Chemical Engineering Journal 131 (2007) 233–244

Thermodynamic modeling and analysis of biomass gasification

for hydrogen production in supercritical water
Youjun Lu, Liejin Guo ∗ , Ximin Zhang, Qiuhui Yan
State Key Laboratory of Multiphase Flow in Power Engineering (SKLMF), Xi’an Jiaotong University, Xi’an, Shaanxi 710049, China
Received 23 February 2006; received in revised form 7 October 2006; accepted 24 November 2006

Abstract
Biomass gasification in supercritical water is a promising technology for hydrogen production by utilizing wet biomass. A new experimental
system of biomass gasification in supercritical water was built in SKLMF. In this paper, a comprehensive thermodynamic analysis, including
chemical equilibrium in the reactor, gas–liquid equilibrium in the high-pressure separator, exergy and energy analysis of the whole system, was
conducted. Chemical equilibrium model is based on minimizing Gibbs free energy. By chemical equilibrium analysis in the reactor, rules of the
main parametric effects on biomass gasification in supercritical water are obtained. Simultaneously, a high-pressure gas–liquid equilibrium model
was proposed based on modified universal functional activity coefficient (UNIFAC) model, Soave–Redlich–Kwong (SRK) equation of state and
modified Huron–Vidal second-order (MHV2) mixing rule. Effects of pressure, temperature and water recycled ratio on gas–liquid equilibrium in
high-pressure separation were discussed. Finally, results from energy and exergy analysis show that energy and exergy efficiencies of the whole
system are in excess of 40% and increase with increasing heat transfer efficiencies. Energy loss of the system is caused mainly by heat transfer and
exergy loss is mainly caused by heat transfer and chemical reaction. Our research provided a thermodynamic tool for improvement of design and
operation optimization of biomass gasification system in SKLMF, which may be also applicable to other biomass gasification system.
© 2006 Elsevier B.V. All rights reserved.

Keywords: Hydrogen production; Supercritical water; Biomass gasification; Chemical equilibrium; Phase equilibrium; Energy; Exergy

1. Introduction much less hydrogen bonds and their strength is much weaker. As
a result, SCW behaves like organic solvents so that many organic
Compared with fossil fuel, biomass is a clean energy with compounds have very high solubility in it. Moreover, gases are
zero CO2 emission, because CO2 is fixed by photosynthesis dur- miscible in SCW. Thus, chemical reaction can be conducted in
ing biomass growth and released again during utilization. Due a single supercritical phase reaction medium. High concentra-
to its low energy density, direct use of biomass is not conve- tions of reactants can often be attained and there are no interphase
nient. Thus, it is necessary to convert biomass to fuel gas, such mass transport processes to hinder reaction rates [2]. As a result,
as hydrogen, which can be used cleanly and highly efficiently in biomass gasification in SCW has a high reaction rate. In addition,
fuel cells. Thermo-chemical gasification is likely to be the most biomass gasification in SCW has high gasification efficiency
cost-effective conversion process. However, a large portion of at much lower temperatures of approximately 673 K compared
biomass is wet, and this causes high drying costs in classical with conventional gasification [3]. Furthermore, biomass gasi-
thermo-chemical gasification process [1]. With the advantage fication in SCW produces higher concentration of hydrogen in
of avoiding drying process, biomass gasification in supercritical product gas, because the high water excess favors the formation
water (SCW) is a promising technology for the utilization of wet of H2 and CO2 instead of CO.
biomass. So far various experimental investigations into gasification
SCW possesses properties much different from those of liq- of biomass model compounds and real biomass in SCW have
uid water. The dielectric constant of SCW is much lower, there is been carried out [1,3–16]. But the work on thermodynamic
analysis of this process is limited. Thermodynamic analysis
is very helpful in providing theoretical guidance for optimiza-
∗ Corresponding author. Tel.: +86 29 8266 3895; fax: +86 29 8266 9033. tion of design and operation of biomass gasification system.
E-mail address: lj-guo@mail.xjtu.edu.cn (L. Guo). Tang and Kitagawa [17] and Yan et al. [18] performed chemical

1385-8947/$ – see front matter © 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.cej.2006.11.016
234 Y. Lu et al. / Chemical Engineering Journal 131 (2007) 233–244

Nomenclature
Subscript
a co-energy parameter in the SRK EOS c critical parameter
aij the molar number of element i in compound j
anm , anm,1 , anm,2 UNIFAC interaction parameters
b co-volume parameter in the SRK EOS equilibrium analysis of hydrogen production by biomass gasi-
bi0 the total molar number of element i in the initial fication in SCW based on Gibbs free energy minimization.
reactant Feng et al. [19] calculated driving forces and phase equilibrium
Cp specific heat capacity for hydrothermal upgrading in sub-critical water and biomass
C1 , C2 , C3 Mathias–Copeman parameters gasification in SCW. Calzavara et al. [20] evaluated biomass
EX , EX,ph , EX,c , EX,Q exergy, physical exergy, chemical gasification in SCW process for hydrogen production and energy
exergy and heat exergy, respectively efficiency of the process was calculated in the ideal case. Mat-
fi fugacity of component i sumura and Minowa [21] conducted fundamental design of
g, gE Gibbs free energy and excess Gibbs free energy SCW fluidized bed for biomass gasification and thermal effi-
H, Ho∗ enthalpy and enthalpy at reference state ciency for the ideal case was also calculated. Yoshida et al. [22]
K equilibrium ratio performed comprehensive comparison of efficiencies and CO2
LHV low heating value emissions between biomass energy conversion technologies and
nj molar number of component j the results show that SCW gasification combined cycle for elec-
P pressure tricity generation is the most efficient for high moisture content
q1 , q2 parameters in MHV2 mixing rule biomass.
Qk van der Waals surface area of group k An experimental system was built in SKLMF to study
ri volume parameters of component i biomass gasification in SCW for hydrogen production, and the
R gas constant objective of this paper is to examine the thermodynamics of
Rk van der Waals volume of group k biomass gasification process in SCW based on the experimen-
S, So∗ entropy and entropy at reference state tal system. Thermodynamic models for chemical equilibrium in
T temperature the reactor and gas–liquid equilibrium in high-pressure separa-
V specific volume tor were developed, and exergy and energy analysis of the whole
xi molar fraction of component i in the liquid phase system were conducted. According to the thermodynamic anal-
yi molar fraction of component i in the gas phase ysis, some advice was present for improvement and operation
zi molar fraction of component i optimization of the experimental system.
Z compress factor
ZO , ZC , ZH , ZN weight fractions of oxygen, carbon, hydro-
2. Experimental system for biomass gasification in SCW
gen and nitrogen, respectively, in the biomass

