Development of a life cycle assessment tool for construction and maintenance

of asphalt pavements
Yue Huang
a,
*
, Roger Bird
b
, Oliver Heidrich
c
a
Scott Wilson Ltd., 12 Regan Way, Chetwynd Business Park, Nottingham NG9 6RZ, UK
b
School of Civil Engineering and Geosciences, Newcastle University, Cassie Building, Claremont Road, Newcastle upon Tyne NE1 7RU, UK
c
Safety, Environment and Quality Management Ltd., Ouseburn Building, Albion Row, Newcastle upon Tyne NE6 1LL, UK
a r t i c l e i n f o
Article history:
Received 29 October 2007
Received in revised form 14 May 2008
Accepted 19 June 2008
Available online 8 August 2008
Keywords:
Asphalt pavements
Life cycle assessment
Recycling
Sustainable construction
a b s t r a c t
The increasing use of recycled materials in asphalt pavements calls for environmental assessment of
such impacts as the energy input and CO
2
footprint. Life cycle assessment (LCA) is being accepted by the
road industry for such purpose. It aims to quantify and collate all the environmental impacts from
the life time of the product or process. This paper reviews relevant LCA resources worldwide, identies
the knowledge gap for the road industry, and describes the development of an LCA model for pave-
ment construction and maintenance that accommodates recycling and up-to-date research ndings.
Details are provided of both the methodology and data acquisition. This is followed by a discussion of
the challenges of applying LCA to the pavement construction practice, and recommendations for fur-
ther work. In the case study, the model is applied to an asphalt paving project at London Heathrow
Terminal-5 (LHR), in which natural aggregates were replaced with waste glass, incinerator bottom ash
(IBA) and recycled asphalt pavements (RAP). Production of hot mix asphalt and bitumen was found to
represent the energy intensive processes. This is followed by data analysis and sensitivity check. Fur-
ther development of the model includes expanding the database to accommodate the recycling and
maintenance practice in the UK, and taking into account the effect that roadwork has on trafc
emissions. The LCA model can be further tested and calibrated as a decision support tool for sustainable
construction in the road industry.
2008 Elsevier Ltd. All rights reserved.
1. Introduction
Recycled and secondary materials are increasingly used in as-
phalt pavements, in terms of tonnage and the variety [1]. The dual
benets of saving landll space and reducing quarry demand,
however, have not come without a cost. Simply diverting the waste
from other industries to aggregates supply has already been
questioned for its energy and CO
2
footprint, waste glass for exam-
ple. The scepticism comes from a mixture of academia, research
consultancy and government organisation, as previous studies have
indicated that using waste glass for construction aggregates overall
consumes more energy and releases more CO
2
than sending them
to landll [24]. Recycling or reuse of asphalt materials needs up-
to-date studies on the associated environmental impacts including
energy use, emissions, leaching, etc. A life cycle approach is gaining
ground in meeting the needs of sustainable construction [5].
Accredited by a number of industries already, life cycle assessment
(LCA) is being accepted and applied by the road industry, to mea-
sure and compare the key life-time environmental impacts of as-
phalt products and laying processes [6].
The life cycle assessment starts with a denition of the aim and
scope of the study. Its main work resides in the development of an
inventory (LCI), in which all the signicant environmental burdens
from the life time of the product or process will be quantied and
compiled. This is followed by an impact assessment (LCIA) calcu-
lating and presenting the result in a predened way that supports
comparison or further analysis. The concept and working phases of
LCA are described in the ISO14040 [7].
The application of LCA in civil engineering, initially as a tool for
assessing solid waste management options, has started only in the
last decade. Relevant practice in roads and asphalt pavements,
notably where recycled and secondary materials are involved, is
limited. Besides giving the knowledge of products environmental
performance, LCA results are also able to support marketing or
environmental labelling. For instance, the ISO14025 Type III Envi-
ronmental Product Declaration (EPD), which enables the informed
comparison between products that full the same function, re-
quires quantied environmental information based on in-
dependently veried LCA results [8].
* Corresponding author. Tel.: 44 (0) 115 907 7000; fax: 44 (0) 115 907 7001.
E-mail address: yue.huang@scottwilson.com (Y. Huang).
Contents lists available at ScienceDirect
Journal of Cleaner Production
j ournal homepage: www. el sevi er. com/ l ocat e/ j cl epro
0959-6526/$ see front matter 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jclepro.2008.06.005
Journal of Cleaner Production 17 (2009) 283296
2. LCA resources for the asphalt industry
In recent years, highway authorities in the UK have preferred
maintenance and rehabilitation over new road construction. The
concept by UK Transport Research Laboratory (TRL) of design for
long-life pavements in light of resource efciency (materials, en-
ergy, etc.) and the requirement for speed repair would conne
a roadwork in the future to the top fewlayers of the pavement only
[9]. A working LCA model shall therefore duly reect the mainte-
nance and recycling in roadwork; and data specic to UK road in-
dustry are preferred. Specically, it should be:
Internationally recognised, including the methodology and
supporting database;
Populated with current and relevant data;
Having as many as possible variables represented in the road
practice;
Forward compatible for data update or formulas revision.
2.1. Existing and developing LCA tools
The US Environmental Protection Agency (EPA) is hosting an
index of LCA resources worldwide including books and journals,
conference proceedings, websites, software and databases, and
case studies since 1998 [10]. European LCA Hub contains similar
resources of LCA tools, services and data [11]. EPAs pilot study in
the late 1990s demonstrated that LCA can help select the envi-
ronmentally preferable method for asphalt pavement treatment
[12]. A hybrid IO (inputoutput) model was used in Japan looking
at the life cycle emissions of CO
2
from a motorway covering both
the construction and operation stage [13]. According to the World
Business Council for Sustainable Development (WBCSD), the po-
tential values of generic data sets, technology assessment and
marketing are viewed by the cement industry as high or mid-
high in using LCA [14]. These works, as they built up the framework
of LCA model and database, paved the way for introducing and
applying LCA to the road sector, based on which the following LCA
tools for roads were developed.
(1) In 19931995, Swedish Environmental Research Institute (IVL)
developed the rst of its kind life cycle inventory (LCI) model of
road construction and maintenance for Swedish National Road
Administration. The 2nd version was released in 2001 [15].
(2) An LCI study focused on the asphalt including the use of
recycled asphalt pavements (RAP) was initiated in 1998 by the
European Asphalt Pavement Association (EAPA) and Euro-
bitume. IVL was commissioned to carry out the project. The 3rd
draft was released in 2005 [16].
(3) In 19971999, Eurobitume conducted a partial LCI study on
bitumen (straight run, paving grade 50/70), covering the life
cycle from crude oil extraction to renery deposit [17].
(4) Technical Research Centre of Finland (VTT) published in 1996
a comparative LCA study on environmental impacts of asphalt
Transport Transport
Production of
secondary aggregates
Transport
Transport
Production of
bitumen
Transport
Transport
Transport
Transport
Transport
Transport
Transport
Production of cold mix asphalt
Production of
natural aggregates
(crushed rocks for
example)
Production of hot mix asphalt
Excavator/Wheel
loader operation
Crusher/Screening
machine operation
Drying and heating
Mixing
Mixing
Surface planing Asphalt paving
Asphalt rolling
Production of
emulsifier
Waste collection
Plant recycling
Production of emulsion
Application of
tack coat
Crude oil extraction
Oil refining
Bitumen storage
Petroleum
products
Transport
Pavement recycling
Transport
Transport
Pavement Life Expectancy
Fig. 1. Unit processes in asphalt pavement construction.
Y. Huang et al. / Journal of Cleaner Production 17 (2009) 283296 284
and concrete pavements [18]. Later in 2001, an LCA model was
developed by VTT for Finnish National Road Administration
addressing the use of industrial by-products (coal y ash, blast
furnace slag, etc.) in roads [19].
(5) In 2005, an LCA model of road construction using municipal
solid waste incineration (MSWI) bottom ash was developed by
the Technical University of Denmark (DTU) [20].
(6) In the UK, the Built Research Establishment (BRE) published the
Environmental Proles in 1998 providing a database on
environmental performance of building materials and prod-
ucts, as well as the methodology for applying LCA to the con-
struction sector [21].
2.2. Need for a new LCA model
There are important ndings from previous LCA studies,
which can be taken as a starting point for further applying LCA to
Cold in-situ recycling
Hot in-situ recycling
Remix
Overlay
Ex-situ recycling
Mix the old materials
with new and paving
Compaction
Applying tack coat
Transport
to disposal

