Project C3 - Future role of rail in integrated transport policy

INVESTIGATORS: Professor Chris Nash, Dr Mark Wardman (University of Leeds); Professor Mike McDonald, Professor John Preston (University of Southampton)
RESEARCHER(S): Dr M. Beecroft, Dr Y Li (University of Southampton); Dr.G Whelan, Dr A.S.Fowkes, Dr A E Whiteing, Dr W F Lythgoe; Mr. J Shires; Ms B Menaz; Mr D Johnson (University of Leeds)

Industrial Collaborators: Passenger Demand Forecasting Council; Seminars hosted by Strategic Rail Authority, Department for Transport

Original objectives: This project aimed to identify alternative future roles for the rail system and sought to model them to understand better their advantages and disadvantages. More specifically, the objectives were:

• Identify alternative aspirations for the future of rail in an integrated transport policy
• Understand capabilities of existing models to forecast such changes
• Identify appropriate model developments for this purpose
• Enhance existing models to enable appraisal of alternative aspirations
• Consider the implications of the appraisal results for future rail research needs

Summary of outcomes:

• Identify alternative aspirations for the future of rail in an integrated transport policy. This task identified three ‘external’ scenarios against which a range of policy options or strategies can be assessed. These are: (i) continue as now, (ii) high GDP growth and (iii) low GDP growth. Against these backgrounds we propose to examine a series of alternate road and rail strategies including road pricing, significant road investment, rail rationalisation and significant rail investment (Report C3/2).
  
  
• Understand capabilities of existing models to forecast such changes. For the passenger sector, it has not been possible to identify a single model which is suitable for all policy tests and therefore our preferred approach is to make use of existing research findings as well as proposing the development of a new modelling framework which brings together the strengths of existing models and allows flexibility in parameterisation, functional specification and application. A similar review of existing models was undertaken for the freight sector with the recommendation to develop a new version of the LEFT model as a strategic model tailored to our specific needs. (Report C3/3).
  
  
• Enhance existing models to enable appraisal of alternative aspirations. (i) Passenger Markets. For passenger forecasting two new models were developed; a spreadsheet model based on ticket sale data for various categories of route and existing knowledge of relevant elasticities, and a more sophisticated model looking in particular at station catchments areas and able to model specific changes on the flows between 400 major stations within Britain. The first strand of the new passenger modelling work looked at how forecast background changes to population, employment, car ownership and household income might influence the demand for rail travel. Projections of each of these explanatory variables to the year 2031 were drawn from the Department for Transport’s planning data and mapped to station catchment areas. Across the country as a whole there are modest forecast changes to rail demand as a result of changes to population, employment and car ownership, although there is significant variation between variables and flows.
  
  Population change is likely to increase the demand for rail travel by 6% across the country as a whole by 2031, with stronger growth for flows based in London and the South East (around 10% growth). Forecast changes to employment levels in the South East are likely to have a strong influence on rail demand increasing trips to London by over 9%, however on other flows the impact of employment is limited.
  
  With the exception of Central London, car ownership levels are forecast to increase with a reduction in the proportion of households without access to cars and a reduction in the demand for travel by rail. This impact is strongest for Non-London short distance flows and weakest (marginally positive) for London based flows. Overall the impact of changing levels of car ownership on rail demand is estimated to be a decrease of over 8%.
  
  Overall, the positive impacts of population and employment change are offset by the negative impacts of increased car ownership. At a flow category level however, non-London flows show a 10-15% reduction in rail demand due to socio-demographic change and London based flows a 10-15% increase in demand by 2031. The changes brought about by socio-demographic change are therefore estimated to be relatively small, especially when compared with the levels of growth to rail demand driven by relatively modest economic growth. Changes to GDP (and GVA) equal to two percent per year are forecast to generate a more than doubling of demand by 2031 without binding capacity constraints.
  
  Policy tests involved the specification of three strategies each for fare, service quality (generalised journey time), road pricing and socio-demographic change, yielding a total of 81 states to be tested. Outside of the rail sector, the modelling exercise considered the demand implications of comprehensive road user charging including a revenue neutral scheme and a revenue raising scheme.
  
  The results for the 81 policy tests show a considerable variation across flow categories and policy tests. In aggregate unconstrained demand forecasts vary between a small reduction and a four fold increase in demand by 2031 depending upon the forecasting assumptions. However in all but the most extreme unfavourable assumptions there is considerable growth. Forecast growth is strongest for flows to London and weakest for non-London short distance travel. For a central scenario regarding other variables, an annual fares increase in excess of 3% in real terms would be necessary to choke off the growth in demand.
  
  It is interesting to note that the strength of the economy is the key driver of rail demand followed by service quality and fares. The impact of national road user charging on rail demand is modest due to the offsetting effects of increased road operating cost and improved road journey times.
  
  A final set of new modelling work involved the identification and improvement to service quality on existing low quality flows. A total of 231 flows from the top 4,000 flows by volume were identified as having very low levels of service relative to other similar flows and demand estimated following an assumed improvement in quality to bring these services in line with other flows on the network. The above forecasts suggest very major long term growth, particularly on heavily used routes to London.
  
