Some Real Applications of MicrosimulationSummaryIn the September issue of Traffic Engineering and Control, the article An Introduction to Microsimulation suggested that the technique offers a unified approach to the analysis of traffic flow because it directly models its components instead of their mathematical proxy. Microsimulation* does not need to explicitly model vehicle platoons, shock waves, queuing behaviour at specific junction types, the local and strategic effects of bus priority schemes, or indeed the whole range of effects which result from traffic management measures at detailed, local, corridor or wide area levels. These phenomena occur because thay are a natural consequence of the modelling process. Microsimulation is synonymous with microsimplification, and schematically represents the individual vehicle driver in the individual vehicle reacting to other individuals and the immediate external constraints on freedom of movement. This simple definition of what actually constitutes road traffic is consistent with the view from behind the driver's wheel. The real components which contribute to traffic movement, such as kerb lines, traffic signals, bus lanes and no entry signs are either not moving, and therefore easy to represent in modelling terms, or, in the case of drivers, so heavily constrained that their movement can be readily described and tracked. Beyond the gruelling conditions of the supermarket car-park, the Toyota four-wheel drive Shopping Trolley has little opportunity to behave much differently from the clapped out Mini, and so road traffic distills to little more than drivers' aggression and awareness, which depend on what was eaten for breakfast as much as the vehicle driven. A randomised variation on normal driving behaviour and vehicle kinematics has proved sufficient to enable a wide range of commissions undertaken by SIAS using its S-Paramics microscopic traffic simulation software to be validated in circumstances where other contemporary systems might struggle. In its early use, S-Paramics was generally deployed where an alternative purpose-built tool broke down while operating outside its validated locus. It quickly became apparent that microsimulation offered a more convincing representation of road traffic across the board, and the technique is now in general use on commissions for central and local government and the private sector. This article includes examples of microsimulation projects ranging from individual junctions to wide area problems of rural and urban congestion, covering both common and unusual situations. Traffic Impact Assessment A stimulating impetus to microsimulation at SIAS has come from private sector clients, such as ASDA Stores and the brewing and leisure group, Scottish and Newcastle. Such organisations are very focussed, and hard task masters in covering all the angles to avoid failed planning applications. Microsimulation has proved popular with them because it delivers unequivocal results fast, and allows them as much participation as they have time for. It helps them to put across their case convincingly, and is easy to audit. A typical example concerns the proposal by ASDA Stores Limited to construct a new superstore on a site at Shirley, in Solihull, W. Midlands. ASDA's proposals include the reconstruction of the existing signalised crossroads at an exceptionally busy junction to form a new roundabout, with an additional linked signalised junction to provide access to the superstore (Figure 1) . In addition to traffic capacity benefits, the new junction will provide for fully-controlled pedestrian crossings on all approaches. Previous analysis undertaken with TRANSYT software had led to the conclusion that the proposed layout would not operate to the satisfaction of the local highway authority. SIAS was asked to re-examine the situation by applying S-Paramics to the data prepared for the previous study in order to rigorously test design options. Microsimulation is characterised by its realtime visualisation facility, used here to refine the road layout detail and signal settings. A significant increase in capacity for the new roundabout was achieved beyond that of the current junction layout, despite the inclusion of pedestrian phases and store access. The S-Paramics exercise was conducted in consultation with Solihull Metropolitan Borough Council and their independent auditors. The road layout proposals have been developed to ensure that the traffic aspects satisfy the planning application criteria and service the ASDA superstore while providing additional local road network benefits extending to Shirley town centre. Paisley Town Centre If a history of microsimulation were ever to be written, then Paisley should occupy Chapter 1. Paramics cut its teeth on Paisley, and the developers are indebted to Renfrewshire Council and Renfrewshire Enterprise for their act of faith. With a population of 80,000, Paisley has filed application to be Scotland's next city, and in early 1996 it became the first major urban area in the World to have its road network represented by a microscopic traffic simulation model. As the principal town to the west of Glasgow, Paisley has a complex mix of problems stemming from the inclusion of the M8 motorway, the dense nature of the town centre, and the