Multilevel optimization for integrated CO2 capture systems under techno-policy transitions

초록

Membrane CO2 capture processes are an environmentally benign and low-carbon option removing CO2 from major point sources. However, the subambient pressure of CO2 product typically requires energy intensive liquefaction for transportation, utilization, or storage. Consequently, the optimal CO2 purity and recovery from the membrane process cannot be determined without considering the downstream liquefaction process, highlighting the need for integrated optimization. This study proposes a two-level nested optimization framework for an integrated CO2 membrane separation and liquefaction process treating approximately 40,000 Nm3/h of flue gas originating from natural gas combined cycle. At the upper level, the optimized CO2 purity and recovery from the membrane process were determined. At the lower level, a deterministic global solver optimizes the membrane superstructure and a Bayesian optimization algorithm optimizes the liquefaction process. The framework optimizes the integrated process under six current and future scenarios involving advanced membrane properties and carbon emission penalties. This optimization simultaneously determines scenario dependent CO2 recovery and purity by evaluating technoeconomic tradeoffs between membrane specific selectivity and permeability and liquefaction process cost. The results show that the economically optimal membrane purity and recovery points shift with carbon emission penalties and membrane properties. With the optimum futuristic membrane scenario at a carbon emission penalty of 160 USD/t-CO2, the membrane process achieves 93.0% CO2 purity and 77.7% CO2 recovery. The integrated process recovers 72.3% of the CO2 in the flue gas. This represents a significant shift toward higher CO2 recovery compared to the current baseline of 93.2% purity and 20.0% recovery in the membrane process under near-term membranes. Futuristic membranes reduce the total annualized cost by 4.5-8.6% compared to near-term benchmark scenario. This proposed twolevel nested optimization framework provides essential insights for designing integrated capture processes as technology and policy evolve.

키워드