Journal of Industrial and Engineering Chemistry xxx (xxxx) xxx

Contents lists available at ScienceDirect

Journal of Industrial and Engineering Chemistry

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

Performance investigation of MEA-1DMA2P solution for CO2 capture: A

rate-based modelling, optimization, and cost analysis
Morteza Afkhamipour a , Mohammad Shamsi b,* , Ebad Seifi a , Abbas Hekmatiyan c ,
Meisam Ansarpour d, Tohid N.Borhani d,*
a
National Iranian Gas Company, South Pars Gas Complex, Asaluyeh, Iran
b
Process Engineering Department, Faculty of Chemical Engineering, Tarbiat Modares University, Tehran, Iran
c
Department of Chemical Engineering, Faculty of Engineering, Arak University, Arak, Iran
d
School of Engineering, Centre for Engineering Innovation and Research, School of Engineering, Computing and Mathematical Sciences, University of Wolverhampton, UK

A R T I C L E I N F O A B S T R A C T

Keywords: This study presents a framework for evaluating the performance of MEA-1DMA2P solution in CO2 absorption for
CO2 capture a structured packed bed by integrating a rate-based model with statistical optimization. The developed model is
MEA-1DMA2P solution validated with experimental data. Sensitivity analysis identified mass transfer correlations as critical factors
Sensitivity analysis
influencing CO2 concentration profiles, with kinetic models having a less significant impact. Operating param-
Statistical optimization
Cost analysis
eters showed that CO2 removal efficiency increases with higher amine flow rate, column pressure, and amine
temperature but decreases with higher CO2 gas concentration. Efficiency notably declines when the amine
temperature exceeds 333 K. Statistical optimization was used to analyze the combined influence of operating
parameters on CO2 removal performance. The response values for statistical optimization were obtained using
the rate-based model, identifying the ranking of parameter impacts: amine flow rate had the most significant
influence, followed by CO2 partial pressure, blended amine temperature, and CO2 loading of blended amine. Cost
analysis was also done by scaling up the absorption system using an MEA-1DMA2P solution. The cost of CO2
capture based on the absorber column was estimated as 40.35 $/tonne of CO2, which is acceptable compared to
the literature and other solvents and processes for CO2 capture.

[3]. It must be mentioned that different amine types showed some

Introduction favorable characteristics, which depend on their chemical structure. The
trend for CO2 solubility in amine solvents typically follows this order:
Chemical absorption is a proven method for CO2 capture from in- primary amines exhibit the lowest, followed by hindered primary
dustrial exhaust gases [1,2]. The efficiency of this technology mainly amines, then secondary amines, and finally tertiary amines have the
relies on carefully choosing a solvent with the correct chemical and highest [6]. Despite each amine’s advantages, it is difficult to find high
physical characteristics. In fact, a suitable solvent must possess the CO2 absorption with low regeneration energy in a single amine [3].
following features: high CO2 solubility, low regeneration energy, fast Thus, for the purpose of increasing the CO2 absorption while reducing
kinetic reaction with CO2, low corrosivity and degradation [3]. These the heat of regeneration, amine blends have been considered by re-
characteristics significantly influence the operational cost of the process searchers as an effective method [1]. This method enables to combine
in terms of column height, solvent utilization, and energy requirement the advantage of different solvent to create a more effective system.
[4]. Standing out among CO2 capture options, amine-based solvents Some advantages that may result from amine blending are reduction of
offer impressive performance due to their rapid mass transfer rates, solvent corrosivity and degradation, high absorption rate, and low en-
economic viability, and capacity to handle significant exhaust gas loads ergy requirement for solvent regeneration [7]. Thus, this study examines
[5]. However, the recovery and regeneration of amine solutions require the CO2 absorption process using a mixture of monoethanolamine
significant energy, making the process highly energy-intensive. Addi- (MEA) and1-dimethylamino-2-propanol (1DMA2P).
tionally, the degradation of amines poses a significant challenge, further
reducing the economic efficiency and overall performance of the process

* Corresponding authors.
E-mail addresses: [email protected], [email protected] (M. Shamsi), [email protected] (T. N.Borhani).

https://doi.org/10.1016/j.jiec.2024.12.011
Received 4 October 2024; Received in revised form 19 November 2024; Accepted 5 December 2024
Available online 6 December 2024
1226-086X/© 2024 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights are reserved, including those for text and data
mining, AI training, and similar technologies.

Please cite this article as: Morteza Afkhamipour et al., Journal of Industrial and Engineering Chemistry, https://doi.org/10.1016/j.jiec.2024.12.011
M. Afkhamipour et al. Journal of Industrial and Engineering Chemistry xxx (xxxx) xxx

Nomenclature ΔHrX (abs) Heat of absorption of CO2 (kJ/kmol CO2)

ΔHlv Heat of condensation of water (kJ/kmol H2O)
Symbols ΔZ Differential height of the packed column
Ac Cross-sectional area of the column (m2)
ae Effective surface area for mass transfer (m2/m3) Greek letters
ai Activity for componenti αCO2 CO2 loading
ap Specific surface area of packing (m2/m3) ϕi Fugacity coefficient for component i in the solution
Ci Molar concentration of component i (kmol/m3) ϕsH2 O Fugacity coefficient of pure water at its vapour pressure
CL Molar concentration of liquid phase (kmol/m3) βij Tuneable parameter between component i and component
CG Molar concentration of gas phase (kmol/m3) j
CpG Molar heat capacity of the gas (kJ/kmolK) γi Activity coefficient for component i
CpL Molar heat capacity of the liquid phase (kJ/kmolK) ∊ Pump efficiency
Cpi Molar heat capacity of component i in the liquid phase (kJ/ μi Viscosity of component i (Pa.s)
kmolK) ρi Density of component i (kg/m3)
Dij Diffusivity of component i in component j (m2/s) σi Surface tension of component i (mN/m)
ECO2 Enhancement factor σL Surface tension of liquid phase (mN/m)
G Molar gas flowrate (kmol/s)
Abbreviations
H Henry’s law constant (kPam3/kmol)
CEPCI Chemical Engineering Plant Cost Index
hG Heat transfer coefficient in gas (kJ/Km2s)
OAD Orthogonal array design
I Solution’s ionic strength
OF Objective function
Kj Chemical equilibrium constant for reaction j
SNR Signal-to-noise ratio
kL,i Physical liquid-side mass transfer coefficient of component
◦

i (m/s) Chemicals
kG,i Gas-side mass transfer coefficient of component i (m/s) 1DMA2P 1-dimethylamino-2-propanol
KG,i Overall mass transfer coefficient of component i in the gas MEA Monoethanolamine
phase (kmol/m2skPa) MEACOO− Carbamate ion
L Molar liquid flowrate (kmol/s) MEAH+ Carbamic acid
mj Molality concentration of component j (kmol/m3) HCO−3 Bicarbonate ion
Ni Molar flux of component i (kmol/m2s) CO2−
3 Carbonate ion
PsH2 O Vapour pressure of pure water (kPa)
Dimensionless groups
P*i Equilibrium partial pressure of i (Partial pressure of i in gas
FrL Froude number (u2L ap /g)
phase in equilibrium with the liquid phase (kPa))
Pi Partial pressure of component i in the bulk gas phase (kPa) ReG Reynolds number of gas phase (uG ρG /ap μG )
TL Liquid-phase temperature (K) ReL Reynolds number of liquid phase (uL ρL /ap μL )
TG Gas-phase temperature (K) ScG,i Schmidt number of component i in gas phase (μG,i /ρG,i DG,i )
uG Superficial gas velocity (m/s) ScL,i Schmidt number of component i in liquid phase
uL Superficial liquid velocity (m/s) (μL,i /ρL,i DL,i )
xi Liquid mole fraction of component i WeL Webber number (u2L ρL /ap σ)
yi Gas mole fraction of component i
Zi Electrical charge for ion i

Performance of MEA-1DMA2P solution increasing the absorption capacity. Combining the advantages of both
solvents will provide interesting features such as fast reaction, high rate
As mentioned earlier, amine blends have the advantage of combining for mass transfer, high CO2 loading, reduced solvent regeneration en-
the favorable characteristics of each solvent in the mixture. Thus, it is ergy, and cost-effective process. The following section will review some
essential to know the features of each solvent. MEA is a primary alka- experimental investigations of this amine blend.
nolamine mainly used as a solvent for CO2 capture in the post-
combustion process [8,9]. The main advantages of MEA are its high
Experimental investigations of MEA-1DMA2P mixture
reactivity with CO2, high absorption rate, low price and availability,
maturity of the process, and data compared to other amines. These
Gao et al. [11] investigated the performance of different tertiary
characteristics provide fast reaction, high mass transfer rate, and cost-
amines, including 1DMA2P, mixed with MEA aqueous solution. The
effective process, which make MEA an attractive solvent for most in-
molar ratio of the system was 5 mol of MEA for 1 mol of tertiary amines.
dustrial processes. However, the primary drawbacks of this solvent
Their results showed that combining MEA and 1DMA2P offered superior
include its high degradation and corrosivity, the high energy require-
CO2 capture due to its high equilibrium, rapid absorption, and swift
ment for regeneration, and low CO2 loading [3,10].
regeneration. Zhang et al. [12] have investigated the formation of bi-
1DMA2P is considered as a tertiary alkanol amine; in contrast to
carbonate ions (responsible for low energy of solvent renewal) in CO2
MEA, it requires low energy for regeneration and high CO2 loading. The
loaded aqueous MEA solution and two types of amines blended which
performance of 1DMA2P is compared with MEA and MDEA [7]. The
are MEA-MDEA and MEA-1DMA2P. The molar ratio of amine blended
outcomes of the experiments revealed that 1DMA2P involves a lower
was kept at 5/1. The experimental results showed that MEA-1DMA2P
reaction heat and higher CO2 equilibrium solubility but a lower mass
generate a higher bicarbonate ion at CO2 loading of 0.34 mol CO2/
transfer rate compared to MEA. Thus, 1DMA2P is a potential solvent
mole amine. This strengthen the idea that MEA-1DMA2P is an adequate
suitable for reducing the heat duty of solvent regeneration and
solvent for lowering energy need for amine renewal. In a further

