A Data-efficient Capacity Estimation Method for Behind-the-Meter Solar PV Systems

Behind-the-Meter (BtM) solar PV systems are becoming more popular, yet accurate capacity information for these systems remains unavailable to key stakeholders, such as Distribution System Operators (DSOs). This information is crucial for DSOs to manage grid stability, optimize energy distribution and plan infrastructure upgrades effectively, especially as BtM adoption continues to grow. This paper presents a data-efficient capacity estimation method for BtM solar PV systems, based on the relationship between electricity consumption at daytime and at nighttime. Using cross-validation and comprehensive Sensitivity Analysis (SA) on a real-world dataset from the Netherlands, the study shows that the proposed method requires only three months of data to achieve an R<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> above 0.9 with appropriately chosen hyperparameters, specifically irradiance thresholds for daytime and nighttime. This data efficiency highlights the method’s strong potential for deployment in real-world production settings.