A bi-objective Markov decision process approach to explore retrofit design pathways for maritime decarbonization

Abstract

Retrofits to alternative fuels like methanol represent a strong candidate for complying with current and upcoming environmental regulations. The decision to retrofit to methanol power, propulsion and energy systems introduces uncertainties linked to technology integration, conversion costs, and environmental performance. This paper proposes a bi-objective Markov decision process as the foundation of a design tool aimed to support retrofit decisions within a vessel’s lifecycle. This approach aims to quantify the trade-off between the conflicting objectives of (a) emissions reduction and (b) retrofit costs minimization. The bi-objective formulation is evaluated against the equivalent single-objective formulations of the problem to assess the effect on the feasible design pathways. Depending on the initial vessel preparation level for a retrofit, this paper quantifies the uncertainty in the retrofit pathways during the lifecycle through the metrics of optimal state and optimal action. A case study has been applied to a notional short-sea vessel. Results indicate differing solutions when comparing the bi-objective to two individual single-objective scenarios. Preparation for a retrofit and the selection and usage of a methanol dual fuel configuration indicate a promising strategy to satisfy both objectives for the vessel’s lifecycle timespan.