A peer-to-peer demand side deviation management method for photovoltaic integrated industrial parks

The clearing and settlement mechanisms of demand response (DR) significantly influence the economic incentives of industrial loads, especially under high-penetration renewable energy scenarios where insufficient generation-side flexibility leads to stricter deviation penalties. To address the resulting economic losses and reduced participation, this paper proposes a two-stage P2P deviation management framework for flexible loads and energy storage systems (ESSs) in an industrial park. First, considering the inherent response delay, speed, and power of flexible loads, the DR trajectory is divided into partial and full response phases, enabling accurate modeling of demand-side uncertainty and deviation. Second, based on this response-phase modeling, a day-ahead deviation optimization strategy is formulated using mixed-integer second-order cone programming. The strategy jointly determines response phase durations, deviation directions, and ESS charging/discharging schedules to reduce expected deviations and enhance complementary effects among loads, while reserving sufficient flexibility for real-time correction. Third, following the day-ahead DR schedules, a real-time P2P deviation compensation model between DR nodes and ESS nodes is developed based on the alternating direction method of multipliers (ADMM). This model coordinates complementary deviation matching among loads and ESS compensation for residual deviations, while balancing distribution network losses and ESS profitability. Case studies on an industrial park with 58% photovoltaic penetration demonstrate that the proposed method effectively reduces DR deviation penalties and improves the benefits of loads and ESSs.