by roughly 5 percent.
Similar work for Europe finds the impact of density to
be comparable (4.5 percent).
Table 1. Percentage of urban areas around world’s major
regions (United Nations, 2014)
Region 1950 1970 2014 2050
Africa 14.1 23.5 40 56
Asia 17.5 23.7 48 64
Europe 51.3 62.8 73 82
Latin America 41.4 57.1 80 86
North America 63.9 73.8 81 87
Oceania 62.4 71.2 71 74
Currently, urban areas generate 60-80 percent
of all greenhouse gas emissions and consume 75
percent of natural resources (UNEP, 2012). Therefore,
it can be asserted that urbanization has enormous
environmental consequences, putting pressure on
 Energy sources analysis from the perspective of sustainable development
energy and water resources; thus it can be stated that
sustainable cities are going to be vital for the future, as
urbanization expands.
3. Fundamental natural resources
Natural resources are regarded as “fixed
endowments”, being used to fuel industrial
economies. Environmental pollution became an
important global concern because of the rapid growth
of industrialization and urbanization. The intensive
exploitation of fundamental natural resources (water,
air, land, ecosystems) contributed in disturbing the
ecological balance and the environmental quality
(Fulekar, 2010). In this section the main energy
sources are being discussed.
3.1. Energy production sources – generalities
Smalley (2004) stated that the growing energy
demand of the world is going to need a new sustainable
energy source. The risk of catastrophic climate disruption
increases due to anthropogenic impact on the environment,
especially the changes of the Earth atmosphere
composition (the concentration of CO2, N2O, CH4).
Several studies report that CO2 is the primary
greenhouse gas emitted due to human activities. Fossil
fuels combustion in the electricity generation process is
the largest source of CO2 emissions in the world (IEA,
2011). Globally, the atmospheric CO2 concentration for
July 2014 was 399 ppm far above the maximum CO2
concentration of about 280 ppm registered before the
Industrial Revolution in the 19th century; today’s rate of
increase is more than hundred times faster than the
increase that took place when the last ice age ended
(Scripps Institution of Oceanography, 2014).
Nonetheless, globally, the energy systems are
at different stages of development and also at different
stages of achieving sustainability. According to the
World Energy Council (WEC) the energy
sustainability is based on three core dimensions:
energy security, social equity and environmental
impact mitigation (WEC, 2012). During the last
century, the world energy consumption has increased
significantly. Table 2 shows the world energy
consumption during the last four decades in relation
with the world’s population. Also, presented are data
regarding the world carbon dioxide emissions due to
oil, gas and coal consumption and the main energy
carriers. In order to highlight the importance of
Table 2. Global energy consumption [tons oil equivalent] and CO2 Emissions (BP, 2011)
Population
Year
[109 inh.] [106 tons oil equivalent]
1970 3.7
1980 4.45
1990 5.28
2000 6.08
2010 6.9
*
equivalent/inhabitant
energy, a recent statistic (United Nations, 2014)
regarding world energy consumption in 2011 per
world’s regions is presented in Table 3. It can be
noticed that world's largest energy consumer is Asia
Pacific region, accounting for 40.5 percent of the
global energy consumption and over 70 percent of the
global coal consumption. Asia Pacific was also an
undisputed leader in oil and hydroelectric
consumption. Europe and Eurasia is the leading region
for nuclear power consumption, natural gas and
renewable energy (BP, 2014). Table 3 shows a
significant disparity between the highest per capita
energy consumption region and the lowest, reflecting
disparities of development between various regions in
the world.
According to IEA (2011), in the last three
decades the per capita energy consumption increased
about 10 percent while world population increased 27
percent. Middle East recorded the most significant
growth of energy consumption per capita, with 170
percent. By 2035 in industrialized regions, such as
North America, the energy use per capita is estimated
to decline at an annual average of 0.5 percent per year
because the population rate growth is going to be
higher than the energy use rate. In developing regions
such as East and South-East Asia, by 2035, the energy
use per capita is estimated to rise at an annual average
of 3 percent per year (IEA, 2011).
According to Observ’ER (2013), the structure
of electricity production from renewable sources in
2012, was as following: the main source was
hydroelectricity with a 78% share; wind power
reclaimed the second place summing 11.4% of the
renewable total. It is essential to mention that in 2002
wind power generated was 52.5 TWh and in 2012 the
same source generated 543.3 TWh, an increase of over
ten times. These were followed by biomass power
(6.9%), solar power (2.2%), geothermal power (1.5%)
and marine energies (0.01%).
3.2. Energy production sources analysis
In 2012 (Observ’ER, 2013) world energy
production summed 22613 TWh, from which the
renewable electricity production weighted in with
4699.2 TWh, representing 20.8% of 2012 yearly total.
Fossil fuels generated 15394.3 TWh, representing
about 68% of global electricity production. The other
conventional source, the nuclear fission generated
about 11% of electricity production.
Energy consumption CO2 Emissions [106
[tons oil eq./inh.*] tons CO2]
4945.3 1.337 14992.5
6624 1.489 19322.4
8108.7 1.536 22613.2
9382.4 1.543 25576.9
12002.4 1.739 33158.4
1731
 Rusu et al./Environmental Engineering and Management Journal 17 (2018), 7, 1729-1741

Table 3. World energy consumption at the end of 2013 (BP, 2014; United Nations, 2014)

Source of energy
C [t oil
Region Natural Nuclear Hydro-
Oil Coal Renewable * Total equivalent/capita]
gas energy electricity
Europe and Eurasia 878.6 958.3 508.7 263 201.3 115.5 2925.3 3.33
North America 1024.2 838.6 488.4 213.7 156.3 65.4 2786.7 5.78
Africa 170.9 111 95.6 3.1 25.7 1.7 408.1 0.36
Latin America 311.6 151.8 29.2 4.7 158.1 18.3 673.5 1.34
Middle East 384.8 385.5 8.2 0.9 5.7 0.2 785.3 3.38
Asia Pacific 1415 575.2 2696.5 77.8 308.7 78.2 5151.5 1.28
Total 4185.1 3020.4 3826.7 563.2 855.8 279.3 12730.4 1.76
Consumption by
32.87 23.73 30.06 4.42 6.72 2.19 100 -
fuel [%]
*
includes wind, geothermal, solar, tidal energy etc.

Fig. 1 presents in points A, B, C and D, the 2013) and is expected to increase by close to 18
global cumulative installed wind power, geothermal, million barrels/day over the period 2014-2035,
solar and nuclear capacity. reaching almost 107.3 million barrels/day by 2035
(OPEC, 2012). Coal is currently the second largest
3.2.1. Fossil fuel energy source of primary energy worldwide. Likewise, coal is
Currently, fossil fuels represent the cornerstone powering the largest growing economies because it is
of global electricity production with more than two- widely distributed across the globe, abundant,
thirds of the total. Fossil fuels grew between 2002 and accessible and it is also the least subsidized of all fuel
2012 from a 64.99 percent share to a 68.08 percent sources (WEC, 2013). Natural gas is expected to play
share (Observ’ER, 2013). The main fossil fuel an increasingly important role in meeting the energy
reserves are unevenly distributed on Earth due to the demand in the world. At the moment there is an
geologic history of the areas. This is presented in important potential for shale gas, but are considerable
Table 4. The global oil demand was estimated to uncertainties about the size of the resources and the
average 89.7 million barrels/day in 2013 (OPEC, economics of developments (OPEC, 2012).

66,000
282,587

Electric capacity [MWe]

50,583
Direct uses capacity [MWth]
198,001

CAPACITY [MW]
CAPACITY [MW]

120,624

27,825
73,938

15,200
47,620

11,224
10,906
8,664
31,100

8,064

8,933
7,072

7,972
17,400

6,833
5,834
10,200

4,794
1,950
1,300

2,110
6,100
3,490
2,510

1,180

1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 1975 1980 1985 1990 1995 2000 2005 2010 2012
YEAR YEAR

(a) (b)
328,000

340,000

349,000

368,000

375,000

364,000

Solar heat [GWth] 268

250,000

PV electricity [GW] 245

INSTALLED CAPACITY [MW]
Capacity [GW]

196
172
135,000

147 152
128
98.4 111 102
71,000

69.7
47.6 55.2
16,000

40
4,800

23.2
9.4 15.8
830

2.2 2.8 3.9 5.3 6.9

2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2012

Year YEAR
(c) (d)
Fig. 1. Global cumulative installed wind power, geothermal, solar and nuclear capacity (modified after GWEC, 2014; GEA, 2013;
IEA, 2011-2013; EPIA, 2014; IAEA, 1960-2012): (a) Global cumulative installed wind power capacity; (b) Global geothermal
installed capacity; (c) Global solar installed capacity; (d) World nuclear installed capacity

1732
 Energy sources analysis from the perspective of sustainable development
Table 4. Distribution of fossil fuel reserves on Earth in 2013 (BP, 2014)
Oil
Region
[109 t] [%]
Europe and Eurasia 19.9 8.8
North America 35 13.7
Africa 17.3 7.7
Latin America 51.1 19.5
Middle East 109.4 47.9
Asia Pacific 5.6 2.5
Total 238.2 100
The global natural gas demand is estimated to
increase by 2.4% per year to reach 3962 billion cubic
meters in 2018, at a barely slower rate than coal (2.6%
per year) but faster than oil (0.7% per year) (IEA,
2013). Despite the shale gas revolution in the United
States where over the past years took place a switch
from coal to natural gas in power generation, the use
of coal is expected to rise by over 50 percent by 2030,
especially in developing countries that face major
challenges in providing energy to their populations,
and is expected to replace oil as the world’s largest
source of primary energy (WEC, 2013).
3.2.2. Hydroelectric energy
Hydropower is a renewable energy source
based on the natural water cycle. To meet the
increasing world electricity demands the electricity
generation from hydropower produces about 16
percent of the world electricity. During 2012 an
estimated 30 GW of new hydropower capacity came
on line, about 51% of it was in China. In 2012 the
world production of hydroelectricity reached 3748
TWh/year, three times more than that of 1970s. Table
5 presents the hydropower technical potential in terms
of installed capacity, undeveloped potential for 2012
and estimated for 2035.
Hydropower is the largest current source of
renewable energy in the electricity sector with a
significant potential. It is noticed that the percentage
of undeveloped potential range from 18.35% in
Europe to 93.28% in Africa, indicating a lot of
opportunities for hydropower development in the
world.
3.2.3. Wind energy
Wind energy has been used by humankind for
thousands of years, at the beginning it was used to
provide mechanical energy and currently to provide
electricity. It has been estimated that the available
wind power that can be converted into other forms of
energy is about 1.26x109 MW. Due to the fact that this
value represents 20 times the rate of the present global
energy consumption, wind energy could meet the
entire global needs (Tong, 2010). Fig. 1a shows the
evolution of global cumulative installed capacity of
wind power during the last two decades. A spectacular
increase of the cumulative installed capacity
worldwide can be noticed; in 2013 was 126 times
larger than that of 1992. Europe was the global leader
in wind energy technology use. Almost 39% of the
world capacity was installed in Europe by the end of
Natural gas Coal
[1012 m3] [%] [106 t] [%]
56.6 30.5 310538 34.8
11.7 6.3 245088 27.5
14.2 7.6 31814 3.6
7.7 4.1 14641 1.6
80.3 43.2 1122 0.1
15.2 8.2 288328 32.3
185.7 100 891531 100
2013; wind is meeting 8% of Europe’s electricity
demand (GWEC, 2014). In 2030 the global
cumulative capacity for installed wind power is
estimated to reach 917798 MW, Asia is expected to be
the global leader with a 41% share (GWEC, 2012).
3.2.4. Geothermal energy
Geothermal resources consist of thermal
energy from the Earth’s interior stored in both rock
and trapped steam or liquid water. Resource utilization
technologies for geothermal energy can be grouped as
types for electrical power generation, for direct use of
the heat, or for mixed heat and power in cogeneration
applications (Goldstein et al., 2011). In Fig. 1b is
presented the global evolution of geothermal power
and geothermal direct use during the last 37 years. A
significant development is observed. In 2012 the
installed electric capacity was almost ten times larger
than that of 1975, and the installed capacity of direct-
use geothermal energy in 2012 was 50 times more than
that of 1975s.
Asia is the leading market for geothermal
power having about 44 percent of the global
geothermal installed capacity (11765 MW) in August
2013 (GEA, 2013). By 2050 geothermal deployment
is estimated to generate 1180 TWhe/yr for electricity,
representing more than three percent of global
electricity demand, and 2100 TWhth/yr for heat
applications, representing about five percent of global
heat demand (Goldstein et al., 2011). Although the
world’s geothermal resources are immense and
ubiquitous, it is difficult to accurately determine the
global geothermal resource potentials, mostly because
the technologies used are continuously evolving (INL,
2006).
3.2.5. Solar energy
Solar energy is the most abundant of all energy
resource on Earth. Annually, the total amount of solar
radiation falling on Earth is more than 7500 times the
world’s annual primary energy consumption of 500 EJ
(WEC, 2010). The most widely used solar
technologies are solar photovoltaic (PV) systems and
solar thermal power plants. In Figure 1c it is shown,
globally, the evolution of solar photovoltaic systems
and solar thermal power plants during the last decade,
being noticed a substantial development. At the end of
2012, Asia (especially China) was the undisputed
global leader in solar thermal energy with a 67 percent
share, followed by Europe with 16 percent share (IEA,
2014).
1733
 Rusu et al./Environmental Engineering and Management Journal 17 (2018), 7, 1729-1741

Table 5. Regional hydropower technical potential and hydroelectricity generation (Kumar et al., 2011; Rajgor, 2013).

Installed capacity Technical 2012 2035

Undeveloped
World region at the end of 2012 potential, installed Total generation Total generation
potential [%]
[GW] capacity [GW] [TWh] [TWh]
Europe 276 338 18.35 745 1062
North America 189 388 51.28 664 771
Asia & Oceania 510 2104 75.76 1654 2320
Latin America 133 608 78.12 611 1054
Africa 19 283 93.28 74 274
Total 1127 3721 69.71 3748 5481

In terms of global cumulative installed PV current agriculture practice and productivity,

capacity, at the end of 2013, Europe was leading the
way with about 59 percent share, covering 3% of the
electricity demand. Nevertheless, in 2013, for the first
time in more than a decade, the European market for
PV lost its leadership in terms of new installations due
to ongoing rebalancing between Europe and the rest of
the world, and furthermore reflecting the pattern in
electricity consumption (EPIA, 2014).
3.2.6 Bioenergy
Bioenergy is one of the most widely used forms
of renewable energy in the world, providing over 10
percent of world’s energy supply (WEC, 2013).
Biomass is presently the most important renewable
energy carrier globally. It is used in the production of
electricity, heat and transportation fuels. The main
feedstock for electricity and heat biomass generation
comes from agricultural crops and residues, forestry,
wood-processing industries, and organic wastes. In
2012 the global biomass heating capacity was of about
293 GWth, being noticed an important growth in the
last decade, especially in Europe. At the end of 2012
the cumulative biomass power generation capacity
worldwide was approximately 83 GW (IEA, 2013).
There are multiple ways to quantify the global
biomass resource potential. Probably the most
important one is the technical potential, which takes
into account the requirements for maintaining the
economic, ecological and social value of ecosystems
(Chum et al., 2011). Table 6 presents the technical
potential of biomass plantation worldwide determined
by taking into account a specific set of assumptions
with respect to nature protection requirements, the
populations, diet, climate. Currently, it is noticed that
the largest technical potential comes from Africa with
over 40 percent share. Unlike industrialized countries
where the biomass use is average only about 3 percent
of the total primary energy, in developing countries,
biomass contributes about 22 percent to the total
primary energy mix, especially in Asia, Latin America
and Africa where are registered the highest rates of
urbanization in the world (WEC, 2013). Pappas et al.
(2012) mentioned that the main barriers to the
deployment of biomass are low conversion efficiency,
low feedstock availability, lack of supply logistics,
risks associated with intensive farming. The studies
arrived at the conclusion that the sustainable upper
bound of the global technical potential in 2050 can
amount to approximately 500 EJ (Dornburg et al.,
2010), which means there is an important potential to
expand biomass use, especially by increasing plant
productivity, land efficiency, water management and
sustainable use of residues and wastes for bioenergy.
3.2.7. Nuclear energy
Nuclear energy is produced as the result of
nuclear fusion, nuclear fission of a radioactive
element. The amount of energy released by the nuclear
fission of 1 kg of uranium is similar to that obtained
by the combustion of 3000 tons of coal or 14000
barrels of oil (Joesten et al., 2007). In Fig. 1d it is
shown that the global cumulative installed capacity for
nuclear power during the last six decades. It is noticed
a spectacular capacity increase in the 1970s and 1980s,
as countries sought to reduce dependence on fossil
fuels, especially after the oil crisis of the 1970s.

Table 6. Technical potential of biomass plantations worldwide (Chum et al., 2011; Fisher et al., 2009)

Total grass- and woodland Unproductive or very low productive Bioenergy area Technical Potential
Region
area [Mha] areas [Mha] [Mha] [EJ/yr]
Europe and Russia 902 618 122 17
North America 659 391 111 19
Africa 1086 386 275 69
Latin America 765 211 160 45
SE Asia 556 335 14 4
Pacific OECD 515 332 97 17
Middle East and N
107 93 1 0.2
Africa
World 4605 2371 780 171