development activity around Glasgow Airport. Previous transportation studies based on TRIPS and SATURN had not provided sufficient information for a detailed operational design and evaluation assessment, so data was converted for use in an S-Paramics microscopic simulation model. This was developed to test traffic management proposals in the Town Centre Action Plan, involving the removal of through traffic and the use of exclusive bus and taxi routes. The model helped to define traffic control options, and proved to be key to the public consultation process. The police seemed particularly impressed by the fact that Day 1 of scheme implementation saw a radical improvement in the traffic flow, obviating the necessity for the usual on-site iterative adjustments, a factor which has subsequently become a feature of S-Paramics implementations elsewhere. The Market Town of Lanark With a population of 9,000, Lanark may be small, but it's got some excruciating traffic problems. The road network has remained unaltered for decades, and is dissected by the busy A73 through route (Figure 2). This doubles as the High Street, where service vehicles and illegal parking compound the problems, resulting in pedestrian/vehicle conflicts to deter both shoppers and traders. Strathclyde Regional Council had previously conducted a comprehensive traffic review in Lanark, which resulted in a range of short and long term options for a Town Centre Action Plan. This was left to be developed by its successor authority, South Lanarkshire Council, and SIAS carried out traffic studies and developed an S-Paramics microscopic simulation model to test a series of traffic management options for the Plan. Testing included a variety of development options and the phased implementation of a ring road. S-Paramics was able to directly model the complexities of this highly constrained road network, and draw clear distinctions between design options. A principal attraction lay in its clarity of presentation to planning committees and for public consultation, to such an extent that South Lanarkshire is now the first local authority in the UK to have an in-house microscopic modelling capability. Perth Despite its picturesque location on the River Tay, or perhaps because of it, Perth shares with other large towns and cities the blight of traffic congestion. The problem is accentuated by the physical constraints of the river and Perth's location at the hub of strategic traffic movements. In order to plan against this background, Perth and Kinross Council commissioned an S-Paramics microscopic simulation model of the Perth urban area, because of the model's sensitivity to both major and marginal traffic management measures, and its ability to test the Council's wide area traffic management proposals. The visual nature of the output was also seen as essential for presentation to Committee and public consultation. Travel demand data from an earlier SATURN model was enhanced by extensive field surveys, and the road network definition upgraded to the S-Paramics standard, to test a variety of traffic planning options. These included pedestrianisation in the city centre, bus priority measures, revised junction control, a new crossing of the River Tay, and a unified approach to the assessment of the impact of traffic from major new developments. The S-Paramics model included, for the first time, a detailed simulation of the interaction of bus priority measures on other road traffic, including pre-signals for bus lanes. Dundee Down river from Perth lies the Bonnie City of Dundee. Unlike most cities associated with their famous rivers, Dundee does not straddle the Silv'ry Tay, but remains firmly fixed on its north bank, right up to the edge (Figure 3). This results in an unusual road layout, with some real complications bringing the Tay Road Bridge into a tight site between the city centre and the river. An S-Paramics wide-area simulation model has been developed on behalf of Dundee City Council and Scottish Enterprise Tayside in order to test a variety of traffic related planning proposals in and around Dundee city centre. These two bodies have a history of close collaboration on local planning, and consider an S-Paramics traffic model to be a particularly effective communication tool, both for their own working groups and for public consultation. The model is being used to estimate the optimum size and location of new developments by evaluating their traffic impact, and to assess a range of traffic management scenarios in order to accommodate additional pedestrianisation and bus priority measures. Cordoned-out sections of the model will be used for detailed design of local areas, one of which has already been undertaken to assess the traffic implications of a major new development at City Quay in the Dundee docks area (Figure 4), close to the city centre and Tay Road Bridge. The S-Paramics model highlighted problems relating to blocking back from the junction of the A92/Inner Ring Road to other junctions, and a need for additional storage capacity on the Ring Road for traffic requiring access to a new food store within the area. This had not been anticipated in a separate traffic impact assessment conducted using standard junction software. Edinburgh Readers might know that Edinburgh lies to the