Bioresource Technology 102 (2011) 9743–9748

Contents lists available at SciVerse ScienceDirect

Bioresource Technology

journal homepage: www.elsevier .com/locate /bior tech

Kinetic modelling of steam gasification of various woody biomass chars:Influence of inorganic elements

Capucine Dupont ⇑, Timothée Nocquet, José Augusto Da Costa Jr., Christèle Verne-TournonCommissariat à l’Energie Atomique et aux Energies Alternatives, 17 rue des Martyrs, 38054 Grenoble cedex 09, France

a r t i c l e i n f o

Article history:Received 24 April 2011Received in revised form 6 July 2011Accepted 8 July 2011Available online 23 July 2011

Keywords:BiomassSteam gasificationCharInorganic elementsKinetics

0960-8524/$ - see front matter � 2011 Elsevier Ltd. Adoi:10.1016/j.biortech.2011.07.016

⇑ Corresponding author.E-mail address: capucine_dupont@hotmail.fr (C. D

a b s t r a c t

A study was performed on the influence of wood variability on char steam gasification kinetics. Isother-mal experiments were carried out in a thermobalance in chemical regime on various wood chars pro-duced under the same conditions. The samples exhibited large differences of average reaction rate.These differences were linked neither with the biomass species nor age and may be related to the bio-mass inorganic elements. A modelling approach was developed to give a quantitative insight to theseobservations. The grain model was used on one biomass of reference for temperatures between 750and 900 �C and steam partial pressures between 0 and 0.27 bar. The model was applied to the other sam-ples through the addition of an integral parameter specific to each sample. A satisfactory correlation wasfound between this parameter and the ratio pota ssium/silicium. This result highlighted the catalyticeffect of potassium and inhibitor effect of silicium on the reaction.

� 2011 Elsevier Ltd. All rights reserved.

1. Introduction In her very complete review on lignocellulosic chars gasification

In the present energy context, there is a growing interest world-wide for heat, power and biofuels production from biomass (Simset al., 2010), notably through gasification (Karmakar and Datta,2011; Xiao et al., 2011). The knowledge of biomass char steam gas-ification kinetics is of major importance in the design of advancedgasifiers since this reaction is the limiting phenomenon in thetransformation (Dupont et al., 2007) and therefore controls conver-sion. Due to the limited availability of biomass, gasifiers will haveto be supplied with various feedstocks, of different species andcoming from different places of growth.

Previous studies have shown that the kinetics of char steamgasification could be very different according to the biomass witha factor of more than twenty for chars prepared in an identicalway (Moilanen, 2006; Septien et al., 2009). A simple calculationbased on the Arrhenius law shows that even a difference of onlya factor of four between two biomasses would imply to changethe operating reactor temperature of 100 �C to achieve the same le-vel of conversion. It could therefore strongly impact the processcontrol. It is therefore of great importance to study the intrinsickinetics of steam gasification of chars from various biomasses.

When series of chars are prepared and gasified under identicalconditions, the differences of reactivity may only be attributed todifferences of morphological structure and of inorganic elementscontent.

ll rights reserved.

upont).

rate, Di Blasi reports that surface area, and consequently morpho-logical structure, seems to be less influential on gasification reac-tivity than content in inorganic elements, and particularlysoluble minerals (Di Blasi, 2009). This latter parameter seemstherefore to be the most important parameter to consider forunderstanding the differences of gasification behaviour of charsfrom different biomasses.

Numerous studies deal with the influence of the content in inor-ganic elements in the case of gasification of charcoal with steam orCO2, for instance (Everson et al., 2008; Kajitani et al., 2002; Lee andKim, 1995; Ochoa et al., 2001; Zhang et al., 2010) or of biomasschars with CO2, such as (Cetin et al., 2004; Huang et al., 2009;DeGroot and Shafizadeh, 1984; Struis et al., 2002; Seo et al.,2010). There are fewer studies in the case of steam gasification ofbiomass chars (Kajita et al., 2009; Moilanen, 2006; Yip et al.,2009; Zhang et al., 2008; Zhu et al., 2008). Hence, as underlinedby Di Blasi (2009), there is at the moment a significant need to bet-ter understand the influence of the inorganic elements content onsteam gasification kinetics of biomass chars.