1734
 Energy sources analysis from the perspective of sustainable development
The growth stagnated in the 1990s due to
increased concerns about safety following the
Chernobyl accident, and even more in the 21st century
due to the Fukushima accident (Hernández- Ceballos
et al., 2012), but also due to a return to lower fossil
fuels prices. By the year 2030, world nuclear capacity
is expected to grow between about 540 GW net in low
demand case and 746 GW net in the high demand case,
especially in Asia (OECD, 2012). Currently, nuclear
and hydropower are the only low-carbon sources,
providing significant amounts of energy. Energy
generated from nuclear plants avoids annual CO2
emissions of about 2.9 billion tones compared to coal-
fired generation (IEA, 2010a, c). Although a generally
high security standard is required as accidents can
happen with devastating consequences for human
beings and for the ecosystem. Therefore deployment
of nuclear energy should be preceded by complex
comparative analyses of all available options (Vujic et
al., 2012).
4. Energy production and environmental impact
assessment
A major objective of sustainable development
is to guarantee environmental quality. However, an
ideal energy production technology that is at the same
time environment-friendly, risk-free and cost-
effective does not exist. Every technology deployment
implies some trade-offs to be made, in order to ensure
optimal use of energy resources, while restricting
environmental and health impacts (Vujic et al., 2012).
The goal of this section is to review the
environmental impact assessment for water, air, land
and ecosystem for different types of electricity
generation sources.
4.1. Environmental impact assessment for water in
case of energy production
Currently, worldwide, industry accounts for 20
percent of all water consumption, compared to 70
percent for agriculture and 10 percent for domestic use
(UNESCO, 2012).
The global average water footprint related to
consumption was 1385 m3∕y per capita over the period
1996–2005 (Hoekstra and Mekonnen, 2012); the
water footprint is one of the most used indexes when
analyzing the relationship between the total amount of
fresh water that is used to produce the goods and
services consumed by the inhabitants of a nation.
However, freshwater withdrawals have tripled over
the last 50 years. Globally, the annual population
increase of 80 million per year implies an increased
demand for freshwater of approximately 64 billion
cubic meters a year. Considering that there would be
an additional 1850 cubic kilometers of water
consumption by 2030, 30 percent higher than today’s
level, more sustainable water use programs have to be
implemented, especially in the regions with rapid
population growth and with growing water stress
(MGI, 2011). Energy and water are intricately
connected. Electricity generating technologies use
water for different production processes throughout
the entire lifecycle. For assessing the impact regarding
freshwater use two aspects are analyzed: “withdrawal”
(the water removed from the ground or diverted from
a water source for use) and “consumption” (the water
evaporated, transpired, incorporated into products or
crops, or otherwise removed from the immediate
water environment) (Kenny et al., 2009).
The electrical energy necessary for economic
growth requires massive withdrawals of water. Table
7 presents the median lifecycle uses of water for
conventional and renewable electricity generation
technologies.
Table 7. The average water footprint for different
electricity generation technologies (adapted upon Fthenakis
and Kim, 2010; Macknick et al., 2011; Wilson et al., 2012)
Electric fuel Water footprint [l/MWh]
Coal 63382
Hydroelectric 1699649
Natural gas 26986
Nuclear 68191
Geothermal 5823
Solar PV 882
Wind 234
Biomass 46984
Wind and PV technologies are by far the most
“water-friendly” energy options (their lifecycle
median water usage ranging from 234 to 882 l per
MWh produced) while hydropower, coal and nuclear
have the largest water use (ranging from 63382 to
almost 1.7 million l per MWh).
Considering that most of the energy is
generated by conventional technologies that require
large volume of water flows, energetic efficiency and
water conservation programs are essential for
protecting the freshwater resources from the impact of
electricity production. Instead of altering the site’s
hydrology and ecosystems, technologies like solar PV
and wind, which are highly sustainable with respect to
water consumption, should be implemented at larger
scale (Wilson et al., 2012).
4.2. Environmental impact assessment of air in case of
energy production
This section reviewed available estimates for
major air pollutants that are emitted by renewable and
non-renewable electricity generation sources. These
include CO2 emissions, nitrous oxides (NOx), sulphur
dioxide (SO2) and particulate matter (PM) emission.
All these cause negative impacts on human health and
the ecosystems’ quality.
Lifecycle greenhouse gas emissions intensity
in electricity generation depends on the source used. It
is necessary to be evaluated not only the emissions
exhausted during the generation of electricity, but also
emissions from upstream processes (fuel exploration,
1735
 Rusu et al./Environmental Engineering and Management Journal 17 (2018), 7, 1729-1741
mining, fuel transport, plant construction), as well as
downstream processes (decommissioning, waste
management and disposal), in order to assess the
impact of the power generation sources (Weisser,
2007). Fig. 2 illustrates a comparison of greenhouse
gas emissions, normalized in tones of CO2 equivalent
per GWh, for different electricity generation sources.
Fig. 2. Lifecycle greenhouse gas emissions from different
electricity generation sources (adapted upon Chum et al.,
2011; Goldstein et al., 2011; Hsu et al., 2012; Stacey and
Garvin, 2012; Weisser, 2007; WNA, 2011)
It is noticed that fossil fuel sources generate the
highest greenhouse gas emissions on a lifecycle basis,
ranging from 469 tones CO2/GWh (natural gas) to
1001 tones CO2/GWh (coal). Most of lifecycle
greenhouse gas emissions for fossil fuels sources are
registered during the operation of the power plant. The
emissions throughout downstream activities are
insignificant but during upstream activities may be
released considerable emissions depending on
processes involved in extraction, fuel transportation
and fuel preparation (Weisser, 2007). Fossil fuel based
electricity production can be a major source of air
pollution: NOX (lifecycle emissions ranging from 0.4
g/kWh (natural gas) to 4 g/kWh (oil)), SO2 (lifecycle
emissions ranging from 5 g/kWh (natural gas) to 27
g/kWh (coal)), PM (lifecycle emissions ranging from
0.02 g/kWh (natural gas) to 2.3 g/kWh (coal)) (Bauer,
2008; Sathaye et al., 2011).
For geothermal plants, the greenhouse gas
emissions are less than 50 tones CO2/GWh for flash
steam plants and less than 80 tones CO2/GWh for
projected EGS (Enhanced Geothermal System) plants
(Goldstein et al., 2011). Lifecycle NOX emissions are
lower than 0.4 g/kWh. SO2 and PM emissions are
negligible (Sathaye et al., 2011).
PV systems generate negligible greenhouse gas
emissions during the electricity generation. However,
about 70 percent of emissions arise during upstream
processes, especially in the system components
manufacture (Hsu et al., 2012). Solar power
equipment releases 50–1000 times less direct Hg
emissions than traditional electricity generation
(Turney and Fthenakis, 2011). In terms of NOx, SO2
and PM emissions are negligible (Sathaye et al.,
2011).
Lifecycle greenhouse gas emissions from
bioenergy systems range between about 16 and 74
1736
tons of CO2/GWh depending on the combustion
efficiency, power rate, type of feed. The majority of
emissions arise during the upstream processes (Chum
et al., 2011). Lifecycle NOX emissions are about 5
g/kWh, SO2 emissions are approximately 2.5 g/kWh
and PM emissions are about 0.5 g/kWh (Bauer, 2008;
Sathaye et al., 2011). For nuclear energy, the majority
of greenhouse gas emissions occur upstream of
operation, depending on the enrichment process. NOX,
SO2 and PM emissions are negligible (Sathaye et al.,
2011). Lifecycle greenhouse gas emissions from
hydropower projects range between 1-34 tones
CO2/GWh, depending on type of the plant (reservoir
or run-off), its size and usage as well as the electricity
mix used for its operations (Weisser, 2007). Lifecycle
NOX, SO2 and PM emissions are negligible (Sathaye
et al., 2011). Wind energy is currently the most
environmental friendly technology with the median of
greenhouse gas emissions at 12 tones CO2/GWh, 83
times less than coal fired power plants (Stacey and
Garvin, 2012). Lifecycle NOX, SO2 and PM emissions
are negligible (Sathaye et al., 2011).
Hence, for reducing greenhouse gas emission
from electric power generation is required a
substantial increasing of the share of total electricity
generated from renewable source and improving
efficiency of existing power plants (Scripps Institution
of Oceanography, 2014).
4.3. Environmental impact assessment for land use in
case of energy production
In order to assess the land use impact of the
power generation sources, this has to be evaluated not
only the land on which the power plant is situated, but
also the impact of the full land use life cycle. Land
transformation and land occupation are the main
metrics used in the literature to describe and compare
land requirement for different energy production
sources. The former indicates the area of land altered
from an original state, while the latter involves the
duration over which the area of the transformed land
returns to its reference state, measured as a product of
land area (m2) and time (year) (Koellner and Scholtz,
2007). Table 8 summarizes the land transformation
and land occupation for power production using
different conventional and renewable sources
(Fthenakis and Kim, 2009).
It is noticed that land use impacts vary widely
between and within sources, also depending on
regional conditions. From Table 8, it is noticed that, as
concerns the non-renewable sources, coal-fuel cycle
affects the land transformation pattern the most. The
mining, beneficiation, and electricity-generation
stages can involve destruction of existing landscape,
soil erosion, land subsidence, water contamination,
land acidification, removal of vegetative cover and
topsoil (Fthenakis and Kim, 2009; Miranda et al.,
2003; Spath et al., 1999).
As regards renewable sources, hydroelectric
power plants affect the pattern of land transformation
the most (Fthenakis and Kim, 2009). Nuclear power
 Energy sources analysis from the perspective of sustainable development
entails the lowest land transformed per unit of
electricity. However, land transformed accidentally by
disasters in the nuclear-fuel cycle can have dramatic
consequences. In November 2011, the Japanese
Science Ministry reported that the destruction of the
Fukushima Daiichi nuclear power plant contaminated
more than 30000 square kilometers of land with
radioactive cesium (Ishizuka, 2011).
Table 8. Land transformation and land occupation
for electricity production from different sources
(Fthenakis and Kim, 2009)
Energy Land transformation Land occupation
source [m2/GWh] [m2/GWh]
Coal 239-489 1290-25200
Natural
265 4200
gas
Nuclear 109-124 300000
Solar PV 164-552 9900
Wind 1030-3230 2040
Hydro 2350-25000 100000
Biomass 12550 380000
For non-renewable fuel cycles, such as coal
and nuclear energy, the land occupation is dominated
by upstream and downstream processes. The full
recovery of coal-mined lands is expected to be several
hundred years (it would take 200–300 years before the
composition of native forest to be reestablished
(Burger, 2002).
As shown in Table 8, bioenergy requires the
greatest land occupation (380000 m2 year/GWh)
because of the need to cultivate feedstock. It has to be
noticed that inadequate land use change due to biofuel
production can aggravate soil and vegetation
degradation associated with overexploitation of
forests, too intensive crop and forest residue removal
(Koh and Ghazoul, 2008; Robertson et al., 2008).
Wind power-plants entail the lowest land occupation
per unit electricity. Moreover, wind farms can be
located on low quality lands (e.g., brownfields or other
contaminated site), and often be used for a variety of
other productive purposes, including livestock
grazing, agriculture, highways, and hiking trails (Mai
et al., 2012; NREL, 2010). The assessment of
environmental impacts of land transformation and
occupation is complex. Although hydroelectric energy
and bioenergy require large amounts of land, the other
renewable technologies, such as wind energy and solar
PV, involve similar or less disturbance of land than
non-renewable sources (Fthenakis and Kim, 2009).
4.4. Environmental impact assessment for the
ecosystem in case of energy production
In energy systems around the world changes,
which will have implications for ecosystems and
livelihoods take place (Boyle et al., 2003). All energy
sources affect the ecosystem. Although renewable
sources are significantly safer, they can have a
negative impact on the ecosystem. The constructions
of fossil fuel-fired power plants, as well as oil
refineries, destroy habitats for animals and plants,
affecting the biodiversity, due to the inability of many
species to survive in such conditions (Wake, 2005).
Hydrocarbons-based energy sources impact
maritime ecosystems through different ways. Seabird
populations are vulnerable to the effects of oil and gas
pollution, especially from the surface oil pollution.
Fish and shellfish populations, benthic communities
and plankton are vulnerable to impacts from
hydrocarbon pollution, especially from hydrocarbon
spills (BP, 2010; Karacik et al., 2009; Mosbech,
2000). Studies have shown that new road building,
pipeline construction and expanding drilling
operations disrupt migration, habitat, and wintering
grounds for many wildlife species (EPA, 2008; Riley
et al., 2012). Coal mining causes extensive
degradation to natural ecosystems, reducing the
number of tree species in the mined areas and
impairing the plant diversity (Sarma et al., 2010). The
environmental impacts of coal mining-derived runoff
(e.g., acid drainage) are severe, large scale, and long
lasting, especially on aquatic fauna and flora
(Bernhardt and Palmer, 2011; Ochieng et al., 2010).
Hazardous air pollutants emitted from fossil fuels-
fired power plants deposit into rivers, lakes and oceans
and are taken up by underwater fauna and flora. These
pollutants lead to bioaccumulation of toxic metals,
environmental acidification, contamination of aquatic
life, can harm plants and the wildlife (EH&E, 2011).
Hydropower, usually, requires the reservoir
formation which involves special biotope and habitat
loss due to inundation, climate change and can cause
the relocation of communities living nearby the lake
(Salvarli, 2006). When the oxygen-poor water from
the bottom of the lake is released through a dam into
the river, downstream habitat condition change (e.g.,
it can kill fish that are accustomed to warmer, oxygen-
rich water) (Hamilton, 2013). The construction of
onshore wind farms alters ecosystems through the
clearing of vegetation and can lead to habitat loss and
fragmentation, especially on forested ecosystems
(Katsaprakakis, 2012).
Wind energy developments cause fatalities of
birds and bats through collision, being responsible for
0.003 percent of bird mortalities caused by human
activities (EWEA, 2009). The construction of offshore
wind farms may disturb benthic communities and
marine mammals, changing the existing biodiversity
in the area and creating a new local ecosystem
(EWEA, 2009). The construction of a geothermal
plant and roads alters ecosystems through the clearing
of vegetation. During well testing, due to the noise,
wildlife in the vicinity of the drill rig can be
moderately affected (Noorollahi and Yousefi, 2003).
Solar PV may cause disturbance during the
installation stage. Also the PV panels cast shadows
and may change the microclimate, causing an
unstudied effect on vegetation (Turney and Fthenakis,
2011). Studies revealed that central concentrator
power systems could pose a danger to birds
(OECD/IEA, 1998), and flying insects can also be
burnt when flying in the vicinity of the reflector’s area,
1737
 Rusu et al./Environmental Engineering and Management Journal 17 (2018), 7, 1729-1741
but the loss of the insect population is negligible
(Tsoutsosa et al., 2005).
Diversion of crops into bioenergy production
can influence food commodity prices and food
security (Headey and Fan, 2008). Growing demand for
bioenergy put pressure on the degradation and loss of
more remote habitats and landscapes (Firbank, 2008).
The insertion of invasive or genetically modified
species causes the decline of biodiversity, especially if
forest, peat, grass and wetlands are used for feedstock
production and if large monoculture plantations are
created (CBD, 2008; Sathaye et al., 2011). Regarding
nuclear energy, the loss of biodiversity is frequently
associated with contamination due to uranium mining
and milling processes. The setting up of a nuclear plant
disrupts the ecological balance of the region (Bond et
al., 2003; Teixeira et al., 2012).
5. Conclusions
This paper has provided an overview of the
current state of world natural resources used to
produce energy, and the energy consumption degree
across different regions of the world, as well as the
main conventional and renewable energy technologies
as regards specific sustainable factors. Global natural
resources have been diminishing over the years,
mainly due to factors such as fast growing population
and industrialization. Worlds’ economy requests ever-
increasing amounts of energy, contributing thus to an
increasing use of natural resources.
Taking into account the identified fossil and
uranium resources, at current rates of consumptions,
these are sufficient for a few decades of supply for
global power plants; most of the energy is generated
by conventional sources that have significant
environmental impact, and because of this, the energy
transition from fossil fuels to renewable sources of
energy should be urgently embarked upon.
Sustainable development requires adequate
methods to measure and compare the environmental
impacts for energy production technologies. An ideal
energy production technology does not exist; each has
its advantages and disadvantages. Renewable
technologies, excepting hydropower, are by far the
most “water-friendly” energy options; thereby their
increasing utilization could reduce the stress in water
resource availability in many regions of the globe.
Fossil fuels generate the highest greenhouse gas
emissions on a lifecycle basis, especially during the
operation of the power plant, being responsible for
global warming, one of the major environmental
problems that the humanity is currently facing. Wind
energy, on the other hand, has minimum impact on the
environment in terms of greenhouse gas emissions.
A substantial decreasing of greenhouse gas
emissions cannot be obtained if renewable energy
resources are not enhanced in share of generation mix.
As regards land transformation, hydroelectric plants
mostly affect the pattern of land transformation.
Despite having the minimum impact on the
environment in terms of land transformation, nuclear
1738
technologies are not universally considered
sustainable due to the environmental and health
hazards they bring along, as shown from the recent
accident in Fukushima Daiichi, which seems to have
halted their expansion.
Unlike bioenergy which requires the greatest
land occupation, wind power plants entail the lowest
land occupation per unit electricity. Finally, in terms
of impact on the ecosystem, studies have shown that
the renewable technologies are considerably safer than
the conventional ones.
References
ASCE, (2010), Policy Statement 418 - The Role of the Civil
Engineer in Sustainable Development, American
Society of Civil Engineers, On line at:
http://www.asce.org/Public-Policies-and-Priorities.
Axinte E., Vasies G., Teleman E.C., Rosca V.E., Axinte
A.O., (2015), Including solar panels on parking areas in
the sustainable development concept. Wind effects,
Environmental Engineering and Management Journal,
14, 2255-2262.
Bauer C., (2008), Life Cycle Assessment of Fossil and
Biomass Power Generation Chains, Paul Scherrer
Institut, Villigen, Switzerland.
Bernhardt E.S., Palmer A.M., (2011), The environmental
costs of mountaintop mining valley fill operations for
aquatic ecosystems of the Central Appalachians, Annals
of the New York Academy of Sciences, 1223, 39-57.
BNEF, (2014), Global trends in clean energy investment –
Q4 2013 fact pack, Bloomberg New Energy Finance,
On line at: www.bnef.com.
Bond A., Bussell M., O' Sullivan P., Palerm J., (2003),
Environmental impact assessment and the
decommissioning of nuclear power plants - a review
and suggestion for a best practicable approach,
Environmental Impact Assessment Review, 23, 197-
217.
Boyle G., Everett R., Ramage J., (2003), Energy for a
Sustainable Future, Oxford University Press, Oxford.
BP, (2010), Environmental Impact Assessment Scoping
Report, QD-BP-EV-REP-0011, British Petroleum, On
line at: www.bp.com.
BP, (2011), Statistical review of world energy, British
Petroleum, On line at: www.bp.com/statisticalreview.
BP, (2014), BP Statistical Review of World Energy June
2014, British Petroleum, On line at:
www.bp.com/statisticalreview.
Bruntland G., (1987), Our Common Future: The World
Commission on Environment and Development, Oxford
University Press, Oxford, 1987.
Burger J.A., (2002), Achieving forestry goals on mined
land: Mine site conditions, species composition, and
stand development of economically viable hardwoods
forests, Proc. of the Market-Based Approaches to
Mined Land Reclamation and Reforestation: A
Technical Interactive Forum, vol. 1, Fort Mitchell,
Kentucky, 15-16 May 2002, 149-160, On line at:
http://www.mcrcc.osmre.gov/MCR/Resources/reforest
ation/Market_Based_Mine_Land_Reclamation.pdf.
Casper J. K., (2010), Fossil Fuels and Pollutions: The
Future of Air Quality, Infobase Publishing, New-York.
CBD, (2008), The potential impacts of biofuels on
biodiversity - matters arising from SBSTTA
recommendation XII/7, UNEP/CBD/COP/9/26,
Convention on Biological Diversity, On line at:
 Energy sources analysis from the perspective of sustainable development
http://www.cbd.int/doc/meetings/cop/cop-
09/official/cop-09-26-en.pdf.
Chalvatzis K.J., Hooper E., (2009), Energy security vs.
climate change: theoretical framework development
and experience in selected EU electricity markets,
Renewable Sustainable Energy Reviews, 13, 2703–
2709.
Chum H., Faaij A., Moreira J., Berndes G., Dhamija P.,
Dong H., Gabrielle B., Goss Eng A., Lucht W., Mapako
M., Masera Cerutti O., McIntyre T., Minowa T.,
Pingoud K., (2011), Bioenergy, In: IPCC Special
Report on Renewable Energy Sources and Climate
Change Mitigation, Edenhofer O., Pichs-Madruga R.,
Sokona Y., Seyboth K., Matschoss P., Kadner S.,
Zwickel T., Eickemeier P., Hansen G., Schlomer S.,
von Stechow C. (Eds.), Cambridge University Press,
Cambridge, 209-332.
Dornburg V., van Vuuren D., van de Ven G., Langeveld H.,
Meeusen M., Banse M., van Oorschot M., Ros J., van
den Born G. J., Aiking H., Londo M., Mozaffarian H.,
Verweij P., Lysen E., Faaij A., (2010), Bioenergy
revisited: Key factors in global potentials of bioenergy,
Energy & Environmental Science, 3, 258-267.
EH&E, (2011), Emissions of hazardous air pollutants from
coal-fired power plants, Environmental Health &
Engineering Inc. Report 17505, Needham.
EPA, (2008), An Assessment of the Environmental
Implications of Oil and Gas Production: A Regional
Case Study, Environmental Protection Agency,
Washington D.C, On line at:
https://archive.epa.gov/sectors/web/pdf/oil-gas-
report.pdf.
EIA, (2013), International Energy Outlook 2013, U.S.
Energy Information Administration, Washington DC,
On line at:
https://www.eia.gov/outlooks/ieo/pdf/0484(2013).pdf.
EPIA, (2014), Global Market Outlook for Photovoltaics
2014-2018, European Photovoltaic Industry
Association, Renewable Energy House, Brussels, On
line at: https://resources.solarbusinesshub.com/solar-
industry-reports/item/global-market-outlook-for-
photovoltaics-2014-2018.
EWEA, (2009), Wind energy – the facts: a guide to the
technology, economics and future of wind power,
European Wind Energy Association, Earthscan,
London.
Firbank L.G., (2008), Assessing the Ecological Impacts of
Bioenergy Projects, BioEnergy Research, 1, 12–19.
Fisher G., Hizsnyik E., Prieler S., Shah M., van Velthuizen
H., (2009), Biofuels and Food Security, The OPEC
Fund for International Development (OFID) and
International Institute of Applied System Analysis
(IIASA), Vienna, On line at:
http://www.ofid.org/LinkClick.aspx?fileticket=4Pjzcr
L3584%3D&tabid=112&mid=552.
Fthenakis V., Kim H.C., (2009), Land use and electricity
generation: a life-cycle analysis, Renewable and
Sustainable Energy Reviews, 13, 649-668.
Fthenakis V., Kim H.C., (2010), Life-cycle uses in U.S.
electricity generation, Renewable and Sustainable
Energy Reviews, 14, 2039-2048.
Fulekar M.H. (Ed.), (2010), Bioremediation Technology for
Hazardous Wastes – Recent Advances, Springer,
Heidelberg.
GEA, (2013), 2013 Geothermal Power: International
Market Overview, Geothermal Energy Association,
Washington D.C.
Goldstein B., Hiriart G., Bertani R., Bromley C., Gutierrez-
Negrin L., Huenges E., Muraoka H., Ragnarsson A.,
Tester J., Zui V., (2011), Geothermal Energy, In: IPCC
Special Report on Renewable Energy Sources and
Climate Change Mitigation, Edenhofer O., Pichs-
Madruga R., Sokona Y., Seyboth K., Matschoss P.,
Kadner S., Zwickel T., Eickemeier P., Hansen G.,
Schlomer S., von Stechow C. (Eds.), Cambridge
University Press, Cambridge, 401-436.
Grübler A., Fisk D., (2012), Energizing Sustainable Cities:
Assessing Urban Energy, Routledge, Abingdon.
GWEC, (2012), Global Wind 2011 Report, Global Wind
Energy Council, Brussels.
GWEC, (2014), Global Wind 2013 Report, Global Wind
Energy Council, Brussels.
Hamilton M.S., (2013), Energy Policy Analysis: A
Conceptual Framework, 2nd ed., M.E. Sharpe, New
York.
Headey D., Fan S., (2008), Anatomy of a crisis: the causes
and consequences of surging food prices, Agricultural
Economics, 39, 375-391.
Henderson J. V., (2003), Urbanization in developing
countries, World Bank Research Observer, 17, 89-112.
Hernández-Ceballos M. A., Hong G. H., Lozano R. L., Kim
Y. I., Lee H. M., Kim S. H., Yeh S.-W., Bolívar J. P.,
Baskaran M., (2012), Tracking the complete revolution
of surface westerlies over Northern Hemisphere using
radionuclides emitted from Fukushima, Science of the
Total Environment, 438, 80-85.
Hoekstra A.Y., Mekonnen M.M., (2012), The Water
Footprint of Humanity, Proceedings of the National
Academy of Sciences, 109, 3232–3237.
Hoekstra A.Y., Chapagain A. K., (2007), Water footprint of
nations: Water use by people as a function of their
consumption pattern, Water Resources Management,
21, 35-48.
Hsu D.D., O’Donoughue P., Fthenakis V., Heath G. A., Kim
H. C., Sawyer P., Choi J.-K., Turney D. E., (2012), Life
cycle greenhouse gas emission of crystalline silicon
photovoltaic electricity generation, Journal of
Industrial Ecology, 16, S122-S135.
Hughes J., Cain L.P., (2003), American Economic History,
6th Edition, Addison Wesley, Boston.
IAEA, (1960-2012), International Atomic Energy Agency
Bulletin, On line at: www.iaea.org.
IEA, (2010a), Energy Technology Perspectives 2010.
Scenarios & Strategies to 2050, International Energy
Agency, Paris, On line at:
https://www.iea.org/publications/freepublications/publ
ication/etp2010.pdf.
IEA, (2010b), Technology Roadmap. Nuclear Energy,
International Energy Agency Publications, Paris, On
line at:
https://www.iea.org/media/freepublications/technolog
yroadmaps/TechnologyRoadmapNuclearEnergy.pdf
IEA, (2010c), Projected Costs of Generating Electricity,
OECD/International Energy Agency, Paris, On line at:
https://www.oecd-nea.org/ndd/pubs/2005/5968-
projected-costs.pdf.
IEA, (2011), World Energy Outlook 2011, International
Energy Agency, Paris, On line at:
https://www.iea.org/publications/freepublications/publ
ication/weo2011_web.pdf.
IEA, (2013), Medium-Term Gas Market Report 2013:
Market Trends and Projections to 2018, International
Energy Agency, International Energy Agency
Publications, Paris, On line at:
http://www.iea.org/publications/freepublications/publi
cation/MTGMR2013_free.pdf.
IEA, (2014), Solar Heat Worldwide, International Energy
Agency – Solar Heating & Cooling Programme, On line
1739
 Rusu et al./Environmental Engineering and Management Journal 17 (2018), 7, 1729-1741
at: https://www.iea-
shc.org/data/sites/1/publications/Solar-Heat-
Worldwide-2016.pdf.
INL, (2006), The Future of Geothermal Energy, INL/EXT-
06-11746, Idaho National Laboratory, Massachusetts
Institute of Technology, Cambridge, On line at:
http://mitei.mit.edu/system/files/geothermal-energy-
full.pdf.
Ishizuka H., (2011), Radioactive cesium blankets 8% of
Japan’s land area, Asahi Shimbun, November 21, On
line at:
http://ajw.asahi.com/article/0311disaster/fukushima/A
J201111210014.
Joesten M., Castellion M.E., Hogg J.L. (2007), The World of
Chemistry, 4th Edition, Thomson Brooks/Cole,
Belmont.
Karacik B., Okay O.S., Henkelmann S., Bernhöft.K- B.
Schramm W., (2009), Polycyclic aromatic
hydrocarbons and effects on marine organisms in the
Istanbul Strait, Environment International, 35, 599–
606.
Katsaprakakis D.A., (2012), A review of the environmental
and human impacts from wind parks. A case study for
the Prefecture of Lasithi, Crete, Renewable and
Sustainable Energy Reviews, 16, 2850– 2863.
Kenny J.F., Barber N.L., Hutson S.S., Linsey K.S., Lovelace
J.K., Maupin M.A., (2009), Estimated Use of Water in
the United States in 2005, U.S. Geological Survey
Circular 1344, Reston, On line at:
http://pubs.usgs.gov/circ/1344/pdf/c1344.pdf.
Koh L.P., Ghazoul J., (2008), Biofuels, biodiversity, and
people: Understanding the conflicts and finding
opportunities, Biological Conservation, 141, 2450-
2460.
Koellner T., Scholz R.W., (2007), Assessment of land use
impacts on the natural environment. Part 1: an
analytical framework for pure land occupation and land
use change, The International Journal of Life Cycle
Assessment, 12, 16-23.
Kumar A., Schei T., Ahenkorah A., Caceres Rodriguez R.,
Devernay J.-M., Freitas M., Hall D., Killingtveit A., Liu
Z., (2011), Hydropower, In: IPCC Special Report on
Renewable Energy Sources and Climate Change
Mitigation, Edenhofer O., Pichs-Madruga R., Sokona
Y., Seyboth K., Matschoss P., Kadner S., Zwickel T.,
Eickemeier P., Hansen G., Schlomer S., von Stechow
C. (Eds.), Cambridge University Press, Cambridge,
437-496.
Macknick J., Newmark R., Heath G., Hallett K.C., (2011),
A Review of Operational Water Consumption and
Withdrawal Factors for Electricity Generating
Technologies, NREL/TP-6A20-50900, National
Renewable Energy Laboratory, Golden, On line at:
http://www.nrel.gov/docs/fy11osti/50900.pdf.
Mai T., Wiser R., Sandor D., Brinkman G., Heath G.,
Denholm P., Hostick D. J., Darghouth N., Schlosser A.,
Strzepek K., (2012), Exploration of High-Penetration
Renewable Electricity Futures, Vol. 1 of Renewable
Electricity Futures Study, NREL/TP-6A20-52409-1,
Golden, CO: National Renewable Energy Laboratory,
On line at: http://www.nrel.gov/docs/fy12osti/52409-
1.pdf.
MGI, (2011), Resource Revolution: Meeting the World’s
Energy, Materials, Food and Water Needs, McKinsey
Global Institute, London.
Michalena E, AngeonV., (2009), Local challenges in the
promotion of renewable energy sources: the case of
Crete, Energy Policy, 37, 2018–2026.
1740
Miranda M., Burris P., Bincang J.F., Shearman P., Briones
J.O., Viña A.L., Menard S., (2003), Mining and Critical
Ecosystems: Mapping the Risks, World Research
Institute, Washington, DC., On line at:
http://defiendelasierra.org/descargas/mining_critical_e
cosystems_full.pdf.
Mosbech A., (2000), Predicting Impacts of Oil Spills – Can
Ecological Science Cope? A case study concerning
birds in Environmental Impact Assessments, National
Environmental Research Institute, Denmark.
NREL, (2010), Brownfields' Bright Spot: Solar and Wind
Energy, National Renewable Energy Laboratory, On
line at:
http://www.nrel.gov/news/features/feature_detail.cfm/
feature_id=1530.
Noorollahi Y., Yousefi H., (2003), Preliminary
Environmental Impact Assessment of a Geothermal
Project in Meshkinshahr, NW-Iran, Proc. of the
International Geothermal Conference, Reykjavík, S12,
1-11.
Observ’ER, (2013), Energy production in the world: general
forecasts, Worldwide Electricity Production from
Renewable Energy Sources: Stats and Figures Series:
Fifteenth Inventory – Edition 2013, On line at:
http://www.energies-renouvelables.org/observ-
er/html/inventaire/Eng/methode.asp
Ochieng G.M., Seanego E.S., Nkwonta O.I., (2010) Impacts
of mining on water resources in South Africa: A review,
Scientific Research and Essays, 5, 3351-3357.
OECD/IEA, (1998), Benign energy? The Environmental
Implications of Renewable, Organisation for Economic
Co-operation and Development, OECD/IEA, Paris.
OECD, (2010), Cities and Climate Change, OECD
Publishing, Paris.
OECD, (2012), Uranium 2011: Resources, Production and
Demand, A Joint Report by the OECD Nuclear Energy
Agency and the International Atomic Energy Agency,
Paris.
OPEC, (2012), World Oil Outlook 2012, Organization of the
Petroleum Exporting Countries, On line at:
http://www.opec.org/opec_web/static_files_project/me
dia/downloads/publications/WOO2012.pdf.
OPEC, (2013), Monthly Oil Market Report – October 2013,
Organization of the Petroleum Exporting Countries, On
line at:
www.opec.org/opec_web/static_files_project/media/d
ownloads/publications/MOMR_October_2013.pdf.
Pappas C., Karakosta C., Marinakis V., Psarras J., (2012), A
comparison of electricity production technologies in
terms of sustainable development, Energy Conversion
and Management, 64, 626-632.
Rajgor G., (2013), China continues its dominance of the
global hydropower market, Renewable Energy Focus,
14, 26-27.
Riley T. Z., Bayne E. M., Dale B. C., Naugle D. E., Rodgers
J. A., Torbit S. C., (2012) Impacts of crude oil and
natural gas developments on wildlife and wildlife
habitat in the Rocky Mountain region. The Wildlife
Society Technical Review 12-02. The Wildlife Society,
Bethesda, Maryland, USA., On line at:
http://wildlife.org/documents/technical-
reviews/docs/Oil%20and%20Gas%20Technical%20R
eview_2012.pdf.
Robertson G.P., Dale V.H., Doering O.C., Hamburg S.P.,
Melillo J.M., Wander M.M., Parton W.J., Adler P.R.,
Barney J., Cruse R.M., Duke C.S., Fearnside P.M.,
Follett R.F., Gibbs H.K., Goldemberg J., Mladenoff
D.J., Ojima D., Palmer M.W., Sharpley A., Wallace L.,
 Energy sources analysis from the perspective of sustainable development
Weathers K.C., Wiens J.A., Wilhelm W.W., (2008),
Sustainable biofuels redux, Science, 322, 49-50.
Rosenthal S., Strange W.C., (2004), Evidence on the Nature
and Sources of Agglomeration Economies, In:
Handbook of Regional and Urban Economics,
Henderson J.V., Thisse J.F. (Eds.), vol. 4, Elsevier,
Amsterdam, 2119-2171.
Salvarli H., (2006), Some aspects on hydraulic energy and
environment in Turkey, Energy Policy, 34, 3398-3401.
Sarma K., Kushwaha S.P.S., Singh K.J., (2010), Impact of
coal mining on plant diversity and tree population
structure in Jaintia Hills district of Meghalaya, north-
east India, New York Science Journal, 3, 79-85.
Sathaye J., Lucon O., Rahman A., Christensen J., Denton F.,
Fujino J., Heath G., Kadner S., Mirza M., Rudnick H.,
Schlaepfer A., Shmakin A., (2011), Renewable Energy
in the Context of Sustainable Development, In: IPCC
Special Report on Renewable Energy Sources and
Climate Change Mitigation, Edenhofer O., Pichs-
Madruga R., Sokona Y., Seyboth K., Matschoss P.,
Kadner S., Zwickel T., Eickemeier P., Hansen G.,
Schlomer S., von Stechow C. (Eds), Cambridge
University Press, Cambridge, 707-790.
Scripps Institution of Oceanography, (2014), Atmospheric
CO2 for July 2014, On line at: www.co2now.org.
Shamsuddin A., Tasbirul I., Mohd A.K., Nissar M.K.,
(2014), Exploitation of renewable energy for
sustainable development and overcoming power crisis
in Bangladesh, Renewable Energy, 72, 223-235.
Sheau-Ting L., Weng-Wai C., Mohammed A.H., (2016),
Determinants of energy conservation intention:
evidence from Malaysian universities, Environmental
Engineering and Management Journal, 15, 231-243.
Smalley R. E., (2004), Our energy challenge, On line at:
http://americanenergyindependence.com/energychalle
nge.aspx.
Spath P.L., Mann M.K., Kerr D.R., (1999), Life Cycle
Assessment of Coal-fired Power Production,
NREL/TP-570-25119, National Renewable Energy
Laboratory, Golden, On line at:
http://www.nrel.gov/docs/fy99osti/25119.pdf.
Stacey L.D., Garvin A.H., (2012), Life cycle greenhouse gas
emissions of utility-scale wind power, Journal of
Industrial Energy, 16, S136-S154.
Tantau A.D., Chinie A.C., Borisov D., (2017), Triggering
events for corporate entrepreneurship leading to
investments in renewable energy, Environmental
Engineering and Management Journal, 16, 2857-2868.
Teixeira T.P., Neves L.M., Araujo F.G., (2012), Thermal
impact of a nuclear power plant in a coastal area in
Southeastern Brazil: effects of heating and physical
structure on benthic cover and fish communities,
Hydrobiologia, 684, 161-175.
Tong W., (2010), Wind Power Generation and Wind
Turbine Design, WIT Press, Southampton.
Tosovic R., Dasic P., Ristovic I., (2016), Sustainable use of
metallic mineral resources of Serbia from an
environmental perspective, Environmental Engineering
and Management Journal, 15, 2075-2084.
Tsoutsosa T., Frantzeskaki N., Gekas V., (2005),
Environmental impacts from the solar energy
technologies, Energy Policy, 33, 289–296.
Turney D., Fthenakis V., (2011), Environmental impacts
from the installation and operation of large-scale solar
power plants, Renewable and Sustainable Energy
Reviews, 15, 3261–3270.
UNEP, (2012), Sustainable, Resource Efficient Cities –
Making it Happen!, United Nations Environment
Programme, Paris.
UNESCO, (2012), The United Nation World Water
Development Report 4, The United Nations
Educational, Scientific and Cultural Organization,
Paris.
United Nations, (2014), Department of Economic and Social
Affairs, Population Division. World Urbanization
Prospects: The 2014 Revision, New York.
Vujic J., Antic D.P., Vukmirovic Z., (2012), Environmental
impact and cost analysis of coal versus nuclear power:
The U.S. case, Energy, 45, 31-42.
Wake H., (2005), Oil refineries: a review of their ecological
impacts on the aquatic environment, Estuarine, Coastal
and Shelf Science, 62, 131–140.
Weisser D., (2007), A guide to life-cycle greenhouse gas
(GHG) emissions from electric supply technologies,
Energy, 32, 1534-1559.
Wilson W., Leipzig T., Griffits-Sattenspiel B., (2012),
Burning Our Rivers: The Water Footprint of
Electricity, River Network, Portland, On line at:
http://climateandcapitalism.com/wp-
content/uploads/sites/2/2012/06/Burning-Our-
Water.pdf.
WNA, (2011), World Nuclear Association Report:
Comparison of Lifecycle Greenhouse Gas Emissions of
Various Electricity Generation Sources, World Nuclear
Association, London.
WEC, (2010), 2010 Survey of Energy Resources, World
Energy Council, London.
WEC, (2012), Time to Get Real – the case for sustainable
Energy Policy, World Energy Council, London.
WEC, (2013), World Energy Resources: 2013 Survey,
World Energy Council, London.
Zaharia C., (2004), Energy and Environment (in Romanian),
“A.I. Cuza” University Press, Iaşi, Romania
1741
1742
Environmental Engineering and Management Journal July 2018, Vol.17, No. 7, 1743-1752
http://www.eemj.icpm.tuiasi.ro/; http://www.eemj.eu

