A Digital Twin of Battery Energy Storage Systems Providing Frequency Regulation

Battery energy storage systems (BESSs) are an important part of the modern electrical grid. They allow seamless integration of renewable energy sources (RES) into the grid by mitigating the variability of RES power production that depends on the availability of natural resources. However, the BESS operation can be disturbed in various ways, e.g. by equipment fault and cyberattacks.

To keep the work of a BESS that provides frequency control services predictable and reliable, a BESS digital twin is proposed in this paper. It supplies the battery owner with an up-to-date battery behavior forecast that can be further applied to intelligent condition monitoring, fault detection, battery management as well as cyberattack detection and mitigation.

A digital twin modeling includes three major steps: data gathering, preprocessing, and forecast. The BESS state of charge (SOC) data generated from utilizing real frequency data to a utility-scale BESS providing frequency regulation in the Nordic area is utilized to evaluate the quality of the BESS digital twin modeling.

The steps of data preprocessing are tailored for SOC forecast. We proposed a BESS digital twin that forecasts SOC by applying artificial intelligence (AI)-based methods. The demonstrative case study is presented to illustrate the framework implementation for a BESS providing frequency regulation.