ABSTRACT A multi‐period mixed‐integer non‐linear programming model is proposed to optimally allocate battery energy storage systems (BESSs) in networks with photovoltaic generation. The solution methodology employs modified grey wolf optimisation (MGWO) to address the discrete part of the problem and a mathematical programming to handle its continuous part, resulting in a hybrid approach labelled as MGWO‐H. The methodology considers the possibility of using BESSs to provide both active and reactive power. The proposal is based on generation and network load forecasts and is solved considering a 24‐h time horizon to minimise energy losses. To demonstrate the applicability and effectiveness of the proposed model, several tests were carried out on two benchmark test systems. The results obtained using the proposed MGWO‐H are compared with two additional proposals: a hybrid approach similar to MGWO‐H that uses the original GWO, labelled as GWO‐H, and a multi‐period optimal power flow. In the two benchmark test systems of 33 and 141 buses, the results indicated that power losses were reduced by up to 30% and 13%, respectively, when BESSs provide only active power, and by 42% and 16.74% when BESSs provide both active and reactive power.
Silva et al. (Wed,) studied this question.