Greek symbols Fig. 1 shows the schematic of experimental system for

Φi fugacity coefficient of component i biomass gasification in SCW in SKLMF. The system includes
(i)
Γ k , Γk activity coefficient of group k and activity coef- mainly reactor, preheater, heater exchanger, high-pressure sepa-
ficient of group k in molecule i, respectively. rator (SPE1), low-pressure separator (SPE2), and so on. Biomass
Θm molar fraction of surface area of group m feedstock at room temperature is pressurized by high-pressure
ε0,i standard chemical exergy of a pure chemical com- pump 1, and then enters the reactor. At the same time, water
pound i with larger flow rate is compressed by the high-pressure pump
γi activity coefficient of component i 2, and then heated to high temperature in the heat exchanger and
ηen , ηex energy and exergy efficiency, respectively preheater. At the inlet of reactor, biomass feedstock with small
μi chemical potential of component i flow rate and high-temperature water with larger flow rate mix
(i) together, so biomass feedstock is heated quickly to supercriti-
υk number of structural groups of type k in molecule
cal temperature. Faster heating of biomass to high temperature
i
increases the biomass gasification efficiency according to our
ω acentric factor
previous study [16]. After leaving the reactor, the high temper-
Ψ nm UNIFAC group interaction parameter between
ature fluid is cooled in the heat exchanger firstly with the heat
groups n and m
being recycled, and is then cooled to environmental temperature
Superscripts in the cooler. In SPE1, product CO2 is separated from product H2
C combinatorial part by high-pressure water absorption because solubility of CO2 in
g gas phase high-pressure water is much larger than that of H2 . Gas phase in
l liquid phase SPE1 is mainly composed of H2 . H2 from SPE1 is then decom-
R residual part pressed. CO2 absorbed in the liquid phase in SPE1 is finally
* ideal gas released in SPE2. When thermodynamic analysis is conducted,
it is assumed that the H2 is purified to a level suitable for a H2 /O2
 Y. Lu et al. / Chemical Engineering Journal 131 (2007) 233–244 235

Fig. 1. Schematic of biomass gasification process for hydrogen production in SCW.

fuel cell (greater than 99.9% with CO concentration less than where R is ideal gas constant, μ0i (T ) the chemical potential of
5 ppm) by pressure swing absorption (PSA). component i in standard state, and fi is partial fugacity of com-
ponent i. fi is calculated by the equation of state proposed by
3. Thermodynamic model and method DUAN [23]. More details of the model were described in our
previous work [18].
3.1. Chemical equilibrium
3.2. Gas–liquid equilibrium
Gibbs free energy of a system, with fixed T and P, can be
expressed as a linear combination of chemical potential of each At gas–liquid equilibrium, the fugacity of component i in gas
component in the system. phase equals to that in liquid phase. At given temperature T and
pressure P, the following equations can be obtained:
n

g= μ j nj (1) g
fi = fil , i = 1, 2, . . . , n (4)
j=1
fi is calculated by Soave–Redlich–Kwong (SRK) equation of
where nj and μj are molar number and chemical potential of state:
component j, respectively.
Equation of element conservation is described as: RT a
P= − (5)
V − b V (V + b)
n

aij nj − bi0 = 0, i = 1, . . . , l (2) where parameter b for mixture is derived from linear mixing
j=1 rule:
c

where aij is molar number of element i in compound j, and bi0 is b= zi bii (6)
the total molar number of element i in the initial reactant. i=1
Gibbs free energy is the minimum when a multicomponent
system reaches chemical equilibrium. Minimizing Gibbs free in which parameter bii is for corresponding pure component
energy of a system, with fixed T and P, is a simple constrained RTci
optimization problem. The constraints can be removed with the bii = 0.08664 (7)
Pci
method of Lagrange multipliers.
Chemical potential of component i (μi ) can be calculated by Parameter a in Eq. (5) for pure component is obtained from
the following expression:
R2 Tci2
aii = 0.4286 [f (Tri )]2 (8)
μi (T, P) = μ0i (T ) + RT lnfi (3) Pci
236 Y. Lu et al. / Chemical Engineering Journal 131 (2007) 233–244

Table 1 Table 2
The Ci constants and critical parameters of pure components Molecular surface area and volume values
Component C1 C2 C3 Tc (K) Pc (MPa) Gas

H2 0.1332 0.0000 0.0000 33.2 1.297 H2 CO CO2 CH4 C2 H4 C2 H6 H2 O

CO 0.5836 0.0000 0.0000 132.9 3.496
Rk 0.8320 2.0940 2.592 2.244 3.1482 3.6044 0.9200
CO2 0.8653 −0.4386 1.3447 304.2 7.376
Qk 1.1410 2.1200 2.522 2.312 2.9700 3.392 1.4000
CH4 0.5472 −0.3992 0.5751 190.6 4.600
C2 H4 0.8479 −0.3421 0.6603 282.4 5.040
C2 H6 0.6853 −0.4284 0.7382 305.4 4.848
H2 O 1.0873 −0.6377 0.6345 647.3 22.048 The residual part in Eq. (12) is represented as
 (i) (i)
lnγiR = υk (lnΓk − lnΓk ) (15)
k
where Tr = T/Tc , f(Tr ) is given by Mathias and Copeman [24]
 √ √ 2 √ 3
1 + C1 (1 − Tri ) + C2 (1 − Tri ) + C3 (1 − Tri ) , Tri < 1
f (Tr ) = √ (9)
1 + C1 (1 − Tri ), Tri > 1
(i)
C1 , C2 and C3 shown in Table 1 were estimated from the Γ k and Γk are activity coefficient of group k and that of
pure-component vapor pressure. group k in molecule i, respectively.
Parameter a for mixture is calculated by modified   
Huron–Vidal second-order (MHV2) mixing rule [25],   Θm Ψkm
lnΓk = Qk 1 − ln Θm Ψmk −  (16)
 c
  c
 m m n Θn Ψnm
 