To estimate the influence of inorganic elements on gasificationreactivity, most authors use the same biomass and then chooseeither to impregnate the solid of inorganic elements (Hawleyet al., 1983; Huang et al., 2009) or to leach it with aqueous or acidsolution (Kajita et al., 2009; Marquez-Montesinos et al., 2002; Yipet al., 2009). Possible differences of morphological structure be-tween the biomass samples are avoided thanks to these methods.However, biomass structure may be modified by these treatmentsand the location of the impregnated inorganic elements may not becomparable with those of the indigenous ones. Moreover, the

9744 C. Dupont et al. / Bioresource Technology 102 (2011) 9743–9748

efficiency of leaching is not total (Etiégni and Campbell, 1991;Marquez-Montesinos et al., 2002; Yip et al., 2009). The other op-tion is to directly compare different biomass feedstocks with dif-ferent inorganic elements contents (Moilanen, 2006). This mayinduce some bias due to the differences of morphological structurebut has the advantage of considering inorganic elements in theirnatural form and location inside the solid.

Generally speaking, the alkaline (sodium, potassium) and alka-line earth (calcium, magnesium) metallic (AAEM) species are rec-ognized to increase the reaction rate (Zhang et al., 2008; Zhuet al., 2008), on the contrary to silicium.

Although qualitative evidence has largely been given on the ma-jor importance of the inorganic elements content on char gasifica-tion rate, none of the existing kinetic models is able to take thisparameter into account. Hence, all these models are only relatedto one biomass. Only Zhang et al. (2008) have recently made aninteresting attempt to develop a quantitative approach of the influ-ence of the inorganic elements in the description of steam gasifica-tion of biomass chars. Two parameters, called c and p, were addedin the random-pore model (Bhatia and Perlmutter, 1980), as previ-ously made by Struis on coal chars (Struis et al., 2002). The origi-nality lies in the correlation of these parameters with thepotassium content of the twelve biomasses used for the develop-ment of the model. However, the correlation was not explicitly gi-ven and it is therefore difficult to test the model performance andits physical meaning.

Based on this background, the present study aims at betterunderstanding steam gasification kinetics of biomass chars anddeveloping a kinetic model valid for various biomass chars, by con-sidering in a quantitative way the influence of the inorganic ele-ments contained in the biomass chars. To achieve this goal,gasification experiments were performed in thermobalance onchars from 21 different biomass samples under different operatingconditions of temperature and steam partial pressure.

2. Methods

2.1. Biomass samples

The biomass used in this study consisted in 21 samples of woodchips, from common species of woods, coming from various places

Table 1List of samples and of their composition in major inorganic elements.

Sample name Species Si Na

mg/kg dry biomass

1 Spruce 661 332 Beech, hornbeam, oak 370 83 Beech, oak 651 124 Pine 291 115 Beech 584 136 Softwoods 3619 867 Hornbeam 224 48 SRFa poplar 498 309 Birch 85 10

10 Beech 757 2711 Oak 119 1112 SRCb willow 841 1413 SRC poplar 642 1914 Poplar 763 4815 Reference sample (beech) 260 1816 Oak 112 617 Pine, spruce 375 1218 Pine 252 1919 SRF poplar 366 2020 SRC poplar 632 4821 Hornbeam, oak 108 8

a SRF: Short Rotation Forestry.b SRC: Short Rotation Coppice.

in France and of different ages, including Short Rotation Forestryand Short Rotation Coppice.

The samples properties were measured following the Europeanstandards on biofuels. In particular, ash content was measured at550 �C and ash composition was measured for the following ele-ments (SiO2, Al2O3, Fe2O3, TiO2, CaO, MgO, K2O, Na2O). The main re-sults are given in Table 1. These properties were presented anddiscussed in detail elsewhere (Dupont et al., 2010).

