Abstract
Saturation flow is traditionally treated as a stabilized maximum rate at which queuing vehicles can be discharged from a lane after the signal changes to green. Traffic engineers in many countries have been using this parameter to estimate the capacities of traffic lanes at signalized intersections. The estimated capacities in turn are used to estimate other performance measures and to classify levels of service. This traditional method of analysis is an efficient way for capacity estimation if the queue discharge rate rises quickly to a steady maximum after onset of the green signal. Actual queue discharge patterns, however, often do not display an easily identifiable steady maximum rate. This characteristic renders the use of saturation flow troublesome and prone to errors that can lead to poor planning and operational decisions. The pitfalls in the current use of saturation flow are examined, and ways of alleviating the problems are suggested.