Examination of core concepts in the Highway capacity manual

49080563Final contract report.January 2002.PDFTech Reporthttp://ntl.bts.gov/lib/37000/37000/37080/02-CR4.pdfVTRC 02-CR3VTRC 02-CR4FHWA/VTRC 02-CR3FHWA/VTRC 02-CR4VirginiaVirginia Transportation Research CouncilSmith, Brian L.,Virginia Transportation Research CouncilVirginia. Dept. of TransportationUnited States. Federal Highway AdministrationUS Transportation CollectionThe Highway Capacity Manual (HCM) is one of the most widely used traffic engineering guidance documents in the world. It was originally published in 1950, and has been under constant revision since. Unfortunately, due to past cost and time constraints associated with traffic data collection, much of information in the manual is based on research conducted using relatively small data sets. This calls into question the statistical significance of some of the manual's material. The Virginia Smart Travel Laboratory is a nationally unique research facility. The distinguishing feature of the laboratory is its direct connection to operational VDOT transportation management systems. This gives the laboratory access to unprecedented quantities of traffic data. The purpose of this research project is to use this data to investigate a key concept of the HCM: freeway traffic lane distribution. An important consideration of transportation management is the distribution of lane use by vehicles. This distribution plays a significant role in how traffic management devices, such as variable message signs, lane control signals, and ramp meters are utilized. Unfortunately, according to the HCM, "when two or more lanes are available for traffic in a single direction, the distribution in lane use varies widely ... there are not "typical" lane distributions." An investigation of this concept using a large set of data from freeways in the urbanized Hampton Roads region of Virginia led to the following conclusions: The distribution of vehicles along a specific link of a freeway system does tend to follow predictable trends by time-of-day. A missing data estimation procedure can be developed that exploits the consistency of lane distribution by time-of-day and location. This estimation methodology proved to accurately estimate missing detector data, generally producing results within the 6%-8% error range. Finally, the report presents the following recommendations to VDOT. VDOT should collect and archive traffic data at the lane level to support future applications, such as the missing data estimation methodology. VDOT should use the lane distribution-based missing data estimation methodology described in this report in Smart Traffic Centers and permanent count stations located on freeways. VDOT should formally transmit this report to TRB for committee consideration as the next version of the HCM is developed