Urban transit decarbonization via stochastic programming and statistical inference under uncertainty

Urban transit decarbonization is integral to achieving a net-zero public transportation systems. This work proposes an optimization model for bus fleet transition planning, involving purchases and allocation to routes, fueling and charging infrastructure, and financing. The model adopts stochastic programming to address decision-making under uncertainty and is formulated as a mixed-integer linear program. A confidence interval estimation method is derived to accommodate diverse decision values and non-uniform scenario probabilities, alongside an efficient scenario construction approach. A case study of the Metro Vancouver regional bus network is conducted to explore transition pathways for adopting battery electric and hydrogen fuel cell buses. Results indicate that shifting to a battery electric fleet is more cost-effective overall, while the hydrogen pathway demands smaller infrastructure investments. The competitiveness of hydrogen could significantly improve if the substantial potential for cost reductions is realized. A mixed fleet can integrate the advantages of both pathways.