Study of variable speed limits highlights the trade-off of throughput and safety

By Sam Sklar
Varying a posted speed limit based on current conditions can improve operations and safety in congested work zones, a team from Missouri found.  This study from the University of Missouri-Columbus, called “Evaluation of Variable Advisory Speed Limits in Congested Work Zones,” explores this phenomenon in a congested work zone in St. Louis, Missouri. The findings? Results depend on whether DOTs are looking to prioritize safety over capacity.
The posted and legal speed limit on a road is rarely the same as the design speed or normal operating speed on a typical day, but these three “speeds” should be related. However, in practice and law, the speeds often are only loosely tied together, and can be far apart based on a number of variables. Although it is not legally enforceable, transportation agencies may also post an advisory speed limit for areas with a sharper-than-normal curve or other hazardous conditions. Some states also use advisory speed limits in poor visibility and/or weather conditions.
The three authors of this paper, Drs. Edara, Sun, and Hou, explore a variable advisory speed limit within a carefully defined corridor in St. Louis. The authors assert that by actively varying the advisory speed limits on a set of congested work-zone lanes, Missouri’s DOT could better manage traffic for the road network. Giving drivers an opportunity to clearly and safely maneuver through a potential area of conflict is one effective way to manage operations as opposed to costly enforcement or static advisory speed limits, which do not change to reflect current conditions.
The authors’ goal was to measure the effects of the flexible speed advisories on safety and operations. Within their parameters (a few miles on a single stretch of one highway in one city) they found the following:

  • Queue lengths dropped 39-53 percent
  • Throughput reduced 7-11 percent
  • Travel time increased 4-8 percent
  • Speed differentials decreased by 10 mph
  • Rear-end conflicts dropped 30 percent
  • Lane changing conflicts dropped 20 percent

The first three findings measure operations performance: the number of cars the road can move and how quickly the road can move them. The next three outcomes measure safety performance: are cars moving at a consistent speed, and can drivers then make safer, more measured decisions in a moment’s notice? Speed plays a role in all these measures, affecting both an individual vehicle and the entire system. The research illustrates the trade-off between operations and safety performance.
The numbers above tell a story, too. The managed corridor, almost by definition, slowed traffic by between 4 and 11 percent; a 30-minute trip became a 32- or 36-minute trip. For commuters on incredibly tight schedules this window becomes problematic, but for most commuters the addition of another song to their commute would most likely go unnoticed. However, transportation agencies are keenly aware that there are cumulative effects to even small delays and they try to avoid them if possible.
More striking is the effect on safety. A 30 percent drop in rear-end collisions is a significant result. More deliberate speed practices gave way to more confident and safer lane changes. It will be hard to tell if applying variable advisory speed limits to more chokepoints or conflicted intersections would continue to produce similar safety results or if additional management would only make a system more complicated, and eventually less safe. Scaling this activity will require more study and more testing.  This type of scenario is one where autonomous vehicles would be preferable over human drivers, varying their speeds automatically in response to conditions. Humans are notoriously impatient and unable to gauge appropriate speeds for conditions.
Using a variable advisory speed limit in areas of potential conflict is one part of a larger transportation strategy. It is up to DOTs to engage with stakeholders and colleagues to fully realize a roadway strategy that is best for everyone.
The full technical study can be found here.
Sam Sklar is a Program Associate at SSTI.