Scholarlink

<p>Green hydrogen from solar-powered electrolysis is a key decarbonization pathway, but economic viability remains challenging, with current green hydrogen projects often reporting cost values above 3 $/kg H<sub>2</sub>. This paper introduces, for the first time in photovoltaic (PV)–electrolyzer–storage systems, an integrated framework that combines constrained Bayesian optimization with dynamic multi-state power management to minimize the Levelized Cost of Hydrogen (LCOH) by jointly optimizing component sizing and operation. The approach is supported by validated models, including a single-diode PV model (&lt; 1% current error) and proton exchange membrane (PEM) electrolyzer polarization curves (a root mean square error of 13.85–37.82&#xa0;mA·cm⁻<sup>2</sup>). Applied to four Moroccan sites, the framework yields LCOH values of 1.88–2.78 $/kg H<sub>2</sub>, positioning these systems near the lower bound of current global cost projections. Dakhla offers the lowest LCOH, benefiting from a high solar resource (2595 kWh/m<sup>2</sup>/year global horizontal irradiance) and an optimal configuration of 501 PV modules, 111 electrolyzer cells, and a 23 kWh battery. Sensitivity analysis identifies battery capacity and PV–electrolyzer coupling as key cost drivers, with optimized systems achieving coupling factors of 83.5–92.5%. Overall, a 48% variation in LCOH across sites highlights the importance of site-specific optimization for Morocco’s emerging hydrogen economy and demonstrates the transferability of the proposed framework to other high-solar-resource regions.</p> Graphical abstract <p></p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Constrained Bayesian optimization and dynamic power management for solar-driven green hydrogen production in Morocco

  • 期刊论文

  • 发表日期：

    2026年3月31日

查看全文
  • Saad Elmamoun,
  • Anas El Maakoul,
  • Zineb Bouhssine,
  • Alain Degiovanni

摘要

Green hydrogen from solar-powered electrolysis is a key decarbonization pathway, but economic viability remains challenging, with current green hydrogen projects often reporting cost values above 3 $/kg H2. This paper introduces, for the first time in photovoltaic (PV)–electrolyzer–storage systems, an integrated framework that combines constrained Bayesian optimization with dynamic multi-state power management to minimize the Levelized Cost of Hydrogen (LCOH) by jointly optimizing component sizing and operation. The approach is supported by validated models, including a single-diode PV model (< 1% current error) and proton exchange membrane (PEM) electrolyzer polarization curves (a root mean square error of 13.85–37.82 mA·cm⁻2). Applied to four Moroccan sites, the framework yields LCOH values of 1.88–2.78 $/kg H2, positioning these systems near the lower bound of current global cost projections. Dakhla offers the lowest LCOH, benefiting from a high solar resource (2595 kWh/m2/year global horizontal irradiance) and an optimal configuration of 501 PV modules, 111 electrolyzer cells, and a 23 kWh battery. Sensitivity analysis identifies battery capacity and PV–electrolyzer coupling as key cost drivers, with optimized systems achieving coupling factors of 83.5–92.5%. Overall, a 48% variation in LCOH across sites highlights the importance of site-specific optimization for Morocco’s emerging hydrogen economy and demonstrates the transferability of the proposed framework to other high-solar-resource regions.

Graphical abstract