  In addition to the new modelling work, the project also reviewed existing work on the demand and capacity implications of the two major investments currently being considered as solutions to this problem; a new high speed line between London and the north and a cross London rail route, on both of which issues ITS has had some involvement in research. On the evidence of these studies, there appears to be a strong case for the construction of a new north-south high speed rail link and of either the Crossrail or Superlink proposals, although further research on the best ways of using these proposals to deal with capacity problems is certainly neededThe passenger models and complete results are described in report RRUK C3/4.
  
  
• (ii) Freight Markets. Required outputs of freight models were tonnes and tonne-km, for a range of policies and strategies tested against one or more background scenarios, with detail on train, wagon and HGV kilometres run also being important if any wider economic appraisals were to be conducted. Work was based on ITS Leeds’ existing strategic freight transport model LEFT2, the new model being called LEFT3. This has no geography or network assignment stage, but road and rail freight traffic is split by 9 distance bands. Further disaggregation is by 7 broad commodity groups and by whether we deem the traffic ‘Train-friendly’(TF) or ‘Train-unfriendly’(TU). This latter distinction depends on the number (if any) of collection or delivery legs involved in the rail movement alternative. Full details of the results and a comprehensive description of the LEFT3 model is contained in Report C3/5.
  
  Base data was for 1998, and ‘do-nothing’ bases were constructed for 2010 and 2016. In 2016 we formed two bases (low & high) depending on economic growth scenarios. Three ‘strategies’ were considered: (O) Do Nothing. This is merely the base forecast for tonnes and tonne kms allocated to the appropriate vehicle types for that year. (A) Pro-Rail. This strategy includes a range of policies broadly favouring rail and leaves out policies broadly favouring road. Included policies involve investment in infrastructure and rolling stock to improve journey times, reliability and rail capacity. Also included are policies raising the user cost of road freight transport to cover its marginal social cost. (B) Pro-Road. This strategy includes a range of policies broadly favouring road and leaves out policies broadly favouring rail, e.g. the Working Time Directive (i.e. this strategy effectively implies its removal by 2010). Amongst included policies are investment in the road infrastructure to improve journey times and also an increase in the heaviest permitted vehicles on the road. Also included are the rationalisation of wagon load rail freight and a doubling of rail track access charges.
  
  The Pro-rail strategy has the effect of removing around 16% of tonne kilometres from road. Rail increases its share of tonne kilometres by around 31-33% over the three scenarios, and tonnes by 18-20%. Rail trips are also increasing in length, by between 11-12% and road trips falling by 15-16%, indicating that the extra traffic gained from road is long distance. Average length of haul overall falls by some 10%, as goods are sourced closer to their point of use. For the Pro-road strategy, there is an overall increase in tonne kilometres of between 3-4% across the scenarios. Road increases tonne kilometres by 5-6%. The largest component policy is allowing heavier road vehicles, which accounts for almost half of the overall effect of the Pro-road strategy. Rail’s tonne kilometres falls by around 10% over all scenarios, and tonnes by 6-7%.
  
  
• Consider the implications of the appraisal results for future rail research needs. We believe the models we developed in this project offer the capability to test the demand implications of alternative long term scenarios and strategies within a relatively simple framework, unlike the very much more complex models which exist within the industry for more detailed forecasting exercises. It did not prove possible within the resources of this project to integrate results on cost modelling and external costs with these models to obtain a complete appraisal of the options examined, and that remains a priority for further research.Both in the freight and passenger side the modelling exercises suggest good long term prospects for the rail industry depending on the policies selected, particularly on key passenger flows to and from London and within the South East, and for bulk freight. The obvious problem this poses is one of capacity. Further developments are needed to integrate rail capacity into the model in a way that is consistent with the strategic aims of the model, and then to use the model to test a range of capacity enhancing investments, alongside other ways of coping with demand such as through pricing policies.

OUTPUTS

Project Reports:

• Fowkes, A.S., Johnson, D.H., Whiteing, A.E., Nash, C.A. (2006). Freight Modelling Report. RRUK C3/5
• Whelan, G.A., Lythgoe, W., Menaz, B., Nash, C.A., Wardman, M. Li, Y., Preston, J.M. (2006). Rail Passenger Modelling. RRUK C3/4
• Fowkes, T., Johnson, D., Nash, C.A., Whelan, G.A., Whiteing. A.E. (2005). Modelling Proposals. RRUK Report C3/3
• Menaz, B., Nash, C.A., Shires, J.D., Whiteing, A.E., Beecroft, M. (2004). Rail/Road Strategies & External Scenarios. RRUK Report C3/2
• McDonald, M., Crockett, J., Beecroft, M., Nash, C., Menaz, B., Fowkes, T. (2003). The Role of Rail in Integrated Transport. RRUK Report C3/1 (prepared for RSSB)

Publications:

• Wardman, M. (2006). Demand for Rail Travel and the Effects of External Factors. Transportation Research E, 42(3) 129-148.5. Wardman, M., Whelan, G. (2004). Estimating the Effects of Service Quality Changes on the Demand for Rail Travel. Paper presented at AET European Transport Conference.
• Whelan, G.A., Wardman, M.R., Lythgoe, W. (2005). Examining the influence of station catchment areas on the demand for rail travel. Presented to the European Transport Conference, Strasbourg.

More detailed information and access to UK events associated with this project will be available to members of the Theme Network