Modelling Studies on Reactive Absorption of Carbon Dioxide in Monoethanolamine Solution from Flue Gas in Coal Based Thermal Power Plants

In this paper the detailed theoretical investigation on absorption of carbon dioxide, from flue gas in coal based thermal power plants has been presented. For absorption studies, monoethanolamine solution is considered as a solvent. The mathematical model for the absorption column has been developed by considering thin film model approach. Unified method is used for an overall estimation of carbon dioxide absorption. The carbon dioxide concentration profile at a given stage, using the thin film layer model, has been predicted at 298 K and 318 K. From the results it is noted that carbon dioxide concentration decreases from interface concentration at equilibrium to a minimum of 0 kmol/m up to a distance of ±2 micrometers. Overall estimation of carbon dioxide absorption has been carried out using the unified model approach. The total amount of carbon dioxide absorbed in absorption column is estimated to be 95.60% of the inlet carbon dioxide with 30 trays, L/G ratio of 8.5 and carbon dioxide flow rate of 95.74 kmol/m. The results revealed that reactive absorption is very effective in absorbing carbon dioxide into monoethanolamine solvent.


INTRODUCTION
India is amongst the top four Carbon dioxide (CO2) producing countries with China, US, and countries under European Union, it produced about 2,341,000 kt of CO2 in 2014 [1].In India, 38% of total CO2 emission is due to electricity production [2].71% of the total electricity supplied in India is generated using thermal power plant of which 62% is fulfilled using coal based thermal power plants [3].This number is huge because India requires a large amount of electricity which is produced by power plants present across the country.
The CO2 emission from thermal power plants in India was estimated to be 4.98 × 10 8 metric tons in the year 2009-2010 [4].India has about 116 coal based thermal power plants with near about 429 units.
The total power capacity from all these plants comes around 167,707.88 MW per year [5].The impact of this huge CO2 emission is rising of global temperature, i.e. global warming, which is posing great threat not only to the environment but to human life as well [6].
Therefore, there is a need to study post combustion treatment technologies for CO2 removal.Among the existing technologies for CO2 removal, absorption is one of the most suitable technology [7].In this paper the authors have carried out detailed theoretical investigation on reactive absorption of CO2 using Monoethanolamine (MEA) solution as the solvent.The method involved in this process is reactive absorption, i.e. the absorption of CO2 from the gaseous stream into MEA accompanied by the reaction where CO2 and MEA combine to form ester (R1HCOO − ) and primary amine (R1NH3 + ) which enhances the mass transfer between the two.In order to understand the state of art in CO2 absorption, the literature survey has been carried out in presented in following section.