Cold planing
Paving new
materials
Transport to
asphalt plant

Production of
asphalt
(including RAP)
Transport
and paving
Heat and scarify
Repave
Mix and lay the
old materials
Apply
stabilising agent
Recycler operation
2-Mixing 1-Milling
3-Paving 4-Grading
Cement Lime
Emulsion Foam bitumen
Fig. 2. Treatment of pavement surface and placement of asphalt.
Table 1
Worksheets in the LCA model
Worksheet Description Sub-worksheet
Process parameters Data on transport distance and fuel efciency, energy consumption of unit processes in
a pavement project
Energy in transport
Energy in materials production
Energy in pavement construction
Pavement parameters Data on pavement dimension and materials recipe, determine the materials tonnage in
a pavement project
Pavement dimension
Materials recipe
Pavement life time
Unit inventory Inventory gures for unit operation of transport, materials production and pavement
construction
Energy production
Combustion of fossil fuel
Transport vehicle operation
Construction vehicle operation
Project inventory Unit inventory data are aggregated into the unit of the pavement project Production process
Transport process
Construction process
Characterisation results Inventory results are assigned to dened impact categories, characterised by selected models
and presented by category indicators
Global warming
Acidication
Photo-oxidant formation
Human toxicity
Eco-toxicity
Eutrophication
Y. Huang et al. / Journal of Cleaner Production 17 (2009) 283296 285
the road and asphalt industry. Nevertheless, the problems of
simply applying one of those LCA models described above to the
UK road sector are stated below [22]. The barriers can be grouped
into 5 categories.
Category 1: relevance
Low relevance to the road and asphalt industry, such as the
BREs model;
Data from non-UK sources may not represent the UKs in-
dustry average. This is of particularly a concern when ap-
plying the model to real case studies;
Category 2: adaptability
Some data are quite old, or the underlying assumptions and
calculation formulas unknown, such as Eurobitumes LCI
results;
Some data are drawn on a xed material recipe, haulage
distance, production process or machinery that cannot be
generalised for use in other studies;
Category 3: compliance
Model or database developed before the ISO14040 was
issued (in 1997) and revised (in 2006) may not be able to
fully comply with it;
Category 4: scope
The models above are generally focused on one or a few
environmental impacts, such as energy and air emissions in
the VTTs model, and leaching in the DTUs model;
The inclusion of recycled materials is varied, but generally
limited, such as RAP in the IVLs model, MSWI bottom ash
in the DTUs model;
Category 5: availability
Practical models are not accessible, due to commercial
restriction.
The fact that there is nothing quite right available on the market
has generated the need for developing a new LCA model on top of
existing resources that can enhance the level of acceptance of the
results in the UK.
3. Unit process denition and inventory development
A owchart of the unit processes in asphalt pavement con-
struction is illustrated in Fig. 1.
The owchart outlines the main phases of the construction
process. At the same time, it is a brief description which does not
Spreadsheets in LCA Model
Process Parameters (WS1)
Data Input
Progress
Data collection and
validation
Pavement Parameters (WS2)
Unit Inventory (WS3)
Project Inventory (WS4)

Characterisation Results (WS5)
Relating data to unit
process
Relating data to
functional unit


Data aggregation
Refining system
boundary
Energy
consumption
Transport
distance
Pavement
dimension
Materials
recipe
Energy
production&
combustion
Transport
vehicle
operation
Characterisation model
Construction
vehicle
operation

Life Cycle Impact
Assessment
Goal & Scope
Definition

Materials
production
Pavement
construction
Transport
L
i
f
e

C
y
c
l
e

I
n
v
e
n
t
o
r
y

A
n
a
l
y
s
i
s

Fig. 3. Structure of the LCA model and procedures for inventory analysis.
Table 2
Conversion factors and caloric values of fossil fuels [24]
Electricity Diesel Burning oil LPG (liqueed petroleum gas) Natural gas Coal
Unit MJ/kWh MJ/kg MJ/kg MJ/kg MJ/m
3
MJ/kg
Value 3.6 45.7 46.2 49.5 39.6
a
25.6
a
When extracted.
Y. Huang et al. / Journal of Cleaner Production 17 (2009) 283296 286
tell details of the process or what machinery is used. For instance,
the treatment of old pavement surface and placement of fresh as-
phalt may take a number of different forms including ex situ (plant)
recycling, hot in situ (place) recycling, cold in situ (place) recycling,
overlay, etc. (see Fig. 2).
Based on the guidelines in the ISO14044 [23], the construction
process needs to be dened on a level that data for unit processes
are: (1) specic to a pavement project therefore the assumptions
made can be minimal and (2) easy to collect and aggregate. The
experiences of obtaining data and establishing the inventory in
developing this LCA model include:
Identify the appropriate data source; data should be up-to-
date, recognised and the reference accessible;
The boundary of data and any assumptions made should be
unambiguously stated. Pay extra attentionwhen data required
in a process come from more than one place, as the data
boundary and underlying assumptions in there may be
different;
In the case that alternative data exist, state which one will be
used in the model and justify the selection. It is advisable to
run the model later using the defeated data as well, for sen-
sitivity check and data review;
Present the data alongside their sources in both the manu-
script and computing tool.
This model, during its development, has been applied to, and
tested by, 3 real case studies of asphalt paving projects in the UK,
which in return build up the scope and adaptability of the
computing tool. An example is given for the London Heathrow
Terminal-5 (LHR) case study in Section 5.
Case study 1 (AprilNovember 2005): asphalt in-lay at Chapel
Ash, Wolverhampton, alternative asphalt recipe and layer
thickness;
Case study 2 (JanuaryFebruary 2007): asphalt paving on an
access road, London Heathrow Terminal-5, use of glass, IBA
and RAP in base and binder course;
Case study 3 (MayJuly 2007): rehabilitation of A34, Stoke-on-
Trent, effect of speed delivery of the roadwork on reducing the
trafc emissions.
4. LCA model description
Microsofts spreadsheet, Excel, is selected for calculation and
graphical presentation of inventory results in this LCA model. The
model consists of 5 worksheets (WS): process parameters, pave-
ment parameters, unit inventory, project inventory and character-
isation results. Data in worksheet process parameters and
pavement parameters are specic to a project. Worksheet unit
inventory is made of calculating formulas and the life cycle in-
ventory of unit processes. Inventory results in the unit of that
pavement project are presented in the project inventory work-
sheet. The inventory loadings are characterised for impact assess-
ment; the characterisation models and factors can be found in the
characterisation results worksheet. Data in these worksheets are
linked by calculation formulas. For instance, when energy data on
Q
U
A
R
R
Y

