Energy Conversion and Management 174 (2018) 552–564

Contents lists available at ScienceDirect

Energy Conversion and Management

journal homepage: www.elsevier.com/locate/enconman

Performance analysis of a water–gas shift membrane reactor for integrated T

coal gasification combined cycle plant
⁎
Wataru Yonamine, Sivasakthivel Thangavel, Hidenori Ohashi, Chihiro Fushimi
Department of Chemical Engineering, Tokyo University of Agriculture and Technology, 2-24-16, Naka-cho, Koganei, Tokyo 184–8588, Japan

A R T I C LE I N FO A B S T R A C T

Keywords: In integrated gasification combined cycle (IGCC) systems, the water–gas shift reaction, which promotes the
Water–gas shift membrane reactor conversion of CO present in syngas mixtures into hydrogen, is an important step for hydrogen production.
Hydrogen separation Application of the water–gas shift membrane reactor (WGSMR) in IGCC systems is an attractive option for CO2
Membrane reactor modeling capture compared with conventional methods because of smaller heat loss in gas purification and high CO
Integrated gasification combined cycle
conversion by selectively removing hydrogen from the reaction zone through the membrane. In this study, we
Sensitivity analysis
proposed and evaluated commercial-scale WGSMR models combined with IGCC using reported laboratory-scale
experimental data to optimize their operational parameters. Various models were developed using the Aspen
Plus® Ver. 8.6 process simulator to investigate the impacts of hydrogen separation, pressure loss, and the flow
direction between the sweep gas on the permeate side and syngas on the retentate side on the WGSMR per-
formance with respect to CO conversion, H2 yield, and reactor temperature. The membrane reactor model gave
approximately 20% higher CO conversion than a reactor model without H2 separation and approximately 4%
lower CO conversion than a membrane reactor model with a pressure drop. A counter-current membrane reactor
model gave approximately 2% higher CO conversion than a co-current model; the H2 yield on the permeate side
was 9.3% higher in the counter-current model by separation of H2 through the membrane. A sensitivity analysis
indicated that a high flow rate and low pressure of sweep gas are advantageous for H2 recovery, and high
catalyst loading and high syngas inlet temperature are preferable for higher CO conversion.

1. Introduction continuously removes H2 from the products of the water–gas shift

Coal is the largest energy source for electricity generation world-
wide; however, the burning of coal is one of the major contributors to
climate change. Crises of fossil fuel depletion and environmental de-
gradation mean that efficient utilization of energy and technology im-
provements have become important agendas [1]. Coal gasification is an
important technology for the efficient production of hydrogen and
electricity [2–4]. In recent years, the integrated coal gasification com-
bined-cycle (IGCC) process, which produces H2 and CO in a coal gasifier
and generates electricity by combusting the produced H2 in gas turbines
and fuel cells, is one of the successful technologies to replace conven-
tional coal-fired power plants [5–7]. In particular, IGCC has greater
advantages when low-rank coals and biomass are used as energy
sources [8–10] and has scope for further development [11–16]; how-
ever, owing to the strong demand for reduction of CO2 emissions from
power generation plants, carbon capture and storage (CCS) has become
a required technology in recent years [17].
The water–gas shift membrane reactor (WGSMR), which
(WGS) reaction, is a promising means of syngas-to-hydrogen conversion
with favorable H2 production and/or CCS capabilities [18,19]. The
WGS reaction is represented by Eq. (1) [20]:
CO + H2O ⇌CO2 + H2 ΔH298K = –41.1 kJ/mol (1)
WGSMR is well suited to IGCC applications owing to its simple
operating process [21–24]. In WGSMR, syngas (including sulfur and
steam) flows directly into the reactor and the WGS reaction takes place
on a catalyst that has sulfur tolerance around 300–400 °C. The H2 and
CO2 produced are simultaneously separated by a membrane. The high
pressure and temperature of the syngas are advantageous to producing
hydrogen gas and transporting it across the membrane [25]. The
driving forces to separate H2 in a WGSMR can be further enhanced if a
sweep gas is used. Iulianelli et al. [26] analyzed recent advances in
WGSMR, especially those that can operate at reaction temperatures
(370–400 °C) with high hydrogen selectivity. Extensive laboratory-scale
experimental studies were conducted. Augustine et al. [27] investigated
a palladium alloy (Pd-alloy)-based WGSMR. They observed that such

⁎
Corresponding author.
E-mail address: cfushimi@cc.tuat.ac.jp (C. Fushimi).

https://doi.org/10.1016/j.enconman.2018.08.022
Received 29 May 2018; Received in revised form 6 August 2018; Accepted 6 August 2018
Available online 23 August 2018
0196-8904/ © 2018 Elsevier Ltd. All rights reserved.
W. Yonamine et al. Energy Conversion and Management 174 (2018) 552–564

Nomenclature Ji″N″ permeation flow of component i in module ″N″ in reactor

[kmol/h]
Abbreviations K equilibrium constant [–]
kj thermal conductivity of layer j (stainless tube and in-
ASU air separation unit sulator) [W/(m K)]
EOS equation of state L reactor length [m]
IGCC integrated gasification combined cycle N Nth module [–]
LHV lower heating value Δp pressure drop [Pa]
PFR plug-flow reactor Pe, i hydrogen permeability in component i [mol m/(m2 s Pa)]
S:C steam:carbon ratio pi partial pressure of component i [Pa]
TMOS tetramethoxysilane pi,perm partial pressure of component i on permeate side [Pa]
WGS water–gas shift pi,ret partial pressure of component i on retentate side [Pa]
WGSMR water–gas shift membrane reactor Q heat loss from the reactor [W]
R′ gas constant [J/(mol K)]
Symbols R″N″ reaction rate for hydrogen product [kmol/h]
r1 radius of inner side in reactor [m]
A0 contact area in each membrane module [m2] rj radius of layer j (stainless tube and insulator) [m]
Cp heat capacity of gas [J/(K mol)] T reactor temperature [K]
dp particle diameter of catalyst [m] Th temperature of hot side in reactor [K]
F flow rate of gas [kmol/h] Tl temperature of cold side in reactor [K]
FCO,in flow rate of CO in syngas inlet [kmol/h] U overall heat-transfer coefficient [W/(m2 K)]
FCO,out flow rate of CO in syngas outlet [kmol/h] u superficial gas velocity [m/s]
FH2,perm flow rate of H2 on permeate side [kmol/h]
FH2,ret flow rate of H2 on retentate side [kmol/h] Greek alphabets
FR″N″A hydrogen flow rate in stream R″N″A to Nth module [kmol/
h] αA/B separation coefficient between A and B [–]
ΔH298K enthalpy of reaction [kJ/mol] γ thickness of membrane [m]
hh heat transfer coefficient of hot side [W/(m2 K)] ε voidage of catalyst bed [–]
hl heat transfer coefficient of cold side [W/(m2 K)] µ viscosity of fluid [Pa s]
JH2 hydrogen permeation rate [kmol/h] ρf density of fluid [kg/m3]