Literature studies
Several studies have been carried out over the years on absorption and reactive absorption of CO2 using various solvents such as Methyldiethanolamine (MDEA), MEA, ammonia, Diethanolamine (DEA), blend of MDEA and ethanolamine.Zhang and Chau-Chyun [8] used electrolyte Nonrandom Two-Liquid (NRTL) activity coefficient model to develop a rigorous and thermodynamically consistent representation for the MDEA-water-CO2 system.The model has been validated for predictions of Vapor-Liquid Equilibrium (VLE), heat capacity, and CO2 heat of absorption of the MDEA-water-CO2 system temperature range of 313 K to 393 K. Liu et al. [9] studied absorption of CO2 absorption in ammonia.They concluded that ammonia could be better solvent for CO2 absorption.However, they have mentioned further experimental investigations are necessary.Lawal et al. [10] presented a study on chemical absorption of CO2 in MEA solution based dynamic modeling of the absorber and regenerated columns linked together.They concluded that the model predicts the absorber and regenerator temperature profiles and CO2 profiles very well.Molina and Bouallou [11] investigated kinetics of CO2 absorption into mixed solutions of MDEA and DEA.Their study primarily focused on to optimize the blend composition to capture CO2.
In most of these studies, a single stage CO2 absorption using thin film mass and heat transfer laws have been considered.These studied predicted the distance, from gas-liquid interface to the point at which CO2 get absorbed in the solvent, using either equilibrium stage or rate based model.
A study done using gPROMS showed rate based model gives better prediction than the equilibrium model for CO2 absorption using MEA.It also states that effect of L/G ratio is more sensitive than effect of flow rate change [12] which was verified in this study as well.Another study rate based model was validated using 0.1 MW pilot plant in South Korea which showed good agreement with data except when the L/G ratios taken were low [13].Study of CO2 absorption using aq.MEA using a packed tower has been done which were in agreement with industry scale pilot plant data [14].
Dynamic modeling of CO2 absorption from coal-fired power plants into an aqueous monoethanolamine solution has been carried out using Aspen Tech software [15].The authors have studied the transient simulation of absorption column and the results are validated with experimental data.A detailed a state-of-the-art review on post-combustion CO2 capture with chemical absorption is presented by Wang et al. [16].The authors have concluded that more efforts in future should be directed to reduce energy combustion in post combustion CO2 capture with chemical absorption.Dynamic behavior of coal-fired power plants with post combustion CO2 capture is studied by Wellner et al. [17].The authors used a model-based control strategy based on existing directives for power plants with post combustion CO2 capture.Aboudheir et al. [18], investigated the kinetics of the reaction between CO2 and high CO2-loaded, concentrated aqueous solutions of MEA for the temperature range from 293 to 333 K.They proposed a new termolecular-kinetics model, for CO2-MEA solution, which proved to be better than previously published kinetic models.Zhang and Chen [19] have performed simulation studies with both the rigorous rate-based model and the traditional equilibrium stage model for CO2 absorption with MEA.The model results were validated with the pilot plant data from recently published literature.They concluded that the rate-based model yields better predictions compared to equilibrium stage model.Plaza et al. [20], presented results of CO2 absorption in MEA with a new model that uses a rigorous thermodynamic model from the published literatures.
From above mentioned literature survey it is noted that researchers across the globe are striving to find the unique model for prediction of the performance of absorption column for the CO2 absorption.Authors have used rate based and equilibrium NRTL models and simulated them using various software like gPROMS and ASPEN plus.Most of papers don't provide the data for overall estimations of concentration profile of CO2 (performance of the column) and its variation with L/G ratio and number of stages.
In view of above, this study tried to provide the theoretical investigation of the absorption tower to predict the performance of the single stage as well as the absorption column for CO2 absorption.Further, it provides the stage concentration profiles of CO2 at various temperatures, inlet flow rates and liquid to gas ratios.The liquid phase CO2 concentration on a particular stage is estimated using interfacial mass transfer concepts and by applying Henry's law at the gas-liquid interface.Further the prediction of overall estimate of the CO2 absorption has been done using unified method.The overall objective of this study is to predict and understand the dynamics of absorption column for the removal of CO2.The outcome of the study can be used for improvements in the operations of the commercial absorption columns for CO2 absorption as well as for the designing of new column for the same.

MODEL DEVELOPMENT
CO2 containing outlet stream of desulphurization unit is the input to the absorption column.This stream consists of CO2 mixed with compounds like Sulphur dioxide (SO2) and Nitrogen oxides (NOx).SO2 also has a high tendency to react with aqueous MEA solution and interfere in the CO2 absorption and therefore a desulphurization unit has to be installed before the absorption column to bring the concentration of SO2 down enough so that it wouldn't have much effect on CO2 absorption.A maximum of 10 ppmv of SO2 concentration and less than 20 ppmv for NOx concentration [18] is acceptable for the absorption column to work efficiently.
A column with multiple trays is considered in which the treated flue gas is moved counter currently with MEA solution.The CO2 present in the flue gas reacts with the solvent in a pseudo first order kinetics where carbamate formation takes place [19].This reaction enhances the mass transfer from gas phase to the liquid phase, thus the absorbed CO2 is separated from the flue gas as it moves upwards in the absorption column.To find the amount of CO2 which gets absorbed horizontally (on a stage) undergoing mass transfer forming a thin film between gas and liquid phase is considered under the following assumptions: • The gas phase offers no resistance to mass transfer, the resistance offered is by the liquid film only; • No vaporization of solvent, since the vapor pressure of MEA is very less at the temperature rage considered for this work; • Henry's law is applicable at the interface of the gas and liquid phase; • Temperature is assumed to be constant throughout the complete process; • Gas liquid equilibrium is assumed at each stage.N-stage equilibrium model is developed using material balance equation, equilibrium relations and summation equations as follows: Material balance: Equilibrium relation: Summation equation:

Reaction involved
The reaction of CO2 with MEA involves the following reactions [19] into ester and amine:

Parameters used
Parameters involved in this model are diffusivity of CO2 and Nitrogen dioxide (N2O) in MEA and water, Henry's constant at gas-liquid interface, density of aqueous MEA, rate and equilibrium constant of the reactions.The relations for above said parameters are provided below.Diffusivity equation.The diffusivity coefficient of CO2 in MEA was estimated using the N2O analogy developed by Ko [20]: The equations of the diffusivity used were given by Versteeg and Vanswaaij [21]: The values of constants in diffusivityrelations are taken from Ko et al. [20], and are given in Table 1.Density relations.The density of aq.MEA solution was taken as 1,013 kg/m 3 using the relations from Weiland et al. [22].
Solubility relation (Henry's law).The Henry's law constant is estimated using N2O analogy by Clarke [23]: These constants are function of temperature Penttila et al. [24], and are provided in Table 2.There have been many discussions regarding the order of rate of reaction with respect to both CO2 and MEA.In most of literature studies, the order of reaction with respect to CO2 is assumed to be 1 but the order of reaction with respect to MEA was found to be in the range 1 to 2. However, in most of the studies reaction rate constant has been taken as 1 st order with respect to MEA.Thus, the overall order was found to be 2 [25].But in this study the authors have considered pure MEA with no vaporization in gas phase due to very less vapor pressure of MEA (40 pascal at 20 °C), hence in this study a pseudo first order reaction is considered.The rate constant was found using following relation given by Aboudheir et al. [26]: k = 4.61 × 10 9 e (−4,412/r) ( Equilibrium constant was calculated with the help of following relation [27,28]: The values of these constants are given in Table 3.