“Gheorghe Asachi” Technical University of Iasi, Romania

HEAVY METALS CONTENT IN THE SPROUTS OF Glyceria maxima

(Hartm.) Holmb. AND IN RIVER SEDIMENTS (NORTHERN POLAND)

Agnieszka Parzych1∗, Małgorzata Cymer2

1Environmental
Chemistry Research Unit, Institute of Biology and Environmental Protection,
Pomeranian Academy in Slupsk, Arciszewskiego 22b St., 76-200 Slupsk, Poland
2Department of Land Reclamation and Environmental Development, University of Warmia and Mazury in Olsztyn,

Plac Łódzki 2, 10 - 719 Olsztyn, Poland

Abstract

The aim of this study was to evaluate the content of Zn, Mn, Cu and Ni in the aboveground and underground shouts of Glyceria
maxima and in the bottom sediments of the Słupia River. These studies allow for assessment of the existing and potential hazards
resulting from toxic influence of heavy metals on water environment and human health. The concentration of research elements
was determined by atomic absorption spectrometry (AAS). Results indicate that the concentration of heavy metals in the examined
bottom sediments remained within the limits of the geochemical background for majority of the determined elements. The
concentrations of Zn and Cu, sporadically exceeded the level of the geochemical background at the stations in the central part of
the city, and in the case of Ni in all researched positions. Following the LAWA classification, the bottom sediments within the
Słupsk area were classified in the first class, as slightly contaminated. It was found the G. maxima underground sprouts accumulated
several times more of Zn, Mn, Cu and Ni than their aboveground sprouts. The indices of quality of the tested bottom sediments
and the enrichment factors of the sprouts with heavy metals indicate that the Słupia River is the least contaminated along the
segments of the river with a low or moderate transformation of the river bed and the largest load of Zn, Mn, Cu and Ni is found at
the stations with a high transformation of the river bed. The strong positive correlations between the concentrations of Mn in bottom
sediments and the level this element in aboveground and underground sprouts indicate that the sprouts of Glyceria maxima are
potentially useful in biomonitoring environmental contamination with Mn.

Key words: bottom sediments, heavy metals, indicators of quality of sediments, macrophytes

Received: February, 2014; Revised final: October, 2014; Accepted: November, 2014; Published in final edited form: July 2018

1. Introduction 2006). Natural processes and industrial and

Due to their common character of application
in different branches of economy, heavy metals are
hazardous to the natural environment (Baptista Neto
et al., 2000; Tam and Wong, 2000; Tang et al., 2010).
As natural components of ecosystems, they are
necessary in minute quantities for correct functioning
of organisms. However, their excessive concentration
in the environment is harmful. At relatively high
concentrations, they disturb functioning of
ecosystems, being harmful for living organisms
(Hlihor et al., 2009; Obolewski and Glińska-Lewczuk,
∗
Author to whom all correspondence should be addressed: e-mail: parzycha1@op.pl, Phone: +48 59 84 05 347
agricultural activity constitute the source of heavy
metals. At present, the intensity of use of river
catchments as well as the inflow of sewage have
substantial impact on the chemical composition of
sediments (Salati and Moore, 2009). It leads to
increase of heavy metal content in fluvial deposits
which are important elements of the water
environment. Their quantity often exceeds the
concentration provided by the geochemical
background and is a strong indicator of rivers
pollution (Klavins et al., 2000; Kobus et al., 2016).
Most heavy metals introduced to rivers are connected
 Parzych and Cymer/Environmental Engineering and Management Journal 17 (2018), 7, 1743-1752
and transported with the suspended matter whose
deposition leads to origination of bottom sediments in
surface waters, both with fast and slower flow. A part
of such chemical compounds is deposited at bottom
sediments (Kruopiene, 2007).
Heavy metals are biologically transformed and
can be maintained in bottom sediments for a long time
in tens or hundreds of years (Bettinetti et al., 2003; Liu
et al., 2009). When accumulated in bottom sediments,
they can constitute secondary sources of pollution
which are still harmful to water ecosystems (Yuan et
al., 2004). The quality of sediments is a good indicator
of pollution of water environment, as well as the
quantity of emitted pollutants in the river catchment
(Klavins et al., 2000; Vandecasteel et al., 2004;
Wiśniowska-Kielian and Niemiec, 2005). Tests of
benthos revealed that the polluted sediments can cause
acute and chronic toxic effects water organisms,
especially fish (Obolewski and Glińska-Lewczuk,
2006; Obolewski et al., 2016).
The pollution of bottom sediments has impacts
on the reduction of many species of water fauna which
are important from the point of view of exploitation
and ecology, such as snails and fish. Heavy metals are
biologically accumulated in plant and animal tissues,
so the toxic hazard increases at subsequent levels of
the trophic chain (Klink et al., 2013; Kurilenko and
Osmolovskaya, 2006; Loska and Wiechuła, 2003).
Plants react in various ways to increased concentration
of heavy metals in the environment. Absorption and
bioaccumulation of trace elements by plants is a part
of the natural cycle. In biological identification of
water ecosystems, the plants of coastal zone
(macrophytes) are used, which, along with
phytoplankton, macrozoobenthos and fish, constitute
basic elements of ecological evaluation of the status of
rivers (Klink et al., 2013, Obolewski and Glińska-
Lewczuk, 2006; Skorbiłowicz, 2003; Skorbiłowicz
and Skorbiłowicz, 2011). Macrophytes reflect the
status of heavy metal pollution of water reservoirs
very well (Samecka-Cymerman and Kempers, 1996,
Miretzky et al., 2004, Aksoy et al., 2005).
The heavy metals content in aquatic plants may
exceed many times their content in the surrounding
water environment, the wide range of variability is
caused by biology and ecology of particular species.
Some aquatic plants are used in environmental
research (Salt and Kramer 2000, Baldantoni et al.,
2004, Mishra and Tripathi, 2008). Glyceria maxima
plays the role of bioindicator of water pollution
(Tanner 1996, Sriyaraj and Shutes 2001, Rabajczyk
and Jóźwiak 2009, Teuchies et al., 2013). Control of
the chemical composition of bottom sediments and
aquatic plants in municipal areas is very important due
to the harmfulness of heavy metals (Dauvalter and
Rognerud, 2001; Xiangdong et al., 2000). These
studies allow for specification of the existing and
potential hazards resulting from toxic influence of
heavy metals on water environment and human health.
The research aimed at (1) evaluation of the
content of Zn, Mn, Cu and Ni in above-ground and
1744
underground sprouts of Glyceria maxima as well as in
the bottom sediments of the Słupia River, (2)
specification of accumulation coefficients of the
above mentioned heavy metals in sprouts of Glyceria
maxima, and (3) evaluation of the quantity of the
bottom sediments within the area of the city of Słupsk
from the point of view of the concentration of selected
heavy metals in them.
2. Material and methods
2.1. Research area
The Słupia River is situated in the central part
of Pomerania (Northern Poland). It is a lowland
watercourse 138.6 km in length and with a catchment
area of 1620 km2. In the north the Słupia River
catchment borders the Baltic Sea catchment; in the
west, the Wieprza River catchment; in the South, the
Brda River catchment, and from the east, Łeba and
Łupawa. The Słupia River headwater is at the
Kashubian Lake District close to Sierakowska Huta at
the height of 178 m a.s.l.
The river flows to the Baltic Sea in Ustka. The
width of the river bed varies from 7 m in the upper part
of the river to 40 m at its mouth, where the average
flow is 15.5 m³.s-1. The river stream is strongly
meandered (Piechura et al., 1997). The area of the City
of Słupsk covers a 8 km stretch of the Słupia River
with various levels of transformation of the water bed
(low, moderate or high) (Table 1) (Obolewski and
Glińska-Lewczuk, 2006), whose banks are covered
with numerous macrophytes.
2.2. Research methods
The research was done in July 2013 at the area
of 10 stations situated within the city limits. Samples
of bottom sediments and aboveground sprouts (leaves,
stems) and underground ones (rootstock) of Glyceria
maxima, (Fig. 1) were taken at the river bank zone of
the Słupia River where the sedimentation of the
suspended material takes place. The bottom sediments
were collected with the use of the Eckman sampler
from the depth of 0-15 cm.
After their transport to laboratory, the samples
were dried at a temperature of 65oC, they were sieved
through a 1mm and ground in a mortar. Before
analysis, the deposit samples were tested as to the
content of CaCO3 (with 0,1M HCl). In the bottom
sediments, the granulometric composition was marked
by means of the gradation test, as well as active acidity
(pH, H2O), exchange acidity (pH, KCl), and the
organic matter content by the method of heat loss in a
muffle furnace at the temperature 550oC. The samples
of Glyceria maxima at the area of each station were
taken for testing from several sprouts creating mixed
samples both from aboveground and underground
sprouts. After their transport to laboratory, the plant
material was cleaned of mineral parts of the soil and
flushed in distilled water.
 Heavy metals content in the sprouts of Glyceria maxima (Hartm.) Holmb. and in river sediments (Northern Poland)

Table 1. Characteristic of sampling points

The degree of The dominant of macrophytes

Type
Stations transformation Glyceria Typha Phragmites Phalaris
ground
river bed maxima latifolia australis arundinacea
1 low silt + + +
2 low silt + + +
3 low silt + +
4 moderate sand + + + +
5 high stone + + +
6 high stone + + + +
7 high stone + + + +
8 moderate sand + +
9 moderate stone + +
10 low silt +

Fig. 1. Sediment sampling points

The aboveground sprouts (leaves, stems) were
separated from the underground ones (rootstock),
dried to stable mass at the temperature of 65oC, and
then homogenized in a laboratory grinder. The content
of Zn, Mn Cu and Ni were determined by AAS method
(Perkin Elmer, Aanalyst 300) in a solution after
mineralization in a mixture of the concentrated HNO3
and 30% H2O2. The analyses were performed in the
oxy-acetylene flame according to Ostrowska et al.
(1991). The tests were carried out following the
original standards (Merck KGaA,1g/1000ml).
2.3. Elaboration of results
The results obtained were verified by means of
statistic methods using Statistica software (7.1.). The
values of minimum, maximum, mean, standard
deviation, Spearman correlation coefficients and
coefficients of variation (CV) are presented.
Classification of bottom sediments was made in
according with the classification of LAWA (1998),
which divides the sediments on the purity class as the
increasing content of heavy metals (Table 2). It is
assumed that the metal content in the dry matter of
bottom sediments for classes I and I-II at the level of
the geochemical background, while class II to IV
indicate to the increasing influence of anthropogenic
pollution of bottom sediments of rivers, and thus of the
whole aquatic ecosystem.
For evaluation of the level of contamination of
sediments, classification of water sediments in Poland
on a basis of geochemical criteria was used
(Bojakowska and Sokołowska, 1998). The indexes
calculated were pollution load index (PLI),
Geoaccumulation index (Igeo) and enrichment factor
(EF) according to Eqs. (1-4).

1745
 Parzych and Cymer/Environmental Engineering and Management Journal 17 (2018), 7, 1743-1752

Table 2. Classification of LAWA over 50%, and medium grained sand (0.5-0.25 mm) –
over 23%. At the research stations 5, 6, 7 and 9 small
Purity class quantities of gravel (7.9 %) and stone (0.5 %) were
I I-II II II- III III- IV found. At all stations, the lowest contribution had a
III IV fraction of dust (<0.05 mm) – average 1.2 %, (Fig. 2).
Zn ≤ ≤ ≤ ≤ ≤ ≤ >
The tested sediments did not contain CaCO3.
100 200 400 800 1600 3200 3200
Cu ≤ ≤ ≤ ≤ ≤ ≤ >
They were characterized by the a slightly alkaline
20 40 80 160 320 640 640 reaction both in the case of pH (H2O), and pH (KCl),
Ni ≤ ≤ ≤ ≤ ≤ ≤ ≤ (Table 3).
30 60 120 240 480 960 960 The average reaction was from 7.00 to 8.19,
Class I Uncontaminated reflecting variability at the level of 4.6 % within the
Class I-II Uncontaminated / Moderately area of 10 research stations. The content of organic
contaminated matter in the bottom sediments of the Słupia River was
Class II Moderately contaminated substantially diverse with the levels from 0.56 to 7.94
Class II-III Moderately contaminated / Heavily %, and variation coefficient was 117.0 % (Table 3).
contaminated
Class III Heavily contaminated
Class III- Heavily / Very heavily contaminated
IV
Class IV Very heavily contaminated

PLI = n CF1 ×CF2 × CF3 × ... × CFn (1)

where:
n – number of analyzed positions;
CF – pollution factor, (Chakravarty and Patgiri, 2009)

Ce
CF = (2)
(Ce )EC
Fig. 2. Granulometric composition sediments of the River
Ce Słupia, point (mean), rectangle (standard deviation),
I geo = (3) whiskers (minimum-maximum)
1.5(Ce )EC
The highest level of organic matter content was
 Ce  in the samples of sediments from stations 4 and 10,
  characterized by strong river meanders which
 C Al  S
EF = (4) welcomed substantial accumulation. The lowest
 Ce  quantities of accumulated organic matter (<1%) were
 
 C Al  EC
found at the stations 2, 3, 5, 6 and 8, with a prevailing
process of free transport of organic matter down the
river stream.
where:
Zinc content in the bottom sediments of the
(Ce)s - concentration of elements in the sample,
Słupia River varied substantially (CV= 82.2 %) with
(CAl)s – concentration of aluminum in the sample,
levels from 12.4 to 104.7 mg∙kg-1, average 34.1 mg∙kg-
(Ce)EC – geochemical background 1
(Table. 4). In 25 % of tested samples, the zinc content
was between 20 and 25 mg∙kg-1 (Fig. 3). The zinc
3. Results and discussion
concentration in the river sediments within the city
limits of Słupsk in 2013 exceeded the value of
The bottom sediments in the Słupia River were
geochemical background (>48 mg∙kg-1) only at
characterized by diverse granulometric composition.
stations 4 and 9.
Fine sand constituted the largest part (0.25-0.1 mm) –

Table 3. The organic matter content and pH of bottom sediments

pH pH Organic
(H2O) (KCl) matter
Average ± SD 7.84 ± 0.36 7.66 ± 0.40 2.14 ± 2.51
Median 7.95 7.79 0.91
Minimum 7.00 6.68 0.56
Maximum 8.19 8.08 7.94
CV, % 4.6 5.2 117.0

1746
 Heavy metals content in the sprouts of Glyceria maxima (Hartm.) Holmb. and in river sediments (Northern Poland)

The natural content of Zn in sediments of Table 4. Heavy metals (mg∙kg-1) contents in bottom
surface water (lakes and rivers) in Poland is about 74 sediments of the river Słupia
mg∙kg-1 (Lis and Pasieczna, 1995). As a geochemical
background of zinc in the bottom sediments of the Zn Mn Cu Ni
Vistula River and its tributaries, the value of 110 Average ± 34.1 ± 152.1 ± 8.4 ± 11.9 ±
SD 28.0 111.5 4.4 2.8
mg∙kg-1 (Helios-Rybicka, 1991) was adopted.
Median 22.8 119.0 7.5 11.7
A slightly different type of aquatic ecosystems Minimum 12.4 58.1 2.6 8.5
of rivers and lakes are old river beds. This is Maximum 104.7 376.7 15.1 16.8
particularly evident in the assessment of the CV, % 82.2 73.3 52.8 23.5
concentration of individual metals in bottom
sediments, which is usually more. Concentration of Zn Only at two stations (4, 10) Mn content
in the bottom sediments of old river beds of the Słupia exceeding 300 mg∙kg-1 was discovered (Fig. 3).
River was on average 142.6 mg∙kg-1 and resulted from Obtained manganese contents in bottom sediments of
accumulation and the lack of water outflow the Słupia River did not exceed the level of
(Obolewski and Glińska-Lewczuk, 2013). Zinc is a geochemical background (500 mg∙kg-1). Slightly
metal commonly introduced to surface waters along higher manganese concentrations were found in the
with municipal sewage and surface flow. Burning of old river beds of the Słupia River – 645 mg∙kg-1
coal is a source of zinc in the environment along with (Obolewski and Glińska-Lewczuk, 2006), which
use of zinc covered water pipes. Zn concentration in could have been a result of accumulation connected
bottom sediments of the Słupia River does not have with the lack of outflow of water.
negative impact on living organisms. According to the The Oder River bottom sediments contained
classification Lawa (1998) the Słupia River sediments 770 mg∙kg-1, and the Rhine River deposits even 960
should be classified to the 1st quality class as to the mg∙kg-1Mn (Kucharzewska et al., 1991). Manganese
zinc content – uncontaminated sediments (Table 2). of the bottom sediments is characterized by high
Manganese content was between 58.1 and mobility (Kabata-Pendias and Pendias, 1999), which
376.7 mg∙kg-1, average 152.1 mg∙kg-1. The manganese is especially stimulated by pH~8. The main source of
concentration variability coefficient in the bottom manganese in environment is washing out of rocks and
deposits was 73.3 % (Table. 4). For 50 % of the soils. Anthropogenic sources comprise mainly coal
examined stations Mn content was from 50 to 100 and gasoline.
mg∙kg-1.

