Estimating Passenger Demand for Fast Rail
Abstract
When timetabling new train services, planners are often interested in how passenger demand will be distributed between available trains. An express train, for example, may attract a higher passenger load than a stopping service. However, if express trains run infrequently, passengers may prefer to take a stopping service to avoid an unreasonably early or late
departure time.
This paper describes a technique for estimating the number of passengers travelling on individual train services, given a timetable and a pattern of passenger movements. The technique, often referred to as the ‘Rooftop Model’, has been used by British Rail since the 1970s to assist in the planning of high-speed rail services (Tyler and Hassard 1973, Shilton 1982). The model has not yet had widespread application in Australia, and one of the objectives of this paper is to inform Australian rail planners about the model’s principles.
The Rooftop Model was recently used by consultants Sinclair Knight Merz to estimate the demand for proposed new services delivered as part of the Victorian Regional Fast Rail Project. The paper explains how the model was adapted to the Victorian regional context and developed into a fully-featured tool for testing Fast Rail timetables.
The paper first looks at a hypothetical example of a passenger’s train-selection process, then provides a mathematical description of the Rooftop Model. Later sections of the paper describe practical aspects of the model’s application to Regional Fast Rail planning.