In time limited overtaking restrictions for trucks on motorways : effect on the number of injury accidents. This report discusses the effect on traffic safety of overtaking restrictions for trucks on motorways. For this purpose, literature is reviewed and Flemish data are analyzed. The international literature does not allow a univocal conclusion. As an effect of the overtaking restriction for trucks, speed of trucks decreases, but speeds of the cars and of the traffic as a whole can increase or decrease. Also spacing can increase or decrease. Often homogeneity of the global traffic flow increases. But also the number of overtaking by private cars increases as does the speed variance between cars and trucks. To what extend these different effects counterbalance one another, is not evident. The effect of overtaking restrictions on the number differs between study sites: both an increase and a decrease of accident numbers have been reported. Only one study found one significant result, being a decrease of the number of accidents. A second purpose of this study was to calculate the effect of the restriction of overtaking by trucks on the number of accidents on Flemish motorways. The effect is calculated with the method presented by Hauer (1997). This method corrects both for the general accident trend on Flemish motorways and for regression to the mean. This correction is based on the use of a comparison group. The number of accidents of this comparison group was calculated using an accident risk model, that has been worked out especially for the present study. The model is based on all the injury accidents of the Flemish highways that happened between 6-10h and 16-19u. Since it is possible that the restriction of overtaking for trucks also influences the number of accidents in which no trucks were directly involved (e.g. by improving homogeneity of the global traffic flow) all injury accidents are considered in the analyses. Possible problems with the overdispersion factor are accounted for by a sensitivity analysis of the dispersion factor. The calculated results have a major disadvantage. The after period contains only one year. As a consequence, the number of accidents in the after period is limited. Moreover, chance influences the results more than could be hoped for, since an occasional extreme result is not been blunted by data of a second year. Hence, the results have to be interpreted with care. In general, we can conclude that the impact of the measure depends strongly on the location. On the link segments of only one location we find a statistically significant decrease of the number of accidents with 82%. On the entry zones of two locations a statistically significant decrease of the number of accidents with 78-82% is found. On the 42 other segments no significant results have been found. In the most extreme cases, the number of injury accidents is found to decrease with 85% or to increase with 175%. In most cases, the confidence interval of the estimate of the effect was that large, that increases nor decreases were statistically significant. Using meta-analysis the results of several locations can be taken together. Although none of the accident increases in itself was significant, the increases dominate in the meta-analysis. Meta-analysis of the link zones results in a statistically significant increase of the number of accidents with 36%. Meta-analysis of the entry zones gives a statistically significant increase of the number of accidents with 47%. For the exit zones, the dispersion factor could not be found during the modelling process. A sensitivity analysis of the dispersion factor resulted in accident increases ranging from 30% to 63%. The largest increases were statistically significant. It is concluded that the impact of the overtaking restriction for trucks on motorways depends strongly on the specific location. Apart from the restriction as such, other factors influence the impact on traffic safety. These factors could be infrastructural, but also not yet identified human or environmental factors. In a limited number of locations a positive result is found, but the meta-analysis does not at all support the idea that the measure increases traffic safety. Based on the present results, abolition of the measure is not recommended for those places where there was a positive effect. Further research should show under which conditions the measure has a positive impact on traffic safety. But the after period of the present study is short, only one year. This limits the generality of the results. Yet, from the present results and the review of the international literature, it can be concluded that overtaking restriction for trucks on motorways is a measure that should be handled with care. Depending on the location, the number of accidents can decrease or increase. And since in the meta-analysis the effect of the measure is negative, general introduction is not recommended. When sufficient data are available, the present study should be extended. New data can raise the number of segments with an overtaking restriction and teh after period of the presently uses segments can be extended. Such extension would also allow examination under which infrastructural conditions the overtaking restriction increases or reduces traffic safety. (Author/publisher) For the complete report see http://www.steunpuntverkeersveiligheid.be/en/node/281
Samenvatting