The severity of flood disasters in underground powerhouses is increasing with the expansion of pumped storage power stations, and the current emergency drainage system of pumped storage power stations cannot meet the large discharge and high-lift demands during periods of abnormal water inflow. This paper proposes a phased parallel emergency drainage method for pumped storage power stations to address the problem of ineffective and slow drainage during underground facility floods. Furthermore, a multi-objective optimization model for emergency drainage is established to solve critical parameters in the drainage plan.

This paper proposes a multi-objective optimization method for emergency drainage in pumped storage power stations. First, based on the engineering characteristics of the underground powerhouses of pumped storage power stations, the emergency drainage process is divided into three stages: early, middle, and late. Second, emergency drainage methods are developed for each stage. During the early stage of drainage, for a small amount of accumulated water in the area above the top floor of the generator, submersible pumps and drainage pipelines are installed in the entrance tunnel to discharge the water to the exit of the tunnel. The middle stage of drainage focuses on draining the accumulated water from the generator floor; multiple drainage vehicles are connected in series to drain the water to the surface. Late-stage drainage is aimed at the accumulated water below the generator floor. A small volume floating pump is used to extract water from each layer, which is then drained to the generator floor. Here, a water storage tank is set up to temporarily store the water extracted by the floating pump; the water is drained in parallel through the entrance tunnel and the maintenance and leakage drainage pipes. Finally, the proposed drainage method is mathematically modeled to minimize drainage time and operating costs as objectives. Furthermore, the non-dominated sorting genetic algorithm-Ⅱ (NSGA-Ⅱ) is employed to obtain the optimal solution for the drainage flow rate at each stage.

The phased series-parallel comprehensive emergency drainage method proposed in this paper employs movable drainage equipment, enabling rapid deployment by onsite emergency rescue teams. In the case of a pumped storage power station, 34 drainage schemes are optimized, all of which meet the constraint conditions. Among them, the shortest emergency drainage time is about 52 h, and the minimum operating cost for drainage equipment is approximately 1.38 million RMB, providing a basis for selecting drainage schemes and procuring or customizing drainage equipment.

The optimal solution set of drainage flow rates corresponding to each drainage stage can be obtained by updating only the self-attribute data of the power station in the model input using the NSGA-Ⅱ-based multi-objective optimization model for emergency drainage of pumped storage power stations. This method reduces emergency drainage time and equipment operating costs while improving the emergency response level of flood accidents in pumped storage power stations.