b i o m a s s a n d b i o e n e r g y 3 5 ( 2 0 1 1 ) 2 9 1 0 e2 9 2 4

Available at www.sciencedirect.com

http://www.elsevier.com/locate/biombioe

Biomass gasication in a 100 kWth steam-oxygen blown circulating uidized bed gasier: Effects of operational conditions on product gas distribution and tar formation
Xiangmei Meng*, Wiebren de Jong, Ningjie Fu, Adrian H.M. Verkooijen
Faculty of Mechanical, Maritime and Materials Engineering, Process and Energy Department, Energy Technology section, Delft University of Technology, Leeghwaterstraat 44, 2628 CA Delft, The Netherlands

article info
Article history: Received 8 December 2010 Received in revised form 9 March 2011 Accepted 11 March 2011 Available online 9 April 2011 Keywords: Dry distillers grains with solubles (DDGS) Circulating uidized bed Steam-oxygen blown gasication Tar formation Sulphur Carbon conversion efciency

abstract
Biomass gasication is one of the most promising technologies for converting biomass, a renewable source, into an easily transportable and usable fuel. Two woody biomass fuels Agrol and willow, and one agriculture residue Dry Distillers Grains with Solubles (DDGS), have been tested using an atmospheric pressure 100 kWth steam-oxygen blown circulating uidized bed gasier (CFB). The effects of operational conditions (e.g. steam to biomass ratio (SBR), oxygen to biomass stoichiometric ratio (ER) and gasication temperature) and bed materials on the composition distribution of the product gas and tar formation from these fuels were investigated. Experimental results show that there is a signicant variation in the composition of the product gas produced. Among all the experiments, the averaged concentration of H2 obtained from Agrol, willow and DDGS over the temperature range from 800 to 820  C was around 24 vol.%, 28 vol.% and 20 vol.% on a N2 free basis, respectively. A fairly high amount of H2S (w2300 ppmv), COS (w200 ppmv) and trace amounts of methyl mercaptan (<3 ppmv) on a N2 free basis were obtained from DDGS. Due to a relatively high content of K and Cl in DDGS fuel, an alkali-getter (e.g. kaolin) was added to avoid agglomeration during gasication. Higher temperatures and SBR values were favorable for increasing the mole ratio of H2 to CO and the tar decomposition but less advantageous for the formation of CH4. Meanwhile, higher temperatures and SBR values also led to higher gas yields, whereas a higher SBR caused a lower carbon conversion efciency (CCE%), cold gas efciency (CGE%) and heating values of the product gas due to a high steam content in the product gas. From solid phase adsorption (SPA) results, the total tar content obtained from Agrol was the highest at around 12.4 g/Nm3, followed by that from DDGS and willow gasication. The lowest tar content produced from Agrol, willow and DDGS using Austrian olivine (Bed 1) as bed materials was 5.7, 4.4 and 7.3 g/ Nm3, values which were obtained at a temperature of 730, 820 and 730  C, SBR of 1.52, 1.14 and 1.10, and ER of 0.36, 0.39 and 0.37, respectively. 2011 Elsevier Ltd. All rights reserved.

1.

Introduction

Due to the limited supply of conventional fossil fuels, the steep rise of energy consumption and rapid economic

developments, alternative renewable energy sources need to be widely explored in order to renew the energy sources and to comply with a sustainable development. Biomass, referring to all organic materials that originate from plants or

* Corresponding author. Tel.: 31 0152786987. E-mail address: X.Meng@tudelft.nl (X. Meng). 0961-9534/$ e see front matter 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.biombioe.2011.03.028

b i o m a s s a n d b i o e n e r g y 3 5 ( 2 0 1 1 ) 2 9 1 0 e2 9 2 4

2911

animals, is traditionally being used as energy source for cooking and heating, particularly in the developing countries. Biomass also plays an essential role in the energy supply of some European counties. For examples, 11% of the national energy supplies in Austria, 17% of the national energy supplies in Sweden and 20% of the primary energy in Finland were obtained from biomass [1]. Currently the interest in biomass utilization such as power generation and syngas production has continuously increased due to its signicant environmental benets. Transformation of biomass fuels into high quality energy carriers and other commercial products can be efciently achieved via either biological or thermochemical processes [2]. Among various biomass conversion technologies within thermochemical and biochemical platforms, biomass gasication has received the highest interest since it offers high conversion efciency and increases options for combination with various high efciency power generation systems using gas engines, gas turbines and fuel cells [3]. During biomass gasication, several parameters such as gasier type, reaction temperature, biomass fuels properties, bed materials and gasifying agent have a substantial inuence on product gas composition, carbon conversion efciency and tar formation. The reactors used for biomass gasication are quite diverse of which the design ultimately determines other parameters such as gasication temperature and the properties of the biomass used. Two of the most frequently used reactor types for biomass gasication are the xed bed and uidized bed (FB) reactor [4]. Current development activities on large scale biomass gasication have been mainly devoted to uidized bed technologies, since FB gasiers have excellent heat and mass transfer between the gas and solid phases with the best temperature distribution and they meet the challenges of wide variations in fuel quality with a broad particle size distribution. The gasifying agents for FB biomass gasication can be either air, steam, pure oxygen or their combination [5]. Air is the mostly used agent nowadays at demonstration or commercial scale because of its extensive low-cost availability. However, the product gas from FB airblown biomass gasication normally containing 50 vol.% N2 has a low heating value (4e7 MJ/Nm3) and it can be used only for electricity production or heat generation [6]. Steam gasication can produce a medium caloric value gaseous fuel (10e14 MJ/Nm3) with a relatively high H2 content and this has been deeply studied experimentally by many researchers [7e9]. However, steam gasication is a more complex endothermic process, which requires heat input as the heat necessary to the process is not directly supplied by the partial combustion of the feedstock during the gasication process at temperatures in excess of 800  C. One easy way for supplying heat is to add some air or oxygen to the gasier, so that an autothermal gasication process can be achieved [5,10e13]. However, due to the high price of the oxygen, the product gas distribution under different operation conditions for steam/ oxygen mixtures as oxidizer has been less studied. Gil et al. [5] pointed out that the concentration of H2 in the raw gas ranges between 14 and 30 vol.% depending on the operational conditions. Lv et al. [11] generally reported that a higher reactor temperature, proper oxygen to biomass stoichiometric ratio (ER), steam to biomass ratio (SBR) and smaller biomass particle size will contribute to increase the H2 production.

Siedlecki et al. [13] reported that the use of magnesite as bed material can signicantly increase the concentration of H2 compared with normal sand as bed material. Besides, biomass gasication in the entrained-ow gasier using oxygen [14] and air [15] has been also studied. Zhou et al. [14] reported that higher temperature favored H2 and CO production and an introduction of O2 to the gasier improved the carbon conversion, but lowered the H2/CO ratio of the syngas. However, according to Higman and Van der Burgt [16], entrained-ow gasiers generally operated at high temperature therefore requiring a very small biomass particle size of 100 mm or less to promote mass transfer and high oxygen for maintaining operation temperature. Furthermore, as one of most problematic by-product from biomass gasication, tar is highly dependent on the operational conditions such as reaction temperature, ER and gasifying agent used etc. To avoid various problems associated with tar condensation, formation of tar aerosols and polymerization to more complex structures, tar reduction is essentially important before the nal utilization of product gas. Several reviews of the current knowledge on the elimination of tars have been published [17e19]. According to Devi et al. [17], tar removal technologies can broadly be divided into two approaches: treatments inside the gasier (primary methods) and hot product gas cleaning after the gasier (secondary methods). The primary options include: (a) the proper selection of the operating conditions; (b) the use of a proper bed additives or a catalyst during gasication; and (c) a proper gasier design. Secondary methods include physical removal, thermal conversion and catalytic destruction of tars. Corella et al. [20] made an excellent comparison between the two methods and found no signicant difference in their effectiveness concerning tar reduction. The effectiveness of dolomite in a second reactor was only slightly higher than the effectiveness of the in-bed location. Sutton et al. [19] reported that a suitable combination of different primary and secondary treatments is likely to improve gasier performance and produce a syngas with minimum tar concentration. Primary methods are gaining much attention as these may eliminate or strongly reduce the need for downstream cleanup. Since extensive studies have been published on biomass gasication with air, steam and air steam as gasifying agent, this paper is mainly focused on understanding how different operational conditions with an emphasis on SBR affect the product gas distribution and tar formation produced during steam-oxygen blown biomass gasication. In this work, three different fuels: Agrol, willow and DDGS (Dry Distillers Grains with Solubles) have been tested on an atmospheric 100 kWth steam-oxygen blown CFB gasier. Among these three fuels, willow is a common woody biomass, Agrol is a commercial solid as a kind of wood pellets which are made from pure sawdust and shavings from sawmills using 100% virgin timber from sustainably managed plantation forestry, while DDGS is a by-product of the so-called dry-grind process to produce ethanol from wheat. During past years, the rapid growth in the production of ethanol led to a parallel rise in DDGS production. To avoid the adverse effect on ethanol production which could be caused by saturated DDGS market [21], the potential applications of DDGS need to be explored and

2912

b i o m a s s a n d b i o e n e r g y 3 5 ( 2 0 1 1 ) 2 9 1 0 e2 9 2 4

therefore its potentiality for gasication to produce a gaseous fuel is investigated in this work.

