USA Banner

Official US Government Icon

Official websites use .gov
A .gov website belongs to an official government organization in the United States.

Secure Site Icon

Secure .gov websites use HTTPS
A lock ( ) or https:// means you’ve safely connected to the .gov website. Share sensitive information only on official, secure websites.

U.S. Department of Transportation U.S. Department of Transportation Icon United States Department of Transportation United States Department of Transportation

Public Roads - November/December 2012

November/December 2012
Issue No:
Vol. 76 No. 3
Publication Number:
Table of Contents

They're Small But Powerful

by Wei Zhang, Joe Bared, and Ramanujan Jagannathan

An FHWA study offers recommendations for constructing mini-roundabouts to reduce congestion and improve safety at intersections throughout the United States.

This mini-roundabout in Maryland contains a painted splitter island, a common feature of the mini-roundabouts constructed in this State. An FHWA study looked at the performance and safety features of mini-roundabouts like this one.

At many intersections in the United States, especially on roads that experience moderate congestion, traditional stop controls work well. This situation is changing, however, in many urban and suburban areas, where growing urbanization has resulted in traffic congestion and safety problems.

One potential solution might be the mini-roundabout. Other countries have been designing and constructing mini-roundabouts since the 1970s, but very few of these intersections are found in the United States. As used here, a mini-roundabout refers to a single lane with an inscribed circular diameter between 50 and 90 feet (15 and 27 meters). Its defining feature is traversable central and splitter islands to accommodate large vehicles.

"Traffic engineers and community planners are starting to recognize the mini-roundabout design as a low-cost solution for improving intersection capacity and safety without the need for acquiring additional right-of-way," says Monique Evans, director of the Office of Safety Research and Development (R&D) at the Federal Highway Administration (FHWA).

To assess the viability of mini-roundabouts for use in the United States, in 2009 the FHWA Office of Safety R&D initiated a study -- Field Testing, Marketing, and Crash Analyses of Mini-Roundabouts -- to evaluate the operational and safety performance of these intersections. A review of design guides for mini-roundabouts from European countries, followed by visits to existing small roundabouts in the United States, provided insights into how the large roundabout's kid brother could relieve congestion and enhance safety on U.S. roads. Most of the mini-roundabouts in the United States lack a fully traversable feature, thereby limiting their capability to handle large vehicles longer than 25 feet (7.6 meters).

Mini-Roundabout Design Considerations

The FHWA research team kicked off the study with a review of Germany's design guide for mini-roundabouts, which calls for the following geometric design criteria:

  1. Inscribed diameter between 42.6 and 78.7 feet (13 and 24 meters)
  2. Circular roadway width between 14.8 and 19.7 feet (4.5 and 6 meters)
  3. Central island maximum height of 4.7 inches (12 centimeters)
  4. Central island minimum curb height of 1.6 or 1.9 inches (4 or 5 centimeters)

The German highway code recommends mini-roundabouts for urban areas where the speed limit is 31 miles per hour, mi/h (50 kilometers per hour, km/h) or less. Evaluation revealed an average reduction of 30 percent in the crash rate -- from 0.79 to 0.56 crashes per million vehicles -- after conversions of intersections into mini-roundabouts.

Because of a lack of field evaluation results and formal guidelines in the United States, implementations of U.S. mini-roundabouts vary in design features. Some function as traffic-calming circles in residential areas, but many are not designed and constructed to handle heavy traffic demand.

Commonly asked questions about the suitability of a mini-roundabout include the following: How much traffic can a mini-roundabout handle? Will schoolbuses have difficulty navigating through mini-roundabouts? What are the impacts of onstreet parking curbside on approach roads? How should the safety of pedestrians and bicyclists be addressed?

The intent of the FHWA study is to address these and other relevant questions objectively, assess the pros and cons of existing designs, and develop a set of guidelines that will achieve design objectives effectively. Among the general design objectives is ensuring that vehicles up to the size of a single-unit truck can circulate around the central island while at the same time providing a way for larger vehicles to traverse the central and splitter islands. Other goals include reducing the number of intersection conflict points, increasing intersection capacity, and improving safety for all modes of traffic.

