Carbon dioxide capture methods for industrial sources.
MetadataShow full item record
In order to reduce the rate of climate change, particularly global warming, it is imperative that industries reduce their carbon dioxide (CO2) emissions. A promising solution of CO2 emission reduction is Carbon dioxide Capture and Storage (CCS) by sequestration, which involves isolating and extracting CO2 from the flue gases of various industrial processes, and thereafter burying the CO2 underground. The capture of CO2 proved to be the most challenging aspect of CCS. Thus, the objective of this research was to identify the most promising solution to capture CO2 from industrial processes. The study focussed on capturing CO2 emitted by coal power plants, coal-to-liquids (CTL) and gas-to-liquids (GTL) industries, which are common CO2 emitters in South Africa. This thesis consists firstly of an extensive literature review detailing the above mentioned processes, the modes of CO2 capture, and the various CO2 capture methods that are currently being investigated around the world, together with their benefits and drawbacks in terms of energy penalty, CO2 loading, absorption rate, capture efficiency, investment costs, and operating costs. Modelling, simulation, and pilot plant efforts are also described. The study reviewed many CO2 capture techniques including solvent absorption, sorbent capture, membrane usage, hydrate formation, and newly emerging capture techniques such as enzyme based systems, ionic liquids, low temperature cryogenics, CO2 anti-sublimation, artificial photosynthesis, integrated gasification steam cycle (IGSC), and chemical looping combustion The technique of solvent absorption was found to be the most promising for South African industries. Vapour-liquid-equilibrium (VLE) measurements of solvent absorption using amine blends were undertaken, using blends of methyl-diethanol amine (MDEA), diethanol amine (DEA) and water (H2O) with composition ratios of 25: 25: 50 wt% and 30: 20: 50 wt% respectively, and with CO2 and N2 gases at CO2 partial pressures of 0.5 to 10.5 bar. Experiments were conducted under system pressures of 5 to 15 bar and temperatures of 363.15 and 413.15 K, using a static analytic apparatus. CO2 liquid loading results were analysed and discussed. The experimental data were regressed in Matlab (R2009b) using the Posey-Tapperson-Rochelle model and the Deshmukh-Mather model. The Matlab programmes are presented along with the regressed binary interaction and model parameters. The accuracy of model predictions are discussed. Thereafter an Electrolyte-NRTL model regression and simulation of the absorption process was conducted using Aspen Plus V 7.1. for flue gas compositions, solvent compositions, temperature, and pressure conditions similar to that of process operating conditions. CO2 loading, design factors, CO2 recovery, and CO2 purity results were analysed and compared where appropriate, with experimental results. Finally a general preliminary energy efficiency and cost analysis was conducted based on the simulation results. The main conclusions reached are that the amine solvent blend containing 25:25:50 wt% of MDEA:DEA:H2O, produced higher CO2 loadings for its respective system conditions than other solvents studied and those found in literature. However, absorption of CO2 was found to be highly dependent on system temperature and pressure. The Deshmukh-Mather model provided higher accuracy than the Posey-Tapperson-Rochelle model, producing CO2 loading predictions with a relative error not exceeding 0.04%, in 1.5 to 3 minutes using a dual core processor. Aspen absorption simulations provided significantly lower CO2 loading results than those experimentally obtained, due to the low contact time achieved and higher temperature dependence in the proposed absorption process. Process improvements were highlighted and implemented to increase CO2 recovery and purity. Energy penalty values were found to be higher than those found in literature, but room for process and design improvement was identified and recommendations were given. Investment cost estimates were found to be justifiable and within reason. Limitations of the simulation were also identified and discussed.