Analysis and modeling of peer-to-peer electricity trading systems in the context of energy communities.

In energy fields, electricity gird is always a significant area to analysis and improve. Recently, with the increasing public concern about clean and renewable energy, and the fast developing technology support, users are more flexible to generate and utilize renewable energy sources(RES) as solar energy, and more active to participate in the energy market. Therefore, instead of solely electricity consumers or suppliers, prosumers, who have the capacity to generate and demand in energy consumption, become common to find in the market. Prosumer community is composed of a number of prosumers who seek for a collaborative benefit, and it is an essential representation of the prosumers activities that gains many attentions. Moreover, peer-to-peer(P2P) electricity market also emerges to fulfil the demand of prosumers to trade energy in a more direct and efficient way. In such a market, prosumers or other direct participants can arrange their demand and surplus source and destination in a bottom-up way. Thus to explore a possibility future that energy filed can exploit, a combination of these new trends is emphasized in this thesis. To model a community-based P2P trading market properly, a regulatory framework that contains both sides features is established. Under the framework, based on the previous study in this fields, the principles that guide the activities in the market are summarized and compared. In practical simulation, this thesis strives to explore the possible influences brought by community typologies and sharing mechanisms. To tackle the overall P2P trading market, the simulation is run with mathematical day-ahead scheduling model and post-clearing model. Therefore the model is simplified with a Mixed Integer Linear Programming(MILP) problem and a negotiation process. The decision-making of community members is framed in an agent-based model(ABM) for a distributed management. The results are obtained from different community markets. It shows the effectiveness of model simulation in terms of scheduling and post-clearing mechanisms. A preferable condition of community composition and pricing method is pointed out for a collective benefit. The performance of typologies and sharing mechanisms is evaluated, and it is compared in community trading behavior and economic profit from the perspective of community, prosumers and consumers.