Estimating Aggregated Behind-the-Meter Photovoltaic Generation Using Smart Meter Data

Abstract

Unobserved behind-the-meter (BTM) photovoltaic (PV) systems have increased rapidly in distribution systems over the past decade. Various PV configurations in the distribution system make it challenging to capture the characteristics of aggregated BTM PV generation. To improve the performance of estimating aggregated BTM PV generation using smart meter data, the paper proposes a method that groups consumers before estimation. This method identifies nonuniform PV configurations through the correlation analysis between net load and calculated plane-of-array (POA) irradiance. It divides consumers into several groups with similar PV configurations and disaggregates PV generation from net load for each group. The case study on a real-life dataset demonstrates that performing estimation on groups of consumers with similar PV configurations improves the estimation performance.