Net benefit-oriented condition-based maintenance for lithium-ion battery packs in SGLS systems: Combining degradation updating and decision-making

Abstract

Driven by global sustainability goals, the integration of renewable energy into power grids has significantly increased the demand for advanced battery management solutions. In source-grid-load-storage (SGLS) systems, effective operation and maintenance (O&M) of lithium-ion battery packs (LiBPs) are critical for balancing energy supply, ensuring operational reliability, and enhancing economic viability. However, existing maintenance strategies often fail to address the combined impacts of benefits, risks, and costs and instead rely on inflexible criteria, such as fixed failure thresholds, which are insufficient for managing batteries. Additionally, these strategies lack adaptability and do not incorporate real-time data, limiting their effectiveness in managing the stochastic dependence and inherent randomness of battery degradation. To address these limitations, this paper presents a dynamic condition-based maintenance (DCBM) strategy. This approach employs degradation modeling and parameters updating via a multivariate Wiener process, utilizing real-time data to refine decision-making. It introduces a novel net benefit-oriented model that integrates energy storage benefits, risk losses, and maintenance costs. By framing the problem as a Markov decision process (MDP), an improved algorithm is developed to optimize decisions throughout the battery’s lifecycle. Numerical analyses demonstrate that the proposed approach manages battery degradation uncertainties more effectively than traditional methods. This research provides an economically viable strategy for maintaining battery energy storage systems (BESSs), incorporating financial, safety, and maintenance considerations, thereby contributing to broader sustainability and efficiency goals.