You are on page 1of 14
<a href=Chemical Engineering Science 170 (2017) 639–652 Contents lists available at ScienceDirect Chemical Engineering Science journal homepage: www.elsevier.com/locate/ces Performance evaluation of biogas upgrading by pressurized water scrubbing via modelling and simulation Christophe E. Wylock , Wojciech M. Budzianowski Université Libre de Bruxelles, Transfers, Interfaces and Processes (TIPs), Av. F. Roosevelt 50, CP 165/67 - 1050 Brussels, Belgium Consulting Services, Poleska 11/37, 51-354 Wrocław, Poland Renewable Energy and Sustainable Development (RESD) Group, Poleska 11/37, 51-354 Wrocław, Poland highlights Modelling of a scrubber-flash-stripper HPWS system for biogas upgrading. A simulation procedure for optimized dimensioning of upgrading plant is proposed. The influence of the water recirculation parameters on the scaling is studied. Investigation of the operating and physico-chemical parameters on the performances. article info Article history: Received 11 July 2016 Received in revised form 12 November 2016 Accepted 9 January 2017 Available online 10 January 2017 Keywords: Biogas Biomethane Water scrubbing Scrubber-flash-stripper configuration Process scaling abstract The water scrubbing of biogas is an efficient, cheap and environmental friendly process to remove CO from biogas in order to upgrade it into biomethane. This work deals with the modelling and simulations of a scrubber-flash-stripper high pressure water scrubbing process, which is currently the most mature technology. The model and its associated simulation procedure can be used to estimate appropriate device sizes as well as to assess the performance of a given plant. Thanks to this tool, an optimized con- figuration is computed for a reference case. A sensitivity analysis is also realized in term of scrubbing effi- ciency and CH slippage. The influence of parameters related to the water recycling as well as the operating parameters is investigated. Moreover, the sensitivity of the model with respect to the solubility and transfer coefficient parameters is analyzed, in order to gain insight on the parameter to focus in the framework of a model calibration by experimental results. 2017 Elsevier Ltd. All rights reserved. 1. Introduction Biogas is a renewable and sustainable fuel rich in methane (CH ) generated through anaerobic digestion of organic matter. Digestible organic materials may come from various sources: ded- icated crops, agricultural residues, household wastes, water treat- ment sludge’s, etc. Biogas is a potential substitute to natural gas which attracts significant attention across countries importing fos- sil fuels. However, such raw biogas is of low pressure, low specific gravity and large specific volume. In addition, it contains a large part of carbon dioxide (CO ), which lowers its calorific value, flame velocity and flammability limits compared to natural gas. Besides, the transport of CO diluted biogas over long distance is more costly compared to e.g. natural gas pipeline transportation. This ⇑ Corresponding author at: Consulting Services, Poleska 11/37, 51-354 Wrocław, Poland. E-mail address: wojciech.budzianowski@gmail.com (W.M. Budzianowski). http://dx.doi.org/10.1016/j.ces.2017.01.012 0009-2509/ 2017 Elsevier Ltd. All rights reserved. explains why the generated biogas is commonly burnt in-situ by combined heat and power (CHP) systems, which enables to convert 35–40% of biogas energy into usable electricity. A part of the cogenerated heat is used for meeting the digester needs but most is often dissipated and wasted because heat users are located far from biogas CHP plants. The dilution of flue gases by combusted CO makes it more difficult to recover energy of flue gases which lowers thermal efficiency of biogas-fired CHP systems. To overcome these issues, several processes have been created over recent years to upgrade biogas into biomethane. Upgrading basically relies on removing most of the CO and potentially other trace compounds from the biogas ( Sun et al., 2015; Abatzoglou and Boivin, 2009; Ryckebosch et al., 2011; Budzianowski, 2012 ). Such biomethane, which can be used directly as automotive fuel or being injected into the natural gas grid, is claimed as an important renewable fuel for Europe ( Thrän et al., 2016 ). The biogas upgrading processes are based on absorption, adsorption, cryogenic or membrane technology ( Beil and Beyrich, " id="pdf-obj-0-5" src="pdf-obj-0-5.jpg">

Contents lists available at ScienceDirect

Chemical Engineering Science

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

<a href=Chemical Engineering Science 170 (2017) 639–652 Contents lists available at ScienceDirect Chemical Engineering Science journal homepage: www.elsevier.com/locate/ces Performance evaluation of biogas upgrading by pressurized water scrubbing via modelling and simulation Christophe E. Wylock , Wojciech M. Budzianowski Université Libre de Bruxelles, Transfers, Interfaces and Processes (TIPs), Av. F. Roosevelt 50, CP 165/67 - 1050 Brussels, Belgium Consulting Services, Poleska 11/37, 51-354 Wrocław, Poland Renewable Energy and Sustainable Development (RESD) Group, Poleska 11/37, 51-354 Wrocław, Poland highlights Modelling of a scrubber-flash-stripper HPWS system for biogas upgrading. A simulation procedure for optimized dimensioning of upgrading plant is proposed. The influence of the water recirculation parameters on the scaling is studied. Investigation of the operating and physico-chemical parameters on the performances. article info Article history: Received 11 July 2016 Received in revised form 12 November 2016 Accepted 9 January 2017 Available online 10 January 2017 Keywords: Biogas Biomethane Water scrubbing Scrubber-flash-stripper configuration Process scaling abstract The water scrubbing of biogas is an efficient, cheap and environmental friendly process to remove CO from biogas in order to upgrade it into biomethane. This work deals with the modelling and simulations of a scrubber-flash-stripper high pressure water scrubbing process, which is currently the most mature technology. The model and its associated simulation procedure can be used to estimate appropriate device sizes as well as to assess the performance of a given plant. Thanks to this tool, an optimized con- figuration is computed for a reference case. A sensitivity analysis is also realized in term of scrubbing effi- ciency and CH slippage. The influence of parameters related to the water recycling as well as the operating parameters is investigated. Moreover, the sensitivity of the model with respect to the solubility and transfer coefficient parameters is analyzed, in order to gain insight on the parameter to focus in the framework of a model calibration by experimental results. 2017 Elsevier Ltd. All rights reserved. 1. Introduction Biogas is a renewable and sustainable fuel rich in methane (CH ) generated through anaerobic digestion of organic matter. Digestible organic materials may come from various sources: ded- icated crops, agricultural residues, household wastes, water treat- ment sludge’s, etc. Biogas is a potential substitute to natural gas which attracts significant attention across countries importing fos- sil fuels. However, such raw biogas is of low pressure, low specific gravity and large specific volume. In addition, it contains a large part of carbon dioxide (CO ), which lowers its calorific value, flame velocity and flammability limits compared to natural gas. Besides, the transport of CO diluted biogas over long distance is more costly compared to e.g. natural gas pipeline transportation. This ⇑ Corresponding author at: Consulting Services, Poleska 11/37, 51-354 Wrocław, Poland. E-mail address: wojciech.budzianowski@gmail.com (W.M. Budzianowski). http://dx.doi.org/10.1016/j.ces.2017.01.012 0009-2509/ 2017 Elsevier Ltd. All rights reserved. explains why the generated biogas is commonly burnt in-situ by combined heat and power (CHP) systems, which enables to convert 35–40% of biogas energy into usable electricity. A part of the cogenerated heat is used for meeting the digester needs but most is often dissipated and wasted because heat users are located far from biogas CHP plants. The dilution of flue gases by combusted CO makes it more difficult to recover energy of flue gases which lowers thermal efficiency of biogas-fired CHP systems. To overcome these issues, several processes have been created over recent years to upgrade biogas into biomethane. Upgrading basically relies on removing most of the CO and potentially other trace compounds from the biogas ( Sun et al., 2015; Abatzoglou and Boivin, 2009; Ryckebosch et al., 2011; Budzianowski, 2012 ). Such biomethane, which can be used directly as automotive fuel or being injected into the natural gas grid, is claimed as an important renewable fuel for Europe ( Thrän et al., 2016 ). The biogas upgrading processes are based on absorption, adsorption, cryogenic or membrane technology ( Beil and Beyrich, " id="pdf-obj-0-16" src="pdf-obj-0-16.jpg">

