Ex-Post Analysis on Effect of Policy Measures on Travel Time Reliability

Globalisation of economy and changes in personal lifestyle have fuelled the importance of reliability in transport network and services. All travellers have experienced congestion and delays. Sometimes it can have consequences, for example missing an appointment or arriving late. Congestion occurs when demand exceeds capacity. This often happens in peak periods, and is often expected. However this can also happen in case of unexpected disruptions, for example in adverse weather conditions or when accidents occur. In order to eliminate disruptions to the highway network when congestion levels arise, it is sometimes intervened with implementation of transport and infrastructure projects and policy measures. The social and economic benefits of these measures are usually only related to reductions in average travel times in cost benefits analysis (CBA). However an important element could be missing from these estimates, as the total net benefits (or costs) could be incorrect when not addressing the reliability of travel times as well. Today, travel time reliability is accounted for implicitly in Dutch CBAs by adding a 25% surcharge on travel time gains. This method is questionable. It is expected that road traffic in the Netherlands will grow in coming years. To be able to manage impending development, it will be needed to intervene with extra measures. As one of the mobility policy goals for the year 2020 is to increase reliability of the road network, determining effects of policy measures on travel time reliability becomes a very important topic. The main scope of this research is to determine effects of policy measures on travel time reliability. First of all, a definition and method to measure the reliability was needed in order to conduct the analysis. As a means to do so, available literature on the matter was reviewed. There it became clear that there exist many indicators to measure reliability of travel times. Reliable travel times are considered to be travel times that show little variation and can be well predicted. Travel time reliability is though a relative concept as what is considered an acceptable travel time can be different from person to person. Indicators for travel time reliability are usually related to characteristics of the day-to-day travel time distribution. The indicators are different from one another and capture different parameters of the distributions. This is why it was decided to test a range of indicators for the analysis. The data behind the analysis consisted of route travel times of the Dutch highway network, from the period 2001 to 2009. This data enabled an expost study of the effects of different policy measures, implemented in the network, on reliability. The methodology chosen for the anlysis was before and after studies of specific implementations of policy measures. The reason for choosing that methodology was because the network is complex, and the routes are different from one another, making the effects of the policy measure dependent on location and prevailing traffic conditions. Five different implementations of policy measures were chosen for the analysis, they were; \u95 Rush hour lane on A13 \u95 Permanent lane on A4 \u95 Enforced speed restriction on A10 \u95 Dynamic route information panel on A12 \u95 Ramp meters on A10 The chosen indicators for reliability were computed for periods before and after implementations of these measures. Time-space plots of vehicle speeds on the locations of these measures were also plotted. This gave a better understanding of prevailing traffic conditions. The results of the research showed that there is no clear-cut relationship between a certain policy measure and travel time reliability. It is all dependent on location and circumstance of each measure. The indicators for reliability also proved to be inconsistent with each other. The indicators tell different stories. The choice of indicator should be reflected in what definition is put into reliability, and what the expected impacts of the measure analysed are. The results also questioned the legitimacy of accounting for reliability in CBA by means of a 25% surcharge on travel time gains, as improvements were seen in mean travel time for some cases but not in the reliability indicators, and vice versa. In order to to compute the economic impacts of reliability, a value of reliability is needed. As of yet, no generally accepted values exist that are used in official national CBA. What this value should be is beyond the scope of this research.Transport & PlanningCivil Engineering and Geoscience