q1 αmix − zi αii + q2 αmix −
2
zi αii
2
where
i=1 i=1 
j υ m xj
(i)
  Q m Xm
gE
c
 b Θm =  , Xm =   (j) (17)
= + zi ln (10) n Qn Xn j n υn x j
RT bii
i=1
Qk is van der Waals surface area of group k. Values of Qk and
where αmix = a/bRT, αii = aii /bii RT, q1 = −0.478, q2 = −0.0047 Rk are shown in Table 2. Ψ nm in Eq. (16) can be calculated from
and gE is excess Gibbs energy, which is given by
a 
nm
Ψnm = exp − (18)
 c T
gE
= zi lnγi (11) where anm is interaction parameter between groups n and m
RT
i=1 in the modified UNIFAC model. Further, values of all gas–gas
interaction parameters were assigned to be zero. To describe the
where γ i is the activity coefficient of the component i, and γ i ,
temperature dependence of the interaction parameters (anm ), two
is obtained from the modified universal functional activity coef-
terms were used
ficient (UNIFAC) model [26]. Some groups, such as H2 , CO,
CO2 , CH4 , C2 H4 , C2 H6 and H2 O, are defined in the modified anm = anm,1 + anm,2 (T − T0 ) (19)
UNIFAC model.
The activity coefficient is expressed as T0 is a reference temperature (298.15 K). anm,1 and anm,2 [27]
shown in Table 3 were estimated from experimental data.
lnγi = ln γiC + lnγiR (12) The fugacity coefficient of component i is described as
   
fi RT 1 α
The first term on the right-hand side of Eq. (12) represents lnφi = ln = ln + − bii
the combinatorial part of the activity coefficient and the second zi P P(V − b) V −b V +b
   
term refers the residual part. ∂(nα) V +b
In the modified UNIFAC model, the combinatorial part is − ln (20)
∂ni T,nj V
described as

lnγiC = 1 − ϕi + lnϕi (13) Table 3

Modified UNIFAC interaction parameters
where n m anm,1 anm,2 amn,1 amn,2

ri
2/3  (i)
H2 H2 O 1586.0 3.924 949.9 −0.3100
ϕi =  2/3
, ri = υ k Rk (14) CO H2 O 1455.0 −2.906 494.0 0.1390
j xj r j k CO2 H2 O 1067.0 −0.418 226.6 −0.2410
CH4 H2 O 1608.0 −2.059 499.2 −0.2550
(i) C2 H4 H2 O 1354.0 −1.542 346.5 −0.3326
In Eq. (14), Rk is van der Waals volume of group k and υk C2 H6 H2 O 1529.0 −3.081 405.0 0.0930
is the number of structural groups of type k in molecule i.
 Y. Lu et al. / Chemical Engineering Journal 131 (2007) 233–244 237

Table 4 where V is the specific volume of real gas and is calculated by

Deviations between experimental data and predicted results of the Model state equation of SRK (Eq. (5)).
System T (K) P (bar) Number of data (ln K)aver a Ref. Reaction equation of biomass combustion is
points (N)
x y x
H2 –H2 O 311–589 3–138 13 0.05 [28]
CHx Oy (s) + 1 + − O2 (g) → CO2 (g) + H2 O(g)
4 2 2
H2 –CO–H2 O 3–138 311–589 15 0.07 [28] (25)
CH4 –H2 O 323–589 14–169 16 0.06 [29]
CO2 –H2 O 302–477 7–202 8 0.04 [29] And the lower heating value of biomass is given by
N C 
a ( lnK)aver = |lnKj,cal − lnKj,exp /CN|, where Kj is equilib-
i=1 j=1
rium ratios of component j, C is number of components.
LHVbiomass ≈ −Vi Hi[298.15 K,1 atm] (26)

from which the enthalpy of real biomass at reference state can

∂(nα)/∂ni can be calculated from the MHV2 mixing rule (Eq. be calculated.
(10)), using: The entropy of real gas is represented by
∂(nα)
(q1 + 2αq2 ) = q1 αii + q2 (α2 + α2ii ) + lnγi S = S∗ + SR (27)
∂ni
b bii S* is the entropy of ideal gas
+ln + −1 (21)  T ∗
bii b Cp P
S ∗ = So∗ + dT − R ln (28)
T0 T Po
Table 4 shows deviation between model predicted and exper-
imental data [28,29]. In all cases the average deviation of the where So∗ is entropy at reference state. Residual entropy is given
logarithm of the equilibrium ratios is small, i.e. predicted values by
are in agreement with experimental results.  V   
∂P R V
SR = − dV + R ln (29)
3.3. Enthalpy and entropy of real fluid and biomass ∞ ∂T V V V o