membranes are highly hydrogen-selective, which makes them one of
the most popular membrane technologies for replacing the water–gas
shift reactor. Basile et al. [28] studied the performance of a WGSMR in
the temperature range of 320 to 400 °C and reaction pressures up to
0.17 MPa. They concluded that optimal CO conversion could be
achieved with a high degree of H2 removal from the system. Mendes
et al. [29] investigated the behavior of a WGSMR with respect to H2
recovery in the low-temperature range (200 to 300 °C). They found that
higher CO conversion was achieved at lower temperatures owing to the
equilibrium advantage. Different membrane materials for WGSMR have
also been examined [27,30–33]. The performance and economic at-
tractiveness of metallic and non-metallic membrane reactors used for
H2 production from coal plants were investigated by Dolan et al.
[34,35]. Recently, proton ceramic membrane reactors have been pro-
posed and applied in simultaneous steam methane reforming and WGS,
coal steam gasification and WGS [36–38].
Some researchers investigated the WGSMR with an IGCC system
[39–43]. Al-Zareer et al. [44] carried out modelling and performance
analysis of an IGCC system with a WGSMR for hydrogen production.
They investigated the effects of mass flow rate of oxygen, steam, and
coal type. They concluded that the parameters that maximize the hy-
drogen production rate differ from those that maximize overall net
energy efficiency of the IGCC system. Lotric et al. [45] simulated a
WGSMR for an IGCC plant with CO2 capture. They found that greater
than 70% of CO conversion to H2 was achieved in the first 40% of
reactor length and the hydrogen yield was dominated by the permea-
tion rate of hydrogen. Franz et al. [46] evaluated the pre-combustion
CO2 capture in IGCC power plants and reported that efficiency losses in
the WGSMR can be reduced as low as 5.8% for a CO2 degree of se-
paration of 90%. Maas and Scherer [47] investigated the pre-combus-
tion CO2 capture with ceramic membranes in a lignite-fired IGCC plant.
Their results showed that the net energy efficiency of 41.97% in the
IGCC system with a WGSMR and a CO2 separation of 97.6% could be
achieved. Bracht et al. [48] investigated a WGSMR for CO2 removal in
IGCC systems, finding that the net thermal efficiency of the IGCC
system was 42.8% (lower heating value, LHV basis) with a CO2 re-
covery of 80%.
Numerous studies of WGSMR operations have been carried out,
most of which focused on laboratory-scale experiments, computational
fluid dynamics (CFD) [20], and economic analyses [46,47]. The hy-
drodynamics are important for analyzing detailed heat and mass
transport in a WGSMR [20]; however, CFD capability is limited by
massive computational time and cost. The application of CFD to IGCC
has therefore been limited to CO2 capturing units and reactors and has
not yet been conducted to analysis of the entire process of IGCC with a
WGSMR in a commercial scale. Nevertheless, integration of WGSMR
into the commercial IGCC process and its detailed analysis are indis-
pensable for its practical use. Although a process simulator is beneficial
for such whole-process calculation, its application to this technology
has been quite limited.
In the present study, we attempted to model, analyze, and optimize
the performance of a WGSMR in commercial-scale IGCC applications
using a process simulator (Aspen Plus® V8.6) for the first time. Four
models were created to analyze the influences of (i) pressure loss, (ii)
direction of flow (i.e., co-current and counter-current), and (iii) oper-
ating parameters (i.e., sweep gas flow rate, sweep gas pressure, catalyst
loading, and syngas inlet temperature) on CO conversions and H2
yields. In the IGCC process, integration of a WGSMR depends on the
membrane materials (polymeric, ceramic, or metallic); thus, non-me-
tallic ceramic membranes (made from tetramethoxysilane (TOMS)
[49]) were assumed to be used because they are inexpensive compared
with metallic membranes and can operate around 300–400 °C.
2. Methodology and models
The main input data were temperature, pressure, molar flow rate,

553
W. Yonamine et al. Energy Conversion and Management 174 (2018) 552–564

Table 1 Table 2
Comparison of water–gas shift membrane reactor (WGSMR) models. Simulation parameters and composition of syngas [53].
Model Flow direction H2 separation* Pressure drop* Condition Unit Value

1a (base case) Co-current ○ × Total volume flow rate 3

Nm /h 284 800
1b Co-current × × Total mole flow rate kmol/h 13 821
2a Co-current ○ ○ Reactor inlet temperature °C 350, 375, 400
3a Counter-current ○ × Pressure MPa 2.7

Composition
* ○: with; ×: without. CO % 28
CO2 % 4
and composition of syngas. Chemical reaction formulae were integrated H2 % 10
to create the WGSMR model. This study used the extended function as a H2O % 3
N2 % Balance (54.787)
calculator in Aspen Plus® for permeation evaluation of WGSMR using a
COS % 0.0130
program created in Fortran. The features of the four models are listed in HCl % 0.0200
Table 1. The models were compared in three different ways: H2S % 0.0800
NH3 % 0.1000
(i) Case 1: Model 1a vs. Model 1b to investigate the effect of the H2
separation membrane;
(ii) Case 2: Model 1a vs. Model 2a to investigate the effect of pressure 30 mm). The reactor was wrapped with an insulating material, the
drop; thickness of which was 11 mm. Each tubular membrane was 4000 mm
(iii) Case 3: Model 1a vs. Model 3a to investigate the effect of flow in length, 200 mm in inner diameter, and 202 mm in outer diameter.
direction. The reactor had an inlet and outlet (each of 500 mm in length). The
total length of the WGSMR was therefore 5000 mm. The catalyst par-
ticles were packed throughout the reactor, including the inlet and outlet
2.1. Process description
sections. In the co-current model, an inert sweep gas (N2) was fed to the
inlet of the membranes at 500 mm from the WGSMR and permeated H2
Fig. 1 shows a block flow diagram of the IGCC system with a
was recovered with the sweep gas at the outlet of the membranes at
WGSMR. In this study, an air-blown gasifier (Joban-Kyodo thermal
4500 mm from the WGSMR. In the counter-current model, the sweep
power plant No. 10), made by Mitsubishi Heavy Industrial (MHI), was
gas inlet was at 4500 mm and the outlet was at 500 mm.
assumed. The WGSMR was integrated into this IGCC system after a
syngas cooler (SGC). The flow rates, pressure, temperature, and com-
positions of syngas are shown in Table 2. It was assumed that the syngas 2.2. Heat balance
did not contain any fly ash or heavy metals. Hydrogen from the WGSMR
was supplied to the hydrogen gas turbine unit. The sweep gas (i.e., Heat balance of the reactor was given by:
nitrogen) was assumed to be supplied from an air separation unit
(ASU). (Difference in outlet and inlet temperatures [K])
Fig. 2 shows a schematic view of the plug-flow membrane reactor × (Heat capacity of gas [J/(K mol)]) × (Flow rate of gas [mol/s])
considered in this study. Three ceramic tubular membranes were in- = (Heat of reaction [W])−−(Heat loss from the reactor Q [W]). (2)
serted into a fixed bed packed with sour-shift CoMo catalyst particles
within a stainless steel container (inner diameter: 1520 mm; thickness: To calculate heat loss from the reactor Q, the reactor outside

