Evaluation of Scheme Forcasting
Stefanos Psarras, AECOM, Russell Martin, Highways England
The paper investigates the benefits of full variable demand models, compared with the fixed demand or elasticity models and the appropriateness of an approach for ‘contingency’ schemes, involving a small modelled area with an elasticity model.
Highways England’s ‘Road Investment Strategy’ (RIS) seeks to deliver a long-term programme for motorway and major road improvements. This RIS represents a step-change in investment in the coming years. To help assess and justify these investment decisions, Highways England has commissioned five Regional Traffic Models, covering England, to develop an evidence base for the assessment of improvement schemes.
Given the geographical scope of these models, they are large and can take a long time to run (estimated 3-6 days). Highways England therefore wishes to understand how the modelling outcomes might be influenced by using the models in different ways, some of which would reduce model run times, and hence help reduce project programmes, costs and accelerate project delivery.
Highways England commissioned Atkins and AECOM to conduct research into the benefits of, and need for, full variable demand models (VDMs), compared with the alternative approaches of fixed demand or elasticity models. In addition, the research looks at the appropriateness of an approach adopted by Highways England for ‘contingency’ schemes, involving a small modelled area with an elasticity model. AECOM, once again being commissioned by Highways England to build on previous research, implemented different variable demand model formulations and combinations, in order to provide evidence and advice on how the Regional Traffic Models could be used when assessing schemes.
Two schemes with associated models suitable for the purpose of the research have been selected to implement different modelling techniques and sensitivity tests. The M25 Junction 16-23 and the A46 Newark to Widmerpool Improvement scheme were identified as suitable for the following reasons:
• the schemes have been modelled using a full variable demand model within DIADEM;
• the reports show that the convergence criteria have been met in the forecast year; and
• detailed economic analysis has been undertaken using TUBA.
The models and the sensitivity tests have been compared and assessed considering:
• Trip total impacts;
• Trip distance impacts;
• Network flows and flow changes;
• Demand impact by distance;
• Cost-Benefit Analysis regarding scheme journey time savings, user and provider benefits and user time benefits by sector movement;
• Accident benefits; and
• Convergence and stability of impacts.
Association for European Transport