The enthalpy and entropy of water are calculated by a modi-

The enthalpy of real gas is represented by
fied formula based on the data formulation IAPWS 1995 [30]
H = H∗ + HR (22)
H = Ho∗ + [H̄(T, P) − H̄ (298.15 K, 1 atm)] (30)
The first term on the right-hand side of Eq. (22) is the enthalpy
S = So∗ + [S̄(T, P) − S̄ (298.15 K, 1 atm)] (31)
of ideal gas
 T where Ho∗ and So∗ are the enthalpy and entropy of water at refer-
H ∗ = Ho∗ + Cp∗ dT (23) ence state, respectively, H̄(T, P) and S̄(T, P) are the enthalpy and
T0
entropy of water at T, P given by the data formulation IAPWS
where Ho∗
is enthalpy at reference state, and Cp∗ is specific heat 1995, respectively.
capacity of ideal gas which is a function of temperature. Table 5
displays the enthalpy and entropy at the reference state, and heat 3.4. Exergy of real fluid and biomass
capacity of ideal gases.
The second term on the right-hand side of Eq. (22) is residual Exergy of a material stream includes chemical exergy (EX,c )
enthalpy, which is expressed as and physical exergy (EX,ph ), and total exergy of a material stream
 V    is given as
∂P
HR = T − P dV + RT (Z − 1) (24)
∞ ∂T V EX = EX,c + EX,ph (32)

Table 5
Enthalpy and entropy at the reference state, and heat capacity of an ideal gas
Cp∗ = A + BT + BT 2 + BT 3 (J/(mol K)) Ho∗ (kJ/mol) So∗ (J/(mol K))

A B × 102 C × 105 D × 108

H2 29.062 −0.82 0.199 0.0 0.0 130.59

O2 25.594 13.251 −0.421 0.0 0.0 205.03
CO 26.537 7.683 −0.1172 0.0 −110.52 197.91
CO2 26.748 42.258 −1.425 0.0 −393.51 213.64
CH4 25.36 1.687 7.131 −4.084 −74.85 186.19
C2 H4 3.798 15.65 −8.346 1.756 52.28 219.45
C2 H6 8.181 16.147 −4.007 −0.694 −84.67 229.49
H2 O(l) −285.84 69.94
238 Y. Lu et al. / Chemical Engineering Journal 131 (2007) 233–244
 
Physical exergy is resulted from the difference in temper- where in Ei and out Ek are the exergy flow of all input and
ature and pressure between operation condition and reference output material streams, respectively. I is the internal exergy loss
environmental condition. The physical exergy of a pure com- due to irreversibility.
pound in a mixture can be easily calculated using enthalpy and Energy efficiency and energy loss ratio are defined, respec-
entropy data for the given system: tively, as
EX,ph = (H − Ho ) − T0 (S − So ) (33) En,out
ηen = × 100% (41)
where H and S are enthalpy and entropy of a system at given En,in
temperature and pressure, and Ho and So are the values of these and
functions at the environmental temperature and pressure. Value
of the physical exergy of biomass is assumed to be zero in this (En,loss )i (En,loss )i
σen =  = × 100% (42)
paper. (En,loss )i En,loss
The physical exergy of gas mixture is derived from the con-
Similarly, exergy efficiency and exergy loss ratio are defined,
ventional linear mixing rule
 respectively, as
i
EX,ph = yi EX,ph (34) Ex,out
i ηex = × 100% (43)
Ex,in
and the chemical exergy of gas mixture is given by
  and
EX,c = yi ε0,i + RT0 yi lnyi (35)
(Ex,loss )i (Ex,loss )i
i i σex =  = × 100% (44)
(Ex,loss )i Ex,loss
where ε0,i is the standard chemical exergy of a pure chemi-
cal compound i. ε0,i is equal to the maximum amount of work
4. Results and discussion
obtainable when a compound is brought from the environmen-
tal state, characterized by the environmental temperature T0 and
4.1. Chemical equilibrium in the reactor
environmental pressure P0 , to the dead state, characterized by
the same environmental conditions of temperature and pressure,
When the calculation is conducted, H2 , CH4 , CO2 , CO, C2 H4 ,
but also by the concentration of reference substances in standard
C2 H6 , H2 O and solid carbon are considered in the chemical
environment. A standard environment model given by Szargut
equilibrium model. Wood sawdust, which is represented by a
was used in this paper.
general formula of CH1.35 O0.617 , is used as a typical gasification
The chemical exergy of biomass is hard to define and there-
material in this paper. The predicted results show that the yields
fore, the statistical correlation of Szargut and Styrylska was used
of C2 H4 , C2 H6 and solid carbon are less than 10−5 mol/kg dry
[31]:
biomass, much less than other species, so these two species can
ε0,biomass = βLHVbiomass (36)
where LHVbiomass is the lower heating value, and
1.0412 + 0.2160(ZH /ZC ) − 0.2499ZO /ZC [1 + 0.7884ZH /ZC ] + 0.0450ZN /ZC
β= (37)
1 − 0.3035ZO /ZC
ZO , ZC , ZH and ZN are the weight fractions of oxygen, carbon, be neglected and the product gas were considered to include
hydrogen and nitrogen, respectively, in the biomass. mainly H2 , CH4 , CO and CO2 .
Heating is needed in the biomass gasification system, and the
thermal exergy is
   4.1.1. Effects of temperature on chemical equilibrium
T0 Fig. 2 shows the variation of equilibrium gas yields at with
EX,Q = 1− δQ (38)
T temperature ranging from 673 K to 1073 K. The yields of H2 and
CO2 increase with the increasing temperature, while the yield
3.5. Energy and exergy efficiencies of CH4 decreases sharply. Therefore, higher temperature favors
H2 production. CO yield is very small, about 10−3 mol/kg dry
The equation of energy conservation for stable material biomass. As temperature increases from 673 K to 1073 K, the
stream is given by CO yield increases first and then decreases. CO yield reaches
maximum at about 823 K. Note that H2 yield increases very
Q̇ = Ẇ + Ḣ (39)
slowly at high temperature and become nearly unchanged since
The equation of exergy balance is about 923 K, and then gas product contains almost only H2 and
  CO2 . The maximal H2 yield of 88.623 mol/kg dry biomass is
Ei = Ek + I (40) obtained. Consequently, from the viewpoint of thermodynamics,
in out further increase of temperature is unnecessary for H2 production.
 Y. Lu et al. / Chemical Engineering Journal 131 (2007) 233–244 239

biomass is fed with the flowrate of 2 kg/h and the concentra-

tion of 30 wt%, and the flow rate of high temperature water
is 10 kg/h, the concentration of biomass in reactor is actually
5 wt%. As a result, the biomass feedstock at high concentration
can be gasified at lower temperature.