It has to be kept in mind that these values should be taken withcaution due to the hardly-avoidable uncertainty of properties mea-surement of heterogeneous solid.

For all samples, ash was mainly composed of calcium. Potas-sium and silicium were the other two important components, fol-lowed by magnesium. The ash content as well as the relativeamounts of elements were quite different among the samples.

No correlation could be found between properties and species.The older the tree, the lower the ash content. This seems logicalsince the proportion of bark, which contains the highest amountof ash in wood, is higher in young wood.

2.2. Experimental procedure

2.2.1. General descriptionTo ensure good representativeness, 30 g of each sample of wood

chips was sampled following the standard XP CENT/TS 14780.The sample was pyrolysed under N2 atmosphere in a low heat-

ing rate furnace (a few �C min�1). It was kept at the final tempera-ture of 450 �C during 4 h. Then the produced char, which countedfor about 25%w of the initial biomass, was ground with a mortarand sieved below 50 lm. Char was assumed to be ground homoge-neously enough to prevent any segregation when sieving. Char gas-ification with steam took place in a Thermo Gravimetric Analysis(TGA) device operating at atmospheric pressure (SETARAM Setsyscoupled with steam generator Wetsys). The device is depicted inFig. 1. Five milligrams of the sample was placed in the crucible ofthe thermobalance. This crucible was a cylinder of 2.5 mm heightand 8 mm diameter. The sample was heated at a rate of24 �C min�1 to the gasification temperature under a N2 gas flowof 0.05 L min�1 to end pyrolysis. After the gasification temperaturewas reached and no mass loss was observed, the gas was switchedto a mixture of H2O/N2 (0–27 vol%, with an uncertainty below 4%)

K Ca Mg Al Ash

wmf%

236 1003 151 95 0.6653 2665 358 26 1.1561 858 207 57 0.8112 2177 263 57 0.6403 3349 427 43 0.9

1520 6576 733 507 3.3523 2451 126 25 0.6

1175 12,248 692 98 4.1328 1746 249 17 0.7

1442 3795 532 24 1.8561 4545 262 16 1.5654 2926 273 79 1.5

1784 7661 675 86 2.51855 15,879 1600 102 4.3

403 1196 313 38 0.6531 4894 386 21 1.6

1250 1909 198 64 1.1164 1157 297 58 0.5

1222 10,034 778 86 2.72019 6987 753 158 2.6

549 6679 253 11 2.1

Fig. 1. Scheme of the TGA device.

Fig. 2. Mass loss versus time for a typical gasification run (T = 800 �C; P = 0.27 bar).

C. Dupont et al. / Bioresource Technology 102 (2011) 9743–9748 9745

with the same total flow rate and gasification occurred. When thereaction was finished, the furnace was brought to ambienttemperature.

Based on the measurements of mass loss versus time, the gasi-fication rate r could then be derived and expressed by the variationof the conversion X versus time:

r ¼ dXdt

ð1Þ

With the following expression of conversion:

X ¼ mi �mðtÞmi �mf

ð2Þ

where mi, m(t) and mf are the masses of char before gasification, atthe time t and at the end of gasification, respectively.

The average reactivity rinteg was also defined between twostages of conversion X1 and X2 by:

rinteg ¼R tX2

tX1

rðtÞdt1�XðtÞ

tX2 � tX1

ð3Þ

Note that there is no standardized definition of reactivity. In partic-ular, authors choose their own limits of conversion (Barrio, 2002),which sometimes makes results comparison difficult.

2.2.2. Regime of reactionPreliminary experiments and calculations have shown that un-

der the operating conditions (T = 750–900 �C; particle size < 50 lm;sample mass = 5 mg), the intrinsic kinetics regime was reached andthat there were no limitations by heat and mass transfers (Nocquet,2009).

