Assessment model of process regulation capability of multi-energy enterprises

Abstract

Accurate quantification of the process-level regulation capability of multi-energy industrial enterprises is crucial for participating in electricity demand response and promoting renewable energy consumption. Addressing the lack of systematic assessment methods for single processes and complete production lines, this paper proposes a quantitative assessment model coupling multi-timescale robust optimization with feasible operating domains (FOR). A multi-energy flow coupling model establishing electricity, heat, gas, and cooling balances is divided into normal, peak-valley response, photovoltaic consumption, and emergency supply modes. A vertex enumeration algorithm extracts the FOR under each mode, and a three-layer scheduling framework integrates day-ahead two-stage robust optimization, intraday distributionally robust optimization, and real-time model predictive control. A cascaded elastic coefficient method aggregates production line-level adjustment capabilities, identifying finishing rolling as the key constraint. Experimental results on a hot rolling steel line show that the peak-valley response mode FOR volume increases by 240.3% to 956.4 MW2 compared to the normal mode, cascaded aggregation identifies a 35.7% capacity loss caused by finishing rolling, and the daytime layer computation time of 8.7 seconds represents an 80.9% reduction compared to traditional robust optimization. This method provides engineering decision-making tools for assessing industrial regulation potential.