2.
2.1.

Gasication experiments
Biomass fuels, bed materials and additive

Agrol, willow, and DDGS used for the experiments were obtained from Lantmannen, Sweden. Chemical compositions of these fuels and their ash (obtained at 550  C) were analyzed by Forschungszentrum Julich GmbH, Germany, and are pre sented in Table 1. The following analytical methods were used for chemical analysis: Inorganic main components of the fuel were analyzed by Inductively Coupled Plasma Optical Emission Spectroscopy (ICP-OES). C, H, N, O and S were analyzed by CHNSO-IR spectrometry (LECO) analyzer, where C, H and S oxide formed were measured by the IR-Absorption of the combustion gases CO2, H2O and SO2, N was quantied based on the thermal conductivity of N2 and O2 was measured separately. Chlorine was analyzed by Cl-Ion chromatography (IC) after Wickbold combustion. The water content was measured after drying the milled DDGS in a vacuum oven at 105  C for 60 h. It can be seen in Table 1 that DDGS sample ashes are rich in K, Na and Mg, but the total quantied ash composition only sums up to 72%, which could be due to the unavailable analysis of P2O5. For all fuels, except for C, H, O, N, S, Cl main elements, some amounts of K, Na, Ca Si and so on were also detected. Giuntoli et al. [21] reported that there is ca. 29% P2O5 in the ash of another, similar DDGS sample. Agrol and willow are fairly

Table 1 e Chemical composition of biomasses and ashes of biomasses obtained at 550  C. Types of fuel Agrol Willow DDGS
12 67.2 15.5 4.82 48.2 6.54 31.2 5.52 0.76 0.21 0.05 0.01 2.87 0.26 38.07 8.87 0.15 10.04 1.86 e 0.33 0.02 10.04

Moisture (wt.% a.r.) 8 8 Proximate Analysis (wt.%, dry) Volatile matters 74.7 69.8 Fixed carbon 16.0 20.1 Ash content [%] 0.14 2.52 Ultimate Analysis (wt.%, dry) C 51.0 50.3 H 6.26 6.17 O 38.2 37.4 N 0.15 0.69 S 0.002 0.002 Cl 0.01 0.01 Elemental Ash Analysis (wt.% of Ash at 550  C) 2.68 4.23 Al2O3 BaO 0.41 0.06 CaO 33.30 26.31 1.32 1.54 Fe2O3 K2O 16.50 13.13 MgO 7.79 2.89 MnO 3.96 0.30 1.00 1.67 Na2O 6.46 33.80 SiO2 0.06 0.18 TiO2 C 5.24 3.61 Cl 0.07 0.11 2.46 2.27 SO3

clean and contain almost no sulphur and relatively low ash content. The proximate analysis of three fuels was carried out by using SDT Q600 from TA Instrument. This instrument is capable of simultaneously performing both differential scanning calorimetry (DSC) and thermo gravimetric analysis (TGA). Some technical details of this instrument are: Platinum/ PlatinumeRhodium (Type R) thermocouple, heating rate from ambient to 1000  C at 0.1e105  C/min and sample pans made of platinum of 40 mL, alumina of 110 mL, 40 mL, and 90 mL. Four different bed materials have been used during gasication. One bed material is untreated Scandinavian olivine (Bed 4), which contains 44.16 wt.% MgO, 43.9 wt.% SiO2, 9.44 wt.% Fe2O3 and trace NiO, MnO and Al2O3 etc. The other two bed materials are Austrian olivines which were received from Biomasse Kraftwerk Gussing GmbH & Co KG, Austria. The difference between these two Austrian olivines is that one was pre-treated (Bed 1) by about 1000 redox cycles in a fast internally circulating uidized bed (FICFB) real gasication process and the other one was untreated natural olivine (Bed 2). The untreated natural olivine (Bed 2) contains 48e50 wt.% MgO, 39e42 wt.% SiO2, 8e10.5 wt.% Fe2O3 and traces of Cr2O3, Mn3O3, CaO and Al2O3 etc. Another bed material (Bed 3) is a mixture of quartz sand and pre-treated olivine with a mass ratio around 50:50. The quartz sand consists of more than 99% SiO2, with trace amounts of Fe2O3, CaO, K2O, Na2O and MgO. The reason olivine has been chosen as bed material is due to its high attrition resistance and its attractive chemical composition (being a natural mineral containing magnesium, iron and silica). Since DDGS fuel has a relatively high K and Cl content, kaolin was chosen as an alkali-getting additive to prevent agglomeration. Kaolin can be transformed to metakaolinite particles which potentially adsorb potassium species. The main constituent of the kaolin is the mineral kaolinite (Al2Si2O5(OH)4), with a small amount of halloysite (Al2Si2O5(OH)4(H2O)2) [13,22]. Kaolin used in this work is a ne powder with a mean particle size of 12 mm. Around 50% of the particles is in the size range of 0.063e63 mm. The composition is 57.5% SiO2, 37.5% Al2O3, 3.1% K2O, 0.9% Fe2O3, and trace amounts of CaO, Na2O, MgO and TiO2. Kaolin was continuously added together with biomass with an amount of 3e10% of total feeding during DDGS gasication.

2.2.

Gasication setup and experimental procedure

The gasication experiments have been carried out using an atmospheric pressure 100 kWth steam-oxygen blown CFB gasier which is located in the laboratory of the Energy Technology Section of the Process and Energy Department at Delft University of Technology. Some technical details of this facility are available in a few papers which have been published elsewhere [13,23]. The schematic diagram of the facility is presented in Fig. 1 and its main characteristics are: a riser length of 5.5 m, a riser inner diameter of 83 mm, a downcomer inner diameter of 54 mm, uidization medium pre-heater (6 kW), electrical trace heating of the whole rig (15 kW).The feeding system can supply biomass at a maximum rate of 20 kg/h. The measurement equipment consists of ow meters (Endress and Hauser AT70 thermal ow meters for all primary ows, except for steam, where an Endress and Hauser Prowirl 72 vortex ow meter is used), thermocouples (K types),

b i o m a s s a n d b i o e n e r g y 3 5 ( 2 0 1 1 ) 2 9 1 0 e2 9 2 4

2913

Fig. 1 e Schematic diagram of the 100 kWth steam-oxygen blown CFB gasier test-rig.

differential pressure meters and weighing devices. There are two high temperature lters ceramic tissue candle lter (BWF, Germany) and a SieSiC ceramic candle lter (Pall Filter systemseWerk Schumacher, Germany) connected in parallel downstream of the cyclone which can be switched during operation. To investigate the effects of different operational parameters (SBR, ER and the temperature) and bed materials on product gas composition and tar formation, several different tests have been carried out with Agrol, willow and DDGS and the ranges of experimental conditions (e.g. SBR, ER, temperature and N2 purging ow) are presented in Table 2. From Table 2, it can be seen that DDGS using Bed 1 and 2 has been tested only within a lower temperature range from 700 to 760  C, as it is very difcult to reach a higher temperature during operation because of bed agglomeration risks.

CH4, C2H4, C2H2, NH3, H2O). The schematic drawing of the gas sampling line downstream of the CFB gasier is shown in Fig. 2. To avoid the coarsest particles from penetrating the sampling line, the gas analysis probe point is located into the direction of main gas ow. During the experiment, a sample ow of the product gas is continuously extracted from the main stream downstream of the gas outlet of the cyclone. The gas sample is then led through a primary condenser (w25  C) to remove the condensables with the highest boiling point, predominately heavy tar and some water, before leading the gas via a gas pump to the gas analysis equipments. After passing through a secondary condensation unit which included a condenser surrounded by ice and then followed by two asks lled with silica gel, the dried gas is characterized with several analysis methods. The gas sampling line is heated by using a trace heating cable (Horst, type HSS-450  C). Additionally, the particle lter vessel is heated by using a heating jacket (Tyco IJ-GL glass silk heating jacket). The temperature of both the sampling line and particle lter vessel is maintained at 300  C using temperature controllers. Tar measurements have been done using the solid phase adsorption (SPA) method developed by KTH [24] and the sampling point is located immediately downstream of the lter outlet. Tar components which were quantitatively analyzed by KTH are presented in Table 3. In Table 3, tar components were divided into class 2e5 based on tar classication suggested by Kiel et al. [25]. The non-identied peaks (unknown) were quantied using an internal standard. For the analysis of the wet gas, both condensers were removed and a heated line (170  C) was connected, leading the gas to the FTIR analyzer. Water content was measured both gravimetrically and by FTIR.

