"The Winter Model" ("Vintermodellen") - is a computer-based system for assessing the most important socio-economic consequences of different strategies for winter road maintenance. A basic part of the model is measurement and analysis of vehicle speed reduction in different winter road conditions, compared with speeds in dry, bare road conditions. The model is developed for roads with a maximum width of 9.5 metres, however, it is important to include larger and more trafficked roads in a further development. In this study, measurements have been performed on two wider roads: a 13 metre road, i.e. the roadway is 7.5 m wide, and the shoulders are each 2.75 m wide; and a 2+1-road, i.e. "collision-free road" with a cable barrier separating the traffic directions, and alternately two lanes in each direction. The vehicles were divided into three categories: passenger cars, trucks with no trailers (incl. buses), and trucks with trailers.

The Norwegian Public Roads Administration has expressed a wish that the expertise of VTI in infrastructure maintenance should be utilised to improve the decision base for the best possible operation and maintenance of the Norwegian road network. VTI was therefore commissioned to estimate the effects for road users of a number of operation and maintenance measures and strategies. This report gives an account of one of the projects. At VTI, within the project Tema Vintermodell, a methodology has been developed to study the accessibility of vehicular traffic in different road conditions. Vehicle speeds are continually measured during the winter season with simultaneous observations of the prevailing weather and road conditions. The reductions in speed in different road conditions, in relation to speeds on dry bare roads, can thus be calculated. The Norwegian Public Roads Administration considers that, for Norwegian conditions, it is most important to study the road network carrying low volumes of traffic, with an average annual daily traffic (AADT) less than ca 1,000 vehicles.

Syftet med Vintermodellen är att beräkna och värdera de väsentligaste konsekvenserna för trafikanter, väghållare och samhälle av olika strategier och åtgärder inom vinterväghållningen. Huvudrapporten är i sig en sammanfattning av de rapporter som beskriver Vintermodellens olika delmodeller. Navet i Vintermodellen är Väglagsmodellen, som utgående från väderdata, vidtagna väghållningsåtgärder och trafik beräknar väglaget timme för timme under vintersäsongen. Väglagsmodellen styr beräkningarna i de olika effektmodellerna: Olycksmodellen, Framkomlighetsmodellen, Fordonskostnadsmodellen, Miljömodellen och Modellen för väghållarkostnader. I Olycksmodellen beräknas olyckskvoter, olyckstyper och konsekvenser, allt kopplat till olika väglag och deras varaktigheter.

I Framkomlighetsmodellen beräknas olika väglags effekt på medelhastigheter och restider.

I Fordonskostnadsmodellen beräknas kostnader för bränsleförbrukning och korrosion på grund av vägsalt.

I Miljömodellen beräknas konsekvenserna för vägnära vegetation av vägsalt.

I Modellen för väghållarkostnader beräknas dels direkta kostnader för åtgärderna, dels kostnader för skador och slitage på beläggning, vägmarkeringar etc. som följd av vinterväghållningsåtgärder.

The aim of the Winter Model is to estimate and put a value on the most important impacts of the strategies and measures in winter road management for road users, road management authorities and society at large. The main report itself is a summary of the reports that describe the submodels in the Winter Model. The hub of the model is the Road Condition Model which, on the basis of weather data, undertaken road management measures and traffic, calculates road conditions hour by hour during the winter season. The Road Condition Model controls calculations in the effect models: Accident Model, Accessibility Model, Vehicle Cost Model, Environment Model and Model for Road Management Costs. The Accident Model calculates accident rates, accident types and consequences, all coupled with different road conditions and their duration. The Accessibility Model calculates the effect of different road conditions on mean speeds and trip times. The Vehicle Cost Model calculates the costs of fuel consumption and corrosion due to road salt. The Environment Model calculates the impacts on roadside vegetation due to road salt. The Model for Road Management Costs calculates both the direct costs of the measures and the costs of damage to, and wear of, road surfacing, road markings etc as a result of road management measures.

Road users are concerned by ice and snow on roads and streets. The main problems are increased accident risk and impaired accessibility. To prevent - or at least decrease - the difficulties, road administrators perform various maintenance actions. The actions are advantageous for the road users, but involve costs for the road administrators and negative effects for the environment. To optimize maintenance efforts the use of management systems should be applied. The Winter Model project will result in a model for assessing the most important effects and their monetary value of alterations of winter maintenance strategies and operations in Sweden. The effects are assessed for road users, road administrators, and environment. For road users, the main effects concern accessibility (in terms of vehicle speed and flow) and safety. By using simultaneous monitoring of road surface conditions and traffic, the relationship between speed and different roadway conditions has been established. The speed reductions due to seven specified roadway conditions (moist, wet, ice or snow) relative to the speed at dry, bare conditions are generally significant. The reduction can be as great as 20 %. No relationship for traffic flow could be established. The average accident rate (accidents per million vehicle kilometers) during a winter season can be 16 times larger in black ice conditions than in dry road conditions. The accident rate in ice and snow conditions has an exponential relation to the duration of the condition, which means the shorter the duration, the higher the accident rate.

The relation between accessibility and different winter road conditions was

studied in the present project. The fundamental hypothesis was that drivers

adapt their behaviour to prevailing weather and road conditions in the

winter. The objective of this project was threefold:

Firstly, to record and develop methods for capturing and preparing field

data.

