U2Demo_D2.3_P2P trading matching and energy sharing models

This deliverable presents the outcomes of Task 2.3 of the U2Demo project. It presents analytical models (from conceptualization to mathematical formulations) that evaluate and quantify the implications of various local energy sharing and trading models. These models capture how allocation rules, pricing approaches, and clearing and matching algorithms affect outcomes at the individual and community levels within local energy communities. This deliverable makes three main contributions. First, it introduces a structured classification of local energy sharing and trading mechanisms. This classification distinguishes between collective and individual arrangements, as well as centralized and distributed configurations. This framework identifies six fundamental model types that capture the diversity of implementations observed in practice and proposed in the literature. Second, the deliverable develops mathematical formulations for the four model types requiring explicit market tooling to enable the local sharing and trading of energy: (i) allocation of capacity, energy, and revenues from collective assets, (ii) centralized individual energy sharing through internal pricing, (iii) auction-based centralized individual trading, and (iv) pool-based distributed peer-to-peer trading. Third, these four models are applied to a common case study based on the Dutch Scheveningen pilot. This case study represents a real-world energy community comprising participants with collective and individual assets and heterogeneous profiles. We apply the models to different test cases by altering the pilot’s system configuration. As such, we can examine the outcomes under various conditions. We evaluate the outcomes using a set of KPIs that cover the community’s procurement costs, as well as fairness and sustainability metrics. Results indicate that both centralized and decentralized collective allocation models provide substantial cost savings and improve self-sufficiency relative to a baseline without energy sharing or trading. Theoretical allocation methods, such as the optimal excess method and the Shapley value, ensure efficiency and individual rationality, though they are computationally demanding. Heuristic methods, particularly multi-round static sharing keys, are a practical alternative for larger communities with flexible assets. Internal pricing mechanisms, such as the mid-market rate and the supply-demand ratio, also deliver substantial cost reductions relative to the baseline without energy sharing. This particularly holds for communities dominated by individual participants that are facing fixed retail contracts. Indeed, internal pricing mechanisms require careful handling of collective assets and negative price conditions to avoid inefficiencies. Although auction-based mechanisms can provide additional flexibility in how energy is allocated within a community, their effectiveness depends heavily on the participants’ bidding behavior. In our case study, the incremental benefit over internal pricing or collective management was minimal. This suggests that the additional complexity and participant involvement may not always be justified. Nevertheless, this finding is highly system-specific, and the decision to implement auction-based mechanisms should be evaluated on a case-by-case basis. Pool-based distributed peer-to-peer trading reliably lowers total costs but distributes benefits unevenly due to individual transaction prices, creating potential equity concerns. Although these mechanisms are effective at reducing the total energy bill of the community, they may require additional fairness measures to ensure community members’ satisfaction. Overall, centralized and decentralized energy management schemes and internal pricing mechanisms closely approximate the community’s best achievable outcome, at least in the considered case study. Auction-based and distributed P2P approaches offer additional flexibility and customization but introduce complexity, participant engagement requirements, and potentially distributional trade-offs.