Performance evaluation of biogas upgrading by pressurized water scrubbing via modelling and simulation

Christophe E. Wylock a , Wojciech M. Budzianowski b ,c ,

<a href=Chemical Engineering Science 170 (2017) 639–652 Contents lists available at ScienceDirect Chemical Engineering Science journal homepage: www.elsevier.com/locate/ces Performance evaluation of biogas upgrading by pressurized water scrubbing via modelling and simulation Christophe E. Wylock , Wojciech M. Budzianowski Université Libre de Bruxelles, Transfers, Interfaces and Processes (TIPs), Av. F. Roosevelt 50, CP 165/67 - 1050 Brussels, Belgium Consulting Services, Poleska 11/37, 51-354 Wrocław, Poland Renewable Energy and Sustainable Development (RESD) Group, Poleska 11/37, 51-354 Wrocław, Poland highlights Modelling of a scrubber-flash-stripper HPWS system for biogas upgrading. A simulation procedure for optimized dimensioning of upgrading plant is proposed. The influence of the water recirculation parameters on the scaling is studied. Investigation of the operating and physico-chemical parameters on the performances. article info Article history: Received 11 July 2016 Received in revised form 12 November 2016 Accepted 9 January 2017 Available online 10 January 2017 Keywords: Biogas Biomethane Water scrubbing Scrubber-flash-stripper configuration Process scaling abstract The water scrubbing of biogas is an efficient, cheap and environmental friendly process to remove CO from biogas in order to upgrade it into biomethane. This work deals with the modelling and simulations of a scrubber-flash-stripper high pressure water scrubbing process, which is currently the most mature technology. The model and its associated simulation procedure can be used to estimate appropriate device sizes as well as to assess the performance of a given plant. Thanks to this tool, an optimized con- figuration is computed for a reference case. A sensitivity analysis is also realized in term of scrubbing effi- ciency and CH slippage. The influence of parameters related to the water recycling as well as the operating parameters is investigated. Moreover, the sensitivity of the model with respect to the solubility and transfer coefficient parameters is analyzed, in order to gain insight on the parameter to focus in the framework of a model calibration by experimental results. 2017 Elsevier Ltd. All rights reserved. 1. Introduction Biogas is a renewable and sustainable fuel rich in methane (CH ) generated through anaerobic digestion of organic matter. Digestible organic materials may come from various sources: ded- icated crops, agricultural residues, household wastes, water treat- ment sludge’s, etc. Biogas is a potential substitute to natural gas which attracts significant attention across countries importing fos- sil fuels. However, such raw biogas is of low pressure, low specific gravity and large specific volume. In addition, it contains a large part of carbon dioxide (CO ), which lowers its calorific value, flame velocity and flammability limits compared to natural gas. Besides, the transport of CO diluted biogas over long distance is more costly compared to e.g. natural gas pipeline transportation. This ⇑ Corresponding author at: Consulting Services, Poleska 11/37, 51-354 Wrocław, Poland. E-mail address: wojciech.budzianowski@gmail.com (W.M. Budzianowski). http://dx.doi.org/10.1016/j.ces.2017.01.012 0009-2509/ 2017 Elsevier Ltd. All rights reserved. explains why the generated biogas is commonly burnt in-situ by combined heat and power (CHP) systems, which enables to convert 35–40% of biogas energy into usable electricity. A part of the cogenerated heat is used for meeting the digester needs but most is often dissipated and wasted because heat users are located far from biogas CHP plants. The dilution of flue gases by combusted CO makes it more difficult to recover energy of flue gases which lowers thermal efficiency of biogas-fired CHP systems. To overcome these issues, several processes have been created over recent years to upgrade biogas into biomethane. Upgrading basically relies on removing most of the CO and potentially other trace compounds from the biogas ( Sun et al., 2015; Abatzoglou and Boivin, 2009; Ryckebosch et al., 2011; Budzianowski, 2012 ). Such biomethane, which can be used directly as automotive fuel or being injected into the natural gas grid, is claimed as an important renewable fuel for Europe ( Thrän et al., 2016 ). The biogas upgrading processes are based on absorption, adsorption, cryogenic or membrane technology ( Beil and Beyrich, " id="pdf-obj-0-30" src="pdf-obj-0-30.jpg">

a Université Libre de Bruxelles, Transfers, Interfaces and Processes (TIPs), Av. F. Roosevelt 50, CP 165/67 - 1050 Brussels, Belgium b Consulting Services, Poleska 11/37, 51-354 Wrocław, Poland c Renewable Energy and Sustainable Development (RESD) Group, Poleska 11/37, 51-354 Wrocław, Poland

highlights

Modelling of a scrubber-flash-stripper HPWS system for biogas upgrading. A simulation procedure for optimized dimensioning of upgrading plant is proposed. The influence of the water recirculation parameters on the scaling is studied. Investigation of the operating and physico-chemical parameters on the performances.

article info

Article history:

Received 11 July 2016 Received in revised form 12 November 2016

Accepted 9 January 2017 Available online 10 January 2017

Keywords:

Biogas Biomethane Water scrubbing Scrubber-flash-stripper configuration Process scaling

abstract

The water scrubbing of biogas is an efficient, cheap and environmental friendly process to remove CO 2 from biogas in order to upgrade it into biomethane. This work deals with the modelling and simulations of a scrubber-flash-stripper high pressure water scrubbing process, which is currently the most mature technology. The model and its associated simulation procedure can be used to estimate appropriate device sizes as well as to assess the performance of a given plant. Thanks to this tool, an optimized con-

figuration is computed for a reference case. A sensitivity analysis is also realized in term of scrubbing effi- ciency and CH 4 slippage. The influence of parameters related to the water recycling as well as the operating parameters is investigated. Moreover, the sensitivity of the model with respect to the solubility and transfer coefficient parameters is analyzed, in order to gain insight on the parameter to focus in the framework of a model calibration by experimental results.