Secondly, field measures were to be carried out, and measurement data

processed and analysed. The intention was to include different types of

roads.

Thirdly, the hypothesis was tested.

Varje år drabbas trafikanterna av olägenheter på grund av is och snö på våra

vägar, gator och gång- och cykelbanor. Problemen är främst ökad olycksrisk

och försämrad framkomlighet. För att förebygga eller i varje fall minska

problemen, vidtar väghållarna driftåtgärder av olika slag, som till exempel

snöröjning och halkbekämpning. Åtgärderna är till nytta för trafikanterna,

men innebär kostnader för väghållarna och negativa effekter för miljön. Det

är därför angeläget att utveckla system för konsekvensbeskrivning av olika

driftstrategier och -metoder för att göra det möjligt att minimera de

samhällsekonomiska kostnaderna. För investeringsplanering av vägar och gator

har man sedan länge haft kalkylmodeller för beräkning av olika effekter:

förutom kostnader för planerings- och byggnadsprocesserna även förändringar

av trafikalstring, restider, olycksrisker och mycket annat. Drift- och

underhållssidan är i detta avseende mycket eftersatt, och det är naturligtvis

en svaghet, när det gäller att konkurrera om begränsade medel med ekonomiska

argument. Bristerna gäller inte minst vinterväghållningen

Ice and snow on roads and streets cause problems for road-users, requiring maintenance actions such as snow-ploughing, salting and sanding. Such actions influence road safety, accessibility, vehicle costs and the environment, involving heavy expenses for the road administrators. Consequently, it is important to develop effective strategies and methods for maintenance actions and to produce knowledge of the effects of different winter standards for road users and road administrators, as well as society in general. The aim of this project is to create a Winter Model (Winter Maintenance Management System, WMMS). The project began in 2001 and has financial support up to 2005. This report is a summary of the achievements in stage 1, during 2 001 to 2003. Stage 2 will be carried out in 2004 to 2005.

Road users are concerned by ice and snow on roads and streets. The main problems are increased accident risk and impaired accessibility. To prevent - or at least decrease - the difficulties, road administrators perform various maintenance actions. The actions are advantageous for the road users, but involve costs for the road administrators and negative effects for the environment. To optimize maintenance efforts the use of management systems should be applied. The Winter Model project will result in a model for assessing the most important effects and their monetary value of alterations of winter maintenance strategies and operations in Sweden. The effects are assessed for road users, road administrators, and environment. For the road users, the main effects comprise accessibility (in terms of vehicle speed and flow) and safety. By using simultaneous monitoring of road surface condition and traffic, the relationship between speed and different roadway conditions have been established. The speed reductions owing to 7 specified roadway conditions (moist, wet, ice or snow) relative to the speed at dry, bare conditions are generally significant. The reduction can be as large as 20 per cent. No relationship for traffic flow could be established. The average accident rate (accidents per million vehicle kilometers) during a winter season can be 16 times larger in black ice conditions than in dry road conditions. The accident rate in ice and snow conditions has an exponential relation to the duration of the condition, i.e. the shorter the duration, the higher the accident rate.

Syftet med denna förstudie var tvåfaldigt. För det första gällde det att undersöka om VTI:s körsimulator med sina begränsningar är lämpad för denna typ av studier. Om detta är fallet, kan de komma att bli ett hjälpmedel för Vägverket att bedöma de sammantagna effekterna av olika vägytetillstånd.

Det andra syftet var att beskriva i vilken utsträckning vägytans tillstånd påverkar förarnas körbeteende.

Förstudien begränsades till att omfatta studium av körbeteendet på spåriga vägar, där spåren låg i nivå med den omgivande vägytan, men där buller och vibrationer ökade påtagligt.

Simulatorstudien visar små men ofta signifikanta skillnader i såväl körbeteende som subjektivt bedömda effekter.

Ingen skillnad i genomsnittlig hastighet kan konstateras, däremot är hastighetsvariationen större vid spårig väg.

Beträffande sidoläge, så skiljer sig inte de genomsnittliga värdena i nämnvärd grad mellan spårig och slät vägyta. Man tycks försöka undvika att köra i spåren, på den jämna vägen körde man längre sträcka med ett sidoläge som motsvarar spårläget.

Varken för självrapporterad arbetsbelastning eller för självrapporterad ansträngning föreligger någon signifikant skillnad mellan spårig och jämn väg. Däremot visade sig skattningen av hur besvärlig man ansåg körningen vara ge signifikanta skillnader i bedömningen: spårig väg ansågs generellt vara besvärligare.

The aim with the pilot study was twofold. Firstly, to examine if the VTI driving simulator with its restraints is suited for this kind of studies. Secondly, to describe to what extent the condition of the road surface influence the driver behaviour.

The study was confined to studying driver behaviour on rutted roads, where the ruts were on a level with the surrounding road surface, but where noise and vibrations increased obviously.

The study shows small but often significant differences in driving behaviour as well as subjectively assessed effects.

No difference in mean speed can be established; however, the speed variance is greater for the rutted road.

The mean values of lateral position differ very little between rutted and smooth road; however, subjects tended to avoid driving in the ruts. On the smooth road they drove a long stretch with a lateral position corresponding to the position of the ruts.

Neither for self-reported work load nor for self-reported effort is there any significant difference between rutted or smooth road. The assessment of how difficult the driving task was, however, showed significant differences: rutted road was generally considered as more difficult.