Long green times and cycles at congested traffic signals

Field data were collected and simulation experiments based on traffic at an intersection in Virginia were conducted to test the hypothesis that headways increase with long green times and to test the common assumption that throughput increases with longer cycles. The results showed that headways increased with long green times as a result of departing turning vehicles and that this effect could cause a significant increase in overall average approach headways. The results also showed that maximum throughput, defined as the point where additional offered load could not be served, did not increase with longer cycles. With values derived from the field data, increasing the cycle did not increase throughput. In simulation, increasing the cycle caused a reduction in throughput as a result of increasing the effect of departing turning traffic on the average headway.