2
M. Afkhamipour et al. Journal of Industrial and Engineering Chemistry xxx (xxxx) xxx

experiment, Afkhamipour et al. [13] focused on determining the employs the Deshmukh–Mather (D-M) model within its rate-based
maximum amount of CO2 that could dissolve in a mixture of MEA and model. In the D-M model, water serves as the solvent for this mixture,
DMA2P at equilibrium. A range of temperatures (313.15–333.15 K), CO2 while amines and ions comprise the dissolved solutes. Given the re-
partial pressures (3–186 kPa), and a total mixed amine concentration of actions, the following equations define the equilibrium constant for each
2.5 M were used in their experiments. The findings revealed that an individual reaction:
increasing the amount of 1DMA2P relative to MEA in the mixture led to aH+ a1DMA2P
a greater capacity for CO2 absorption. Furthermore, the E-UNIQUAC K1 = (1)
a1DMA2PH+
model was employed to forecast the experimental outcomes, exhibiting
an acceptable degree of accuracy with an AARD of 22.4 %. Ling et al. aH+ aMEA
K2 = (2)
[14] developed an empirical model to predict the rate of mass transfer aMEAH+
and CO2 equilibrium solubility in MEA-1DMA2P. The accuracy of this
model has been verified with the experimental data using a packed bed aHCO−3 aMEA
K3 = (3)
column. This shows an AARD of 2.64 % and 15.97 % for equilibrium aMEACOO−
solubility and mass transfer, respectively. Moreover, the experimental
aH+ aHCO−3
data give the parameters that might be considered for an effective pro- K4 = (4)
aCO2 aH2 O
cess, such as CO2 partial pressure, liquid temperature, and amine flow
rate. aH+ aCO2−3
K5 = (5)
aHCO−3
Motivation of the current study
aH+ aOH−
K6 = (6)
This research used a rate-based model to accurately predict CO2 aH2 O
capture performance by employing a mixture of MEA and 1DMA2P in
conjunction with a sensitivity analysis of the kinetic and mass transfer where ai denotes the activity for ion and molecular species, based on the
correlations. Furthermore, a parametric study was done to examine the molality concentration (mi ), defined as:
influence of critical variables on absorption efficiency. Also, an eco- ai = γi mi (7)
nomic analysis was conducted to estimate the cost of CO2 capture. Ul-
timately, statistical optimization studies were done to evaluate the here, γ i represents the activity coefficient. The constants (Kj ) for equi-
collective impact of multiple operational factors on absorption librium reactions (R1-R6) are temperature-dependent, and can be
efficiency. expressed through the following equation:
A
Methodology ln(Kj ) = + Bln(T) + C(T) + D (8)
(T)
This part of the study presents the procedures and equations used for
The parameters A, B, C, and D within Eq. (8) were obtained from pre-
thermodynamic and rate-based modelling, parametric study, sensitivity
viously published references. They are listed in the supporting infor-
analysis, statistical optimization and economic assessment.
mation (Table S1).
Furthermore, to ascertain the unknown concentrations for all spe-
Thermodynamic modelling cies, the process of mass balance and charge conservation within the
system is fundamental. This can be expressed as:
The required equations for development of thermodynamic model
are represented in this section. [1DMA2P]0 = [1DMA2P] + [1DMA2PH+ ] (9)

Chemical reaction of CO2-MEA-1DMA2P-H2O system [MEA]0 = [MEA] + [MEAH+ ] + [MEACOO− ] (10)

CO2 absorption in the MEA-1DMA2P mixture occurs via a series of ( ) [ ]
chemical reactions [13]. These reactions are as follows: αCO2 [1DMA2P]0 + [MEA]0 = [CO2 ] + [HCO3 − ] + CO3 2− + [MEACOO− ]
(11)
K1
1DMA2PH+ ↔ 1DMA2P + H+ (R1) [ ]
[MEAH+ ] + [H+ ] + [1DMA2PH+ ] = [HCO3 − ] + [OH− ] + 2 CO3 2−
K2
(12)
MEAH+ ↔ MEA + H+ (R2) + [MEACOO− ]

K3 The notation uses [1DMA2P]0 and [MEA]0 to denote the initial concen-
MEACOO− + H2 O ↔ MEA + HCO−3 (R3)
trations of 1DMA2P and MEA, while αCO2 indicates the CO2 loading. In
K4
addition to (R1)-(R6) the following reactions also must be considered:
CO2 + H2 O ↔ HCO−3 + H+ (R4)
H2 O (g)⇌H2 O (aq) (R7)
K5
HCO−3 ↔ CO2−
3 +H +
(R5) CO2 (g)⇌CO2 (aq) (R8)
K6
H2 O ↔ OH− + H+ (R6) MEA(g)⇌MEA (aq) (R9)

Deshmukh–Mather model equations 1DMA2P(g)⇌1DMA2P (aq) (R10)

In the CO2 absorption process, which involves both physical and For phase equilibrium calculations, the vapor phase presence of
chemical equilibria, CO2 initially enters the liquid phase from the gas 1DMA2P and MEA can be omitted; hence, (R9) and (R10) are not
phase physically. Subsequently, various chemical species, including ions considered in this study. Therefore, the following equations have been
and molecular components, are produced in succession through chem- utilized for deriving phase equilibrium for H2O and CO2 [13]:
ical reactions. Selecting an appropriate thermodynamic model is crucial
for accurately predicting experimental data [6,15]. Therefore, the study

3
M. Afkhamipour et al. Journal of Industrial and Engineering Chemistry xxx (xxxx) xxx

P*CO2 ϕCO2 = γCO2 HCO2 xCO2 (13) loading of CO2. The following steps outline the solution procedure for
the thermodynamic model equations (Eqs. (1)-(18)), incorporating an
P*H2 O (ϕH2 O /ϕsH2 O ) = γ H2 O PsH2 O xH2 O (14) optimization and calculation approach illustrated by the flowchart in
Fig. 1.
P*CO2 stands for the equilibrium partial pressure of CO2 in the gas phase,
• Initial guesses of zero were assumed for fij and gij in Eq. (17).
ϕCO2 is a term for fugacity coefficient, and γCO2 denotes the activity co-
• The OF (Eq. (18)) was minimized based on CO2 loading data.
efficient of CO2 relative to an unsymmetrical reference state. Addition-
• Optimization terminated when the OF reached the minimum value.
ally, HeCO2 stands for Henry’s law of CO2 in water, P*H2 O is the
• Once the tuneable parameters were fine-tuned, they were incorpo-
equilibrium partial pressure of H2 O, and the term ϕH2 O /ϕsH2 O represents rated into the D-M model.
the ratio of the fugacity coefficient of water in the solution to the • Non-linear equations Eqs. (1)-(14) via Newton’s method, following
fugacity coefficient of pure water at its vapor pressure. Furthermore, these steps:
γ H2 O is utilized as the activity coefficient of H2 O, and PsH2 O signifies pure 1. The initial CO2 loading guess for Newton’s method was based on the
water’s vapor pressure, which is calculated using the Antoine equation. concentration of H+.
It was found that at low to moderate pressures, both Eqs. (13) and 2. All species’ concentrations are obtained by solving Eqs. (1)-(12)
(14) were valid with the Poynting factor set to 1. Further, variables such simultaneously using the initial concentration of H+ from step 1. The
as ϕCO2 and ϕH2 O /ϕsH2 O were assumed to be unity. Furthermore, activity concentration of H+ linked to solution pH, ranged between 10− 7 and
coefficients are included for every chemical species to account for the 10− 14 .
non-ideal behavior of the mixture in the D-M model. The extended 3. An approximation of Eq. (15) neglecting higher-order terms was
Debye-Hückel equation is employed to determine these activity co- employed to generate initial estimates of species concentrations.
efficients [16], which is as follows: 4. These estimates were refined using optimized parameters to obtain
√̅̅ final concentrations.
AZ2i I ∑
lnγi = − √̅̅ + 2 βij mj (15) 5. The model’s accuracy was assessed by comparing its predicted CO2
1+B I j∕
=w loading values with those obtained experimentally.

Eq. (15) encompasses both electrostatic and van der Waals forces within Table S4 in the supporting information reports results for binary
the system. Parameter A, influenced by temperature, has been obtained interaction parameters for the CO2-MEA-1DMA2P-H2O system corre-
from previous studies [17], while parameter B remains constant at 1.2 sponding to potential binary interactions (Table S2).
[18]. Furthermore, the solution’s ionic strength (I) is defined as: Fig. 1 presents the flowchart depicting the model utilized to deter-
1∑ mine CO2 loading in the CO2-MEA-1DMA2P-H2O system. The CO2 ab-
I= mj Z2j (16)
2 i sorption heat ΔHrX (abs) expression was derived from the Gibbs-
Helmholtz equation, and the results incorporated into the liquid tem-
In Eq. (16), Zi represents the electrical charge for ions, while the term βij perature equation within the rate-based model, as shown below:
in Eq. (15) signifies the tuneable parameters between molecules and
ions. Due to the extensive nature of the reaction system involving CO2 d(lnP*CO2 )
ΔHrX (abs) = − R () (19)
and the 1DMA2P-MEA solution, a total of 11 species are involved,
d T1
resulting in the assessment of 121 interaction parameters for the
chemical species. Determining the appropriate number of tuneable pa-
rameters for a model that utilizes experimental data for the CO2- The equilibrium partial pressure of CO2 (P*CO2 ) was assumed to depend
1DMA2P-MEA-H2O system is challenging using conventional optimi- on temperature (T) and CO2 loading (αCO2 ) through the following
zation algorithms [19–22]. Using activity coefficients in the D-M model equation [23]:
allows the number of parameters needed to adjust for interactions be-
b dαCO2
tween components to be reduced to 21 βij . The complexity of the model lnP*CO2 = a + + cαCO2 + + eα2CO2 (20)
T T
is reduced by making some assumptions. Water does not affect other
components, interactions have equal effects in both directions (βij = Taking the derivative of Eq. (20) gives:
βji ), and attractions/repulsions between similar species (like charges
ΔHrX (abs) = − R(b + dαCO2 ) (21)
and molecules) are negligible. Additionally, interactions of scarce H+
and OH− ions are neglected. The CO2-1DMA2P-MEA-H2O system is characterized by fitting co-
Table S2 in the supporting information presents potential binary efficients a, b, c, d, and e of Eq. (20), which have values of − 23.48,
interaction parameters within the D-M model. Eq. (15) incorporates − 4943.69, 160.15, − 1584.80, and − 115.44, respectively. It is worth
these parameters, which are adjusted using a temperature-dependent noting that the fitted data for Eq. (20) was obtained from the equilib-
linear equation as follows: rium solubility data of a 6 kmol/m3 of MEA/1DMA2P (5/1) solution
βij = fij + (gij ×T) (17) [14].

