Loss Analysis in Radial Inflow Turbines for Supercritical CO2 Mixtures

Abstract Recent studies have indicated the potential of CO2-mixtures to lower the cost of concentrated solar power plants. Based on aerodynamic and cost considerations, radial inflow turbines (RIT) can be a suitable choice for small to medium sized sCO2 power plants (about 100 kW to 10 MW). The aim of this paper is to quantify the effect of doping CO2 on the design of RITs. This is achieved by comparing the 1D mean-line designs and aerodynamic losses of pure sCO2 RITs with those of three sCO2 mixtures containing tetrachloride (TiCl4), sulphur dioxide (SO2), and hexaflourobenzene (C6F6). Results show that the optimal turbine designs for all working fluids will have similar rotor shapes and velocity diagrams. However, factors such as the clearance-to-blade-height ratio, turbine pressure ratio, and the difference in the viscosity of the fluids cause variations in the achievable turbine efficiency. Once the effects of these factors are eliminated, differences in the total-to-static efficiency amongst the fluids may become less than 0.1%. Moreover, if rotational speed limits are imposed, then greater differences in the designs and efficiencies of the turbines emerge amongst the fluids. It was found that limiting the rotational speed reduces the total-to-static efficiency in all fluids; the maximum reduction is about 15% in 0.1 MW CO2 compared to the 3% reduction in CO2/TiCl4 turbines of the same power. Among the mixtures studied, CO2/TiCl4 achieved the highest performance, followed by CO2/C6F6, and then CO2/SO2. For example, 100 kW turbines for CO2/TiCl4, CO2/C6F6, CO2/SO2, and CO2 achieve total-to-static efficiencies of 80.0%, 77.4%, 78.1%, and 75.5% respectively. Whereas, the efficiencies for 10 MW turbines are 87.8%, 87.3%, 87.5%, and 87.2%, in the same order.