4.1.3. Effects of oxygen addition on chemical equilibrium

Biomass gasification in SCW is an endothermic reaction, so
the process requires extra heat to drive the chemical reaction.
Generally, heat is supplied to the reactor from external heat
resource, but exergy and energy loss will be caused due to heat
transfer. Also, fast heating and complete gasification of biomass
are difficult to achieve with external heating. Considering that
biomass can be oxidized to be CO2 and H2 O by oxygen in SCW
Fig. 2. Equilibrium gas yields in the reactor as a function of temperature for
and generate heat, oxygen can be added to the reactor to real-
biomass gasification at 25 MPa with 5 wt% dry biomass content. ize internal heat supply for the biomass gasification reaction.
High heat transfer efficiency and gasification efficiency are also
4.1.2. Effects of feedstock concentration on chemical gained.
equilibrium Some experimental studies on partial oxidation of biomass
Fig. 3 displays the effects of feedstock concentration in the in SCW for hydrogen production have been conducted recently
reactor on equilibrium gas yield at 873 K and 25 MPa. As shown [32,33]. Here, the effects of oxygen addition on the equilibrium
in Fig. 3, H2 and CO2 yields decrease gradually with increas- gas yield are investigated by thermodynamic calculation. Equiv-
ing feedstock concentration, while CH4 and CO yields increase. alence ratio (ER) represents the amount of oxygen addition, and
The product gas consists of mainly H2 and CO2 when biomass is defined as
feedstock concentration is low, but the CH4 yield becomes weight oxygen/weight dry biomass
ER = (45)
remarkable when the feedstock concentration is high. The objec- stoichiometric/biomass ratio
tive of biomass gasification in SCW is to produce H2 , so it is
Fig. 4 shows the effects of oxygen addition on equilibrium
better for the system to operate with high biomass feedstock con-
gas yield at 25 MPa with various temperatures. It can be seen
centration and produce as little CH4 as possible. Since CH4 yield
that H2 and CH4 yields decrease with the increasing amounts
decreases with increasing temperature, as mentioned above, high
of oxygen addition under the same temperature and pressure.
reaction temperature seems necessary to achieve high H2 yield
Fig. 5 displays the variation of external energy requirement in
with high feedstock concentration. To realize the effective gasi-
the gasification reactor with oxygen ER for equilibrium state at
fication of biomass with high feedstock concentration at lower
873 K, 25 MPa. The external energy requirement decreases with
temperature, a special layout of the experimental system was
increasing oxygen ER. But even when ER is to 0.5, which means
proposed. As shown in Fig. 1, high concentration biomass feed-
only a half of H2 yield can be obtained compared with that of no
stock mixes with high temperature water at the inlet of reactor
oxygen addition, the external energy requirement is still greater
and gets diluted, while water and heat were recycled. With
than 1000 W. As a result, to realize full self-heating, larger ER
this layout, biomass is fed to the system at high concentration,
value is needed and leads to even less hydrogen production. In
while gasified in the reactor at low concentration. For example,
fact, most heat generated from biomass oxidation was absorbed
by the water but not used effectively for the reaction. In the
system proposed by Hong et al. [32], auxiliary fuel, such as waste
ethanol, is added to heat the reactor internally by its oxidation
reaction and high hydrogen yield is obtained.

4.2. Gas–liquid phase equilibrium in high-pressure

separator

It is assumed that chemical equilibrium is reached when

biomass feedstock with concentration of 5 wt% is gasified in
the reactor at 873 K, 25 MPa, and then reaction products enter
the high-pressure separator.

4.2.1. Effects of pressure on gas–liquid equilibrium

Fig. 6 displays the effects of pressure in SPE1 on hydrogen
Fig. 3. Equilibrium gas yields in the reactor as a function of dry biomass content recovery ratio (defined as the amount of hydrogen in gas phase
for biomass gasification at 25 MPa, 873 K. of SPE1/total amount of hydrogen in the product gas), gas com-
240 Y. Lu et al. / Chemical Engineering Journal 131 (2007) 233–244

Fig. 6. Effects of operation pressure in SPE1 on separating H2 from CO2 at

298 K: (a) gas composition and hydrogen recovery ratio in gas phase in SPE1,
(b) gas composition in liquid phase in SPE1.
Fig. 4. The equilibrium hydrogen yield (a), methane yield (b) as a function of
reaction temperature with different oxygen equivalence ratio. The dry biomass
content in the reactor is 5 wt% and the reaction pressure is 25 MPa.
position in gas phase and liquid phase in SPE1 at 298 K. It is
shown in Fig. 6(a) that the molar fraction of hydrogen in the gas
phase increases from 65.56% to 92.41% and the molar fraction
of CO2 decreases sharply from 33.11% to 6.12% with the pres-
sure in SPE1 increasing from 0.1 MPa to 30 MPa. The Henry
constants of CO and CH4 are all greater than that of H2 , so most
of CO and CH4 leave the high-pressure separator with the gas
phase stream and contaminate the H2 . It is can be also seen that
hydrogen recovery ratio decreases and the molar fraction of CH4
has a little tendency to increase with the increasing pressure.
Fig. 6(b) shows that the molar fraction of CO2 in the liquid
phase decreases and the molar fraction of H2 increases with
the pressure in SPE1. Combination of Fig. 6(a and b) suggests
that increasing the pressure in SPE1 favors the purity of H2
in the gas phase but decreases the hydrogen recovery ratio, so
appropriate operation pressure of SPE1 must be selected. The
predicted results show that H2 of 82.45% and recovery ratio of
88.15% are obtained at 15 MPa, 298 K.

