Conference Papers 2023

Seminar
Day 1 (6 Sep 2023), Session 3, Overcoming Barriers to Forecasting, 16:30 - 18:00

Status
Accepted, awaiting documents

Submitted by / Abstract owner
Ann Marsden

Authors
Ann Marsden

Short abstract
This paper discusses the methodological challenges when using generalised journey time (GJT) and associated values of travel time savings (VTTS) across modes in forecasting rail demand and user benefits for new rail stations or lines.

Abstract
Where there are incremental changes in rail demand following, for example, a timetable change then that demand and associated user benefits will reflect the change in GJT by rail mode. Changes in GJT are calculated between the Do Minimum (DM) and Do Something (DS) and, when combined with rail values of time, used to calculate the benefits of timetable changes.

However, when new services are introduced to the rail network (stations or lines or both) there will be no DM GJT from which to pivot for changes associated with the new services. One possible solution to the issue of no base demand is to develop a gravity model reflecting demographic and economic characteristics of the origins and destinations of journeys and GJTs and fares between those OD pairs. DfT have developed such a model and reported on the model at ETC 2021 and ETC 2022. This paper explores three methodological challenges in developing and using such a model.

The first challenge is what GJT value to use for the DM? DfT Transport Appraisal Guidance (TAG) suggests using 2% of the DS forecast demand for the DM starting point demand. Another possibility, explored and implemented in the DfT gravity model suite, is to use car or bus journey time for the DM GJT between, for example, a new station and interchange station (or stations). One question is what journey time, service interval penalty and interchange penalties (if any) to use for these GJTs?

The second challenge concerns what GJT elasticity should be applied to the change in GJT between the DM and DS where the former reflects bus or road GJT and the latter rail GJT. Using GJT elasticities estimated in a gravity model using a rail base matrix reflect changes in rail journey times. Is it appropriate to then apply these elasticities to changes in GJT between modes?

The third challenge concerns what VTTS should be applied to the changes in GJT between the DS and DM for a switch from bus or car to rail? DfT TAG notes that for any group (bus passengers, car drivers etc.) there will be a distribution of values around the average value for the group and the distributions for each group are likely to overlap. For example, a car driver with an above average value of time for car drivers could switch to rail, where their value of time might be below average. However large changes between modes might alter the modal distributions sufficiently to significantly change the average values in the DM and DS cases. The question is how to manage the application of different VTTS where there are large changes between modes?

The paper will explore the theoretical context of these three methodological challenges and use evidence from the academic literature and policy practice both for the UK and other countries to determine how the issues have been resolved. The discussion can be used to better inform rail industry practitioners and policy makers in their decision making around rail schemes where such changes in GJT are anticipated to occur.

Programme committee
Transport Models

Topic
Human wellbeing – delivering transport schemes for the health of future generations