Water
Gas purification
CO2 +
Coal Exhaust gas-2
unit Impurities
Exhaust
Mill and drying gas-1 Steam STACK
process

Water-gas shift
Gasifier Syngas
process cooler(SGC) membrane reactor
(WGSMR)
N2
Superheated
steam H2 + N2
O2
N2
Electricity
Air separation unit
(ASU) Combined cycle
Steam

Air Water

Fig. 1. Schematic diagram of integrated coal gasification combined cycle (IGCC) with water–gas shift membrane reactor (WGSMR).

554
W. Yonamine et al. Energy Conversion and Management 174 (2018) 552–564

Sweep gas (N2) ceramic membrane [51] was given by:

500 4000 500
Pe A0
Syngas JH2 =
γ
(
pH2,ret −pH2,perm ) (10)
Syngas without H2
where A0 is contact area in each membrane module [m ], γ is the 2

thickness of the membrane [m], pH2,ret and pH2,perm are the partial
1520

pressures of hydrogen on the retentate and permeate sides, respectively

[Pa], and Pe is the hydrogen permeability (assumed as 5.5 × 10–6
I.D. 200 [mol m/(m2 s Pa)]). The thickness of membrane was set at 1.1 mm.
H2 + sweep gas(N2)
30 In a ceramic membrane, nitrogen and carbon dioxide permeate
Ceramic membrane
11 slightly through the membrane. This was expressed in terms of a se-
Fixed bed 4000
paration coefficient αA/B [–] [51]:
Pe,A
O.D. 202 γ Pe,A
1520 αA/B = Pe,B
=
I.D. 200 Pe,B
γ (11)
Fig. 2. Schematic view of water–gas shift membrane reactor in co-current flow where Pe,A and Pe,B are permeation coefficients of components A and B,
(units: mm). respectively. The separation coefficients of hydrogen and CO2 against
nitrogen are 1265 and 3.00, respectively [51].
temperature (Tc) was taken as 20 °C and the Eq. (3) was used:
2πL 2.5. Co- and counter-current models for mass balance calculation
Q= (Th−Tc),
1
hh r 1
+
N
∑j=1 ( 1
kj
rj+1
ln r
j )+ 1
hc r N + 1 (3) In this study, the WGSMR model was described by sequential plug-
flow reactors and separators in the Aspen®. The process flow diagrams
where hh and hc are the heat-transfer coefficients of the hot and cold of the WGSMR models for co- and counter-current flows are shown in
sides, respectively [W/(m2 K)], rj is the radius of layer j [m], kj are the Fig. 3(a) and (b), respectively. The WGSMR design was divided into
thermal conductivities of the stainless tube and insulator [W/(m K)], L three parts: the first part was the reaction area from 0 to 500 mm (i.e.,
is the reactor shell length [m], and Th and Tc are the temperatures of the reactor inlet), the second part was the separation–reaction area from
hot and cold sides [K], respectively. 500 to 4500 mm along the reactor length; this was further divided into
By ignoring boundary film resistance, we obtained: forty 100-mm modules, each comprising a plug-flow reactor and hy-
2πL drogen separator, and the third part was the reaction area from 4500 to
Q= (Th−Tc );
1
ln r2 +
r 1 r
ln r 3 5000 mm (i.e., reactor outlet). The WGS reaction in the plug-flow re-
k st kin (4)
1 2
actors (PFR″N″) was first computed and then separation by membrane
Q = U1 A0 (Th−Tc ) = U1 × 2πLr3 × (Th−Tc ); (5) permeation (SEP″N″) was estimated. Here, the combination of PFR″N″
and SEP″N″ is referred to as the Nth module (N = 1–40). The details
1 regarding permeation in the SEP were coded by using Fortran program
U1 = r3 r r3 r
k st
ln r2 + kin
ln r 3 (6) routines. The performances of the PFR and SEP were repeatedly cal-
1 2
culated using the Aspen model. The calculations Table 3 shows the si-
The thermal conductivities in this study were set to 20.1 W/(K m) mulation parameters. The Peng–Robinson equation of state in Aspen
for SUS304 as the reactor shell material and 3.98 W/(K m) for the in- Plus® [52] was used for calculation of the gas properties. The inlet
sulator. The thickness of the SUS shell was assumed to be 30 mm and temperatures of the syngas to the WGSMR were taken as 350, 375, and
that of the insulator was 11 mm. U1 was calculated as: 400 °C because a conventional COS converter in a gas purification
1 system is usually operated around 400 °C. The maximum allowable
U1 =
(760 + 30 + 11) × 10−3 (760 + 30) × 10−3 (760 + 30 + 11) × 10−3 (760 + 30 + 11) × 10−3 syngas outlet temperature was set to 550 °C, owing to the heat tolerance
ln + ln
20.1 760 × 10−3 3.98 790 × 10−3
of the reactor material. The S:C ratio (steam-to-carbon ratio of syngas)
≈ 231.1 W/m2. (7)
was set to 3 to prevent deactivation of the catalyst by carbon deposits
[20].
2.3. Kinetics of water–gas shift reaction To calculate the permeation flow through the membrane in the Nth
module (cf. Fig. 3(a) and (b)), the mass balance of hydrogen was first
The reaction kinetics of the WGSMR with a CoMo catalyst [50] can calculated using Eq. (12):
be expressed as: F R′ ′N + 1′ ′A = F R′ ′N ′ ′A + R N ′ ′−JH 2′ ′N ′ ′ (12)
″
−60.3 0.75 0.31 −0.072 −0.09 ⎛ 1 pCO2 pH2 ⎞ where FR″N″A [kmol/h] is the hydrogen flow rate in stream R″N″A to the
r = 0.008exp⎛ ⎞pCO pH O pCO pH ⎜1− K p p ⎟,
⎝ RT ⎠ 2 2 2
⎝ CO H2 O ⎠ (8) Nth module, R″N″ [kmol/h] is the reaction rate for hydrogen production
in PFR″N″, JH2″N″ [kmol/h] is the H2 permeation rate through the
where R is the gas constant [J/(mol K)], T is the reactor temperature membrane, and FR″N+1″A [kmol/h] is the hydrogen flow rate in stream
[K], K is the equilibrium constant, and pi is the partial pressure of R″N + 1″A to (N + 1)th module. The symbols in Fig. 3 should be noted:
component i [kPa]. The equilibrium constant is given by Eq. (9) [49]: R″N″B represents the gas flow after the WGS reaction before separation;
5693.5 49170 H2″N N–1″ represents gas flow from the (N–1)th to the Nth module on
ln(K ) = + 1.077ln(T ) + 5.44 × 10−4T −1.125 × 10−7T 2−
T T2 the permeate side.
−13.148. (9) In the co-current WGSMR model (Fig. 3(a)), the hydrogen per-
meation rate in the Nth module was modified by combining Eqs. (10)
and (12):
2.4. Membrane permeation calculation
Pe A0
FR N +1 A = FR N A +R N − (pH2 R N B −pH2 N N − 1 )
The hydrogen permeation rate JH2 [mol/s] through the TMOS " " " " " " γ " " " " (13)

