Discussion on Papers 10–11

DR D. COOMBE, Halcrow Fox and Associates

With reference to Paper 10, consistency between traffic forecasting and economic evaluation is very important. Current practice often involves using a model to forecast traffic flows and speeds, followed by COBA to calculate the economic benefits. COBA recalculates the speeds and delays, based on the flows output by the traffic model, and often incorrectly as COBA cannot deal with junction interactions, that is, bottlenecks which meter flow downstream and cause queues to form through upstream junctions. It would be better to use one of the new congested assignment models - for example, SATURN or JAM - which calculate delays reasonably well, followed by matrix based economic evaluation using COBA standard values of time, operating costs and so on. The trip matrix can be fixed, or variable to take account of trip generation/suppression (important), peak spreading (important), distribution and modal split (often of little importance as effects of road schemes). COBA, as it stands, has little value as an urban economic evaluation tool, although I find it is of value in the inter-urban scene.