Scheduling Multiple Industrial Electrolyzers in Renewable P2H Systems: A Coordinated Active-Reactive Power Management Method

• Published in: IEEE transactions on sustainable energy pp. 1 - 15
• Main Authors: Zeng, Yangjun, Qiu, Yiwei, Zhu, Jie, Chen, Shi, Zhou, Buxiang, Li, Jiarong, Yang, Bosen, Lin, Jin
• Format: Journal Article
• Language: English
• Published: IEEE 27.08.2024
• Subjects: Coordinated management; electrolyzers; Hydrogen; Job shop scheduling; Power system management; Productivity; Reactive power; reactive power optimization; Renewable energy sources; renewable power-to-hydrogen (ReP2H); thyristor rectifier; Voltage; voltage security
• ISSN: 1949-3029; 1949-3037
• DOI: 10.1109/TSTE.2024.3450503

Utility-scale renewable power-to-hydrogen (ReP2H) systems typically consist of multiple electrolyzers (ELZs), many of which are powered by thyristor rectifiers (TRs). A TR-powered ELZ has a nonlinear and nondecouplable relation between its active and reactive loads. The on-off switching and load allocation across multiple ELZs impact the efficiency of P2H energy conversion and the active and reactive power flows in the electrical network. Improper scheduling may result in an excessive reactive load from the hydrogen plant, causing voltage violations and increased network losses, which compromise both safety and profitability. To address these issues, this paper first explores the tradeoffs between the energy efficiency and reactive loads of ELZs. Then, we propose a joint active-reactive power management method to coordinate the loads and thermal properties of the ELZs, renewables, energy storage, and var compensation to improve the overall productivity and profitability. Mixed-integer second-order cone programming (MISOCP) is established to achieve these goals, and a decomposition algorithm enables its applicability in large-scale systems. Case studies show that the proposed method, at best, increases the hydrogen yield by 2.49% while reducing network losses by 3.12% compared to the state-of-the-art strategies based on wind and solar power data from Inner Mongolia, China. The optimal var resource configuration for ReP2H systems is also briefly discussed.