Nonlinear Model Predictive Control of Alkaline Electrolyzers to Provide Ancillary Services

Grid-connected power-to-hydrogen (P2H) plants are required to provide ancillary balancing services to the utility power systems. In this context, electrolyzers in the P2H plant must track regulatory load commands such as the automatic gen-eration control (AGC) or peak shaving signals. Large-scale P2H plants commonly use alkaline electrolyzers (AELs). However, the AELs are subject to complicated process constraints related to electrochemistry, thermal balance, and impurity accumulation. If they are not properly controlled in tracking the fast-varying regulatory commands, safety limits can be violated. To address this challenge, this paper proposes a nonlinear model predictive control (NMPC) approach for the AELs to track the regulatory command, considering comprehensive process constraints. In the meantime, the proposed controller optimizes energy conversion efficiency by coordinating the AELs' load levels, temperatures, and hydroaen-to-oxygen (HTO) impurities. Simulation on two 1,000
${\mathrm{N}}\mathrm{m}^{3}{/}\text{hr}$
AELs responding to an AGC signal validates the proposed approach.