Faculty of Engineering


Developing Auckland's cycleway network – how do we prioritise cycle infrastructure projects?

This project focuses on improving the current economic evaluation method to incorporate more comprehensive demand forecasting, and to take into account interdependencies between projects such as joining two existing or proposed cycle paths.

In the New Zealand Transport Strategy 2008, the Government set a target to increase walking, cycling and other active modes of transport to 30% of total trips in urban areas by 2040.

The objectives are to:

  • Assist economic performance.
  • Alleviate congestion.
  • Improve access and mobility for all road users.
  • Improve the reliability and resilience of transport networks.
  • Reduce the use of fossil fuels.
  • Improve public health by increasing use of active transport modes and decreasing air pollution.
b-cycle-network1

As with all public infrastructure requirements, construction of the proposed cycle infrastructure needs to be phased and requires a means for prioritising projects over time. The existing project selection strategy uses benefit cost ratios (BCR), with high BCR projects prioritised on the basis of a simple rank order. Estimated usage over the life of the cycling infrastructure is a major component of the BCR computation, hence demand forecasting is crucial in this analysis.

cl-walk-bike-sign1

Key focus areas/issues


The key focus areas are:

  1. Developing a robust demand forecasting model.
  2. Capturing interdependencies between projects.

Current major developments


A mathematical model to assess user benefits derived from project interdependencies is being developed at The University of Auckland. This enables joint assessment of projects, for example, with high BCRs that may be linked to projects with low BCRs. In some scenarios it may be beneficial to prioritise construction of a low BCR project in order to increase connectivity in the cycle network as a whole.

A practical example is a case where two cycleways, which are inexpensive to construct and have high BCRs, are linked by an overbridge which is relatively expensive to construct. The bridge as a stand alone project may not have a sufficiently high BCR ratio to justify approval using the traditional selection method, but nevertheless can be justified when considering the impact on the whole system. We call this the project bundling or portfolio effect.

Connectivity of cycle paths is a key determinant of usage by cyclists and hence taking into account connectivity allows a more complete analysis of the benefits. To achieve this we estimate the benefits of projects, as represented by BCRs, for individual projects and also for combinations of projects.

A new method for forecasting demand estimates was proposed that:

  • Assesses how cycle friendly a road is depending on traffic, road type, etc.
  • Uses estimates of how many cyclists currently travel from various origins to various destinations on the road and cycle path network
  • Incorporates impact of new cycle paths on the entire network.

This has led to development of a computer programme to obtain the number of cyclists that can be expected to use any road or cycleway in the network. In particular, this enables estimation of the number of cyclists on a cycleway that may be constructed as one of the cycle infrastructure projects.

The consideration of the portfolio effect combined with the new demand forecasting method provides a new, comprehensive approach for the evaluation and selection of cycle infrastructure projects that maximise benefits for Auckland.

Key achievements


cl-cycle-walk-sign-path1

As a case study, our methodology has been applied to a corridor along the Southern motorway between Princes Street in Otahuhu and Carlton Gore Road in Newmarket close to Auckland's central business district (CBD). Assuming several new cycleways can be built along this route, our model forecasts that demand increases when adjacent connected cycleways are bundled together and also with proximity of those new cycleways to the CBD.

Assuming a limited budget that would allow for the construction of only some of the proposed cycleways, our work shows that a maximum benefit, in terms of predicted demand and derived BCR scores, is obtained by constructing the connected cycleways closest to the CBD. This preliminary analysis has been conducted by Uttara Nataraj as part of her student research project. Future work includes application of the methodology to other areas of Auckland and validation against actual projects as and when cyclist usage data becomes available.

For road controlling authorities with budget constraints for new cycleway infrastructure, our methodology provides a project selection and prioritisation approach based on whole network benefits, rather than on an individual project basis. This may be a useful decision making aid for funders in the delivery of best value for money, to meet the latent demand for cycling projects in the Auckland network, to assist in improving community health and ultimately could help in the development of an integrated world class cycleway network for Auckland.

Key people


  • Andrea Raith
    Engineering Science
  • Matthias Ehrgott
    Engineering Science
  • Garry Miller
    Civil and Environmental Engineering
  • Mieszko Iwaskow
    Senior Transport Planner Auckland, NZTA
  • Keith Pauw
    Senior Transport Engineer Auckland, NZTA

Contact


Andrea Raith
Email: a.raith@auckland.ac.nz
Phone: +64 9 373 7599 extn 81977

Related publications


Raith, A., Nataraj, U., Ehrgott, M., Miller, G. and Pauw, K., 2011.  Prioritising cycle infrastructure projects. Australasian Transport Research Forum Proceedings 2011.

Nataraj, U., 2010. Selecting a Portfolio of Cycling Projects. Department of Engineering Science Part IV Project Report, The University of Auckland.

Raith, A., Van Houtte, C., Wang, J.Y.T., Ehrgott, M., 2009. Applying Bi-objective Shortest Path Methods to Model Cycle Route-choice. ATRF 2009 Proceedings.

Raith, A. and Ehrgott, M., 2009. A comparison of solution strategies for biobjective shortest path problems, Computers and Operations Research, Vol 36, no. 4, pp 1299-1331.

Top