Exploring the Upper Capacity Limits on a Single-Lane Roundabout

Development Type:        Existing Music Festival
Services Provided:          Roundabout Operations Review
Client:                                 Minnesota Department of Transportation (MnDOT)

What is the capacity of a single-lane roundabout? This question is becoming more urgent to answer as we to learn about modern roundabouts. Multi-lane and dual-lane roundabouts appear to be distinct from single-lane ones. With a single lane, drivers intuitively appear to understand the circular intersection better. And the driving decision is easier – go if it’s clear and stop if it’s not. There’s no ambiguity where a motorist misunderstands where the car on the inside or outside lane of a multi-lane roundabout is going and makes a poor decision that results in a crash. That’s not to say multi-lane or dual-lane roundabouts aren’t safe, it’s just single-lane roundabouts are safer at this time.

Knowing the capacity of a single-lane roundabout is therefore very important. Where the volumes are less than the capacity, intersections can then operate with a single-lane roundabout for as long as possible.

Spack Consulting, through its sister company Traffic Data Inc (TDI), worked with MnDOT to devise a plan to measure roundabout capacity. In this case, WE Fest provided the opportunity to examine two single-lane roundabouts (Highway 59 intersections with Willow Street/Long Lake Road and with Becker County State Aid Highway 22) experiencing higher than normal volumes. This annual event located in Detroit Lakes, Minnesota attracts over 150,000 people over the three-day country music festival. Having lots of drivers in a short timeframe suggested we might hit the roundabout capacity and we wanted to be there to capture it.

Using our own camera technology, we set up COUNTcam traffic video recorders at multiple locations. One camera captured the intersection itself to provide the turning movement counts. One or two more cameras were pointed down each approach to each capture vehicle stacking. We could see queues to approximately 600 feet (24 passenger cars) using the video recorders. In all, 15 video recorders were needed to capture the volumes and stacking at the two intersections.

As with any special event, people arrive early and leave late. To give us the best opportunity for capturing the roundabout traffic at capacity, we, therefore, used the cameras to record operations for six days. After initial set-up, our field personnel completed a mid-week check to ensure everything was still working as expected. To our relief, no vandalism, unusual weather (high winds, etc.), or other special circumstances disrupted things. After pick-ups, we had 15 144-hour-long videos to review.

The analysis included several different components. After counting the volumes over the study time period, we compared the event volumes to the typical volumes through the roundabout intersections. This work included reviewing daily volumes, peak hour volumes, and the percentage of turning movements on each approach. After all, the percentage of left turning traffic makes a difference in capacity at any type of intersection.

The vehicle queues were observed on each approach to each roundabout. The goal was to find the worst, or longest, queue of vehicles for each approach. The stacking exceeded our video recorder’s view on only one approach and for only a four-minute period out of the 144 hours. In addition to the number of vehicles or length of vehicles in the queue, we determined the time needed to go from the back of the queue into the roundabout. We also compared the data to the Intersection Control Evaluation (ICE) reports for each intersection – a sort of how did we (or our colleagues at another firm in this case) do in forecasting volumes and operations.

We attempted to find a correlation between the vehicle queues and the intersection volumes. Using the approach volumes as well as the combined approach plus conflicting volumes, the thought was to see if an easy equation could be determined for future use of roundabout analysis.

Lastly, we completed determinations of volume-to-capacity (v/c) ratios and operational analyses using the Highway Capacity Manual’s (HCM) 6th Edition methodology. Essential, this work allowed us to compare calculated results against observed results.

So, what did we find?

The results were decidedly mixed. We determined the roundabout traffic did not reach the capacity of the intersections even though the event traffic was 60 percent higher than non-event days. Other important conclusions included:

  • Queue Length. Although the maximum queue was relatively long on multiple approaches, the stacking cleared quickly with comparable delay times to that of a traffic signal with a 120-second or longer cycle length.
  • ICE Report. The ICE report did a relatively good job in predicting future acceptable operations.
  • Hourly Approach Volume. The hourly approach volume graphed against the observed maximum queue showed the highest correlation.
  • HCM Predictions. The HCM predicted vehicle queues were less than observed, significantly so in some cases.

While the study did not observe roundabouts at capacity, it did confirm the intersections are operating

acceptably during appreciably higher volume event traffic with queued motorists moving rapidly into the roundabout. The HCM calculations provide a good initial review of single-lane roundabout operations in terms of a simple pass-fail grade. This evaluation tool is important, but more detailed simulations or other analyses may be necessary to review any impacts of vehicle queues.

As with how many research projects seemingly end, further research could be completed on this data set. Modeling simulations could be developed to test the maximum and 95th percentile queues for comparison to the observed and HCM queues. A higher level statistical analysis may also produce a better correlation between vehicle queues and approach volume or another metric. And the observed delays could be calculated for comparison to the HCM or other simulation results.

With so much data, there are seemingly endless ways to spin the numbers and find more interesting nuggets about roundabout operations in Minnesota. Want to read more? Check out our final report completed in cooperation with MnDOT here.