S
I
T
E

A
S
P
H
A
L
T

P
L
A
N
T

R
O
A
D

S
I
T
E

C
R
U
D
E

O
I
L
S
O
U
R
C
E

B
I
T
U
M
E
N
E
M
U
L
S
I
O
N

P
L
A
N
T

E
M
U
L
S
I
F
I
E
R
P
R
O
D
U
C
T
I
O
N

Bitumen
Asphalt
Emulsifier
Hotmix process
Cold mix process
Surface dressing
RAP
(ex-situ)
Emulsion
Aggregates
Fig. 4. Transport in asphalt pavement construction.
Table 3
Emission standards for combustion of fossil fuels
Machinery in pavement project Reporting detail Name of CORINAIR activity Chapter
Materials plant Combustion of LPG/burning oil for asphalt/cement production Cement/asphalt concrete plant Group 3
a
: B331
Transport vehicles Heavy-duty (>3.5 t) diesel trucks Road transport Group 7
b
: B710
Construction vehicles Diesel engines in crushing equipment, paver, roller, etc. Other mobile sources and machinery Group 8
c
: B810
a
Group 3: combustion in manufacturing industry.
b
Group 7: road transport.
c
Group 8: other mobile sources and machinery.
Y. Huang et al. / Journal of Cleaner Production 17 (2009) 283296 287
a pavement process are altered, the project inventory and charac-
terisation results will change accordingly. The structure of the LCA
model and relations between worksheets in it are designed to the
features of a road project, and follow the ISO14040 norms (see in
Table 1 and Fig. 3).
4.1. Process parameters
The process parameters worksheet includes data on transport
distance (km), fuel efciency of transport vehicles (litre/kmor litre/
kmt) and energy consumption per unit of materials production
(MJ/t) and pavement construction (MJ/m
2
) in a pavement project.
Data on energy consumption include both the amount and energy
type. The caloric value (see Table 2) dened by UK Department of
Trade and Industry (DTI) is used in this model to convert the vol-
ume of combusted fossil fuels into the universal energy unit (MJ). It
is noted that the selection of use of an energy type is often limited
in a specic industry, and the upstream emission loadings for these
energy types are different. Parameters can be grouped into energy
in materials production, energy in transport and energy in
pavement construction.
This model uses data from UK plants and contractors, as well as
other European studies reviewed above. Details can be found in
PhD thesis at Newcastle University, UK [25]. Data on energy con-
sumption are also available from other sources including US De-
partment of Energy, National Crushed Stone Association (NCSA),
Canadian National Research Council (NRC), etc. [26]. One of the
main elements of LCA is transparency, in that, a user or reviewer of
the model should be able to tell where the data in the model came
from, and what assumptions were used in making calculations. This
is duly followed in the development of this model. Data analysis
including sensitivity check on data sources can be carried out, as
shown later in the case study.
The mileage and vehicles for transport depend on the construc-
tion process and materials in use (see Fig. 4), so does the fuel ef-
ciency. Whencalculating the diesel consumption, transport vehicles
are normallyassumed to run at full load and empty onreturn. Fuel
consumption in these cases needs to be differentiated. A key feature
of the computing tool is the adaptability. Assumptions made like
above can be changed to reect the particulars in a project. Cargo
ship and rail locomotive may be employed for long-distance, heavy-
load haulage, depending on the availability and economics.
In asphalt laying assembly, the roller will pass up and down over
the freshly paved material for a certain number of passes as spec-
ied for that material. It must roll before the temperature drops too
much, so it can never get too far behind the paver, or too close
where the material is still too soft. Therefore the working speed of
paver and roller in a pavement project is restrained by each other;
the selected gure for LCA calculation must refer to the manage-
ment of that laying assembly [27].
4.2. Pavement parameters
The pavement parameters worksheet includes data on pave-
ment dimensions (surface area, layer thickness) and materials
recipe (ratio of coarse and ne aggregates, ller, bitumen, etc.). The
information it has on materials tonnage, together with data in
process parameters, will determine the workloads in a pavement
project for inventory calculation. Parameters can be grouped into
pavement dimension, materials recipe and pavement life time.
Data on asphalt tonnage in a project are normally available from
material supplier. This spreadsheet is using conditional formatting
which is able to warn the user of any illogical data input, for ex-
ample, the sum of components tonnage (or percentage) does not
equal to the total weight (or 100%). Bitumen emulsion is seen in
tack coat, chip seal (surface dressing) or cold mix asphalt. Both the
bitumen and emulsier contents are varied between these appli-
cations. Emulsion usage in tack coat and chip seal is measured in
the unit of kg/m
2
, whilst in cold mix asphalt, by weight ratio.
Different service life should be applied to the asphalt layers in the
pavement structure, for example, 12 years for surface course, 15
years for binder course, etc. Pavement life expectancy is an im-
portant factor affecting the inventory results, for it effectively in-
uences the denition of system boundary and functional unit in
the LCA study.
4.3. Unit inventory
In the unit inventory worksheet, an environmental input and
output inventory is built up for the unit processes in a pavement
project. Available inventory data for some primary processes (e.g.
energy production, vehicle engine operation) are presented rst,
followed by progressive calculations to get the inventory data on
other processes in the pavement project. Emissions from a process
have two aspects. One is the process itself (e.g. diesel engine oper-
ation, gas oil combustion); the other is the production of energy
consumed in that process. Figures on energy consumption of vehi-
cles andequipments come fromthe process parameters worksheet.
This worksheet can be grouped into energy production, combus-
tion of fossil fuel, transport vehicle operation and construction
vehicle operation, which is described in more detail below.
4.3.1. Energy production
Data on production of electric power come from the Union of
the Electricity Industry (EURELECTRIC), using the industry average
of 15 European countries [28]. A later version of 2005 is available,
only to EURELECTRIC members. The breakdown of fuel used in
electricity generation in the UK is available since 2002 [24], yet the
emissions (complete inventory) from the generation process are
Table 5
EU emission standard for stage III controlled (20 kW<P <560 kW) diesel engines
Engine power (kW) Emission (g/kWh) FC
a
(g/kWh)
NO
x
N
2
O CH
4
CO NMVOC PM NH
3
020 14.1 0.35 0.05 8.38 3.82 2.22 0.002 271
2037 6.40 0.35 0.05 5.50 1.10 0.60 0.002 269
3775 4.00 0.35 0.05 5.00 0.70 0.40 0.002 265
75130 3.50 0.35 0.05 5.00 0.50 0.30 0.002 260
130560 3.50 0.35 0.05 3.50 0.50 0.20 0.002 254
>560 14.4 0.35 0.05 3.00 1.30 1.10 0.002 254
a
FC: fuel consumption.
Table 4
EU emission standard (Euro IV) for heavy-duty diesel engines
Tier Test Emission (g/kWh)
Euro IV ESC&ELR
a
CO HC NO
x
PM Smoke (m
1
)
1.5 0.46 3.5 0.02 0.5
a
ESC: European Stationary Cycle; ELR: European Load Response.
Y. Huang et al. / Journal of Cleaner Production 17 (2009) 283296 288
not known. Data on production of diesel are taken from the IVLs
study covering the life stage of crude oil extraction, rening and
transport to the consumer [16]. The inventory loadings for pro-
duction of burning oil and LPG are assumed to be the same as
diesel. Alternative sources of inventory data on energy production
(electric power, natural gas and petroleum oil) may include the
National Atmospheric Emissions Inventory (NAEI) report [29], and
the BUWAL250 (database in SimaPro7).
4.3.2. Combustion of fossil fuel
Natural gas and petroleum oil (burning oil, LPG, etc.) are com-
busted in plants (for asphalt, emulsion, etc.) and construction ve-
hicles (paver, remixer, etc.) for heating purpose. Diesel is consumed
by engines in transport vehicles (truck, locomotive, etc.) and con-
struction vehicles (paver, roller, etc.). Emission limits on these fossil