Development of model using unified method
In this method, N numbers of stages are considered for absorption of CO2 using MEA as the solvent.Flue gas and MEA are flowing counter currently in the tower where the flue gas is fed from the bottom most tray (Vn), where as pure MEA is added from the top tray.At each tray, fractional amount of CO2 gets absorbed into MEA and flue gas which is now CO2 lean moves to the next upper plate.The process continues up to the N th stage, post which the lean flue gas moves into the stack for removal.The schematic of i th stage is shown in Figure 1.
The overall mass balance equation gives: The component balance gives: where Vi decreases by NaAVm amount at each stage because of CO2 absorption and its reaction with monoethanolamine.Equilibrium relation: Figure 1.N-stage equilibrium model The process depends on the stripping factor (Si), (Vi Ki)/Li, which makes the equation: Yi+1,jVi+1,j + yi,jVi,j [(Li,j / Gi,jKi,j) −1] + yi−1, jVi−1,j [(Li,j / Gi,jKi,j)] + Fi = 0 (14) or: where Si,j is the stripping factor, i.e: Si,j = (Gi,j × Ki,j) / Li,j Solution procedure.The model equations are solved using the equilibrium matrix method (Thomas algorithm) to estimate overall absorption CO2 in the tower.It is simplification of Gaussian elimination method which is used to solve system of equations, using forward and backward substitutions.For this purpose the MATLAB software has been used.The K value in the equilibrium relation is calculated using the Henry's constant at the interface of the two medium.

RESULTS AND DISCUSSION
The simulation studies for the absorption column were carried outto find the mole fraction of CO2 absorbed at each stage.The overall absorption of CO2 in absorption column is also estimated.

Parameters values
The parameters values used, at two temperatures 298 K and 318 K, in simulation studies are mentioned in Table 4.The parameters viz., rate constant, Henry's constant, diffusion coefficient of CO2 and MEA are used for this study.Variation of rate constant with temperature.As the temperature increases the rate constant also increases which in turn enhances the reaction rate of CO2 with MEA solution and thus increases absorption.
Variation of Henry's constant with temperature.The value of Henry's constant increases at the interface of gas and liquid with increase in temperature.
Variation of diffusion coefficient with temperature.Diffusion coefficients also increase with temperature and hence increase the diffusion of the acidic gas into the liquid solution.

Stage-wise variation in concentration profile
The effect of change in the inlet gas temperature and inlet CO2 composition, on the concentration profile of CO2 from the gas-liquid interphase to bulk MEA at a given stage is presented in Figure 2 and Figure 3, respectively.These studies were carried out at constant inlet CO2 flow rate of 75.74 kmol/m 3 and L/G ratio of 8.5.Effect of temperature on concentration profile (4 = 0.8). Figure 2 presents the effect of temperature on CO2 profile from gas-liquid interphase to bulk liquid phase.From Figure 2 it is noted that 2.35 × 10 −2 kmol/m 3 of CO2 gets absorbed completely into the aqueous MEA solution at a distance of 2.1 × 10 −6 m from gas-liquid interface at temperature 318 K.However, at 298 K, 1.76 × 10 −2 kmol/m 3 of CO2 gets absorbed completely into aq.MEA solution at a distance of 1.5 × 10 −6 m from the gas-liquid interface.
Thus, from above results it is noted that with increase in temperature the concentration of CO2 increases at the interface which gets absorbed within film thickness.This is due to the fact that with increase in temperature, the solubility of the acidic gas increases in the solvent.Hence, using the thin film layer model, CO2 concentration was found to decrease from interface concentration at equilibrium to a minimum of 0 kmol/m 3 up to a distance within ±2 micrometers.
Effect of inlet CO2 composition on concentration profile (T = 318 K).The Figure 3 shows effect of change in inlet CO2 composition on CO2 profile from gas-liquid interphase to bulk liquid phase.From Figure 3 it is observed that as the inlet CO2 composition in the flue gas is increased from 0.5 to 0.8 the interface concentration of CO2 also increased from 2.35 × 10 −2 kmol/m 3 to 2.75 × 10 −2 kmol/m 3 at that particular stage.However, there is no change in the distance from gas-liquid interface for CO2 concentration to reduce to zero.

