Exploring multi-objective metaheuristics to optimise resource allocation in energy communities

Abstract

In the transition to a sustainable urban model, renewable energy communities where participants act as both producers and consumers have gained substantial importance. In this context, the main challenge lies in the optimal sizing of energy generation and storage units to achieve a balance between economic profitability and independence from the conventional grid. In this study, we define the resource management problem for the initial setup of an energy community, optimising the number of solar panels, wind turbines, and battery capacity per member. Besides, the objective functions considered include the levelized cost of energy (LCOE), self-sufficiency, and self-consumption, evaluated under four different interaction models, from full independence to shared battery use. While metaheuristics are commonly employed in energy management problems, the increasing variety of algorithms complicates the selection of an effective optimisation method. We implemented several multi-objective metaheuristics, including Particle Swarm Optimisation, Marine Predators Algorithm, Whale Optimisation Algorithm, and Equilibrium Optimiser. Additionally, we introduce the Political Optimizer, specifically adapted for multi-objective optimisation, which provides an additional tool for this kind of problems. Results from simulations demonstrate that the Multi-Objective Equilibrium Optimiser with Archive Evolution Path achieved the lowest LCOE, 0.0736 $/kWh, and the highest self-sufficiency rate, 0.8873, with shared excess energy production. When implementing shared storage, the best configurations reached an LCOE of 0.0725 $/kWh, a self-consumption rate of 1.0000 and a self-sufficiency rate of 0.9509. Finally, the largest emission reductions were also observed with shared storage, and reached up to 38.1%.