Multi-objective optimization and performance evaluation of variable nozzle radial-inflow turbine in organic Rankine cycle

As the core component of the organic Rankine cycle, the variable nozzle radial-inflow turbine can achieve superior off-design performance by adjusting its nozzle angle. However, traditional methods treat turbine design and working fluid selection independently, requiring significant computational resources and time. In this study, a turbine optimization model is developed by integrating its design model into the thermo-economic model of the organic Rankine cycle. Furthermore, an encoding method is proposed to transform the working fluid into a decision variable, enabling simultaneous determination of the optimal working fluid and aerodynamic parameters of the turbine through multi-objective optimization. Additionally, the off-design performance of the optimized turbine under different nozzle angles and operating conditions is evaluated using computational fluid dynamics and the entropy production method. The results indicate that R600 is the optimal working fluid for the turbine, with an optimal rotational speed of 25 254 rpm, a nozzle outlet angle of 18.92°, and a nozzle inlet radius of 0.1289 m. Under these optimal conditions, the system achieves a net power output of 358 kW, with a thermal efficiency of 12.12%, a levelized energy cost of 0.0328 $/kWh, and a turbine dimension factor of 3.85 × 10−7 m/W. Under off-design conditions, flow losses in the turbine primarily originate from the rotor, with passage vortices near the suction side of the blades playing a dominant role. The turbine achieves a maximum efficiency of 81.5% at a nozzle angle of 26° and an outlet pressure of 0.2 MPa. The off-design performance of the turbine can be significantly improved by selecting a suitable nozzle exit angle according to the operating conditions.