2.4.

Investigating variables denition

2.3.

Product gas and tar measurement

The product gas composition was analyzed by using different analytical instruments which include HartmannBraun Uras10P NDIR (online CO2, CO), HartmannBraun Magnos6G PM (online O2), Varian CP4900 m-GC (semi-online CO, CO2, H2, CH4, benzene, toluene and xylenes (BTX)), Varian GC 450 (semionline CO, CO2, H2, CH4, BTX, H2S, COS and methyl mercaptan)

To characterize the gasication conditions different ratios are applied. The ER (oxygen to biomass stoichiometric ratio) was calculated as the ratio of oxygen supplied to the oxygen required for the complete stoichiometric combustion of the biomass on a daf (dry ash free) basis. SBR (steam to biomass mass ratio) was calculated as the ratio of steam supplied to biomass supplied on an a.r. (as received) basis. In the following equations, mi, daf is the mass fraction of i on a daf basis and Fm, i is mass ow rate of i [kg/h]. mO2, air is the mass fraction of oxygen in the air. Fm;oxygen =Fm;fueldaf Actual ER Fm;oxygen =Fm;fueldaf Stoich where

Fm;oxygen Fm;fueldaf

!
Stioch

n

mC;daf MWO2 MWC

mH;daf MWO2 4MWH

mN;daf MWO2 2MWN mO2 ;air

'   mS;daf MWO2 mO;daf MWS

and Fourier transform infrared (FTIR) spectrophotometer from ThermoElectron Nicolet 5700 (semi-online CO2, CO, COS,

Two additional important parameters are the carbon conversion efciency (CCE %) and the cold gas efciency (CGE %). CCE

2914

b i o m a s s a n d b i o e n e r g y 3 5 ( 2 0 1 1 ) 2 9 1 0 e2 9 2 4

% is dened as the ratio of carbon which is converted from the added fuel into gaseous carbon components (gas tar) to the carbon in the added fuel [26]. CCE% ! mC; gastar Fm; gastar 100% mC; fueldry Fm; fueldry

CGE% of the gasication is dened as the ratio of the sum of the energy in product gases to the energy of biomass input (biomass energy). The CGE% applied in this work is based on the lower heating value (LHV) of the product gas and is dened as [13]: ! LHVgas Fm;gas 100% CGE% LHVfueldry Fm; fueldry The higher heating value (HHV) of biomass fuels on a d.b. (dry basis) was calculated using a unied formula suggested by Channiwala and Parikh [27,28], and the lower heating value (LHV) of the fuel can be estimated by:

formation produced from Agrol, willow and DDGS are presented in the following parts. The changing trends of the product gas composition, reactor temperature and dp cell during experiment as measured versus time during a days experiment (23rd April, 2010) is presented in Figs. 3 and 4. T_Aver_Riser and T_Aver_Reactor represent the average temperature of the riser, and of riser and downcomer, respectively. From these two Figures, some clear and important points can be concluded: 1) A relatively stable temperature prole was established in the whole reactor, which can be seen from the value of T_Aver_Riser and T_Aver_Reactor. The standard deviation of T_Aver_Riser and T_Aver_Reactor during stable operation condition was within 7  C; 2) The concentrations of CO and H2 obtained from Agrol, willow and DDGS were signicantly different. No signi-

HHV 0:3491C 1:1783H 0:1005S 0:1034O 0:0151N 0:0211AMJ=kg MJ=kg LHV HHV 9mH hfg

where C, H, O, N, S and A are percentages of mass fraction carbon, hydrogen, oxygen, nitrogen, sulphur and ash in the dry fuel. mH is the mass fraction of hydrogen in the fuel and hfg is the enthalpy of vaporization of water (w2.26 MJ/kg).

3.

Results and discussion

To obtain a clean bio-syngas and to increase the net energy conversion efciency, the gasication operating conditions need to be optimized. The effects of reaction temperature, ER, SBR and bed materials on the product gas composition and tar

cant difference was observed for the concentration of CH4 for all the fuels and that of CO2 for Agrol and willow; 3) The reactor temperature remained fairly stable and similar during Agrol and willow gasication, while a sharply decreasing trend was observed by using DDGS fuel, which indicates of different gasication behavior; 4) No signicant uctuation was observed in the measured values of all dp cells during Agrol gasication. A gradually increasing trend of dp 7 and 9 values (downcomer) was found during willow gasication, and almost all dp-cells started to behave differently by changing from willow to DDGS.

Table 2 e Overview of process conditions setting for Agrol, willow and DDGS gasication. Exp. NO
Date Time start Time end Duration Oxidant Fuel Additive Bed materials Temperature ( C) Steam ow (kg/h) Biomass ow(kg/h) Oxygen ow (kg/h) L-valve ow (kg/h) Purging N2 ow(kg/h) Fluidized velocity (m/s) SBR () ER () a b c d

D1
15-10-09 12:00 14:00 2:00 Steam-O2 DDGS None Bed 4a 780e830 10.6e10.9 13.0e13.6 4.9e5.0 1.4 2.47 2.8e3.3 0.81e0.83 0.38

D2
21-04-10 12:00 15:40 3:40 DDGS Kaolin Bed 1 700e760 14.8e16.8 15.27 5.5e2.7 1.85 2.7 3.3e3.9 0.97e1.10 0.37

D3
23-04-10 16:30 18:10 1:40 DDGS Kaolin Bed 2 700e760 14.5e16.8 15.27 5.49e5.5 1.77 2.7 3.3e3.9 0.95e1.08 0.36

A1
25-03-10 16:20 18:20 2:00 Agrol None Bed 3 800e830 11.1e11.6 10.1e11.1 5.0e5.5 1.4 2.7 2.9e3.3 1.00e1.15 0.40

A2
13-04-10 12:50 17:10 4:20 Agrol None Bed 1c 700e830 14.7e15.4 10.1e10.8 4.5e5.7 1.4 2.34 3.5e4.0 1.36e1.52 0.36e0.42

A3
15-04-10 11:50 17:00 5:10 Agrol None Bed 1 700e830 13.6e14.7 10.1e11.7 4.7e5.2 1.4 2.7 2.9e3.8 0.97e1.45 0.35e0.38

A4
23-04-10 11:00 13:30 2:30 Agrol None Bed 2d 700e830 11.5e14.6 11.72 4.80e4.90 1.4 2.7 3e3.6 0.98e1.25 0.33

W1
25-03-10 12:20 16:20 4:00 Willow Kaolin Bed 3b 700e830 11.6e15.9 11.1 5.0e6.0 1.4 2.7 2.9e3.9 1.00e1.15 0.40e0.47

W2
19-04-10 10:50 17:10 6:20 Willow Kaolin Bed 1 700e830 10.9e14.8 11.7e12.7 5e5.6 1.4 2.7 2.9e3.8 0.90e1.20 0.38e0.39

W3
23-04-10 13:50 16:30 2:40 Willow Kaolin Bed 2 700e830 11.5e14.5 12.71 4.95e5.0 1.4 2.7 3e3.6 0.9e1.14 0.34

Untreated, used Scandinavian olivine. Mixed sand and used treated Austrian olivine. Used, treated Austrian olivine. Fresh Austrian olivine.

b i o m a s s a n d b i o e n e r g y 3 5 ( 2 0 1 1 ) 2 9 1 0 e2 9 2 4

2915

Fig. 2 e Schematic drawing of the dry gas sampling line downstream of the 100 kWth CFB gasier.

3.1. Effects of different operational parameters on product gas 3.1.1. Effect on product gas composition

concentration of H2 increased with increasing the temperature because higher temperature favored endothermic char gasication reactions and steam reforming and cracking of light hydrocarbons and tars. The different behavior of various gaseous species at lower temperatures could be due to the variation in reactivities of chars produced during the pyrolysis step [32] and at higher temperatures could be due to the intensied effect of steam on the decomposition of higher molecular mass components [33]. Gil et al. [34] reported that as ER or SBR was increased the concentrations of CO, CH4 and C2hydrocarbons decreased due to partial oxidation and steam reforming reactions. Wang and Kinoshita [35] reported that by varying SBR from 0.4 to 1.0, the concentration of H2 increased and that of CO2 remained roughly the same, while the concentration of CO, CH4 and other light hydrocarbons produced decrease slightly.