fuel combustion processes are specied in the European Environ-
ment Agencys (EEAs) EMEP/CORINAIR Emission Inventory
Table 6
Classication and characterisation (selected)
a
Impact category Inventory loading Unit of characterisation factor Value of characterisation factor Source
Depletion of minerals Aggregates tonne minerals 1
Bitumen 1
Depletion of fossil fuels Energy (MJ) MJ
b
1 BRE [21]
Global warming CO
2
kg CO
2
-eq. (100 years) 1 IPCC [38]
CH
4
23
N
2
O 296
Stratospheric ozone depletion kg CFC
11
-eq. WMO [39]
Acidication SO
2
kg SO
2
-eq. 1 IIASA [40]
NO
x
0.7
c
NH
3
1.88
Photo oxidant (ground-level ozone, or fog)
formation
SO
2
kg C
2
H
4
-eq. 0.048 CML [41]
NO
x
0.028
c
CO 0.027
CH
4
0.006
NMVOC 1.0
Human toxicity Emission to air SO
2
kg 1,4-dichlorobenzene-eq. 0.096 CML [41]
NO
x
1.2
CO 2.4
HC
d
5.7E05
NMVOC 0.64
PM
10
0.82
NH
3
0.1
Heavy metals
e
5.1E05
Emission to fresh water HC
d
2.8E05
Heavy metals
e
950.6 (As)
22.9 (Cd)
12.3 (Pb)
1426.0 (Hg)
Eco-toxicity
f
Emission to air NMVOC kg 1,4-dichlorobenzene-eq. 3.2E11 CML [41]
HC
d
1480
Heavy metals
e
7.8E04 (As)
3.7E05 (Cd)
2.4E03 (Pb)
4.1E05 (Hg)
Emission to fresh water HC
d
1.1E04
Heavy metals
e
4.0E04 (As)
7.4E04 (Cd)
3.7E02 (Pb)
7.2E04 (Hg)
Eutrophication NO
x
kg PO
4
-eq. 0.13
c
CML [41]
NH
3
0.35
COD 0.022
Phosphate 1
Nitrate 0.1
Noise Noise/1000 vehicle km DALY 1.3(26)E03
g
SAEFL [42]
Depletion of landll space Solid waste m
3
landll space
a
IPCC: Intergovernmental Panel on Climate Change; WMO: World Meteorological Organisation; IIASA: International Institute of Applied System Analysis; CML: Institute of
Environmental Sciences, Leiden University; EMEP: Convention on Long-range Transboundary air pollution; SAFEL: Swiss Agency for the Environment, Forests and Landscape;
DALY: Disability Affected Life Years.
b
In normalisation phase (see Table 7), tonne of oil equivalent (TOE, 1TOE 41,868 MJ) is used, to be consistent in the unit with other environmental loadings, such as the
quarry depletion and emissions.
c
Figure for NO
2.
d
Figure for carcinogenic PAH (polycyclic aromatic hydrocarbons).
e
Figure for the total of Arsenic (As), Cadmium (Cd), Mercury (Hg) and Lead (Pb).
f
Figures in Eco-toxicity are the mean characterisation factor of fresh water aquatic eco-toxicity, marine aquatic eco-toxicity and terrestrial eco-toxicity.
g
Figure in the bracket is for night-time (22pm6am) journey (sleep disturbance); gure outside is for daytime (6am22pm) journey (communication disturbance). As for
comparison, the DALY of truck emissions (CO, NO
x
, HC and PM
10
) per 1000 vehicle kilometre is 1.14E03. If the time of the day of the transport is unknown, a day/night-time
split of 95:5 is assumed. If the road trafc data are provided in the unit of tonne kilometre, the following loading factors are assumed for conversion: 3.8 t for a 16 t truck, 7.0 t
for a 26 t truck and 10.8 t for a 40 t truck [43].
Y. Huang et al. / Journal of Cleaner Production 17 (2009) 283296 289
Guidebook [30]. It has 11 groups in it addressing the combustion
process in different industries. Those relevant to this LCA study are
shown in Table 3.
To avoid condentiality restrictions and the differences between
machine manufacturers, these emission limits, if available, are used
in this LCA model as the inventory loadings of fossil fuel consump-
tion. Alternative emission limits on diesel engines (including CO
2
,
CO, HC, NO
x
and particulates) are dened by European Automobile
Manufacturers Association (ACEA) and United Nations Economic
Commissions for Europe (UNECE), in the unit of g/km [31,32].
4.3.3. Transport/construction vehicle operation
A number of vehicle features (fuel type, age, mileage, etc.), op-
erational conditions (load, road layout, speed, acceleration, trafc
ow, etc.) as well as environmental factors (altitude, ambient
temperature, etc.) have an effect on the vehicles exhaust emissions
level; there is a lack of relevant model or database that measure
such emissions from heavy-duty trucks [33]. EU emission limits
(see Table 4, effective from October 2005) on heavy-duty diesel
engines are used in this LCA model as the inventory for diesel en-
gine operation of transport vehicles [34]. The limits (see Table 5,
effective from January 2006) on stage III controlled diesel engines
are used as the inventory for diesel engine operation of construc-
tion vehicles [30]. Missing data (e.g. SO
2
) are supplemented by the
data from the IVLs study [16].
Transport and construction vehicles are assumed in the LCA
study to run at their operating capacity as specied by contractors.
Transport vehicles for example, the unit inventory loadings of
transport vehicles operation (g/km) can be calculated by multi-
plying the fuel consumption (MJ/km) by the sum of engine oper-
ation (g/MJ) and fuel production (g/MJ [16]).
Unit inventory g=km
FC MJ=km engine g=MJ fuel production g=MJ
FC (MJ/km): fuel consumption, the amount of fuel added to
the engine, using lower heating value (Table 2), converted
from l/km;
Engine (g/MJ): emissions from engine operation, g emissions
per MJ work energy output from the shaft of the engine,
converted from g/kWh, Table 4 for transport vehicles and
Table 5 for other construction vehicles.
4.4. Project inventory
In the project inventory worksheet, unit inventory data for
materials production, transport and pavement construction are
aggregated into the unit of the pavement project, based on the
workloads calculated from pavement parameters (for materials
tonnage and pavement area) and process parameters (for trans-
port distance). The results can be grouped into materials pro-
duction, transport and materials placement. At the end of the
worksheet is the total of each environmental input (e.g. energy,
aggregates) and output (e.g. CO
2
, PM) for that pavement project.
The percentage that each process accounts for the total is also
available.
4.5. Characterisation results
A consensus has been formed on 6 key impact areas by the UK
asphalt industry, after a review workshop set up by the Rened
Bitumen Association (RBA), Quarry Products Association (QPA) and
the Highways Agency, and published by UK TRL [35]. This model
also refers to the review of existing LCIA methods [36], and the
Table 7
Normalisation factors
Impact category Characterisation result UK total Year Data source
Depletion of minerals Aggregates 214 Mt 2004 QPA [45]
Depletion of fossil fuels TOE 243,963E 03 2005 DTI [24]
Global warming CO
2
559.223 Mt 2004 NAEI [46]
Stratospheric ozone depletion CFC
11
88,000 t
a
2003 WMO [39]
Acidication SO
2
979 t 2003 NAEI [25]
Photo oxidant (ground-level ozone, or fog) formation NMVOC (replace C
2
H
4
)
b
1089 t 2003 NAEI [25]
Human toxicity NH
3
(replace 1,4-Dichlorobenzene)
b
300 t 2003 NAEI [25]
Eco-toxicity Heavy metals (replace 1,4-dichlorobenzene)
b
339.479 t
c
2001 Environment Agency [47]
Eutrophication NO
x
(replace PO
4
)
b
1570 t 2003 NAEI [25]
Noise DALY 499.4 billion vehicle km 2005 Department for Transport [48]
Depletion of landll space Landll disposal 75 Mt
d
2002 Environment Agency [49]
UK population 60,209,500 2005 Ofce of National Statistics [50]
a
Global gure.
b
Data on UK total emissions of C
2
H
4
and PO
4
are difcult to obtain. Therefore in the normalisation phase, NMVOC and NO
x
are appointed instead as the indicator for ground-
level ozone formation and eutrophication, respectively. Characterised results presented as C
2
H
4
-equivalent and PO
4
-equivalent are then converted, using the characterisation
factor in that impact category, into NMVOC-equivalent and NO
x
-equivalent, respectively. For the same reason, NH
3
and Heavy metals are used in the LCA study to replace 1,4-
Dichlorobenzene as the indicator for human toxicity and eco-toxicity, respectively.
c
Same as in Table 6, data on heavy metals refer to the total of Arsenic (As), Cadmium (Cd), Mercury (Hg) and Lead (Pb). The normalisation factors for each are presented in
Table 8.
d
Total tonnage of waste sent to landll.
Table 8
UK total of heavy metals to air and water [47] (Unit:tonne)
Arsenic Cadmium Mercury Lead Total
Emission to air N/A 5.070 8.820 194.000 207.89
Release to water 110.560 1.074 19.710 0.245 131.589
Total 110.560 6.144 28.530 194.245 339.479
Impact Area
Noise
H
i
g
h