The fundamental difference between mini-roundabouts and all other types of modern roundabouts is reflected in the design concept. For mini-roundabouts, the roundabout elements such as approach lane, entrance, central island, circulating lane, splitter islands, and guidance signs are designed and placed in such a way to enable 97 percent of vehicles such as cars, vans, SUVs, and small trucks to pass through by staying in the circulatory roadway. The traversable central island and splitter islands are left for very large vehicles, such as full-size schoolbuses, to drive through. In implementation, this difference in design concept can result in a significant difference in cost.

Measurements and Peak-Hour Traffic Demands at Five U.S. Mini-Roundabouts

Location Central Island Splitter Island Peak-Hour All-Entering Volume(VPH)
Diameter (ft) Height (in) Length (ft) Height (in)
Location 1 24 6 205 6 241
Location 2 27 6 205 6 243
Location 3 14 0/6* 70 4 467
Location 4 36 0 64/152** 0 N/A
Location 5 17 0 26 6 N/A
Note: Intersection locations were intentionally made anonymous.
*Vertical height above adjacent ground of the central island's edge and crown at this location.
**This location has two splitter islands with two different lengths.
Source: FHWA.

Survey of Current U.S. Implementation

The FHWA research team visited six small roundabouts in five States. Five locations are in Delaware, Maryland, Texas, and Washington State. The sixth mini-roundabout studied by the team is not described because it is on a college campus and serves mainly bicycle traffic. The researchers also studied photos of mini-roundabouts from other cities.

This survey showed that mini-roundabouts in the United States are constructed at junctions where the peak-hour traffic demand is less than 500 vehicles per hour. Costs of the mini-roundabouts ranged from approximately $20,000 to $250,000 per intersection. The objectives for constructing mini-roundabouts varied considerably for these installations but included the following:

  • Improve access for pedestrians and bicyclists.
  • Encourage more school-aged children to walk to school.
  • Reduce the occurrence of speeding and running STOP signs.
  • Reduce the noise caused by vehicles braking hard and accelerating at intersections.
  • Preserve the values of the neighboring properties.
  • Increase intersection capacity.
  • Improve intersection safety.

At the mini-roundabout in location 2 in Texas, the researchers observed that a splitter island curb looks difficult to traverse due to the height of the curb, as does the central island. At some locations, such as location 5, the mini-roundabout has KEEP RIGHT signs mounted at the edge of the central island, preventing large trucks from traversing it. At other locations, mini-roundabouts have tall trees planted at the center of the central islands. These practices can make the central and splitter islands difficult to traverse and might create traps for large vehicles. A photograph taken by one of the FHWA researchers shows a long truck making a right turn at a mini-roundabout in location 4 in Maryland and illustrates the need for traversable splitter and central islands.

The researchers believed this splitter island at the mini-roundabout at location 2 in Texas might be difficult to traverse because of the high curbs. Unless splitter islands are traversable, mini-roundabouts can become traps for large vehicles.

Painted flush splitter islands are common in some regions of the United States. Flush splitter islands are fully traversable, but less effective in enforcing the design intent. For example, at some mini-roundabout locations, the researchers observed several drivers of passenger cars illegally crossing the flush splitter island to make left turns, rather than circulating through the mini-roundabout to make the turn. At the mini-roundabout in location 4 in Maryland, when it was first completed, vehicles were seen cutting through the flush central island and creating unsafe traffic conditions. The State's department of transportation (DOT) mitigated the problem by installing flexible posts (a quick and low-cost solution) on the central island.