3.1.2.
Effects of temperature, SBR, ER and bed materials on the composition of product gas obtained from Agrol, willow and DDGS is shown in Figs. 5e7, respectively. The concentration of N2 for Agrol bed 1 and willow-bed 1 at ca.780  C was 25 vol.% and 21 vol.% on a dry basis, respectively. From Figs. 5e7 it can be seen that there is a signicant variation among the product gas composition produced from different fuels as a function of temperature. Generally high temperatures favored the formation of H2 from all fuels. The concentration of H2 obtained from willow was higher than that from Agrol, while CO, CH4, C2H2 and C2H4 had an opposite trend. The formations of CO, CH4, C2H2 and C2H4 behaved differently over different temperature ranges. The averaged H2 to CO mole ratio obtained from Agrol, willow and DDGS over the temperature range from 800 to 820  C was around 1.0, 1.4 and 0.7, respectively. With increasing SBR, the concentrations of CO, CH4, C2H4 and C2H2 all gradually decreased, while the concentration of H2 increased with a varying degree depending on fuel types.The concentration of CO2 obtained from willow was the highest, followed by that from Agrol and DDGS, but no signicant change was observed with a variation of SBR. However, when SBR was increased from 0.98 to 1.16 at a temperature of 800  C, the concentrations of CO, CH4 and C2H2 obtained from Agrol Bed 2 all slightly increased. Higher ER values increased the concentration of CO2 in the product gas but decreased the concentrations of CO, CH4 and H2. The effect of different bed materials on the product gas composition can hardly be determined due to the difference of SBR, ER and the temperature. At a similar operational conditions (e.g. 820  C, SBRw1.15 and ERw0.4), the concentration of H2 obtained from willow using Bed 1 (w30 vol.%, dnf.) was higher than using Bed 3 (w26 vol.%, dnf.) indicating that olivine can improve H2 production. Without feeding any kaolin into the gasier, a severe agglomeration was observed during DDGS gasication using Bed 4. Most experimental results agreed well with those reported by other researchers.Turn et al. [29] reported that higher hydrocarbon (C2H2, C2H4, C2H6) concentrations decreased as reactor temperature increased but no noticeable conversion of C2H2 and C2H4 was observed at lower reactor temperatures from 750 to 800  C. Gil et al. [5], Kumar et al. [30] and Franco et al. [31] all reported that the

Effect on sulphur species

Because Agrol and willow both contain less than 0.002% sulphur, almost no sulphur species were detected during experiments. However, a high content of H2S and well measurable amounts of COS and methyl mercaptan were determined during DDGS gasication. With increasing temperature, the concentration of H2S remained fairly stable, whereas the concentrations of COS and methyl mercaptan slightly decreased. Since steam-oxygen blown CFB gasication of DDGS fuel no reported results are available in the literature, here a simple thermodynamic equilibrium model established using Factsage software has been used to predict the concentration of H2S, COS and methyl mercaptan. A detailed description about the modeling using Factsage can be found in reference [36]. The predicted concentration of H2S was slightly higher than the measured value, while COS and methyl mercaptan concentrations were lower. The measured average concentrations for H2S, COS and methyl mercaptan are around 2300, 200 and 0.8 ppmv, respectively, while the predicted ones are 2600, 40 and 0 ppmv, respectively. Thus during experiments, less H2S will be produced than predicted by chemical equilibrium calculations, which probably resulted in higher concentrations of COS, methyl mercaptan and other sulphur species. Attar and Dupuil [37] reported that part of the sulphur is retained in the solid as alkali sulphide due to the reaction of H2S with the alkali minerals which prevents

Table 3 e Tar components quantitatively analyzed from SPA. Class

Class 2 Class 3 Class 4

Name
Heterocyclic aromatics Aromatics (1 ring) Light PAHs (2, 3 ring)

Tar components
phenol, o-cresol, m-cresol and p-cresol benzene, toluene, m/p-xylene, o-xylene indan, indene, naphthalene, 2-methylnaphthalene, 1-methylnaphthalene, biphenyl, acenaphthylene, acenaphthene, uorene, phenanthrene, anthracene Fluoranthene, pyrene

Class 5

Heavy PAHs (>3-ring)

2916

b i o m a s s a n d b i o e n e r g y 3 5 ( 2 0 1 1 ) 2 9 1 0 e2 9 2 4

Fig. 3 e Gas composition and reactor temperature trend lines as measured during the experiment on 23 April 2010.

them from entering the gaseous phase. The mineral elements such as Fe, Ni, Si, Al, Fe, Na, K, Ti, Mg, Ca present in the fuel may act as catalysts and affect the concentrations of sulphur species during gasication. The presence of high levels of Ca, K and Na in the biomass could lead to a high retention of S and Cl in the ash [36]. Therefore, the lower measured H2S could be attributed to the retention of sulphur in the ash because of the presence of bed materials and minerals in DDGS fuel. However, due to the low sulphur content in biomass fuel and limited measurement capacity and accuracy, equilibrium modeling is already a valuable tool to get an insight into the

behavior of sulphur species during biomass gasication. With a variation of SBR, the concentration of H2S and COS obtained from DDGS using Bed 1 and Bed 2 showed a similar change trend. For instance, an increase of SBR from around 1.0 to 1.1 led to a signicantly decrease in the concentrations of H2S obtained from DDGS using Bed 1 from 2700 to 1800 ppmv accompanied with an slightly increase in COS concentration from 116 to 125 ppmv. Predicted results from the Factsage equilibrium model for DDGS gasication using Bed 1 showed that when SBR was increased from 0.98 to 1.09, a slight decrease in the concentrations of H2S and COS from 2640 to

Fig. 4 e Several dp-cell trend lines as measured during the experiment on 23 April 2010.

b i o m a s s a n d b i o e n e r g y 3 5 ( 2 0 1 1 ) 2 9 1 0 e2 9 2 4

2917

H2
45 Agrol-Bed1 40

CO

CO2

CH4

C2H2+C2H4
10 Willow-Bed1

Operation conditions (-)

Other gas composition (vol.dnf.%)

35 30

6 25 20 4 15 10 5 0 0 2

=1 .4 5

=1 .1 3

=0 .9 7

=1 .2 5

=0 .9 3

=1 .2 2

=0 .9

=0 .3 8, SB

=0 .3 8, SB

=0 .3 5, SB

=0 .3 5, SB

=0 .3 8, SB

=0 .3 8, SB

=0 .3 9, SB

,E

,E

,E

,E

,E

,E

,E

77

77

81

81

78

78

82

Fig. 5 e Effects of different operational conditions on the composition of product gas from Agrol and willow using bed 1.

2600 ppmv and 28 to 25 ppmv was observed which partly agree with experimental results. A similar predicted result has been achieved for DDGS gasication using Bed 2. A higher ER led to a decrease in H2S concentration but an increase in the concentration of COS. These experimental results only partly agreed with predicted results from Factsage equilibrium model. However, these predicted trends agree with experimental and predicted values reported by Dias and Gulyurtlu [38]. The authors reported that the concentration of H2S increased from 672 ppmv up to 1204 ppmv with increasing ER from 0 to 0.4, and the concentration of sulphur oxidizes such as SO2 and COS also increased in different proportions.

3.1.3. gas

Effect on CCE%, CGE% and heating values of product

The operational conditions also largely inuence CCE%, CGE% and LHV of the product gas. Generally a higher temperature and a higher SBR value led to a higher gas yield. The effects of operational parameters on CCE%, CGE% and heating values of product gas obtained from some Agrol, willow and DDGS testing are given in Table 4. From Table 4 it can be seen that an increase in SBR led to a higher yield of the product gas, but simultaneously also caused a lower CCE%, CGE% and LHV of the product gas. However, higher temperature largely improved CCE%, CGE% and LHV of the product gas. These

H2 50 Agrol-Bed1 45

CO

CO2

CH4

C2H2+C2H4 10 Willow-Bed2 Willow-Bed3

Operation conditions (-) Agrol-Bed3

Other gas composition (vol.dnf.%)

40 35 30 25 20 15 10 5 0

82

,E

=0 .3 9, SB

=1 .0 4

=1 .1 5

.9 1

.1 4

R= 1. 16

.9 8

R= 0

R= 1

R= 0

=0 .4 ,S B

,S B

,S B

,S B

,S B

=0 .3 4

.3 3

.3 3

=0

=0

=0

,E R

,E R

,E R

,E R

,E R

80 0 C

80 0 C

80 0 C

Fig. 6 e Effects of different operational conditions on the composition of product gas from Agrol and willow using bed 2 and 3.