To find the best values for the tuneable parameters, an optimization Rate-based model
technique called the Nelder-Mead simplex algorithm was used. This
method is available in MATLAB’s fminsearch function. These parame- This work explores CO2 capture in a packed column using a blended
ters minimized the objective function (OF), which quantifies the amine solution (1DMA2P-MEA) through a rate-based model. The model
discrepancy between measured and calculated CO2 loading data as integrates heat and mass transfer principles, a robust thermodynamic
follows: framework, estimations of key physical and chemical properties, along
n
[ ]
Exp 2 with an advanced mass transfer enhancement factor model and internal
1 ∑ αCal
CO2 − αCO2
OF = (18) hydraulic calculations. The model’s predictions are compared to
N i=1 αExp
CO experimental data using numerical techniques, serving as a validation
2

process. Subsequently, it is used to analyse the column’s performance,

where, αExp Cal
CO2 is experimental loading of CO2 and αCO2 is calculated including concentration and temperature profiles, the sensitivity of mass

4
M. Afkhamipour et al. Journal of Industrial and Engineering Chemistry xxx (xxxx) xxx

Fig. 1. Flow chart for thermodynamic modelling.

transfer correlations, and the influence of operating conditions on CO2 Model theory
capture efficiency. A mathematical model for a packed absorption column can be
derived using mass and energy balance principles. It treats the gas and
Assumptions of the rate-based model liquid phases as separate entities, establishing individual balances
The most important assumptions used in the model are: within a small differential section of the column (refer to Fig. 2 for this
differential height (ΔZ)). In this figure, surface III represents a volume
• Steady state is considered. element of the packed column with a height of dz. This surface includes
• The gas and liquid streams are assumed to move through the packed gas and liquid layers, shown by surfaces I and II.
column plug-like. The model also considers the pressure drop within
the column to be insignificant Governing equations
• The heat transfer between the column and its surroundings is The governing equations required for the rate-based model are listed
neglected. In other words, the behavior of the column is adiabatic. as follows:
• The model assumes the effective surface area for both mass transfer Gas flowrate mass balance:
and heat transfer within the packing material is the same.
dG
• Due to the low pressure encountered during CO2 absorption after = − (NCO2 + NH2 O )ae Ac (22)
dZ
post-combustion process, the model treats the gas phase as ideal.
CO2 mass balance in the gas phase:

5
M. Afkhamipour et al. Journal of Industrial and Engineering Chemistry xxx (xxxx) xxx

Fig. 2. Packed column and its element.

dyCO2 [ ( ) ]a A
e c
= (NCO2 ) yCO2 − 1 + (NH2 O )yCO2 (23) dxMEA ae Ac
dZ G = [NCO2 (xMEA + 1) + NH2 O (xMEA ) ] (28)
dZ L
H2O mass balance in the gas phase:
Mass balance for 1DMA2P:
dyH2 O [ ( ) ]a A
e c
= (NH2 O ) yH2 O − 1 + (NCO2 )yH2 O (24) dx1DMA2P ae Ac
dZ G = [NCO2 (x1DMA2P + 1) + NH2 O (x1DMA2P ) ] (29)
dZ L
Liquid flow rate mass balance:
Mass balance for MEAH+:
dL
= − (NCO2 + NH2 O )ae Ac (25) dxMEAH+ ae Ac
dZ = [NCO2 (xMEAH+ − 1) + NH2 O (xMEAH+ ) ] (30)
dZ L
CO2 mass balance in the liquid phase:
Mass balance for 1DMA2PH+:
dxCO2 ae Ac
= [(xCO2 − 1)NCO2 + (xCO2 + 1)NH2 O ] (26) dx1DMA2PH+ ae Ac
dZ L = [NCO2 (x1DMA2PH+ − 1) + NH2 O (x1DMA2PH+ ) ] (31)
dZ L
H2O mass balance in the liquid phase:
Mass balance for HCO3 − :
dxH2 O ae Ac
= [(xH2 O + 1)NCO2 + (xH2 O − 1)NH2 O ] (27) dxHCO−3 [ ( ) ]a A
dZ L = NCO2 xHCO−3 − 1 + NH2 O (xHCO−3 )
e c
(32)
dZ L
Mass balance for MEA:
Gas temperature:

dTG TG ae Ac ( )T a A
G e c hG (TG − TL )ae Ac
= (NCO2 + NH2 O ) − NCO2 CpCO2 + NH2 O CpH2 O − ( ) (33)
dZ G CpG G CpG G

6
M. Afkhamipour et al. Journal of Industrial and Engineering Chemistry xxx (xxxx) xxx

quantifies the significant increase in the mass transfer rate caused by the
Liquid temperature: chemical reaction between CO2 and the amine solution. This factor is

dTL TL ae Ac ( ) T a A h (T − T )a A [
G e c G G L e c ] ae Ac
= (NCO2 + NH2 O ) − NCO2 CpCO2 + NH2 O CpH2 O − − NCO2 ΔHrX (abs) + NH2 O .ΔHlv (34)
dZ L L.CpL L.CpL LCpL

typically calculated by solving simultaneous reaction–diffusion partial

In these equations ae is the effective surface area for mass transfer (m2/ differential equations within the liquid film or by employing mass
m3), Ac is the cross-sectional area of the column (m2), Ni molar flux of transfer theories found in existing literature, such as surface renewal
component i (kmol/m2.s) (Ni = KG,i (Pi − P*i )), CpG molar heat capacity of theory, film theory, and penetration theory [10,29,30]. In this study, Eq.
the gas (kJ/kmol.K), CpL molar heat capacity of the liquid (kJ/kmol.K), (36) indicates that ECO2 is solely dependent on the dimensionless number
Cpi molar heat capacity of component i in the liquid phase (kJ/kmol.K), Hatta.
TL liquid-phase temperature (K), TG gas-phase temperature (K), yi the √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
gas mole fraction of component i, and xi the liquid mole fraction of k2,1DMA2P− MEA •Cbulk 1DMA2P− MEA • DCO2 ,1DMA2P− MEA
ECO2 = (36)
component i. hG heat transfer coefficient in gas (kJ/K.m2.s), L molar kL,CO2
◦

liquid flowrate (kmol/s), G molar gas flowrate (kmol/s), ΔHlv the heat of
condensation of water (kJ/kmol H2O), and ΔHrX (abs) heat of absorption The kinetic constant for the second-order reaction of CO2 with 1DMA2P-
of CO2 (kJ/kmol CO2). MEA solution, denoted as k2,1DMA2P− MEA , defines the reaction rate. The
diffusivity of CO2 in the 1DMA2P-MEA solution is denoted by
Mass transfer correlations DCO2 ,1DMA2P− MEA .Cbulk
1DMA2P− MEA is the concentration of 1DMA2P-MEA in
Existing correlations were employed within a rate-based model to the liquid. The physical mass transfer coefficient in the liquid phase is
quantify mass transfer coefficients in both liquid and gas phases, along denoted by kL,CO2 . The K2,1DMA2P for the CO2-H2O-1DMA2P system was
◦

with the effective interfacial area (refer to Table 1). Eq. (35) depicts the used with the equation proposed by Kadiwala et al. [31].
two-film model, expressing the overall mass transfer coefficient (KG,CO2 ) ( )
− 7523
by considering individual mass transfer resistances in the liquid and gas K2,1DMA2P = 1.9 × 109 exp (37)
T
phases.
1 1 HCO2 While for the CO2-H2O-MEA system, the kinetic constant for the second-
= +
KG,CO2 kG,CO2 ECO2 kL,CO2
◦
order reaction was derived from various literature sources. Table S3
(35)
1 1
=
KG,H2 O kG,H2 O
Table 2
Within the two-film model, the gas-side mass transfer coefficient, Experimental operating conditions for CO2-MEA-1DMA2P-H2O system in a DX-
Type structured packing column.
denoted by kG,CO2 , and the physical liquid-side mass transfer coefficient,
denoted by kL,CO2 . These parameters can be calculated using correlations
◦
Conditions Values and units

presented in Table 1. ECO2 is enhancement factor related to the increase MEA-1DMA2P concentration 6.0 kmol/m3
of mass transfer rate in the chemical reaction processes [24]. Henry’s Feed temperature for amine 293–333 K
Flow rate of amine 2.9–5.8 m3/m2.hr
law constant is denoted by HCO2 .
CO2 partial pressure 12–20 kPa
CO2 loading 0.204–0.437 mol CO2/mole amine
Enhancement factor Gas flow rate 33.49–48.07 kmol/m2.hr
The factor known as the mass transfer enhancement factor (ECO2 )

Table 1
Three cases of selected mass transfer correlations for sensitivity analyses.
Case Authors Correlations Ref.
[ ]0.8
1 Rocha et al. DG,i ρG s(uLe + uGe )
kG,i = 0.054 Sc0.33
G,i
[25]
s μG
[ ]
Rocha et al. 0.9DL uLe 0.5 uG uL
kL,i = 2
◦
uGe = and uLe = [25]
πs ∊(1 − hL )sinα ∊hL sinα
( )
Rocha et al. ae 29.12u0.4 ρL 0.15
L vL s
0.2 0.159
= Fse Ft = Fse [25]
ap (1 − 0.93cosθ)(sinα)0.3 ∊0.6 σL g
[ ]
2 Bravo et al. DG,i ρG deq (uLe + uGe ) 0.8 0.33
kG,i = 0.0338 ScG,i [26]
deq μG
Bravo et al. [
DL,i
]0.5
uG
( 2 )0.33
9Γ g
kL,i = 2
◦
u , uGe = and uLe = [26]
πs Le ∊sinα 8ρL μL
( )− 0.18 ( )0.233
Hanley and Chen ae ρ μG
= 2.308Re−G 0.274 Re0.246L We0.248
L Fr−L 0.161 G [27]
ap ρL μL
( )
3 Hanley and Chen CL DL,i
kL,i = 12Re1L Sc0.33 [27]
◦
L,i
de
( )
Hanley and Chen 0.33 CG DG,i
kG,i = 0.3516Re0.5 G ScG,i [27]
de
( )
Henriques de Brito et al. ae ρ uL 0.3
= 0.465 L [28]
ap μL ap