L
o
w

Priority
Global Regional Local & Site
Minerals
Acidification
Fossil fuel
Landfill
Eutroph.
Global warming
Ozone depletion
Human-tox.
Eco-tox,
Fog
Fig. 5. Grouping and weighting of environmental impact categories.
Y. Huang et al. / Journal of Cleaner Production 17 (2009) 283296 290
methods recommended by UK BRE and the ISO14047 [21,37]. Based
on the ndings from above studies, 11 impact categories are se-
lected for use in this model (see Table 6), also presented in the
characterisation results worksheet are the selected assessment
method (characterisation model and characterisation factor). Al-
ternative models for some impact categories are identied in the
thesis [25].
Inventory loadings in this model are allocated in their full
amount to relevant categories as if they all go through the serial
processes [37]. The characterised result for an impact category is
the total of all the individually characterised loadings in that cat-
egory (see equation below). There are LCI loadings that have not
been assigned to, and characterised in, any of the impact categories.
Impact assessment of these loadings is expected in light of on-going
development of the environmental assessment (LCIA) method. A
number of emissions to water (e.g. BOD, Chloride) apply to this
case.
Characterisation result
X
i
Inventory loading
i
Characterisation factor
i
4.6. Optional phases after characterisation
Depending on the scope of an LCA study, the characterisation
results can be further divided by a reference value (normalisation),
which in this LCA study is the characterisation factor per UK capita.
The latest gures from the literature are presented in Table 7. This
LCA study proposes a grouping and weighting method, in accor-
dance with the Eco-points (see Fig. 5) developed by UK BRE for the
construction industry [44]. It is specied in the ISO14044 that
weighting is not recommended for use in comparative LCA study
[23]. Identication of signicant areas based on the inventory
loadings, and data analysis (including completeness check, sensi-
tivity check and consistency check) also can be carried out for
a pavement project, as shown later in the case study.
5. Case study: asphalt paving at London Heathrow (LHR)
Terminal-5
5.1. Project background, goal and scope denition
Previous LCA studies have questioned the environmental ben-
ets of using waste glass for construction aggregates in terms of
carbon footprint [3], especially when the recycling involves
a transportation of waste glass of more than 3040 km [51]. This
case study investigated the life cycle environmental impacts of
asphalt paving at LHR Terminal-5 access road in which natural
aggregates were partially replaced with waste glass, incinerator
bottom ash (IBA) and reclaimed asphalt pavement (RAP), and
compared the results to the pavement of the same size and function
but made using virgin aggregates only. This is followed by a dis-
cussion and data analysis referring to the most signicant variables
in this project. This case study is to test and calibrate the LCA model
described above. The ndings, presented as inventory loadings
(LCI), can be benecial to road engineers or researchers dealing
with recycling in roads.
Production of
aggregates
Production of
bitumen
Production of
asphalt
Placement of
asphalt
Production of
emulsion
Planing of old
asphalt surface
Placement of
tack coat
Glass
recycling
System Boundary I
n
c
i
n
e
r
a
t
i
o
n

Landfill
Indicates alternative
practice for comparison
Production of
emulsifier
Glass
collection