80 0 C

80 0 C

81 5 C

,E R

=0

.4 3

.3 4

,S B

R= 1

.0 7

2918

b i o m a s s a n d b i o e n e r g y 3 5 ( 2 0 1 1 ) 2 9 1 0 e2 9 2 4

H2 45

CO

CO2

CH4

COS (ppmv)

H2S(ppmv) /10

CH3HS (ppmv)100 250

40 200

Other Gas Concentration (vol.dnf.%)

35

30 150 25

20 100 15

10

50

0 790C,ER=0.38,SBR=0.81 800C,ER=0.38,SBR=0.81 810C,ER=0.38,SBR=0.82 820C,ER=0.38,SBR=0.81

Fig. 7 e Effects of different operational conditions on the composition of product gas from DDGS using bed 4.

results are in agreement with those reported by other researchers. Gil et al. [5] reported that by increasing GR (the ratio of steam and oxygen supplied to biomass fuel supplied on a.r. basis) or decreasing SOR (the mole ratio of steam supplied to oxygen supplied), the LHV of product gas decreased. The gas yield on a dry basis increased with an increment of GR and bed temperature. A higher temperature in the gasier bed or lower GR led to higher thermal efciencies. Wei et al. [7] reported that the gas yield increased with reactor temperature due to enhanced endothermic steam reforming, cracking reactions of the tar and char gasication at elevated temperatures. Turn et al. [29] reported that increasing the temperature of the gasifying agents led to an increase in the heating value of the product gas and reduced the tars, soot and char residues. Kinoshita et al. [39] reported that the product gas yield and CCE% increase as temperature increases since higher temperatures facilitate tar conversion. Boateng et al. [40] reported that the yield and LHV of the product gas, CCE% and CGE% increased with increasing gasication temperature from 700 to 800  C. From Table 4, it can also be seen that CCE%, CGE%, the yield and heating values of the product gas obtained from willow were slightly higher than those obtained from Agrol. For instance, under similar operational conditions (e.g. SBR 1.2, ER 0.38, 780  C, Bed 1),

the yield of product gas produced from willow and Agrol was 2.63 and 2.48 Nm3/kg fuel on a daf. basis, respectively. Furthermore, it seemed that the product gas yield produced from DDGS was much higher than those from Agrol and willow. For different bed materials, it seemed Bed 2 was better than Bed 1 on the enhancement of the product gas yield. That could be due to some accumulation of ash in the V-cone after several time measurements leading to a slight higher measured pressure drop, since the experiments using Bed 2 was carried out at the last day.

3.2. Effects of different operational parameters on tar formation/reduction 3.2.1. Effect of temperature on tar formation

A summary of process parameters (SBR, ER, temperature and bed materials) for all SPA samples from Agrol, willow and DDGS is presented in Table 5. The analyzed tar SPA results are shown in Figs. 5e8, where Total Tar/2, Total Unknown, Sum > C10H8 and Sum > C10H8UN represent a half of the total tar content, the total unknown tar content, the total tar content heavier than naphthalene and the total unknown tar content heavier than naphthalene, respectively.

Table 4 e CCE%, CGE% and HHV, LHV of the product gas produced from some Agrol, willow and DDGS gasication tests. Fuel ()
Agrol Agrol Willow Willow Willow Willow DDGS

Bed types ()
Bed 1 Bed 1 Bed 1 Bed 1 Bed 2 Bed 3 Bed 2

T(C) ( C)
770 770 780 780 800 820 750

ER ()
0.38 0.38 0.38 0.38 0.34 0.43 0.36

SBR ()
1.21 1.45 1.13 1.22 1.14 1.04 1.08

CCE% (%)
84.2 74.0 90.5 86.9 91.8 93.5 96.6

CGE% (%)
52.5 47.0 56.2 55.2 62.1 55.7 71.5

HHV_gas MJ/kg
3.9 2.9 4.2 4.0 4.3 4.0 4.6

LHV_gas MJ/kg
3.6 2.7 3.8 3.6 4.0 3.7 4.3

Product gas/Biomass (Nm3/kg daf.)

2.48 3.08 2.55 2.63 2.81 2.50 2.82

H2S,COS,CH3HS Concentration (ppmv,dnf.)

b i o m a s s a n d b i o e n e r g y 3 5 ( 2 0 1 1 ) 2 9 1 0 e2 9 2 4

2919

Table 5 e Process parameters setting for all SPA samples from Agrol, willow and DDGS gasication.
SPA Sample Fuel SBR () ER() Temperature( C) Bed materials 0413A Agrol 1.52 0.36 730 Bed 1 0419A Fuel SBR () ER() Temperature( C) Bed materials Willow 1.19 0.38 740 Bed 1 0325 A Fuel SBR () ER () Temperature ( C) Bed materials Agrol 1.15 0.40 820 Bed 3 0413B Agrol 1.42 0.33 730 Bed 1 0419B Willow 0.99 0.38 740 Bed 1 0423A Agrol 1.25 0.33 770 Bed 2 0413C Agrol 1.35 0.42 780 Bed 1 0419C Willow 1.27 0.38 740 Bed 1 0423B Agrol 0.98 0.33 800 Bed 2 0413D Agrol 1.35 0.42 820 Bed 1 0419D Willow 0.93 0.38 780 Bed 1 0423C Agrol 1.15 0.33 800 Bed 2 0415A Agrol 1.45 0.38 770 Bed 1 0419E Willow 1.13 0.38 780 Bed 1 0423D Willow 0.90 0.34 800 Bed 2 0415B Agrol 1.21 0.38 770 Bed 1 0419F Willow 1.22 0.38 780 Bed 1 0423E Willow 1.14 0.34 800 Bed 2 0415C Agrol 1.13 0.38 775 Bed 1 0419G Willow 0.90 0.39 820 Bed 1 0421A DDGS 1.10 0.37 730 Bed 1 0415D Agrol 0.97 0.35 815 Bed 1 0419H Willow 1.04 0.39 820 Bed 1 0421B DDGS 0.98 0.37 740 Bed 1 0415E Agrol 1.16 0.35 810 Bed 1 0419I Willow 1.14 0.39 820 Bed 1 0423F DDGS 0.95 0.36 750 Bed 2 0415F Agrol 1.25 0.35 810 Bed 1 0325W Willow 1.04 0.43 820 Bed 3 0423G DDGS 1.08 0.36 750 Bed 2

From Fig. 8 (Agrol gasication using Bed 1), we can see that with increasing temperature the total tar content produced decreased. The total amounts of class 2 and 4 tars were much higher than those of class 3 and 5 tars. There observations agreed well with the results reported by other researchers. Gil et al. [34] and Kinoshita et al. [39] all reported that phenol and cresol were predominant only at temperature below 800  C, while naphthalene and indene were the major components at 900  C. Narvaez et al. [6] reported that the temperature not only inuenced tar formation but also the tar properties by affecting the chemical reactions involved in the whole gasication network. Brage et al. [41] reported that with increasing
Class 2 7.0 Class 4 Total Unknown Increase T Sum>C10H8

temperature the differences in molecular thermal stability lead to a progressive accumulation of non-oxygenated aromatics and favor the formation C2H2 and C2H4 at the expense of phenols. Van Paasen and Kiel [42] reported that tar concentration decreased with temperature varying from 750 to 950  C, and simultaneously tar compositions shifted from alkyl-substituted poly-aromatic hydrocarbons (PAHs) to nonsubstituted PAHs. Furthermore, from samples 0413C and D, it seemed that a higher temperature favored the decomposition of the heavier tar fraction class 5. When the temperature was increased from 780 to 820  C, more than 40% of class 5 tar and 60% of Sum > C10H8UN was reduced, and 26% increase in the
Sum>C10H8UN Class 3 Class5 Total Tar/2 400

Decrease SBR&ER

Decrease SBR (780 C)

Increase SBR (810 C)

6.0

350

300 5.0 250 4.0 200 3.0 150 2.0 100 1.0

50

0.0 0413A 0413B 0413C 0413D 0415A 0415B 0415C 0415D 0415E 0415F Sample Names(-)

Fig. 8 e The analyzed SPA results of tar samples from Agrol gasication using bed 1.

