The Double Crossover Diamond

by Joe Bared and Don Saiko

Introducing an innovative interchange for grade-separated crossroads that promises to speed construction, save money, and increase safety.

This aerial photograph shows the Nation's first DCD interchange, located in Springfield, MO. (Inset) An aerial of the conventional interchange in Springfield prior to construction of the DCD interchange.

Here's a riddle. What comes from France and can lead the motorist safely through multiple conflict points? The correct answer is the double crossover diamond (DCD) interchange, also called the diverging diamond.

Many interchanges in mostly urban areas are congested and experiencing high crash frequencies. The conventional diamond interchange constitutes 62 percent of interchange types in the United States. Diamond interchanges are grade-separated and require bridge structures that are costly to widen by adding new lanes, if increased traffic capacity is needed. The DCD design is a variation of the diamond interchange that can alleviate all three of these problems (congestion, high crash frequencies, and cost to add traffic capacity).

The DCD design has existed in France for 30 years, according to Michel Labrousse, director, Centre d'Etudes Techniques de l'Equipement Normandie-Centre, who provided records, signal layouts, traffic flow, and crash data from the ground-breaking interchange in Versailles, France. Installations exist also at Seclin and Perreux-sur-Marne. The DCD interchange in Versailles improved the traffic capacity over the previous design, and it has a low crash rate.

Until recently, few U.S. engineers were aware of the DCD and had rejected its unconventional design. This story is reminiscent of U.S. experience with roundabouts, which took about three decades to adopt from Europe and Australia.

Recently, researchers from the Federal Highway Administration's (FHWA) Turner-Fairbank Highway Research Center (TFHRC) have begun to promote the DCD in the United States. Inspired by the findings reported in a paper by Gilbert Chlewicki, "New Interchange and Intersection Designs: The Synchro-nized Split-Phasing Intersection and the Diverging Diamond Interchange," included in the 2003 Proceedings of the 2nd Urban Street Symposium, FHWA researchers began modeling and developing simulations to study how the DCD design might work in the United States. This research culminated in the Missouri Department of Transportation (MoDOT) becoming the first State agency to embrace the new design by opening the first DCD in the United States on June 21, 2009.

How does the DCD work? The geometric design channelizes vehicular traffic on a grade-separated crossroad from the right side of the road to the left side and then back again at the ramp terminals.

Performance of DCD Versus Conventional Diamond Interchange

Traffic Flow on the Double Crossover Diamond

What distinguishes the DCD from the conventional diamond interchange is that it combines left-turning traffic with through traffic. This is accomplished by having both left-turn and through vehicles cross over to the opposite sides of the roadway at the ramp terminals. The result is that northbound traffic traveling over the bridge travels on the roadway on the left (between the two ramp terminals), and the southbound traffic also travels on the roadway to the left.

Only two signal phases are needed instead of three or four. In the DCD, the crossover junctions are signalized. However, after the crossed-over vehicles have passed the first ramp terminal, left-turn and through movements proceed without having to stop. With the DCD configuration, traffic engineers set the signal control phasing so that vehicles are required to stop at only one of the signals along the arterial road. This design eliminates the left-turn signal phase from the arterial road and also the need for a ramp to store vehicles waiting to go left.

While one direction of the arterial through traffic is proceeding, one side of the off-ramp traffic turning left from the freeway proceeds on green to queue up between the two ramp terminals. For example, while northbound arterial traffic is proceeding, westbound off-ramp traffic from the freeway turning left (south) is also free to enter and wait between the ramp terminals. The same is true for the eastbound off-ramp traffic. When the off-ramp left turns are signalized, off-ramp traffic is likely to stop twice, once at the ramp signal and another time at the crossover intersection.

Benefits of the DCD Design

The DCD design offers advantages in operational capacity, safety, environment, and cost compared with the conventional diamond interchange.