555
W. Yonamine et al. Energy Conversion and Management 174 (2018) 552–564

(a) Syngas from

gasifier Nth module
R "N" A

SYNGAS IN

R "N+1" A
WSTM PFR0 (inlet) PFR "N" PFR41 (outlet)
R "N" B

Syngas without H2
Steam SEP "N"
H2 "N"
H2 + sweep gas
Sweep gas H2 "N N-1" H2 "N+1 N" to combine
from ASU cycle

40
(b) Syngas from
gasifier

Nth module
R "N" A

SYNGAS IN

R "N+1" A
WSTM PFR0 (inlet) PFR "N" PFR41 (outlet)
R "N" B

Syngas without
Steam SEP "N" H2
H2 "N"

H2 + sweep gas H2 "N+1 N"

Sweep gas
to combine cycle from ASU

H2 "N N-1"

40

Fig. 3. Water–gas shift membrane reactor (WGSMR) model with (a) co-current and (b) countercurrent flow.

Table 3 where pH2R N B and pH2 NN − 1 are the partial pressures of H2 in streams
Water–gas-shift membrane reactor (WGSMR) parameters. R″N″B and "H"2″N N–1″," respectively.
" By using Eq. (13), the hydrogen
Parameter Value
mass balance was calculated from the first to the 40th module, the sum
of which is shown in Eq. (14):
Catalyst CoMo catalyst (SSC-1) [14]
40 40
Overall heat-transfer coefficient of reactor 231 Pe A0
[W/(m2 K)] FR
" "
40 A = FR 1 A
" "
+ ∑ RN −
γ
∑ (pH2 R N B −pH2 N N − 1 )
N =1 N =1
" " " " (14)
Catalyst loading [kg/(10 cm-shell length)] 65
Operating temperature [°C] 350–550
Steam:carbon (S:C) ratio 3:1 We calculated the N2 and CO2 mass balances by following the same
Sweep gas pressure [MPa] 0.1 procedure.
Sweep gas flow [kmol/h] 5000 In the counter-current WGSMR model (Fig. 3(b)), the hydrogen flow
Number of membranes [–] 3 rate in the stream R″N″A was calculated by using hydrogen partial
Membrane composition TMOS ceramic
pressure, as shown in Eq. (15):
Membrane geometric size Thickness: 1.1 mm
Area: 2.51 m2
Pe A0
Inner diameter (I.D.).: 200 mm FR N +1 A = FR N A + RN − (pH2 R N B −pH2 N + 1 N )
Outer diameter (O.D.):
" " " " γ " " " " (15)
202 mm
where pH2 R N B and pH2 N + 1 N are the partial pressures of H2 in the
" " and H ″N+1
streams R″N″B " " respectively. By using Eq. (15), the
N″,
2

556
W. Yonamine et al. Energy Conversion and Management 174 (2018) 552–564

hydrogen mass balance was calculated from the first to the 40th 3. Results and discussion
module, as given by Eq. (16):
3.1. Reactor performance of co-current model with and without H2
40 40
PA separation in the case of no pressure drop (Case 1)
FR 40 A = FR 1 A + ∑ RN − e 0 ∑ (pH2 R N B −pH2 N + 1 N )
" " " " N =1
γ N =1
" " " " (16)
We first investigated the effect of the hydrogen separation mem-
In both the co- and counter-current models, the Wegstein method brane on the WGSMR performance for the case in which the pressure
was used for the mass balance calculation. drop in the WGSMR was ignored. Fig. 4(a)–(c) show the CO conver-
sions, reactor temperatures, and hydrogen yields on the permeate side,

(a)

550
T=400 (with separation)
T=400 (without separation)
500 T=375 (with separation)
T=375 (without separation)
T=350 (with separation)
Temperature [ ]

450 T=350 (without separation)

400

350

300
(b) 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000
Distance from reactor entrance [mm]
5000

T=350 T=375 T=400

4000
H2 Yield [kmol/h]

3000

2000

1000

0
(c) 500 1000 1500 2000 2500 3000 3500 4000 4500
Distance from reactor entrance [mm]
Fig. 4. Performances of water–gas shift membrane reactor (WGSMR) Models 1a and 1b (Case 1: Effect of H2 separation in co-current model): (a) CO conversion, (b)
reactor temperature, and (c) H2 yield.

557
W. Yonamine et al. Energy Conversion and Management 174 (2018) 552–564

Table 4 FCO, in−FCO, R N+1 A

CO conversion = " "
Parameters for pressure-drop calculation. FCO, in (17)
Parameter Unit Value
where FCO, in [kmol/h] is the flow rate of CO in the syngas inlet and
Particle diameter m 0.1 FCO R N+1"A [kmol/h] is the flow rate of CO in the stream R″N+1″A in
"
Loading density kg/m3 79.08 the Nth module. H2 yield [kmol/h] is the flow rate of H2 on the
Apparent density of catalyst kg/m3 800 permeate side FH 2 N + 1 N , which permeated and was recovered through
Roughness mm 0.0457 " "
the separation membrane. In Model 1b, the permeation rates of hy-
drogen, nitrogen, and carbon dioxide were set to zero. In both models,
respectively, for the co-current WGSMR model with and without a CO conversions increased monotonically because of the progress of the
hydrogen separation membrane (Models 1a and 1b) when the syngas WGS reaction. At each syngas inlet temperature (350, 375, and 400 °C),
inlet temperature was 350, 375, and 400 °C. In Fig. 4(a), the equili- Model 1a gave much higher CO conversions than Model 1b because the
brium CO conversions obtained by “REquil” in ASPEN Plus® at 400, WGS reaction was promoted by withdrawing of the H2. At a syngas inlet
450, and 500 °C are also shown for reference. The CO conversion was temperature of 400 °C, CO conversion reached 87.3% at the outlet in
defined at each point by Eq. (17): Model 1a and 66.4% in Model 1b. A higher syngas inlet temperature
gave higher CO conversion because of the increase in reaction rate.