Indicates transport
Indicates omission
Fig. 6. System boundary of LHR terminal-5 project.
Fig. 7. Pavement structure in LHR terminal-5 project.
Y. Huang et al. / Journal of Cleaner Production 17 (2009) 283296 291
5.1.1. Data source, quality and allocation
Data needed for this LCA study are obtained primarily from
materials suppliers and contractors, Aggregate Industries UK Ltd.
(AI) in the project. The missing data come from those justied for
use in the model described above, which are a combination of lit-
erature and other European LCA results. Quality of these data is
analysed later in the interpretation phase. As the inherent prop-
erties of glass have not been changed by recycling, waste glass input
in this LCA study is counted the same way as stone aggregates.
Embedded energy in glass manufacture is therefore not included in
this LCA study.
5.1.2. System boundary
Product systems are dened as the asphalt layers (surface
course, binder course and base) in the LHR Terminal-5 project
constructed partially using glass, IBA and RAP (referred as new
proposal), compared to the asphalt layers of the same size and
function but containing virgin aggregates and binder only (referred
as conventional proposal): this is a comparative LCA study. It is
assumed that using those recycled materials has no measurable
effects on the asphalt layers life expectancy or technical constraints
on reuse or recycling when these layers are replaced.
The upstream boundary for recycled materials is set at the col-
lection point: bottle banks for glass, incinerators for IBA and road
site of the old asphalt pavement for RAP. Alternative ways of dis-
posal include transport of those materials to landll (see Fig. 6).
Boundaries, assumptions and data options made for conventional
materials are described above in the model. The transport of bi-
tumen and emulsier to emulsion plant is not included in this
study, for data are not available.
5.13. Functional unit
The function of asphalt surface is to provide a safe, comfortable,
economical and durable driving. Functional unit is dened as the
30,000 m
2
of the asphalt surface. This case study assumes the same
durability of asphalt layers between the two proposals. This is
reected in the denition of functional unit that includes elements
in only the construction stage. The pavement layers included for
study consist of 35 mm Stone Mastic Asphalt (SMA) surface course,
77 mm High Modulus Bituminous (HMB) binder course and
205 mm HMB base. Quarry aggregates, bitumen and emulsion,
waste glass, IBA and RAP are used (see Fig. 7). Technical assessment
of the Smatex for surface course is seen in UK TRL report [52].
5.1.4. Method of impact assessment
The nature and amount of work in the life cycle impact assess-
ment phase have been discussed with the client when dening the
scope of the study, and was considered purely an environmental
analysis process that needs very little input from road engineers
once the life cycle inventory is complete. Based on the clients in-
struction, and the fact that most of the procedures would repeat
those described earlier in the LCA model development (Section 4.5
and 4.6), this case study will proceed straight into the in-
terpretation phase, from the completion of the inventory analysis.
5.2. Inventory analysis
5.2.1. Pavement parameters
Pavement dimension, asphalt recipe and materials tonnage are
seen in Fig. 7. It is assumed that the only difference between the
two proposals is the use of waste glass, IBA and RAP. 10% IBA and
10% glass were used to replace coarse and ne aggregates, re-
spectively, in both binder course and base. 25% RAP (48% coarse,
47% ne, and 5% binder) was also used in both layers. The break-
down of materials usage in this project is seen in Table 9.
5.2.2. Process parameters
The transport vehicles and distance, and fuel consumption of
construction vehicles are seen in Table 10 and Table 11.
Table 9
Materials in LHR termianl-5 project (unit: tonne)
Natural aggregates Glass
c
IBA RAP Primary bitumen
Coarse Fine Filler Total
New proposal Surface course 1379.7 283.5 113.4 1776.6 113.4
Tack coat 7.2
a
Binder course 2402.6 1394 123 2530.2 139.4 240.3 1025
b
165.2
Tack coat 7.2
a
Base 7490.7 4391.1 516.6 8021.8 439.1 749.1 3228.8
b
476.2
Total 12328.6 578.5 989.3 4253.8 769.2
Conventional proposal Total 18094.6 824.8
a
Specic density of bitumen and emulsion is assumed to be 1 kg/litre.
b
US National Cooperative Highway Research Program (NCHRP) research indicated that residue binder in the RAP might need to be counted as active binder when the
replacement rate of RAP exceeds 20% [53]. UK TRL suggested a 50% recovery rate of the binder in porous asphalt that is recycled into thin surfacing accounting for up to 30% of
the newmixture [54]. In this case study, 50% residue binder in the RAP is assumed to be recovered in the newasphalt mixture. The rest 50% is counted as hardened black rock,
the aggregates portion of the new mixture.
c
Fuel consumption for collection of waste glass is 0.4 MJ/t [55]. The processes (crushing, screening, etc.) and energy input for making glass aggregates are assumed the same
as for making natural aggregates.
Table 10
Transport parameters
Freight Origin Destination Mileage Vehicle type Fuel consumption Payload Fuel efciency
Aggregates Bardon Hill quarry West Drayton 120 mi (193.1 km) Train 89 L/mi (5.05.6 L/km) 1729 t 0.85 L/t
a
Bitumen Southampton West Drayton 80 mi (128.7 km) Truck Data missing, the gures for 14 t truck
from the IVLs study are used for
calculation
Emulsion
b
York (to Crawley rst) LHR T5 230 mi (370.1 km) Truck
Asphalt West Drayton LHR T5 4 mi (6.4 km) Truck
Glass
b
Brentford West Drayton 11 mi (17.7 km) Truck
IBA
b
Edmonton West Drayton 20 mi (32.2 km) Truck
a
Fuel consumption in empty journey is counted in the calculation of fuel efciency.
b
The suppliers of emulsion, glass and IBA are Colas, Day Group Ltd. and Ballast Phoenix Ltd., respectively.
Y. Huang et al. / Journal of Cleaner Production 17 (2009) 283296 292
5.2.3. Inventory loadings calculation
Environmental inputs (raw materials, energy) and outputs
(emissions to air and water, solid waste) are calculated for the LHR
Terminal-5 project. For comparison, the tonnage of glass, IBA and
RAP in the asphalt is in turn set to zero, hypothetically (with the
asphalt recipe design altered, replacing that recycled portion with
virgin aggregates), to see how the LCI loadings change as a result.
Some of the inventory loadings in these scenarios are presented in
Table 12. Complete inventory results are shown in the spreadsheet.
It can be seen that RAP replacement has the greatest effects,
compared with glass and IBA. There are two reasons for the sig-
nicance: (1) its tonnage and (2) its dual effects of replacing ag-
gregates and reducing the input of primary bitumen, an energy
intensive product. Glass replacement causes more energy use and
emissions. This is due to the high consumption rate of fuel in waste
glass collection (442 MJ/t), compared to that of 42 MJ/t for aggre-
gates quarrying. Energy and CO
2
loadings for these recycling sce-
narios are presented in Fig. 8.
5.3. Interpretation
Not all the data required for this LCA study were available from
contractors. Some (e.g. transport) were obtained from industry
average not specic to this project; other data gap (e.g. glass/IBA
aggregates) has affected the denition of system boundary. Data
used in this case study were of mixed age, accuracy and applica-
bility. A full inventory of environmental loadings of asphalt prod-
ucts and processes is welcome for LCA use; very often, however,
only energy data are available.
5.3.1. Identication of signicant areas
In the LHR Terminal-5 project, asphalt mixing, bitumen and
aggregates production consumed approximately 62%, 23% and 6%,
respectively, of the energy total and resultantly, produced more
emissions than other processes. The use of recycled materials re-
duced, by about 7%, the primary bitumen input. Another signicant
benet of the recycling was the saving of 5766 t of natural aggre-
gates, and diverting 579 t and 989 t of waste glass and IBA, re-
spectively, from landll. Old asphalt planing is the desirable type of
recycled aggregates considering quantity, transport, resource ef-
ciency and recyclability of the asphalt layers. Further research is
needed to determine the recovery rate of residue binder from the
RAP.
Transport of aggregates accounted for more than 61% of all
diesel use for transport. This is due to the long haulage distance
(193 km) and materials tonnage. Railway locomotive with a higher
fuel efciency (0.17 MJ/t km) than trucks (0.460.94 MJ/t km)
was used for aggregates transport. Glass and IBA were obtained
from fairly local sources; the haulage of RAP which was applied on
site was not counted in the calculation. The road site in this project
was fairly close (6.4 km) to the hot mix asphalt plant, which
explained the diesel use of only 17% for asphalt transport.