2.2.3. RepeatabilityThe repeatability of the TGA experiments was checked by dou-

bling each test. The difference between two tests was alwaysbelow 0.2% of the average reaction rate. The influence of the sam-pling and pyrolysis steps was also checked by performing the prep-aration of one sample twice. The differences in the TGA resultswere in the range of uncertainty, namely below 0.2%.

3. Results and discussion

3.1. Comparison of reaction rates

As shown in Fig. 2 and previously mentioned, it can be seen thatsteam was injected only when there was no more mass loss due topyrolysis. The mass loss rate during gasification appeared to benearly constant during most of the reaction, except at its end, whena sharp increase could be observed, in agreement with some previ-ous studies (Marquez-Montesinos et al., 2002).

In order to consider the constant part of the curve, the averagereaction rate rinteg, given by Eq. (3), was calculated for conversionsbetween 1% and 80%, as previously done by Barrio (2002).

The values obtained for the different biomass samples at 800 �Cand PH2O ¼ 0:27 bar are shown in Fig. 3a. The mean value was ofabout 6% min�1, with a variation of a factor of 3.5 among the sam-ples. These values were in the order of magnitude of those ofMoilanen (2006) and led to a characteristic time of steam gasifica-tion of about 103 s at 800 �C, which confirms that this reaction oc-curs very slowly under typical gasifiers conditions. As explained inintroduction, the difference of reaction rate between the samples isequivalent to a difference of temperature to get the same level ofconversion with the different samples. Here, there was a factor of3.5 between the samples, which corresponds to about 90 �C of dif-ference of temperature. Such a difference may be quite significantin terms of process control.

This difference of reaction rate did not appear to be correlatedwith the species themselves. Indeed, for instance, the two samplesof beech numbered 5 and 15 had different reaction rates. No trendappeared neither regarding the wood family, namely softwoodsand hardwoods, which confirms the findings of De Groot (1984).The age of the wood did not seem to have any influence sincethe sample of SRC poplar 13 and of ‘‘old’’ poplar 14 had nearlythe same reaction rates, whereas the SRC poplar 20 had a signifi-cantly faster reaction rate. As already observed by De Groot(1984) and mentioned by Di Blasi (2009), these differences mustthen be attributed to differences in inorganic elements content.

Average reaction rates on the whole range of conversion ap-peared to be very variable among the biomass samples. It may beinteresting to see whether these differences were sensible alongthe whole range of conversion or more specifically in narrowerranges of conversion. This may help to better understand themechanisms.

The average reaction rates have been plotted for the differentbiomass samples in Fig. 3b–e for limited ranges of conversion,namely 1–20%, 20–40%, 40–60% and 60–80%.

Globally speaking, the average reaction rate tended to increasewith conversion. For conversions ranging from 1% to 20% and 20%

0

2

4

6

8

10

12

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 2Sample number

Ave

rag

e re

acti

on

rat

e b

etw

een

X=1

-80%

(%

.min

-1)

a

0

1

2

3

4

5

6

7

8

9

10

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

Sample number

Ave

rag

e re

acti

on

rat

e b

etw

een

X=2

0-40

% (

%.m

in-1

)0

1

2

3

4

5

6

7

8

9

10

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21A

vera

ge

reac

tio

n r

ate

bet

wee

n X

=20-

40 %

(%

.min

-1)

c

0

1

2

3

4

5

6

7

8

9

10

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

Sample number

Ave

rag

e re

acti

on

rat

e b

etw

een

X=4

0-60

% (

%.m

in-1

)

0

1

2

3

4

5

6

7

8

9

10

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

Ave

rag

e re

acti

on

rat

e b

etw

een

X=4

0-60

% (

%.m

in-1

)

d

0

5

10

15

20

25

30

35

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

Sample number

Ave

rag

e re

acti

on

rat

e b

etw

een

X=6

0-80

% (

%.m

in-1

)

0

5

10

15

20

25

30

35

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

Ave

rag

e re

acti

on

rat

e b

etw

een

X=6

0-80

% (

%.m

in-1

)

e

0

1

2

3

4

5

6

7

8

9

10

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

Sample number

Ave

rag

e re

acti

on

rat

e b

etw

een

X=1

-20

% (

%.m

in-1

)

b

Fig. 3. Average reaction rate of the samples at 800 �C and PH2 O ¼ 0:27 bar for conversions between: (a) 1% and 80%; (b) 1% and 20%; (c) 20% and 40%; (d) 40% and 60%; (e) 60%and 80%.

