Optimisation of charging infrastructure for battery electric multiple units: A coupled Simulation–Optimisation framework applied to three UK railway routes

Abstract

Battery electric trains are increasingly proposed as a cost-effective alternative to full railway electrification on routes where capital constraints prohibit continuous overhead line installation. This paper presents a coupled simulation–optimisation framework that combines a physics-based MATLAB train dynamics model, incorporating Davis resistance, aerodynamic drag, gradient forces, auxiliary heating/cooling loads, and regenerative braking, with a GAMS mixed-integer programme that minimises 20-year whole-life infrastructure cost subject to SoC feasibility constraints. The framework is applied to three UK routes of differing topology, and results show that partial OHL electrification is always cost-optimal. Sensitivity analyses identify grid connection cost, ambient temperature, and battery degradation as the most consequential design parameters. Annual carbon savings relative to diesel multiple units are significant but depend on regional electricity grid carbon intensity. The findings further suggest that BEMU deployment is unlikely to increase passenger fares and may instead enable fares to be held constant or reduced.