Improved loss evaluation models for sCO2 centrifugal compressors based on qualitative and quantitative analysis

The supercritical CO2 (sCO2) Brayton cycle has garnered considerable attention for its compact equipment, low compression-power, and high efficiency. The centrifugal compressor serves as the cycle's pivotal component. However, conventional one-dimensional (1D) loss models previously developed for air compressors could not be suitable for sCO2 compressors. In this paper, a loss quantification strategy based on the entropy-generation is presented to optimize empirical coefficients of 1D loss models. Firstly, flow losses in both air and sCO2 compressors are predicted using empirical loss models. Secondly, the distribution of entropy-generation rate in various regions of both air and sCO2 compressors is quantified by three-dimensional (3D) numerical simulations with validated by experimental data. Thirdly, air compressors' empirical coefficients are correlated with the distribution of entropy-generation rate. Finally, modified coefficients for sCO2 compressors are proposed using the entropy-generation rate distribution and the presented correlation. Results show that for the air compressor, relative errors between 1D and 3D loss-weight coefficients remained within 5 % for most loss mechanisms, while for the sCO2 compressor, the incidence loss coefficient deviates by up to −22 % caused by overestimating losses when using empirical coefficients. After modifying coefficients of loss models for sCO2 compressors, the prediction error of total-pressure-ratio decreased from about 1.8 % to 0.6 %. Presented coefficients and strategy are suitable for sCO2 compressors to improve prediction accuracy and extend operating range of empirical loss models.