2017 Elsevier Ltd. All rights reserved.

1. Introduction

Biogas is a renewable and sustainable fuel rich in methane (CH 4 ) generated through anaerobic digestion of organic matter. Digestible organic materials may come from various sources: ded- icated crops, agricultural residues, household wastes, water treat- ment sludge’s, etc. Biogas is a potential substitute to natural gas which attracts significant attention across countries importing fos- sil fuels. However, such raw biogas is of low pressure, low specific gravity and large specific volume. In addition, it contains a large part of carbon dioxide (CO 2 ), which lowers its calorific value, flame velocity and flammability limits compared to natural gas. Besides, the transport of CO 2 diluted biogas over long distance is more costly compared to e.g. natural gas pipeline transportation. This

Corresponding author at: Consulting Services, Poleska 11/37, 51-354 Wrocław, Poland. E-mail address: wojciech.budzianowski@gmail.com (W.M. Budzianowski).

0009-2509/ 2017 Elsevier Ltd. All rights reserved.

explains why the generated biogas is commonly burnt in-situ by combined heat and power (CHP) systems, which enables to convert 35–40% of biogas energy into usable electricity. A part of the cogenerated heat is used for meeting the digester needs but most is often dissipated and wasted because heat users are located far from biogas CHP plants. The dilution of flue gases by combusted CO 2 makes it more difficult to recover energy of flue gases which lowers thermal efficiency of biogas-fired CHP systems. To overcome these issues, several processes have been created over recent years to upgrade biogas into biomethane. Upgrading basically relies on removing most of the CO 2 and potentially other trace compounds from the biogas (Sun et al., 2015; Abatzoglou and Boivin, 2009; Ryckebosch et al., 2011; Budzianowski, 2012). Such biomethane, which can be used directly as automotive fuel or being injected into the natural gas grid, is claimed as an important renewable fuel for Europe (Thrän et al., 2016). The biogas upgrading processes are based on absorption, adsorption, cryogenic or membrane technology (Beil and Beyrich,

  • 640 C.E. Wylock, W.M. Budzianowski / Chemical Engineering Science 170 (2017) 639–652

Nomenclature k Latin q ½ A molar concentration, mol/m 3 interfacial area density, m 1 specific
Nomenclature
k
Latin
q
½
A
molar concentration, mol/m 3
interfacial area density, m 1
specific heat capacity, kJ/(kg K)
#
s
gas loss, Nm 3 /h
density, kg/m 3
global mass transfer rate, kg/s
mass transfer rate density, kg/(m s)
C p
X
column cross-section area, m 2
D
diameter, m
D
Subscript and superscript
g
G
gas phase
H
L
liquid phase
H
in
inlet
k
G
out
outlet
k
L
diffusive coefficient, m 2 /s
gravity acceleration, m/s 2
Henry constant, (Pa m 3 )/mol
height, m
transfer coefficient in the gas film, m/s
transfer coefficient in the liquid film, m/s
chemical equilibrium constant
liquid-side global mass transfer coefficient, m/s
CO 2
carbon dioxide
K
CH
methane
K L
4
HCO
bicarbonate ion
M
3
2
CO
carbonate ion
p
molar mass, kg/mol
gas total pressure, Pa
partial pressure of species i, Pa
mass flow rate, kg/s
superficial velocity, m/s
absolute temperature, K
3
G
FLS
flash tank
p
G;i
RAW
raw biogas
Q
SCR
scrubber
u
STR
stripper
T
w
pure water
x
y
z
mass fraction, kg/kg
molar fraction, mol/mol
vertical coordinate, m
r
reaction
abs
absorption
Greek
DE
g
enthalpy variation, kJ/mol
removal efficiency, %

2013). In this work we investigate the gas-liquid absorption by water, called hereafter water scrubbing. Indeed, water is a cheap and environmental friendly solvent for removing CO 2 . It makes use of the higher solubility of CO 2 (and other compounds like H 2 S) than the solubility of CH 4 in water. The ratio of Henry con- stants between CO 2 and CH 4 is about 30 which makes water scrub- bing of CO 2 from biogas relatively efficient. The most mature technology is the scrubber-flash-stripper high pressure water scrubbing (SFS-HPWS), which is sketched in Fig. 1. The CO 2 capture from the biogas is performed in a scrubber, which

640 C.E. Wylock, W.M. Budzianowski / Chemical Engineering Science 170 (2017) 639–652 Nomenclature k Latin q

Fig. 1. Block-diagram of a scrubber-flash-stripper configuration of a pressurized water scrubbing process.

commonly consists of a packed column. Due to the small CO 2 sol- ubility in water, the absorption rate is enhanced by setting its oper- ating pressure between 0.8 and 1.2 MPa, in order to increase CO 2

partial pressure in the biogas. This enables to reach high purity bio- methane at the top of the scrubber. The dissolved CH 4 within the CO 2 -rich water leaving the scrubber is also increased under elevated pressure operation. Therefore, the scrubbing solvent needs to be subsequently separated in a flash tank operating at a reduced pressure of about 0.11–0.2 MPa. The released gas, rich in CH 4 and CO 2 is then mixed with the raw biogas and reinjected into the scrubber. To regenerate the scrubbing water, the CO 2 has to be released. To achieve this solvent regeneration a second low pres- sure packed column is used as a stripper. In this case, CO 2 is des- orbed from water at ambient temperature using air as stripping agent. This considerably reduces energy requirements for solvent regeneration compared to solvents with chemical compounds that strongly bind CO 2 and requires consequently higher stripping temperature. Several scrubber-flash-stripper HPWS plants already exist at the commercial scale. Some examples of such plants as well as their advantages and drawbacks are presented in Budzianowski et al. (2017). Despite, it seems that very few HPWS models are pro- posed in the literature. In addition, the available experimental data are generally sparse and often focused only on the scrubber analysis. To scale appropriately the gas-liquid devices and the associated equipment’s (pump, compressor, etc.) and to identify optimum operating conditions for the process, mathematical modelling and numerical simulation are particularly valuable to reach high- performance SFS-HPWS plant. The industrially validated models may be also useful in process control, including on-line monitoring of plant performance. This work deals with the mathematical mod- elling and the numerical analysis of the SFS-HPWS plant. The appropriate column dimensions are related to the characteristics

of the biogas to upgrade (flow rate,

concentration...),

the targeted

C.E. Wylock, W.M. Budzianowski / Chemical Engineering Science 170 (2017) 639–652

641

CO 2 abatement and the effective transfer rates. The relevant trans- fer phenomena are therefore described at the local scale to reach reliable estimations of diameter, height and operating conditions suitable for the plant scale. Besides the model development, two procedures are proposed, depending on the user’s goal: (1) scale a plant to ensure a given performance or (2) evaluate the performance of a given plant oper- ating with given conditions. This work notably analyses the influ- ence of the parameters related to the scrubbing water recycling on the scrubber and stripper sizes and the sensitivity of the perfor- mance with respect to the operating conditions. It aims to get fur- ther insight into the operations of SFS-HPWS plants by examining the trends and proposing some guidelines. The developed model explicitly considers mass transfer between gas and liquid and hence it is a so called non- equilibrium rate based model which offers improved reliability in simulations of mass transfer in biogas upgrading plants. A new procedure for experimental model validation and calibration was developed. Such models used with present-day computers will be suitable tools for improving resource and energy efficiency of processing industries. Their use will contribute to reducing indus- trial emissions and making existing industries economically viable. Also such models will allow for facilitated introduction of innova- tive processes to the industrial practice at a very low cost. There- fore, the present study can be considered as cutting edge research in process engineering and science.