(a) (b)

(c) (d)
Fig. 3. Histograms of heavy metals contents in bottom sediments of River Słupia. Black line – values of geochemical background

1747
 Parzych and Cymer/Environmental Engineering and Management Journal 17 (2018), 7, 1743-1752
Copper content in bottom sediments of the
Słupia River was on average 8.4 mg∙kg-1 with the
values ranging from 2.6 to 15.1 mg∙kg-1. Cu
concentration showed variability at the level of 52,8%
within the area of research stations (Table 4). The
copper content in the sediments was low and was
within the limits of the geochemical background (6
mg∙kg-1) except the stations 4, 5, 6, 7, 9 and 10. The
highest content of Cu (15.1 mg∙kg-1) was found at
station 5, where substantial transformation of the river
bed took place (concrete covering), (Fig. 3). Slightly
higher levels of Cu (34.9 mg∙kg-1) were revealed in
bottom sediments of the old river beds of the Słupia
River by Obolewski and Glińska-Lewczuk (2006),
which could have resulted from the increased
accumulation of organic matter caused by the lack of
water outflow (Bettinetti et al., 2003; Liu et al., 2009).
The copper content in the water sediments is
connected with the kind of the parent rock and usually
is several mg∙kg-1. According to Lis and Pasieczna
(1995) the natural content of Cu in the river sediments
usually does not exceed 20 mg∙kg-1. Contamination of
environment with copper, including river sediments,
takes place mainly due to mining and processing of
copper ore, but can also be a result of applied
pesticides, additives to fodder in agriculture and a
result of burning fossil fuels. Discharge of untreated
sewage from galvanizing plants and their improper
storage makes ions of copper penetrate the bottom
waters and sediments (Bojakowska and Sokołowska,
1998). Due to its small content in tested sediments
(Cu<36 mg∙kg-1), copper has no negative impact on
living organisms.
Nickel content in the bottom deposits of the
Słupia River was from 8.5 to 16.8 mg∙kg-1, average
11.9 mg∙kg-1. The concentration variation coefficient
of Ni was the lowest from among tested heavy metals
and was 23,5 % (Table 4). The largest quantity of
nickel (>15 mg∙kg-1) was found at the stations 4 and
10, as well as in the case of manganese (Fig. 3). At all
stations, the content of Ni exceeded the level of the
geochemical background (>5 mg∙kg-1), but according
to classification Lawa (1998) it belonged to the class I
of the clarity (<30 mg∙kg-1), Table 2. Higher Ni
content was found in the sediments of the old river
beds of the Słupia River – 32.2 mg∙kg-1 (Obolewski
Table 5. The correlation coefficient between the researched parameters (n = 30, p <0.05, rcrit. = 0.30) in the bottom sediments of
Słupia
Organic
matter
Organic matter -
pH (H2O) -0.78
pH (KCl) -0.81
Zn 0.29
Mn 0.90
Cu 0.53
Ni 0.91
rcrit. - critical values of Spearman’s correlation coefficient referred to Ramsey (1989), bold values are statistically significant
1748
and Glińska-Lewczuk, 2013), and most probably, it
was an effect of increased accumulation of organic
matter and the lack of water outflow (Bettinetti et al.,
2003; Liu et al., 2009). The natural nickel content in
the bottom sediments of the Vistula River was 40
mg∙kg-1 (Helios-Rybicka, 1991), and in the case of
other rivers of Poland, about 10 mg∙kg-1 (Lis and
Pasieczna, 1995). Burning of coal and liquid fuels is a
source of nickel in the natural environment. Effluents
from the municipal and industrial dump sites
constitute additional sources of nickel contamination
of water sediments, as a result of dumping of
accumulators and nickel-cadmium batteries and
discharge of waste from galvanizing plants used
catalysts. A series of statistically significant
relationships were found among the tested parameters
(Table 5).
Along with the increase of organic matter
content in bottom sediments, decrease of pH was
observed (r=-0.78 and r=-0.81, p<0.05, n=30) together
with increase of concentration of Mn, Cu and Ni
(r=0.90, r=0.53 and r=0.91, p<0.05, n=30). Active
acidity and exchange acidity were also significantly
related to one another. Increase of acidity had a
significant impact on release of ion forms of Mn, Cu
and Ni from bottom sediments. It was also found, that
with the increase of concentration of Zn and Mn, an
increase of Cu concentration was found. A positive
and statistically vital correlation was also found
between Mn and Ni and Cu and Ni (respectively r=
0.90 and r=0.47, p<0.05, n=30), (Table 5).
The values of geo-accumulation indexes were
between: 0<Igeo<1, which according to a 7-grade
classification of bottom sediments LAWA (1998),
Eqs. 1, 3 and 4 shows moderate contamination of the
bottom sediments of the Słupia River in relation to Zn,
Mn, Cu and Ni. The values of enrichment factors EF,
show the lack of contamination of bottom sediments
of the Słupia River in case of Zn and Mn (EF<1) and
a slight contamination with copper and nickel (Table
6). Enrichment factors (EF>1) in the case of Cu and
Ni reveal anthropogenic contamination with these
elements. Also the indexes of pollution load (PLI)
show slight contamination of the Słupia River bottom
sediments with Zn and Mn (PLI<1) and Cu and Ni
(PLI<2.33), (Table 6).
pH pH Zn Mn Cu Ni
H2O KCl
-
0.98 -
-0.28 -0.24 -
-0.65 -0.67 0.21 -
-0.44 -0.40 0.41 0.43 -
-0.76 -0.75 0.10 0.90 0.47 -
 Heavy metals content in the sprouts of Glyceria maxima (Hartm.) Holmb. and in river sediments (Northern Poland)

Table 6. Quality indicators of bottom sediments of the
Słupia River
Zn Mn Cu Ni
PLI 0.57 0.25 1.21 2.33
Igeo 0.0099 0.0004 0.1548 0.3167
EF 0.71 0.30 1.39 2.38
The heavy metal content in the tissues of the
water plants and in the bottom sediment of the rivers
reflects the impact of anthropomorphic pressure on
natural environment. The zinc content was from 17.9
to 53.0 mg∙kg-1 in aboveground sprouts and from 15.8
to 119.5 mg∙kg-1 in underground sprouts of Glyceria
maxima, showing variability at the level of 43.3 % and
56.0 % (Table 7). To cover physiological needs of
most plants, the concentration in leaves at the level of
15-30 mg∙kg-1 is sufficient (Kabata-Pendias and
Pendias, 1999). The plants absorb Zn in the quantities
proportional to its concentration in soil (Kozanecka et
al. 2002). An average zinc content in aboveground
parts of the plants which are not under the influence of
pollution is 10-70 mg∙kg-1 (Kabata-Pendias and
Pendias, 1999), which shows a permissible level of Zn
in aboveground sprouts of Glyceria maxima covering
the banks of the Słupia River.
Manganese content in Glyceria maxima was on
average 261.7 mg∙kg-1 in aboveground sprouts and
441.2 mg∙kg-1 in underground ones. Variation in
concentration at 10 stations, at 68.4% and 86.8%, was
much higher than in the case of Zn. According to
Kabata-Pendias and Pendias (1999), the demand of
plants for manganese varies and in most cases 10-25
mg∙kg-1 is sufficient. A concentration about 500
mg∙kg-1 can be toxic for most of the plants. Increased
Mn content in the sprouts of Glyceria maxima in
relation to physiological demand can show a positive
impact of that macrophyte on purification of water and
bottom sediments of the Słupia River. The main factor
deciding about the availability of heavy metals for
plants is the reaction of bottom sediments.
Solubility of metals is low as to neutral and
alkaline reactions (Table 3), and increases along with
lowering of the value of pH (Smal and Salomons,
1995). Increase of mobility of Zn and Mn is most
effective with pH=6, while Cu and Ni at pH=5.5.
Manganese, however, is characterized by increased
solubility also in alkaline environment (pH~8),
(Alloway, 1995, Martinez and Motto, 2000).
Cu content in aboveground and underground
sprouts of Glyceria maxima was 10.3 mg∙kg-1 and 21.1
mg∙kg-1 respectively and shows variability within the
area of 10 stations of 21.9% and 55.7% respectively
(Table 7). Copper in plants displays low mobility
(Kabata-Pendias and Pendias, 1999). The
physiological demand of most plants is below 4-5
mg∙kg-1 and is quite variable depending on the part of
the plant, its development stage, variety and species.
Its average content in the aboveground parts of the
plants is from 5-20 mg∙kg-1 (Kabata-Pendias and
Pendias, 1999). The average nickel content was 17.4
mg∙kg-1 and 24.9 mg∙kg-1 respectively in aboveground
and underground sprouts of Glyceria maxima.
Variability coefficients within the area of the research
stations were 25.8% and 18.5% respectively (Table 7).
Physiological nickel content in plants at the
area which was not under the impact of pollution was
most often: 0.1-5.0 mg∙kg-1 (Kabata-Pendias and
Pendias, 1999), and in municipal agglomerations such
values can be higher since nickel easily undergoes bio
accumulation especially in water plants which are
sensitive bio-indicators of waters (Sarosiek and
Wożakowska-Natkaniec, 1993). The slightly higher
Ni content than physiological demand in the sprouts
of Glyceria maxima therefore indicates some pollution
of bottom sediments with that element. Enrichment
factors (EF) of determination of heavy metals in
aboveground sprouts had the values lower than 1.9,
and in the underground sprouts < 3.3 (Table 7). All of
the determined heavy metals were accumulated much
easily in underground sprouts than in the aboveground
ones of Glyceria maxima.

Table 7. Heavy metals (mg∙kg-1) contents, coefficients of variation (%) and the enrichment factor (EF) in the aboveground
and underground sprounts of Glyceria maxima

Zn Mn Cu Ni
Glyceria maxima - aboveground sprouts
Average ± SD 29.9±12.7 261.7±179.1 10.3±2.3 17.4±4.5
Median 23.4 200.9 10.1 17.2
Minimum 17.9 98.2 7.6 11.9
Maximum 53.0 622.7 13.7 24.3
CV, % 43.3 68.4 21.9 25.8
EF 1.2 1.9 1.7 1.5
Glyceria maxima - underground sprouts
Average ± SD 59.3±33.2 441.2±383.1 21.1±11.7 24.9±4.6
Median 57.3 242.7 19.3 25.8
Minimum 15.8 130.3 12.3 17.8
Maximum 119.5 1276.3 51.5 30.3
CV, % 56.0 86.8 55.7 18.5
EF 2.5 3.1 3.3 2.2

1749
 Parzych and Cymer/Environmental Engineering and Management Journal 17 (2018), 7, 1743-1752
The values of enrichment factors (EF>1)
indicate slight contamination with the tested metals.
The relationships among the determined elements
made up the following decreasing series:
Mn>Zn>Ni>Cu.
Between the content of Zn, Mn, Cu and Ni in
the bottom sediments of the Słupia River and the level
of concentration of heavy metals in the sprouts of
Glyceria maxima, a series of statistically significant
relationships were determined. With the increase of
concentration of manganese in underground sprouts,
the Mn content increase in aboveground sprouts. A
reverse dependency was revealed in the case of
copper, which could have resulted from the low
mobility of Cu in plant tissues (Kabata-Pendias and
Pendias, 1999).
These relationships are consistent with the
results obtained by Aksoy et at. (2005), which
indicates a relatively low mobility of Cu in aquatic
plants and accumulation it in large quantities in the
rhizomes and roots. In the case of manganese
concentration, a statistically significant relationship
between Mn content in the sediments and its
concentration in underground and aboveground
sprouts of Glyceria maxima was revealed (Table 8).
No similar dependencies were found between Zn and
Cu content which can suggest a limited availability of
those elements for the plants from of bottom
sediments with pH~8, (Table 3).
According to Kabata-Pendias (2001) many
factors influence intake of heavy metals by plants. pH
of soil is one of them (Takač et al, 2009). In alkaline
environment (Table 3) most heavy metals is non-
available for plants. Increase of mobility of Zn is most
effective with pH<6.2 (Martinez and Motto, 2000),
and manganese in alkaline environment (pH~8),
(Kabata-Pendias and Pendias, 1999). Along with the
increase of concentration of tested metals in bottom
sediments of the Słupia River, the values of EF factors
decreased, which is manifested by negative values of
correlation coefficients (Table 8).
Table 8. The correlation coefficients (r) of heavy metals
contents in the relationship: aboveground
prouts/underground sprouts, bottom sediment/aboveground
sprouts, bottom sediment/ underground sprouts, bottom
sediment/EF (aboveground sprouts), bottom sediment/EF(underground
sprouts), (n=30, p <0.05, rcrit. = 0.30)
r in relation: Zn Mn Cu Ni
aboveground sprouts/ 0.21 0.64 -0.45 0.20
underground sprouts
bottom sediment/ -0.02 0.87 -0.20 0.58
aboveground sprouts
bottom sediment/ 0.19 0.75 0.06 -0.04
underground sprouts
bottom sediment/ -0.72 -0.34 -0.81 -0.31
EF(aboveground sprouts)
bottom sediment/ -0.50 -0.17 -0.62 -0.70
EF(underground sprouts)
1750
4. Conclusions
The bottom sediments of the Słupia River were
classified as slightly polluted. The indexes of the
bottom sediments quality and heavy metals
enrichment factors of the sprouts indicate that the
lowest pollution is found at the section of the Słupia
River with low and moderate transformation of the
river bed.
All of the determined heavy metals were
accumulated much easily in underground sprouts than
in the aboveground ones of Glyceria maxima. The
values of enrichment factors (3.3>EF>1.2) indicate
slight contamination with the tested metals. The
research results indicate that the sprouts of Glyceria
maxima are potentially useful in biomonitoring
environmental contamination with Mn.
References
Aksoy A., Demirezen D., Duman F., (2005),
Bioaccumulation, detection and analyses of heavy
metal pollution in Sultan marsh and its environment,
Water Air and Soil Pollution, 164, 241-255.
Alloway B. J., (1995), Soil Processes and the Behavior of
Metals, In: Heavy Metals in Soils, Alloway B.J. (Eds.)
2nd Edition, Blackie, Glasgow, 7-28.
Baldantoni D., Alfani A., Di Tommasi P., Bartoli G., Virzo
de Santo A., (2004), Assessment of macro and
microelement accumulation capability of two aquatic
plants, Environmental Pollution, 130, 149-156.
Baptista Neto J.A., Smith B.J., Mcallister J.J., (2000), Heavy
metal concentrations in surface sediments in a
nearshore environment, Jurujuba Sound, Southeast
Brazil, Environmental Pollution, 109, 1-9.
Bettinetti R., Giarei C., Provini A., (2003), A chemical
analysis and sediment toxicity bioassays to assess the
contamination of the River Lambro (Northern Italy),
Archives of Environmental Contamination and
Toxicology, 45, 2-80.
Bojakowska I., Sokołowska G., (1998), Geochemical
cleanliness class deposits water, Przegląd Geologiczny,
46, 49-54.
Chakravarty M., Patgiri A.D, (2009), Metal pollution
assessment in sediments of the Dikrong River, N.E.
India, Journal of Human Ecology, 27, 63-67.
Dauvalter V., Rognerud S., (2001), Heavy metal pollution in
sediments of the Pasvik River drainage, Chemosphere,
42, 9-18.
Helios-Rybicka E., (1991), Phase–specific bonding of heavy
metals in the sediments of the Vistula River, Poland,
Applied Geochemistry, 2, 45-48.
Hlihor R.M., Apostol L.C., Smaranda C., Pavel L.V.,
Caliman F.A., Robu B.M., Gavrilescu M., (2009),
Bioavailability processes for contaminants in soils and
their use in risk assessment, Environmental
Engineering and Management Journal, 8, 1199-1206.
Kabata-Pendias A., (2001), Trace Elements in Soils and
Plants, 3rd Edition, CRC Press, Boca Raton, Florida.
Kabata-Pendias A., Pendias H., (1999), Biogeochemistry of
Trace Elements, PWN, Warszawa.
Klavins M., Briede A., Rodinov V., Kokorite I., Parele E.,
Klavina I., (2000), Heavy metals in rivers of Latvia,
Science of the Total Environment, 262, 175-183.
 Heavy metals content in the sprouts of Glyceria maxima (Hartm.) Holmb. and in river sediments (Northern Poland)
Klink A., Wisłocka M., Musiał M., Krawczyk J., (2013),
Micro– and trace– elements accumulation in Typha
angustifolia L. and Typha latifolia L. organs and their
use in bioindycation, Polish Journal of Environmental
Studies, 22,183-190.
Kobus S., Glinska-Lewczuk K., Sidoruk M., Skwierawski
A., Obolewski K., Timofte C.M., Sowinski P., (2016),
Effect of hydrological connectivity on physico-
chemical properties of bottom sediments of floodplain
lakes - a case study of the Lyna River, northern Poland,
Environmental Engineering and Management Journal,
15, 1237-1246.
Kozanecka T., Chojnicki J., Kwasowski W., (2002), Content
of heavy metals in plant from pollution - free regions,
Polish Journal of Environmental Studies, 11, 395-399.
Kruopiene J., (2007), Distribution of heavy metals in
sediments of the Nemunas River (Lithuania), Polish
Journal of Environmental Studies, 16, 5, 715-722.
Kucharzewska K., Borkowski J., Kucharzewski A., (1991),
The Content of Heavy Metals in Alluvial River Oder
River (in Polish), In: National Conference: Geological
Aspects of Environmental Protection, AGH, Kraków
(Ed.), 172-176.
Kurilenko V.V., Osmolovskaya N.G., (2006), Ecological
Biogeochemical role of macrophytes aquatic
ecosystems of urbanized territories (an example of
small bodis of St. Petersburg), Russian Journal of
Ecology, 37, 147,
https://doi.org/10.1134/S1067413606030015.
LAWA, (1998), Assessment of the Water Quality of Running
Waters in the Federal Republic of Germany - Chemical
Water Quality Classification, Targets for the Protection
of Above-Ground Inland Waters, Band 2, Berlin.
Lis J., Pasieczna A., (1995), Geochemical Atlas of Poland,
PIG, Warszawa.
Liu J., Li Y., Zhang B., Cao J., Cao Z., Domagalski J.,
(2009), Ecological risk of heavy metals in sediments of
the Luan River source water, Ecotoxicology, 18, 748-
758.
Loska K., Wiechuła D., (2003), Application of principal
component analysis for the estimation of source heavy
metal contamination in surface sediments from Rybnik
Reservoir, Chemosphere, 51, 723-733.
Martinez C.E., Motto H.L., (2000), Solubility of lead, zinc
and copper added to mineral soils, Environmental
Pollution, 170, 153-158.
Miretzky P., Saralegui A., Fernandez C.A., (2004), Aquatic
macrophytes potential for the simultaneous removal
heavy metals (Buenos Aires, Argentina), Chemosphere,
57/8, 997-1005.
Mishra V.K., Tripathi B.D., (2008), Concurrent removal and
accumulation of heavy metals by the three aquatic
macrophytes, Bioresource Technology, 99, 7091-7097.
Obolewski K., Glińska-Lewczuk K., (2006), Contents of
heavy metals in bottom sediments of oxobow lakes and
the Słupia River, Polish Journal of Environmental
Studies, 15, 440-444.
Obolewski K., Glińska-Lewczuk K., (2013), Distribution of
heavy metals in bottom sediments of floodplain lakes
and their parent river – a case study of the Słupia,
Journal of Elementology, 18, 673-682.
Obolewski K., Strzelczak A., Glinska-Lewczuk K.,
Osadowski Z., Astel A., Timofte C.M., (2016),
Ecohydrological relationships between benthic
communities and environmental conditions in the
spring areas, Environmental Engineering and
Management Journal, 15, 1281-1291.
Ostrowska A., Gawliński S., Szczubiałka Z., (1991),
Methods of analysis and assessment of soil land plant
properties. Catalogue, Environmental Protection,
Warszawa, 101-103.
Piechura J., Walczowski W., Beszczyński-Moller A.,
(1997), On the structure and dynamics of the water in
the Slupsk Furrow, Oceanologia, 39, 35-54.
Rabajczyk A., Jóźwiak M.A., (2008), The possibilities of
using macrophytes as bioindicators of heavy metals
occurring in sediments, Monitoring Środowiska
Przyrodniczego, 9, 19-26.
Ramsey P.H., (1989), Critical values for Spearman’s rank
order correlation, Journal of Statistic Education, 14,
245-253.
Salati S., Moore F., (2009), Assessment of heavy metal
concentration in the Khoshk River water and sediment,
Shiraz, Southwest Iran, Environmental Monitoring and
Assessment, 164, 677-689.
Salt D.E., Kramer V., (2000), Mechanisms of metal
Hyperaccumulation in Plants, In: Phytoremediation of
Toxic Metals: Using Plants to Clean Up the
Environment, Raskin I, Ensley B.D. (Eds.), Wiley and
Sons, New York, 231-246.
Samecka-Cymerman A., Kempers A.J., (1996),
Bioaccumulation of heavy metals by aquatic
macrophytes around Wrocław, Poland, Ecotoxicology
and Environmental Safety, 35, 242-247.
Sarosiek J., Wożakowska-Natkaniec H., (1993), Chromium
and Nickel in Plants of the Family Lemnaceae and their
Environment, In: Chrome, Nickel and Aluminum –
Environmental Problems and Methodical, Kabata-
Pendias A. (Eds.) Zesz. Nauk. Kom. Człowiek i
Środowisko, 49-54.
Skorbiłowicz E., (2003), Heavy metal content in bottom
deposits and water plants in the Ełk River on urbanized
area, Ecological Chemistry and Engineering, 10, 9,
937-945.
Skorbiłowicz E., Skorbiłowicz M., (2011), Metals in grain
fractions of bottom sediments from selected rivers in
north-eastern Poland, Physics and Chemistry of the
earth, 36, 567-578.
Smal H., Salomons W., (1995), Acidification and its Long-
Term Impact on Metal Mobility, In: Biogeodynamics of
Pollutants in Soils and Sediments, Springer-Verlag,
Berlin, 193-212.
Sriyaraj K., Shutes R.B.E., (2001), An assessment of the
impact of motorway runoff on a pond wetland and
stream, Environment International, 26, 433-439.
Takač P., Szabová T., Kozáková L., Benková M., (2009),
Heavy metals and their bioavailability from soils in the
long-term polluted Central Spis region of SR, Plant,
Soil and Environment, 55, 167-172.
Tam N.F.Y., Wong, Y.S., (2000), Spatial variation of heavy
metals in surface sediments of Kong mangrove
swamps, Environmental Pollution, 110, 195-205.
Tang B., Shan B., Zhang H., Mao Z., (2010), Heavy metal
sources and associated risk in response to agricultural
intensification in the estuarine sediments of Chaohu
Lake Valley, East China, Journal of Hazardous
Materials, 176, 945-951.
Tanner C.C., (1996), Plants for constructed wetland
treatment systems – A comparison of the growth and
nutrient uptake of eight emergent species, Ecological
Engineering, 7, 59-83.
Teuchies J., Jacobs S., Oosterlee L., Bervoets L., Meire P.,
(2013), Role of plants in metal cycling in a tidal
wetland: Implications for phytoremediation, The
Science of the Total Environment, 445-446, 146-154.
1751
 Parzych and Cymer/Environmental Engineering and Management Journal 17 (2018), 7, 1743-1752
Vandecasteel B., Quataert P., De Vos, B., Tack F.M.G.
(2004), Assessment of the pollution status of alluvial
plains: a case study for the dredged sediment derived
soils along the Leie River, Archives of Environmental
Contamination and Toxicology, 47, 14-22.
Wiśniowska-Kielian B., Niemiec M., (2005), Heavy metals
contents in sediments of selected tributaries of river
Dunajec, Journal of Elementology, 10, 435-443.
Wolska L., Mędrzycka K., (2009), Assessment of
ecotoxicity of bottom sediments from the ports Marine
1752
in Gdansk and Gdynia, Ochrona Środowiska, 31, 49-
52.
Xiangdong Li., Shen Zhenguo, Onyx W.H. Wal, Yok-
Sheung Li., (2000), Chemical partitioning of heavy
metal contaminants in sediments of the Pearl River
Estuary, Chemical Speciation Bioavailability, 12, 17-
25.
Yuan C., Shi, J., He B., Liu, J., Liang, L., Jiang G., (2004),
Speciation of heavy metals in marine sediments from
the East China Sea by ICP-MS with sequential
extraction, Environment International, 30, 769-783
Environmental Engineering and Management Journal July 2018, Vol.17, No. 7, 1753-1764
http://www.eemj.icpm.tuiasi.ro/; http://www.eemj.eu