5.3.2. Sensitivity check
The sensitivity check aims to determine the inuence of varia-
tions in data source, methodology and assumptions on the in-
ventory results. Normally it is carried out after the identication of
signicant areas. Materials production, notably the hot mix asphalt,
bitumen and aggregates, represented in the LHR Terminal-5 project
most of the energy use and emissions. The production process was
also found to be where more alternative data exist. Sensitivity
check is therefore carried out, on the effect of data source on the
inventory loadings of energy and CO
2
. A variation of 10% is con-
sidered by the authors as being signicant for this case study.
It can be seen from Table 13 and Table 14 that the source of data
on materials production does not have signicant effects on the
energy or CO
2
total in the inventory. However, the projects energy
total would be more than halved had the aggregates and bitumen
(emulsion) been mixed by the cold method. This is of course,
based on the assumption that aggregates grading and bitumen
content remain the same, and durability of the asphalt is not
changed as a result.
6. Discussion
It is the case studies that built up the scope and capacity of this
LCA model during its development. The case studies show how
more details of the asphalt pavement project went into the model,
how a case study learned from the previous ones and applied the
ndings, and howthe LCA model and case studies will benet from
each others advancement. The main improvements include the
following aspects. The key elements for a quality LCA study of as-
phalt pavements are summarised in Fig. 9.
Table 11
Machinery parameters
Width of
screed/roller (m)
Working speed
(m/hr)
Fuel
consumption (L/hr)
Emulsion applier 2 220 6.0
Paver 4.9 300
a,b
16.87
Roller 1.7 6000
b
12.5
a
Normally, the paver runs at a speed of 1012 m/min on surface and binder
course, and 810 m/min on the base.
b
Paving and rolling is nished in 2 passes on the base.
Energy and CO2 Loadings for Comparison
363 357 365 375 371
869 853 874 900 888
0
200
400
600
800
1000
T5 No Glass No IBA No RAP No Recycling
Energy Total (TOE) CO2 emission (t)
Fig. 8. Comparison of energy and CO
2
between recycling scenarios.
Table 12
Inventory loadings of LHR terminal-5 and alternative scenarios
Scenario of using recycled materials Inventory loadings (selected)
Aggregates (t) Bitumen (t) Energy (TOE) CO
2
(t) SO
2
(kg) NO
x
(kg) Solid waste (t)
Glass, IBA, RAP as in T-5 12328.6 769.2 363 869 2230 4310 5803.3
IBA, RAP as in T-5, no glass 12907.1 769.2 357 853 2220 4260 5224.8
Glass, RAP as in T-5, no IBA 13317.9 769.2 365 874 2240 4400 4813.8
Glass, IBA as in T-5, no RAP 16526.8 824.8 375 900 2360 4700 1548.3
Virgin materials only 18094.6 824.8 371 888 2370 4740 19.8
Y. Huang et al. / Journal of Cleaner Production 17 (2009) 283296 293
The communication skills with contractors, make concise and
to-the-point questionnaire for project specic data;
The level of details and sophistication of the model, when the
calculation process becomes swift and more adaptive at the
same time;
The incorporation of up-to-date and alternative data sources,
rather than rely on a low number of references which is prone
to limit the scope and accuracy of any LCA study;
The presentation of results, featured by graphic illustration and
data quality analysis (check for completeness, sensitivity and
consistency);
The development of the LCIA phase, provide both the manda-
tory and optional elements addressed in the ISO14040 series.
7. Conclusion and recommendations
Life cycle assessment makes an important part of the life cycle
approach as a tool to support decision making. This paper in-
troduced the concept of LCA; reviewed the existing LCA tools
worldwide; identied the knowledge gap for the UK road industry;
and described the development of an LCA model for asphalt
pavement construction. Details were provided of both the meth-
odology and data sourcing. A case study of applying the LCA model
to a real asphalt paving project in the UK was provided, referring to
the most signicant variables in the project. This was followed by
data analysis and sensitivity check, and a discussion of the expe-
riences obtained from this and other case studies.
To duly reect the current practice in UK road construction, the
LCA model should accommodate maintenance and recycling sce-
narios, with data specic to the UK road industry. The model should
represent as many as possible variables in a pavement project,
whilst remaining exible for data update and formula revision. A
practical model must be populated with good quality data. It must
also be tested and calibrated through real case studies. Data in this
LCA model come from a mixed source of UK plants, EU standards
and relevant European LCA results.
Still there is a roomfor improving both the wealth and quality of
data fed to the LCA study. Its application in road practice is rela-
tively new; inventory data for some materials and processes are yet
to be available. Many processes that do not have energy input were
scoped out of the LCA study, simply because the required emission
data were not documented. Companys Integrated Pollution Pre-
vention and Control (IPPC) application documents for their sites or
plants could be potential sources of data for those process related
emissions. On the other hand, innovative asphalt materials and
laying techniques emerge in response to the industry improve-
ment, which calls for an expanding database for LCA practitioners
that can accommodate these novelties. Where the required data for
a unit process come from more than one source, the compatibility
(date, boundary, underlying assumptions, etc.) of the data needs to
be studied. Data acquisition for LCA is further hurdled by com-
mercial restriction on some proprietary data. In summary, the main
challenges of applying LCA to pavement construction practice in-
clude the following aspects, which make the areas for further work:
Include the non-energy (process) related emissions in the
model;
Look for energy or inventory data on more secondary aggre-
gates in the asphalt;
Predict the life expectancy, and the way of disposal, of pave-
ment layers made using recycled materials;
Include the effect of road maintenance works on the trafc and
resultantly, the fuel use and emissions, helped by micro-sim-
ulation model (e.g. VISSIM).
Despite the challenges above, LCA is being accepted by the
road industry to measure and compare the key life time environ-
mental impacts of its products and construction processes, and
use the results for internal review or environmental labelling.
Table 13
Sensitivity check on data source and mixing method for asphalt production
Total energy
use (TOE)
Hot mix
production
Cold mix
production
Deviation
(, %)
Sensitivity
AI data 363 165 198, 54.6% Signicant
EAPA data 365
IVL data 164
Deviation (, %) 2, 5.5% 1, 0.6%
Sensitivity Insignicant Insignicant
Table 14
Sensitivity check on data source for bitumen and aggregates production
Total CO
2
emissions (tonne)
Deviation
(, %)
Sensitivity
Bitumen production Eurobitume data VTT data
869 908 39, 4.3% Insignicant
Aggregates production AI data VTT data
869 847 22, 2.5% Insignicant
Knowledge of
the Industry
Presentation of
LCA Results
Compliance with
ISO Standards
Quality of Data
Powerful
Computing Tool
Communication
Speed Quality
LCA Model
Development
Transparency
TARGET
Fig. 9. Elements for a successful LCA study of asphalt pavements.
Y. Huang et al. / Journal of Cleaner Production 17 (2009) 283296 294
Recommended applications of LCA in road paving include the
comparison of
Different asphalt composition and materials usage;
Recycled materials with virgin aggregates (like the LHR Ter-
minal-5 case study);
Different recycled materials (glass, RAP, etc.);
Different laying or recycling techniques (hot ex situ, cold in
situ, etc.) and maintenance options (depth, interval, etc.);
Asphalt with concrete (standard recipe for both, same function
in the pavement layer).
Despite previous LCA studies that have questioned the envi-
ronmental benets of recycling some waste (e.g. glass) for use as
aggregates based on such impacts as the carbon footprint, it is
advisable to study by LCA the environmental impacts of the
recycling in pavement projects. The reason is that these two
types of LCA studies normally do not share the same functional
unit or system boundary. For instance, in LCA of glass, the
functional unit is packaging a certain volume of liquid, while the
functional unit in LCA of roads is the provision of a certain area
of asphalt surface on the carriageway. Therefore, the LCA results
of the close-loop recycling do not negate the environmental
benets of the open-loop recycling that could be identied
where the quarrying limits, transport scenarios, landll re-
striction, etc. are taken into account.
Acknowledgments
An earlier version of this paper was presented at the 6th In-