Some mini-roundabouts were built as low-cost streetscape projects. As a result, these mini-roundabouts included the following design elements that resulted in undesirable field performance:

  • Stop bars corresponding to the former STOP signs remained in place after construction, along with the yield signs at the entrances to the mini-roundabouts. The FHWA team observed some drivers treating the intersection as stop controlled rather than yield controlled.
  • Intersection markings for pedestrian crossings were not removed and relocated at least 20 feet (6 meters) before the yield lines, creating conditions conducive to vehicle-pedestrian conflicts inside the intersection. According to the National Cooperative Highway Research Program's (NCHRP) Report 672, Roundabouts: An Informational Guide, pedestrian crossings should be placed inside the approach lane instead of at the intersection to reduce potential vehicle-pedestrian conflicts.
  • Central islands were too small and the circulating lane too wide, coupled with flush-painted splitter islands, making it too easy for drivers to make a left turn by crossing the splitter island rather than circulating the central island.
  • Curbside, onstreet parking allowed all the way to the intersection was coupled with flush-painted splitter islands, tempting drivers to make a U-turn before the intersection when spotting an empty parking space, rather than circulating around the central island to make the U-turn.

The above observations showed that central and splitter islands need to be traversable to accommodate large vehicles, but they should be raised, or delineated by flexible posts, rather than being flush only. The islands need to create physical channelization and discourage small vehicles from mounting them.

Intersection Traffic-Handling Capacity

The research team recorded field videos of traffic passing through the mini-roundabout at location 4 in Maryland during the afternoon peak-traffic period on a typical weekday. The researchers also manually logged 5-minute traffic counts at that location. The peak-hour traffic demand was 924 vehicles per hour (VPH). The corresponding traffic videos showed no signs of a queue forming.

This mini-roundabout in location 5 contains traffic signs in the central island, making it difficult for large vehicles to traverse the island. Trucks over two axles are prohibited on this roadway.

The researchers derived two potential maximum peak-hour traffic capacities in the following manner. First, they multiplied the peak 5-minute traffic counts from 4:15 p.m. to 4:20 p.m. by 12, giving a potential capacity of 1,140 VPH. From the traffic scenes in the videos, the level of service corresponding to this level of traffic demand would be an "A," meaning the intersection delay is 15 seconds or less. Next, the researchers multiplied the 5-minute peak counts on all of the individual approaches by 12, giving a potential capacity of 1,368 VPH.

The researchers derived the capacity under the first scenario by assuming that the traffic pattern from 4:15 p.m. to 4:20 p.m. would repeat itself for the entire hour. The traffic scenes corresponding to the first scenario are visible in the videos, which showed no queuing.

For the second scenario, the team derived the capacity by assuming the 5-minute peak traffic from each approach would happen simultaneously, and that period would repeat itself for the entire hour. The traffic conditions of the second scenario can be created by adding 0, 12, and 7 more vehicles to leg 1, leg 2, and leg 3, respectively, to the traffic demand of the first scenario, which is the equivalent of adding 0, 2.4, and 1.4 more vehicles per minute to leg 1, leg 2, and leg 3. Judging from the available gaps observed in the videos, this mini-roundabout would have no problem handling these additional traffic demands.

At some evaluation sites where the researchers collected data before construction, stop-controlled intersections experienced recurring congestion when the traffic entering from all directions exceeded 900 VPH.

A member of the FHWA research team took this photograph of a long truck trying to make a right turn at the mini-roundabout at location 4.

From the videos of the mini-roundabouts, the FHWA team estimated the parameters of driver behavior, such as gap and followup headway (the spacing in seconds between the leading vehicle's rear bumper and the following vehicle's front bumper) that most motorists would accept when entering the mini-roundabout. The team then conducted numerous simulation analyses by saturating one entrance with simulated vehicles and gradually increasing the circulating traffic flow to assess the capacity. These simulation studies indicated that a mini-roundabout can carry up to 1,000 VPH (circulating traffic plus entering traffic) per approach. In June 2012, at the Transportation Research Board's Urban Street Symposium, T. Lochrane presented these results in a separate paper, "Traffic Capacity Models for Mini-Roundabouts in the United States: Calibration of Driver Performance in Simulation," coauthored by N. Kronprasert, J. Bared, D. Dailey, and W. Zhang.

