Revealing design archetypes and flexibility in e-molecule import pathways using Modeling to Generate Alternatives and interpretable machine learning

TL;DR

This work addresses the fragility of cost-optimal designs for hydrogen import pathways by introducing Modeling to Generate Alternatives (MGA) to enumerate near-optimal, diverse configurations across eight carrier/transport pathways in a Morocco–Belgium corridor. It combines a TRACE(PyPSA) energy-system optimization with MGA to map a large near-optimal design space and employs interpretable ML (k-prototypes clustering and CART) to extract actionable design rules and archetypes. The key findings reveal a broad near-optimal space with substantial flexibility in renewable generation and storage—most technologies are not strictly required, except for the electrolyzer in hydrogen pathways—and identify three design archetypes across carriers: wind-dominated, PV-dominated, and mixed configurations. This approach provides decision-makers with robust, context-aware options suited to regulatory, spatial, and stakeholder constraints, and offers a framework that can be extended to other energy-transition planning problems.

Abstract

Given the central role of green e-molecule imports in the European energy transition, many studies optimize import pathways and identify a single cost-optimal solution. However, cost optimality is fragile, as real-world implementation depends on regulatory, spatial, and stakeholder constraints that are difficult to represent in optimization models and can render cost-optimal designs infeasible. To address this limitation, we generate a diverse set of near-cost-optimal alternatives within an acceptable cost margin using Modeling to Generate Alternatives, accounting for unmodeled uncertainties. Interpretable machine learning is then applied to extract insights from the resulting solution space. The approach is applied to hydrogen import pathways considering hydrogen, ammonia, methane, and methanol as carriers. Results reveal a broad near-optimal space with great flexibility: solar, wind, and storage are not strictly required to remain within 10% of the cost optimum. Wind constraints favor solar-storage methanol pathways, while limited storage favors wind-based ammonia or methane pathways.

Figures (9)

• Figure 1: Overview of the hydrogen import supply chain. Renewable electricity from solar PV and wind turbines is used to produce hydrogen via PEM electrolysis, with seawater desalination providing the required feedwater. Battery storage balances the short-term variability of renewable generation, while hydrogen storage buffers supply before conversion. The produced hydrogen can then enter one of four carrier-specific synthesis pathways: (i) compression or liquefaction for direct hydrogen transport, (ii) catalytic methanation with captured $\text{CO}_2$ to produce synthetic methane, (iii) combination with nitrogen from an air separation unit in a Haber-Bosch process to produce ammonia, or (iv) catalytic synthesis with $\text{CO}_2$ to produce methanol. The resulting carriers are transported either by ship or pipeline to the import terminal. At the import side, carriers go through reconversion: regasification for liquid hydrogen, reforming for methane and methanol, and cracking for ammonia, so that all pathways ultimately deliver gaseous hydrogen as a uniform end product.
• Figure 2: Renewable potential of Morocco, showing high wind and solar resources with average capacity factors of 44% and 29%, respectively. The monthly average of solar energy remains stable year-round, whereas wind energy is more variable, with the highest average potential during spring.
• Figure 3: Hydrogen pathways (shipping and pipelines) are associated with the lowest LCOH, while non-hydrogen carrier pathways exhibit nearly twice the cost. Among non-hydrogen options, ammonia has the lowest LCOH, while methane has the highest. Excluding the reconversion step for non-hydrogen carriers significantly reduces the LCOE difference, bringing them closer to hydrogen and making methane cheaper than methanol. This contrasts with the case including the conversion step, where methanol performs better as a hydrogen carrier, whereas methane appears more suitable as a direct energy source.
• Figure 4: Installed capacities for the optimized pathways, showing that the hydrogen pathway requires the lowest overall capacities, while methanol has the largest energy storage and electricity generation capacities. Methane exhibits the highest electrolyzer capacity. The high storage and generation requirements for methanol come from the minimum power/hydrogen intake flow constraint of the methanolization unit.
• Figure 5: Cost breakdown of the optimized pathways, showing that wind dominates the costs for hydrogen, ammonia, and methane, while PV is the main cost driver for methanol. Non-hydrogen carriers face higher overall costs due to additional conversion units (e.g., ammonia synthesis, direct air capture), which add to the already higher costs of shared technologies such as wind and electrolysis.