Optimization of two-stage membrane system for CO2 capture from flue gas

ANA MARISA ARIAS; MIGUEL MUSSATI; NICOLÁS J. SCENNA; SERGIO F. MUSSATI

Zagreb

Encuentro; XXIV CROATIAN MEETING OF CHEMISTS AND CHEMICAL ENGINEERS; 2015

Croatian Society of Chemical Engineers, Croatian Chemical Society

Membranes have been successfully applied for several gas and/or liquid separations and purifications. They involve low cost and high energy-efficiency. However, only few articles address with the study of the potential application of membranes for CO2 capture.This work deals with the optimization of a two-stage membrane system for CO2 capture (Fig. 1). Using a NLP model, the optimization problem stated as follows:Optimization problem.To maximize {CO2 total recovery}Subject to: hj(x) = 0 , j= 1,...,Jgk(x) ≤ 0 , k= 1,...,KAM1+ AM2 ≤ PAMFigure 1. Two stage membrane system.where x denotes the vector of continuous variables (pressures, compositions, flow-rates, membrane areas, power consumption) which are simultaneously optimized. hj(x) refers to equality constraints (momentum and mass balances) and gk(x) represents a set of inequality constraints defined in order to circumscribe a feasible operating region. Given the flue gas specification (composition, temperature and flow-rate), the optimal design task involves the determination of the operating conditions and how the total membrane area PAM should be allocated in the first (AM1) and second stage (AM2) to maximize the CO2total recovery. Optimizations are performed for a wide range of PAM. The model is implemented into General Algebraic Modeling System (GAMS), which is a high-level modeling system for mathematical programming and optimization. A detailed discussion of the optimization results is presented.