• Energy Research
• 2025-2025close
• EU close
• CH close
• Applied Energy close

This paper investigates the use of deep multi-agent reinforcement learning (MARL) for the coordination of residential energy flexibility. Particularly, we focus on achieving cooperation between homes in a way that is fully privacy-preserving, scalable, and that allows for the management of distribution network voltage constraints. Previous work demonstrated that MARL-based distributed control can be achieved with no sharing of personal data required during execution. However, previous cooperative MARL-based approaches impose an ever greater training computational burden as the size of the system increases, limiting scalability. Moreover, they do not manage their impact on distribution network constraints. We therefore adopt a deep multi-agent actor-critic method that uses a centralised but factored critic to rehearse coordination ahead of execution, such that homes can successfully cooperate at scale, with only first-order growth in computational time as the system size increases. Training times are thus 34 times shorter than with a previous state-of-the-art reinforcement learning approach without the factored critic for 30 homes. Moreover, experiments show that the cooperation of agents allows for a decrease of 47.2% in the likelihood of under-voltages. The results indicate that there is significant potential value for management of energy user bills, battery depreciation, and distribution network voltage management, with minimal information and communication infrastructure requirements, no interference with daily activities, and no sharing of personal data. Applied Energy, 377, Part A ISSN:0306-2619 ISSN:1872-9118

Decentralized multi-energy systems (MESs) are a key element of a future low-carbon energy supply. Here, we address the crucial role of grid-connected and off-grid MESs in achieving a low-carbon future, particularly relevant for regions like the Mediterranean with high renewable energy potential and carbon-intensive grid networks, using a mixed integer linear program for optimal economic and environmental MES design considering location-specific regulations. The results reveal that substantial cost (up to 30%, potentially saving 1.6 million) and greenhouse gas (GHG) emission reductions can be reached in Mediterranean regions with sufficient solar and wind energy resources currently relying on fossil fuel-based generators. However, our case study shows that the actual cost and emission reductions are most likely limited due to location-specific regulations (limiting cost savings to 0.8 million), especially those that constrain solar photovoltaic and onshore wind. Off-grid energy systems might be suitable decarbonization options in Mediterranean regions, to avoid absorbing current GHG-intensive power from the local power grid, under marginal cost increase (15%) compared to grid-connected cost optimization. However, off-grid MESs require substantial upfront investments and exhibit some environmental trade-offs, especially on material utilization, which could be overcome by balanced autonomy. Overall, truly sustainable and secure decentralized energy systems can only be reached by considering life cycle environmental impacts, social acceptance, and regulations during the design phase. Applied Energy, 377 ISSN:0306-2619 ISSN:1872-9118