The research team conducted capacity analyses of several potential sites using the critical lane-volume and simulation methods and concluded that mini-roundabouts can conservatively handle 1,600 VPH (the sum of entering traffic demand for all approaches), while providing an adequate level of service.

Recommended Design Elements

The review of German design guides and the results of the U.S. field survey suggest that mini-roundabouts are a viable design option for high-traffic intersections of collector roads when four conditions are met. The proposed mini-roundabout is an intersection of two- or three-lane collector roads, the posted speed limit is 35 mi/h (56 km/h) or less, the traffic demand from the major and minor approaches is comparable, and the roads have a low percentage of single-unit trucks or larger vehicles (5 percent or less).

Traffic Counts at Location 4 Mini-Roundabout

5-Minute Volumes

P.M. Leg 1 Leg 2 Leg 3 Total
3:45 0 0 1 1
3:50 15 13 16 44
3:55 9 29 19 57
4:00 29 38 14 81
4:05 24 (41) 30 95
4:10 19 29 15 63
4:15 (39) 29 27 95
4:20 26 28 (34) 88
4:25 18 27 15 60
4:30 35 25 17 77
4:35 25 29 12 66
4:40 27 27 24 78
4:45 23 27 22 72
4:50 24 25 25 74
4:55 21 25 14 60
5:00 28 33 29 90
5:05 34 27 24 85
5:10 29 25 25 79
5:15 14 20 34 68
5:20 12 23 23 58
5:25 18 18 26 62
5:30 22 26 18 66
5:35 30 16 13 59


Actual Hourly Volumes

P.M. Leg 1 Leg 2 Leg 3 Total
3:45 to 4:45 266 315 224 805
3:50 to 4:50 289 342 245 876
3:55 to 4:55 298 354 254 906
4:00 to 5:00 310 350 249 909
4:05 to 5:05 309 345 264 918
4:10 to 5:10 319 331 258 908
4:15 to 5:15 329 327 268 924
4:20 to 5:20 304 318 275 897
4:25 to 5:25 290 313 264 867
4:30 to 5:30 290 304 275 869
4:35 to 5:35 277 305 276 858
Potential Maximum Volume_1:  
  468 348 324 1,140
Potential Maximum Volume_2:  
  468 492 408 1,368
This table shows the 5-minute and peak-hour traffic counts, plus two forms of potential peak-hour traffic counts.
Source: FHWA.

Although the FHWA simulation analysis of mini-roundabouts indicates a potential capacity of up to 1,000 VPH per entrance, actual field capacity also depends on the presence of other modes of traffic, such as pedestrians and bicyclists.

The design of the central and splitter islands should fulfill the following requirements:

  • Be fully traversable by large vehicles.
  • Discourage drivers of small vehicles from trying to traverse the central island.
  • Design the curb height to avoid causing difficulties for snowplowing operations (in States where it snows).
  • Prohibit curbside, onstreet parking 100 feet (30 meters) from the intersection or from the starting point of the splitter island, whichever is greater.

Properly raised central and splitter islands are preferred to help ensure driver compliance with the design intent. The central island could be painted yellow to make it more conspicuous. Flush central islands are acceptable but might have to be combined with retroreflective flexible posts to realize the design intent.

The FHWA research team developed three mini-roundabout design templates using a WD-50 truck (50-foot, five-axle tractor cab and trailer) as the design vehicle and assuming a typical two-lane urban street in the United States. These streets normally have approach widths of 24 feet (7 meters). The approach widths of three-lane streets are 36 feet (11 meters) with corner radii of 30 feet (9 meters). The inscribed circular diameters of all three mini-roundabout templates would fit entirely within existing intersection boundaries.

