Hydrogen Cost Projections Overestimate Learning Speeds

Hydrogen cost models risk over-optimism by misapplying cumulative capacity growth to manufacturing experience

Hydrogen electrolyzer cost projections using cumulative installed capacity doubling as the learning curve metric are overstating potential cost reductions. A 2025 European study of electrolyzer projects (2005–2025) shows raw cost declines of 23.3% across all projects, but normalized for project size, learning rates drop significantly: 13.3% for PEM electrolyzers and 17.6% for alkaline. Crucially, cumulative capacity doubling does not equal manufacturing experience doubling due to stack-size growth. A 5 GW to 50 GW capacity increase (3.32 doublings) requires only one stack count doubling if average stack size grows from 1 MW to 5 MW. This means current cost models underestimate how much manufacturing scale must actually occur to achieve meaningful cost reductions. The IEA 2025 breakdown shows electrolyzer stacks represent 15–20% of total capital cost. A 20% stack cost reduction would only lower overall project costs by 4%, highlighting that cost projections must separately model stack manufacturing, electrochemical performance, and scale effects. This matters for hydrogen’s role in decarbonizing hard-to-electrify sectors: over-optimistic cost projections could delay deployment when actual manufacturing experience growth is slower than current models assume. Next, studies should track stack-level manufacturing experience separately from project scale to avoid further overestimation. The 2025 European study focuses only on electrolyzer projects and may not reflect all hydrogen production pathways, and stack-size assumptions could vary in practice.

Source: CleanTechnica