Class 3,5 tar (mg/m3)

Tar content(g/m3)

2920

b i o m a s s a n d b i o e n e r g y 3 5 ( 2 0 1 1 ) 2 9 1 0 e2 9 2 4

fraction of class 3 tar was also observed. These observations indicate that a higher temperature converts higher moleculeweight compounds into smaller molecules. These results seem to partly agree with those reported by Han and Kim [43]. They reported that an increase in the temperature had a positive effect on the decomposition of class 1 and 2 tars, while the concentrations of class 3 and 5 tars increased with temperature enhancement. Kiel et al. [25] reported that by varying the temperature no clear change trend of the class 5 tar was found. The concentration of class 4 tar increased with increasing gasication temperature at ER 0.4, which was largely due to the simultaneous occurrence of many effects such as polymerization reactions. Class 4 tar comprises of a mixture of alkyl-substituted tars and polyaromatic hydrocarbons which shown a different behavior. Without considering the effects of other operational parameters, a higher temperature led to higher proportions of class 4 and 5 tars. With increasing temperature from 770 to 810  C (sample 0415A to F), the proportional contributions of class 4 and 5 tars increased from 40 to 55% on the average. For Agrol gasication using Bed 1, the lowest tar content of 5.7 g/Nm3 and highest one of 12.4 g/Nm3 were measured at the temperature of 730  C with SBR 1.52 and ER 0.36, and at 780  C with SB 1.35 and ER 0.42, respectively. These results were surprising because normally at a higher temperature the tar content should be lower. However, as it is well-known, the total tar content produced from biomass gasication does not depend on only the temperature, but also other parameters such as SBR, ER etc. Somehow lower tar content can be achieved at lower temperature with a suitable combination of other critical operational parameters. Rapagna et al. [44] also observed a much lower tar concentration of around 2.4 g/Nm3 at the exit of the gasier at a temperature of 770  C using olivine as bed material. Li et al. [45] reported that the tar yield

from biomass gasication decreased when the temperature increased from 700 to 815  C. Similar trends were also observed during willow gasication using Bed 1. From Fig. 9 (willow gasication using Bed 1), it can be seen that the total tar content produced from willow gasication decreased with increasing temperature from 740 to 820  C. Due to different fuel properties (e.g. ash content), the total tar content produced from willow was signicantly lower than that from Agrol. The lowest and highest tar contents produced from willow were 4.4 and 6.7 g/Nm3, which were measured at a temperature of 820  C with SBR 1.14 and ER 0.39 (sample 0419I), and at 740  C with SBR 1.19 and ER 0.38 (sample 0419A), respectively. During Agrol and willow gasication, a higher temperature led to a higher concentration of class 5 tar components. Regarding this point, Van Paasen and Kiel [43] pointed out that two opposite mechanisms can determine the production of class 5 tar with increasing temperature. Class 5 tar compounds could be produced either from the decomposition of heavy large class 1 tar compounds, or from lighter tar compounds due to PAH growth reactions. The inuence of temperature on tar formation produced from DDGS was difcult to be determined from Fig. 10. However, the total tar content produced from DDGS gasication using Bed 1 was 7.3 g/Nm3 (sample 0421A) which was comparatively lower, because at the similar operational conditions the total tar content produced from Agrol and willow gasication was around 10.2 (sample 0415C) and 6.7 g/m3 (sample 0419A). Furthermore, at a fairly low temperature of 730  C, the contents of class 2 and 4 tars were only 1 and 1.85 g/ Nm3, respectively, which is below 2 g/Nm3 being considered as an important limit for many downstream applications [46].

3.2.2.

Effect of SBR and ER on tar formation

From Fig. 8, it can be clearly seen that an increment of SBR signicantly promoted tar decomposition during Agrol

Class 2 4.0

Class 4

Total Unknown

Sum>C10H8

Sum>C10H8UN

Class 3

Class5

Total Tar/2 450 400 350 300

Increase SBR (740 C)

Increase SBR (780 C)

Increase SBR (820 C)

3.0

250 2.0 200 150 1.0 100 50 0.0 0419B 0419A 0419C 0419D 0419E Sample Names(-) 0419F 0419G 0419H 0419I 0

Fig. 9 e The analyzed SPA results of tar samples from willow gasication using bed 1.

Class 3,5 tar (mg/m3)

Tar content(g/m3)

b i o m a s s a n d b i o e n e r g y 3 5 ( 2 0 1 1 ) 2 9 1 0 e2 9 2 4

2921

Class 2 Sum>C10H8UN

Class 4 Class 3

Total Unknown Class5 Increase SBR (bed 2)

Sum>C10H8 Total Tar/2

5.0

Decrease SBR (bed 1)

300

4.0

250

Tar content(g/m3)

200 3.0 150 2.0 100 1.0

50

0.0 0421A 0421B Sample Names(-) 0423F 0423G

Fig. 10 e The analyzed SPA results of tar samples from DDGS gasication using bed 1 and 2.

gasication using Bed 1. With increasing SBR from 1.13 to 1.45 at 770  C (sample 0415C to A) and SBR from 0.97 to 1.25 at 810  C (sample 0415D to F), the total tar content both decreased from around 10.2 to 7.2 g/Nm3. Generally with increasing SBR the concentrations of class 3, 4 and 5 tars all decreased, while the concentration of class 2 increased. However, at 770  C, the lowest content of class 5 tar was found at an SBR value of 1.21 instead of 1.45. Furthermore, a slight increase in the concentrations of class 3 and 5 tars was also observed when increasing SBR from 0.97 to 1.16 at 810  C (sample 0415D to E). The highest tar content obtained from Agrol was around 12.4 g/Nm3 (sample 0413C) which was measured at 780  C and SBR 1.35 and ER 0.42. However, under a less favorable tar decomposition circumstance, a much lower tar content of 8.7 g/Nm3 (sample 0415B) was also observed. These two samples were taken at different days and there may be slight difference inside the gasier (e.g. char content). Similar trends were observed for willow and DDGS gasication using Bed 1. In Fig. 9, with increasing SBR from 0.93 to 1.22 at 780  C (sample 0419D to F), the total tar content produced from willow decreased from 6.6 to 4.7 g/Nm3. With increasing SBR from 780 to 820  C (samples 0419D to I), the concentrations of class 2 to 5 tars decreased with a varying reduction degree from 20 to 66%, but at 740  C (samples 0419B to C) the concentration of class 2 tar increased with a variation of SBR from 0.99 to 1.27. Furthermore, at 820  C, the lowest concentrations of class 3 and 5 tars were observed at SBR of 1.04 (sample 0419H) instead of at 1.14 (sample 0419I). In Fig. 10, at a temperature of 730  C when SBR was increased from 0.98 to 1.1 (sample 0421B to A), the total tar content produced from DDGS decreased from 9.1 to 7.3 g/Nm3. More than 45% of class 4 and 60% of class 5 tar were reduced, but the concentration of class 2 largely increased from 0.6 to 1.0 g/Nm3 and almost no change in the content of class 3 tar. Similar results were observed for tar compounds produced from Agrol, willow and DDGS gasication using Bed 2 (see in Figs. 10 and 11). However, a fairly dramatic change was observed about the class 2 produced from Agrol and willow.

With increasing SBR from around 0.9 to 1.15 (sample 0423B to G), the total tar content and the concentrations of class 4 and 5 tars produced from three fuels all decreased. The lowest tar concentrations obtained from Agrol, willow and DDGS were 9.7, 5.3 and 6.7 g/Nm3, respectively (sample 0423 A, E and G). Furthermore, it seemed that the increment of SBR from 0.98 to 1.15 at a temperature of 800  C had negligible inuence on the decomposition of tar produced from Agrol. The total tar content remained around 10 g/Nm3, and similar results were also observed when using Bed 1. The effect of ER on tar formation can not be concluded from the SPA results. From above-stated results, we can see that the compositions and content of tar produced from three fuels were fairly different, which can be attributed to the difference among their properties. The results reported in this work are similar to those reported by other researchers. Aznar et al. [47] reported that with varying GR from 0.7 to 1.2 more than 85% reduction in total tar was achieved. The different behavior of class 2 to 5 tars produced from Agrol, willow and DDGS can be attributed to their different properties such as the distribution of cellulose, hemicellulose and lignin and ash contents. According to Paasen and Kiel [43], the nal tar composition depends on the lignocelluloses composition of biomass fuels. The difference in the concentration of class 2 tar can be largely due to the difference in structure between cellulose and lignin, while the difference in the concentrations of class 4 and 5 tars might be attributed to differences in lignocelluloses composition of the feedstock. Since class 4 and 5 tars are mostly PAH compounds which come from aromatic functional group in the molecular structure of lignin.

3.2.3.

Effect of the types of bed materials on tar formation

From Figs. 8e11, it is difcult to conclude whether Bed 1 or Bed 2 was better on tar decomposition. For instance, the total tar content obtained from willow using Bed 2 was 5.3 g/Nm3 at 800  C with SBR 1.14 and ER 0.34 (sample 0423E), which was slightly lower than that of 5.4 g/Nm3 using Bed 1 at 780  C with

Class 3, 5 tar (mg/m3)

2922

b i o m a s s a n d b i o e n e r g y 3 5 ( 2 0 1 1 ) 2 9 1 0 e2 9 2 4

Class 2

Class 4

Total Unknown

Sum>C10H8

Sum>C10H8UN

Class 3

Class5

Total Tar/2

7.0

Bed3 (Agrol & willow)

Higher SBR

Increase SBR (Agrol) Bed2

Increase SBR (willow)

450 400 350

6.0

5.0

4.0

250 200 150

3.0

2.0 100 1.0 50 0 0325 A 0325W 0423A 0423B Sample Names(-) 0423C 0423D 0423E

0.0

Fig. 11 e The analyzed SPA results of tar samples from Agrol and willow gasication using bed 2 and 3.