9746 C. Dupont et al. / Bioresource Technology 102 (2011) 9743–9748

to 40%, this value was of about 3% min�1, while it was of about5% min�1 for conversions ranging from 40% to 60% and of about10% min�1 for conversions between 60% and 80%. This was inagreement with previous authors’ observations. As mentioned byprevious authors (Di Blasi, 2009), this may be related to develop-ment of the char surface area or to the concentration increase ofinorganic elements in the char along the reaction, which wouldstrengthen their catalytic effect.

It could also be seen that the differences between the samplesincreased with conversion. Hence, a factor of three could be ob-served between minimum and maximum values of reaction ratesfor conversions ranging from 1% to 20% and from 20% to 40%. Forconversions ranging from 40% to 60%, this factor was of 5 and forconversions ranging from 60% to 80%, this factor was of 10. Assum-ing as previously that the difference of reaction rate between thebiomass samples is mainly due to their content in inorganic ele-ments, this means that the influence of the inorganic elements var-ies with conversion. The higher the conversion, the higher theconcentration of inorganic elements and therefore the higher theircatalytic or inhibitor effect.

3.2. Kinetic modelling

Based on the previous results, the objective of this section is todevelop a modelling approach able to predict the different reaction

rates of the samples. This modelling approach is based on theassumption that the differences of reaction rates are mainly relatedto the inorganic elements content in biomass.

3.2.1. Kinetic modelling principleThe approach is based on several main steps.

� First, a sample of reference was selected and a model was devel-oped that described the gasification kinetics of this sample.� This model was then applied to the description of the kinetics of

the other samples through a corrective factor.� A correlation of this mathematical factor was finally determined

with biomass properties, namely inorganic elements.

3.2.2. Kinetic modelling on the reference sampleThe beech sample number 15 was chosen as reference since this

species is one of the most common woods in France and that itscontent in inorganic compounds is very low.

Four main models can be generally found in the literatureregarding biomass char gasification chemical kinetics: the volu-metric model, the grain model (Szekely and Evans, 1970), the ran-dom-pore model (Bhatia and Perlmutter, 1980) and the Langmuir–Hinshelwood model (Hinshelwood, 1940; Langmuir, 1922). Thelast one is specific since it is able to take into account the influenceof reactive gas such as H2 on the kinetics. Among the other three

0

20

40

60

80

100

t (s)

X (%

) ExperimentModel

0

20

40

60

80

100

0 500 1000 1500 2000

0 500 1000 1500 2000 2500 3000 3500 4000

t (s)

X (%

) ExperimentModel without aModel with a

a

b

Fig. 5. Conversion versus time: (a) obtained experimentally and through model onthe reference sample; (b) obtained experimentally and through model withoutparameter a and with parameter a on the sample of spruce.

C. Dupont et al. / Bioresource Technology 102 (2011) 9743–9748 9747

ones, only the grain model and the random-pore model take intoaccount the porous structure of the particle. Comparisons of theperformance of these models are scarce (Fermoso et al., 2010;Seo et al., 2010). The random-pore model, which has the highestnumber of adjustable parameters, appears to give the best resultsand to describe the evolution of the solid structure with conver-sion. However, this parameter could neither be accurately mea-sured nor calculated in the experiments related in the presentstudy. Hence, the grain model was selected in the present study.

The kinetic parameters were obtained through experiments atdifferent temperatures (750; 800; 850; 900 �C). The values werefound to be in excellent agreement with Arrhenius law, with acoefficient of correlation of 0.996. They were also found to be sim-ilar to the literature values presented in Di Blasi’s review.