ternational Conference on Sustainable Aggregates, Pavement En-
gineering and Asphalt Technology, JMU Liverpool, UK, 21st22nd
February 2007. Financial support and intellectual inputs from Ag-
gregate Industries UK Ltd. are greatly appreciated. We would like to
thank Dr. Paul Phillips and Bob Allen for providing technical advice
and data for case studies throughout this research.
References
[1] Greeman A. Road recycling is widespread. Road recycling a supplement to
world highways; June 2007.
[2] Dacombe PJ, Krivtsov V, Banks CJ, Heaven S. Use of energy footprint analysis
to determine the best options for management of glass from household
waste. In: International conference on sustainable waste management and
recycling, London; 2004.
[3] Thornton Grant, Hollins Oakdene. The impact of the carbon agenda on the
waste management business; 2006.
[4] WRAP. Environmental benets of recycling an international review of life
cycle comparison for key materials in the UK recycling sector. The Waste and
Resources Action Program; 2006.
[5] WSSD. Plan of implementation. World summit on sustainable development.
Johannesburg; 2002.
[6] Bird R, Clarke R, Donnelly T, Heidrich O, Huang Y. Life cycle and sustainability
indices for road paving materials [unpublished report]. School of Civil Engi-
neering and Geosciences, University of Newcastle upon Tyne; 2004.
[7] BS EN ISO14040. Environmental management - life cycle assessment - prin-
ciples and framework. British Standard Institution; 2006.
[8] BS EN ISO14025. Environmental labels and declarations type III environ-
mental declarations principles and procedures. British Standard Institution;
2006.
[9] TRL. Guidance on the development, assessment and maintenance of long-life
exible pavements. TRL Report 639. Transport Research Laboratory; 2005.
[10] EPA. LCA resources. Available from: http://www.epa.gov/NRMRL/lcaccess/re-
sources.html; 2007 [accessed 03.04.07].
[11] EUROPA. LCA resources directory. Available from: http://lca.jrc.ec.europa.eu/
lcainfohub/directory.vm; 2007.
[12] Schenck R. Using LCA for procurement decisions: a case study performed for
the US environmental protection agency. Environmental Progress 2000;19(2):
1106.
[13] Inamura H. Life cycle inventory analysis of carbon dioxide for a highway
construction project using inputoutput scheme: a case study of the Tohoku
expressway construction works. Graduate School of Information Sciences,
Tohoku University; 1999.
[14] WBCSD. Toward a sustainable cement industry (Substudy 6): what LCA can tell
us about the cement industry. World Business Council for Sustainable De-
velopment; 2002.
[15] Stripple H. Life cycle assessment of road a pilot study for inventory analysis.
2nd ed. Swedish Environmental Research Institute (IVL); 2001.
[16] Stripple H. Life cycle inventory of asphalt pavements. 3rd ed. Swedish Envi-
ronmental Research Institute (IVL); 2005.
[17] Eurobitume. Partial life cycle inventory or eco-prole for paving grade bi-
tumen; 1999.
[18] Hakkinen T, Makela K. Environmental adaption of concrete environmental
impact of concrete and asphalt pavements. Technical Research Centre of Fin-
land; 1996.
[19] Mroueh U-M, Eskola P, Laine-Ylijoki. Life-cycle impacts of the use of in-
dustrial by-products in road and earth construction. Waste Management
2001;21:2717.
[20] Birgisdottir H. Life cycle assessment model for road construction and use of
residues from waste incineration. Institute of Environment & Resources,
Technical University of Denmark; 2005.
[21] Howard N, Edwards S, Anderson J. BRE methodology for environmental pro-
les of construction materials, components and buildings, built research es-
tablishment; 1999.
[22] Huang Y, Bird R, Allen B. A life cycle assessment tool for recycling in
pavement construction. In: SETAC Europe 13th LCA Case Study Sympo-
sium with Focus on the Building and Construction Sector, Stuttgart,
Germany; 2006.
[23] EN ISO 14044 BS. Environmental management life cycle assessment re-
quirements and guidelines. British Standards Institution; 2006.
[24] DTI. Digest of United Kingdom energy statistics 2006. Department of Trade
and Industry; 2006.
[25] Huang, Y. Life cycle assessment of use of recycled materials in asphalt pave-
ments. Ph.D. thesis, School of Civil Engineering and Geosciences, Newcastle
upon Tyne, Newcastle University; 2007.
[26] Zapata P, Gambatese JA. Energy consumption of asphalt and reinforced con-
crete pavement materials and construction. Journal of Infrastructure Systems
March 2005:920.
[27] Lavin P. Asphalt pavements: a practical guide to design, production and
maintenance for engineers and architects. London: Spon; 2003.
[28] EURPROG. Programmes and prospects for the european electricity sector.
EURPROG Report, nal version. 26th ed. Union of the Electricity Industry;
1998.
[29] NAEI. UK emissions of air pollutants 1970 to 2003. National Atmospheric
Emissions Inventory; 2005.
[30] EEA. EMEP/Corinair emission inventory guidebook. European Environmental
Agency; 2005.
[31] UNECE. Protocol to the 1979 convention on long-range transboundary air
pollution to abate acidication, eutrophication and ground-level ozone. ECE/
EB.AIR/72. United Nations Economic Commissions for Europe; 1999.
[32] ACEA. Reducing CO2 emissions: working together to achieve better results.
Available from: http://www.acea.be/home_page; 2007 [accessed 25.06.07].
[33] TRL. A review of available road trafc emission models. TRL report 457.
Transport Research Laboratory; 2000.
[34] European Union. Directive 2005/55/EC of the European Parliament and of the
Council; 2005.
[35] TRL. A model set of asphalt sustainability indicators. TRL report 638. Transport
Research Laboratory; 2005.
[36] Pennington DW, Potting J, Finnveden G, Lindeijer E, Jolliet O, Rydberg T, et al.
Life cycle assessment (Part2): current impact assessment practice. Environ-
mental International 2004;30:72139.
[37] EN ISO 14047 BS. Environmental management life cycle impact assessment.
examples of application of BS EN ISO14042. British Standards Institution;
2003.
[38] IPCC. IPCC third assessment report, intergovernmental panel on climate
Change; 2001.
[39] WMO. Scientic assessment of ozone depletion: 2006 [prepared by the Sci-
entic Assessment Panel on the Montreal Protocol on Substances that Deplete
the Ozone Layer]. World Meteorological Organisation; 2006.
[40] Huijbregts M, Krewitt W, Verkuijlen E, Reijnders L, Heijungs R. Spatially ex-
plicit characterization of acidifying and eutrophying air pollution in life-cycle
assessment. Journal of Industrial Ecology 2000;4(3):7592.
[41] CML. CML impact assessment. Leiden; 2004.
[42] Muller-Wenk R. Attribution to road trafc of the impact of noise on health.
Environmental Series No.339. Swiss Agency for the Environment, Forests and
Landscape (SAEFL); 2002.
[43] Muller-Wenk R. Life-cycle impact assessment of road transport noise. dis-
cussion Paper No.077. Institute for Economy and the Environment, University
of St. Gallen; 1999.
[44] Dickie I, Howard N. BRE Digest446: assessing environmental impacts of
construction industry consensus BREEAM and UK ecopoints. Built Research
Establishment; 2000.
[45] QPA. Resource conservation. Available from: http://www.qpa.org/sus_natres_
cons01.htm; 2007 [accessed 04.01.07].
[46] NAEI. Greenhouse Gas Inventories for England, Scotland, Wales and Northern
Ireland: 19902004. National Atmospheric Emissions Inventory; 2006.
[47] EA. Pollution inventory data report: trends and analysis 19982001. The En-
vironment Agency; 2004.
Y. Huang et al. / Journal of Cleaner Production 17 (2009) 283296 295
[48] DfT. Transport statistics great Britain 2006 Edition. Department for Trans-
port; 2006.
[49] The Environment Agency. Waste management options. Available from: http://
www.environment-agency.gov.uk/subjects/waste/1031954/1031960/
?version1%26lang_e; 2007 [accessed 06.01.07].
[50] Ofce of National Statistics. Population estimates. Available from: http://www.
statistics.gov.uk/CCI/nugget.asp%3FID6; 2007 [accessed 04..01.07].
[51] Hopkins A, Foster C. Achieving life cycle benets through practical sustainable
wastemanagement opportunities, benets andbarriers associatedwiththe use
of glass waste as aggregates in coated roadstone. RMC Environment Fund; 2003.
[52] TRL. Assessment of SMATEX a Bardon thin asphalt surface course. TRL report
435. Transport Research Laboratory; 2000.
[53] NCHRP. Recommended use of reclaimed asphalt pavement in the SUPERPAVE
mix design method: technicians manual. NCHRP Project 912. National Co-
operative Highway Research Program; 2001.
[54] TRL. Feasibility of recycling thin surfacing back into thin surfacing systems.
TRL report 645. Transport Research Laboratory; 2005.
[55] Edwards DW, Schelling J. Municipal waste life cycle assessment (Part 2)
transportation analysis and glass case study. Transactions of the Institution of
Chemical Engineers. 1999;77:25974.
Y. Huang et al. / Journal of Cleaner Production 17 (2009) 283296 296