Optimal Power Allocation Strategy for Hydrogen Production System of ALK-PEM Electrolyzer Considering Second-level Response Characteristics

Most electrolyzer scheduling models operate at minute-to-hour resolutions and thus overlook the second-level dynamics required to track fast renewable fluctuations, leading to avoidable wind curtailment and revenue loss for off-grid wind-to-hydrogen systems. This paper proposes an optimal power allocation framework for a hybrid electrolyzer system composed of alkaline (ALK) and proton exchange membrane (PEM). ALK provides low-cost, large-capacity baseline production but responds slowly, whereas PEM delivers second-level agility to absorb high-frequency power variations. A two-stage strategy is designed to couple efficiency-oriented power allocation with refined power dynamic regulation under explicit start/stop, ramp-rate, and capacity constraints. Moreover, a dual-layer array-rotation mechanism is incorporated to balance operating stress across stacks and prevent unit-specific overuse. Case studies using measured data from a practical project in Northeast China show that, compared with traditional power allocation methods, the proposed method increases hydrogen production by 3.3%-5.6%, reduces wind curtailment by 70.6%-94.7%, and improves hydrogen revenue by 21.9%-32.0%, enhancing renewable utilization and profitability.