Modelling of percent time spent following using spatial measurement approach for two-lane highways

Abstract

Percent Time Spent Following (PTSF) is used as a key service measure for two-lane two-way highways, as recommended by the U. S. Highway Capacity Manual (U. S. HCM). PTSF is the average percent of total travel time that vehicles must travel in platoons behind slower vehicles due to inability to pass. Despite its acceptance as performance indicator, PTSF is difficult to measure directly in the field. Hence, it is estimated using a surrogate measure, i.e. based on the percent of vehicles travelling with headways less than 3 s as recommended by the U. S. HCM or use of a model derived from such estimates or a simulation approach. However, the applicability of the 3 s headways as the sole criterion in the surrogate measure is still debatable. By definition, PTSF is a spatial variable, whereas its estimation is based on fixed point, which may not be a representative of a segment‟s performance. Therefore, this approach fails to incorporate the travel time associated with PTSF. Hence, it is desirable to develop a spatial measurement approach for PTSF that takes into account of travel time and to derive a new PTSF model. This could substantiate the application of the 3 s surrogate measure. This study explores the application of test vehicle method for spatial estimation of PTSF and derives a model based on traffic flow and roadway geometric variables. Data for the study were sampled from twenty four (24) directional segments of two-lane two-way highways in Johor and Pahang States, Malaysia. A test vehicle equipped with a GPS-based speed acquisition and video recording system was used for the data sampling. PTSF was estimated as the average percent of total travel time spent by the test vehicle behind slower vehicles at headways less than 3 s. An empirical model to estimate PTSF for ranges of traffic flow and roadway geometric conditions with a reasonable accuracy was developed. A statistical analysis showed that there is no significant difference between PTSF obtained from the model derived in this study and those from the surrogate measure at α = 0.05. This finding supports the application of the surrogate measure based on the U. S. HCM which had been debated for long. It equally serves as a response to the contentious issue, which had not received the desired attention from experts hitherto. However, a comparison between PTSF from this study model and each of the Malaysian Highway Capacity Manual (MHCM) and the U. S. HCM models revealed that both MHCM and U. S. HCM models overestimate PTSF significantly at α = 0.05. This implies that PTSF based on the MHCM and U. S. HCM models would lead to erroneous designation of a road segment‟s level of service, which would in turn suggest for a premature facility improvement with attendant unjustified expenditures.