Orateur
Description
In the context of the global transition towards net-zero emissions, local energy markets (LEMs) offer a practical and effective approach for integrating the increasing penetration of distributed energy resources, such as intermittent renewable generation, energy storage systems, and flexible loads. By facilitating active participation from small-scale consumers, producers, and prosumers in energy trading, LEMs enhance local energy balance and system efficiency. Moreover, clean energy carriers like hydrogen are being progressively incorporated into LEMs, driven by various government incentives and policy support. However, most existing research focuses predominantly on electricity trading, with limited attention to hydrogen and its associated market mechanisms. The development of integrated electricity-hydrogen markets remains insufficiently explored, particularly under uncertainty.
Therefore, we propose a novel multi-period coupled electricity-hydrogen local market framework that considers interactions on both the supply and demand sides—unlike the majority of existing studies which typically consider only single-sided energy management or isolate electricity and hydrogen markets. To capture renewable energy uncertainty, we propose a data-driven chance-constrained approach that leverages the advanced machine learning technique - normalizing flow, enabling representation of temporal correlations in renewable generation without relying on predefined distributional assumptions. Furthermore, we develop a decentralized market clearing mechanism based on a modified alternating direction method of multipliers with additional proximal terms, which guarantees convergence and optimality in theory even when the objective function lacks strongly convexity. Numerical simulations validate the effectiveness of the proposed approach, demonstrating a 17% reduction in total system costs compared to scenarios without local electricity trading.
Acknowledgements: Financial support from the EPSRC (under projects EP/T022930/1 and EP/W003317/1) is gratefully acknowledged.