7
M. Afkhamipour et al. Journal of Industrial and Engineering Chemistry xxx (xxxx) xxx

Table 3 Table 4
Used experimental data for CO2-MEA-1DMA2P-H2O system in a DX-Type Key physicochemical properties utilized for gas and liquid phases.
structured packing column. Property Equation/Comment Ref
Parameter Run-4 Run-5 Run-19 Run-24 ( )
MEA density ρMEA = 1.19 − 4.29 × 10− 4 T − 5.66 × 10− 7 T2 × 1000
[32]
G (kmol/m2⋅hr) 44.21 44.21 44.21 44.21 (kg/m3)
L (m3/m2⋅hr) 5.82 3.70 3.70 3.70 Water density ρH2 O = 780.82 + 1.73T − 0.0034T2
[32]
CMEA (kmol/m3) 5 5 5 5 (kg/m3)
C1DMA2P (kmol/m3) 1 1 1 1 1DMA2P ρ1DMA2P = 1138.20 − 0.982T
[33]
TL-in (K) 313.15 313.15 313.15 313.15 density (kg/
TG-in (K) 298.15 298.15 298.15 298.15 m3)
αlean (mol/mol) 0.282 0.276 0.204 0.437 Mixture xMEA MMEA + xH2 O MH2 O + x1DMA2P M1DMA2P
ρMixture =
MMEA MH O M1DMA2P [34]
αrich (mol/mol) 0.475 0.479 0.484 0.490 density (kg/ xMEA + xH2 O 2 + x1DMA2P
CO2 concentration along the packed column (%) m3) ρMEA ρH2 O ρ1DMA2P
Z (m) 0 15.00 12.00 15.00 15.00 MEA surface σMEA = 107.91 − 0.1458T + 4 × 10− 6 T2
[35]
0.23 10.56 10.08 10.86 14.39 tension
0.44 6.62 7.44 7.53 14.49 (mN/m)
0.6 4.39 5.92 5.20 14.09 Water surface σH2 O = 76 − 0.169(T − 273.15)
[35]
0.86 1.67 2.99 1.77 13.79 tension
1.08 0.56 1.67 1.06 13.18 (mN/m)
​ 1.25 0.15 0.76 0.15 11.36 1DMA2P σ1DMA2P = 95.872 − 0.185T
[35]
surface
tension
(supporting information) lists these equations. (mN/m)
Mixture σMixture = σIdeal + σNon− Ideal σIdeal = σMEA xMEA +
[35]
surface σ1DMA2P x1DMA2P + σH2 O xH2 O σNon− Ideal =
Experimental data used in the rate model tension xMEA x1DMA2P GMEA− 1DMA2P +
The experimental data presented by Ling et al. [14] was used to (mN/m) x1DMA2P xH2 O G1DMA2P− H2 O +
perform model development and validation. In fact, as mentioned in xMEA xH2 O GMEA− H2 O GMEA− 1DMA2P =
− 2.57 − 190
section 1.2, Ling et al. [14] studied the performance of the MEA- − 0.432TG1DMA2P− H2 O = +
wMEA + w1DMA2P w1DMA2P
1DMA2P mixture for post-combustion CO2 capture. To do so, they − 1.51
0.41TGMEA− H2 O = − 0.298T
determined through different experiments the physical solubility, wMEA
( )
equilibrium, and mass transfer of CO2 in the blended MEA-1DMA2P MEA viscosity − 7 2
μMEA = 0.078 +6 × 10 T /1000The coefficients are
[36]
(Pa.s) regressed using data presented in ref. [36]
solution with a molar ratio of 5/1(5 kmol of MEA to 1 kmol of
1DMA2P μ1DMA2P = (6.8873 − 0.018T)/1000
1DMA2P). To evaluate the impact of various operating parameters on [37]
viscosity
system performance, a range of temperatures, flow rates, and CO2 par- (Pa.s)
tial pressures were investigated. Table 2 presents the experimental
( )
Water 3703.6
μH2 O = exp − 52.843 + +5.866logT − 5.879 × [24]
operating conditions for the absorber column [14]. To assess the vali- viscosity T
29 10
dation of the rate-based model, the published experimental data from 10− T
Mixture μMixture = xMEA μMEA + x1DMA2P μ1DMA2P + xH2 O μH2 O
four runs (4, 5, 19, and 24) [14], as presented in Table 3, were utilized. [30]
viscosity
(Pa.s)
Physicochemical properties Henry’s N2O analogy method:
[14]
Table 4 shows the key physicochemical properties used for modelling constant
HCO2 in MEA− 1DMA2P =
(kPa.m3/
CO2 capture with MEA-1DMA2P solution.
kmol) HCO2 in H2 O
HN O in MEA− 1DMA2P HN2 O in H2 O = 8.55 ×
HN2 O in H2 O 2
Solving the rate-based model (
− 2284
)
106 exp HCO2 in H2 O = 2.82 ×
The differential equations (Eqs. (22)-(34)) of the rate-based model (
T
)
were solved simultaneously using the finite difference approach and − 2044
106 exp HN2 O in MEA− 1DMA2P = 1.56 ×
T
shooting method to simulate and optimize the absorber. Due to counter- ( )
− 3047.80
current flow in the column, this computational method was employed. 8
10 exp
T
The absorber receives gas and liquid streams with well-defined inlet Diffusivity of − 2119
DCO2 − H2 O = 2.35 × 10− 6 exp( )DCO2 − Mixture = [38]
conditions. However, the concentrations, temperatures, and flow rates CO2 (m2/s)
( )0.8
T
of the gas and liquid exiting the top and bottom of the column remain DCO2 − H2 O
μH2 O
μMixture
unknown. In such cases, the shooting method iteratively adjusts the Molar heat Cp,MEA = 7000.128T0.552 Cp,1DMA2P = 0.936 + 2.17 ×
initial values and repeatedly guesses and corrects liquid conditions capacity (J/ 103 TCp,Mixture = xMEA Cp,MEA + x1DMA2P Cp,1DMA2P +
[3940]
leaving the column. In this method, the absorber column height is dis- kmol.K) xH2 O Cp,H2 O
cretized into 100 differential elements, and the unknown operating
variables of the liquid stream exiting the column are estimated initially.
Fig. 3 details the application of the shooting method to solve and adjust and the differences, and the next guess (Lb3 ) can be determined utilizing
for the initial liquid phase variables, including temperature, liquid flow linear function as follows:
rate, and other components in rich amine solution. By considering initial Lb2 − Lb1 ( )
guesses (2 guesses) for example here liquid flow rate (Lb1 and Lb2 ) at the Lb3 = Lt − Lt,cal1 + Lb1 (38)
Lt,cal2 − Lt,cal1
stream exiting from bottom of column, the top inlet liquid flow rate
(Lt,cal1 and Lt,cal2 ) obtained using the finite difference method, can be Overall program algorithm for rate-based model
compared with the real flow rate (Lt ) at the top of column. The differ- The rate-based calculations for CO2 capture using amines become
ences between these liquid flow rates represent the discrepancy between complex and time-consuming due to the occurring chemical reactions.
the real conditions and the calculated values by the finite difference Therefore, a computer program is used. This program is written in the
method. These differences are plotted on the vertical axis, and points A MATLAB programming environment (Fig. 4).
and B are formed by these differences along with the two initial guess The program steps are as follows:
values. The line AB represents the relationship between the guess values

8
M. Afkhamipour et al. Journal of Industrial and Engineering Chemistry xxx (xxxx) xxx

Fig. 3. Schematic of the shooting method for initial value adjustment.

1. Choose a value for ΔZ of the packed column.

2. Enter the input gas feed variables, including temperature, flow rate,
and components.
3. Guess the values for the temperature and flow rate of liquid phase at
the outlet of the absorption column.
4. Determine all the physicochemical properties for both gas and liquid
phases.
5. Compute the mass transfer enhancement factor and determine
KG,CO2 .
6. Calculate the mass transfer flux for the components transferring
between the liquid and gas phases.
7. Use the finite difference technique to estimate how mole fraction,
flow rate, and temperature change with position in gas and liquid
phases.
8. Conclude the algorithm’s execution by determining the discrep-
ancies between the results of the model and experimental data of CO2
mole fraction.

Parametric study

The impact of important operating parameters was first considered

as one factor, followed by investigating the simultaneous influence of
these parameters on the removal efficiency. Effect of changing amine
flow rates, amine temperature, the amount of CO2 entering the absorber,
different column pressure values on the efficiency of CO2 removal is
examined in this section.

Sensitivity analysis

For accurate rate-based modelling of packed columns, it is crucial to

analyse how sensitive the model is to changes in effective surface area
correlations, mass transfer coefficients, and kinetic parameters. There-
fore, various cases presented in Table 3 were utilized to perform the
sensitivity analysis.