2. Mathematical modelling

2.1. Scrubber and stripper

It is well known that CO 2 absorption in water is coupled with chemical reactions: the dissolved CO 2 may react with water mole-

Table 1

Literature correlations used in the model.

cules and hydroxide ion (OH ) to form bicarbonate (HCO ) and carbonate (CO ) ions, according to the following scheme (Danckwerts and Sharma, 1966; Cents et al., 2005; Danckwerts, 1970; Vas Bhat et al., 2000) :

2

3

3

CO 2 þ H 2 O HCO þ H þ CO 2 þ OH HCO HCO þ OH CO þ H 2 O H 2 O H þ þ OH

3

3

3

2

3

ð1Þ

ð2Þ

ð3Þ

ð4Þ

Nevertheless, due to the absence of any counter-ion, the pH of scrubbing water is always below 7. The contribution of the reaction (2) is very small due to the negligible OH concentration and the reaction (3) is unlikely to occur. Only reaction (1) may have an effect on the absorption rate. Its reaction rate is given by Danckwerts and Sharma (1966), Cents et al. (2005) by k H 2 O ½CO 2 , with k H 2 O 0.02 s 1 . This enables the estimation of its related Hatta number (Trambouze and Euzen, 2002) by:

Ha ¼

p ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

D L;CO 2 k H 2 O

k L

ð5Þ

where D L;CO 2 is the diffusive coefficient of CO 2 in water and k L is the liquid-side mass transfer coefficient, as defined in Table 1. Ha is then estimated to 0.05, which is significantly smaller than the threshold value 0.3. The reaction is therefore much slower than the mass transport and it does not enhance the transfer. The effect of the reaction rate is therefore neglected for the mass transfer rate computation. Despite the fact that the CH 4 solubility in water is much smaller than CO 2 , the high pressure used in HPWS induces a substantial CH 4 dissolution. The CH 4 transfer may influence in the computa-

Correlation References Packing characteristics for 25 mm unglazed ceramic Intalox saddle d P = 25 10
Correlation
References
Packing characteristics for 25 mm unglazed ceramic Intalox
saddle
d P = 25 10 3 m
Trambouze and Euzen (2002) (Tables 4.5 and 4.8)
A P = 250 m 2 /m 3
F p = 320 m 2 /m 3
r c = 61 10 3 N/m
Henry coefficients
2044
Versteeg and Van Swaaij (1988), in Vas Bhat et al. (2000), Wylock et al.
H CO 2 ¼
3:59 10 7 exp
T
(2011)
H CH 4 ¼ 1:3817 10 8 exp 1600
1
1
Sander (1999)
T
298:15
Diffusive coefficients
2:35 10 6 exp
Versteeg and Van Swaaij (1988), in Vas Bhat et al. (2000)
2119
D L;CO 2 ¼
T L
h
i 1:8769
Guo et al. (2013)
T
1:95 10 9
L
D L;CH 4 ¼
229:8 1
1:75
Reid et al. (1986), in Wylock et al. (2014)
7:84 10 5 T
G
D G;CO 2 ¼
p G
1:5
Cowie and Watt (1971)
D G;CH 4 ¼ 4:29 10 4 T
G
p G
Equilibrium constants
K 1 ¼ exp
12092:1
36:786lnðT Þ þ 235:482
Edwards et al. (1978), in Vas Bhat et al. (2000), Wylock et al. (2011)
q w
T
ð
5839:5
þ22:4773logðT Þ 61:2062
Tsonopoulos (1976), in Vas Bhat et al. (2000), Wylock et al. (2011)
K w ¼ 10
T
Þ q
2
w
Column cross section area and pressure drop
q ffiffiffiffiffi
q ffiffiffiffiffi
q L
Trambouze and Euzen (2002) (Table 4.6)
¼
Q L
q G
p 2 ¼ u u L
G
q G
Q G
q L
0:25
ln p 1;fl ¼ 0:1117 4:012p
2
0:2 u 2
l
q G q W
q L q L
L
p 1 ¼ F g P
l
G
W
Mass transfer parameters
0:1
0:2
0:05
2
2
Onda et al. (1968), in Trambouze and Euzen (2002) (Table 4.7)
0:75
u
u
r c
q L
A Pp
A ¼ A P 1 exp 1:45
L
L
r
u L q L
l L A P
rA P
g
1:33
0:5
0:33
2
u
k
l
u L q L
A P
L
0:0051ðA P d P Þ 0:4
L
L
A P
D L ¼
l L A P
q L D L
g
0:7
0:5
k
l
u G q G
D G ¼ 5:23ðA P d P Þ 2
G
G
A P
l G A P
q G D G
  • 642 C.E. Wylock, W.M. Budzianowski / Chemical Engineering Science 170 (2017) 639–652

tion of the CO 2 molar ratio and it is necessary to evaluate the dis- solved concentration in scrubbing water to control the flash tank efficiency. The gas-liquid transfer of both CO 2 and CH 4 is therefore considered. Their mass transfer rate density, s CO 2 and s CH 4 respec- tively, are computed according to the two-film theory (Whitman, 1923) by the classical following expressions (Trambouze and Euzen, 2002; Coulson and Richardson, 1999):

s CO 2 ¼ M CO 2 A XK L;CO 2 ðH CO 2 p G;CO 2 ½CO 2 L Þ

s CH 4 ¼ M CH 4 A XK L;CH 4 ðH CH 4 p G;CH 4 ½CH 4 L Þ

ð6Þ

ð7Þ

A is the interfacial area density and X is the cross-section area of the column. M i , H i , p G;i , ½i L and K L;i denotes the molar mass, the Henry constant, the gas bulk partial pressure, the liquid bulk molar concentration and the liquid-side global mass transfer coefficient of the species i, respectively. Eqs. (6) and (7) assume that CO 2 and CH 4 concentrations in gas and liquid phases are uniform at a given column height. The global transfer coefficients are computed as classically from

the transfer coefficient in each phase by K L;i ¼

k G;i k L;i

k G;i þH i

k L;i , where k L;i

and k G;i are the transfer coefficient in the liquid and gas films, respectively. It is assumed that water is not transferred (i.e. no water evaporation neither condensation). The governing equations of the scrubber/stripper models derive from the classical one- dimensional global and species mass balances on an infinitesimal element of the column height in both phases. The equations describing the CO 2 and CH 4 mass flow rate evolutions in liquid and gas phases write:

dðQ G x G;CO 2 Þ

dz

¼

Q G

dx

G;CO

2

dz

þ x G;CO 2

dQ G

dz

¼ s CO 2

dðQ G x G;CH 4 Þ

dz

¼ Q G

dx

G;CH

4

dz

þ x G;CH 4

dQ G

dz

¼ s CH 4

dðQ L x L;CO 2 Þ

dz

¼

Q L

dx

L;CO

2

dz

þ x L;CO 2

dQ L

dz

¼ s CO 2

dðQ L x L;CH 4 Þ

dz

¼ Q L

dx

L;CH

4

dz

þ x L;CH 4

dQ L

dz

¼ s CH 4

ð8Þ

ð9Þ

ð10Þ

ð11Þ

z is vertical coordinate in the column, Q denotes the mass flow rate and x the mass fraction. Since the inert part of the gas and liquid flow rates, Q G ð1 x G;CO 2 x G;CH 4 Þ and Q L ð1 x L;CO 2 x L;CH 4 Þ respectively, remains constant, the two following equations can be derived:

