How to Quickly Evaluate the Thermodynamic Performance and Identify the Optimal Heat Source Temperature for Organic Rankine Cycles?

Abstract

Organic Rankine cycle (ORC) is a promising technology to convert low- and medium-temperature energy into power. Identifying the optimal working fluids and heat source temperature are always the focuses in the ORC field. This paper presents a new methodology to evaluate the thermodynamic performance of ORC with different working fluids and identify the optimal heat source temperature. Initially, the parameterization model is developed to characterize the working fluids by thermodynamic property parameters including critical temperature (Tc), critical pressure (pc), acentric factor (ω), and ideal gas isobaric heat capacity (cp0). Subsequently, the simultaneous optimization of thermodynamic property parameters and cycle parameters is conducted to obtain the thermodynamic performance limits of simple and regenerative ORCs at six typical geothermal heat source temperatures. By comparing the thermodynamic performance limits of ORC under different heat source temperatures, the optimal heat source temperature is identified. Then, ten commonly used working fluids are selected as reference working fluids, and the thermodynamic property parameters comparisons between reference and ideal working fluids, which can be characterized by the optimized thermodynamic property parameters, are investigated. Finally, multiple linear regression models are developed to evaluate the thermodynamic performance. The numerical differences of thermodynamic property parameters between the ideal reference and reference working fluids are chosen as initial variables, while the thermal efficiency and volumetric power output are used as thermodynamic performance indicators. The results show that the optimal heat source temperature is 250 °C, which is independent of cycle configuration. The thermodynamic performance of ORCs can be evaluated accurately by the multiple linear regression models. The maximum relative error of the multiple linear regression models is 3.02%. Moreover, Tc is the most dominant thermodynamic property parameter.