Statistical optimization

The statistical optimization was performed using the Taguchi

method. This powerful statistical method is based on the Taguchi
Fig. 4. Flowchart of the rate-based model calculations.
orthogonal array design (OAD) matrix, which provides the best setting
parameters for experimental design and enables the determination of
the optimum points for process optimization [41]. This work employs a
statistical optimization technique, the Taguchi method, to identify the
operating conditions that maximize CO2 capture efficiency for an

9
M. Afkhamipour et al. Journal of Industrial and Engineering Chemistry xxx (xxxx) xxx

Fig. 5. Flowchart of the CO2 capture optimization algorithm based on the Taguchi method.

absorber column [42]. Based on the flowchart (Fig. 5), steps were used
Table 5
to optimize CO2 removal efficiency.
Process and total equipment costs [44].
Fig. 5, a detailed description of using the OAD method for simulta-
neously assessing the impact of important operational parameters, Equipment Installation factor Cost ($)

including amine temperature, amine solution flow rate, CO2 loading, Packed column 1.7 5,830.83
and CO2 mole fraction in the flue gas, along with four levels for each Pump 2.0 49,126.66
Storage tank (×2) 2.7 178,501.37
factor, on the CO2 absorption efficiency by the amine solution has been
Process equipment cost (A) − Sum of all process equipment
investigated. In fact, the OAD-L16 method was employed for finding the General facilities (B) − 10–20 % A
points in which optimized operational parameters. Initially, in the first Total equipment cost (TEC) − A+B
stage, the OF and their levels need to be determined. Subsequently, the
determination of the four-level range for each factor used for optimi-
zation was conducted. Finally, by designing the orthogonal array matrix Economic assessment
L16 and implementing it with a rate-based model, the signal-to-noise
ratio (SNR) values for designing the orthogonal array matrix were In this section, the capital cost (CAPEX) and operating cost (OPEX)
calculated. are estimated. The Chemical Engineering Plant Cost Index (CEPCI) for
Therefore, the OF is formulated to maximize the CO2 removal effi- the cost inflation from the reference year to 2023 (797.9) is considered
ciency in an absorber column. The SNR values are obtained by applying [43]. The utilization of the CEPCI for the purpose of predicting equip-
Eq. (39) to maximize absorption efficiency. ment expenditures across different time frames has been widely
[ ] accepted as the conventional approach for cost escalation.
1∑n 1 Capital and operating costs are estimated for the absorption section
SNR(i) = − 10log10 (39)
n i=1 RE2(i) only. Therefore, as there is no stripper column in this study, the cost of
the used solvent increases rapidly. On the other hand, no energy is
The L16 orthogonal array design is implemented with n runs. CO2 required for the solvent regeneration and CO2 compression sections. The
removal efficiency (RE(i) ) is measured for each run (i = 1 to 16) and total CAPEX includes all key process equipment items for the units
serves as the response value. The operating variables that influence shown in Fig. 3 of the Ling et al. [14]. The OPEX includes fixed general
absorption efficiency were optimized by graphing the SNR values under maintenance costs, including labour, non-income government taxes, and
optimal conditions across different levels of the operating parameters. general insurance. The variable operating costs include cooling water
This model utilized MATLAB and Minitab software for all calculations. costs. The latter incorporates operating costs for the solvent absorption
system and, therefore, includes the cost of the solvent. The energy costs

10
M. Afkhamipour et al. Journal of Industrial and Engineering Chemistry xxx (xxxx) xxx

Table 6 The variable cost is based on the annual operating time of the process
Itemised set up cost factors. and the utilities cost (raw material, electricity, etc). In this work, it is
Cost elements value assumed the process to operate 300 days a year and 18 h a day, which
give a total operating hour of 5400 hr/year. Here, the electricity cost
Instrumentation (C) 15 % TEC
Piping (D) 20 % TEC will mainly be due to the liquid pump requirement, which can be
Electrical (E) 7 % TEC determined as follows [50]:
Total installed cost (F) A+B+C+D+E
Start-up costs (G) 8 % TIC L×g×ρ×z
P= (41)
Engineering (H) 5 % TIC ∊
Owners’ costs (I) 7 % (F + G + H)
Engineering, procurement, construction, and owner’s cost (J) F+G+H+I where L points to the solvent flowrate, g is acceleration gravity, ρ is the
Project contingency (K) 10 % J solvent density, z refers to the distance from the solvent storage to the
CAPEX J+K
bottom of the column assuming 10 m distance from solvent storage to
the top of the reactor, and ∊ is the pump efficiency which is considered
for the capture facility are also part of the variable operating costs. 75 %.

Capital cost Results and discussions

The capital cost comprises the expenses associated with process
equipment and installation, as detailed in the subsequent sections. This section investigates the findings from thermodynamic and rate-
based modelling and explores the results of sensitivity analysis and
Process equipment cost process optimization.
The equipment costs, including absorption and bubble vessels,
pump, and storage tank, are attained by James et al. [44]. The total Thermodynamic model results
capital investment was the sum of the equipment costs multiplied by the
installation factor [44] (see Table 5). Due to the inaccessibility of in- For CO2 capture with rate-based models, the CO2 partial pressure in
formation regarding equipment supplier and manufacturer data, several the 1DMA2P-MEA solution is crucial for the mass transfer driving force.
references have been developed to estimate the equipment cost. Hence, Experimental absorber column data is available for concentrations of 1/
according to James et al. [44], cost is associated with the type and size of 5 M 1DMA2P to MEA, and Ling et al. [14] have also reported equilib-
the equipment. The vessel weight can be determined as follows: rium data for this concentration of 1DMA2P-MEA solution. The activity
Shell mass = π × D × h × tw × ρ (40) coefficient and interaction parameters, given by Eqs. (15) and (17), were
used to minimize interactions in the liquid phase in conjunction with
where D, h, tw and ρ are the column diameter, length, thickness (taken as experimental data provided by Ling et al. [14]. Regression data from the
5 mm for corrosive process), and material density (8000 kg/m3 for 316 D-M model were utilized to calculate the loading of CO2 and the con-
stainless steel). centrations of species produced from the reaction between 1DMA2P-
MEA and CO2. The D-M model’s predicted CO2 loading, and experi-
Set up cost mental CO2 loading are illustrated in Fig. 6 with an AARD of 1.92 %, the
The itemization of total capital cost is justified on conventional model demonstrated a high level of accuracy.
Chemical Engineering methodologies as expounded in Peters et al. [45]. An excellent model can predict results outside the defined range.
Table 6 displays the elements in the capital cost computations. These Therefore, the proposed thermodynamic model offers the advantage of
encompass the overall equipment cost and establishment costs. Addi- predicting CO2 absorption capacity behavior under various partial
tionally, it is assumed that retrofitting the CO2 capture facility exerts no pressure and temperature conditions. According to the 3D plot in Fig. 7,
influence on the operational or capital cost of the extant process plant. CO2 loading exhibits a trade-off between temperature and CO2 partial
The expenses linked with modifying the extant plant are assumed to be pressure. Lower temperatures favor higher CO2 loading. Conversely,
encompassed within the capital and operational costs of the capture higher pressures can improve CO2 loading.
facility.
Rate-Based model results
Operating cost
Operating costs are composed of variable and fixed costs. Variable A comparison between predicted and experimental CO2 mole frac-
operating costs include electricity and solvents. Fixed operating costs tion profiles is shown in Fig. 8. The predictions are based on the model
include tax, maintenance, insurance, supervision, and labor. Table 7 employing correlations from case 2 in Table 1. In the study done by Ling
provides details. et al.[14], 24 runs are available, of which runs 4, 5,19, and 24 were
utilized for validation and model accuracy. According to Fig. 8, excellent
Table 7 results of model relative to the experimental data were obtained. In this
Itemised operating cost factors. figure, run 4 exhibits a CO2 concentration profile ranging from 0.15 at
Operating cost Item Cost Ref. the column bottom to 0.0015 at the top. Similarly, run 5 shows a range
of 0.12–0.0076 from bottom to top. Also, validation results for runs 19
Fixed operating Labour 80,000 $/person/ year [46]
cost Supervision 15 % of operating labour
and 24 are presented in Fig. S1 (supporting information).
Maintenance and 6 % of capital cost Fig. 9 depicts the operational and equilibrium CO2 concentration
repairs profiles throughout the column. The ongoing CO2 absorption by the
Operating supplies 15 % of Maintenance and amine solution within the column results in the consumption of CO2; this
repairs
phenomenon leads to a higher driving force for CO2 mole fraction at the
Tax and insurance 2 % of capital cost
Administrative costs 20 % of operating labour bottom of the column in comparison to the top. Consequently, the CO2
Variable operating MEA 1981 $/m3 [47] removal efficiency is enhanced as the driving force diminishes along the
cost Water 0.02 $/m3 [48] height of the column. The driving force for mass transfer progressively
Electricity 0.13 $/kWh [48] weakens from the bottom to the top of the packed column, ultimately
1DMA2P 2655 $/m3 [49]
reaching equilibrium.

11
M. Afkhamipour et al. Journal of Industrial and Engineering Chemistry xxx (xxxx) xxx

Fig. 6. Validation of thermodynamic model for CO2-H2O-MEA-1DMA2P system.

Fig. 7. Prediction of equilibrium CO2 loading in MEA-1DMA2P system at different CO2 Partial pressures and temperatures.

Fig. 10 presents H2O concentration profiles along the absorber col- chemical reactions that consume it.
umn in liquid and gas phases. Unfortunately, experimental data for H2O Fig. 11 illustrates the profiles for concentrations of the MEA and
concentration along the column were unavailable. The analysis assumes 1DMA2P solutions along the absorber column’s height. From a height of
a dry feed gas stream entering the absorber column. The CO2 capture 1.25 m to approximately 0.2 m, the amine concentration shows slight
reaction with the amine solution, along with the temperature increase in variation. The zone between 0.2 m from the bottom and the column
the bottom section of the column, raises the water’s partial pressure. outlet (rich amine) has the highest mass transfer accompanied by re-
Consequently, the water mole fraction throughout the packing column actions. So, driving force occurs, and the amine concentration decreases
shows a corresponding rise. Subsequently, upon contact with the amine to 0.026 (mole fraction). Therefore, the amine is consumed through
solution along the column, it reaches a constant saturation level. For the reactions and converted into species with higher concentrations of
liquid phase water, recognized as a solvent in the amine system, its MEAH+, 1DMA2PH+, and A critical aspect of the kinetic model involves
concentration decreases from top to bottom in the column due to utilizing the species concentrations within the 1DMA2P-MEA-H2O-H2O

12
M. Afkhamipour et al. Journal of Industrial and Engineering Chemistry xxx (xxxx) xxx

Fig. 8. Validation of rate-based model based on the experimental data for runs 4 and 5 [14].

