Machine Learning Based Capacity Estimation of Behind-the-Meter Battery Energy Storage Systems

As the widespread integration of renewable energy (RE) technologies like solar photovoltaics (PVs) continues to accelerate, there is a growing demand for energy storage solutions, notably battery energy storage systems (BESS), to address the intermittent and variable nature of RE generation. Behind-the-meter (BTM) battery energy storage systems (BTMBESS) has emerged as an effective means to manage energy consumption and store the excess RE generation at the local level. The accurate knowledge of available BTM-BESS is necessary to optimise the system operation and for grid infrastructure planning. This article proposes a machine learning (ML) based approach to estimate the installed capacity of BTM-BESS at customer premises. The proposed approach leverages customer historical data, such as, energy consumption, peak load, and installed PV capacity to estimate the installed capacity of BTM-BESS. The proposed model is rigorously validated using real-world data from residential customers, and results indicate that the ML-based approach can provide high accuracy estimation of installed capacity of BTM-BESS.