Q G

dx G;CO 2 þ dx G;CH 4

dz

dz

þ ð1 x G;CO 2 x G;CH 4 Þ dQ dz G

¼ 0

Q L

dx L;CO 2 þ dx L;CH 4

dz

dz

þ ð1 x L;CO 2 x L;CH 4 Þ dQ dz L

¼ 0

ð12Þ

ð13Þ

By injecting Eq. (12) in Eqs. (8) and (9) and Eq. (13) in Eqs. (10) and (11), they reduce to:

dx CO 2 ;G

dz

¼

s CO 2 ð1 x CO 2 ;G Þ þ s CH 4 x CO 2 ;G

Q G

dx CH 4 ;G

dz

¼

s CH 4 ð1 x CH 4 ;G Þ þ s CO 2 x CH 4 ;G

Q G

dx CO 2 ;L

dz

¼

s CO 2 ð1 x CO 2 ;L Þ þ s CH 4 x CO 2 ;L

Q L

dx CH 4 ;L

dz

¼

s CH 4 ð1 x CH 4 ;L Þ þ s CO 2 x CH 4 ;L

Q L

ð14Þ

ð15Þ

ð16Þ

ð17Þ

where the locale value of Q G and Q L are computed from the inert gas and liquid conservation. It is considered that gas phase follows the Raoult’s law, there- fore a species partial pressure p G;i is directly deduced from its molar fraction y G;i , which is related to its mass fraction by

y G;i ¼

x G;i

M i P j

x

G;j

 
 

M

j

as p G ¼ p G;in þ Dp Dz

. The total pressure of the gas p G is considered as a lin-

ear function of the column height. It is thus computed from the inlet gas pressure and the pressure drop induced by the packing

z.

In the liquid phase, on the one hand, the concentration of the dissolved CH 4 is related its liquid mass fraction by

½CH 4 L ¼ x CH 4 ;L

q L

M

CH 4

, with q L the liquid density. On the other hand,

the dissolved CO 2 takes part to reactions. Although the reactions do not induce any transfer rate enhancement, they may have an influence by modifying the actual CO 2 concentration. ½CO 2 L is not directly deduced from the total dissolved CO 2 mass fraction x L;CO 2 ; , because a small fraction of CO 2 is converted into HCO The dissolved CO 2 is shared according to:

3 .

x L;CO 2

q L

M

CO 2

¼ ½CO 2 L þ ½HCO

3

L

ð18Þ

Considering that the chemical equilibria are reached in the bulk of liquid phase at steady-state anywhere the column, ½CO 2 L is com- puted from x L;CO 2 using the equilibrium expression

K 1 ¼

½HCO

3

L ½H þ L

½CO 2 L

ð19Þ

and the electroneutrality equation neglecting the OH concentra-

tion, leading to:

½CO 2 L ¼ x L;CO 2

q L

1

2

M

CO 2

K 1 þ

s ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi !

K 1

4x

L;CO

2

q

L

M

CO

2

þ K

1

ð20Þ

2.2. Flash tank

The flash tank is modelled as a perfectly mixed continuous device, with one inlet (the liquid leaving the scrubber) and two outlets (the gas returning to the scrubber and the liquid entering the stripper). The following simplifications are assumed:

(i) the separation between the phases is perfect (no liquid is entrained by the gas), (ii) the thermodynamic equilibrium is reached instantaneously between the gas and the liquid, (iii) the

pressure drop induced by the expansion is negligible (p FLS is con-

stant), (iv) the flash temperature is constant (isothermal), (v) the water vapor content in the gas is negligible since the temperature is low.

The governing equations of the proposed model include therefore:

Q

L;in x L;CO 2 ;in

¼

Q L;out x L;CO 2 ;out þ

Q

G;out x G;CO 2 ;out

Q

L;in x L;CH 4 ;in ¼

Q L;out x L;CH 4 ;out þ

Q G;out x G;CH 4 ;out

Q L;in ð1 x L;CO 2 ;in x L;CH 4 ;in Þ ¼ Q L;out ð1 x L;CO 2 ;out x L;CH 4 ;out Þ

x G;CO 2 ;out þ x G;CH 4 ;out ¼ 1

½CO 2 L;out

¼

H CO 2 p G;CO 2 ;out

½CH 4 L;out ¼ H CH 4 p G;CH 4 ;out

ð21Þ

ð22Þ

ð23Þ

ð24Þ

ð25Þ

ð26Þ

C.E. Wylock, W.M. Budzianowski / Chemical Engineering Science 170 (2017) 639–652

643

  • 2.3. Model parameters

The expressions correlating the packing characteristics and the

physico-chemical are found in the literature. They are presented in

Table 1.

It is worth to mention that the correlations used for the transfer

coefficients are the correlations of Onda et al. (1968), which were

developed from experimental data for various liquids and operat-

ing conditions. Even though some recent works like Nock et al.

(2014) tend to show that the more recent correlations of Billet

and Schultes (1993) are slightly better, the correlations of Onda

et al. remain the most commonly used for several industrial pro-

cesses. The accuracy depend on the system studied and for exam-

ple in the system ammonia-carbon dioxide-water-air similar to

that of the current study the correlation of Onda et al. was found

the most accurate from several correlations compared

(Budzianowski and Kozioł, 2006).

3. Simulation procedure

The simulations are performed using the computational soft-

ware Matlab. The model can be used to estimate appropriate

device size for reaching a given performance as well as to compute

the performance of a given plant.

  • 3.1. Plant scaling

This procedure is used if the goal is to determine the required

scrubber and stripper sizes, scrubbing water flow rate and strip-

ping air flow rate for reaching an imposed CO 2 abatement at the

scrubber gas outlet. In this case, three regulation parameters

related to the scrubbing water recycling have to be specified: the

CO 2 mass fraction in the water entering and leaving the scrubber

and in the stripping air leaving the stripper: x CO 2 ;L;in;SCR , x CO 2 ;L;out;SCR

and x CO 2 ;G;out;STR , respectively.

The scaling is realized in two stages. The first one computes the

characteristics of the scrubber in coupling with the flash tank.