(a)

550
T=400 (without pressure loss)
T=400 (with pressure loss)
500 T=375 (without pressure loss)
T=375 (with pressure loss)
T=350 (without pressure loss)
Temperature [ ]

T=350 (with pressure loss)

450

400

350

(b) 300
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000
Distance from reactor entrance [mm]
4000
T=400 (without pressure loss)
T=400 (with pressure loss)
T=375 (without pressure loss)
3000 T=375 (with pressure loss)
H2 Yield [kmol/h]

T=350 (without pressure loss)

T=350 (with pressure loss)
2000

1000

(c) 0
500 1000 1500 2000 2500 3000 3500 4000 4500
Distance from reactor entrance [mm]
Fig. 5. Performances of water–gas shift membrane reactor (WGSMR) Models 1a and 2a (Case 2: Effect of pressure drop in co-current model): (a) CO conversion, (b)
reactor temperature, and (c) H2 yield.

558
W. Yonamine et al. Energy Conversion and Management 174 (2018) 552–564

(a)

550
T=400 (countercurrent)
T=400 (cocurrent)
500 T=375 (countercurrent)
T=375 (cocurrent)
T=350 (countercurrent)
Temperature [ ]

450 T=350 (cocurrent)

400

350

(b) 300
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000
Distance from reactor entrance [mm]

(c)

Fig. 6. Performances of water–gas shift membrane reactor (WGSMR) Models 1a and 3a (Case 3: Effect of flow direction, i.e. co-current vs. counter-current): (a) CO
conversion, (b) reactor temperature, and (c) H2 yield.

Here, the equilibrium CO conversions at 400, 450, and 500 °C were
95.2%, 93.0%, and 90.5%, respectively, because conversion of the
products of the exothermic WGS reaction decrease at higher tempera-
tures according to Le Chatelier's principle. The calculated values were
much lower than the equilibrium CO conversions, implying that the
WGS reaction was kinetically controlled under these conditions. The
reaction temperature monotonically increased (cf. Fig. 4(b)) because
the WGS reaction is exothermic. The syngas outlet temperature reached
530 °C in Model 1a and 489 °C in Model 1b when the inlet temperature
was 400 °C. The temperatures of Model 1a were higher than those of
Model 1b at each syngas inlet temperature because the CO conversion
of the former was higher.
As can be seen in Fig. 4(c), the H2 yield on the permeate side in
Model 1a monotonically increased and higher temperature gave a
higher H2 yield. Only H2 yields at 500–4500 mm in Model 1a are shown
because the H2 yield was zero at 0–500 and 4500–5000 mm in Model 1a
and at 0–5000 mm in Model 1b because of no H2 separation in these
areas. At this contact point (i.e., 500 mm), the hydrogen partial pres-
sure on the retentate side was zero and the steepest rise in H2 yield was
observed. In the WGSMR of Model 1a, H2 yield was dominated by the
kinetics of the WGS reaction and permeation rate. The maximum H2
yield at the permeate outlet of 3.78 × 103 kmol/h was obtained at a
syngas inlet temperature of 400 °C (the corresponding temperature was
530 °C, as shown in Fig. 4(b)). It was found that hydrogen separation

559
W. Yonamine et al. Energy Conversion and Management 174 (2018) 552–564

Fig. 7. Hydrogen partial pressures on retentate and permeate sides for (a) co-current flow (Model 1a) and (b) counter-current flow (Model 3a).

Table 5 fraction of fluid) of the catalyst bed [–], dp is the particle diameter of
Parameters for sensitivity analysis. the catalyst [m], u is the superficial gas velocity [m/s], and ρf is the gas
Operating parameter Unit Base case model Parameter range
density [kg/m3]. Table 4 shows the values used to calculate the pres-
(Model 1a) sure drop.
Fig. 5(a)–(c) show the CO conversion, reactor temperature, and H2
Sweep gas flow kmol/h 5000 1100–20,000 yield, respectively, in Models 1a (without pressure drop) and 2a (with
Sweep gas pressure MPa 0.1 0.01–10
Catalytic loading kg/(10 cm- 65 0–250
pressure drop in the fixed bed) for the co-current model. CO conver-
shell length) sions in Model 2a were lower than those of Model 1a at each syngas
Syngas inlet temperature °C 375 325–425 inlet temperature because the hydrogen partial pressure on the re-
to reactor tentate side, which is the driving force for H2 gas permeation (cf. Eq.
(10)), was decreased by the pressure drop across the catalyst bed. In the
case of the syngas inlet temperature of 400 °C in Model 2a, CO con-
and high syngas inlet temperature are very effective to obtain higher
version was 83.8%, the H2 yield was 3.62 × 103 kmol/h, and the syngas
CO conversion and H2 yield.
outlet temperature was 524 °C. The temperature increments of Model
2a were lower than those of Model 1a at each syngas inlet temperature
3.2. Effects of pressure drop on reactor performances (Case 2) due to the slower reaction rate. The effects of pressure loss on the CO
conversion, H2 yield, and reactor temperature were less significant than
The effect of pressure drop on the performance of the co-current that of H2 separation by the membranes.
WGSMR was investigated by comparing the base case model (Model 1a)
with Model 2a (cf. Table 1). The pressure drop across the fixed bed of 3.3. Reactor performances of co- and counter-current flows without
the CoMo catalyst particles (i.e., retentate side) was calculated from pressure drop (Case 3)
Ergun’s law (Eq. (18)):
Δp μu (1−ε )2 ρf (1−ε ) Reactor performances of the co-current (Model 1a) and counter-
= 150 + 1.75 u2 current (Model 3a) WGSMR were analyzed. Fig. 6(a) and (b) show the
L dp2ε 3 dp ε (18)
CO conversions and reactor temperatures, respectively, at each syngas
where Δp is the pressure drop across the fixed bed [Pa], L is bed length inlet temperature. The CO conversion increased monotonically because
[m], µ is the viscosity of the fluid [Pa s], ε is the voidage (or volume of progress of the WGS reaction. Very small differences between the co-

