Determining Peak Hour Factors for Capacity Analysis

Considering the peaking effect over a short time period is important because queuing might build up and take substantial time to discharge. Traditionally, the peak-hour factor has been used to quantify such a peaking effect. The Highway Capacity Manual (HCM) suggests a design value of 0.92 for congested urban areas and 0.88 for rural areas, if no field measurements are available. These default values give general guidelines, but they might be too coarse for practical usage. An effort is made to model the actual peak-hour factors as a function of volume-to-capacity ratio and the functional classification of roadways. A total of 1,669 data points were obtained for analysis. The results show that, among several functional forms, the simple power function established with functional classification of roadways can be used to explain 46% of data variation, which appears to be acceptable, given the significance of data variability. The 95th percentile confidence intervals on the mean estimates and the predictive limits are also provided. Compared to the HCM default value, the recommended peak-hour factors, in general, result in higher average intersection delays with the optimal signal control. Finally, model validation using data collected from two other geographical areas indicates the proposed prediction model is transferable.