Solid particle erosion (SPE) poses a significant challenge to the efficient and safe operation of energy recovery turbines, necessitating urgent research and optimization. This paper proposes three optimization schemes (P1, P2, P3) aimed at addressing the issues of particle erosion and low aerodynamic performance in two-stage energy recovery turbines. A multi-objective optimization method is constructed by combining experimental design, a genetic algorithm and CFD gas-solid two-phase flow simulation. Through MOP calculations, the appropriate number of sample points and construction method of response surface were determined for two-stage ERT. The sensitivity analysis indicates that the number of blades significantly affects both the efficiency and erosion rates, whereas the blade profile parameters have a lesser impact. In the first-stage optimization of P1, the optimized blades increase the total-to-total isentropic efficiency from 89.27 % to 90.94 %. The erosion degrees of the first- stage stator and rotor blades decrease by 51.94 % and 64.08 %, respectively. After a comprehensive evaluation of the three schemes, P1 was selected as the best optimization solution because of its significant improvement in efficiency and effective mitigation of the erosion problem. The P2 and P3 schemes, however, have too many optimization variables or objectives, making it difficult to achieve deep collaborative optimization of local erosion resistance and overall aerodynamic performance of two-stage energy recovery turbines.