560
W. Yonamine et al.
and counter-current models were observed with respect to these values
at each syngas inlet temperature. At the reactor outlet, the CO con-
versions and reactor temperatures in Model 3a were slightly higher
than those of Model 1a. When the syngas inlet temperature was 400 °C,
CO conversion reached 88.9% and the syngas outlet temperature
reached 533 °C at 5000 mm in the counter-current model. In the co-
current model, CO conversion was 87.3% and the syngas outlet tem-
perature was 530 °C at 5000 mm. Fig. 6(c) shows the H2 yield of Models
1a and 3a. In the counter-current model, the sweep gas was fed at
4500 mm and the direction of the permeate side flow was from
4500 mm to 500 mm. The counter-current model (Models 1a and 3a)
gave a higher H2 yield, although there were only slight differences in
CO conversion and reactor temperature. When the syngas inlet tem-
perature was 400 °C, the H2 yield was 3.78 × 103 kmol/h at 4500 mm
in the co-current model and 4.12 × 103 kmol/h at 500 mm in the
counter-current model.
To analyze the reason for the improved H2 yield, the hydrogen
partial pressures on the permeate and retentate sides were analyzed.
Fig. 7(a) shows those values or the co-current model (Model 1a). The
maximum difference in hydrogen partial pressure was obtained at the
inlet of the membrane (i.e., 500 mm), because the hydrogen partial
pressure increased due to the WGS reaction from 0 to 500 mm. The
Fig. 8. Sensitivity analyses for CO conversion, H2 recovery ratio, and syngas outlet temperature: effects of (a) sweep gas flow, (b) sweep gas pressure, (c) catalyst
loading, and (d) syngas inlet temperature.
561
Energy Conversion and Management 174 (2018) 552–564
hydrogen partial pressure on the permeate side increased mono-
tonically at 500–4500 mm because of separation of hydrogen by the
membranes. On the retentate side, the hydrogen partial pressure de-
creased steeply at 500–1000 mm because of its permeation through the
membranes. The partial pressure then gradually increased at higher
reaction temperatures because of progress of the WGS reaction. In the
counter-current model (Model 3a, Fig. 7(b)), the hydrogen partial
pressure increased monotonically from 4500 to 500 mm on both the
permeate and retentate sides. At the outlet of the permeate, the hy-
drogen partial pressure in the counter-current model (cf. 500 mm in
Fig. 7(b)) was higher than that of the co-current model (cf. 4500 mm in
Fig. 7(a)), indicating a higher permeation rate in the former. The re-
action and permeate rates are considered to be the main factors that
determine hydrogen partial pressure. This WGSMR system was con-
trolled by the reaction kinetics because the permeation rate was much
larger than the reaction rate, i.e., the product hydrogen rapidly per-
meated through the membrane under this condition. The difference in
hydrogen partial pressure in the counter-current flow model was
greater than that of the co-current flow model. The counter-current
model therefore provides a larger pressure difference and enables fast
permeation and reaction.
W. Yonamine et al.
Fig. 8. (continued)
3.4. Optimum operating conditions for water–gas shift membrane reactor
The sensitivity of four parameters (sweep gas flow rate, sweep gas
pressure, catalyst loading, and syngas inlet temperature to the reactor)
on the reactor performances (CO conversion, H2 recovery ratio, and
reactor temperature) were analyzed for Model 1a (base case). The
sweep gas parameters (i.e., sweep gas flow rate and sweep gas pressure)
could affect membrane separation performance, so it was appropriate to
assess the H2 recovery ratio, which is the ratio of the amount of H2
recovered by the membrane to that produced by the WGS reaction. This
parameter was defined by Eq. (19):
FH 2, perm, out
Hydrogen recovery ratio =
FH 2, perm, out + FH 2, ret , out
where FH 2, perm, out and FH 2, ret , out [kmol/h] are the flow rates of H2 at the
outlet of the permeate and retentate sides, respectively.
Table 5 shows the ranges of the parameters used for this sensitivity
analysis. Fig. 8(a) shows the effect of sweep gas flow rate on the reactor
performance. As the sweep gas flow rate increased, CO conversion and
H2 recovery ratio increased due to the increase in the difference in
Energy Conversion and Management 174 (2018) 552–564
hydrogen partial pressure between the retentate and permeate sides,
i.e., the sweep gas flow rate affected the separation performance of the
membrane. Consequently, the H2 recovery ratio varied considerably
compared with the CO conversion.
Fig. 8(b) shows the effect of sweep gas pressure on the reactor
performance. The highest sweep gas pressure gave a lower of CO con-
version and lower H2 recovery ratio. Thus, a lower sweep gas pressure
gave performance improvement of the WGSMR. The H2 recovery ratio
was very sensitive to changes of sweep gas pressure when the pressure
exceeded 0.1 MPa, because a high sweep gas pressure induces high H2
partial pressure on the permeate side (sweep gas flow rate was constant,
regardless of sweep gas pressure), resulting in lower H2 permeation.
Fig. 8(c) shows the effect of catalyst loading on the reactor perfor-
(19) mance. The horizontal axis indicates the catalyst loadings per 10 cm-
shell length. It was observed that a higher catalyst loading mono-
tonically increased the CO conversion and syngas outlet temperature.
The relationship between catalyst loading and pressure loss was a trade-
off. Fig. 8(d) shows the effect of syngas inlet temperature. The higher
syngas inlet temperature gave higher CO conversion and syngas outlet
temperature. The CO conversion and syngas outlet temperature were
562
W. Yonamine et al.
sensitive to changes in catalyst loading and syngas inlet temperature;
however, the H2 recovery ratio was insensitive to these changes because
they do not affect membrane performance.
The results of the sensitivity analysis showed that a large sweep gas
flow rate and lower sweep gas inlet pressure led to a higher H2 recovery
ratio, and a higher syngas inlet temperature and higher catalyst loading
