Behind-the-Meter Solar Disaggregation: The Value of Information

In recent decades, the increasing adoption of rooftop solar photovoltaic (PV) systems has posed new challenges for distribution system operators (DSOs). DSO typically only have access to each customer's net generation/consumption. However, since these PV systems are commonly installed behind customers' meters, the DSOs are unable to monitor their generation, leading to a lack of observability that is critical for maintaining grid security. One potential solution to this issue is behind-the-meter solar disaggregation techniques, which can estimate behind-the-meter activity without the need for additional meter installations. These techniques utilise various sources of information, such as weather data and customer reactive power consumption, which can be obtained through third-party providers or direct contracts with customers. However, operators must determine which pieces of information are the most beneficial before investing in them. This paper studies seven different behind-the-meter solar disaggregation techniques that use varying sets of information and compare them based on their scalability, complexity, and required data. We also conduct numerical experiments using two publicly available datasets to evaluate the accuracy of these techniques.