Fig. 5. The external energy requirement in the gasification reactor as a function

4.2.2. Effects of temperature on gas–liquid equilibrium
of oxygen equivalence ratio. It is assumed that the temperature and pressure of
biomass gasification system are 873 K and 25 MPa, respectively. The mass flow Fig. 7 displays the effects of the SPE1 operation temperature
rate of biomass feedstock (pump 1 in Fig. 1) is 2.0 kg/h at 298 K and feedstock on the high-pressure separation process with the SPE1operation
concentration is 30 wt%. The mass flow rate of the pure water stream (pump 2 pressure of 15 MPa. As shown in Fig. 7(a), as operation temper-
in Fig. 1) inputting the reactor is 10 kg/h and the temperature is 873 K. ature increases, the molar fraction of H2 in gas phase decreases
 Y. Lu et al. / Chemical Engineering Journal 131 (2007) 233–244 241

Fig. 8. Effects of water recycle ratio on gas composition and hydrogen recov-
ery ratio in gas phase in SPE1. Operation pressure is 15 MPa and operation
temperature is 298 K.

recycled ratio increases. Note that H2 molar fraction increases

more and more slowly because most CO2 is dissolved in water
when water recycled ratio is low. Since the amount of H2 dis-
solved in water increases with increasing water recycled ratio,
higher water recycled ratio leads to lower hydrogen recovery.
Consequently, it is not necessary to add extra water into SPE1
when the gas and water ratio is small.

4.3. Energy and exergy analysis of the whole system

To conduct the energy and exergy analysis of the system, a set

of typical operating parameters were chosen for the calculation.
Fig. 7. Effects of operation temperature in SPE1 on separating H2 from CO2 :
(a) gas composition and hydrogen recovery ratio in gas phase in SPE1, (b) gas
Operating temperature and pressure of the biomass gasification
composition in liquid phase in SPE1. system are 873 K and 25 MPa, respectively. The mass flow rate of
biomass feedstock (pump 1 in Fig. 1) is 2.0 kg/h, and feedstock
concentration is 30 wt%. The mass flow rate of water stream
while the molar fraction of CH4 and CO2 increase, and the hydro- (pump 2 in Fig. 1) is 10 kg/h and the temperature is 873 K.
gen recovery ratio also increases. Purity of H2 in the gas phase It is assumed that external heat resource is used to meet heat
is 86.24% at 283 K and 75.7% at 333 K, respectively. Fig. 7(b) requirement of the system, that biomass gasification in the reac-
shows the variation of gas composition in liquid phase with oper- tor reaches chemical equilibrium and that heat transfer efficiency
ation temperature. Note that the molar fraction of CO2 increases of the reactor equals to that of the preheater.
and that of H2 decreases with increasing temperature. As a result,
proper operation temperature of SPE1 should be selected to con- 4.3.1. Energy and exergy losses of the system
sider both H2 purity and hydrogen recovery ratio. H2 and CH4 in Table 6 displays the results of energy and exergy analysis of
the liquid phase can be separated in SPE2 and combust with oxy- the biomass gasification system. It is shown that energy effi-
gen to produce heat, which can be recycled for the gasification ciency of the system is 44.21% and the exergy efficiency is
system to reduce external heat input. 42.26% under the calculation conditions. Energy loss is caused
mainly by heat transfer and the loss from heat exchanger, cooler,
4.2.3. Effect of water recycled ratio on gas–liquid preheater and reactor takes up 94.67% of the total. Energy loss
equilibrium from heat exchanger is the largest part and that from cooler is the
The amount of CO2 dissolved in water is limited, therefore, second. Exergy loss represents the irreversibility of the system.
extra water need to be added into SPE1 when the amount of Exergy loss of the biomass gasification system is caused mainly
CO2 is large. Water recycled ratio (defined as the mass flow from reactor, heat exchanger and preheater. Exergy loss from
rate of pump 3 in Fig. 1/the total mass flow rate of pump 1 these three units takes up 81.49% of the total, with that from
and pump 2) is another important operation parameter of SPE1. reactor taking up 36.88%, heat exchanger 32.01% and preheater
Fig. 8 displays the effects of water recycled ratio on gas compo- 12.6%. Exergy loss of the reactor is resulted from the irreversibil-
sition of SPE1 at 15 MPa, 298 K. It can be seen that the molar ity of both chemical reaction and heat transfer, while that of heat
fraction of H2 increases, the molar fraction of CO2 decreases, exchanger and preheater is only from heat transfer irreversibil-
while the hydrogen recovery ratio decreases sharply as water ity. Therefore, heat transfer efficiency of the biomass gasification
242 Y. Lu et al. / Chemical Engineering Journal 131 (2007) 233–244

Table 6
Exergy and energy analysis of hydrogen production by biomass gasification in
SCW
Ex,loss (W) σ ex (%) En,loss (W) σ en (%)

