Metaheuristic methods applied to the pumps and turbines configuration design of water pumped storage systems

The increasing penetration of fluctuating Renewable Energy Sources (RES), particularly in small isolated power systems, is raising some problems in the operational management of the system. The simpler way to solve the problem is to perform RES curtailment, but this is not the right decision from an environmental point-of-view. In this context, the storage alternative is becoming more and more a cost-effective option. This paper presents an integrated techno-economic model, whose purpose is to come up with the optimal configuration of a Water Pumped Storage System (WPSS). The expected output is the WPSS design configuration, i.e. the pump number and size and turbine number and size, which leads to the maximum Net Present Value (NPV) of the investment project. To reach the proposed objective, the model uses meta-heuristic optimization techniques, namely Particle Swarm Optimization (PSO) and Genetic Algorithms (GA), with a Unit Commitment and Economic Dispatch module (UC + ED) of the thermal-based electrical generation units, in order to calculate the system operational costs. The Terceira Island, Azores archipelago, Portugal, is the case study presented as an application of the developed model. PSO and GA converge to the same results, pointing to an optimal WPSS configuration of 2 pumps of 490 kW and 1 turbine of 500 kW, with a 263 k€ positive NPV, therefore showing the economic and technical feasibility of a WPSS, in this application.