Investigation of Some Unresolved Issues in Freeway Capacity

Abstract

<p>Three issues related to the capacity of uninterrupted flow facilities are addressed using 52 days of detector data from one freeway, the Queen Elizabeth Way (QEW) in Mississauga, west of Toronto. The first issue is that of a possible capacity drop in the bottleneck after the queue forms upstream. The results show that there is a drop, of 98 passenger cars per hour per lane (pcphpl).</p> <p>The second issue tackled is the distributional characteristics of queue discharge flow. Daily average morning peak flow rates and the peak 15-minute data were examined. The frequency histogram is not close to a Normal one. The mean and the median, contrarily, are very close to each other. The use of the mean value was therefore deemed appropriate. The distribution of speed was also looked at. At a distance of 1.5 kilometres from the bottleneck, the observed average travel speed of vehicles discharging from an upstream queue was 74 km/hour.</p> <p>The third issue focuses on the definition of capacity, conceptually and numerically. Thirty-second, 5-minute and peak 15-minute flow data were used. Capacity flows were observed under two different conditions, free-flow and congested, or forced-flow. The findings that evolved from the data analysis indicate that if capacity refers to the highest flows sustained for at least 15 minutes, as suggested in the Highway Capacity Manual (HCM), then it is conceptually sound to define capacity as pre-queue high flow. The practical implications of the capacity concept, however, make it very useful and equally valid to define capacity as the maximum queue discharge flow.</p> <p>Using the mean values of 30-second flow data, the pre-queue capacity is 2,306 pcphpl which is rounded off to 2,300 pcphpl; for queue discharge flow, it is 2,200 pcphpl. (These values were weighted by the duration whereas the capacity drop value above was not).</p>