“Gheorghe Asachi” Technical University of Iasi, Romania

RESEARCH ON WASTEWATERS BIOREMEDIATION

WITH AQUATIC SPECIES FOR CONSTRUCTED WETLANDS

Dana Malschi1∗, Liviu Muntean1, Ioan Oprea 1, Carmen Roba1, Gabriela Popiţa1,
Lucrina Ştefănescu1, Bianca Malschi Florian2, Elena Rînba3
1Babeş-Bolyai University, Faculty of Environmental Sciences and Engineering,
30 Fântânele Street, 400294, Cluj-Napoca, Romania
2National Environmental Protection Agency, 99-101 Dorobanţilor, Cluj-Napoca, Romania
3Babeş-Bolyai University, Botanical Garden, 42 Republicii, Cluj-Napoca, Romania

Abstract

With the aim of contribute to a bioremediation technology drafting, this paper presents a preliminary research on bioremediation
of wastewaters polluted with heavy metals and other toxic contaminants. Bioremediation tests were performed with aquatic species
Lemna minor, Vallisneria spiralis, Eichhornia crassipes, Hydrilla verticillata, Pistia stratiotes, Cladophora glomerata, on samples
of contaminated waters resulting from the toxic pond of waste landfill Şomârd-Mediaş, Sibiu County, Romania, during the 2012-
2013 period. The studies were conducted in microcontainers with contaminated water using constructed wetlands. In order to check
the phytoextraction and bioremediation potential of these species, a comparative analysis of water samples was performed, based
on the determination of heavy metals and different pollution parameters. The results on phytoextraction tests have shown that: L.
minor plants were able to extract Cu, Fe and Pb; V. spiralis plants were able to extract Cu, Fe, Zn and Ni; H. verticillata, E.
crassipes, P. stratiotes and green algae C. glomerata were able to extract Cu, Fe and Zn. This research has shown that the aquatic
plants grown for two weeks in the toxic pond water have important effects on wastewater bioremediation. All species have shown
a significant decrease of NH4+ concentration. L. minor, V. spiralis, H. verticillata, E. crassipes, P. stratiotes have determined the
increase of Eh and the decrease of CCO-Mn, TDS, EC. Stronger effects of bioremediation in the tests with ½ diluted water from
the toxic pond were recorded. The results prove the significance of the bioremediation of contaminated waters using aquatic species
for constructed wetlands in order to implement this biotechnology.

Key words: bioremediation, Eichhornia crassipes, Lemna minor, Pistia stratiotes, Vallisneria spiralis

Received: February, 2014; Revised final: September, 2014; Accepted: October, 2014; Published in final edited form: July 2018

1. Introduction associations and microorganisms involved in

Studies on the bioremediation of the
contaminated sites entail a wide approach due to the
opportunities of drafting the adequate technologies for
biological depollution (Barbu and Sand, 2004; Buta et
al., 2011; Elekes et al., 2010; Gavrilescu, 2010; Glick
and Stearns, 2011; Malschi, 2009, Malschi et al.,
2013; Oros, 2002, 2011; Pivetz, 2001).
Bioremediation of contaminated aquatic
ecosystems may be conducted by using aquatic plant
biological depollution. Waters polluted with heavy
metals may be efficiently treated by artificial
constructed wetland ecosystems. The removal of
heavy metals by constructed wetland has been the
subject of significant researches (Dickinson et al.,
2009; Jing et al., 2001; Juang and Chen, 2007; Kuschk
et al., 2005; Rai, 2008; Wang et al., 2016).
Numerous experimental results highlighted the
effect of the common duckweed Lemna minor L.,
Araceae family in phytoremediation, by pollutants

∗
Author to whom all correspondence should be addressed: e-mail: danamalschi@yahoo.com, Phone: +40745 367373, Fax: +40264-30 70 30
 Malschi et al./Environmental Engineering and Management Journal 17 (2018), 7, 1753-1764
bioaccumulation, metals accumulation and
wastewater treatment (Dosnon-Olette et al., 2011;
Malschi et al., 2013; Rahmani and Sternberg, 1999;
Rahman et al., 2007; Willey, 2007). Bioremediation
research has revealed the effect of aquatic species such
as eel grass Vallisneria spiralis L. (Hydrocharitaceae)
(McCutcheon and Schnoor, 2004; Masi et al., 2002;
Núñez et al., 2011; Sim et al., 2007 etc.) on metals
hyperaccumulation (Malschi et al., 2013; Oprea et al.,
2013); water hyacinth Eichhornia crassipes (Mart.)
Solms (Pontederiaceae) on biosorption and metals
accumulation (Buta et al., 2011); waterthyme Hydrilla
verticillata (L.f.) Royle (Hydrocharitaceae), water
lettuce Pistia stratiotes L. (Araceae) (Stoica et al.,
2009) and green algae Cladophora glomerata (L.)
Kütz. (Chlorophyta) (Yalçına et al., 2008) on metals
accumulation and wastewater treatment.
Studies on biological pollution indicators in the
aquatic environment reveal an important interest due
to the significance of biomonitoring in pollution
assessment, determination of ecotoxicity and
development of bio- and eco-technologies (Glick and
Stearns, 2011; Malschi, 2009; Oros, 2002, 2011; Oros
and Drăgici, 2002; Pavel et al., 2012).
Constructed wetlands are used for the
decontamination of industrial wastewater and leachate
from wastes landfills (Jing et al., 2001; Kuschk et al.,
2005; Masi et al., 2002; Sim et al., 2007). The
constructed wetlands provide the biological processes
determining biological degradation, transformation,
biosorption and phytoextraction of pollutants by
aquatic species (Dyck et al., 2009; Juang and Chen,
2007). The use of constructed wetlands (composed of
water plants, green algae, cyanobacteria and various
species with high biological effect of pollutants
bioaccumulation) can remove various heavy metals
and other pollutants (Jing et al., 2001; Kuschk et al.,
2005; Malschi, 2009; Malschi et al., 2013; Oprea et
al., 2013; Xu et al., 2016).
This paper presents preliminary research to
prove the importance of the contaminated waters
bioremediation technology using aquatic species for
constructed wetlands. The novelty of the paper
consists in controlling the heavy metals
phytoextraction potential of the tested species and
measuring the bioremediation potential of waste
waters quality parameters. In order to check the
phytoextraction and bioremediation potential of these
species, the experimental tests were performed with L.
minor, V. spiralis, E. crassipes, H. verticillata, P.
stratiotes, C. glomerata.
2. Material and methods
During the 2012-2013 period, researches on
bioremediation of polluted waters were conducted in
the Biotechnology Laboratory of the Faculty of
Environmental Science and Engineering within the
Babeş-Bolyai University, Cluj-Napoca. The polluted
water samples were collected from the toxic pond of
municipal and industrial waste landfill Şomârd-
Mediaş, Sibiu County.
1754
Various toxic pollutants and heavy metals have
accumulated in this pond causing an important
environmental problem, which has been intensely
studied (Mihăiescu et al., 2010; Muntean, 2003; Oprea
et al., 2010).
Lemna minor, Vallisneria spiralis, Eichhornia
crassipes, Hydrilla verticillata, Pistia stratiotes and
Cladophora glomerata have been used to study the
bioremediation and phytoextraction of water
pollutants, in micro containers tests such as
constructed wetlands. The biologic material of the
aquatic species has been grown under greenhouse
conditions at the Botanical Garden of the Babeş-
Bolyai University, Cluj-Napoca. During the
experiment, the best conditions for plant growing in
laboratory were provided in transparent plastic pots
adequate as volume for the waters as follows: 700 mL
for L. minor variants, 1 liter for P. stratiotes variants,
2 liters for the variants of E. crassipes, V. spiralis, H.
verticillata, 400 mL for C. glomerata variants. The
experimentation time was of two weeks for each
aquatic species (24.10.2012-07.11.2012) with
additional light for 16 hours daily. Three types of
experimental waters for each species were placed in
micro containers: 1. drinking plain water as blank; 2.
water collected from the toxic pond of Şomârd waste
landfills; 3. water with the 1/2 dilution from the toxic
pond of Şomârd waste landfills. The development of
plants phytotoxic effects of water samples subjected to
testing and the phytoremediation effects were
highlighted by photographing the visible differences
between the samples compared to the situation in the
unpolluted blank sample.
Before and after phytoremediation and
phytoextraction tests, analysis of water quality
parameters of the samples were conducted. A
multiparameter type WTW Multi 350 was used for the
following parameters: the values of O2, Eh-redox
potential, CCO-Mn, EC-electrical conductivity, TDS-
total dissolved solids, S–salinity.
The concentrations of anions (F-, Cl-, NO2-,
NO3-, SO42-) and cations (Li+, Na+, K+, Mg2+, Ca2+,
NH4+) have been determined using ICS 1500 Dionex
Ion Chromatography (IC) analysis method. Water
samples were filtered with qualitative filter paper 125
mm and before analyzing were diluted to an electrical
conductivity lower than 100 µS/cm. An analytical
column IonPac AS23 for ion exchange was used for
the separation of the major anions. The suppressor was
an ASRS 300 ULTRA II (4 mm) Anions Self
Regenerating Suppressor. As anion mixture standard
Seven Anion Standard II (Dionex, USA) was used.
The injection volume was 25 μL, and the flow rate 1.0
mL/min, the eluent was an aqueous solution consisting
of 4.5 mM Na2CO3 and 0.8 mM NaHCO3.
For the separation of the major cations, an
Analytical column IonPac CS12A for ion exchange
was used and a CSRS 300 ULTRA II (4 mm) Cations
Self Regenerating Suppressor. Six Cation II Standard
(Fluka, Germany) has been used as cation standard
mixture. The eluent was an aqueous solution
consisting of 20 mM methanesulfonic acid.
 Research on wastewaters bioremediation with aquatic species for constructed wetlands

Heavy metal concentrations (Pb, Cu, Zn, Ni,
Fe) have been determined with atomic absorption
spectrometry (AAS) (ZEEnit 700 Analytik Jena)
method for determination of trace elements from
solution (flame method, mg/L). Before analyzing, the
water samples were acidified with HNO3 65% at a pH
between 2 and 3 and filtered with qualitative filter
paper of 125 mm.
3. Result and discussions
Previous studies performed in 2009-2010 on
the bioaccumulation potential of pollutants
recommend the cultivation of L. minor plants as a
method of bioremediation. Bioaccumulation of heavy
metals and metalloids was studied on L. minor plants
grown in polluted waters. L. minor was noticed to be
a very good bioaccumulator for Cr, Mn, Zn, As, Ba
and a moderate bioaccumulator of Cu, Pb, V from the
polluted waters in the waste heaps and tailings dams
areas of mining operations, toxic ponds and waste
platforms in the chemical industry (Malschi et al.,
2013). The study on the parameters of polluted water
revealed an extremely high level of heavy metals (Fe,
Cd, Zn) pollution in the water samples collected from
Şomârd-Mediaş municipal waste landfill, in Sibiu
county, also mentioned in previous studies (Mihăiescu
et al., 2010; Muntean, 2003; Oprea et al., 2010). The
results obtained after applying the Pb, Cd, Zn
phytoextraction and bioremediation tests (Oprea et al.,
2013) proved the role of the wastewater
bioremediation treatment using L. minor and V.
spiralis as component of constructed wetlands in lab.
micro containers. After two weeks of bioremediation
with V. spiralis it was noticed that these plants have a
great capacity to biotransform the pollutants and to
make them form a viscous oil film on the water
surface.
The new tests of bioremediation and
phytoextraction with L. minor (Fig.1), P. stratiotes
(Fig. 2), E. crassipes (Fig. 3), V. spiralis (Fig. 4), H.
verticillata (Fig. 5), C. glomerata (Fig. 6A) were
performed on waste water samples collected from
historically polluted sites, from the toxic pond of
municipal and industrial waste landfill Şomârd, Sibiu
county, during 2012-2013. In order to study the
phytoextraction and bioremediation potential of
aquatic species, a comparative analysis was performed
between the samples of polluted water and control
unpolluted water, in terms of pollution parameters
modifications depending on plants activity.
The experimental data obtained indicate that
significant effects on heavy metals phytoextraction
and biosorption were noticed in L. minor, V. spiralis,
E. crassipes, H. verticillata, P. stratiotes, C.
glomerata. Biotransformation of some toxic
contaminants and their removal from the water in a
film of oil on the water surface has been shown for V.
spiralis, H. verticilata, E. crassipes, C. glomerata
(Fig. 6B). Present experimental results have illustrated
that L. minor, V. spiralis, E. crassipes, H. verticillata,
P. stratiotes, C. glomerata have a very important role
in the bioremediation of wastewater pollution with
various toxic compounds and heavy metals, and are
recommended for use in micro ecosystems of
constructed wetlands (Fig. 7).
3.1. Phytoextraction/biosorption of heavy metal with
aquatic species with plants and green algae
The concentrations of heavy metals in water
samples collected from the Şomârd/Mediaş landfill
toxic pond in 2012 exceeded the maximum permitted
limits for surface waters (Ord. 161, 2006). Toxic pond
water has such concentrations of heavy metals that it
fitted the 5th water quality class for Cu and Pb, the 3rd
water quality class for Ni and the 2nd water quality
class for Zn, Fe.
By biosorption and phytoextraction on the
experimental tests (during the 24.10.2012-07.11.2012
period), the plants of L. minor, V. spiralis, H.
verticillata, E. crassipes, P. stratiotes and the green
algae C. glomerata were able to extract copper. Only
L. minor plants extracted lead and only V. spiralis
plants extracted nickel. The plants of L. minor, V.
spiralis, H. verticillata, E. crassipes, P. stratiotes and
green algae C. glomerata were able to extract iron. V.
spiralis, H. verticillata, E. crassipes, P. stratiotes
plants and C. glomerata were able to extract zinc (Fig.
8).

1 2 3 1 2 3
(a) (b)

1 2 3 1 2 3
(c) (d)
Fig. 1. Phytoextraction and biomonitoring with Lemna minor in water from Şomârd waste landfills: (a) Phytoextraction tests
during 23-29.10.2012; (b) Phytoextraction tests during 23.10-6.11.2012; (c) Biomonitoring tests/29.10.2012; (d) Biomonitoring
tests/6.11.2012. 1=Control, 2=water from toxic pond, 3=the 1/2 dilution of water from the toxic pond

1755
 Malschi et al./Environmental Engineering and Management Journal 17 (2018), 7, 1753-1764

1 2 3 1 2 3
(a) (b)

Fig. 2. Phytoextraction tests with Pistia stratiotes in water from Şomârd-Mediaş waste landfills: (a) Before
phytoextraction/23.10.2012; (b) After phytoextraction tests/6.11.2012. 1=Control, 2=water from the toxic pond, 3= the 1/2
dilution of water from the toxic pond

1 2 3
(a)

1 2 3 1 2 3
(b) (c)

Fig. 3. Phytoremediation tests with Echhornia crassipes in water from Şomârd-Mediaş waste landfills.
a) and b) Phytoextraction tests/29.10.2012; c) After phytoextraction/13.11.2012. 1=Control, 2=water from the toxic pond, 3= the
1/2 dilution of water from the toxic pond

1 2 3 1 2 3
(a) (b)

Fig. 4. Phytoremediation tests with Vallisneria spiralis in water from Şomârd-Mediaş waste landfills. a) Before
phytoextraction/24.10.2012; b) After phytoextraction/6.11.2012. 1=Control, 2=water from the toxic pond, 3=the 1/2 dilution of
water from the toxic pond

1756
 Research on wastewaters bioremediation with aquatic species for constructed wetlands

1 2 3 1 2 3
(a) (b)

Fig. 5. Phytoremediation tests with Hydrilla verticillata in water from Şomârd-Mediaş waste landfills.
a) Phytoextraction/29.10.2012; b) After phytoextraction/6.11.2012. 1=Control, 2=water from the toxic pond, 3= the 1/2 dilution of
water from the toxic pond

1 2 3 1 2 3
(a) (b)

Fig. 6. Bioremediation with Cladophora glomerata in water from Şomârd waste landfills
a) Bioremediation tests with Cladophora glomerata, after phytoextraction/15.11.2012. 1- control, 2 - water from the toxic pond,
3 - the 1/2 dilution of water from the toxic pond, b) Biotransformation of pollutants from waste water after phytoremediation.
Isolation of pollutants oily film after the treatment with: 1=Cladophora glomerata, 2=Hydrilla verticillata; 3=Valisneria spiralis

1. Bioremediation tests /14-25.11.2012

2. Bioremediation tests /14-25.11.2012 2. Photos of lab tests / 25.11.2012

Fig. 7. Bioremediation in laboratory micro-constructed wetlands consisting in aquatic species associations: Lemna minor,
Valisneria spiralis, Eichhornia crassipes, Hydrilla verticillata, Pistia stratiotes, Cladophora glomerata in toxic pond polluted
water from Şomârd-Mediaş waste landfills. 1- bioremediation tests /14-25.11.2012 / with species association in 1/2 dilution toxic
pond water. 2 - bioremediation tests /14-25.11.2012 / with species association in 1/4 dilution toxic pond water

Preliminary results on phytoextraction tests plants have extracted Cu, Zn, Fe, Ni, H. verticillata, E.
during 24/10/2012 - 07/11/2012 have shown that: L. crassipes, P. stratiotes and C. glomerata have
minor plants have extracted Cu, Fe and Pb, V. spiralis extracted Cu, Zn and Fe (Fig. 8, Table 1)

1757
 Malschi et al./Environmental Engineering and Management Journal 17 (2018), 7, 1753-1764

Fig. 8. Laboratory tests with Lemna minor, Vallisneria spiralis, Eichhornia crassipes, Hydrilla verticillata, Pistia stratiotes,
Cladophora glomerata used in heavy metal biosorption from the Şomârd landfill toxic pond water /24.10.2012-07.11.2012.
Heavy metal concentrations:
a) Before phytoextraction / 24.10.2012;
b) After phytoextraction with plants / 07.11.2012

Table 1. Chemical quality standards for some heavy metals increase of SO42-and decrease of Cl-, F-. V. spiralis
to set out the ecological status of surface waters / maximum plants in drinking water type determined the increase
limits / quality classes (Ord. 161, 2006) concentrations of F-, SO42-, decrease of Cl- and fixing
N to NO2- during testing. In the toxic pond water and
Quality
U/M Quality classes in 1/2 dilution toxic pond water, V. spiralis determined
indicator
Heavy
the increase of SO42- and decrease of Cl-, F-. H.
metals
1 2 3 4 5 verticillata determined the increase of Cl-, F-
> concentration, decrease of SO42-. H. verticillata
Cu µg/L 20 30 50 100 determined the increase of SO42- and decrease of Cl-,
100
> F- in the toxic pond water and the ½ dilution toxic pond
Zn µg/L 100 200 500 1000
1000 water. The E. crassipes plants determined the increase
Pb µg/L 5 10 25 50 > 50 of Cl-, F-, SO42- concentrations in drinking water,
Cd µg/L 0,5 1 2 5 >5 while E. crassipes plants determined the increase of
> SO42-, Cl- and a decrease of F- in the toxic pond water
Ni µg/L 10 25 50 100
100 and the 1/2 dilution toxic pond water.
Fe2+ + > The activity of P. stratiotes plants in drinking
µg/L 300 500 1000 2000
Fe3+ 2000
water determined the increase of Cl-, F-, SO42-
concentrations and fixing of N to NO2-. P. stratiotes
3.2. Bioremediation of wastewater from the toxic pond plants have determined the increase of SO42- and
of Şomârd/Mediaş landfills with aquatic species L. decrease of F-, Cl- in the toxic pond water, while in the
minor, V. spiralis, E. crassipes, H. verticillata, P. 1/2 dilution toxic pond water, the plants have
stratiotes, C. glomerata determined the increase of SO42-, Cl- and a decrease of
F-. C. glomerata produced the increasing of Cl-, F-,
In 2012, the water from Şomârd/Mediaş SO42- concentrations and fixing N to NO2-. C.
landfills toxic pond had very high concentrations of glomerata algae have determined the increase of Cl-,
anions Cl-, F-, SO42- fitting the 5th quality class (poor) SO42- and decrease of F- in toxic pond water and in ½
for surface waters (Ord. 161, 2006). After the dilution toxic pond water (Tables 2, 3).
phytoremediation experimental tests, in case of the Very high concentrations of SO42-, Cl- and F-
control samples with drinking water, the N-NO2- were in toxic pond waters before the experiment. After
appears in Lemna, Vallisneria, Pistia, Cladophora. phytoremediation experiments, the concentration of
The presence and activity of these species have led to SO42- increased. The concentration of Cl-, F- and salt
the accumulation of nitrogen in the form of NO2-, in were variable in different waters types of tests (Tables
unpolluted waters. The environmental effect of 2, 3). Preliminary research on phytoremediation of
nitrogen fixation does not occur in the toxic pond environmental status of waters have shown that: in the
waters (Tables 2, 3). The presence of L. minor plants toxic pond water and the 1/2 dilution toxic pond water,
in drinking water determined the increase of L. minor, V. spiralis and H. verticillata, plants have
concentrations of Cl-, F-, SO42- and fixing N to NO2-. determined the increase of SO42- and decrease of Cl-,
In the toxic pond water and in the ½ dilution toxic F- anions, while E.crassipes and C. glomerata have
pond water, L. minor plants have determined the determined the increase of Cl-, SO42- and the decrease

1758
 Research on wastewaters bioremediation with aquatic species for constructed wetlands

of F-. P. stratiotes plantsproduced the increase of SO42-
and decrease of Cl-, F- anions in the toxic pond water.
3.3. Cations bioremediation with aquatic species
P. stratiotes has decreased the concentration of
K+, Ca2+, Mg2+, Na+, Li+ and more pronounced Mg2+.
L. minor caused the decrease of Ca2+, Mg2+
concentration and the increase of K concentration; V.
+

spiralis increased amount of K+, Ca2+, Mg2+; while H.