The team used the following control parameters in producing the design templates:

  • Circulating drive lane width -- between 14 and 16 feet (4 and 5 meters) wide
  • Mini-roundabout entrance lane -- 10 to 11 feet (3 to 3.4 meters) wide
  • Pedestrian crossing -- 10 feet (3 meters) wide and placement at least 20 feet (6 meters) before the yield line
  • Splitter island -- 4 feet (1.2 meters) minimum width

For the design of a mini-roundabout, the research team suggests making the central island as large as possible after achieving the desired width for the circulating lane. Also, the central island should be raised to a maximum height of 5 inches (13 centimeters) with a slope of 1:15 (vertical: horizontal) to allow water to drain away from the central island.


The field evaluations conducted to date yielded several observations. First, the objectives for implementing mini-roundabouts vary greatly, depending on a project's location and traffic, safety, or other issues of concern. Second, the cost of mini-roundabout design and construction also varies greatly, depending on the condition of the existing pavement and the need for relocating or repairing subsurface facilities such as utility lines and drainage systems. Third, raised central and splitter islands function better for channelizing traffic and ensuring driver compliance with the intent of the mini-roundabout design.

In sum, mini-roundabouts can be a preferred design option on two-lane and three-lane road junctions currently controlled by STOP signs (or traffic signals) that are experiencing recurring congestion. This conclusion is especially applicable to junctions with existing traffic demands in excess of 900 VPH from all directions and projected traffic demands up to 1,600 VPH.

Observation suggests that mini-roundabouts also can improve traffic flow at intersections with lower traffic demands. Regarding a mini-roundabout at Stevensville, MD, Eduardo Arispe, operations research analyst with FHWA, says, "The traditional stop sign [control] on Thompson Creek Road was easy to understand but [tended to] cause an unnecessary queue of vehicles at a place with very low traffic volume. Although it took [drivers] some time to adjust to the new mini-roundabout design, the traffic flow is much smoother as a result."

This rendering shows a fully conforming mini-roundabout for a 24- by 36-foot (7.3- by 11-meter) intersection.

Anecdotal testimony also suggests that a mini-roundabout in Lake Stevens, WA, has been greeted with enthusiastic approval. Brooke Severns with the Seattle real estate division of Safeway, Inc., says that a manager of the grocery store near the roundabout reports, "There have been absolutely no complaints from our customers, just the opposite; everyone finds the traffic to be moving much smoother in and out of our store parking lot and the surrounding roads." In recent months, Severns adds, sales at this location have increased. "I wouldn't hesitate to say that [the roundabout] contributed to a portion of that increase," she says. "I think we would certainly do this in another location after our experience here."

Still, not all is roses. Mick Monken, director of public works with the Lake Stevens city government, reports, "I have received a unique complaint from four citizens following the opening of the mini-roundabout accessing the shopping center. That is, since the city completed these roundabouts, finding a parking space has become difficult as more people are now using these stores."

In this particular case, the full parking lot indicates a successful outcome since one objective of the project was to retain businesses and attract new ones.

Wei Zhang is the intersection safety R&D program manager for FHWA's Office of Safety R&D. He has worked for FHWA for 8 years. In his current position, he develops products and programs that help reduce crashes, fatalities, and injuries at intersections and interchanges. He holds a Ph.D. from the University of Minnesota.

Joe Bared is the concepts and analysis team leader in the FHWA Office of Operations R&D. He has worked for FHWA for more than 22 years. In his current position, he manages research related to modeling and simulation of intelligent transportation systems technologies. Previously, he managed the development of the first roundabout guide in the United States and the new FHWA publication, Alternative Intersections/Interchanges: Informational Report (FHWA-HRT-09-060). He holds a Ph.D. from the University of Maryland.

Ramanujan Jagannathan is program manager for Federal traffic operations at Vanasse Hangen Brustlin, Inc., with more than 10 years of experience in analyses relating to traffic operations and safety and pedestrian safety. He holds an M.Sc. from Virginia Tech.

For more information, contact Wei Zhang at 202-493-3317 or, Joe Bared at 202-493-3314 or, or Ramanujan Jagannathan at 703-847-3071 x5240 or