Optimization on high-pressure side of pump-turbine runner based on high efficiency and stability criterion via multi-objective genetic algorithm method

Abstract

Pumped storage power plants are doomed to play a more important role in peak-valley shifting hence the demand for operation stability is gradually regarded as an equal important criterion as high efficiency. In our paper, new concepts, namely “swept,” “bowed (lean),” and “twisted” are introduced to systematic innovative design the geometry of high-pressure side (HPS) of a pump-turbine runner. Thereafter, a multi-objective optimization process, which consists of design of experiment (DoE), meta model generation, multi-objective genetic algorithm (MOGA), self-organization map (SOM) and takes both high efficiency and stability discipline into account is produced. The final optimization plan is selected and testified numerically. Compared to the original runner, the efficiency is improved from 91.7% to 91.9% at pump mode and improved from 92.3% to 92.8% at turbine mode. Moreover, the S margin is increased from 89.34° to 90.23° at small GVO and increased from 79.56° to 80.29° at large GVO. Besides, hump unsteady region is completely eliminated. Moreover, the flow characteristic comparation is conducted based on local hydraulic loss rate (LHLR) method. For design points, the optimized runner obviously decreases the hydraulic loss in hub and middle part of runner domain for turbine mode and slightly decrease the hydraulic loss in stay vane region for pump mode. For unsteady characteristics, HPS can better adjust the hydraulic loss distribution in corresponding discharge operating points, largely decreasing S unsteady characteristic in turbine mode and eliminating hump unsteady characteristic in pump mode. We believe that the methods proposed in our paper can bring the design of hydraulic machinery to a new level.