Variable advisory speed limit systems reduce variability in speeds and improve safety at congested work zones.

An investigation of work zone traffic control in Missouri.

Date Posted
04/29/2014
Identifier
2014-B00907
TwitterLinkedInFacebook

Evaluation of Variable Advisory Speed Limits in Work Zones

Summary Information

This study evaluated the mobility and safety impacts of using variable advisory speed limit (VASL) systems on interstate highways in Missouri. Case studies were conducted and simulation modeling performed to evaluate performance across a wide range of treatment scenarios. The first set of case studies focused on VASL in congested and uncongested urban work zones on I-270 near St. Louis. The second set examined VASL in rural work zones on US-54 and US-63. Simulation modeling was also used to expand the range of treatment scenarios evaluated.

Urban Study

In 2010, permanent VASL dynamic message signs (DMS) were installed on I-270 to control traffic and adjust speed limits to accommodate prevailing traffic conditions. In June 2012, researchers used traffic data collected at or downstream of static speed limit signs over a three day period to assess system impacts in areas with and without congestion. At sites with congestion, the system did not operate continuously, but operated periodically to encourage drivers to reduce speeds gradually as they approached work zone bottlenecks. At sites without congestion, the VASL system operated continuously to detect average speeds and remind motorists of safe travel speeds.

Rural Study

To determine the impacts of VASL on rural work zones data were collected at two locations in central Missouri. The first site on US-54 located south of Jefferson City was monitored on September 21, 2011. The second site on US-63 located south of Columbia was monitored on March 13, 2012. Since traffic conditions did not warrant lower advisory speed limits in these areas, researchers evaluated the effectiveness of the system to display static speed limits and reinforce posted speed limits in control areas. These work zones generally experienced less average daily traffic, fewer lanes, less traffic, and more trucks compared to the urban work zones.

Simulation

Two traffic simulation models were constructed to represent congested work zone conditions with and without VASL.

FINDINGS

Urban Case Studies
  • At uncongested sites on I-270, the average speeds with VASL were lower by 2.2 mi/hr, than without VASL.
  • The standard deviation of speeds with VASL was higher, by 4.4 mi/hr, than without VASL.
The increase in standard deviation is possibly due to the advisory nature of VASL. Since they are not enforceable some drivers complied while others did not. Compliance rates inside the work zone were not high with or without VASL. Still, the compliance with VASL was about eight times greater than without VASL. At congested sites on I-270 work zones where bottlenecks were expected, field studies found that the VASL were effective in making drivers slow down as they approached the work zone, thus reducing sudden changes in speeds.

Rural Case Studies
  • The uncongested traffic conditions at rural work zones on US-54 and US-63 did not warrant varying the speed limits. The VASL system complemented existing static speed limits and displayed the work zone speed limit.
  • Both sites showed reductions in mean speed, variance, and 85th percentile speed downstream of the VASL sign indicating that drivers paid attention to the advisory speeds.
  • At the US-54 site, the speed reduction from the VASL sign to the taper was gradual when VASL was deployed.
Simulation

The traffic simulation model produced mixed results with both positive and negative mobility and safety benefits. By varying the compliance rate and truck percentages, 10 different evaluation scenarios were generated. The use of VASL resulted in:
  • 40 to 58 percent decrease in average queue length
  • 6 to 13 percent reduction in throughput
  • 20 to 29 percent decrease in number of stops
  • 1.5 to 10 percent increase in travel times.
The effects on the predicted number of rear-end and lane-changing conflicts, however, varied depending on the proportion of trucks in the traffic stream.
  • With 10 percent trucks in the traffic stream the number of conflicts increased, but with 15 percent trucks the number of conflicts decreased.
The varied results prompted researchers to explore new ways to improve the VASL algorithm. The 5-minute (P5) algorithm was developed to account for varied congestion levels. Modeling results using this algorithm suggested that VASL could improve throughput by 11.5 percent, improve travel times by 1.5 percent, and reduce rear-end conflicts by 20 to 31 percent for traffic with 15 to 10 percent trucks, respectively. Similarly, lane change conflicts would be expected to decrease.
Goal Areas
Deployment Locations