Assessment of pressure swing adsorption as CO2 capture technology in coal-fired power plants

摘要

Coal-based power generation is responsible for a significant share of CO2 emissions ona global scale. Technologies to drastically reduce coal carbon footprint are critical formeeting mitigation targets. Absorption, whether chemical or physical depending on theprocess framework, is commonly regarded as the most mature technology in thiscontext. Nevertheless, absorption suffers from some drawbacks, such as high energyrequirements and corrosion of process equipment. Adsorption is considered as apromising alternative, with potential for reducing energy penalty, environmental impactand cost of CO2 capture.The main objective of this thesis is to assess the viability of a process relying onadsorption, i.e. pressure swing adsorption (PSA), as CO2 capture technology in coalfiredpower plants.In order to get a comprehensive overview on the prospects of PSA, different cases wereconsidered. Post-combustion CO2 capture was studied by integrating a PSA unit into anadvanced supercritical pulverized coal (ASC) plant. Pre-combustion CO2 capture wasstudied by integrating a PSA unit into an integrated gasification combined cycle (IGCC)plant. Proper designs for these process frameworks were defined, taking into accountcharacteristics, requirements and constraints of the systems. PSA is a discontinuousprocess, made of different steps undergone by each column of a PSA train. A dynamicmodel was built, based on material, energy and momentum balances. The developeddynamic model was then linked to the steady-state model of the power plant, byexploiting appropriate process scheduling and the cyclic steady state (CSS) condition ofthe PSA process (a condition in which the process transient behavior becomes steadythroughout different cycles). The resulting composite model allowed performingsimulations and analyses on a system level.The post-combustion case (ASC + PSA) showed competitive energy performance. Thenet electric efficiency obtained was 34.8%, whereas the reference plant without CO2capture had 45.1%. The CO2 capture requirement was also fulfilled with more than 90%CO2 sequestrated. A comparison with chemical absorption - performing with 34.2% netelectric efficiency - confirmed the competitiveness of PSA. A serious issue ascertainedconcerned the required footprint of the PSA unit. A first estimation suggested thenecessity of more than 260 adsorption columns for processing the entire flue gascoming from the boiler. The feasibility of PSA in the post-combustion case appearedless attractive because of the number of vessels needed. The pre-combustion case (IGCC + PSA) returned good results for all the performanceindicators investigated. A comparison with physical absorption showed that PSA isslightly outperformed in terms of energy efficiency (36.2% versus 37.1%, with thereference plant without CO2 capture having 47.3%), CO2 recovery (86.1% versus 90.6%)and footprint. However, the performance gap was evaluated to be rather small, thusadditional investigations were carried out in this process framework.Improvements in the performance of the pre-combustion case were sought byconsidering two domains, the process and the adsorbent material. Several possibleprocess configurations were analysed and a range of results obtained. Improved energyperformance could be obtained but to the detriment of the CO2 separation performanceand vice versa. Modifications in the adsorbent material properties (attempting tosimulate different adsorbents and/or advancements in the materials) showed asignificant influence not only on the gas separation process but on the whole plantperformance. The utilization of improved adsorbents demonstrated the capability to givea substantial contribution to close the gap with absorption, though it may not besufficient. None of the cases studied succeeded to fully match absorption-basedperformance both in terms of energy and CO2 capture efficiency. Further, an approachto exploit possible synergies between the two studied domains and realize the fullpotential of PSA in this framework was outlined. It consisted of tuning the materialproperties according to a specific process configuration. The results achieved wereencouraging as net electric efficiencies up to 37.1% were obtained without drasticdecrease in the CO2 capture efficiency.The knowledge developed in the pre-combustion process framework suggested a furthercase which was believed interesting for PSA. An IGCC plant was defined coproducingpower and ultrapure H2 with CO2 capture. The system is of interest both because itallows capturing CO2 and because differentiating the plant products can beadvantageous in terms of flexible operation. Two novel process configurations weredeveloped, entirely relying on PSA. The first consists of two consecutive PSA stages(Two-train PSA), while the second configuration carries out both CO2 separation and H2purification within a single PSA stage (One-train PSA). Both these configurationssucceeded to provide a varying power-to-hydrogen output ratio - the net power outputcould be reduced from 346 MW to 300 MW by increasing the ultrapure H2 throughput -with a constant coal feed to the gasifier and retaining plant efficiency on a good level.The common process design for an IGCC coproduction layout encompasses absorptionfor CO2 capture and PSA for H2 production. With regard to that, a comparative analysisseems to confirm the expected advantages brought by the utilization of PSA as the onlygas separation technology. A higher integration level could be achieved, allowingsignificant energy savings. The assessment of PSA in this framework was concluded tobe promising and worth further analyses. Summing up, it was demonstrated that PSA can be successfully integrated in coal-firedpower plants as CO2 capture technology. However, the analyses carried out showed alsothat PSA is generally outperformed by absorption in an overall evaluation taking intoaccount different performance indicators. Potentials and limits of the technology havebeen highlighted and recommendations for optimizing the performance have beenoutlined. The knowledge developed can be useful to address further work on PSAtechnology, especially in those specific frameworks (e.g. coproduction of power and H2)where PSA can reach competitiveness.