Pump 1a 22.62 0.44 23.40 0.43

Reactorb 1882.84 36.88 971.01 17.90
Heater exchangerc 1634.21 32.01 2581.77 47.61
Preheater 643.18 12.60 534.66 9.86
Pump 2 161.56 3.16 177.12 3.27
Cooler 100.71 1.97 1046.59 19.30
Valve 1 83.31 1.63 26.76 0.49
Valve 2 97.25 1.90 0.44 0.01
SEP1 + SEP2d 187.65 3.68 23.75 0.44
PSAe 292.23 5.72 37.69 0.69
Σ 5105.56 100 5423.19 100
ηex (%) 42.46
ηen (%) 44.21
a The energy efficiency of all high pumps is 30%, the temperature of the initial
biomass feedstock is 298 K, and LHV of biomass is 18425.97 kJ/kg.
b The heat resource temperature of reactor and preheater is 1273 K, and the Fig. 10. Effects of heat transfer efficiency of reactor and preheater on the energy
conversion efficiency of gasification system. The heat transfer efficiency of heat
heat transfer efficiency is 75%.
c The heat transfer efficiency is 75%, and the high temperature fluid is cooled exchanger is 75%, and the temperature of external heat resource is 1273 K.
to the temperature of 373 K.
d The operation temperature and pressure are 298 K and 15 MPa, respectively.
ciency in the heat exchanger. The increasing tendency is even
The water recycled ratio is 0, and the gas compositions of gas phase and liquid
more obvious with higher heat transfer efficiency in the heat
phase in high-pressure separator are calculated by gas–liquid equilibrium model.
e The feed pressure of PSA is 4 MPa, the tail gas pressure is 0.1 MPa, exchanger. Fig. 10 displays the effects of heat transfer efficiency
and the hydrogen recovery ratio is 90%. The energy requirement of PSA is in reactor and preheater. It can be seen that increase of heat
4.46 kWh/kmol CO2 . transfer efficiency in reactor and preheater can also results in
the increase of total energy and energy efficiencies.
system is the key to the improvement of total energy and exergy
efficiencies. Effects of heat transfer efficiency of reactor, heat 4.3.3. Effects of heat resource on total energy and exergy
exchanger and preheater were analyzed as follows. efficiencies
Table 7 shows the effect of the temperature of external heat
4.3.2. Effects of heat transfer efficiency on total energy and resource on total energy and exergy efficiency. As is shown
exergy efficiencies in the table, total energy efficiency keeps unchanged with the
Fig. 9 shows the effects of heat transfer efficiency in heat variation of the temperature of external heat resource, because
exchanger on the total energy and exergy efficiencies. As energy input and output of the system does not change. How-
expected, total energy and energy efficiencies of the biomass ever, heat at high temperature includes more exergy than that at
gasification increase with the increase of heat transfer effi- low temperature, so the exergy efficiency decreases with increas-
ing temperature of external heat resource. Solar thermal energy
can be considered to be a heat resource with the temperature of
5800 K. Total exergy efficiency of the gasification system can be
up to 37.8% if solar energy is used as the external heat source.
Solar energy is abundant and clean, therefore, solar energy can
be used as a potential external heat source for the biomass
thermo-chemical gasification process. Coupling of hydrogen
production from biomass gasification in SCW and solar energy
heating will achieve a renewable energy conversion process
indeed.

Table 7
Comparison of energy and exergy efficiency with different heating methodsa
Temperature of heat resource (K)
1073 1273 1473 5800 (solar thermal)

ηen (%) 44.21 44.21 44.21 44.21

ηex (%) 43.75 42.26 41.57 37.80
Fig. 9. Effects of heat transfer efficiency of heat exchanger on the energy con-
version efficiency of gasification system. The heat transfer efficiency of reactor a The heat transfer efficiency of the reactor, preheater and heat exchanger is
and preheater is 75%, and the temperature of external heat resource is 1273 K. 75%.
 Y. Lu et al. / Chemical Engineering Journal 131 (2007) 233–244 243