SBR 1.13 and ER 0.38 (sample 0419E), while a bit higher than that of 4.4g/Nm3 at 820  C with SBR 1.14 and ER 0.39 (sample 0419I). Furthermore, from these three tar SPA samples, it can also be seen that the concentration of class 5 tar using Bed 2 was much higher than that using Bed 1, while the class 3 tar showed an opposite trend. The concentrations of class 3 and 5 tars in samples 0423E, 0419E and 0419I were 72, 64 and 65 mg/ Nm3, and 150, 219 and157 mg/Nm3, respectively. From these results, it can be concluded that Bed 1 showed considerable catalytic reactivity toward the decomposition of heavy PAH compounds (class 5 tar). Some fairly interesting observations were found during Agrol gasication using different bed materials. From SPA samples 0415D and E and samples 0423B and C, it can be seen that when the temperature was lower than 820  C and ER lower than 0.35 the increment of SBR from 0.9 to 1.16 had negligible inuence on tar decomposition and compositions. Under these conditions, the averaged total tar content was around 10 g/Nm3. A combination of higher SBR (>1.4) with lower temperature (w730  C) and intermediate ER (0.33e0.36) could lead to a lower tar content produced from Agrol (samples 0413A to B). A fairly low tar content of 7.7 g/Nm3 was also obtained from Agrol using Bed 3 under a combination of operational conditions of a lower SBR (w1.15) with a higher temperature (>820  C) and a higher ER (w0.4). The effects of Bed 1 and 2 on the compositions of tar can be seen. These two similar olivine bed materials(Bed 1 and 2) have also been studied by Pecho et al. [48] for the optimization of the biomass gasication under sulphur-free (S-free) and H2S enriched conditions. They reported that the catalytic activity of Bed 1 increased signicantly by redox-type pre-treatment and it also acted as an oxygen carrier from combustion zone to the gasication zone, whereas fresh olivine Bed 2 had almost no catalytic activity. Their ndings can be a good reference for this

work. Rauch et al. [49] compared different olivines for biomass steam gasication and found that olivine had high attrition resistance as bed material for uidized beds and the catalytic activity for tar reforming. Bed 3 had less positive inuence on tar decomposition than Bed 1. For instance, at the temperature of 820  C with SBR of 1.04 and ER of 0.4, the total tar content produced from willow gasication using Bed 1 and 3 was 4.5 and 6.9 g/Nm3, respectively. The total amount of class 4 and 5 tars was reduced from 4.6 to 3.1 g/Nm3 and 153 to 62 mg/Nm3, respectively. Regarding to the reactivity of different bed materials on tar decomposition, more experiments need to be carried out in order to get a better insight into their inuence.

4.

Conclusion

The effects of operational conditions (SBR, ER and temperature) and bed materials on the product gas distribution and tar formation produced from Agrol, willow and DDGS gasication have been investigated on an atmospheric pressure 100 kWth steam-oxygen blown CFB gasier. The compositions of the product gas and tar obtained from these fuels are fairly different, but it is difcult to make a clear comparison between tem due to some differences in SBR, ER and the temperature used. Under similar operational conditions the concentration of H2 obtained from willow was much higher than that obtained from Agrol and DDGS. Among all the experiments, the product gas composition obtained from willow over the temperature range from 800 to 820  C consisted of the highest H2 concentration (28 vol.%), followed by Agrol (24 vol.%) and DDGS (20 vol.%) on a N2 free basis, respectively. A fairly high amount of H2S (w2300 ppmv), COS (w200 ppmv) and trace amounts of methyl mercaptan (<3 ppmv) on a N2 free basis

Class 3,5 tar (mg/m3)

Tar content(g/m3)

300

b i o m a s s a n d b i o e n e r g y 3 5 ( 2 0 1 1 ) 2 9 1 0 e2 9 2 4

2923

were also obtained from DDGS. A higher temperatures and SBR were more favorable for H2 production but less advantageous for the formation of CO and CH4, whereas a higher SBR also led to a lower CCE%, CGE% and heating values of the product gas due to high steam content in product gas. Although DDGS fuel has a relatively high K and Cl content, continually adding 3e10% kaolin (based on total feeding rate) into the reactor can successfully avoid agglomeration. Tar produced from the three fuels mainly contains phenol, cresol, naphthalene, indene and pyrene. The total tar content obtained from Agrol is at maximum w12.4 g/Nm3, followed by that from DDGS and willow. A higher temperature and higher SBR were favorable for the tar decomposition. However, it seems that an increment of SBR from 0.9 to 1.16 had negligible inuence on tar obtained from Agrol fuel at a lower temperature (<820  C) and a lower ER (<0.35). The content of class 5 tar obtained using Bed 1 was lower than that obtained using Bed 2, which prove Bed 1 was more reactive on the decomposition of heavy tar compounds. Although at a fairly low temperature of 730  C, the total tar content produced from DDGS using Bed 1 was 7.3 g/Nm3 where the contents of class 2 and 4 tars were only 1 and 1.85 g/Nm3, respectively, which is near 2 g/Nm3 being considered as an important limit for many downstream applications. However, DDGS fuel has a high sulphur content which also leads to a high concentration of H2S in the product gas. So far, the effects of SBR, ER and temperature on the distribution of H2S and COS from DDGS are still not entirely clear. Additional experiments are necessary in order to get a better insight in the reactivity and inuence of different bed materials on tar decomposition in combination with reduction of other contaminants e.g. H2S.

the ratio of oxygen supplied to the oxygen required for the complete stoichiometric combustion of fuel on a daf basis FB uidized bed FTIR fourier transform infrared FICFB fast internally circulating uidized bed GC gas chromatograph GHG greenhouse gas emissions GR the ratio of steam and oxygen supplied to biomass fuel supplied on a.r. basis HHV higher heating value [MJ/kg] LHV lower heating value [MJ/kg] NDIR non dispersive infrared analyzer PM paramagnetic analyzer PAHs polycyclic aromatic hydrocarbons SPA solid phase absorption SBR the ratio of steam supplied to biomass fuel supplied on a.r. basis SOR the mole ratio of steam supplied to oxygen supplied TGA thermogravimetric analyzer Fm, oxygen oxygen supplied mass ow rate [kg/h] Fm, fuel (daf) biomass fuel supplied mass ow rate on a daf basis [kg/h] Fm, gas gas mass ow rate [kg/h] the enthalpy of vaporization of water [MJ/kg] hfg mass fraction of different element (C, H, O, N, S) in mi.daf fuel on a daf basis mO2, air mass fraction of O2 in air Molar weight of different element (C, H, O, N, S) MWi

ER

references

Acknowledgments
The authors thank Lantmannen, Sweden for supplying Agrol, willow and DDGS fuels, Markus Koch from Biomasse Kraftwerk Gussing GmbH & CoKG for delivering olivine bed mate rials, Michael Muller from Institute of Energy Research, Forschungszentrum Julich GmbH for performing fuels char acterization analysis, Claes Brage from KTH for analyzing all SPA tar samples. The European Commission is acknowledged for co-nancing the 7th Framework Project, related to this research: GreenSyngas (Project NO. 213628).

Nomenclature

a.r. BTX CCE% CGE% CFB CHP DDGS DSC daf dnf d.b.

as received benzene, toluene, xylenes carbon conversion efciency cold gas efciency circulating uidized bed combined heat and power dry distillers grains with solubles differential scanning calorimetry dry ash free dry nitrogen free dry basis

[1] Faaij APC. Bio-energy in Europe: changing technology choices. Energy Pol 2006;34:322e42. [2] Caputo AC, Palumbo M, Pelagagge PM, Scacchia F. Economics of biomass energy utilization in combustion and gasication plants: effects of logistic variables. Biomass Bioenergy 2005; 28(1):35e51. [3] Bin Zainal Alauddin ZA, Lahijani P, Mohammadi M, Mohamed AR. Gasication of lignocellulosic biomass in uidized beds for renewable energy development: a review. Renew Sustain Energy Rev 2010;14(9):2852e62. [4] Warnecke R. Comparison of xed bed and uidized bed gasier. Biomass Bioenergy 2000;18(6):489e97. [5] Gil J, Aznar MP, Caballero MA, Frances E, Corella J. Biomass gasication in uidized bed at pilot scale with steamoxygen mixtures. Product distribution for very different operating conditions. Energy Fuels 1997;11(6):1109e18. [6] Narvaez I, Oro A, Aznar MP, Corella J. Biomass gasication with air in an atmospheric bubbling uidized bed. Effect of six operational variables on the quality of the produced raw gas. Ind Eng Chem Res 1996;35(7):2110e20. [7] Wei L, Xu S, Zhang L, Liu C, Zhu H, Liu S. Steam gasication of biomass for hydrogen-rich gas in a free-fall reactor. Int J Hydrogen Energy 2007;32(1):24e31. [8] Umeki K, Yamamoto K, Namioka T, Yoshikawa K. High temperature steam-only gasication of woody biomass. Appl Energy 2010;87(3):791e8. [9] Janqswang W, Klimanek A, Gupta AK. Enhanced yield of hydrogen from wastes using high temperature steam gasication. J Energy Resour Tech Trans ASME 2006;128: 179e85.

