(56g) Computer-Aided Solvent and Process Design for Carbon Capture

The design of novel solvents for process improvements has stemmed much interest in recent years (Bardow et al., 2010, Folić et al., 2007, Oyarzun et al., 2011). Solvents are used in many industrial processes for separation of useful products from a feed. Consequently improvements in solvent performance are sought after for a more efficient and cost effective process. The properties desired in a solvent vary from process to process and include selectivity, volatility, flammability, toxicity, viscosity, heat capacity and reaction kinetics. Often a trade-off has to be reached between the desired properties as not all desired properties may be achievable within one solvent.

In order to optimize system performance, both process and solvent parameters must be optimized. The simultaneous optimization of a process and solvent is a challenging task. Although process parameters may be continuously optimized, solvent properties are discrete and therefore pose a major hurdle in the optimization process. Complex mixed-integer nonlinear programming (MILP) methods have to be implemented in order to optimize the nonlinear process problem and choose solvent parameters from a discrete solvent property space. This problem has traditionally been circumvented by decoupling the solvent and process optimization processes and by optimizing the solvent to specific desired property targets. The Computer Aided Molecular Design- Continuous Molecular Targeting (CAMD-CoMT) framework proposed by Bardow et al is an effective way of tackling the integrated process and solvent design problem and does not require decoupling the two optimization problems. Instead, the optimization problem assumes a hypothetical solvent with continuous solvent parameters based on a physically based thermodynamic model. The perturbed-chain-statistical associating fluid theory (PC-SAFT) equation of state is used to calculate the solvent properties based on five solvent parameters.

In the current paper, we calculate optimal process parameters and determine two optimal solvents (pure and binary mixture) for a carbon capture process in contrast to the pure solvent search in presented in (Bardow et al., 2010). The process is modelled in ASPEN Plus (Aspentech, USA), while an optimization is executed in Matlab (Mathworks, USA) to calculate the vapour liquid equilibrium properties of the solvent using the PC-SAFT thermodynamic model. This simultaneous process and solvent optimization enhances the performance of the system considerably, and has good potential for application across a myriad of other process systems.

BARDOW, A., STEUR, K. & GROSS, J. 2010. Continuous-Molecular Targeting for Integrated Solvent and Process Design. Industrial & Engineering Chemistry Research, 49,2834-2840.

FOLIĆ, M., ADJIMAN, C. S. & PISTIKOPOULOS, E. N. 2007. Design of solvents for optimal reaction rate constants. AIChE Journal, 53,1240-1256.

OYARZUN, B., DARDOW, A. & GROSS, H. 2011. Integration of process and solvent design towards a novel generation of CO2 absorption capture systems. Energy Procedia, 4, 282-290.