Evaluating Active Traffic Management Systems on the M25 Motorway

Evaluating Active Traffic Management Systems on the M25 Motorway


S Moriarty, T Wang, Kellogg Brown & Root, UK



Demand for travel using the M25 motorway around London continues to grow. The UK Government?s integrated transport policy supports action to manage the demand for road journeys where transport problems are intense. One possible method, suggested by a Government funded study, is to control and prioritise the limited available road space through technological innovation. The study advocated road user charging as its favoured longer term option.

This brief note presents a summary of a recent work in assessing the impact of management system options in terms of traffic operational conditions on the M25 between junction 16 and junction 21 during the peak period. A range of modelling outputs was used to assess the merits of the alternative active traffic management systems. For each traffic management system, key indicators such as variability in journey time and traffic levels on the M25 and on approach roads were examined. In addition sensitivity tests were conducted into the variability of the micro-simulation.

The GETRAM traffic micro-simulation suite was used to test the systems considered. The GETRAM package, consisting of a graphical editor TEDI, and the micro-simulator AIMSUN, was used to develop a model with the study area covering the motorway network in the north-west quadrant of the M25 between the M40 and the M1 corridors. The traffic flow demand data for the micro-simulation modelling was extracted from a larger separate calibrated/validated NAOMI version 5 SATURN model. The traffic demand flows were profiled over a three hour period of 07:00 to 10:00 hour period. The demand was disaggregated into different vehicle categories representing cars with 0, 1, 2 or more passengers, light goods vehicles with 0, 1, 2 or more passengers, and heavy goods vehicles (HGV). In addition a matrix of bus and coach services (PSV) was synthesised from service patterns and frequencies. The base year GETRAM model was calibrated to traffic flows and journey times in 2001 for the average morning peak period between 07:00 and 10:00 hours.

The calibrated validated 2001 base year model was used to prepare a 2011 forecast year model, which was used for systems testing. Forecast demand data for 2011 was derived from the NAOMI model.

For the micro-simulation tests, all of the systems tested included the proposed widening of the M25 from 3 to 4 lanes, and the following traffic managements system were considered separately:

* Mainline metering
* Ramp metering
* Dedicated public transport lane (i.e. bus/coach only lane)
* High occupancy vehicle lane (i.e. lane for vehicles with 2 or more occupants)

In the mainline metering system, non-priority traffic, such as cars and light goods vehicles, would pass through a metering plaza, which consisted of traffic signals which controlled the passage of vehicles on each lane so that each vehicle would incur a time penalty. Priority traffic, such as heavy goods vehicles and public service vehicles, would bypass the metering plaza using segregated lanes, which were free-flow. The mainline meter needs a considerable distance on the approaches to allow priority and non-priority traffic to separate into their appropriate traffic stream. For the north-west quadrant of the M25, the only suitable location where there was a sufficient length of approach to accommodate a plaza within the existing highway boundary was between J16 and J17.

In the vicinity of the plaza area there was local widening to five lanes. The mainline meter had two priority vehicle lanes and three non-priority metered lanes. The matrix signs on the approaches to the plaza were used to reduce the traffic speed on the approaches from 70mph to 30mph. The access to the segregated priority lanes was from the inside lane of the motorway. This widened out to two lanes on the segregated bypass and had a speed limit of 70mph. Beyond the mainline meter, the priority lanes would merge with the three non-priority lanes. The priority lanes were restricted to HGV and PSV vehicles but all traffic could use the non-priority lanes.

The signalised mainline meter was set at a rate such that the flow permitted would be the same as the flow that would have occurred on the M25 between J16 and J17, if road user charging had been introduced. From a consideration of the NAOMI model assignment with road user charging, the flow rate was set at 4,175 vehicles per hour for each carriageway for each direction of the M25.

In the ramp metering system, metering points were installed on the entry slip-roads to the M25 at J17 to J20 inclusive (although the northbound on-slip at J18 was excluded for road safety reasons). The meters were used to control access to the M25 in order to manage demand.

The ramp meter flow rate was set at a rate similar to the traffic flow that would have occurred on each slip-road, if road user charging had been introduced. The following flow rates were determined by examining the forecast traffic flows entering the M25 from the slip-roads using the NAOMI v5 model. The range of flow-meter levels varied between 400 and 1,750 vehicles per hour for each on-slip entering the M25.

In the dedicated public transport lane system the fourth lane was reserved for the exclusive use of public transport vehicles, although these vehicles could use any available lane. The other vehicles were confined to three lanes so they would not receive any benefit from the additional capacity provided by the fourth lane, except that they may have less competition from coaches and buses. The options with the dedicated public transport lane located on either the inside or outside lane were considered.

In the high occupancy lane system, the outside fast lane was reserved for vehicles with 2 or more occupants. Other traffic was restricted to the remaining three lanes but these lanes could also be used by high occupancy vehicles.

The main findings of the study were as follows:

* It was possible to identify the impact of the different active traffic management systems.
* Of the traffic management systems modelled on the M25 motorway between J16 and J21, the ramp metering system seemed the most effective in controlling traffic operations by promoting smooth traffic flows on the M25 active mainline.
* Of the systems considered the dedicated bus lane on the outside lane seemed to be the least effective. In congested traffic conditions, public transport vehicles found it difficult to weave across the congested three non-priority lanes, resulting in a low percentage usage of the dedicated public transport lane. The main line metering system performed poorly as it generated additional delays to all traffic, which were attributable to a number of reasons including weaving.


Association for European Transport