The 1985 Highway Capacity Manual (HCM) contains a procedure for estimating stopped delays at signalised intersections. These estimates are to be used to assess the levels of service at an intersection. The assessed levels of service, in turn, provide a basis for making decisions concerning geometric designs and signal operations. To facilitate sound decision making, the HCM procedure must be able to produce accurate delay estimates. This paper evaluates the reliability of the HCM procedure, based on field data, and discusses needed modifications. The evaluation reveals that the procedure tends to overestimate stopped delays at reasonably well-timed signal operations. The discrepancies between the HCM estimates and the observed delays can be very large even when correct cycle lengths and green durations are used as inputs. Such large discrepancies are attributable in part to the progression adjustments recommended in the HCM procedure. Given actual cycle lengths and green durations, the procedure's ability to correctly identify the levels of service is good. However, a reliable method is needed for estimating average cycle lengths and green durations for traffic-actuated signal operations.
Abstract