Fig. 9. Prediction of the model for operating and equilibrium concentrations.

system. The concentrations of critical ionic species in the CO2-1DMA2P- greater the CO2 loading becomes. This is due to CO2 absorption by the
MEA-H2O system are predicted and visualized in Fig. 12. Concentrations amine solution, which consumes CO2, leading to a corresponding rise in
of OH− and H+ ions, as well as CO3 2− and MEACOO− salts, have been the concentrations of ionic species MEA+, 1DMA2PH+, and HCO3 − in
disregarded, focusing solely on predicting the concentrations of MEAH+, the rich amine solution.
1DMA2PH+, and HCO3 − . As observed, moving down the column, the Fig. 13 illustrates the CO2 loading profile within the packed column.

13
M. Afkhamipour et al. Journal of Industrial and Engineering Chemistry xxx (xxxx) xxx

Fig. 10. Prediction of the model for H2O mole fraction in liquid and gas phases.

Fig. 11. Prediction of the model for amine concentrations.

Fig. 12. Prediction of the model for ionic concentrations.

The CO2 loading starts at 0.26 at the column inlet and progressively absorption is strongest at the bottom of the column, where the gas and
increases as the amine solution absorbs CO2 while flowing downwards. liquid are furthest from equilibrium. This leads to a rise in the temper-
As the solution reaches the bottom, the CO2 capture reaction reaches its ature of the liquid as it travels down the column, absorbing heat from the
maximum efficiency. exothermic reaction between CO2 and the amine solution. The signifi-
Fig. 14 reveals the temperature gradient between the gas and liquid cant temperature difference at the column bottom heats the gas phase up
phases as they progress along the column. The driving force for CO2 to 0.1375 m due to a chemical reaction between CO2 and amine

14
M. Afkhamipour et al. Journal of Industrial and Engineering Chemistry xxx (xxxx) xxx

Fig. 13. Prediction of the rate-based model for CO2 loading.

Fig. 14. Prediction of the model for temperature profiles.

Fig. 15. Prediction of the model for gas and liquid flow rates.

solution. Subsequently, a decrease in absorption along the column from to bottom as CO2 is absorbed.
bottom to top leads to a decrease in temperature gradient, and the Fig. 16 explores how varying the inlet gas temperature affects the
profiles reach an equilibrium state. temperature profiles of the gas phase throughout the column. As a result
Fig. 15 illustrates the gas and liquid flow rate profiles within the of the exothermic reactions occurring primarily in the bottom region of
packed column. The gas flow rate exhibits a decreasing trend as it as- the absorber, the temperature bulge occurs at this section of the absorber
cends the column. However, a distinct bulge appears at the bottom column. At TG = 298 K, the bulge temperature is approximately 318 K,
section due to the combined effects of reaction conditions and the rise in while at TG = 323 K, it increases to 324. Therefore, with an increase in
temperature. Conversely, the liquid flow rate steadily increases from top TG, the bulge temperature also rises, but at lower temperatures, the

15
M. Afkhamipour et al. Journal of Industrial and Engineering Chemistry xxx (xxxx) xxx

Fig. 16. Prediction of the rate-based model for different inlet gas temperature to the absorber column.

Fig. 17. Prediction of the rate-based model for different inlet liquid temperature to the absorber column.

Fig. 18. The impact of amine flow rate on the CO2 removal efficiency.

16
M. Afkhamipour et al. Journal of Industrial and Engineering Chemistry xxx (xxxx) xxx

range of temperature variations is greater than at higher temperatures. Fig. 20 explores how the amount of CO2 entering the absorber (inlet
For instance, at TG = 298 K, the temperature variation range is from 298- concentration based on mole fraction) affects CO2 capture efficiency.
313 K, whereas at TG = 323 K, it changes from 324-313 K. The TG curve The results show that as the CO2 concentration increases, the amine
initially increases from the bottom of the column for all three temper- solvent becomes less effective at absorbing it, leading to a decrease in
atures. As the gas and liquid phases reach similar temperatures at the top overall CO2 capture efficiency within the column. Fig. 21 illustrates the
of the column, the driving force for CO2 absorption diminishes, leading CO2 removal efficiency profiles for different column pressure values,
to a decrease in absorption. All three profiles converge to a similar value showing that as column pressure rises, removal efficiency increases.
with an increase in column height. However, the process at high pressures also leads to increased gas
Fig. 17 investigates how the initial temperature of the inlet liquid compression costs, so considering the CO2 capture process working at
influences the liquid’s temperature profile throughout the column under atmospheric pressure, it is better to optimize other operational
various operating temperature conditions. As anticipated, a direct cor- conditions.
relation exists between the inlet and outlet liquid temperatures.
Consequently, the outlet liquid temperature rises proportionally with Sensitivity analysis results
the increase in inlet liquid temperature. It should be noted that the heat
exchange between the gas and liquid phases and the heat produced from The findings regarding the CO2 mole fraction distribution along the
the reaction between CO2 and the solvent results in a relatively constant column are presented. Fig. 22 displays various CO2 concentration pro-
temperature level from the top to the bottom of the column. Subse- files in the gas phase corresponding to cases 1–3 of the effective surface
quently, from a height of 0.4, the temperature gradually increases, area correlations. Compared to cases 1 and 3, case 2 significantly
exhibiting the most significant temperature changes and gradients in the matches the experimental data observed along the column. The best
lower section of the absorber. result was observed with the Hanley and Chen [27] correlation.
Fig. 23 also illustrates how various mass transfer correlations affect
Parametric study results the CO2 mole fraction along the column. As can be seen, the Bravo et al.
[26] correlation has a more significant impact on the CO2 concentration
Fig. 18 shows how varying amine flow rates impact the efficiency of profile compared to other correlations presented in Table 3.
CO2 removal at different heights along the column. Fig. 18 reveals that Fig. 24 presents the influence of kinetic correlations on the distri-
raising the amine flow rate through the column enhances mass transfer bution of CO2 mole fractions along the column’s height. According to the
and CO2 capture efficiency. However, exceeding the optimal amine flow obtained results, all eight cases (1–8) demonstrated excellent agreement
rate may reduce the contact time between phases, mass transfer, and with the experimental data throughout the column, and it cannot be
CO2 removal efficiency. Since amine solutions are costly, an excessive definitively concluded which correlation is better since the kinetics does
increase may not be economically viable. Moreover, excessively raising not significantly affect the rate-based modelling in this system.
the amine flow rate can lead to flooding within the column, resulting in a
significant increase in solvent regeneration energy consumption, leading Statistical optimization results
to a decrease in absorption capacity. Therefore, optimizing this crucial
input parameter for the column is necessary to achieve economic effi- The primary distinction between the classical method, univariate
ciency and obtain the optimal point for improving the overall system analysis on the CO2 removal efficiency and the analysis of the simulta-
performance. neous impact of multiple operational parameters lies in the number of
Fig. 19 shows that a hotter liquid (higher temperature) means the simulation runs necessary. The classical method was employed solely to
column captures less CO2. This is because the reaction between CO2 and assess the influence of highly significant parameters, as detailed in
the amine solution releases heat (exothermic), and a hotter liquid section 3–3. This section delves into the concurrent effects of various
already has less ability to absorb CO2. This leads to a reduction in the operational parameters, aiming to minimize the number of required
efficiency of CO2 removal. In an exothermic reaction, the temperature simulation runs and optimize the process. According to the flowchart
should decrease to improve the reaction progress. Therefore, the presented in Fig. 5, the predicted results of the simultaneous effects of
gas–liquid temperature is an essential parameter for CO2 removal effi- operational variables of the column on CO2 absorption efficiency, using
ciency that needs to be optimized. The maximum absorption efficiency implementation of the orthogonal array design matrix L16 in a rate-
throughout the column can be achieved by controlling this temperature. based model, have been calculated Table 8. SNR data were obtained

Fig. 19. The impact of amine temperature on the CO2 removal efficiency.

17
M. Afkhamipour et al. Journal of Industrial and Engineering Chemistry xxx (xxxx) xxx

Fig. 20. The impact of CO2 mole fraction on the CO2 removal efficiency.

Fig. 21. The impact of absorber column pressure on the CO2 removal efficiency.

Fig. 22. Sensitivity analysis of effective surface area on the distribution of CO2 mole fractions along the absorber column’s height.

18
M. Afkhamipour et al. Journal of Industrial and Engineering Chemistry xxx (xxxx) xxx

Fig. 23. Sensitivity analysis of mass transfer coefficient on the distribution of CO2 mole fractions along the absorber column’s height.

Fig. 24. Sensitivity analysis of kinetic correlations on the distribution of CO2 mole fractions along the absorber column’s height.

with Minitab software. The absorption efficiency and SNR for each row include degrees of freedom, sum of squares, mean square, F-statistic, and
of this table have been computed and listed accordingly in the last two p-values for all factors. Table 10 reveals that all operational parameters
columns of Table 8. The range of variables for optimization is listed in with a p-value below 0.05 statistically influence CO2 absorption effi-
Table S5. ciency. Therefore, in this study, the influence of all parameters is
Table 9 identifies the operating conditions that maximize CO2 ab- considerable. It is worth noting that if the P values for all parameters are
sorption efficiency within the absorber column. These optimal condi- greater than 0.05, a new OAD is required for the modelling-optimization
tions correspond to the entries with the highest SNR values. According to framework (indicated by the phrase “No” in the flowchart).
the outcomes of this table, the optimal points for the CO2 mole fraction In Table 11, the CO2 removal efficiency is predicted using the
in the inlet gas, amine flow rate, amine temperature, and CO2 loading optimal points of operational parameters with a rate-based model and
are determined to be 16 kPa (level 4), 13 m3/m2.h (level 3), 318 K (level statistical optimization method. The analysis confirms that the optimal
4), and 1 mol CO2 /mol amine (level 1), respectively. In Fig. 25, the absorption efficiency using the rate-based model and Taguchi method is
values of SNR for absorption efficiency are plotted vs. the levels of predicted to be 99 % and 100 %, respectively, with only a one-percent
operational parameters, where the maximum values indicate the variation.
optimal values of these parameters. Based on the difference between the
minimum and maximum values of SNR in Table 9, the operational fac- Economic assessment results
tors are ranked.
Table 10 presents a comparison of key parameters influencing CO2 According to the data provided in Tables 5-7, the CAPEX is $607,145
capture efficiency after optimizing and ranking. These parameters and the OPEX is $4,765,901 for the studied CO2 capture scaled-up