The influence of pressure was studied by varying steam partialpressure between 0 and 0.27 bar. As shown in Fig. 4, an increase insteam partial pressure logically led to a decrease of the timeneeded for complete conversion. This influence seemed relativelylimited. Quantitatively speaking, it seemed to be well-describedthrough a power law with a constant exponent value during nearlythe whole conversion, except at very high level of conversion, inagreement with the results of Marquez–Montesinos. This valuewas of about 0.6, which was close to the values found by Kojima(0.41), Barrio (0.51) and Hemati (0.73) and equal to the one foundby Marquez–Montesinos.

The kinetic law was then as follows:

dXdt¼ 8:77� 104 exp

�167000RT

� �P0:6

H2Oð1� XÞ2=3

In terms of average reaction rate between 1 and 80%, the error be-tween experiments and model was below 15%, which is very satis-factory. The global trend of the conversion curve was also correctlymodelled, except at high conversions, as shown in Fig. 5a.

3.2.3. Application of the model to the whole range of biomass samplesAn integral parameter ai specific of each sample was added in

the law to minimize the sum of the differences between the exper-imental conversion rate and the modelling one.

dXdt¼ 8:77� 104 exp

�167000RT

� �P0:6

H2Oaið1� XÞ2=3

It has to be noted that this parameter is integral. It is therefore notable to describe the influence of conversion on the differences be-tween biomasses, which was observed in experiments, but it hasthe advantage of being very simple. As shown in Fig. 5b, resultswere greatly improved through this model with integral parameterand appeared to be very satisfactory. Hence the addition of this sim-ple parameter seemed to be sufficient for good prediction of the or-der of magnitude of the rate.

0

0,5

1

0 2000 4000 6000 8000

t(s)

X(%

)

0.02 0.15 0.10 0.050.2 0.20

50

100 7 PH2O(bar)

Fig. 4. Conversion versus time for different steam partial pressures.

Correlations of this mathematical parameter were searchedwith physical parameters in order to explain differences of reactionrates between biomasses. As previously mentioned, both structure,that is surface area, and inorganic elements amounts may have aninfluence.

Here it was assumed that all biomasses had the same structure.This assumption seems reasonable since only woody biomass wasconsidered. There may be some differences of structure betweensoftwoods and hardwoods, but this did not result in clear differ-ences of reaction rates between these two families. Moreover, asmentioned in the background, studies reported that surface areahad less influence on gasification reactivity than inorganicelements.

It was also assumed that the inorganic elements remained inbiomass chars and were not volatilized under the explored rangeof temperatures. This seems reasonable since authors asserted thatthe volatilization of biomass inorganic compounds occurred athigher temperatures than those of these experiments (Eversonet al., 2008; Marquez-Montesinos et al., 2002; Ochoa et al.,2001). Experiments were also performed on chars pre-heated un-der N2 at 900 �C and then gasified at 800 �C. No significant differ-ence was found with results obtained following the usualprocedure, which consisted in heating the chars only up to800 �C. Based on the assumption that the reaction rate was mainlyinfluenced by the inorganic content of the char, this result tendedto show that there was no significant volatilization of the inorganicelements in this range of temperatures.

Various correlations of the parameter ai were tested versus bio-mass inorganic elements content (potassium, calcium, silicium,sum of potassium and calcium, ratio potassium/silicium, ratio alka-line compounds/basic compounds, etc.). As shown in Fig. 6 a linearcorrelation versus the ratio potassium/silicium was found to be

R2 = 0,77R2 = 0,77R2 = 0,77

0 1 2 3 4 5 6

0.25

0

0.5

0.75

1

1.25

1.5

1.75

2

a i(-)

mK/mSi (kg.kg-1)

R2=0.77

Fig. 6. Correlation between the corrective factor ai and the ratio potassium/siliciumof biomass.