In 2012, the water from the toxic pond had a
small amount of Mg2+, very high concentrations of
Na+, K+, NH4+ fitting the 5th quality class (poor) and
moderate amount of Ca2+ (Tables 4, 5), fitting the 3rd
quality class of surface waters and concentrations
(Ord. 161, 2006). Aquatic plants grown for two weeks
in unpolluted water (plain water) caused different
changes as compared to the cation concentrations in
the control water samples (before bioremediation
tests).
L. minor, E. crassipes, P. stratiotes have
caused the decrease of K+ and Ca2+ concentration. V.
spiralis, H. verticillata and C. glomerata caused the
appearance of K+ and the increase of the Ca2+ amount.
The presence of aquatic plants reared for two weeks in
toxic pond water has caused important changes
compared to the control water samples.
verticillata and E. crassipes decreased Ca2+ and
increased K+, Mg2+. The E. crassipes increased Na+
and Li+. C. glomerata was revealed from the increase
amount of K+, Ca2+, Mg2+, Na+, Li+ (Tables 4, 5).
Preliminary results on bioremediation have
shown that the presence of aquatic plants reared in
water from the toxic pond has caused important effects
on cations concentration. All species have resulted in
significantly decreased amount of NH4+. L. minor, H.
verticillata, Eichhornia crassipes have decreased Ca2.
With important effect of increasing cations
concentrations have been noted: L. minor, V. spiralis,
H.verticillata, E. crassipes, C. glomerata which have
increased amount of K+. The presence of plants grown
for two weeks in water from the toxic pond with ½
dilution water determined changes as compared to the
control samples.

Table 2. Laboratory tests for phytoremediation with aquatic species Lemna minor, Vallisneria spiralis, Eichhornia crassipes,
Hydrilla verticillata, Pistia stratiotes, Cladophora glomerata of the anions environmental status of water
from the Şomârd waste landfills toxic pond

Anions concentrations F- Cl- NO2- NO3- SO42-

mg/L mg/L mg N/L mg N/L mg /L
A. Before bioremediation tests / 24.10.2012 / control water samples
Control-Drinking plain water 1.53 16.65 - 6.02 41.98
Water from the toxic pond 137.05 1662.09 - - 2785.43
1/2 dilution water from the toxic pond 60.63 810.73 - - 1307.87
B. After bioremediation tests / 7.11.2012 with:
Lemna minor
Control-Drinking water 2.02 27.30 19.64 - 52.48
Water from the toxic pond 41.21 1443.08 - - 3053.09
1/2 dilution water from the toxic pond 21.49 746.85 - - 1503.59
Vallisneria spiralis
Control-Drinking water 1.69 12.91 8.83 - 53.01
Water from the toxic pond 39.98 1525.09 - - 3031.76
1/2 dilution water from the toxic pond 22.08 736.70 - - 1417.86
Hydrilla verticillata
Control-Drinking water 1.84 26.10 - - 25.94
Water from the toxic pond 38.84 1514.99 - - 3230.43
1/2 dilution water from the toxic pond 21.77 785.50 - - 1563.50
Eichhornia crassipes
Control-Drinking water 4.29 84.58 - - 321.81
Water from the toxic pond 45.26 1803.99 - - 3744.97
1/2 dilution water from the toxic pond 26.36 847.15 - - 1783.26
Pistia stratiotes
Control-Drinking water 2.07 25.86 44.23 - 66.38
Water from the toxic pond 41.43 1436.59 - - 3042.64
1/2 dilution water from the toxic pond 23.02 850.47 - - 1822.10
Cladophora glomerata
Control-Drinking water 2.09 35.11 30.57 - 96.36
Water from the toxic pond 48.92 2048.04 - - 4543.16
1/2 dilution water from the toxic pond 25.71 1169.99 - - 2525.91

1759
 Malschi et al./Environmental Engineering and Management Journal 17 (2018), 7, 1753-1764

Table 3. Chemical and physicochemical quality standards for some nutrients to set out the ecological status of surface
waters/maximum limits/quality classes (Ord. 161, 2006)

The quality indicator Quality classes

U/M
Nutrients 1 2 3 4 5
Nitrites (N-NO2) mg N/L 0.01 0.03 0.06 0.3 >0.3
Nitrates (N-NO3 mg N/L 1 3 5.6 11.2 >11.2
Chlorides (Cl-) mg/L 25 50 250 300 >300
Sulfates (SO42+) mg/L 60 120 250 300 >300

Table 4. Laboratory tests for phytoremediation with aquatic species Lemna minor, Vallisneria spiralis, Eichhornia crassipes,
Hydrilla verticillata, Pistia stratiotes, Cladophora glomerata of the cation environmental status of water from the Şomârd waste
landfills toxic pond

Cations concentrations Li+ Na+ K+ Mg2+ Ca2+ NH4+

mg/L mg/L mg/L mg/L mg/L mg/L
Before bioremediation tests / 24.10.2012 / control water samples
Control-Drinking water - 119.17 - 3.36 25.66 -
Water from the toxic pond 15.35 2372.78 528.99 5.28 242.29 32.45
1/2 dilution water from the toxic pond 8.32 1387.03 293.00 4.63 136.17 26.67
B. After bioremediation tests /7.11.2012 with:
Lemna minor
Control-Drinking water - 111.03 45.46 3.05 24.09 -
Water from the toxic pond 14.32 2293.52 546.20 4.30 209.37 25.20
1/2 dilution water from the toxic pond 6.02 866.65 268.57 0.06 92.21 17.25
Vallisneria spiralis
Control-Drinking water - 104.49 10.32 2.05 38.09 -
Water from the toxic pond 15.41 2447.90 648.09 6.79 276.42 27.91
1/2 dilution water from the toxic pond 6.71 1147.35 328.29 1.73 119.96 20.27
Hydrilla verticillata
Control-Drinking water - 120.81 23.66 2.96 32.25 -
Water from the toxic pond 14.06 2353.30 588.14 7.99 209.96 26.48
1/2 dilution water from the toxic pond 7.39 1207.47 305.84 2.39 137.37 -
Eichhornia crassipes
Control-Drinking water - 123.97 14.39 2.74 18.03 -
Water from the toxic pond 17.49 2845.40 638.56 11.65 194.34 13.55
1/2 dilution water from the toxic pond 8.19 1292.19 321.34 2.97 147.01 -
Pistia stratiotes
Control-Drinking water - 120.54 28.44 2.72 19.55 -
Water from the toxic pond 13.09 2035.26 512.69 0.07 232.71 0.75
1/2 dilution water from the toxic pond 8.01 1333.96 327.10 5.35 117.34 -
Cladophora glomerata
Control-Drinking water 0.60 35.19 41.18 4.25 60.17 -
Water from the toxic pond 17.69 2854.91 660.62 5.59 250.37 10.47
1/2 dilution water from the toxic pond 10.48 1903.78 417.70 13.60 187.54 -

Table 5. Chemical and physicochemical quality standards for some cations to set out the ecological status of surface
waters/maximum limits/quality classes (Ord. 161, 2006)

The quality indicator U/M Quality classes

1 2 3 4 5
Na+ mg/L 25 50 100 200 >200
Mg2+ mg/L 12 50 100 200 >200
Ca2+ mg/L 50 100 200 300 >300
NH4+ mg/L 0.4 0.8 1.2 3.2 >3.2

1760
 Research on wastewaters bioremediation with aquatic species for constructed wetlands

L. minor decreased amount of K+, Ca2+, Mg2+, and P. stratiotes caused the increase of K+. C.
Na+, Li+; V. spiralis and H. verticillata caused the glomerata caused the increase the concentration of K+,
decrease of Ca2+, Mg2+, Na+, Li+ and the increase of Ca2+, Mg2+, Na+, Li+.
K+; while E. crassipes decreased amount of Mg2+. P.
stratiotes increased the concentration of Mg2+, K+ and 3.4. Bioremediation of wastewater by means of
decreased Ca2+ (Tables 4, 5). Stronger effects of aquatic species: L. minor, V. spiralis, E. crassipes, H.
bioremediation in the tests with aquatic plants grown verticillata, P. stratiotes, C. glomerata
for two weeks in ½ dilution water from the toxic pond
as compared to the control samples were recorded. H. The toxic pond water from the Şomârd/Mediaş
verticillata, E. crassipes, P. stratiotes and C. waste landfills, as compared to the control unpolluted
glomerata have decreased to zero the amount of NH4+. water samples, shows lower values of O2 mg/L and of
L. minor and V. spiralis significantly decreased Eh - oxidation-reduction potential and higher values
amount of NH4+. L. minor decreased the amount of of CCO-Mn, of EC - electrical conductivity, of TDS -
Ca2+, Na+, K+; V. spiralis and H. verticillata caused the total dissolved solids and of S – salinity, in 2012 (Fig.
decrease of Ca2+ and Na+. V. spiralis, H. verticillata 9).

A.1 C ontrol-Drinking wa ter -80,8 0 0,86 Ox ig en (mg /L )

B .1 L emna minor -80,5 4,29
-65 2,85
B .1 Va llisneria spira lis 2,09 6 C C O-Mn
-76,4
B .1 Hydrilla vertic illa ta 2,4 3,71 E h (mV)
B .1 E ic hhornia c ra ssipes -73,3 5,1 2,29
-54,2
B .1 P istia stra tiotes 5,77 6
B .1 C la dophora g lomera ta -52,6
4,28 3,14
E xperim ental variants

A.2 Wa ter from bund -50,7 0,34 62,86

B .2 L emna minor -86,4 36,57
0,24
B .2 Va llisneria spira lis -88 0,25 36,57
B .2 Hydrilla vertic illa ta -76,7 0,3 29,71
B .2 E ic hhornia c ra ssipes-100,4 0,2 40
B .2 P istia stra tiotes -82 27,43
0,16
B .2 C la dophora g lomera ta -82 37,71
0,5

A.3 Dilution 1/2 bund wa ter -49,3 32

0,26
B .3 L emna minor -80,2 32
0,37
B .3 Va llisneria spira lis -75 18,3
0,24
B .3 Hydrilla vertic illa ta -78,1 19,43
0,31
B .3 E ic hhornia c ra ssipes -92,7 0,12 24,29

-105 -85 -65 -45 -25 -5 15 35 55

E h va lues in mV a nd C C O-Mn a nd Ox ig en va lues in mg /L

(a)

A.1 C ontrol-Drinking wa te r 8,136

0
B .1 L e m na m inor 0,3 8,15
B .1 Va llisne ria spira lis 0,2 7,87
B .1 Hydrilla ve rtic illa ta 0,3 8,07
B .1 E ic hhornia c ra ssipe s 0,3 8,19
B .1 P istia stra tiote s 0,3 7,39
7,21
E xperim ental variants

B .1 C la dophora g lom e ra ta 0,2

7,602
A.2 Wa te r from bund 6,9
B .2 L e m na m inor 6,4 8,25
B .2 Va llisne ria spira lis 6,7 8,28
B .2 Hydrilla ve rtic illa ta 6,7 8,06
B .2 E ic hhornia c ra ssipe s 7,5 8,51
B .2 P istia stra tiote s 7 8,18
B .2 C la dophora g lom e ra ta 8,4 8,18
A.3 Dilution 1/2 bund wa te r 4 7,572
B .3 L e m na m inor 3,4 8,13
B .3 Va llisne ria spira lis 3,6 8,07
B .3 Hydrilla ve rtic illa ta 3,5 8,09
B .3 E ic hhornia c ra ssipe s 4 8,379
0 1 2 3 4 5 6 7 8 9
S ‰ pH pH units / S (‰) va lue s

(b)

1761
 Malschi et al./Environmental Engineering and Management Journal 17 (2018), 7, 1753-1764

284
A.1 C ontrol-Drinking water 338 568
B .1 L emna minor 299 675
B .1 Vallisneria spiralis 340 598 E C µS /cm
B .1 Hydrilla verticillata 316 680
B .1 E ichhornia crassipes 351 632 TDS mg /L
B .1 P istia stratiotes
B .1 C ladophora g lomerata 315 701
E xperim ental variants

630
6085 12170
A.2 Water from bund
B .2 L emna minor 5680 11360
B .2 Vallisneria spiralis 5950 11900
B .2 Hydrilla verticillata 5905 11810
B .2 E ichhornia crassipes 6645 13290
B .2 P istia stratiotes 6175 12350
B .2 C ladophora g lomerata 7360 14720
3615
A.3 Dilution 1/2 bund water 3140 7230
B .3 L emna minor 6280
B .3 Vallisneria spiralis 3315 6630
B .3 Hydrilla verticillata 3245 6490
B .3 E ichhornia crassipes 3385 67706770
B .3 P istia stratiotes 3385
B .3 C ladophora g lomerata 3340 6680
0 2000 4000 6000 8000 10000 12000 14000 16000

TDS values in mg /L and E C values in µS /cm

(c)

Fig. 9. Laboratory tests for phytoremediation of the environmental status of waters from the Şomârd Mediaş waste landfills toxic
pond, with aquatic species Lemna minor, Vallisneria spiralis, Eichhornia crassipes, Hydrilla verticillata, Pistia stratiotes,
Cladophora glomerata.
A. Before phytoremediation / 24.10.2012; B. After phytoremediation / 7.11.2012.
1. Control-Drinking water; 2.Water from the pond; 3. 1/2 dilution toxic pond water.
Quality parameters: 9 a - dissolved O2 mg/l; CCO-Mn; Eh–redox potential; 9 b - pH; S-salinity; 9 c - TDS-total dissolved solids;
EC-Electrical conductivity

Plants evolution in control samples with
unpolluted water shows that after phytoextraction, L.
minor, V. spiralis, H. verticillata, E. crassipes, P.
stratiotes and C. glomerata have determined the
increase of the dissolved O2, CCO-Mn, EC and S and
the decrease of Eh - redox potential. H. verticillata, E.
crassipes, P. stratiotes and C. glomerata have also
determined the increase of TDS.
Preliminary tests on bioremediation have
shown that in the samples with water from the toxic
pond and with ½ dilution water from the toxic pond:
L. minor plants, V. spiralis, H. verticillata, E.
crassipes, P. stratiotes have determined the increase
of Eh, the decrease of CCO-Mn, TDS and EC. In the
samples with water from the toxic pond: P. stratiotes
have determined the decrease of salinity, as well as a
slight decrease of O2; E. crassipes and C. glomerata
have determined the increase of salinity and
conductivity, as well as a slight decrease of TDS and
O2.
The present studies results and laboratory
experiments on bioremediation as a method of
decontaminating polluted areas from the toxic pond of
municipal landfill waste Şomârd platform-Mediaş,
Sibiu county, recommended the aquatic species L.
minor, V. spiralis, E. crassipes, H. verticillata, P.
stratiotes, C. glomerata for use as constructed
wetlands method (Fig. 9).
4. Conclusions
Research results proved the importance of the
bioremediation applied for constructed wetlands
polluted with heavy metal ions, by using aquatic plants
Lemna minor, Vallisneria spiralis, Eichhornia
crassipes, Hydrilla verticillata, Pistia stratiotes and
green algae Cladophora glomerata, producing the
improvement of environmental status of wastewater.
Strong effects on heavy metals
phytoextraction, the increase of Eh, the decrease of
CCO-Mn, TDS, EC and significant decrease of NH4+
were recorded. These plants and algae can be used in
biotechnological systems of constructed wetlands for
ecological restoration.
References
Barbu C. H., Sand C., (2004), Theory and Practice of
Modern Remediation of Soils Polluted with Heavy
Metals (in Romanian), Alma Mater Press, Sibiu,
Romania.
Buta E., Paulette L., Mihăiescu T., Buta M., Cantor M.,
(2011), The influence of heavy metals on growth and

1762
 Research on wastewaters bioremediation with aquatic species for constructed wetlands
development of Eichhornia crassipes species,
cultivated in contaminated water, Notulae Botanicae
Horti Agrobotanici Cluj-Napoca, 39, 135-141.
Dickinson N.M., Baker A.J.M., Doronila A., Laidlaw S.,
Reeves R.D., (2009), Phytoremediation of inorganics:
realism and synergies, International Journal of
Phytoremediation, 11, 97-114.
Dosnon-Olette R., Couderchet M., Oturan M.A., Oturan N.,
Eullaffroy P., (2011), Potential use of lemna minor for
the phytoremediation of isoproturon and glyphosate,
International Journal of Phytoremediation, 13, 601-
612.
Dyck V.M., Rob V.D., Geel K., (2009), On-site wastewater
treatment in Flanders, opportunities and threats for
constructed wetlands, project TISA, Belgium, On line
at:
http://www.constructedwetlands.net/vandyck_IWA_ne
pal_jaar.pdf.
Elekes C.C., Busuioc G., Ionita Gh., (2010), The
bioaccumulation of some heavy metals in the fruiting
body of wild growing mushrooms, Notulae Botanicae
Horti Agrobotanici Cluj-Napoca, 38, 147-151.
Gavrilescu M., (2010), Environmental biotechnology:
achievements, opportunities and challenges, Dynamic
Biochemistry, Process Biotechnology and Molecular
Biology, Global Science Books, 4, 1-36.
Glick B.R., Stearns J.C., (2011), Making phytoremediation
work better: maximizing a plant’s growth potential in
the midst of adversity, International Journal of
Phytoremediation, 13, 4-16.
Jing S.R., Lin Y.F., Lee D.Y., Wang T.W., (2001), Using
constructed wetland systems to remove solids from
highly polluted river water, Water Science and
Technology: Water Supply, 1, 89–96.
Juang D.F., Chen P.C., (2007), Treatment of polluted river
water by a new constructed wetland, International
Journal of Environmental Science and Technology, 4,
481-488.
Kuschk P., Wießner A., Müller R., Kästner M., (2005),
Constructed Wetlands – Treating Wastewater with
Cenoses of Plants and Microorganisms, A Research
Association at UFZ Centre for Environmental Research
Leipzig-Halle, In the Helmholtz Association, Published
by UFZ Centre for Environmental Research Leipzig-
Halle. Medien GmbH, On line at:
https://www.ufz.de/export/data/2/93454_CW_Booklet.
pdf.
Malschi D., (2009), Biology, Ecophysiology and
Microbiology Elements, Course notes and practical
applications (in Romanian), Electronic Manual, Bioflux
Publishing House, Cluj-Napoca, Romania, On line at:
http://www.editura.bioflux.com.ro/docs/malschi2.pdf.
Malschi D., Roman C., Miclean M., Şenilă M., Ştefănescu
L., Malschi-Florian B., Bolonyi A., Ghira G. B.,
Brăhaiţa I. D., Crihan Al. B., (2013), Study on
phytoextraction bioremediation of heavy metals and
metalloids from industrially polluted lands and waters
using lab tests based on Lolium perenne and Lemna
minor, Environmental Engineering and Management
Journal, 12, 1109-1112.
Masi F., Bendoricchio G., Conte G., Garuti G., Innocenti A.,
Franco D., Pietrelli L., Pineschi G., Pucci B.,
Romagnolli F., (2002), Constructed Wetlands for
Wastewater Treatment in Italy: State-of-art and
Obtained Results, Proc. of the IWA 7th Int. Conf. on
Wetland System for Water Pollution Control,
International Water Association, Arusha, Tanzania,
274-283.
McCutcheon S. C., Schnoor J. L., (2004), Index of Names of
Plants, In: Phytoremediation: Transformation and
Control of Contaminants, McCutcheon S.C., Schnoor
J.L. (Eds.), John Wiley & Sons, Inc., Hoboken, NJ,
USA, 925-939.
Mihăiescu R., Muntean O. L., Bodea C., Modoi C., Malos
C., Mihăiescu T., Arghius V., Rosian Gh., Baciu N.,
(2010), Alternative closing municipal landfills. Case
study: municipal landfill Medias, Sibiu County (in
Romanian), ProEnvironment, 3, 529-533.
Muntean O.L., (2003), A study of environmental
reconversion in Copşa Mică Area (Romania),
Wurzburger Geographische Manuskripte, 63, 63-71.
Núñez S.E.R., Marrugo Negrete J.L., Arias Rios J.E.,
(2011), Hg, Cu, Pb, Cd, and Zn accumulation in
macrophytes growing in tropical wetlands, Water, Air
and Soil Pollution, 216, 361-373.
Oprea I.C., Malschi D., Muntean O.L., (2013), The
biomonitoring and the bioremediation of toxic water
resulting from municipal waste storage of Somard,
Sibiu county, Advances in Environmental Sciences, 5,
189-196.
Oprea I.C., Modoi C., Muntean O.L., (2010), The Integrated
Management of Solid Urban Wastes from the Medias
Municipality, Proc. of the 2010 SIMTEX Conference,
Bucharest, 244-253.
Ord. 161, (2006), Order 161/2006 approving the Norms on
surface water quality classification in order to
determine the ecological status of water bodies,
Published in Romanian Official Monitor no. 511 of 13
June 2006.
Oros V., (2002), Biomonitoring, In: Environmental
Pollution and Monitoring, Transilvania University of
Braşov Publishing House, Brasov, Romania, 60-73.
Oros V., (2011), Elements of Ecotoxicology and
Ecotoxicological Tests, Risoprint Publishing House,
Cluj-Napoca, Romania.
Oros V., Drăgici C., (2002), Waste Management (in
Romanian), Transilvania University of Brasov,
Romania.
Pavel V.L., Diaconu M., Bulgariu D., Statescu F.,
Gavrilescu M., (2012), Evaluation of heavy metals
toxicity on two microbial strains isolated from soil:
Azotobacter sp. and Pichia sp., Environmental
Engineering and Management Journal, 11, 165-168.
Pivetz B.E., (2001), Phytoremediation of Contaminated Soil
and Ground Water at Hazardous Waste Sites, Ground
Water Issue, United States Environmental Protection
Agency, Washington DC, On line at:
https://www.epa.gov/sites/production/files/2015-
06/documents/epa_540_s01_500.pdf.
Rahman M.A., Hasegawa H., Ueda K., Maki T., Okumura
C., Rahman M.M., (2007), Arsenic accumulation in
duckweed (Spirodela polyrhiza L.): A good option for
phytoremediation, Chemosphere. 69, 493-499.
Rahmani G.N.H., Sternberg S.P.K., (1999), Bioremoval of
lead from water using Lemna minor, Bioresource
Technology, 70, 225-230.
Rai P.K., (2008), Heavy metal pollution in aquatic
ecosystems and its phytoremediation using wetland
plants: An ecosustainable approach, International
Journal of Phytoremediation, 10, 133-160.
Sim C.H., Yusoff K.M., Shutes B., Sim H.C., Mansor M.,
(2007), Nutrient removal in a pilot and full scale
constructed wetland. Putrajaya city, Malaysia, Journal
of Environmental Management, 88, 307-317.
1763
 Malschi et al./Environmental Engineering and Management Journal 17 (2018), 7, 1753-1764
Stoica M., Jinescu G., Godeanu M., Nisipeanu S., (2009).
Biosorption of lead on Pistia stratiotes, Proc. XIII
Balkan mineral Processing Congress, Bucharest, 842-
849.
Wang W.D., Wang X.G., Zhou L.C., Liu H.B., Ding Z.Z.,
Liang Y.T., (2016), Wastewater treatment in a
constructed wetland followed by an oxidation pond in a
rural area of China, Environmental Engineering and
Management Journal, 15, 199-205.
Willey N., (2007), Phytoremediation: Methods and
Reviews, Willey N. (Ed.), Humana Press Inc., New
Jersey.
1764
Xu D.F., Li Y.X., Howard A., (2016), Effect of the
interaction of earthworms and substrates on
photosynthetic characteristics, nitrogen and phosphorus
uptake by Iris pseudacorus in constructed wetland
system, Environmental Engineering and Management
Journal, 15, 123-132.
Yalçına E., Çavusoglu K., Maras M.¸ Bıyıkoglud M.,
(2008), Biosorption of Lead (II) and Copper (II) metal
ions on Cladophora glomerata (L.) Kütz.
(Chlorophyta) algae: effect of algal surface
modification, Acta Chimica Slovenica, 55, 228–232.
 Environmental Engineering and Management Journal July 2018, Vol.17, No. 7, 1765-1771
http://www.eemj.icpm.tuiasi.ro/; http://www.eemj.eu

