Energy management of hybrid electrochemical and hydrogen-based systems for decentralized and sustainable energy Systems

Hybrid Energy storage systems (HESSs) are essential to enabling high renewable energy penetration and achieving carbon neutrality. Their function is to absorb surplus energy during periods of excess generation and release it when demand exceeds supply, thus ensuring grid stability and reducing curtailment. As energy networks become increasingly complex, characterized by decentralized resources, prosumers, and multi-energy interactions, the need for advanced supervisory systems capable of optimizing energy flows in real-time has grown. The challenge lies in developing Energy Management Systems (EMSs) that can operate efficiently under stochastic, nonlinear, and dynamic conditions. This thesis aims to contribute to the modeling, analysis, and optimization of hybrid electrochemical–hydrogen energy systems. Specifically, it focuses on developing advanced EMS methodologies that ensure the efficient, reliable, and sustainable operation of future renewable-based energy networks. When addressing energy management, it is crucial to clearly define the application context in which the storage system operates. Each application is characterized by unique requirements, objectives, and operational constraints, which determine the optimal EMS strategies. This work considers three main categories of storage applications: civil applications, light-duty mobility, and heavy-duty mobility. For each category, HESS case studies are analyzed, and the effects of different EMS methodologies are evaluated in terms of energy efficiency, applicability, system degradation, and economic savings over both the short and long term. The three primary EMS categories explored are Rule-Based (RB-EMS), Optimization-Based (OB-EMS), and Learning-Based (LB-EMS) systems. The selection of the most suitable EMS requires careful consideration of several key parameters, including performance, adaptability, computational effort, and ease of implementation.