9748 C. Dupont et al. / Bioresource Technology 102 (2011) 9743–9748

significant under the studied range of ratios, that is potassium/silicium < 5. The kinetic model then became as follows:

dXdt¼8:77� 104 exp

�167000RT

� �0:1812

mk

msiþ 0:5877

� �

� P0:6H2Oð1� XÞ2=3

This correlation highlights the accelerating effect of potassium andthe inhibitor effect of silicium, as qualitatively shown by differentprevious authors. The effect of calcium, which is often reported asa gasification catalyst, was not shown here. This might be due tothe limited range of variation of calcium content in the tested sam-ples and to its global high content in all chars. Future work may dealwith this issue by using, if possible, wood samples with various cal-cium contents. Also, in a more general way, as mentioned above, thepossible inaccuracy of the inorganic elements measurement in bio-mass may be a source of error. To limit this bias, it would be inter-esting to increase the number of tested samples.

Finally, through this study, an original modelling approach hasbeen developed that shows that the rate of steam gasification ofvarious woody biomass chars may be satisfactorily predicted fromtheir composition in inorganic elements.

4. Conclusion

The present study aims at modelling the influence of wood var-iability on char steam gasification kinetics. Experiments in thermo-balance showed a factor of 3.5 between the average reaction rate ofvarious wood biomass chars produced under the same conditions.Hence an original kinetic model was developed to describe the gas-ification rate of the whole set of biomasses by taking into accounttheir inorganic elements content. The rate seemed to be correlatedwith the ratio potassium/silicium under the explored conditions.This shows the catalytic effect of potassium and the inhibitor effectof silicium on steam gasification of biomass chars.

Acknowledgements

The authors gratefully acknowledge FCBA, the association AILE,ONF and UCFF for the supply of the biomass samples.

References

Barrio, M., 2002. Experimental Investigation of Small-scale Gasification of WoodyBiomass. The Norwegian University of Science and Technology.

Bhatia, S.K., Perlmutter, D.D., 1980. A random pore model for fluid–solid reactions: I.Isothermal, kinetic control. AIChE 26, 379–383.

Cetin, E., Moghtaderi, B., Gupta, R., Wall, T.F., 2004. Influence of pyrolysis conditionson the structure and gasification reactivity of biomass chars. Fuel 83, 2139–2150.

DeGroot, W.F., Shafizadeh, F., 1984. Kinetics of gasification of Douglas Fir andCottonwood chars by carbon dioxide. Fuel 63, 210–216.

Di Blasi, C., 2009. Combustion and gasification rates of lignocellulosic chars.Progress in Energy and Combustion Science 35, 121–140.

Dupont, C., Boissonnet, G., Seiler, J.-M., Gauthier, P., Schweich, D., 2007. Study on thekinetic processes of biomass steam gasification. Fuel 86, 32–40.

Dupont, C., Rougé, S., Berthelot, A., Dasilvaperez, D., Graffin, A., Labalette, F.,Laboubée, C., Mithouard, J., Pitocchi, S., 2010. Suitability of wood chips anddifferent biomass feedstocks for use in a semi-industrial plant of BtL productionby gasification. International Journal of Chemical Reactor Engineering 8.

Etiégni, L., Campbell, A.G., 1991. Physical and chemical characteristics of wood ash.Bioresource Technology 37, 173–178.

Everson, R.C., Neomagus, H.W.J.P., Kaitano, R., Falcon, R., du Cann, V.M., 2008.Properties of high ash coal–char particles derived from inertinite-rich coal: II.Gasification kinetics with carbon dioxide. Fuel 87, 3403–3408.

Fermoso, J., Gil, M.V., Pevida, C., Pis, J.J., Rubiera, F., 2010. Kinetic models comparisonfor non-isothermal steam gasification of coal-biomass blend chars. ChemicalEngineering Journal 161, 276–284.

Hawley, M.C., Boyd, M., Anderson, C., DeVera, A., 1983. Gasification of wood charand effects of intraparticle transport. Fuel 62, 213–216.

