Fixed-Demand Multifidelity Modeling and Capacity–Energy Design of Industrial Alkaline Water Electrolysis for Green Ammonia
Abstract Industrial alkaline water electrolysis for green ammonia requires capacity–energy decisions across model fidelities at the same hydrogen duty. A fixed-demand multifidelity framework combines 200 Shortcut simulations with 24 nested Rigorous training cases and 16 held-out Rigorous test cases at 11.12 t h–1 net H2. Fixed-demand fidelity matching removes throughput as a source of apparent performance differences, while response-specific fidelity selection retains low-fidelity information only when it improves prediction. Shortcut information does not preserve the Rigorous specific electricity consumption (SEC) ranking, whereas current density retains a cross-fidelity relationship. The final SEC representation uses Rigorous training data alone and gives a held-out root-mean-square error (RMSE) of 0.068 kWh kg–1 H2. Direct Rigorous calculations confirm minimum-area, knee, and minimum-energy alternatives spanning 52,800–93,600 m2 and 65.28–57.14 kWh kg–1 H2. Within this domain, fixed-demand electrification design requires joint consideration of electricity use and installed electrochemical capacity, with model fidelity selected by response.
Authors
- Gabriel Contreras‐Zarazúa (ORCID: https://orcid.org/0000-0001-6332-0106)
- Luis Germán Hernández-Pérez (ORCID: https://orcid.org/0000-0002-8603-9920)
- Iris Sandoval-Rojas
Institutions
- Universidad Autónoma Metropolitana (MX)
- ETH Zurich (CH)
- Universidad Nacional Autónoma de México (MX)
Publication Details
- Journal
- Industrial & Engineering Chemistry Research
- Published
- 2026-11-16
- DOI
- https://doi.org/10.1021/acs.iecr.6c02534
- Primary Topic
- Hybrid Renewable Energy Systems
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- Schweizerischer Nationalfonds zur Förderung der Wissenschaftlichen Forschung
- Consejo Nacional de Ciencia y Tecnología