Since the gas entering the scrubber is a mixture between the raw

biogas to treat and the gas leaving the flash tank, which depends

on the composition of the liquid leaving the scrubber, several iter-

ations of the scrubber-flash loop are performed to compute the

streams at steady-state. The second stage computes the stripper

characteristics taking into the liquid stream coming from the flash

tank and the CO 2 depletion to reach for reinjection in the scrubber.

3.1.1. Scrubber-flash

The actual gas flow rate entering the scrubber is

Q G;in;SCR ¼ x CO 2 ;G;in;SCR

Q G;RAW þ Q G;out;FLS and its composition is given by

¼

ðQ G;RAW x CO 2 ;G;RAW

þ Q G;out;FLS x CO 2 ;G;out;FLS Þ=Q G;in;SCR and þ Q G;out;FLS x CH 4 ;G;out;FLS Þ=Q G;in;SCR . At the

x CH 4 ;G;in;SCR ¼ ðQ G;RAW x CH 4 ;G;RAW

first iteration, Q G;out;FLS ¼ x CO 2 ;G;out;FLS ¼ x CH 4 ;G;out;FLS ¼ 0 can be used

as initial guesses.

Let x CO 2 ;G;out;SCR be the targeted mass fraction to reach at the gas

scrubber outlet. As an approximation, it can be considered that only

CO 2 is transferred from gas to liquid. Therefore, the global CO 2

scrubbing

rate

can

be

estimated

# CO 2 ;SCR ¼ Q G;in;SCR

1

1 x CO 2 ;G;in;SCR

1 x CO 2 ;G;out;SCR

. By imposing the CO 2 mass fraction in the water

at the inlet and the outlet of the scrubber, x CO 2 ;L;in;SCR and

x CO 2 ;L;out;SCR , respectively, the requested flow rate is estimated by

Q L;in;SCR ¼ # CO 2 ;SCR

1 x CO 2 ;L;out;SCR

x CO 2 ;L;out;SCR x CO 2

;L;in;SCR . Considering that the dissolved

methane in the scrubber water is negligible at the inlet

(x CH 4 ;L;in;SCR 0) and at equilibrium with the entering gas phase

x CH 4 ;L;out;SCR ¼ H CH 4

p G;in;SCR

RT

x CH 4 ;G;in;SCR

q L

x

CO ;G;in;SCR

2

M

CO

2

x

þ

CH ;G;in;SCR

4

M

CH

4

þ

ð1 x

CO ;G;in;SCR

2

x

CH ;G;in;SCR

4

Þ

M

inert

1

A ,

the resulting liquid flow rate leaving the scrubber is then deduced by

Q L;out;SCR ¼ Q L;in;SCR

1 x CO 2 ;L;in;SCR

1 x CO 2 ;L;out;SCR x CH 4 ;L;out;SCR

.

Eqs. (21)–(26) are solved numerically using Q L;in;FLS ¼ Q L;out;SCR ,

x CO 2 ;L;in;FLS ¼ x CO 2 ;L;out;SCR and x CH 4 ;L;in;FLS ¼ x CH 4 ;L;out;SCR , with the routine

fsolve, to

compute Q G;out;FLS , Q L;out;FLS , x CO 2 ;G;out;FLS , x CO 2 ;L;out;FLS ,

x CH 4 ;G;out;FLS and

x CH 4 ;L;out;FLS . The new values of Q G;out;FLS , x CO 2 ;G;out;FLS

and x CH 4 ;G;out;FLS are then used to realize a new iteration. This loop

is realized until the variations between two successive iterations

become slower than a given tolerance.

Once the steady-state is reached, the eventually obtained

Q L;in;SCR is used to calculate the requested scrubber diameter D SCR

for the corresponding superficial velocities in the considered pack-

ing. The diameter is such that the superficial gas velocity corre-

sponds to 60% of the flooding one and it is calculated thank to

the correlations presented in Table 1.

Finally, the minimum scrubber height, H SCR , is estimated by

solving Eqs. (14)–(17) with the ode15s routine, from the bottom

to the top of the scrubber. A much larger height is used as an initial

guess. H SCR is identified as the height matching the outlet criterion,

i.e. x CO 2 ;G;SCR ðH SCR Þ ¼ x CO 2 ;G;out;SCR .

3.1.2. Stripper

Since the composition of the air injected at the stripper inlet are

known (x CO 2 ;G;in;STR ¼ 6 10 4 , corresponding to the molar ratio of

0.04% in the atmospheric air), it is required to impose only the

CO 2 mass fraction to reach at the stripper outlet x CO 2 ;G;out;STR . The

global CO 2 stripping rate to achieve is calculated as

# CO 2 ;STR ¼ Q L;in;SCR x CO 2 ;L;in;SCR Q L;out;FLS x CO 2 ;L;out;FLS . The needed inlet

air flow rate is then deduced by Q G;in;STR ¼ # CO 2 ;STR

1 x CO 2 ;G;out;STR

x CO 2 ;G;out;STR x CO 2 ;G;in;STR

.

Knowing the gas and liquid flow rates, the stripper diameter

and height, D STR and H STR respectively, are estimated using the

same procedure than for the scrubber. The estimated H STR is the

height matching the condition x CO 2 ;L;SCR ðH SCR Þ ¼ x CO 2 ;L;out;SCR , with

x CO 2 ;L;out;SCR ¼ x CO 2 ;L;in;SCR .

3.2. Performance computation

Whether the scrubber and stripper diameters and heights are

known and the scrubbing water and stripping air flow rates are

imposed, or has been estimated by the plant scaling procedure,

the performance of each unit of the plant are computed. Due to

scrubbing water recycling, several iterations of the process are

realized to ensure the steady-state at a given tolerance.

At each iteration, the outlet variables of each unit are computed

according to their inlet variables, with the following sequence.

The gas mixture at the scrubber gas inlet is given

by

Q G;in;SCR ¼ Q G;RAW þ Q G;out;FLS ,

x CO 2 ;G;in;SCR ¼ ðQ G;RAW x CO 2 ;G;RAW þ

Q G;out;FLS x CO 2 ;G;out;FLS Þ=Q G;in;SCR and x CH 4 ;G;in;SCR ¼

ðQ G;RAW x CH 4 ;G;RAW þ

Q G;out;FLS x CH 4 ;G;out;FLS Þ=Q G;in;SCR . The characteristics of the liquid inlet

are given by the imposed Q L;in;SCR and by x CO 2 ;L;in;SCR ¼ x CO 2 ;L;out;STR

and x CH 4 ;L;in;SCR ¼

x CH 4 ;L;out;STR . The evolution profiles of x CO 2 ;G;SCR ,

x CH 4 ;G;SCR , x CO 2 ;L;SCR , x CH 4 ;L;SCR , Q G;SCR and Q L;SCR along the height H SCR

are computed by solving Eqs. (14)–(17) with the bvp4c routine,

enabling the determination of the outlet variables x CO 2 ;G;out;SCR ,

x CH 4 ;G;out;SCR , x CO 2 ;L;out;SCR , x CH 4 ;L;out;SCR , Q G;out;SCR and Q L;out;SCR . At the first iteration, Q G;out;FLS ¼ x CO 2 ;G;out;FLS ¼ x CH 4 ;G;out;FLS ¼ 0 and

x CO 2 ;L;in;SCR ¼ 7 10 7 and x CH 4 ;L;in;SCR ¼ 4 10 12 (corresponding to

the composition at equilibrium with the CO 2 and CH 4 content in

the atmosphere, respectively) can be used as initial guesses. In this

case, a particular procedure is required to ensure the numerical

convergence of the collocation method used in the bvp4c routine.

