Mathematical Modelling and Process Simulation of a Membrane Absorption Process for Pre-combustion CO2 Capture
MetadataShow full item record
The greenhouse gas emissions, particularly CO2, are significantly impacting the global climate change. There are strong desires to find energy efficient and economical solutions for CO2 capture from energy production, mainly through post-combustion, pre-combustion and oxy-fuel combustion technologies. Various separation methods are being increasingly investigated, including physical/chemical absorption, adsorption, membrane separations and cryogenic processes. Membrane absorption (membrane contactor) is a relatively new technology employed to separate CO2 from flue gas or syngas. It combines the benefits of both absorption (high selectivity) and membrane separations (high modularity and compactness) to attain improved CO2 capture. Various membrane contactors have been studied in recent years for CO2 separation from flue gas. A new class of solvents known as ionic liquids (IL) has also been applied as CO2 absorbents to improve the separation performance. ILs possess negligible vapour pressure, high absorption capacity, and good chemical and thermal stability with tunable physical and chemical properties. However, membrane contactors for pre-combustion CO2 capture at elevated pressures and temperatures have rarely been reported. Until now, the work from the MCIL-CO2 project by the Memfo group at NTNU has been a major contribution to the study in this application, and the published reports from this project are mainly the experimental study of the membrane absorber. As part of the MCIL-CO2 project, the objective of the present work is to study the membrane contactor process for pre-combustion carbon capture through modelling and process simulation, as well as to analyse the energy and economic feasibility of this process. The mathematical model developed in this work can predict the performance of the membrane contactors and the influence of various operational parameters can also be forecasted and optimized. Screening of ILs liquids were tested to identify better candidates for the pre-combustion process conditions. Initially, the amino acid based ILs, tetrabutylphosphonium glycinate ([P4444][Gly]) blended with low molecular weight polyethylene glycol (PEG 400) was selected to investigate the CO2 absorption capacity and other physical properties such as density, viscosity and thermal stability. In order to find out an optimal absorbent solution, different blending ratios were studied. The concentration of the blend solvent was optimized with respect to CO2 solubility, regeneration efficiency and cyclic capacity. The solubility of CO2 in [P4444][Gly], PEG 400 and their blends of four different concentrations was measured experimentally for a temperature range of 60-140 oC and up to a pressure of 17 bar. The thermal stability of the amino acid IL-PEG 400 blends were measured up to a temperature of 200 oC. The density of these blends was measured for a temperature range of 20-80 oC at ambient pressure, while their viscosity was tested for the temperature ranging from 20 to 90 oC. The results showed an increase in CO2 solubility by increasing the IL concentration in the blend and reduction of solubility by increasing the temperature. The optimum CO2 absorption was determined to be 30 wt.% of [P4444][Gly] in the blend. The regeneration study of 30 wt.% [P4444][Gly]-70 wt.%PEG 400 for three cycles verified their reusability in the process and confirmed that the reaction between [P4444][Gly] and CO2 can be reversed at 140 oC. The parameter fitting of the experimental data using empirical correlations was also performed. The density correlation defined the experimental data very well with an absolute average relative deviation (AARD) of 0.16 % while viscosity was predicted with 3.7 % squared error. The prediction of the equilibrium partial pressure of CO2 in blend was found out with an accuracy of 3-5 %. In addition, mass transfer study and mathematical modelling of membrane contactor for CO2 absorption has been executed. A one-dimensional mathematical model has been developed to evaluate the separation performance of membrane contactor by forecasting the CO2-outlet concentration both in liquid and gas streams. The model accounts for both wetted and non-wetted modes of membrane. For this study, tubular glass membrane contactor was used and 1-butyl-3-methylimidazolium tricyanomethanide ([bmim][TCM]) was selected based on CO2 absorption capacity, viscosity and good thermal stability. Material balances have been applied over the membrane contactor and established in as the form of differential equations. These equation were solved by method of weighted residual and f-solve in the MATLAB. The overall mass transfer coefficient is estimated by resistance in series model, where a new mass transfer resistance term is added to reflect the non-flat concentration profile in the liquid phase. The developed model was also validated with the experimental data with AARD of 8%. Simulation results indicated that the liquid phase resistance contributes 67 % and 44 % to the total mass transfer resistance for non-wetted and wetted modes of membrane, respectively. The resistance that occurred due to considering transport in liquid phase contributes 31 and 20 % for non-wetted and wetted modes of membrane contactor, respectively. Furthermore, the process design and simulations of pre-combustion CO2 capture process using IL in membrane contactor have also been carried out to analyse the energy and cost assessment of this process. Designing this process on a large industrial scale requires scaling up the membrane contactor followed by integrating it in Aspen-HYSYS simulation software. The tubular membrane contactor was replaced with hollow fibre membrane contactor and the developed mathematical model was used to predict the temperatures of both gas and liquid streams other than the CO2-outlet concentrations. Due to unavailability of the membrane absorber model in Aspen-HYSYS, the developed mathematical model was incorporated to Aspen-HSYS by the Cape-Open simulation compiler. The physical properties of the gas phase were estimated from Peng Robinson equation of state and IL absorbent ([bmim][TCM]) by UNIQUAC. A pressure-swing membrane absorption-desorption process was designed to achieve CO2 separation from the shifted syngas. The feed gas mixture containing only CO2 and H2 was considered to simplify the process. The CO2 gas is absorbed in the membrane contactor and then stripped off in a series of flash separators and pressure reducing valves. The lean absorbent is fed back to membrane absorber by booster pumps. The energy evaluation of this process indicated that CO2 absorption at high pressure is less energy intensive compared to post combustion CO2 capture. The specific energy requirement for this process has been found to be 0.74 MJ/kgCO2, which is much lower than that in conventional physical absorbents in packed column. The economic evaluation was also performed by estimation of total capital cost and operating cost. The capital investment for 0.14M tons of CO2 capture per year was estimated to be ~47.4M $, while operating cost was counted 9.04M $. The specific cost per ton of CO2 was calculated to be 87$.