2.15 Transport models

Validation of transport models

When transportation models are used to generate demand forecasts and assign traffic to transportation networks, documentation should be provided to demonstrate the models have been correctly specified and produce realistic results. The documentation is listed in the series of checklists in worksheet 8.3 and these should be completed for each analysis time period.

The aspects of the models covered by the validation checks are as follows:

• Project model specification - including model type and parameters, data sources, trip matrices, assignment methodology and forecasting checks
• A base-year assignment validation - comprising checks on link and screen-line flows, intersection flows, journey times and assignment convergence
• Strategic demand model checks - incorporating validation of the models and techniques used to produce trip matrices

Model reviewers may also use these checklists to confirm that appropriate documentation has been provided for review purposes.

Checks on output of traffic models

All project benefits calculated using a traffic or transportation model shall be checked to show the results are reasonable. The checks shall be done and reported at two levels - coarse checks and detailed checks.

Coarse checks

The objective of these is to check if the travel time benefits calculated are of the right order of magnitude. Travel time savings per vehicle shall be calculated for both the first year of benefits and a future year by dividing the daily travel time savings by the AADT of traffic traversing the project (worksheet 8.1).

Detailed checks

The objective of these is to ensure the travel times on individual road sections, through critical intersections and for selected journeys through the network, are reasonable. This analysis shall be undertaken for the first year of benefits and for a future year, and for both peak and off-peak periods if appropriate (worksheets 8.2).

These checks shall cover:

• road section speeds for both the do minimum and the project options
• peak-period delays and volumes at critical intersections for both the do minimum and the project options. Delays shall be based on the intersection approach which incurs the greatest delay
• travel times for both the do minimum and the project options for journeys which review the major travel time benefits, based on the travel time savings per vehicle for each journey route
• a comparison of the total travel time savings for journeys which receive the major travel time benefits and the total travel time savings predicted by the traffic model, in the first year of benefits and a future year.

Evaluating congested networks and induced traffic effects

Guidelines are provided in appendix A11 for modelling situations where very high levels of congestion are anticipated over the economic life of the scheme. Professional judgement should be used to determine the appropriate procedures to adopt. In cases where there are excessive or unrealistic levels of congestion in the do minimum network, a number of techniques may be used to generate a realistic and stable representation of the do minimum context. These commonly involve upgrading the capacity of the do minimum network or using some form of growth constraint on the trip matrix, such as matrix capping.

The matrix derived from this process remains the same in both the do minimum and project option, and is then used in the standard fixed trip matrix (FTM) evaluation procedure. Appendix A11 provides details of growth constraint techniques.

In some rare situations, significant levels of congestion may be expected in the project option across important parts of the network (spatially) and affecting a substantial proportion of the economic life of the project (temporally). The resulting induced travel may affect option benefits as well as the choice of roading option in a corridor. Where this is the case, the evaluation should incorporate an analysis of induced traffic effects.

If the expected effects will be sufficiently similar in each option (so that the relative economic performance of the options would not be affected significantly), it may be feasible to choose between options without a separate analysis of induced travel effects.

If significant induced traffic variations between options are not expected, or only a single preferred option is to be evaluated, advice should be sought from Land Transport NZ as to whether an analysis of induced travel effects should be completed. Appendix A11 contains procedures for evaluating induced traffic effects.