led to higher CO conversion. The high catalyst loading may, however,
lead to higher pressure loss and promotion of catalyst sintering. Careful
reactor design is needed.
4. Conclusions
The performance of a WGSMR model (packed bed of CoMo catalyst
particles (5000 mm in length) and TMOS membrane (4000 mm in
length)) composed of 40 modules of plug-flow reactors and H2 se-
parators to convert CO in syngas produced from a large-scale coal ga-
sifier to H2 were evaluated using Aspen Plus® V.8.6 and Fortran. The
following conclusions were drawn:
1. In a co-current model, when the syngas inlet temperatures were 350,
375, and 400 °C, CO conversions were lower than the equilibrium
values calculated by “REquil” in ASPEN Plus®. This indicated that
the WGS reaction was kinetically controlled under these conditions.
Hydrogen separation and high syngas inlet temperature are pre-
ferred to obtain higher CO conversion.
2. The CO conversions and H2 yield on the permeate side decreased by
3–4% due to the pressure loss in the packed bed (i.e., the retentate
side) at each syngas inlet temperature.
3. Counter-current flow (Model 3a) gave a similar CO conversion and
higher H2 yield on the permeate side than the co-current flow
(Model 1a). The increment in H2 yield on the permeate side was
larger than that of CO conversion because of the larger difference in
H2 partial pressure between the retentate and permeate sides in the
counter-current model.
4. An increase in the sweep gas flow rate and a low sweep gas pressure
increased the H2 recovery ratio; high catalyst loading and high
syngas inlet temperature increased CO conversion. When the reac-
tion kinetics are the rate-determining step in the process, reactor
performance can be improved by increasing syngas inlet tempera-
ture and decreasing the pressure drop. These results help to un-
derstand the optimum range of each parameter for better perfor-
mance of the WGSMR.
Acknowledgments
This study was financially supported by a Grant-in-Aid for Scientific
Research B (Kakenhi Kiban B, 17H03451), a Grant-in-Aid from the
Japan Society for the Promotion of Science (JSPS) Fellows (Kakenhi
Gaikokujin Tokubetsu Kenkyuin, 16F16077), and a Project Support
Grant by Tokyo University of Agriculture and Technology (TUAT). We
thank Kathryn Sole, PhD, from Edanz Group (www.edanzediting.com/
ac) for editing a draft of this manuscript.
References
[1] Tzimas E, Mercier A, Cormos CC, Peteves S. Trade-off in emissions of acid gas
pollutants and of carbon dioxide in fossil fuels power plants with carbon capture.
Energy Policy 2007;35:3991–8.
[2] Tsutsumi A. Advanced coal gasification power generation. Chem Eng Jpn
2011;75:578–81. (in Japanese).
[3] Khartchenko NV, Kharchenko VM. Advanced energy systems. 2nd ed. Boca Raton,
FL: CRC Press; 2014.
[4] Bell DA, Towler BF, Fan M. Coal gasification and its applications. Burlington:
William Andrew; 2011.
[5] Giuffrida A, Romano MC, Lozza G. Efficiency enhancement in IGCC power plants
with air-blown gasification and hot gas clean-up. Energy 2013;53:221–9.
[6] Fushimi C, Dewi WN. Energy efficiency and capital cost estimation of superheated
steam drying processes combined with integrated coal gasification combined cycle.
J Chem Eng Jpn 2015;48:872–80.
563
Energy Conversion and Management 174 (2018) 552–564
[7] Giuffrida A, Romano MC, Lozza GG. Thermodynamic assessment of IGCC power
plants with hot fuel gas desulfurization. Appl Energy 2010;87:3374–83.
[8] Hayashi J-i, Hosokai S, Sonoyama N. Gasification of low-rank solid fuels with
thermochemical energy recuperation for hydrogen production and power genera-
tion. Trans IChemE Part B Process Safy Environ Prot 2006;86:409–19.
[9] Guan G, Fushimi C, Tsutsumi A, Ishizuka M, Matsuda S, Hatano H, et al. High
density circulating fluidized bed gasifier for advanced IGCC/IGFC—advantages and
challenges. Particuology 2010;8:602–6.
[10] Guan G, Fushimi C, Tsutsumi A. Prediction of flow behavior of the riser in a novel
high solids flux circulating fluidized beds for steam gasification of coal or biomass.
Chem Eng J 2010;164:221–9.
[11] Kawabata M, Kurata O, Iki N, Tsutsumi A, Furutani H. Advanced integrated gasi-
fication combined cycle (A-IGCC) by exergy recuperation-technical challenges for
future generations. J Power Technol 2012;92:90–100.
[12] Hoya R, Fushimi C. Thermal efficiency of advanced integrated coal gasification
combined cycle power generation systems with low-temperature gasifier, gas
cleaning and CO2 capturing units. Fuel Process Technol 2017;164:80–91.
[13] Tsutsumi A. Advanced IGCC/IGFC using exergy recuperation technology. Clean
Coal Technol J 2004;11:17–22. (in Japanese).
[14] Kawabata M, Kurata O, Iki N, Tsutsumi A, Furutani H. System modeling of exergy
recuperated IGCC system with pre- and post-combustion CO2 capture. Appl Therm
Eng 2013;54:310–8.
[15] Fushimi C, Ishizuka M, Guan G, Suzuki Y, Norinaga K, Hayashi J-i, Tsutsumi A.
Hydrodynamic behaviors of binary mixture of solids in a combined circulating
fluidized bed with high mass flux. Adv Powder Technol 2014;25:379–88.
[16] Bhattacharya A, Datta A. Effects of supplementary biomass firing on the perfor-
mance of combined cycle power generation: a comparison between NGCC and IGCC
plants. Biomass Bioenergy 2013;54:239–49.
[17] Stolten D, Scherer V, editors. Efficient carbon capture for coal power plants.
Weinheim: Wiley-VCH; 2011.
[18] Kunze C, Spliethoff H. Assessment of oxy-fuel, pre-and post-combustion-based
carbon capture for future IGCC plants. Appl Energy 2012;94:109–16.
[19] Carbo MC, Boon J, Jansen D, Van Dijk HAJ, Dijkstra JW, Van den Brink RW, et al.
Steam demand reduction of water–gas shift reaction in IGCC power plants with pre-
combustion CO2 capture. Int J Greenh Gas Con 2009;3:712–9.
[20] Hla SS, Morpeth LD, Dolan MD. Modelling and experimental studies of a water–gas
shift catalytic membrane reactor. Chem Eng J 2015;276:289–302.
[21] Bhattacharyya D, Turton R, Zitney SE. Acid gas removal from syngas in IGCC plants.
Integrated gasification combined cycle (IGCC) technologies. Oxford: Woodhead
