Development of an efficient cross-scale model for working fluid selection of Rankine-based Carnot battery based on group contribution method

Abstract

Rankine-based Carnot battery is promising system with outstanding performances in addressing the challenges of local consumption of renewable energy generation and utilization of low-grade waste heat. A suitable working fluid is fundamental to the Rankine-based Carnot battery cycle and profoundly influences system performance. However, studies on working fluid selection for the Rankine-based Carnot battery are limited to a predefined database. This approach fails to study or design novel working fluids. Therefore, the nine molecular groups used as a set of molecular groups in this paper can cover most organic working fluids in Rankine-based thermodynamic cycle systems. Developing an accurate cross-scale model based on the group contribution method is used in working fluid selection for Rankine-based Carnot battery. Meanwhile, for the sake of testing the accuracy of the proposed model, twenty-one promising organic working fluids were selected to be compared and analyzed with the results calculated by the REFPROP under the typical working conditions of the Rankine-based Carnot battery. The results show that the absolute average relative deviation of the boiling temperature prediction using the multiple linear regression model proposed in this study is only 3.9 %, while the absolute average relative deviation of the critical temperature based on the proposed boiling temperature model is 5 %. Finally, the present work demonstrates excellent accuracy in predicting the thermodynamic performance of this system. The absolute average relative deviation of the coefficient of performance, generation efficiency and power-to-power-efficiency is 4.6 %, 2.0 %, and 6.4 %, respectively.