5. Conclusion [9] T. Yoshida, Y. Matsumura, Gasification of cellulose, xylan and lignin

mixtures in supercritical water, Ind. Eng. Chem. Res. 40 (2001) 5469–
5474.
The thermodynamic models proposed in this paper can be [10] T. Yoshida, Y. Oshima, Y. Matsumura, Gasification of biomass model com-
used to predict product gas composition at chemical equilib- pounds and real biomass in supercritical water, Biomass Bioenergy 26
rium for the gasification in SCW, to assess the H2 purification (2004) 71–78.
at the phase equilibrium in the high-pressure separator, and to [11] H. Schmieder, J. Abeln, N. Boukis, E. Dinjus, A. Kruse, M. Kluth, G.
calculate the energy and exergy efficiencies of the whole sys- Petrich, E. Sadri, M. Schacht, Hydrothermal gasification of biomass and
organic wastes, J. Supercrit. Fluids 17 (2000) 145–153.
tem. The models are able to provide information for design, [12] A. Kruse, T. Henningsen, A. Sinag, J. Pfeiffer, Biomass gasification in
development and optimization of H2 production using biomass supercritical water: influence of the dry matter content and the formation
gasification in SCW. The results from chemical equilibrium of phenols, Ind. Eng. Chem. Res. 42 (2003) 3711–3717.
analysis show that high temperature is in favor of hydrogen pro- [13] A. Sinag, A. Kruse, J. Rathert, Influence of the heating rate and type of
duction. When biomass is gasified with addition of less than catalyst on the formation of key intermediates and on the generation of
gases during hydropyrolysis of glucose in supercritical water in a batch
stoichiometric quantities of oxygen, the reactor can get its heat reactor, Ind. Eng. Chem. Res. 43 (2004) 502–508.
from the in situ exothermic oxidation reaction but H2 and CH4 [14] X.H. Hao, L.J. Guo, X. Mao, X.M. Zhang, X.J. Chen, Hydrogen production
yields decrease with the increasing amounts of oxygen addition. from glucose used as a model compound of biomass gasified in supercritical
Results from phase equilibrium of the high-pressure separator water, J. Int. Hydrogen Energy 28 (2003) 55–64.
show that the increasing pressure and temperature favors the [15] X.H. Hao, L.J. Guo, X.M. Zhang, Y. Guan, Hydrogen production from
catalytic gasification of cellulose in supercritical water, Chem. Eng. J. 110
purity of hydrogen in the gas phase but decreases the hydrogen (2005) 57–65.
recovery ratio, so appropriate operation pressure and tempera- [16] Y.J. Lu, L.J. Guo, C.M. Ji, X.M. Zhang, X.H. Hao, Q.H. Yan, Hydrogen pro-
ture must be selected. Results of energy and exergy analysis of duction by biomass gasification in supercritical water: a parametric study,
the gasification system show that energy and exergy efficiency J. Int. Hydrogen Energy 31 (2006) 822–831.
of the whole system can be more than 40%. The energy loss is [17] H. Tang, K. Kitagawa, Supercritical water gasification of biomass: thermo-
dynamic analysis with direct Gibbs free energy minimization, Chem. Eng.
mainly from heat exchanger, cooler, preheater and reactor, and J. 106 (2005) 261–267.
the exergy loss is from the reactor, heat exchanger and preheater, [18] Q.H. Yan, L.J. Guo, Y.J. Lu, Thermodynamic analysis of hydrogen pro-
which are mostly resulted from heat transfer in the system. Con- duction from biomass gasification in supercritical water, Energy Convers.
sequently, high heat transfer efficiency will result in high energy Manage. 47 (2006) 1515–1528.
and exergy efficiency of the whole system. [19] W. Feng, H.J. Kooi, A.J. Swaan, Biomass conversions in subcritical and
supercritical water: driving force, phase equilibria, and thermodynamic
analysis, Chem. Eng. Process. 43 (2004) 1459–1467.
Acknowledgements [20] Y. Calzavara, C. Joussot-Dubien, G. Boissonnet, S. Sarrade, Evaluation of
biomass gasification in supercritical water process for hydrogen production,
Energy Convers. Manage. 46 (2005) 615–631.
This work is currently supported by the National Key [21] Y. Matsumura, T. Minowa, Fundamental design of a continuous biomass
Project for Basic Research of China through contract No. gasification process using a supercritical water fluidized bed, J. Int. Hydro-
2003CB214500 and the National Natural Science Foundation gen Energy 29 (2004) 701–707.
of China through contract No. 50521604. [22] Y. Yoshida, K. Dowaki, Y. Matsumura, R. Matsuhashi, D. Li, H. Ishitani, H.
Komiyama, Comprehensive comparison of efficiency and CO2 emissions
between biomass energy conversion technologies—position of supercrit-
References ical water gasification in biomass technologies, Biomass Bioenergy 25
(2003) 257–272.
[1] A. Kruse, A. Gawlik, Biomass conversion in water at 330–410 ◦ C [23] Z.H. Duan, N. Moller, J.H. Weare, A general equation of state for super-
and 30–50 MPa: identification of key compounds for indication differ- critical fluid mixture and molecular dynamics simulation of mixture PVTX
ent chemical reaction pathways, Ind. Eng. Chem. Res. 42 (2) (2003) properties, Geochim. Cosmochim. Acta 60 (1996) 1029–1216.
267–279. [24] P.M. Mathias, T.W. Copeman, Extension of the Peng–Robinson equation
[2] P.E. Savage, Organic chemical reactions in supercritical water, Chem. Rev. of state to complex mixture: evaluation of the various forms of the local
99 (1999) 603–621. composition concept, Fluid Phase Equilib. 13 (1983) 91–108.
[3] P.T. Williams, J. Onwudili, Composition of products from the supercritical [25] S. Dahl, M.L. Michelsen, High-pressure vapor–liquid equilibrium with
water gasification of glucose: a model biomass compound, Ind. Eng. Chem. a UMFAC-based equation of state, AIChE J. 36 (1990) 1829–
Res. 44 (2005) 8739–8749. 1836.
[4] D.C. Elliott, L.J. Sealock, E.G. Baker, Chemical processing in high- [26] B.L. Larsen, P. Rasmussen, A. Fredenslund, A modified UNIFAC group-
pressure aqueous environments. 2. Development of catalyst of gasification, contribution model for prediction of phase equilibria and heats and mixing,
Ind. Eng. Chem. Res. 32 (1993) 1542–1548. Ind. Eng. Chem. Res. 26 (1987) 2274–2286.
[5] D.C. Elliott, L.J. Sealock, E.G. Baker, Chemical processing in high- [27] S. Dahl, A. Fredenslund, P. Rasmussen, The MHV2 model: a UNIFAC-
pressure aqueous environments. 3. Batch reactor process development based equation of state model for prediction of gas solubility and vapor-
experiments for organics destruction, Ind. Eng. Chem. Res. 33 (1994) liquid equilibria at low and high pressure, Ind. Eng. Chem. Res. 30 (1991)
558–565. 1936–1945.
[6] D. Yu, M. Aihara, J.M. Antal, Hydrogen production by steam reforming [28] W. Wagner, A. Pruß, The IAPWS formulation 1995 for the thermodynamic
glucose in supercritical water, Energy Fuel 7 (1993) 574–577. properties of ordinary water substance for general and scientific use, J. Phys.
[7] J.M. Antal, S.G. Allen, D. Schulman, X. Xu, R.J. Divilio, Biomass Chem. Ref. Data 31 (2002) 387–535.
gasification in supercritical water, Ind. Eng. Chem. Res. 39 (2000) [29] P.C. Gillespie, G.M. Wilson, Vapor–liquid equilibrium data on water sub-
4044–4053. stitute gas components: N2 –H2 O, H2 –H2 O, CO–H2 O, H2 –CO–H2 O and
[8] T. Minowa, F. Zhen, T. Ogi, Cellulose decomposition in hot-compressed H2 S–H2 O, Gas Processors Association Research Report No. 41, Wilco
water with alkali or nickel catalyst, J. Supercrit. Fluids 13 (1998) 253–259. Research Co., Provo, UT, April 1980.
244 Y. Lu et al. / Chemical Engineering Journal 131 (2007) 233–244

[30] P.C. Gillespie, G.M. Wilson, Vapor–liquid equilibrium and liquid–liquid [32] G.T. Hong, M.H. Spritzer, Supercritical water partial oxidation, in:
equilibrium: H2 O–CH4 , H2 O–CO2 , H2 O–H2 S, H2 O–pentane, Gas Pro- Proceedings of the 2002 U.S. DOE Hydrogen Program Review NREL/CP-
cessors Association Research Report No. 48, Wilco Research Co., Provo, 610-32405, 2002.
UT, April 1982. [33] T. Yoshida, Partial oxidative and catalytic biomass gasification in super-
[31] M.J. Prins, K.J. Ptasinski, F.J.J.G. Janssen, Thermodynamics of gas- critical water: a promising flow reactor system, Ind. Eng. Chem. Res. 43
char reactions: first and second law analysis, Chem. Eng. Sci. 58 (2003) (2004) 4097–4104.
1003–1011.