Work zones as a series of bottlenecks

Studies on queue length and delay estimation have simplified the work zone as the location of a single bottleneck. However, if more than one bottleneck exists, such a simplification introduces considerable error into estimations of queue length and delay and can adversely affect traffic management strategies. The work space and the transition area are the two main locations in which bottlenecks are likely to occur. Drivers who pass a work space may reduce speeds or increase their headway and thus reduce capacity. Lane shifts or merges may also reduce capacity and cause queues. Work zones with inhomogeneous sections often have more than one bottleneck in a single work zone. The objective of this study was to develop a more accurate approach to predict queue length and delay when two bottleneck locations are assumed to exist in a work zone. Analyses were conducted for two-to-one work zones in which one of the two lanes was closed in a given direction. Shockwave theory was applied to study six conditions, each of which represented a relationship between the demand and capacity of the bottlenecks. Because of differing initial conditions and the interaction between the bottlenecks, 21 queue propagation patterns existed. Results from this approach were validated with field data. The average of the queue lengths did not differ significantly from the average of the observed queue lengths. The maximum queue length was overestimated by roughly 4%. This percentage was lower than that reported for queue length (above 20%) in estimates made with popular software packages, such as QUEWZ and QuickZone.