19
M. Afkhamipour et al. Journal of Industrial and Engineering Chemistry xxx (xxxx) xxx

Table 8 process, highlighting the significant financial investment required for

Orthogonal array design (L16) and CO2 absorption efficiency. the implementation and maintenance of this technology. The economic
Flow Temperature CO2 CO2 CO2 SNR indicator used in this paper to estimate the cost of capture is $ per tonne
rate (K) loading partial removal of CO2 avoided. The cost of capture ($ per tonne avoided) is the carbon
(m3/ (mol CO2/ pressure efficiency credit per tonne avoided that would be required as income to match the
m2.h) mol (kPa)
present value of capital and operating costs for the capture project. The
amine)
cost of CO2 avoided is calculated using a discounted cash flow analysis
3 298 0.1 4 72.36393 37.19044 that considers the total project costs (capital and operating) and the net
3 308 0.2 8 70.17032 36.92307
3 313 0.3 12 55.60699 34.90259
CO2 avoided [51]:
3 318 0.4 16 73.77283 37.35793
Present value project costs
8 298 0.2 12 97.20877 39.75411 Cost of CO2 captured =
8 308 0.1 16 99.14314 39.92525 Present value CO2 captured
∑n
8 313 0.4 4 97.01867 39.73711 (Ki + Oi /(1 + d)i )
8 318 0.3 8 99.38075 39.94605 = ∑n i=1 i
(42)
13 298 0.3 16 99.88068 39.98963 i=1 (CO2 captured)i /(1 + d)
13 308 0.4 12 99.80148 39.98274
13 313 0.1 8 99.95583 39.99616 where Ki and Oi are the real capital and operating costs ($ million) in the
13 318 0.2 4 99.9577 39.99632 ith year, d is the discount rate (% pa), n is the total project life, and CO2
18 298 0.4 8 95.59958 39.60912 avoided is the annual amount of CO2 avoided in million tonnes. The
18 308 0.3 4 85.52055 38.64141
18 313 0.2 16 92.25244 39.29956
capital and operating costs of CO2 capture are the incremental costs of
18 318 0.1 12 99.99948 39.99996 building the capture plant at an existing industrial facility. As discussed
by Allinson et al. [52], this correlation has many advantages. These
include the ability for capital expenditure to be incurred at any time
Table 9
throughout the project life. Similarly, operating costs and the amount of
SNR response values (optimal points in bold). CO2 avoided can vary. Several simplifying assumptions are used in this
section. The capital cost is spread over 2 years with a breakdown of 40 %
Level Amine flow Temperature CO2 CO2 partial
rate loading pressure
and 60 %. The operating cost and amount of CO2 avoided remains
constant over the 25-year project life.
1 36.59 39.14 39.28 38.89
The cost of CO2 capture using a mixture of MEA + 1DMA2P aqueous
2 39.84 38.87 38.99 39.12
3 39.99 38.48 38.37 38.66 solution was estimated as 40.35 $/tonne of CO2 which is acceptable
4 39.39 39.33 39.17 39.14 compared to the literature [53–55] and other solvents and processes for
Delta 3.40 0.84 0.91 0.48 CO2 capture. Notably, this cost can vary according to the various oper-
Rank 1 3 2 4 ating conditions, such as solvent flow rates, gas flow rate, and gas
concentration.

Conclusions

Our findings reveal that the D-M as thermodynamic model

Fig. 25. Main effects plot for SNR.

20
M. Afkhamipour et al. Journal of Industrial and Engineering Chemistry xxx (xxxx) xxx

Table 10
ANOVA for SNR related to packed column operating parameters.
Factor Degree of freedom Sum of Adjusted sum of Adjusted mean Factor variance to error variance Percentage
(DF) squares squares square ratio (F-ratio) contribution (P)

Amine flow 3 30.4871 30.4871 10.1624 40.9900 0.0060

Temperature 3 1.5966 1.5966 0.5322 2.1500 0.2730
CO2 loading 3 1.9802 1.9802 0.6601 2.6600 0.2210
Partial 3 0.6132 0.6132 0.2044 0.8200 0.5610
pressure
Residual error 3 0.7437 0.7437 0.2479 ​ ​
Total 15 35.4208 ​ ​ ​ ​

the work reported in this paper.

Table 11
Factor levels for predictions.
Appendix A. Supplementary data
Amine Temperature CO2 CO2 CO2 CO2
flow (K) loading partial removal removal
Supplementary data to this article can be found online at https://doi.
rate (mole pressure efficiency efficiency
(m3/ CO2/ (kPa) by by org/10.1016/j.jiec.2024.12.011.
m2.h) mole Minitab MATLAB
amine) References
13 318 0.1 16 100 % 99 %
[1] B. Aghel, S. Janati, S. Wongwises, M.S. Shadloo, Review on CO2 capture by
blended amine solutions, Int. J. Greenhouse Gas Control 119 (2022) 103715.
[2] M. Shamsi, E. Naeiji, S. Rooeentan, B.F. Shahandashty, P. Namegoshayfard,
accurately predicts equilibrium data within various pressure, tempera- M. Bonyadi, Proposal and investigation of CO2 capture from fired heater flue gases
ture, and loading ranges. The experimental and model data showed a to increase methanol production: A case study, Energy 274 (2023) 127375.
deviation of 1.92 %. Consequently, the D-M model was integrated into [3] T.N. Borhani, M. Wang, Role of solvents in CO2 capture processes: The review of
selection and design methods, Renew. Sustain. Energy Rev. 114 (2019) 109299.
the rate-based model to achieve the results presented. The sensitivity [4] M.T. Mota-Martinez, J.P. Hallett, N. Mac Dowell, Solvent selection and design for
analysis was performed to identify key factors influencing the concen- CO 2 capture–how we might have been missing the point, Sustainable, Energy Fuel
tration and temperature profiles predicted by the rate-based model. The 1 (2017) 2078–2090.
[5] T.N. Borhani, M.R. Abbasi, M. Hosseinpour, M. Salimi, M. Afkhamipour, E. Oko, K.
finding of sensitivity analyses indicated that mass transfer relationships
S. Campbell, N. Kahllaghi, CO2 absorption-desorption cycles: Progress, gaps, and
have a higher impact on the CO2 concentration profiles throughout the future, Carbon Capture Sci. Technol. 13 (2024) 100325.
column compared to kinetic correlations. Examined the effect of indi- [6] A.V. Rayer, K.Z. Sumon, T. Sema, A. Henni, R.O. Idem, P. Tontiwachwuthikul, Part
5c: Solvent chemistry: solubility of CO2 in reactive solvents for post-combustion
vidual operating parameters on the CO2 removal efficiency, holding all
CO2, Carbon Manage. 3 (2012) 467–484.
others constant. For this effect, increasing the amine solution flow rate [7] Y. Liang, H. Liu, W. Rongwong, Z. Liang, R. Idem, P. Tontiwachwuthikul,
and column pressure enhances the CO2 removal efficiency. Conversely, Solubility, absorption heat and mass transfer studies of CO2 absorption into
increasing the amine solution temperature and inlet CO2 gas concen- aqueous solution of 1-dimethylamino-2-propanol, Fuel 144 (2015) 121–129.
[8] P. Luis, Use of monoethanolamine (MEA) for CO2 capture in a global scenario:
tration decreases the efficiency. A statistical analysis of the results Consequences and alternatives, Desalination 380 (2016) 93–99.
revealed that the amine solution flow rate, CO2 loading, amine solution [9] M. Shamsi, E. Naeiji, M. Vaziri, S. Moghaddas, K. Elyasi Gomari, M. Naseri, M.
temperature, and partial pressure of inlet CO2 gas had the greatest in- Bonyadi, Optimization of energy requirements and 4E analysis for an integrated
post-combustion CO2 capture utilizing MEA solvent from a cement plant flue gas,
fluence on the removal efficiency, in decreasing order. The statistical Environ., Devel. Sustain., (2023) 1-27.
analysis and the rate-based models yielded highly consistent outcomes [10] M. Afkhamipour, E. Seifi, A. Esmaeili, M. Shamsi, T.N. Borhani, Comparison of CO2
for optimized removal efficiency, with a discrepancy of only 1 %. The absorption in DETA solution and [bmim]-[PF6] using thermodynamic and process
modelling, Fuel 356 (2024) 129607.
rate-based model, coupled with the D-M thermodynamic model, pro- [11] H. Gao, Z. Wu, H. Liu, X. Luo, Z. Liang, Experimental studies on the effect of
vides a reliable framework for predicting the performance of the amine tertiary amine promoters in aqueous monoethanolamine (MEA) solutions on the
processes. This investigation pinpoints critical operating parameters for absorption/stripping performances in post-combustion CO2 capture, Energy Fuel
31 (2017) 13883–13891.
maximizing CO2 capture efficiency with this promising solvent. Finally,
[12] R. Zhang, Z. Liang, H. Liu, W. Rongwong, X. Luo, R. Idem, Q. Yang, Study of
the cost analysis of CO2 capture was estimated and found to be accept- formation of bicarbonate ions in CO2-loaded aqueous single 1DMA2P and MDEA
able compared to the literature and other solvents and processes for CO2 tertiary amines and blended MEA–1DMA2P and MEA–MDEA amines for low heat
of regeneration, Ind. Eng. Chem. Res. 55 (2016) 3710–3717.
capture.
[13] M. Afkhamipour, M. Mofarahi, A. Rezaei, R. Mahmoodi, C.-H. Lee, Experimental
and theoretical investigation of equilibrium absorption performance of CO2 using a
CRediT authorship contribution statement mixed 1-dimethylamino-2-propanol (1DMA2P) and monoethanolamine (MEA)
solution, Fuel 256 (2019) 115877.
[14] H. Ling, S. Liu, H. Gao, H. Zhang, Z. Liang, Solubility of N2O, equilibrium
Morteza Afkhamipour: Writing – original draft, Supervision, Soft- solubility, mass transfer study and modeling of CO2 absorption into aqueous
ware, Resources, Project administration, Investigation, Formal analysis, monoethanolamine (MEA)/1-dimethylamino-2-propanol (1DMA2P) solution for
post-combustion CO2 capture, Sep. Purif. Technol. 232 (2020) 115957.
Data curation, Conceptualization. Mohammad Shamsi: Writing –
[15] T.N. Borhani, S.A. Nabavi, D.P. Hanak, V. Manovic, Thermodynamic models
original draft, Validation, Software, Resources, Investigation, Formal applied to CO2 absorption modelling, Rev. Chem. Eng. 37 (2021) 931–957.
analysis. Ebad Seifi: Writing – review & editing, Resources, Methodol- [16] P.C. Meier, Two-parameter Debye-Hückel approximation for the evaluation of
ogy, Formal analysis. Abbas Hekmatiyan: Writing – review & editing, mean activity coefficients of 109 electrolytes, Anal. Chim. Acta 136 (1982)
363–368.
Methodology, Investigation. Meisam Ansarpour: Writing – review & [17] K. Thomsen, P. Rasmussen, Modeling of vapor–liquid–solid equilibrium in
editing, Methodology, Investigation. Tohid N.Borhani: Formal anal- gas–aqueous electrolyte systems, Chem. Eng. Sci. 54 (1999) 1787–1802.
ysis, Investigation, Methodology, Writing – review & editing. [18] M. Afkhamipour, M. Mofarahi, A modeling-optimization framework for assessment
of CO2 absorption capacity by novel amine solutions: 1DMA2P, 1DEA2P, DEEA,
and DEAB, J. Clean. Prod. 171 (2018) 234–249.
[19] A.T. Zoghi, M. Shokouhi, Measuring solubility of hydrogen sulphide in aqueous
Declaration of competing interest blends of N-methyldiethanolamine and 2-((2 aminoethyl) amino) ethanol and
correlating by the Deshmukh-Mather model, J. Chem. Thermodyn. 100 (2016)
The authors declare that they have no known competing financial 106–115.
interests or personal relationships that could have appeared to influence