Hinshelwood, C.N., 1940. The Kinetics of Chemical Change. Clarendon Press, Oxford.Huang, Y., Yin, X., Wu, C., Wang, C., Xie, J., Zhou, Z., Ma, L., Li, H., 2009. Effects of

metal catalysts on CO2 gasification reactivity of biomass char. BiotechnologyAdvances 27, 568–572.

Kajita, M., Kimura, T., Norinaga, K., Li, C.-Z., Hayashi, J.-i., 2009. Catalytic andnoncatalytic mechanisms in steam gasification of char from the pyrolysis ofbiomass. Energy and Fuels 24, 108–116.

Kajitani, S., Hara, S., Matsuda, H., 2002. Gasification rate analysis of coal char with apressurized drop tube furnace. Fuel 81, 539–546.

Karmakar, M.K., Datta, A.B., 2011. Generation of hydrogen rich gas through fluidizedbed gasification of biomass. Bioresource Technology 102, 1907–1913.

Langmuir, I., 1922. Transactions of the Faraday Society 17, 621.Lee, W.J., Kim, S.D., 1995. Catalytic activity of alkali and transition metal salt

mixtures for steam–char gasification. Fuel 74, 1387–1393.Marquez-Montesinos, F., Cordero, T., RodrIguez-Mirasol, J., RodrIguez, J.J., 2002. CO2

and steam gasification of a grapefruit skin char. Fuel 81, 423–429.Moilanen, A., 2006. Thermogravimetric characterisations of biomass and waste for

gasification processes. VTT, Espoo.Nocquet, T., 2009. Rapport – Détermination d’un modèle cinétique de

vapogazéification de différentes biomasses forestières à l’aide d’essais enthermobalance. CEA, Grenoble.

Ochoa, J., Cassanello, M.C., Bonelli, P.R., Cukierman, A.L., 2001. CO2 gasification ofArgentinean coal chars: a kinetic characterization. Fuel Processing Technology74, 161–176.

Seo, D.K., Lee, S.K., Kang, M.W., Hwang, J., Yu, T.-U., 2010. Gasification reactivity ofbiomass chars with CO2. Biomass and Bioenergy. Corrected Proof 34, 1946–1953.

Septien, S., Verne-Tournon, C., Nocquet, T., Dupont, C., 2009. Steam gasification charreactivity from different biomass. In: 17th European Biomass Conference,Hamburg.

Sims, R.E.H., Mabee, W., Saddler, J.N., Taylor, M., 2010. ‘‘An overview of secondgeneration biofuel technologies. Bioresource Technology 101 (6), 1570–1580.

Struis, R., von Scala, C., Stucki, S., Prins, R., 2002. Gasification reactivity of charcoalwith CO2. Part II: Metal catalysis as a function of conversion. ChemicalEngineering Science 57, 3593–3602.

Szekely, J., Evans, J.W., 1970. A structural model for gas–solid reactions with amoving boundary. Chemical Engineering Science 25, 1091–1107.

Xiao, X., Meng, X., Le, D.D., Takarada, T., 2011. Two-stage steam gasification of wastebiomass in fluidized bed at low temperature: Parametric investigations andperformance optimization. Bioresource Technology 102, 1975–1981.

Yip, K., Tian, F., Hayashi, J.-i., Wu, H., 2009. Effect of alkali and alkaline earth metallicspecies on biochar reactivity and syngas compositions during steamgasification. Energy and Fuels 24, 173–181.

Zhang, Y., Ashizawa, M., Kajitani, S., Miura, K., 2008. Proposal of a semi-empiricalkinetic model to reconcile with gasification reactivity profiles of biomass chars.Fuel 87, 475–481.

Zhang, Y., Hara, S., Kajitani, S., Ashizawa, M., 2010. Modeling of catalytic gasificationkinetics of coal char and carbon. Fuel 89, 152–157.

Zhu, W., Song, W., Lin, W., 2008. Catalytic gasification of char from co-pyrolysis ofcoal and biomass. Fuel Processing Technology 89, 890–896.