Overall estimate
The overall estimate of CO2 absorption in absorption column has been carried out using unified method.Here at each stage, it is considered that a constant amount (Na × A × Vm) of CO2 is getting absorbed due to the reaction of CO2 with MEA solvent.Where Na is the molar flow rate, A is area and Vm is the molar volume at that temperature.The film area is taken as 2.5 m × 0.4 m and molar volume is taken as 24.86 m 3 /kmol using gas law at 318 K.The major component of inlet flue gas is considered to be CO2 as the outlet from the desulphurization unit is rich in CO2 concentration because most of the SO2 content is removed in the desulfurization unit.
The overall percentage of CO2 absorption is calculated by subtracting outlet mole fraction at stage 1 at the top from inlet mole fraction at the bottom and dividing it by inlet mole fraction.
Effect of change in L/G on overall CO2 absorption.Figure 4 present the effect of change in L/G on overall CO2 absorption.The simulation studies are carried out at inlet CO2 composition of varying in the range from 0.85 to 0.95, inlet CO2 flow rate of 75.74 kmol/m 3 , temperature of 318 K and with 30 numbers of theoretical stages.The details of input/output streams composition along with L/G ratio is provided in Table 5.From the figure it is observed that overall CO2 absorption increases from 66.73% to 95.60% with increase in L/G ratio from 2 to 8.5, respectively.This must be attributed to fact that the increase in solvent rate enhances the absorption rate.It is also seen from the Figure 4 that for L/G ratio 8.5, CO2 profile across the column is almost flat as compared to CO2 profile at L/G ratio 2 and 4.5.Effect of change in no of stages on overall CO2 absorption.The effect of change in no of stages on overall CO2 absorption is presented in Figure 5.The simulation studies are carried out at constant inlet CO2 composition of 0.8, inlet CO2 flow rate of 75.74 kmol/m 3 , temperature of 318 K and with L/G ratio of 8.5.It can be noted from Figure 5 that overall CO2 absorption increases as the number of stages are increased from 15 to 30.

CONCLUSIONS
The expansion of economy and population with current technology leads to higher rates of urbanization, industrialization and deforestation, all factors contributing to an increase in CO2 level in the surrounding.So, steps need to be taken in order to reduce this level wherein the first step that has to be taken would be to trap the industrial CO2 emissions in the most efficient and economical way.This can be achieved by post combustion capture of CO2 using absorption columns.
The film model gives an overview as to how the CO2 absorption takes place for a particular tray and the thickness up to which entireamount of CO2 gets absorbed.From the stage wise CO2 concentration profile it is noted that the absorption of CO2 into MEA is very instantaneous.This may be attributed to the fact that reactive absorption is very effective in CO2 absorption.This study also gives us visualization as to how absorption takes place at each tray throughout the column.This study has revealed the important design variables which controls the overall absorption of CO2.
The unified model approach is best suited for overall estimation of CO2 absorption.From the results it is noted that, 95.60% of the inlet CO2 can be absorbed in absorption column with 30 trays, L/G ratio of 8.5 and inlet CO2 flow rate of 95.74 kmol/m 3 .From the results it is also noted that CO2 absorption in the tower is more sensitive to L/G ratio as compared to composition or flow rate changes which is in accordance with literature studies.

Figure 2 .Figure 3 .
Figure 2. Effect of change in temperature on CO2 concentration profile

Figure 4 .
Figure 4. Effect of change in L/G on overall CO2 absorption

Figure 5 .
Figure 5.Effect of change in no of stages on overall CO2 absorption

Figure 6 .
Figure 6.Effect of change in inlet CO2 flow rates on overall CO2 absorption

Table 1 .
Values of constants in diffusivity relations

Table 2 .
Values of constants for Henry's constant relations

Table 3
Operational details.A pure MEA solution is supplied from top of the absorption column counter currently to the gaseous stream with CO2 as a major component.Mathematical model has been developed for this system.Simulation studies are carried out at two temperatures 298 K and 313 K, with CO2 compositions of 0.6 and 0.8 to find CO2 concentration profile across the gas-liquid interface at a given stage.Simulation studies are further carried out to estimate overall absorption of CO2 in absorption column, at different inlet flow rates of CO2: 65.74, 75.74, 85.74 and 95.74 kmol/m 3 , at different L/G ratios: 2, 6.5, 8.5 and at different number of stages 15, 25 and 30.

Table 4 .
Effect of variation of temperature on different parameters

Table 5 .
Input-output streams data