21
M. Afkhamipour et al. Journal of Industrial and Engineering Chemistry xxx (xxxx) xxx

[20] R.H. Weiland, T. Chakravarty, A.E. Mather, Solubility of carbon dioxide and [37] M. Afkhamipour, M. Mofarahi, Experimental measurement and modeling study on
hydrogen sulfide in aqueous alkanolamines, Ind. Eng. Chem. Res. 32 (1993) CO2 equilibrium solubility, density and viscosity for 1-dimethylamino-2-propanol
1419–1430. (1DMA2P) solution, Fluid Phase Equilib. 457 (2018) 38–51.
[21] A. Benamor, M.K. Aroua, Modeling of CO2 solubility and carbamate concentration [38] G.F. Versteeg, W.P. Van Swaaij, Solubility and diffusivity of acid gases (carbon
in DEA, MDEA and their mixtures using the Deshmukh-Mather model, Fluid Phase dioxide, nitrous oxide) in aqueous alkanolamine solutions, J. Chem. Eng. Data 33
Equilib. 231 (2005) 150–162. (1988) 29–34.
[22] M. Haji-Sulaiman, M.K. Aroua, A. Benamor, Analysis of equilibrium data of CO2 in [39] E.O. Agbonghae, K.J. Hughes, D.B. Ingham, L. Ma, M. Pourkashanian, A semi-
aqueous solutions of diethanolamine (DEA), methyldiethanolamine (MDEA) and empirical model for estimating the heat capacity of aqueous solutions of
their mixtures using the modified Kent Eisenberg model, Chem. Eng. Res. Des. 76 alkanolamines for CO2 capture, Ind. Eng. Chem. Res. 53 (2014) 8291–8301.
(1998) 961–968. [40] A. Nezamloo, Measurements of the Molar Heat Capacity and the Molar Excess
[23] X. Chen, F. Closmann, G.T. Rochelle, Accurate screening of amines by the wetted Enthalpy for Various Alkanolamines in Aqueous Solutions, The University of
wall column, Energy Procedia 4 (2011) 101–108. Regina (Canada), 2013.
[24] M. Llano-Restrepo, E. Araujo-Lopez, Modeling and simulation of packed-bed [41] J. Antony, S. Warwood, K. Fernandes, H. Rowlands, Process optimisation using
absorbers for post-combustion capture of carbon dioxide by reactive absorption in Taguchi methods of experimental design, Work Study 50 (2001) 51–58.
aqueous monoethanolamine solutions, Int. J. Greenhouse Gas Control 42 (2015) [42] T.N. Borhani, E. Oko, M. Wang, Process modelling and analysis of intensified CO2
258–287. capture using monoethanolamine (MEA) in rotating packed bed absorber, J. Clean.
[25] J.A. Rocha, J.L. Bravo, J.R. Fair, Distillation columns containing structured Prod. 204 (2018) 1124–1142.
packings: a comprehensive model for their performance. 2. Mass-transfer model, [43] Chemical Engineering Online. (n.d.). Chemical engineering news and resources.
Ind. Eng. Chem. Res. 35 (1996) 1660–1667. Retrieved November 16, 2024, from https://www.chemengonline.com.
[26] J.L. Bravo, Mass transfer in gauze packings, Hydrocarb. Process. 64 (1985) 91–95. [44] R. James, W. Couper, J. Fair, S. Walas, Costs of individual equipment, Chem.
[27] B. Hanley, C.C. Chen, New mass-transfer correlations for packed towers, AIChE J Process Equipment (2005) 663–669.
58 (2012) 132–152. [45] M.S. Peters, K.D. Timmerhaus, Plant design and economics for chemical engineers,
[28] M.H. De Brito, U. Von Stockar, A.M. Bangerter, P. Bomio, M. Laso, Effective mass- McGraw-Hill International, 2018.
transfer area in a pilot plant column equipped with structured packings and with [46] L.C. Brown, G.E. Besenbruch, R. Lentsch, K.R. Schultz, J. Funk, P. Pickard, A.
ceramic rings, Ind. Eng. Chem. Res. 33 (1994) 647–656. Marshall, S. Showalter, High efficiency generation of hydrogen fuels using nuclear
[29] M. Afkhamipour, M. Mofarahi, Comparison of rate-based and equilibrium-stage power, in, General Atomics, San Diego, CA (United States), 2003.
models of a packed column for post-combustion CO2 capture using 2-amino-2- [47] H. Ali, N.H. Eldrup, F. Normann, R. Skagestad, L.E. Øi, Cost estimation of CO2
methyl-1-propanol (AMP) solution, Int. J. Greenhouse Gas Control 15 (2013) absorption plants for CO2 mitigation–method and assumptions, Int. J. Greenhouse
186–199. Gas Control 88 (2019) 10–23.
[30] M. Afkhamipour, M. Mofarahi, Rate-based modeling and sensitivity analysis of a [48] H. Ali, N.H. Eldrup, F. Normann, V. Andersson, R. Skagestad, A. Mathisen, L.E. Øi,
packed column for post-combustion CO2 capture into a novel reactive 1-dime- Cost estimation of heat recovery networks for utilization of industrial excess heat
thylamino-2-propanol (1DMA2P) solution, Int. J. Greenhouse Gas Control 65 for carbon dioxide absorption, Int. J. Greenhouse Gas Control 74 (2018) 219–228.
(2017) 137–148. [49] Alibaba. (n.d.). Alibaba: Global trade starts here. Retrieved November 16, 2024,
[31] S. Kadiwala, A.V. Rayer, A. Henni, Kinetics of carbon dioxide (CO2) with from https://www.alibaba.com.
ethylenediamine, 3-amino-1-propanol in methanol and ethanol, and with 1-dime- [50] The Engineering ToolBox. Available online https://www.engineeringtoolbox.com/
thylamino-2-propanol and 3-dimethylamino-1-propanol in water using stopped- .
flow technique, Chem. Eng. J. 179 (2012) 262–271. [51] M.T. Ho, G.W. Allinson, D.E. Wiley, Comparison of MEA capture cost for low CO2
[32] C.-H. Hsu, M.-H. Li, Densities of aqueous blended amines, J. Chem. Eng. Data 42 emissions sources in Australia, Int. J. Greenhouse Gas Control 5 (2011) 49–60.
(1997) 502–507. [52] G. Allinson, V. Nguyen, CO2 geological storage economics, in: Greenhouse Gas
[33] Z. Idris, L. Ang, D.A. Eimer, J. Ying, Density measurements of unloaded and CO2- Control Technologies-6th International Conference, Elsevier, 2003, pp. 615-620.
loaded 1-dimethylamino-2-propanol at temperatures (298.15 to 353.15) K, [53] J.C. Farla, C.A. Hendriks, K. Blok, Carbon dioxide recovery from industrial
J. Chem. Eng. Data 60 (2015) 1419–1425. processes, Clim. Change 29 (1995) 439–461.
[34] Y. Zhuang, C. Fan, M. Cao, C. Han, Z. Li, F. Yang, X. Liu, Y. Chen, Density and [54] D. Gielen, CO2 removal in the iron and steel industry, Energ. Conver. Manage. 44
excess molar volume of ternary mixtures monoethanolamine+ ethanol/n-propanol (2003) 1027–1037.
+ H2O at (293.15–323.15) K and 101 kPa, J. Mol. Liq. 314 (2020) 113624. [55] M. Slater, E. West, C. Mariz, Carbon dioxide capture from multiple flue gas sources,
[35] D. Fu, L. Wang, X. Tian, Surface thermodynamics of DMA2P, DMA2P-MEA and Greenhouse Gas Control Technologies-6th International Conference, Elsevier
DMA2P-PZ aqueous solutions, J. Chem. Thermodyn. 107 (2017) 79–84. (2003) 103–107.
[36] A. Sorvari, Annual Transactions of the Nordic Rheology Society, Vol 21, Nordic
Rheology Society (2013).

22