south of Dundee, but perhaps not that it is further west than Bristol (but not at the same latitude, of course). Edinburgh is rarely out of the traffic news, and while much radical action is in prospect, until recently an understanding of the consequences went largely unaided by the available analytical processes. Interminable committee wranglings have now given way to debate enlightened by a real-time visualisation of traffic management options, and decisions are being made with unusual alacrity. An S-Paramics microscopic traffic simulation model of the city centre (Figure 5) has been developed to assist the City of Edinburgh Council in testing a proposal to remove westbound traffic from Princes Street and to provide responses to questions raised through the New Town Forum of Urban Living, and at other public consultations. While complementing the Council's existing modelling strategy, the microscopic model offers features additional to those of its current TRIPS based system, which remains the principal source of input data within a strategic framework. The S-Paramics model enables the detailed examination of a wide range of traffic management and transport policy options including bus priority, parking, and lane restrictions, and can identify the typical effects of urban congestion such as junction interaction including blocking back, buses queuing to access bus stops and consequential lane blocking. Estimates will be made of vehicle emissions resulting from vehicle accelerations and queue idling, and microsimulation offers the potential fr testing the more fundamental options which might emerge with governmental encouragement. A detailed treatment of microsimulation in Edinburgh will be described in a later issue of Traffic Engineering and Control. Microscopic Modelling on the Trunk Road Network S-Paramics microscopic models have been developed for stretches of the M4, M5, M25 and A419 in England for Highways Agency and the Department of Transport (as was), while in Scotland a comprehensive microscopic simulation exercise has been undertaken on the Edinburgh periphery, with emphasis on the notorious western approaches. Following local government reorganisation in April 1996, the National Roads Directorate at The Scottish Office Development Department (NRD) assumed responsibility for the greater part of the A720, the Edinburgh City Bypass. Since completion, the road has acted as an outer distributor and stimulator of development, and is now operating over capacity for long periods throughout the 7-day week. NRD commissioned SIAS to undertake an Investigation of the A720, conceptually a strategic road, but suffering operationally from its mix of functions. The Investigation included engineering and environmental appraisals, and a market research study, and centred on a traffic modelling approach based on an S-Paramics microsimulation model of the entire 25km length of the A720, as well as adjacent sections of the M8, A8 and M9. Effects modelled and reported on included merging behaviour, blocking back, gradients, shock waves, signing, potential ramp metering, lane closures and lane marshalling, incidents and road works. Trunk Road Signalised Roundabout Following the Scottish Office NRD's initial study for the A720 Edinburgh City Bypass, some fundamental proposals have been evaluated using the microsimulation model, including junction redesign, road widening, hard shoulders and climbing lanes. Market research had supported the commonly held view that the unsignalised at-grade Sheriffhall roundabout was dangerous. This lies on the Edinburgh-Dalkeith A68/A7 route, where east-west A720 traffic crossed at high speed, leaving little opportunity for other traffic to enter and circulate safely. It was virtually impossible for cyclists and pedestrians to cross. Market research had indicated that signalisation would be welcomed, but that there was concern over further congestion. The A720 S-Paramics model was used to test traffic signal operation at the Sheriffhall Roundabout (see Figure 3 p.483, September issue) and to investigate whether such a solution could improve safety without adverse capacity effects. The study went further, and for the first time in the UK, a trunk road design was designed and tested by microsimulation, the recent implementation fully justifying the approach. The Sheriffhall design included the introduction of spiral lane markings, additional entry lanes, and the altering of lane widths, signal positions and stop lines. The full programme of tests included optimising the signal timings for four new layouts, each one a stage towards potential grade separation. The model demonstrated how signals could be installed without an increase in congestion, and identified capacity improvements which were carried through to the design specification. Trunk Road Economic Evaluation A large concern at a particularly congested eastbound bottleneck on the A720 Edinburgh City Bypass is the disruption caused by incidents. As in many other trunk road situations under congested conditions, an incident cleared at one point on the network can generate a standing queue within a shock wave over an hour later at some 5km distant, an event which S-Paramics can simulate and evaluate. Such effects can outweigh the conventional economic benefits assessed