Scenario-based multi-criteria evaluation of alternative fuels for international marine transportation

The maritime sector is under mounting pressure to decarbonize, with international shipping responsible for nearly 3% of global greenhouse gas (GHG) emissions. However, the industry faces substantial uncertainty regarding viable next-generation marine fuel pathways, and existing studies largely lack scenario-based assessments that reflect divergent policy and market priorities. In this context, this study applies a multi-criteria decision analysis (MCDA), operationalized through the TOPSIS method, across 19 decision attributes grouped into four categories (environmental, economic, technical readiness, and operability) under three policy-relevant scenarios: balanced, climate-led, and cost-led. The framework evaluates energy transition pathways for international shipping aligned with mid-century climate goals, focusing on leading alternative fuels (ammonia, methanol, hydrogen, and biodiesel). Fuel performance and attributes importance assumptions for the 2050 horizon are synthesized from a broad range of academic and industry sources and assessed within the defined scenarios. Ranking outcomes are further examined using Monte Carlo (MC) simulations that explicitly distinguish uncertainty in future fuel performance from uncertainty in policy and stakeholder priorities. The results show that fuel rankings are highly sensitive to both scenario framing and uncertainty structure, with no single fuel consistently dominating across perspectives. Hydrogen performs best under climate-led priorities, biodiesel dominates under cost-led conditions, and ammonia and methanol emerge as competitive options under balanced scenarios. The TOPSIS results reveal clear rank reversals across scenarios. In the balanced scenario, ammonia ranks first, followed by methanol and hydrogen, with biodiesel ranking last. Under the climate-led scenario, hydrogen emerges as the top-ranked fuel, followed by ammonia and methanol, while biodiesel consistently ranks lowest. In contrast, the cost-led scenario favors biodiesel as the leading option, followed by ammonia and methanol, with hydrogen ranking last. MC sensitivity analyses indicate that these rankings are robust but not deterministic. Rankings are relatively stable under weight-only uncertainty, more sensitive to uncertainty in fuel performance scores, and most variable when uncertainty in both scores and weights is combined. Category-level uncertainty analysis further shows that economic and operability attributes exert the greatest influence on ranking outcomes, while technical readiness and environmental attributes play a more limited role in differentiating fuels at the 2050 horizon.