Indeed, the profile of x CH 4 ;L;SCR with respect to height is non-

  • 644 C.E. Wylock, W.M. Budzianowski / Chemical Engineering Science 170 (2017) 639–652

monotonic, contrary to the other variables. It results from the com-

petition between two opposite contributions to the mass transfer

driving force. On the one hand, the CH 4 molar fraction in the gas

increases with column height, while the total gas pressure

decreases. On the other hand, the dissolved CH 4 concentration in

the entering liquid is almost zero. There is thus a sharp variation

in the dissolved CH 4 concentration at the liquid inlet, which may

disrupt the collocation method. This difficulty is overcome by a

continuation procedure on x CH 4 ;L;in;SCR , which is gradually decreased,

from a value close to equilibrium with x CH 4 ;G;out;SCR to its actual

value. The computed profiles being used as initial guesses for the

next iterations, this procedure is not used anymore after the first

one.

At the flash tank feeded by x CO 2 ;L;in;FLS ¼ x CO 2 ;L;out;SCR ,

x CH 4 ;L;in;FLS ¼ x CH 4 ;L;out;SCR and Q L;in;FLS ¼ Q L;out;SCR , Eqs. (21)–(26) are

solved with the routine fsolve to determine Q G;out;FLS , Q L;out;FLS ,

x CO 2 ;G;out;FLS , x CO 2 ;L;out;FLS , x CH 4 ;G;out;FLS

and x CH 4 ;L;out;FLS

The liquid leaving the flash tank is conveyed to the stripper,

then

x CO 2 ;L;in;STR ¼ x CO 2 ;L;out;FLS ,

x CH 4 ;L;in;STR ¼ x CH 4 ;L;out;FLS

and

Q L;in;STR ¼ Q L;out;FLS . At the stripper gas inlet, Q G;in;STR is and the air

composition is known: x CO 2 ;G;in;STR ¼ 4 10 4 and

x CH 4 ;G;in;STR ¼ 2 10 7 . The evolution profiles of x CO 2 ;G;STR , x CH 4 ;G;STR ,

x CO 2 ;L;STR , x CH 4 ;L;STR , Q G;STR and Q L;STR along the height H STR are com-

puted by solving Eqs. (14)–(17) with the bvp4c routine, enabling

the determination of the outlet variables x CO 2 ;G;out;STR , x CH 4 ;G;out;STR ,

x CO 2 ;L;out;STR , x CH 4 ;L;out;STR , Q G;out;STR and Q L;out;STR .

4. Results and discussion

4.1. Model validation

The scrubbing column model (i.e. mass transfer in a packed col-

umn) validated by using experimental results obtained by different

authors in order to minimise the risk of using incorrect or mislead-

ing data. They are presented in Table 2.

Simulation results are compared to experimental data and aver-

age deviations are calculated. Fig. 2-a compares CO 2 removal from

644 C.E. Wylock, W.M. Budzianowski / Chemical Engineering Science 170 (2017) 639–652 monotonic, contrary to the

Fig. 2. Comparison of experimental data compiled in Table 2 with simulated results

for various biogas upgrading plants: CO 2 removal from biogas (a) and molar fraction

of CO 2 in biomethane (b).

Table 2

Experimental data on collected in the literature for the scrubber model validation.

N

p G;in;SCR

(MPa)

D SCR

(m)

H SCR

(m)

Q G;in;SCR

(m 3 /h)

Q L;in;SCR

(kg/s)

x CO 2 ;G;in;SCR

x CO 2 ;G;out;SCR

p FLS

(MPa)

Ref.

A1

A2

A3

A4

A5

A6

A7

A8

A9

A10

A11

A12

A13

A14

A15

A16

A17

A18

B1

B2

C1

D1

D2

1.0

0.8

1.0

0.3

0.8

0.15

3.0

20

0.15

0.10

0.10

3.5

1.0

1.0

1.0

0.3

242.3

1.00

1.00

1.03

1.03

1.06

1.06

1.08

1.08

1.11

1.11

1.14

1.14

1.17

1.17

1.19

1.19

1.22

1.22

0.42

0.3

13.3

12.06

0.40

0.35

0.40

0.35

0.40

0.35

0.40

0.35

0.40

0.35

0.40

0.35

0.40

0.35

0.40

0.35

0.40

0.35

0.32

0.40

0.452

0.070

0.070

0.065

0.065

0.060

0.060

0.055

0.055

0.050

0.050

0.045

0.045

0.040

0.040

0.035

0.035

0.030

0.030

0.050

0.040

0.108

0.0047

0.0036

Nock et al. (2014), Läntelä and

Luostarinen (2013)

0.3

0.4

Chandra et al. (2012)

Huang et al. (2015)

Götz et al. (2011)

C.E. Wylock, W.M. Budzianowski / Chemical Engineering Science 170 (2017) 639–652

645

biogas for a range of different biogas upgrading plants. The CO 2

removal efficiency from biogas is defined as follows:

g CO 2

x CO 2 ;G;in;SCR x CO 2 ;G;out;SCR

¼

x CO 2 ;G;in;SCR

100%

ð27Þ

Fig. 2-b presents the comparison of experimental and simulated

molar fraction of CO 2 in biomethane (x CO 2 ;G;out;SCR ).

The data presented in Fig. 2-a show a small deviation for CO 2

removal (1.7% in arithmetic average) and those presented in

Fig. 2-b show a limited deviation (14.7%) for CO 2 molar fraction

in biomethane. According to the weak scatter observed in Fig. 2,

it may be concluded that the scrubber model predict rather well

the global behavior of real systems for their given experimental

conditions regarding to the CO 2 separation.

4.2. Reference case

As a reference case, a plant is scaled with the following criteria:

a stream of 250 Nm 3 /h of raw biogas containing 35% molar of CO 2

and 64% molar of CH 4 has to be treated to reduce the CO 2 content

down to 2% molar. The pressure is set to 8 bar at the scrubber inlet,

1.2 bar and at the stripper inlet and 1.1 bar in the flash tank. The

temperature is set to 288.15 K in all devices. As mentioned in Sec-

tion 3.1, three parameters have to be defined in order to close the

model equation system. These are related to the scrubbing water

regeneration and they can be viewed as regulation parameters.