2924

b i o m a s s a n d b i o e n e r g y 3 5 ( 2 0 1 1 ) 2 9 1 0 e2 9 2 4

[10] Lv PM, Chang J, Xiong ZH, Huang HT, Wu CZ, Chen Y, et al. Biomass airsteam gasication in a uidized bed to produce hydrogen-rich gas. Energy Fuels 2003;17(3):677e82. [11] Lv PM, Yuan Z, Ma L, Wu C, Chen Y, Zhu J. Hydrogen-rich gas production from biomass air and oxygen/steam gasication in a downdraft gasier. Renew Energy 2007;32(13):2173e85. [12] Gordillo G, Annamalai K. Adiabatic xed bed gasication of dairy biomass with air and steam. Fuel 2010;89(2):384e91. [13] Siedlecki M, Nieuwstraten R, Simeone E, de Jong W, Verkooijen AHM. Effect of magnesite as bed material in a 100 kWth steam-oxygen blown circulating uidized-bed biomass gasier on gas composition and tar formation. Energy Fuels 2009;23(11):5643e54. [14] Zhou J, Chen Q, Zhao H, Cao X, Mei Q, Luo Z, et al. Biomassoxygen gasication in a high-temperature entrained-ow gasier. Biotechnol Adv 2009;27:606e11. [15] Zhao YJ, Sun SZ, Tian HM, Qian J, Su FM, Ling F. Characteristics of rice husk gasication in an entrained ow reactor. Bioresour Technol 2009;100:6040e4. [16] Higman C, van der Burgt M. Gasication. USA: Gulf Professional Publishing, ISBN 0-7506-7707-4; 2003. [17] Devi L, Ptasinski KJ, Janssen FJJG. A review of the primary measures for tar elimination in biomass gasication processes. Biomass Bioenergy 2003;24(2):125e40. [18] Sutton D, Kelleher B, Ross JRH. Review of literature on catalysts for biomass gasication. Fuel Process Tech 2001; 73(3):155e73. [19] Abu El-Rub Z, Bramer EA, Brem G. Review of catalysts for tar elimination in biomass gasication processes. Ind Eng Chem Res 2004;43(22):6911e9. [20] Corella J, Aznar MP, Gil J, Caballero MA. Biomass gasication in uidized bed: where to locate the dolomite to improve gasication. Energy Fuels 1999;13(6):1122e7. [21] Giuntoli J, de Jong W, Arvelakis S, Spliethoff H, Verkooijen AHM. Quantitative and kinetic TG-FTIR study of biomass residue pyrolysis: dry distillers grains with solubles (DDGS) and chicken manure. J Anal Appl Pyrolysis 2009;85(1e2):301e12. [22] Ohman M, Nordin A, Skrifvars B-J, Backman R, Hupa M. Bed agglomeration characteristics during uidised bed combustion of biomass fuels. Energy Fuels 2000;14(1):169e78. [23] Meng X.M., de Jong W., Fu N.J and Verkooijen A.H.M. Primary results of dried distillers grains with solubles gasication in a 100 kWth steam-oxygen blown circulating uidized bed gasier, the 8th international symposium on gas cleaning at high temperatures. Taiyuan, Shanxi, China; August 22e25 2010. [24] Brage C, Yu Q, Chen GX, Sjostrom K. Use of amino phase adsorbent for biomass tar sampling and separation. Fuel 1997;76(2):137e42. [25] Kiel JHA, van Passen SVB, Neeft JPA, Devi L, Ptasinski KJ, Janssen FJJG, et al. Primary measures to reduce tar formation in uidized bed biomass gasiers, nal report SDE project P1999-012; 2004. ECN-C-04e014. p. 1e108. [26] de Jong W. Nitrogen compounds in pressurized uidized bed gasication of biomass and fossil fuel, PhD dissertation at Delft University of Technology; 2005. [27] Jarungthammachote S, Dutta A. Thermodynamic equilibrium model and second law analysis of a downdraft waste gasier. Energy 2007;32(9):1660e9. [28] Channiwala SA, Parikh PP. A unied correlation for estimating HHV of solid, liquid and gaseous fuels. Fuel 2002; 81(8):1051e63. [29] Turn S, Kinoshita C, Zhang Z, Ishimura D, Zhou J. An experimental investigation of hydrogen production from biomass gasication. Int J Hydrogen Energy 1998;23(8):641e8. [30] Kumar A, Eskridge K, Jones DD, Hanna MA. Steameair uidized bed gasication of distillers grains: effects of steam to biomass ratio, equivalence ratio and gasication temperature. Bioresour Technol 2009;100(6):2062e8.

[31] Franco C, Pinto F, Gulyurtlu I, Cabrita I. The study of reactions inuencing the biomass steam gasication process. Fuel 2003;82(7):835e42. [32] Orfao JJM, Antunes FJA, Figueiredo JL. Pyrolysis kinetics of lignocellulosic materialsdthree independent reactions model. Fuel 1999;78(3):349e58. [33] Qin YH, Feng J, Li WY. Formation of tar and its characterization during airesteam gasication of sawdust in a uidized bed reactor. Fuel 2010;89(7):1344e7. [34] Gil J, Corella J, Aznar MP, Caballero MA. Biomass gasication in atmospheric and bubbling uidized bed: effect of the type of gasifying agent on the product distribution. Biomass Bioenergy 1999;17(5):389e403. [35] Wang Y, Kinoshita CM. Experimental analysis of biomass gasication with steam and oxygen. Solar Energy 1992;49(3): 153e8. [36] Meng X.M., de Jong W and Verkooijen A.H.M. Prediction of sulphur compounds distribution in gasication products of biomass fuels, 17th European biomass conference and exhibition. Germany, Hamburg; 29 Junee3 July 2009. p. 940e7. [37] Attar A, Dupuis F. The rate and the fundamental mechanisms of the reaction of hydrogen sulde with the basic minerals in coal. Ind Eng Chem Process Des Dev 1979; 18(4):607e18. [38] Dias M and Gulyurtlu I. H2S and HCl formation during RDF and coal co-gasication: a comparison between the predictions and experimental results. In: Proceedings of the biomass gasication technologies workshop MRC Gebze Campus-Turkiye; 2008. [39] Kinoshita CM, Wang Y, Zhou J. Tar formation under different biomass gasication conditions. J Anal Appl Pyrolysis 1994; 29(2):169e81. [40] Boateng AA, Walawender WP, Fan LT, Chee CS. Fluidized-bed steam gasication of rice hull. Bioresour Technol 1992;40(3): 235e9. [41] Brage C, Yu Q, Chen G, Sjostrom K. Tar evaluation proles obtained from gasication of biomass and coal. Biomass Bioenergy 2000;18(1):87e91. [42] van Paasen SVB and Kiel JHA. Tar formation in a uidisedbed gasier-impact of fuel properties and operating conditions, ECN-C-04e013; 2004 March. [43] Han J, Kim H. The reduction and control technology of tar during biomass gasication/pyrolysis: an overview. Renew Sustain Energy Rev 2008;12(2):397e416. [44] Rapagna S, Jand N, Kiennemann A, Foscolo PU. Steamgasication of biomass in a uidized-bed of olivine particles. Biomass Bioenergy 2000;19(3):187e97. [45] Li XT, Grace JR, Lim CJ, Watkinson AP, Chen HP, Kim JR. Biomass gasication in a circulating uidized bed. Biomass Bioenergy 2004;26(2):171e93. [46] Aznar MP, Caballero MA, Gil J, Martn JA, Corella J. Commercial steam reforming catalysts to improve biomass gasication with steam oxygen mixtures. 2. Catalytic tar removal. Ind Eng Chem Res 1998;37(7):2668e80. [47] Aznar MP, Corella J, Gil J, Martn JA, Caballero MA, Olivares A, et al. Biomass gasication with steam and oxygen mixtures at pilot scale and with catalytic gas upgrading part I: performance of the gasier. Dev Thermochem Biomass Convers 1997;2:1194e208. [48] Pecho J, Schildhauer TJ, Sturzenegger M, Biollaz S, Wokaun A. Reactive bed materials for improved biomass gasication in a circulating uidised bed reactor. Chem Eng Sci 2008;63(9):2465e76. [49] Rauch R, Pfeifer C, Bosch K, Hofbauer H, Swierczynski D, Courson C. et al. Comparison of different olivines for biomass steam gasication. In: Proceedings of the conference for science in thermal and chemical biomass conversion, vol. 1. Victoria, Canada; 2004. p. 799e809.