“Gheorghe Asachi” Technical University of Iasi, Romania

ENVIRONMENTALLY FRIENDLY CELLULOSIC FIBERS

FROM CORN STALKS

Ana Maria Cheșcă, Bogdan Marian Tofănică, Adrian Cătălin Puiţel,

Raluca Nicu, Dan Gavrilescu ∗

“Gheorghe Asachi” Technical University of Iasi, Faculty of Chemical Engineering and Environmental Protection,
Department of Natural and Synthetic Polymers, 73 Prof.dr.doc. Dimitrie Mangeron Street, 700050 Iasi, Romania

Abstract

One of the main causes of global climate change is deforestation. Wood is largely used in construction, furniture, and pulp and
paper industries, which consume hundred of millions of cubic meter of wood year by year. Forest surfaces decrease rapidly at
global level and this reality is of crucial importance for world global ecological footprint. Shortage of pulpwood for paper industry
became a great challenge and, for this reason, other raw materials must be taken into consideration. Recovered paper is the most
important raw material for paper industry, but low paper grades can be produced only. Collection rate of recovered paper is high
in most countries and there are no additional quantities to assure the large requirements of the paper industry. Today, nonwood
plants are considered more and more important raw material for this industry. Wheat and rice straws are used for decades to produce
chemical pulp in many Asian and European countries.
This paper investigates the possibility of replacing pulpwood with corn stalks as raw material in producing chemical pulp for paper.
Corn stalks are abundant agricultural waste in Romania, since more than 12 million tones of stalks were generated in 2017. It was
found that corn stalks contain less lignin than wood and can be delignified using less chemicals, at lower temperatures and in
shorter time. Moreover, corn stalks can be processed using ordinary reagents like sodium hydroxide, and pulps with different yields
and lignin content can result. Natron pulping of corn stalks offers pulps with reasonable strength properties which can replace wood
pulp in obtaining papers for corrugated board.

Key words: corn stalks, pulping, paper, resources

Received: December, 2017; Revised final: May, 2018; Accepted: June, 2018; Published in final edited form: July 2018

1. Introduction deforestation are associated with overpopulation and

Wood is a valuable raw material and large
quantities are used in many industries like
construction, furniture and pulp and paper. Wood is
also an important fuel being more and more used today
for household heating. Deforestation is currently a
challenging problem, because about 26 ha of forest
disappear every minute at global level (Piazza and
Santanu, 2015). Extension of farming land, excessive
wood harversting, forest fires and urbanization are
among the causes of deforestation. Other reasons of
poverty, as well as corruption in many countries which
leads to immediate profit (Kissinger et al., 2012).
Increasing of carbon footprint and climate
change are among the most important consequences of
deforestation. The woodlands are the major terrestrial
sink of carbon, so that deforestation disrupts carbon
cycle at global level (Fortuna et al., 2012; Gavrilescu
et al., 2017; Ladd et al., 2013; Ward et al., 2015). The
key result is the rise of concentration of carbon
dioxide. By cutting the forests, the soil cannot hold
water and a drier climate become evident (Binkley and

∗
Authors to whom all correspondence should be addressed: e-mail: gda@tuiasi.ro
 Cheșcă et al./ Environmental Engineering and Management Journal 17 (2018), 7, 1765-1771
Fisher, 2012; Gimona et al., 2012). In the last decades,
the woodland of EU countries has continually
decreased and now its surface represents no more than
34% from the EU countries surface. (FAO, 2014).
Pulp and paper sector needs large volumes of
wood. To obtain a tone of pulp, 5-6 m3 of pulpwood
are necessary, so that a pulp mill consumes up to one
million cubic meters of timber per year (Gavrilescu
and Craciun, 2012; Hetemaki et al., 2013). According
to Confederation of European Paper Industries
(CEPI), around 150 million m3 of softwood and
hardwood were consumed in 2016 by the pulp and
paper sector of European Union countries (CEPI,
2017). This high level of wood consumption can not
be overcame without serious environmental
consequences. Wood has become deficient and
expensive and, for this reason, paper producers have
sought solutions to replace it. Recovered paper can
replace wood pulp in production of many paper grades
and this raw material is largely used by paper
manufacturers (Bajpai, 2015; Ghinea et al., 2014;
Puitel et al., 2014; Scott, 2011). However, recovered
paper can be used in production of low quality paper
grades and it is available in limited quantities.
Another solution involves the fabrication of
chemical pulp from nonwood raw materials (Kassim
et al., 2015; Leponiemi, 2008). Nonwood plants are of
increasing interest as a potential cellulosic fiber
resource and these include cereal straw (wheat and
rice), flax, cotton, corn stalks, rapeseed stalks,
bagasse, reed etc. (Marin et al., 2017; Gavrilescu et al.,
2009).
Chemical pulp is largely obtained from wood
by means of suitable delignification processes
(Gavrilescu, 2013). Romania was a forest-rich
country, but the forests surface decreased from 40% in
1900 to 28% in 1948. According to Romsilva, the
surface of Romanian forests accounted 25% in 2013
(Romsilva, 2015). The main causes of deforestation in
Romania in last decades were excessive logging and
deficient reforestation. As a consequence, wood
consumption in industry decreased year by year. Due
to the lack of wood and the continuously increasing of
pulpwood price, Romanian pulp mills were gradually
closed, so that no pulp mill is working anymore today.
For this reason, Romanian paper mills were obliged to
replace the pulpwood with recovered paper as fibrous
source (Vlase et al., 2012).
Paper recycling is environmental friendly, but
involves large investment in new plants and
implementation of appropriate technologies for paper
production. Compared to the chemical pulp,
secondary fibres represent an inferior raw material and
only low-grades quality of paper can be produced.
Because recovered paper is used several times in
producing paper, the cellulosic fibres gradually lost
their papermaking potential. Periodically, a certain
quantity of secondary fibres must be replaced with
virgin fibers (chemical pulp) in order to assure the
required quality of fibrous raw materials. For this
reason, addition of virgin fibers in the fibrous circuit
1766
of paper manufacture is necessary (Gavrilescu et al.,
2014).
Virgin cellulosic fibres are produced from
pulpwood or nonwoods. Corn stalks are among the
nonwood waste with great potential to be used as
resource for virgin fibers. Romania is an important
producer of corn among EU countries, since more than
14 million tones were produced in 2017 (Eurostat,
2018). The corn stalks yield was estimated at more
than 12 million tones and represents a valuable
agricultural waste. There are several traditional ways
to use the corn stalks: to feed animals, for
incorporation into soil to regenerate soil humus, or for
household heating. A new direction refers to chemical
processing of stalks to obtain bio-based chemicals
(Puitel et al., 2012).
The goal of this paper is to investigate the
posibility to replace wood with corn stalks as resource
for fibrous materials required in the production of
paper used in corrugated board manufacturing.
Chemical composition of corn stalks collected from
Moldova region of Romania was determined. Natron
pulping using sodium hydroxide was selected in order
to estimate its delignification potential. A large range
of pulps in terms of yield and lignin content was
obtained. The strength properties of pulps were
determined and the possibility for their use in
obtaining papers for corrugated board is discussed.
2. Materials and methods
Stalks from current production of corn (Zea
mays L.) in Moldova region of Romania were
naturally dried in laboratory and then chipped at 3-5
cm length. The average moisture of stalks was 10%.
Chemical composition of stalks was determined,
according to the standards listed in Table 1.
The corn stalks pulp was obtained using a 10L
laboratory rotating digester with automatic control of
temperature (Puitel et al., 2015). An amount of 400 g
oven dried (o.d.) stalks was filled in the digester and
120-220 g o.d. pulp was obtained, depending on the
pulping conditions. Cooking liquor was prepared in
the laboratory by dissolving NaOH in tap water. The
following cooking conditions were used:
- heating time: 25-40 min., depending of cooking
temperature;
- sodium hydroxide addition: 12-16% on o.d. stalks;
- cooking temperature: 120-160 oC;
- time at cooking temperature: 30-90 min.
Table 1. Standards used for chemical composition
of corn stalks
Component, % Standard
Cellulose TAPPI T 17 wd-70, 1970
Pentosans TAPPI T 223 cm-01, 2001
Lignin TAPPI T 222 om-02, 2002
Extractives TAPPI T 204 cm-07, 2007
Ash TAPPI T 211 om-02, 2002
 Environmentally friendly cellulosic fibers from corn stalks

After digestion, the brown stock was defibered
and washed with water until no color in the resulting
liquor was observed. Washed pulp was screened using
a vibratory screen with 0.25 mm slots. The screened
yield of pulps was gravimetrically determined and the
screened pulps were analyzed for their lignin content
expressed as Kappa number (TAPPI T 236 om-06,
2006). The pulps were refined in a Jokro mill and
transformed in handsheets in accordance with the
Rapid Köthen method (ISO 5269-3, 2008). The pulp
handsheets were tested regarding their strength
properties according to the standards listed in Table 2.
Table 2. Strength properties testing of natron pulp
from corn stalks
Strength property Standard
ISO 1924:3,
Tensile strength
2005
Bursting strength ISO 2758, 2014
Corrugating Medium Test - CMT ISO 7263, 2011
Short-span compression strength - SCT ISO 9895, 2008
3. Results and discussion
The corn at maturity is composed of root,
stalk, leaves, cobs and ear. After harvesting the cobs,
the plant material taken from the field consists of
stalks, leaves and ears, of which the stalks represent
about 85% (by mass). Corn stalks contains pit, which
represents up to 90% of the stalk volume. Chemical
composition of the constituent parts of corn stalks is
presented in Table 3.
Corn stalks have a similar chemical composition
related to other pulping raw materials. A comparison
with Romanian pulpwood main species reveals that
corn stalks have less cellulose, less lignin and higher
content of pentosans and ash (Table 4). Comparative
to wheat straw, corn stalks contain less cellulose, more
lignin and similar percentages of extractives and ash.
The most important advantage of corn stalks is the
lower lignin content than that of wood. This means
that corn stalks can be delignified using less
chemicals, lower temperatures and shorter time,
compared with wood delignification.
The lower content of cellulose, the high content
of extractives and ash are the main drawbacks of corn
stalks as a pulping raw material. Table 5 illustrates the
pulping conditions and the main properties of natron
pulp from corn stalks. As Table 5 shows, the yield and
lignin contents of pulps strongly depend on pulping
parameters. The pulp yield ranges between 29.7-
55.1% and Kappa number between 21.4-64.4 units.
Pulp yield correlates with lignin content as Fig. 1
shows. By using higher alkali addition (14-16%) and
moderate temperatures (140-160oC), corn stalks can
be delignified at low Kappa numbers (no more than 40
units), while the pulp yield ranges 30-40%. If the
cooking is performed at low alkali addition (12%
NaOH) and low temperature (120oC), pulps having
higher lignin content (Kappa number 40-65 units) and
yields (40-55%) are obtained, depending on the
cooking time. It is obvious that to obtain pulps having
better yields, low alkali addition and temperatures
must be used.
It is known that lignin content is the one of
main factor influencing pulp strength properties.
Figure 2 presents the influence of Kappa number on
tensile index of corn stalks pulp. It is evident that
lignin content has a great influence on pulp tensile
index. As Fig. 2 shows, tensile index records the best
values at pulp Kappa number of 35-45 units. A similar
behaviour was found regarding burst index of corn
stalks pulps. As it is presented in Fig. 3, burst index
values are the best at Kappa number of 30-45 units,
and it is clear that tensile and burst indexes show the
same dependence on the pulp lignin content.
It is already known that both burst strength and
short span compression strength are the most
appreciated among strength properties of papers for
corrugated board (Campean et al., 2017; Gavrilescu
and Toth, 2007). In addition, crushing resistance of the
paper flutes is the key parameters of the corrugating
medium (Gavrilescu, 2017). For these reasons, both
short span compression strength (SCT) and crushing
resistance (Corrugating Medium Test – CMT) were
analyzed concerning their dependence on pulp lignin
content. Fig. 4 shows the influence of lignin content
on SCT index of pulp.

Table 3. Chemical composition of corn stalks constituent parts

Chemical composition, %
Component Cellulose Lignin Pentosans Ash
(SiO2)
Stalks 39.4 20.1 25.2 0.81
Pit 31.6 17.0 27.3 0.79
Leaves 31.3 17.4 21.2 7.0

Table 4. Comparison between chemical composition of corn stalk and other raw materials for pulping

Chemical composition, %
Raw material
Cellulose Lignin Pentosans Extractives Ash
Corn stalk 38.0 18.5 26.0 6.0 5.1
Wheat straw* 43.1 17.5 27.7 5.5 5.3
Softwood** 48-53 27-33 7-11 0.5-1.6 0.1-0.4
Hardwood** 44-50 18-25 11-20 0.1-0.6 0.2-0.6
* Data according to Puitel et al. (2012); ** Data according to Petrovici and Popa (1997)

1767
 Cheșcă et al./ Environmental Engineering and Management Journal 17 (2018), 7, 1765-1771

Table 5. Natron pulping condition of corn stalks and pulp properties

AA, Time, Kappa TI, BI, CMT, SCT,

No. Temp. °C Yield, %
% min number N.m/g kPa.m2/g N.m2/g N.m/g
1 12 120 30 55.1 64.4 52.1 2.2 2.40 37.3
2 12 140 30 47.3 48.8 60.3 2.5 2.34 36.5
3 12 160 30 41.1 45.4 63.2 2.4 2.26 35.5
4 12 140 60 42.2 45.7 66.0 2.9 2.22 34.8
5 12 160 90 40.0 42.0 68.3 3 2.20 33.5
6 14 120 60 39.5 41.8 68.1 2.8 2.14 33.8
7 14 140 30 38.3 41.4 67.4 2.8 2.12 33. 5
8 14 140 60 35.0 35.7 65.0 2.9 2.12 33.0
9 14 140 90 31.7 26.8 65.0 3.0 1.84 31.7
10 14 160 60 32.9 29.3 64.0 2.8 1.96 32.5
11 16 120 90 33.1 30.1 64.0 2.8 2.00 29.5
12 16 140 30 32.3 27.5 60.2 2.8 1.94 29.2
13 16 140 60 30.8 23.6 60.5 2.6 1,84 27.2
14 16 160 30 29.7 21.4 54.0 2.3 1.68 25.4
AA - Alkali Addition; TI - Tensile Index; BI – Burst Index; CMT – Corrugating Medium Test index; SCT – Short-span Compressive Test index

60

55

50

45
Pulp yield, %

40

35

30

25

20
15 25 35 45 55 65 75
Pulp Kappa number

Fig. 1. Influence of corn stalk lignin content on pulp yield

70
Tensile index of pulp, N.m/g

65

60

55

50

45

40
15 25 35 45 55 65 75
Kappa number of pulp

Fig. 2. Influence of corn stalk lignin content on tensile index of pulp

1768
 Environmentally friendly cellulosic fibers from corn stalks

Fig. 4 reveals that, by increasing pulp lignin This fact is of particular importance for the paper used
content, SCT index of the pulp handsheets increases, in corrugated board manufacture. It is obviuos that
but no maximum values is recorded. The same handsheets having higher SCT and CMT indexes are
evolution undergoes the dependence of CMT index on obtained from lignin-rich pulps, as resulted from Figs.
pulp lignin content (Fig. 5). CMT index of pulp 4 and 5.
continuously increases as the lignin content of pulp The strength properties of corn pulp presented
increases too. Comparing Figs. 4 and 5 it is in Figs. 3 and 4, comparative with the recommended
unequivocal that SCT and CMT indexes of pulp values of burst and SCT indexes listed in Table 5
depend in a similar manner on lignin content. The prove that corn straws pulp can be a valuable fibrous
explanation of this behavior is that the rigidity of source in producing of papers for corrugated board. In
cellulosic fibres boosts as lignin content increases, and this respect, corn pulp can be an attractive solution for
this determines the enhancing of handsheets rigidity. Romanian paper sector.
3.5