Operational benefits derive from the DCD's ability to combine left-turning traffic with through traffic, thereby eliminating the left-turn-only signal phase of a conventional interchange. The left-turn-only phase is no longer required because both left-turn and through traffic cross over to the left side and at most are required to wait just once at a two-phase signal. This design results in a doubling of throughput of the left-turning arterial traffic and a reduction of total delay when compared with a conventional diamond interchange in high-volume scenarios. At high traffic volumes, the DCD shows about 50 percent less delay in seconds per vehicle than a conventional diamond. Capacity benefits are best when directional traffic is unbalanced because the crossover allows only one movement at a time in comparison to conventional intersections. That means it will be advantageous when the volume of one opposing through movement is greater than the other.

Safety benefits derive from three aspects of the design. First, the DCD has just 14 crossing-path conflicts compared with 26 crossing conflicts in the typical diamond interchange. Safety of the DCD also is enhanced by the reverse curvature preceding the crossover intersections. These curves lead to reduced speeds at the location of the crossing-path conflict points and are expected to lead to fewer crashes.

Another safety benefit is that the DCD increases safety and mobility for pedestrians and bicyclists. Unlike the traditional diamond intersection, through traffic in the DCD often must stop at least once as it traverses the intersection. (The benefit of this design is derived primarily from conditions of unbalanced directional flows. If the traffic is balanced fifty/fifty, the DCD design will not be favorable.) Currently, MoDOT and FHWA are evaluating the safety benefits in terms of crash frequency and conflicts, and data should be available within 2 years. Data on crash reductions are available from the Versailles site, which experienced 11 minor injury crashes in 5 years compared to 23 fatal and injury crashes at a typical U.S. diamond interchange in a similar time period. A typical diamond interchange is a design without frontage roads or without other intersections in close proximity to the ramp terminals.

Environmental benefits. Environmental benefits can only be roughly estimated as the data have not yet been quantified. Initial benefits are realized at the construction stage. These benefits are derived from the compactness of the design that allows for a smaller footprint and therefore more green space, and lower right-of-way costs. In the case of a retrofit, these gains are even greater because excavation and modification of the existing overpass is avoided since space for the additional lanes required by a conventional design is not needed.

Other environmental benefits of DCDs derive from the operational benefits of reduced vehicle idling and fewer incidents.

Cost benefits. In the case of a retrofit, a recent project to convert an existing interchange into a DCD in Springfield, MO, saved $6.8 million compared to a single point urban interchange or widening of a conventional diamond design. The traditional design would have required the existing bridge to be torn down and a new, wider bridge to be built.

Making It Work In the USA

With so many advantages, it would seem that implementing the DCD in the United States would be straightforward. Certainly, U.S. engineers recognize the design's potential, but questions remain. Would U.S. motorists be comfortable with opposing traffic passing them on the right instead of the left? Might some drivers, particularly less experienced or older drivers, have difficulty navigating such an interchange? And would local leaders be comfortable introducing a new and unfamiliar design that had only limited testing in the United States?

Engineers at TFHRC recognize that answers to such questions are needed before more informed recommendations for implementation in the United States can be made. As a first step, FHWA engineers in 2004 began to explore the geometric and operational aspects of the DCD design. To accomplish this, they developed two types of simulations: microsimulations to facilitate modeling of design and operational characteristics, and a full-scale drivable model to allow direct observation of driver performance.

DCD Traffic Simulation -- Micro Models

To start with, FHWA researchers acquired comprehensive data on various operational aspects of the DCD design. Engineers at TFHRC constructed numerous simulations of DCDs and then examined their performance under various operating conditions.

One microsimulation used was VISSIM, a tool for transportation professionals who want to simulate different traffic scenarios (at individual vehicle microscopic levels) before starting implementation. Compared to a conventional interchange, the DCD design showed increases in capacity up to about 30 percent. The greatest advantage was derived when opposing traffic volumes were more unbalanced and left-turning volumes from the arterial and off-ramps were high.

DCD Laboratory Simulation -- Driving Model

Once the TFHRC engineers were satisfied with the model predictions of operational characteristics, the next step was to determine a suitable location for a driving simulation and work with local officials to build it. To this end, FHWA held discussions with MoDOT about a Kansas City site that State transportation engineers were designing as a DCD. The MoDOT engineers visited TFHRC to virtually drive through their proposed design in FHWA's highway driving simulator. As a result, they gained confidence in the design's feasibility and also assurance that drivers would indeed be able to navigate safely through this novel intersection design. The driving model also enabled the engineers to make design modifications to the geometry, signals, and signs.

