Equitable Bus Route Electrification Based on a Mixed Integer Linear Programming Approach

Abstract

While public transportation has seen improvement over time with advancements in vehicle technology and urban planning, low-income populations do not see the full benefit of these advancements. The common approach to transportation planning is to distribute benefits in the most cost-efficient manner, meaning neighborhoods with the best existing infrastructure are likely to receive more timely benefits than low-income areas that require more costly updates. This disparity can be thought of as a lack of equity in transportation planning, where equity means that the population that needs a public service the most should benefit the most from improvement of that service.

This work focuses on improving equity within the proposed electrification of the Massachusetts Bay Transportation Authority (MBTA) bus network in Boston. We are interested in which routes should receive updates first to maximize equity, while understanding that focusing on equity poses an inherent cost trade-off. To solve this problem, an optimal subset of routes must be selected for electrification using an objective function that prioritizes routes with the lowest income riders and the highest levels of pollution from diesel buses.

Assuming an optimal cost structure for the full transition to battery-electric buses, and also assuming that not all depots and routes will be electrified on the same time scale, we use Mixed Integer Linear Programming (MILP) methods and a quantification of transportation equity in various objective functions to decide which bus routes originating from the Cabot depot should be prioritized for electrification benefits from an equity standpoint. We then analyze the sensitivity of our results to changes in the cost constraint and conclude the degree to which equity factors correspond to higher energy transition costs. The results show that high-pollution routes are less attractive from a cost standpoint than low-income ridership routes. It is also shown that a given percentage of total electrification costs can electrify a subset of routes with even larger percentages of total pollution and low-income ridership, meaning that the benefits of including equity factors are high for given cost levels in our problem scope.