by COBA and NESA, and microsimulation provides an opportunity to measure these within complex design contexts, since it is able to reflect marginal network changes which cannot normally be assessed. An analysis using the A720 S-Paramics model evaluated two scenarios against the base, to distinguish between disruption caused by incidents on a section upgraded to dual 2-lane with full hard shoulder, or full dual 3-lane. The A720 experiences on average three incidents per day involving varying degrees of traffic disruption, and the use of the incident data meticulously collected by Lothian and Borders Police ensured accurate incident allocation within the model, which was run continuously for a seven day week. The model output, in terms of total time spent on the Bypass for the base and two design scenarios, underwent a standard cost/benefit analysis, to produce a benefit cost ratio of up to 2.5 and first year rate of return of up to 12.5% derived from incident congestion relief alone. Although only time benefits were assessed, the role of microsimulation in identifying hitherto "missing" economic benefits has been clearly established. Toll Plaza Simulation at the Forth Road Bridge The problems of Edinburgh's western approaches embrace the Forth Road Bridge. In the autumn of 1997 work began on the removal of toll booths for southbound traffic crossing the Bridge, and the upgrade of the northbound toll plaza to accommodate eight toll booths. It was reasonably expected that southbound traffic would flow more freely as a result, and that northbound queuing would be reduced. However, immediately to the south of the Bridge a significant proportion of southbound traffic leaves the A90 for the Echline Roundabout, to merge with other traffic destined for the A8000. During the morning peak period, and at other times at weekends, the queue on the Echline ramp backed onto the Bridge, negating the expected benefit of toll booth removal. Indeed, it became apparent that the removal was likely to contribute to the problem by increasing southbound traffic speeds, resulting in a more intense shock wave generated at the top of the Echline ramp. The Forth Bridge Board commissioned an S-Paramics model to examine the problem by way of microscopic simulation. The model encompassed the southern end of the Bridge, the toll plaza, the Echline Roundabout and approaches (Figure 6). It became quickly apparent that while full signalisation of the roundabout would be beneficial, partial signalisation on the ramp from the Forth Bridge would achieve almost the same effect. This would be to regulate the queue on the ramp and reduce the potential for queueing on the Bridge, while significantly reducing the historic queue of northbound traffic on the A8000. This in turn had potential for generating new problems at the toll plaza, but the simulation model demonstrated that the additional toll booth would create sufficient additional capacity.The single set of traffic lights on the Forth Bridge ramp to the Echline Roundabout became operational in early January 1998, and proved very successful, although subsequently developed into a fully signalised system. Queueing on the Bridge has been significantly reduced, and the anticipated benefit to traffic on the A8000 has been achieved. The problem was defined, solved, and alleviated in the space of just six weeks, a programme made possible by the use of microscopic simulation. Rural Schemes Congestion also occurs in the countryside. William Wordsworth and Beatrix Potter have a lot to answer for, having contributed to the popularity of the Lake District in general, and the route between Ambleside and Windermere in particular. This 6km stretch is known for its summer and holiday traffic queues, and the journey time is regularly extended beyond thirty minutes, queuing traffic often covering the entire route. Cumbria County Council has been invited by the Government to investigate a rural bus priority scheme here as a sustainable transport pilot scheme. Congestion has reduced the attractiveness of buses in the summer, so a scheme has been devised to reduce their journey times. As the concept is new and untested, the Council commissioned a microsimulation model to demonstrate its effectiveness. Due to physical constraints, the scheme covers a 1.5km section only, and involves a 3.5m bus lane with a carriageway reduction to 3m in each direction for other vehicles, requiring the adoption of a 30mph or 40mph limit. In addition to taxis, coaches and HGVs may use the bus lane, because of issues of safe passing. There is a simple diverge at the start, with a bus gate at the end to provide signal priority to the bus lane. The S-Paramics microscopic simulation model was completed within two weeks and used to test the base and the scheme with a 30mph or 40mph limit. It was shown that buses can reduce their journey times, and that new passing opportunities present off-peak benefits to all vehicles. Cars were not significantly delayed, their slight increase in journey times being a result of the speed restrictions. With the assistance of the S-Paramics demonstration it was agreed that the bus priority scheme should be designed in detail and put to the Government for funding. Radical Designs Because microsimulation offers a universal approach to traffic