A Novel Multi-Layer Framework for Dynamic Operation of Prosumers in Peer-to-Peer (P2P) Energy Markets

Abstract

Research in transactive energy systems, in recent years, has been primarily focused on the financial aspects of peer-to-peer (P2P) energy trade with little attention paid to the operational and practical aspects of how this energy trade should occur in a system. Practical prosumer behavior in such systems should be subject to their own internal status and the external conditions of the electrical network. Moreover, for such practical realization of a prosumer to be feasible, they should be primed to operate through system disturbances and parameter uncertainties. In this thesis, a novel mathematical model is presented to enable prosumers to partake in P2P energy trades with full operational freedom over their own consumption, energy storage system (ESS) operation and their distributed power generation capability. The proposed model integrates a physical system (physical layer) with prosumer operations (virtual layer) to evolve a multi-layer framework which allows physical network constraints to be implemented with relative ease. The formulation is implemented on a 33-Bus test system considering various system objectives and the results demonstrably prove the significance and applicability of the proposed framework in P2P energy markets. The multi-layer framework is then further extended to enable the prosumer to respond to uncertainties in the local grid or its distributed energy resources, through an MPC based approach. The considered uncertainties are further split into categories termed as ’known uncertainty’ and ’unknown uncertainty’; with the former referring to forecasting errors and the latter referring to unexpected system disturbances. Several cases are developed considering combinations of system parameters to be uncertain and by introducing disturbances to observe prosumer responses. The simulation results prominently show the prosumers responding to unexpected disturbances by adjusting their behavior and P2P energy trade while maintaining their optimal objectives. Such results demonstrate the viability of this MPC approach for the realization of a practical prosumer.