Modelling Flow Breakdown at Merges on Congested Motorways



Modelling Flow Breakdown at Merges on Congested Motorways

Authors

Clive Gilliam, Hyder Consulting (UK) Ltd, Craig Drury, UK Highways Agency, Ian Black, Independent Consultant

Description

HSMx (Hyder Simulation Model in Excel) is an aggregate simulation model of flow breakdown on motorways that occurs at merges (‘standing wave’ flow breakdown). It can forecast the impacts of motorway merge flow breakdown, which current highway assignment models are unable to do.

Abstract

HSMx (Hyder Simulation Model in Excel) is an aggregate simulation model of flow breakdown on motorways that occurs at merges (‘standing wave’ flow breakdown). It can forecast the impacts of motorway merge flow breakdown, which current highway assignment models are unable to do. With a suitable demand profile HSMx can predict both average travel time and travel time variability by 15 minute time periods over a whole day. When flow breakdown occurs the model takes account of queues and delays arising from previous periods.

HSMx has two key components:
• a “Breakdown Probability Function”, which forecasts the probability of flow breakdown occurring for a given level of merge demand (main line plus on-slip); and
• a “Queue Discharge Function”, which forecasts the rate at which the queue discharges once flow breakdown has occurred.

In a study for the UK Highways Agency the parameters of these two functions were estimated using traffic flows and journey times (derived from MIDAS data) for a set of over 100 links and merge junctions on the UK motorway network. The stability of these parameters for different merge types and motorway carriageway standards and the reasons for their variation will be presented.

Findings will be compared with a small number of related studies in the US and Germany.

A method for incorporating HSMx into standard highway assignment packages and the likely improvement in forecasting accuracy will also be described.

HSMx requires detailed information on demand profiles and the breakdown behaviour of individual junctions. It may appear therefore that the case for the adoption of HSMx falls on the grounds of too much complexity and the absence of essential data. But this is not the case. The study has demonstrated that suitable data is available and, despite quality difficulties and issues such as blocking back, can be successfully analysed. HSMx can deliver accurate predictions of both average travel time and travel time variability, and its computational requirements are modest (operating as it does within Excel).

Publisher

Association for European Transport