flow analysis and design, it is not dependent on deterministic distillations from existing similar circumstances. Microsimulation has a good track record of predicting events not explicitly modelled, and is therefore particularly suitable for the testing of radical road layouts. Highways Agency is examining a number of these, and recent media discussion of a new type of road intersection in the USA prompted SIAS to undertake a short desktop exercise to investigate claims related to increased capacity and safety of what is becoming known here as the Displaced Right Turn (DRT) junction. Although not yet implemented in the UK, the current level of interest shown by planning authorities suggests its acceptance in due course. The DRT design involves moving stop-lines further back from the centre of a controlled junction in order to provide storage capacity for left and straight ahead movements, to enable right turn movements to be undertaken without having to provide a dedicated phase. The layout enables pedestrian crossings and bus-priority measures to be more easily accommodated in the design. Right-turn phases greatly reduce overall capacity and increase the potential for conflict, and the DRT concept helps to minimise these aspects. Most of SIAS's project work relates to client commissions, but the availability of the S-Paramics microscopic system enabled a short exercise to be undertaken with little effort. S-Paramics has already been applied to a wide variety of junction designs including uncontrolled motorway intersections, signalised roundabouts and complex signalised arrangements. It has been observed that, in congested conditions, the characteristics of driver behaviour and vehicle kinematics within S-Paramics are universally applicable, and that calibration to selected locations has generally resulted in successful validation elsewhere, provided that the road network and junction layouts have been properly described within the model. As a result, S-Paramics has not yet failed to be validated against field observations, often where alternative analytical methodology breaks down. It was therefore considered likely to give a good estimate of the performance of the DRT junction, and provide a tool to achieve an optimum design. The layout in the example in Figure 7 is not a refined design, but a first pass at the junction in Figure 1, currently being considered for conversion to a signalised roundabout. Even without optimisation S-Paramics confirms findings in the USA, since the preliminary design in Figure 7 yields a capacity improvement of up to 20%. Other designs for desperate situations might include ramp metering, for which no convincingly authentic analytical methodology has hitherto existed. As part of its A720 Investigation, the Scottish Office NRD required preliminary consideration of ramp metering options. The example in Figure 8 is a detail from the A720 simulation model, and shows the testing of peak period ramp metering at the A720/A702 Lothianburn junction. The objective here, as elsewhere, has been to minimise queuing delays to traffic joining the A720 and consequential blocking across adjacent junctions, whilst enabling smooth merging with main line traffic. The narrow two lanes of the A720 are prone to shock waves triggered by braking near on-slips, leading to considerable blocking back. The detector loop locations, flow threshholds by lane and cycle times have been optimised to reduce these effects. The operational benefits of ramp metering in terms of journey times, journey reliability and safety have been demonstrated within the wider context of the entire A720 corridor and assessed in relation to other ameliorative proposals. It should be emphasised that there are no current plans to introduce ramp metering on the A720. Conclusion Although the equilibrium pipe-flow analogy traffic modelling systems developed in the 1960s to represent traffic flow have undergone some development, they are too often deployed in unsuitable circumstances. Microsimulation picks up where they leave off, where road traffic begins to seize up. Because the game is now less about road building and more about making best use of existing road space, new capacity needs to be found, maintained and ultimately sustained. It appears to be there for the taking, and the above examples have all been about finding the capacity and using it to best effect. It will be a political and marketing job to ensure that this space does not fill up again. ACKNOWLEDGMENTS The author is grateful for material in compiling this article to: ASDA Stores Limited Cumbria County Council Dundee City Council The City of Edinburgh Council Forth Properties Limited Perth and Kinross Council Renfrewshire Council Renfrewshire Enterprise Scottish Enterprise Tayside South Lanarkshire Council The Forth Bridge Board and The National Roads Directorate at The Scottish Office. Footnotes * The first known public sightings of the term 'microsimulation' were in papers by Peter Bonsall in 1979: 'Microsimulation of organised car-sharing: model predictions and policy implications' (TRRL WP/SRB11), and 'Microsimulation of mode choice: a model of organised car-sharing', presented at the 1979 PTRC Summer Annual meeting.
|
Important handbooks available to SIAS clients and S-Paramics licence holders |