The following values have been selected: x CO 2 ;L;in;SCR ¼ 3:6 10 5 ,

x CO 2 ;L;out;SCR ¼ 8 10 3 and x CO 2 ;G;out;STR ¼ 0:27. The scaling leads to

the following characteristics: D SCR ¼ 0:99 m, H SCR ¼ 5:26 m,

Q L;in;SCR ¼ 23:3 kg/s (81.28 m 3 /h), D STR ¼ 0:99 m, H STR ¼ 8:76 m

and Q G;in;STR ¼ 8:3 10 2 kg/s (231 Nm 3 /h).

C.E. Wylock, W.M. Budzianowski / Chemical Engineering Science 170 (2017) 639–652 645 biogas for a range

Fig. 4. Post-computation liquid temperature evolution with height in the scrubber

(a) and the stripper (b).

The performance of the scaled plant are then evaluates. Using

the identified values, it is found that the biogas treatment gener-

ates 166 Nm 3 /h of biomethane. The flash tank efficiency is close

to 99%, and the methane loss (at the stripper outlet) are about

6 10 2 Nm 3 /h.

C.E. Wylock, W.M. Budzianowski / Chemical Engineering Science 170 (2017) 639–652 645 biogas for a range

Fig. 3. Evolution of x CO 2 ;G , x CH 4 ;G , x CO 2 ;L and x CH 4 ;L with the height in the scrubber (a) and in the stripper (b).

  • 646 C.E. Wylock, W.M. Budzianowski / Chemical Engineering Science 170 (2017) 639–652

The evolution of the CO 2 and CH 4 mass fraction in the gas and in

the liquid of the scrubber and the stripper can also be examined.

They are presented in Fig. 3.

4.3. Thermal effect evaluation

Thanks to the mass transfer rate evaluated in the reference case,

the heat release associated with the CO 2 absorption and its reac-

tion with water can be estimated. It is realized to assess the

assumption validity of isothermal transfer. According to

Taghizadeh et al. (2001), the enthalpy of CO 2 absorption in water

is DE abs ¼ 20:3 kJ/mol. The enthalpy of reaction of CO 2 with water

to form HCO 3 is evaluated to DE r ¼ 46 kJ/mol thanks to the

enthalpy of formation provided in Planck (1903). The CO 2 chemical

conversion rate is deduced from the transfer rate by

s CO 2

M CO 2

½HCO

L ½H þ L

K 1 þ½HCO

3

3 L

þ½H þ L . Considering an adiabatic column and neglecting

the heat transfer to the gas phase, the local liquid temperature

variation due to the heat of absorption and reaction is therefore

calculated by:

C p;L Q L

dT L

dz

¼

s CO 2

M

CO 2

DE abs þ DE r

½HCO

3

L ½H þ L

½H þ L !

K 1 þ ½HCO

3

L þ

ð28Þ

where C p;L is the liquid specific heat capacity.

The liquid temperature evolution with the height in the scrub-

ber and the stripper are presented in Fig. 4. It is observed that the

temperature deviation between inlet and outlet is 2.7 K in the

scrubber and 0.7 K in the stripper. It seems therefore relevant to

consider that the transfer is isothermal in the stripper. In the

scrubber, this assumption is not rigorous but seems nevertheless

reasonable. Moreover, this estimation considers an adiabatic

column, whereas in a real system there would be heat transfer

with the ambient air through the column walls. The actual temper-

ature deviation should be therefore smaller than the estimated

one.

4.4. Influence of regulation parameters

The effect of the three regulation parameters x CO 2 ;L;in;SCR ,

x CO 2 ;L;out;SCR and x CO 2 ;G;out;STR are examined in order to assess their

influence on the scaling. Indeed, these parameters have to be

selected appropriately since they may influence the column sizes

(determining the capital expenditure) and flows to ensure (deter-

mining the operating expenditure) for the same imposed CO 2

abatement.

Variations are then applied to x CO 2 ;L;in;SCR , x CO 2 ;L;out;SCR and

x CO 2 ;G;out;STR to run simulations, keeping the target of 2% molar CO 2

at the scrubber outlet. The results are examined in terms of varia-

tions on D SCR , H SCR , Q L;in;SCR , D STR , H STR and Q G;in;STR , which are pre-

sented in Figs. 3–5.

It is shown in Fig. 5 that x CO 2 ;L;in;SCR has a slight influence on the

requested H STR (below 3% for 10% variations of x CO 2 ;L;in;SCR ). For the

other ones, the influence can be considered negligible. Therefore,

CO 2 ;L;in;SCR is not a critical parameter, at least in the investigated

range, and it is low enough to ensure good performance of the

scrubber.

It is observed in Fig. 6 that the values of x CO 2 ;L;out;SCR strongly

influence the requested scrubber height to maintain the same

646 C.E. Wylock, W.M. Budzianowski / Chemical Engineering Science 170 (2017) 639–652 The evolution of the

Fig. 5. Variations of scrubber and stripper characteristics with variations of x CO 2 ;L;in;SCR .

646 C.E. Wylock, W.M. Budzianowski / Chemical Engineering Science 170 (2017) 639–652 The evolution of the

Fig. 6. Variations of scrubber and stripper characteristics with variations of x CO 2 ;L;out;SCR .

C.E. Wylock, W.M. Budzianowski / Chemical Engineering Science 170 (2017) 639–652

647

CO 2 abatement performance, while it slightly affects the other

characteristics. The variations of Q G;in;STR reach about 1% for 10%

variations of x CO 2 ;L;out;SCR . They are below 1% for D SCR , Q L;in;SCR , D STR

and H STR .

Fig. 7 shows the characteristics variations with x CO 2 ;G;out;STR . It

is observed that it is significant for H STR and Q G;in;STR , while it

has no influence on the scrubber characteristics. It shows that

decrease the targeted CO 2 content in the air leaving the stripper

is interesting, in the investigated range, because it requires a

decrease of the stripper height but it is offset by the need of

increasing also the stripping air flow rate to keep constant the

CO 2 abatement.

4.5. Influence of operating conditions

The sensitivity of the plant configuration with respect to the

operating parameters is investigated. Starting from the reference

case identified in 4.2, Q L;in;SCR , Q G ;in;STR , Q G;RAW , x CO 2 ;G;RAW , T, p G;in;SCR ,

p G;in;STR , p G;FLS are varied and their influence on the gas streams at

the scrubber and stripper outlet are computed. The performance

is evaluated on the basis of two criteria: the CO 2 ratio at the scrub-

ber outlet x CO 2 ;G ;out;SCR and the methane loss at the stripper outlet

dened by k CH 4 ;out;STR ¼

Q G;out;STR y CH 4 ;G ;out;STR . They are presented

in relative variations in Figs. 8–15, respectively.

4.5.1. Scrubbing water flow rate

It appears from Fig. 8 that the scrubbing water flow rate Q L;in;SCR

has a considerable influence on the scrubbing performance. Indeed,

even small variations can affect significantly x CO 2 ;G;out;SCR and

k CH 4 ;out;STR . They respectively decrease and increase while Q L;in;SCR

increase, meaning that a high Q L;in;SCR is favorable for the CO 2

scrubbing efficiency but it can be set off by an increase of the

methane loss. Q L;in;