Bi-level optimal operations for grid operator and low-carbon building prosumers with peer-to-peer energy sharing

Low-carbon buildings (LCBs) are normally equipped with distributed energy resources (DERs) such as photovoltaic (PV) panels and batteries, thereby creating LCB prosumers that both produce and consume energy. Peer-to-peer (P2P) energy sharing among LCB prosumers could bring higher economic benefits for themselves, and facilitate better local power balance for the power grid. To fully harness the benefits of P2P energy sharing for both LCB prosumers and the power grid, a bi-level optimization method for LCB prosumers and the power grid operator is proposed in this paper. The grid operator at the upper level imposes the optimal time-varying network charge to LCB prosumers at the lower level to maximize its profit. And LCB prosumers with the objective of minimizing their costs adjust the schedules including P2P energy sharing and their heating loads to respond to the grid operator's optimal network charge prices. Additionally, to further exploit the energy resources including heating loads in LCBs, the model predictive control (MPC) approach is integrated with the bi-level optimization in the presence of uncertainties. The bi-level optimization belongs to the NP-hard category, making it challenging to solve. To reduce the computational complexity, the bi-level optimization problem is converted to the single-level programming using Karush-Kuhn-Tucker (KKT) conditions. Numerical results illustrate that the bi-level optimization method can obtain a balanced scheduling scheme between the grid operator and LCB prosumers. In addition, incorporating the thermal inertia of building heating loads can provide more operational flexibility for both the grid operator and LCB prosumers. Furthermore, the costs of LCB prosumers based on MPC approach can be reduced with the decrease of forecast error level.