Abstract

The amount of studies related to centrifugal compressors operating with supercritical carbon dioxide (S-CO2) is growing over the years, especially due to the development of modern carbon capture and storage (CCS) and Enhance Oil Recovery (EOR) technologies driven by the COP21 countries to assist in the reduction on carbon dioxide emission. Other important advantages of the use of centrifugal compressors are their low capital costs and high efficiency with a simple layout. The present research proposes an optimization procedure to minimize the required power of an existing centrifugal compressor by using an in-house code (developed on MATLAB software) which uses the meanline method (with high fidelity loss models) coupled to Non-dominated Sorting Genetic Algorithm (NSGA-II). The real-gas formulation (Span and Wagner equations of state) is used to calculate the thermo-physical properties of the supercritical carbon dioxide near the critical point. The comparison of the flow patterns between optimized compressor and conventional compressor is performed by using Computational Fluid Dynamics. The preliminary results from the optimization procedure indicated that the required power of the optimized compressor was 12% lower and the isentropic efficiency was 9% higher than the conventional compressor. Finally, the geometrical parameters of the optimized compressor are presented and the main conclusions are pointed out.