Publishing; 2017. p. 385–418.
[22] Seyitoglu SS, Dincer I, Kilicarslan A. Assessment of an IGCC based trigeneration
system for power, hydrogen and synthesis fuel production. Int J Hydrogen Energy
2016;41:8168–75.
[23] Marano JJ, Ciferino JP. Integration of gas separation membranes with IGCC.
Identifying the right membrane for the right job. Energy Proc 2009;1:361–8.
[24] Chein R, Chen YC, Chung JN. Parametric study of membrane reactors for hydrogen
production via high-temperature water gas shift reaction. Int J Hydrogen Energy
2013;38:2292–305.
[25] Al-Zareer M, Dincer I, Rosen MA. Effects of various gasification parameters and
operating conditions on syngas and hydrogen production. Chem Eng Res Des
2016;115:1–18.
[26] Iulianelli A, Pirola C, Comazzi A, Galli F, Manenti F, Basile A. Water gas shift
membrane reactors. Membrane reactors for energy applications and basic chemical
production. Oxford: Woodhead Publishing; 2015. p. 3–29.
[27] Augustine AS, Ma YH, Kazantzis NK. High pressure palladium membrane reactor for
the high temperature water–gas shift reaction. Int J Hydrogen Energy
2011;36:5350–60.
[28] Basile A, Chiappetta G, Tosti S, Violante V. Experimental and simulation of both Pd
and Pd/Ag for a water gas shift membrane reactor. Sep Purif Technol
2001;25:549–71.
[29] Mendes D, Chibante V, Zheng J-M, Tosti S, Borgognoni F, Mendes A, et al.
Enhancing the production of hydrogen via water-gas shift reaction using Pd-based
membrane reactors. Int J Hydrogen Energy 2010;35:12596–608.
[30] Maroño M, Torreiro Y, Montenegro L, Sánchez J. Lab-scale tests of different ma-
terials for the selection of suitable sorbents for CO2 capture with H2 production in
IGCC processes. Fuel 2014;116:861–70.
[31] Basile A, Chiappetta G, Tosti S, Violante V. Experimental and simulation of both Pd
and Pd/Ag for a water–gas shift membrane reactor. Sep Purif Technol
2001;25:549–71.
[32] Criscuoli A, Basile A, Drioli E. An analysis of the performance of membrane reactors
for the water–gas shift reaction using gas feed mixtures. Catal Today
2000;56:53–64.
[33] Brunetti A, Barbieri G, Drioli E, Lee KH, Sea B, Lee DW. WGS reaction in a mem-
brane reactor using a porous stainless steel supported silica membrane. Chem Eng
Process 2007;46:119–26.
[34] Dolan MD, Donelson R, Dave NC. Performance and economics of a Pd-based planar
WGS membrane reactor for coal gasification. Int J Hydrogen Energy
2010;35:10994–1003.
[35] Dolan MD, Song G, Liang D, Kellam ME, Chandra D, Lamb JH. Hydrogen transport
through VNiM alloy membranes. J Membr Sci 2011;373:14–9.
[36] Kyriakou V, Garagounis I, Vourros A, Marnellos GE, Stoukides M. A protonic
ceramic membrane reactor for the production of hydrogen from coal steam gasifi-
cation. J Membr Sci 2018;553:163–70.
[37] Deibert W, Ivanova ME, Baumann S, Guillon O, Meulenberg WA. Ion-conducting
ceramic membrane reactors for high-temperature applications. J Membr Sci
2017;543:79–97.
W. Yonamine et al.
[38] Wang H, Wang X, Meng B, Tan X, Loh KS, Sunarso J, et al. Perovskite-based mixed
protonic–electronic conducting membranes for hydrogen separation: recent status
and advances. J Ind Eng Chem 2018;60:297–306.
[39] Chiesa P, Kreutz T, Lozza G. CO2 sequestration from IGCC power plants by means of
metallic membranes. J Eng Gas Turb Power 2007;129:123–34.
[40] Sofia D, Giuliano A, Poletto M, Barletta D. Techno-economic analysis of power and
hydrogen co-production by an IGCC plant with CO2 capture based on membrane
technology. Comp Aided Chem Eng 2015;37:1373–8.
[41] Rezvani S, Huang Y, McIlveen-Wright D, Hewitt N, Deb Mondol J. Comparative
assessment of coal fired IGCC systems with CO2 capture using physical absorption,
membrane reactors and chemical looping. Fuel 2009;88:2463–72.
[42] Esmaili E, Mostafavi E, Mahinpey N. Economic assessment of integrated coal ga-
sification combined cycle with sorbent CO2 capture. Appl Energy 2016;169:341–52.
[43] Koc R, Kazantzis NK, Nuttall WJ, Ma YH. Economic assessment of inherently safe
membrane reactor technology options integrated into IGCC power plants. Process
Saf Environ Prot 2012;90:436–50.
[44] Al-Zareer M, Dincer I, Rosen MA. Modeling and performance assessment of a new
integrated gasification combined cycle with a water gas shift membrane reactor for
hydrogen production. Comp Chem Eng 2017;103:275–92.
[45] Lotric A, Sekav_cnik M, Kunze C, Spliethoff H. Simulation of water gas shift
membrane reactor for integrated gasification combined cycle plant with CO2 cap-
ture. Strojni_ski Vestnik- J Mech Eng 2011;57:911–26.
[46] Franz J, Maas P, Scherer V. Economic evaluation of pre-combustion CO2-capture in
564
Energy Conversion and Management 174 (2018) 552–564
IGCC power plants by porous ceramic membranes. Appl Energy 2014;130:532–42.
[47] Maas P, Scherer V. Lignite fired IGCC with ceramic membranes for CO2 separation.
Energy Proc 2014;63:1976–85.
[48] Bracht M, Alderliesten PT, Kloster R, Pruschek R, Haupt G, Xue E, et al. Water gas
shift membrane reactor for CO control in IGCC systems: techno-economic feasibility
study. Energy Convers Manage 1997;38:S159–64.
[49] Twigg MV. Catalyst handbook. Oxford: Wolfe Publishing Ltd.; 1989.
[50] Hla SS, Duffy GJ, Morpeth LD, Cousins A, Roberts DG, Edwards JH. Investigation of
the effect of H2S on the performance of an iron/chrome-based high-temperature
water-gas shift catalyst using simulated coal-derived syngas. Catal Commun
2009;10:967–70.
[51] Han HH, Ryu SH, S-i Nakao, Lee YT. Gas permeation properties and preparation of
porous ceramic membrane by CVD method using siloxane compounds. J Membr Sci
2013;431:72–8.
[52] Aspen Physical Property System. Physical property methods, Aspentech, Version
Number:V8.6; May 2014.
[53] Nunokawa M, Kobayashi M, Nakao Y, Akiho H, Ito S. Development of gas cleaning
system for highly-efficient IGCC – Proposal for scale-up scheme of optimum gas
cleaning system based on generating efficiency analysis. Energy Engineering
Research Laboratory, Rep. No. M00016.2010, < http://criepi.denken.or.jp/jp/
kenkikaku/cgi-bin/report_download.cgi?download_name=M09016&report_cde=
M09016; 2010 > (accessed 6 April 2018).