Estimating Behind-the-Meter Storage Capacity: A Data-Driven Approach Facing Time-of-Use Pricing

As the deployment of distributed generators accelerates, the newly installed capacity of behind-the-meter (BTM) storage installations has experienced a significant upsurge. The energy storage economy gains momentum further catalyzing this trend. Nevertheless, the majority of BTM storage systems function autonomously from public utilities, primarily serving as private assets for energy arbitrage, thereby presenting regulatory challenges for the grid operators. The literature has endeavored to tackle these challenges, with a predominant focus on operational strategies. However, the inference of storage capacities to obtain direct insights remains relatively unexplored. This study aims to bridge this research gap by devising a data-driven methodology for estimating storage capacity, with a particular focus on the Time-of-Use pricing schemes commonly utilized by small-scale BTM storage systems. Additionally, the grid operators may face challenges in obtaining sufficient data for estimation. To this end, we examine the theoretical minimum sample size necessary for accurate capacity estimation. Our comprehensive evaluation validates the efficacy of the proposed approach and elucidates the associated complexities of sample size.