Urban railway systems require regular maintenance to uphold safe and efficient operations. System operators are sometimes forced to perform this maintenance during revenue service hours. The resulting changes in revenue service are called a “revenue service adjustment,” or RSA. Because RSAs typically feature planning horizons of months or even years, operators have an opportunity to design them in ways that minimize the level-of-service (LOS) impacts for passengers. This paper presents a framework for operators to consider LOS impacts early in the RSA planning process. First, a taxonomy of service delivery strategies is developed, where various strategies can be characterized by typical LOS impacts and operational considerations to efficiently identify options to consider. Second, a method is presented to develop optimal service plans to deliver the various strategies. Third, the distribution of LOS impacts on various groups of passengers is found for each service plan. Finally, considerations related to work planning and productivity are quantified and balanced against LOS impacts. The framework is demonstrated on a real-world case study from the Washington Metropolitan Area Transit Authority (WMATA) in which a single track needed to be removed from service. The results show that under the same resource constraints, different RSA design decisions can result in a large range of potential wait time impacts, with ratios ranging from 1.15 to 1.34 compared with normal revenue service, as well as a promising daily time period to perform work that balances a large increase in productivity with a smaller increase in LOS impacts.
Moody et al. (Tue,) studied this question.