Dynamic operation and performance evaluation of a novel PV-driven electricity‑hydrogen cogeneration system integrated with SOFC, PEMEC and plasma-assisted ammonia decomposition

This study investigates a photovoltaic (PV)-driven electricity–hydrogen cogeneration system integrating a solid oxide fuel cell (SOFC), a proton exchange membrane electrolyzer (PEMEC), and a plasma-assisted ammonia decomposition (PAD) unit. The system enables cross-sector low-carbon co-supply but must balance fluctuating PV generation and time-varying electricity demand with near-constant hydrogen demand. A bi-level control framework is developed: an upper-layer mixed-integer linear program (MILP) schedules optimal energy-allocation trajectories, while a lower-layer dynamic controller tracks them for the SOFC, PEMEC, and battery. For the primary electricity supplier, an enhanced SOFC dynamic model capturing intra-stack distributed states is embedded to enable thermal-safety-aware, constraint-enforced operation beyond lumped models, and a PAD performance model is derived from experiments. Results reveal an economy–tracking trade-off: an SOFC load change rate limit of 20%/15 min yields an LCOE of 59.9 $/MWh with an average relative power-tracking error of 0.4%, whereas tightening the limit to 5%/15 min reduces tracking error but increases LCOE to 63.2 $/MWh due to larger battery capacity. The PEMEC absorbs surplus PV electricity, with hydrogen production reaching 48.3% of total PV generation. Integrating PAD reduces annual PV curtailment from 39.4% to 32.2%, converting an additional 68 MWh of PV electricity to hydrogen annually. Novelty is demonstrated by quantifying the economy–tracking knee point (20%/15 min) under realistic PV fluctuations and by evaluating PAD's system-level role in curtailment utilization within the same plant-level framework. This system architecture and control framework support low-carbon multi-energy parks with improved renewable utilization and flexible electricity–hydrogen co-supply.