3
Burst index kPa.m2/g

2.5

1.5

1
15 25 35 45 55 65 75
Kappa number of pulp

Fig. 3. Influence of corn stalk lignin content on burst index of pulp

40
37.5
35
SCT index, N.m/g

32.5
30
27.5
25
22.5
20
17.5
15
15 25 35 45 55 65 75
Kappa number of pulp

Fig. 4. Influence of lignin content on SCT index of pulp

2.7

2.4
CMT index, N.m2/g

2.1

1.8

1.5
15 25 35 45 55 65 75
Kappa number of pulp

Fig. 5. Influence of lignin content on CMT index of pulp

1769
 Cheșcă et al./ Environmental Engineering and Management Journal 17 (2018), 7, 1765-1771
Table 5. Recommended values of strength properties of paper used for production of corrugated board (ECO, 2012)
Paper grade Burst index,
kPa.m2/g
Testliner 1 More than 3.0
Testliner 2 2.5-3.0
Testliner 3 2.0-2.5
4. Conclusions
Corn stalks can replace pulpwood in the
production of fibrous materials for paper industry. The
main advantages refer to the fact that corn stalks
contain less lignin than wood and can be delignified
using less chemicals, lower temperatures and shorter
time. The high content of extractives and ash are the
main drawbacks of corn stalks as a raw material for
pulp manufacture.
Natron pulp yield strongly correlates with
lignin content and the values depend on pulping
parameters. It is obvious that to obtain pulps having
better yields, low alkali addition and temperatures
must be used.
The relatively high strength properties of corn
pulp prove that corn straws pulp can be a valuable
fibrous source in producing of papers for corrugated
board. In this respect, corn natron pulp can be an
attractive solution for Romanian paper sector.
Acknowledgments
This work was supported by a grant of the Romanian
National Authority for Scientific Research and Innovation,
CNCS/CCCDI – UEFISCDI, project number PN-III-P2-2.1-
BG-2016-0016, within PNCDI III.
References
Bajpai P., (2015), Green Chemistry and Sustainability in
Pulp and Paper Industry, Springer International
Publishing Switzerland, Heidelberg, New York,
Dordrecht, London.
Binkley D., Fisher R., (2012), Ecology and Management of
Forest Soils, John Wiley and Sons, New York.
Campean T., Grad F., Gradinariu, C., Patrascu C.,
Gavrilescu M., Gavrilescu D.A., (2017), Eco-friendly
corrugated board and sustainable packaging
manufacturing, Environmental Engineering and
Management Journal, 16, 705-714.
CEPI, (2017), Key Statistics 2016 European Pulp and Paper
Industry, Edited by CEPI, Brussels, Belgium, On line
at:
http://www.cepi.org/system/files/public/documents/pu
blications/statistics/2015/Key%20Statistics%202014%
20FINAL.pdf.
ECO, (2012), European Corrugated Base Paper List, On line
at:
http://www.dssmith.com/contentassets/84c1a31eae1c4
cb183d8419e872b0745/cepi-containerboard-list-of-
grades.pdf.
Eurostat, (2018), Agricultural production – crops, Main
statistical findings – Cereals, On line at:
http://ec.europa.eu/eurostat/statisticsexplained/index.p
hp/ gricultural_production_-_crops#Cereals.
1770
Short span compression
index (N.m/g)
More than 17.5
15.5-17.5
13.5-15.5
FAO, (2014), State of the World’s Forests. Enhancing the
Socio-Economic Benefits from Forests, Food and
Agriculture Organization of the United Nations, Rome,
Italy, On line at:
http://www.fao.org/docrep/016/i3010e/i3010e.pdf.
Foresteurope, (2011), Full State of Europe's Forests 2011
Report, On line at: http://www.foresteurope.org.
Fortuna M.E., Simion I.M., Gavrilescu M., (2012),
Assessment of sustainability based on LCA - case of
woody biomass, Cellulose Chemistry and Technology,
46, 493-510.
Gavrilescu D., (2013), Pulping Fundamentals and
Processing, In: Pulp Production and Processing: From
Papermaking to High-Tech Products, Popa V.I. (Ed.),
Smithers Rapra Technology Ltd., Shawbury, UK, 35-
70.
Gavrilescu D., Craciun G., (2012), Kraft Pulping (in
Romanian), Texte Press, Dej, Romania.
Gavrilescu D., Tofanica B.M., Puiţel A.C., Petrea P., (2009),
Sustainable use of vegetal fibers in composite
materials. Sources of vegetal fibers, Environmental
Engineering and Management Journal, 8, 429-438.
Gavrilescu D., Puiţel A.C., Câmpean T., Grad F., (2014),
Sustainable use of recovered paper - Romanian paper
industry. Part II. Environmental impact, Environmental
Engineering and Management Journal, 13, 1909-1915.
Gavrilescu D., Toth S., (2007), Corrugated Cardboard (in
Romanian), T3 Press, Sf. Gheorghe, Romania.
Gavrilescu D., (2017), Quantities and Measurement Units
(in Romanian), Performantica Press, Iasi, Romania.
Gavrilescu M., Campean T., Gavrilescu D.A., (2017),
Extending Production Waste Life Cycle and Energy
Saving by Eco-Innovation and Eco-Design: The Case
of Packaging Manufacturing, In: Nearly Zero Energy
Communities, Proceedings of the Conference for
Sustainable Energy (CSE) 2017, Visa I., Duta A. (Eds.),
Springer Proceedings in Energy, 611-631.
Ghinea C., Petraru M., Simion I.M., Sobariu D., Bressers
H.T.A., Gavrilescu M., (2014), Life cycle assessment
of waste management and recycled paper systems,
Environmental Engineering and Management Journal,
13, 2073-2085.
Gimona A., Poggio L., Brown I., Castellazzi M., (2012),
Woodland networks in a changing climate: Threats
from land use change, Biological Conservation, 149,
93-102.
Hetemaki L., Hanninen R., Moyseev A., (2013), Market and
Market Forces for Pulp and Paper Products, In: The
Global Forest Sector: Changes, Practices, and
Prospects, Hansen E., Panwar R., Vlosky R. (Eds.),
CRC Press, New York, 99-128.
ISO 1924-3, (2005), Paper and board - Determination of
tensile properties, Part 3: Constant rate of elongation
method (100 mm/min), International Organization for
Standardization-TC 6 Paper, Board and Pulps, On line
at:
http://www.iso.org/iso/iso_catalogue/catalogue_tc/cata
logue_detail.htm?csnumber=32206.
 Environmentally friendly cellulosic fibers from corn stalks
ISO 5269-3, (2008), Pulps - Preparation of laboratory sheets
for physical testing, Part 3: Conventional and Rapid-
Köthen sheet formers using a closed water system,
International Organization for Standardization-TC 6
Paper, Board and Pulps, On line at:
http://www.iso.org/iso/catalogue_detail.htm?csnumber
=39340.
ISO 9895, (2008), Paper and board - Compressive strength -
Short-span test, International Organization for
Standardization -TC 6 Paper, Board and Pulps, On line
at:
http://www.iso.org/iso/iso_catalogue/catalogue_tc/cata
logue_detail.htm?csnumber=41400.
ISO 2758, (2014), Paper - Determination of bursting
strength, International Organization for
Standardization-TC 6 Paper, Board and Pulps, On line
at:
http://www.iso.org/iso/iso_catalogue/catalogue_tc/cata
logue_detail.htm?csnumber=61487.
Kassim A., Aripin A., Hatta Z., Daud Z., (2015), Exploring
Non-wood Plants as Alternative Pulps: From the
Physical and Chemical Perspectives, In: ICGSCE 2014
- Proceedings of the International Conference on
Global Sustainability and Chemical Engineering,
Hashim M.A. (Ed.), Springer Science+Business Media,
Singapore, 19-24.
Kissinger G., Herold M., De Sy V., (2012), Drivers of
Deforestation and Forest Degradation: A Synthesis
Report for REDD+ Policymakers, Lexeme Consulting,
Vancouver, Canada, 17-21, On line at:
https://www.forestcarbonpartnership.org/sites/fcp/files
/DriversOfDeforestation.pdf_N_S.pdf.
Ladd B., Laffan S.W., Amelung W., Peri P.L., Silva L.C.R.,
Gervassi P., Bonser S.P., Navall M., Sheil D., (2013),
Estimates of soil carbon concentration in tropical and
temperate forest and woodland from available GIS data
on three continents, Global Ecology and Biogeography,
22, 461-469.
Leponiemi A., (2008), Non-wood pulping possibilities - a
challenge for the chemical pulping industry, APPITA
Journal, 61, 235-243.
Marin N., Puitel A.C., Chesca A.M., Gavrilescu D., (2017),
Response surface modeling of wheat straw pulping
using sodium carbonate and sodium hydroxide
mixtures, Cellulose Chemistry and Technology, 51,
745-753.
Piazza A., Santanu Roy S., (2015), Deforestation and
optimal management, Journal of Economic Dynamics
& Control, 53, 15-27.
Petrovici V., Popa V.I., (1997), Wood Chemistry (in
Romanian), Lux Libris Press, Braşov, Romania.
Puitel A.C., Tofanica B.M., Gavrilescu D., (2012),
Environmentally friendly vegetal fibers based
materials, Environmental Engineering and
Management Journal, 11, 651-660.
Puitel A.C., Câmpean T., Grad F., Gavrilescu D., (2014),
Sustainable use of recovered paper in the Romanian
paper industry. Part II - Environmental impact,
Environmental Engineering and Management Journal,
13, 1909-1915.
Puitel A.C., Marin N., Petrea P., Gavrilescu D., (2015),
Lignocellulosic agricultural residues - a virgin fibre
supply solution for paper-based packaging, Cellulose
Chemistry and Technology, 49, 633-639.
Romsilva, (2015), National Forest - Forest Fund, General
Aspects of Forest, Public Property (in Romanian),
National Forest, Bucharest, Romania, On line at:
http://www.rosilva.ro/articole/prezentare_generala__p
_178.htm.
Scott G.M., (2011), Recovered Paper, In: Waste: A
Handbook for Management, Letcher T.M., Vallero D.
(Eds), Academic Press, San Diego, 137-150.
TAPPI T 17 wd-70, (1970), Cellulose in Wood, Technical
Association of the Pulp and Paper Industry - Tappi Test
Methods, Tappi Press, Atlanta, Georgia, USA, 2011.
TAPPI T 9 wd-75, (1975), Holocellulose in Wood,
Technical Association of the Pulp and Paper Industry -
Tappi Test Methods, Tappi Press, Atlanta, Georgia,
USA, 2011.
TAPPI T 223 cm-01, (2001), Pentosans in wood and pulp,
Technical Association of the Pulp and Paper Industry -
Tappi Test Methods, Tappi Press, Atlanta, Georgia,
USA, 2011.
TAPPI T 222 om-02, (2002), Acid-insoluble lignin in wood
and pulp, Technical Association of the Pulp and Paper
Industry - Tappi Test Methods, Tappi Press.
TAPPI T 211 om-02, (2002), Ash in wood, pulp, paper and
paperboard: combustion at 525°C, Technical
Association of the Pulp and Paper Industry - Tappi Test
Methods, Tappi Press, Atlanta, Georgia, USA, 2011.
TAPPI T236 om-06, (2006), Kappa Number of Pulp,
Technical Association of the Pulp and Paper Industry -
Tappi Test Methods, Tappi Press, Atlanta, Georgia,
USA, 2011.
TAPPI T 204 cm-07, (2007), Solvent Extractives of Wood
and Pulp, Technical Association of the Pulp and Paper
Industry - Tappi Test Methods, Tappi Press, Atlanta,
Georgia, USA, 2011.
Vlase R., Iftimi V., Gavrilescu D., (2012), Sustainable use
of recovered paper – Romanian paper industry,
Environmental Engineering and Management Journal,
11, 1657-1662.
Ward H.C., Kotthaus S., Grimmond C.S.B., Bjorkegren A.,
Wilkinson M., Morrison W.T.J., Evans J.G., Morison
J.I.L., Iamarino M., (2015), Effects of urban density on
carbon dioxide exchanges: Observations of dense
urban, suburban and woodland areas of southern
England, Environmental Pollution, 198, 186-200.
1771
1772
_______________________________________________________________________________
INSTRUCTIONS FOR AUTHORS
1. Introduction
Environmental Engineering and Management
Journal (EEMJ) is an international medium for
publication of Original Papers, Reviews, Case Studies,
Book Reviews on the fundamentals, applications in
environmental engineering and technologies, applied
environmental sciences, environmental health,
management, sustainable development, education for
sustainability. Advertising is also accepted with
contractual payment forms.
Submission of a manuscript implies that the work
described has not been published before (except in the
form of an abstract or as part of a published lecture, or
thesis); that it is not under consideration for publication
elsewhere; that its publication has been approved by all
coauthors, if any, as well.
Papers, books for review, offers to review, suggestions
and commercials (advertising) should be submitted to the
Editor-in-Chief All papers will be published in English.
Non-English speaking authors should seek the advice of
a professional language expert or an English speaker to
help translate the paper.
2. Legal requirements
The author(s) guarantee(s) that the manuscript is/will not
be published elsewhere in any language without the
consent of the copyright holders, that the rights of third
parties will not be violated, and that the publisher will not
be held legally responsible should there be any claims for
compensation.
Authors wishing to include figures or text passages that
have already been published elsewhere are required to
obtain permission from the copyright holder(s) and to
include evidence that such permission has been granted
when submitting their papers. Any material received
without such evidence will be assumed to originate from
the authors.
The author(s) are encouraged to transfer the copyright of
the article to the publisher upon acceptance of an article
by the journal, using the Authors’ Warranty and
Assignment of Copyright agreement. This transfer
enables the Editor to protect the copyrighted material for
the authors, but does not relinquish the author’s
proprietary rights. The publication of an article is
conditioned by the signature of the author to whom
correspondence should be addressed on this Authors’
Warranty and Assignment of Copyright that is provided
by the Editor.
Ethics in Publishing
For information on ethics in publishing for journal
publication see
http://omicron.ch.tuiasi.ro/EEMJ/plagiarism.htm
Conflict of interest
All authors are requested to disclose any actual or
potential conflict of interest such as any financial,
personal or other relationships with other people or
organizations concerning the submitted work that could
inappropriately influence, or be perceived to influence,
their work.
3. Editorial procedure
For original papers, an upper limit of 7000 words is
recommended (including Abstract, Keywords,
References, Figures and Tables), processed with MS
editing facilities.
For review papers (critical evaluation of existing data,
defined topics or emerging fields of investigation, critical
issues of public concern), an upper limit of 15000 words
is recommended (including Abstract, Keywords,
References, Figures and Tables).
Manuscripts should be written in English (American or
British usage is accepted, but not a mixture of these) and
submitted electronically, in .doc format (please do not
use .docx) to the Editor-in-Chief, at one (and only one) of
the following e-mail addresses: eemjournal at yahoo
dot com; eem_journal at yahoo dot com; eemjeditor
at yahoo dot com; eemj_editor at yahoo dot com;
eemjournal at gmail dot com; eemjeditor at gmail dot
com; eemjoffice at gmail dot com.
Please be sure to include your full affiliation and e-mail
address (see Sample manuscript).
When submitting the manuscript, it is mandatory to
include a cover letter to the editor. The cover letter must
state:
- that all authors mutually agree that the manuscript
can be submitted to EEMJ;
- that the manuscript contains the original work of
the authors;
 - the novelty in results/findings, or significance of
results;
- that the manuscript has not already been
published, or is not under consideration for
publication elsewhere.
Manuscripts are evaluated first in the Editorial Office (as
a preliminary condition for acceptance) in terms of
meeting the requirements of the journal, including
attempts of plagiarism. The authors are responsible for
the accuracy of the whole paper and references. Authors
will be notified about the registration of their contribution.
Only those contributions, which conform to the following
instructions, can be considered for the peer-review
process. Otherwise, the manuscripts are returned to the
authors, with observations, comments and annotations.
The peer review process is decisive for paper
acceptance. It could be done in several stages,
depending on the revision quality of the manuscript in
accordance with the requirements of paper evaluators.
Please do not transmit electronic data or requirements to
the publisher until your manuscript has been reviewed
and accepted for publication. Please follow the
instructions below.
A minimum of four suitable potential reviewers should be
provided by the authors. Please provide their name,
email addresses, and institutional affiliation. When
compiling this list of potential reviewers please consider
the following important criteria: they must be
knowledgeable about the manuscript subject area of the
manuscript; they must not be from the authors’ own
institution or country; they should not have recent (less
than five years) joint publications with any of the authors.
However, the final choice of reviewers is at the editors'
discretion.
4. Manuscript preparation
General:
Authors must follow the Instructions for authors strictly,
failing which the manuscripts would be rejected without
review. Editors reserve the right to adjust the formatting
style to conform to the standards of the journal.
Manuscripts should be concise, in 1.5 line spacing, and
should have 2 cm all over margins. The font should be
Times New Roman of size 12 points. The numbering of
chapters should be in decimal form. Ensure that each
new paragraph is clearly indicated, using TAB at 1.25 pts.
The text layout should be in single-column format.
Keep the layout of the text as simple as possible. Most
formatting codes will be removed and replaced on
processing the manuscript. However, do use bold face,
italics, subscripts, superscripts etc. Add line numbering
and page numbers. To avoid unnecessary errors it is
strongly advised to use the 'spell-check' and 'grammar-
check' functions of your word processor.
Title page will contain:
x A concise and informative title (Times New Roman
bold, all caps of size 14 points); the maximum length of
the title should be maximum 100 letters and spaces;
The full name(s) of the author(s) (first name, then last
name, with Times New Roman bold 12 points) should
be written below the title. The affiliation(s) and complete
postal address(es) of the author(s) will be provided
immediately after the full name of the authors and will
be written with Times New Roman 12 points. When the
paper has more than one author, their name will be
followed by a mark (Arabic numeral) as superscript; for
the corresponding author, an asterix will be added using
Word_Insert_Reference_Footnote_Symbol sequence.
Also, the full and e-mail addresses, telephone and fax
numbers of the corresponding author will be provided in
the footer of the first page, as: Author to whom all
correspondence should be addressed: email…..,
Phone…….., Fax…………
x Abstract: each paper must be preceded by an abstract
presenting the most important results and conclusions
in no more than 250 words. Do not include citations in
the Abstract.
x Keywords: three to five keywords should be supplied
after the Abstract for indexing purposes, separated by
comma, ordered alphabetically, using American
spelling and avoiding general and plural terms and
multiple concepts (avoid, for example, "and", "of"). Be
sparing with abbreviations: only abbreviations firmly
established in the field may be eligible.
x The text of the paper should be divided into
Introduction, Materials and methods (or
Experimental), Results and discussion,
Conclusions, References (for papers dealing with
environmental management, policy, education etc., the
Experimental part can be replaced by case-studies
presentation).
1. Introduction
State the objectives of the work and provide an adequate
background, avoiding a detailed literature survey or a
summary of the results.
2. Material and methods (or 2. Experimental)
Provide sufficient detail to allow the work to be
reproduced. Methods already published should be
indicated by a reference: only relevant modifications
should be described.
3. Results and discussion
Results should be clear and concise. Discussion should
explore the significance of the results of the work, not
repeat them. Avoid extensive citations and discussion of
published literature.
4. Conclusions
The main conclusions drawn from results should be
presented in a short Conclusions section. Do not include
citations in this section.
Formulae, symbols and abbreviations:
Formulae will be typeset in Italics (preferable with the
Equation Editor of Microsoft Office 2003) and should be
written or marked as such in the manuscript, unless they
require a different styling. The formulae should be
numbered on the right side, between brackets:
a3 = 3M / 4N (1)
Always refer in the text to Equations as (Eq. 1), Eqs. (1-
4) etc.
The more complex Chemical Formulae should be
presented as Figures.
Abbreviations should be defined when first mentioned in
the abstract and again in the main body of the text and
used consistently thereafter.
SI units must be used throughout.
Footnotes should be avoided.
References:
The list of References should only include works that are
cited in the text and that have been published.
References should be cited in the text in brackets
(Harvard style) as in the following examples:
(Chisti, 1989), (Gavrilescu and Roman, 1996), (Moo-
Young et al., 1999).
References should be alphabetically listed at the end
of paper, with complete details, as follows:
Books: Names and initials of authors, year (between
brackets), chapter title, title of the book (italic), editors,
edition, volume number, publisher, place, page number:
Mauch K., Vaseghi S., Reuss M., (2000), Quantitative
Analysis of Metabolic and Signaling Pathways in
Saccharomyces cerevisiae, In: Bioreaction
Engineering, Schügerl K., Bellgardt K.H. (Eds.),
Springer, Berlin Heidelberg New York, 435-477.
Faber K., (2000), Biotransformations in Organic
Chemistry – A Textbook, vol. VIII, 4th Edition,
Springer, Berlin-Heidelberg-New York.
Handbook, (1951), Handbook of Chemical Engineer, vol.
II, (in Romanian), Technical Press, Bucharest,
Romania.
Symposia volumes: Names and initials of authors, year
(between brackets), paper title, symposium name,
volume number, place, date, page numbers:
Clark T.A., Steward D., (1991), Wood and Environment,
Proc. 6th Int. Symp. on Wood and Pulping
Chemistry, Melbourne, vol. 1, 493-498.
Journal papers: Names and initials of authors, year
(between brackets), full title of the paper, full name of the
journal (italic), volume number (bold), first and last page
numbers:
Tanabe S., Iwata H., Tatsukawa R., (1994), Global
contamination by persistent organochlorines and their
ecotoxicologcial impact on marine mammals, Science
of the Total Environment, 154, 163-177.
Patents: Names and initials of authors, year (between
brackets), patent title, country, patent number (italic):
Grant P., (1989), Device for Elementary Analyses. USA
Patent, No. 123456.
Dissertations: Names and initials of authors, year
(between brackets), title, specification (PhD Thesis, MSc
Thesis), institution, place:
Aelenei N., (1982), Thermodynamic study of polymer
solutions, PhD Thesis, Institute of Macromolecular
Chemistry Petru Poni, Iasi, Romania.
Star K., (2008), Environmental risk assessment
generated by natural hazards, MSc Thesis, Institute of
Hazard Research, Town, Country.
Legal regulations and laws, organizations:
Abbreviated name, year (between brackets), full name of
the referred text, document type, author, URL address:
ESC, (2007), Improving access to modern energy
services for all fundamental challenge, Economic and
Social Council, ENV/DEV/927, On line at:
http://www.un.org/News/Press/docs/2007/envdev927
.doc.htm.
EPA, (2007), Biomass Conversion: Emerging
Technologies, Feedstocks, and Products,
Sustainability Program, Office of Research and
Development, EPA/600/R-07/144, U.S.
Environmental Protection Agency, Washington, D.C.,
On line at:
http://www.epa.gov/Sustainability/pdfs/Biomass%20
Conversion.pdf.
EC Directive, (2000), Directive 2000/76/EC of the
European Parliament and of the Council of 4
December 2000, on the incineration of waste, Annex
V, Official Journal of the European Communities, L
332/91, 28.12.2000, Brussels.
GD, (2004), Governmental Decision No. 1076/2004
surnamed SEA Governmental Decision, regarding the
procedure for strategic environmental impact
assessment for plans or programs, Romanian Official
Monitor, Part I, No. 707 from 5th of August, 2004.
Web references
The full URL should be given in text as a citation, if no
other data are known. If the authors, year, title of the
documents are known and the reference is taken from a
website, the URL address has to be mentioned after
these data:
Burja C., Burja V., (2008), Adapting the Romanian rural
economy to the European agricultural policy from the
perspective of sustainable development, MPRA,
Munich Personal RePEc Archive, On line at:
http://mpra.ub.uni-
muenchen.de/7989/1/MPRA_paper_7989.pdf
Web references must not be listed separately, after the
reference list.
All references must be provided in English with a
specification of original language in round brackets.
Citation in text
Please ensure that every reference cited in the text is
also present in the reference list (and vice versa). Do not
cite references in the abstract and conclusions.
Unpublished results, personal communications as well
as URL addresses are not recommended in the
references list..
Citation of a reference as "in press" implies that the item
has been accepted for publication.
Papers which have been accepted for publication should
be included in the list of references with the name of the
journal and the specification "in press".
References style
Text: All citations in the text may be made directly (or
parenthetically) and should refer to:
- single author: the author's name (without initials, unless
there is ambiguity) and the year of publication: “as
previously demonstrated (Smith, 2007)”; “as Smith
(2007) demonstrated”
- two authors: both authors' names and the year of
publication: (Arnold and Sebastian, 2008; Smith and
Hansel, 2006; Stern and Lars, 2009)
- three or more authors: first author's name followed by
"et al." and the year of publication: “As has recently been
shown (Werner et al., 2005)…”, “Kramer et al. (2000)
have recently shown ...."
Citations of groups of references should be listed
first alphabetically, then chronologically.
Examples: "…..as demonstrated (Aden, 1996a, 1996b,
1999; Allan and Jones, 1995; Han et al., 2007; Weiss
and Schmidt, 1988)”.
Abbreviations
Define all abbreviations at their first mention in the text.
Ensure consistency of abbreviations throughout the
article.
Acknowledgements
Include acknowledgements in a separate section at the
end of the article before the references and do not,
therefore, include them on the title page, as a footnote to
the title or otherwise. List here those individuals who
provided help during the research (e.g., providing
language help, writing assistance or proof reading the
article, funding supports etc.).
Footnotes
Footnotes must be avoided. Do not include footnotes in
the Reference list.
Table footnotes
Indicate each footnote in a table with a superscript
lowercase letter.
Tables
Draw the Tables in grid format using a basic, solid line
style without shadows.
Ensure that the data presented in Tables do not duplicate
results described in Figures or elsewhere in the paper.
Figures
Number Figures consecutively in accordance with their
appearance in the text. All illustrations should be
provided in camera-ready form, suitable for reproduction,
which may include reduction without retouching.
Photographs, charts and diagrams are all to be referred
to as Figure(s) and should be numbered consecutively,
in the order to which they are referred.
Figures may be inserted preferably as black line
drawings. They should be pasted on, rather than taped,
since the latter results in unclear edges upon
reproduction.
Ensure that each illustration has a caption, placed below
the Figure. Supply also captions separately, not attached
to the figure. A maximum limit of 8 Figures are allowed
per manuscript.
A caption should comprise a brief title (not on the Figure
itself) and a description of the illustration. Keep text in the
illustrations themselves to a minimum but explain all
symbols and abbreviations used. Multiple Figures can be
expressed as one Figure (for e.g. 1a, 1b, 1c etc...), while
retaining the maximum limit of 6.
ALL Figures must be submitted in either .jpg or .tiff
format with a very good resolution (but do not submit
graphics that are disproportionately large for the
content).
Figures and Tables must be embedded in the text.
Proofs
Proofs will be sent to the corresponding author (by e-
mail) and should be returned within 72 hours of receipt.
Corrections should be restricted to typesetting errors;
any other changes may be charged to the authors.
Paper in Electronic Format:
Authors are asked to submit their final and accepted
manuscript as an attachment to one of the above-
mentioned e-mail addresses. Use the .doc format (not
.docx !).
5. Page charge
There is no charge per printed page for regular papers.
6. Reprints
The corresponding author will be provided with a .pdf file
of the paper via e-mail, free of charge. A hard copy Of
the issue containing the paper can be provided on
request, for a fee. This request will be formulated when
the final form of the manuscript (the electronic one) is
provided.
7. Additional procedures for the editorial
management of Environmental Engineering
and Management Journal
Starting with volume 13/2014, the members of the
Scientific Advisory Board and the Corresponding authors
will receive each issue of the journal in electronic .pdf
format. Printed copies can be delivered on request.
An author cannot appear on more than two papers per
regular issue. An author cannot appear on more than
three papers in a Guest Editor/ Conference issue.
The evaluation/peer-review process of manuscripts
submitted for Guest Editor / Conference issues
published according to journal Editorial Procedure and
Policy will be handled and evaluated as with regular
manuscripts, so that only consistent papers will be
published. The manuscripts must be sent at least six
months before the presumed data of publication. Those
manuscripts which do not fulfill the journal requirements
will be rejected. This is to discourage superficiality in the
development of the manuscript, from formal and scientific
points of view. No amount of money will be refunded
following the rejection of manuscripts.
No more than 160 pages and no more than 20 papers
are acceptable for regular issues. No more than 180
pages and no more than 25 papers are opportune for
special issues. The number of pages is based on the
published version of an article, not on the submitted
version (for instance, when page breaks are changed
and when images are enlarged for easier viewing).
An issue can include, as an exception, a number of 25-
30 papers as a maximum. This situation could appear
when the authors of an already accepted paper need and
ask for early publication, from the position in the list of
accepted papers. In this circumstance, a yearly
submission to our journal (500 Euro) will be asked for
each already accepted paper, for its publication in
advance.
A newly-received manuscript can be evaluated by
priority, when a deposit of 200 Euro shall be paid by
authors in advance. If the manuscript is deemed
unsuitable for publication, the authors will not be eligible
for a refund.
Failure to pay mandatory charges may result in the paper
being withheld from early publication.
In all cases, the corresponding author will sign an
agreement with the Editor-in-Chief, and the Director of
the EcoZone Publishing House according to journal
Editorial Procedure and Policy.
Automated software is used for checking against
plagiarism. In order to avoid false results, the generated
reports are cautiously checked and confirmed by journal
staff. We do not send plagiarism reports to authors. Any
manuscripts which do not prove to be at least 90%
original will be rejected. The rejection for plagiarism may
arise at any phase of the editorial / review process, even
if the manuscript was accepted for publication. The
authors are solely responsible for the originality of their
submission. Therefore, any manuscript should be
carefully checked for such inconsistencies before
submitting to EEMJ.
The authors are required to follow academic publishing
rules, to adhere to the principles of scientific ethics and
to sign the Authors’ warranty and Copyright transfer.
Detailed information concerning these issues may
be found of the EEMJ website, under Editorial
Procedure and Policy.