Experiment-informed simulation of hydrogen fuel cell auxiliary power units for maritime decarbonisation

The continuous rating of a multi-cell proton-exchange-membrane (PEM) fuel cell stack is normally inferred from mean polarisation behaviour, yet sustained operation is limited by the weakest cell. Here, cell voltages measured on a 20-cell, 250 cm 2 short stack over 27 operating points (0-1.8 A cm −2 ) are used to infer cell-to-cell variation in ohmic resistance and limiting current density, from which a virtual 300-cell, 60 kW-class stack is sampled. Voltage dispersion grows from below 1 mV near open circuit to ∼36 mV at the 62.9 kW peak, and the minimum cell voltage falls below 0.45 V above ∼1.5 A cm −2 , so cell spread, not mean performance, sets the continuous limit. The limit falls at the same set-point in 91% of 10,000 sampled stacks, but a 20-cell stack from the same population overstates it by about 6%. Published degradation behaviour reduces it to 1.10-1.30 A cm −2 at end of life. For a modular marine auxiliary power unit, levelised cost is £0.42-0.82 kWh −1 at £4-£10 kg −1 H 2 and operational carbon intensity is ∼85% below diesel. The method gives stack manufacturers and marine integrators a practical route to rating a full-size stack on its weakest cell before it is built.