Coalition analysis for low-carbon hydrogen supply chains using cooperative game theory

Abstract

Low-carbon hydrogen is a promising option for energy security and decarbonization. Cooperation is needed to ensure the widespread use of low-carbon energy. Cooperation among hydrogen supply chain (HSC) agents is essential to overcome the high costs, the lack of infrastructure that needs heavy financial support, and the environmental failure risk. But how can cooperation be operationalized, and its potential benefits be measured to evaluate the impact of different allocation schemes in low-carbon HSCs? This research works around this question and aims to analyze the potential of cooperation in a generalized low-carbon HSC with limited and critical resources using systems and cooperative game theory. This work is original in several aspects. It evaluates cooperation effects under different benefit allocation schemes while considering infrastructure agents’ dependencies (production, transportation, and storage) and specific traits. Additionally, it provides a transparent, replicable methodology adaptable to various case studies. It is highlighted that HSC coalitions form hierarchies with veto power, pursuing common goals like maximizing decarbonization and demand fulfillment. A cooperative game theory toolbox is developed to evaluate, display, and compare the results of six allocation solutions. The toolbox does not aim to determine the best allocation scheme but rather to support smart decision-making in the bargaining process, facilitating debate and agreement on a trade-off solution that ensures the viability and achievement of long-term coalition goals. It is built on three naïve and three game-theoretical allocation rules (Gately, Nucleolus, and Shapley value) applicable to peer group games with transferable utility. Results are presented for an 8-agent low-carbon HSC along with the total environmental benefit, the allocated individual shares, and numerical indicators (stability, satisfaction, propensity to disrupt), reflecting the acceptability of allocations. Numerical results show that the Nucleolus achieves the highest satisfaction among stable allocations, while the Gately allocation minimizes disruption propensity. Naïve rules yield different outcomes: “equal distribution for producers” carries the highest risk, whereas “equal shares for all agents” and “proportional to individual benefits” rules are stable but perform poorly on other criteria. © 2025 The Authors