In addition to the Kansas City design, TFHRC modeled a comparable conventional diamond interchange. The main findings from 70 volunteer drivers who participated in the experiment showed that they navigated through the DCD correctly, as they did in a comparable diamond interchange. Mean speeds through the DCD were about 24 miles per hour, mi/h (39 kilometers per hour, km/h), compared to 34 mi/h (55 km/h) at the conventional diamond. The reduced speed does not reduce capacity.

Service Volumes of Conventional and DCD Interchange Designs

* (L/T) means that the left-turning traffic, as well as through traffic, uses the lane.
Source: Transportation Research Board.

Missouri Experience

Although MoDOT planned and designed the Kansas City DCD interchange first, the State actually opened the Nation's first DCD at a site in Springfield, MO, in June 2009. The Springfield DCD interchange opened before the Kansas City one because of budgetary situations. MoDOT constructed the Springfield project to alleviate congestion on the heavily traveled Kansas Expressway (Missouri Route 13) at I-44, while providing a pedestrian and bicyclist crossing down the center of the bridge.

MoDOT selected this innovative design because it would be faster to build, cheaper to construct, and safer for motorists and pedestrians. These goals became MoDOT's motto for the project: quicker, cheaper, safer.

Quicker. MoDOT built the project in 6 months instead of 12 to 18 months, mainly because the existing Kansas Expressway bridge over I-44 was rehabilitated and kept in place. A new, much larger bridge would have been needed if a single-point urban interchange had been built instead.

Cheaper. The project cost came in at $3.2 million because the existing bridge was used in place. Reconstructing this interchange as a single-point urban interchange would have raised the cost to about $10 million. MoDOT put the money saved toward other projects.

Safer. During the first 6 months of operation, the DCD reduced crashes by 50 percent between the ramp terminals and by 25 percent between the first intersections north and south of the interchange, compared to the same period in 2008, based on crash data obtained by MoDOT from the city. Most were rear-end crashes, and none were head-on from driving the wrong way. MoDOT accomplished this reduction by eliminating left-turn conflicts and reducing bumper-to-bumper congestion. MoDOT attributes none of the remaining crashes to the DCD design.

The DCD also made this interchange safer by including a 9.5-foot (2.9-meter) walkway down the center of the bridge for pedestrians and bicyclists visiting retail and residential areas to the south and recreational centers (fairgrounds and zoo) to the north. A concrete barrier wall on each side of the walkway separates pedestrians and bicyclists from vehicles. Crosswalks at the signals on each end of the walkway provide a safe way to cross traffic.

Shortly after the interchange opened, MoDOT contracted with a research company to conduct a mail survey to assess customer satisfaction with the DCD project. The contractor mailed surveys to 400 randomly selected Springfield area residents and received 75 responses. Most respondents thought the project made the roadway safer (96.7 percent), more convenient (95.1 percent), less congested (95.2 percent), easier to drive (86.9 percent), and better marked (89.8 percent). They agreed that the DCD was the right transportation solution (89.4 percent).

The relocation of the road next to the fairgrounds and zoo helped to alleviate traffic congestion that used to occur. According to Melinda Arnold, public relations and marketing director at the Dickerson Park Zoo, which is adjacent to the Springfield DCD, "Our experience is that since the DCD opened, there don't seem to be backups during peak travel times."

Sergeant Tom Royal, of the Springfield police department's traffic division, notes that before the DCD opened, the police had to address numerous crashes with vehicles going northbound on Route 13 and making left-hand turns across two lanes of traffic to go west on I-44 at the interchange. Since the DCD interchange opened, "We've seen a drastic reduction in crashes," says Royal. "Crashes took a nosedive, and the reconfiguration has alleviated the traffic congestion."

"The Turner-Fairbank Highway Research Center had a major role in identifying, evaluating, and producing computer simulations; testing design alternatives; and assisting the Missouri Department of Transportation in construction," says Monique Evans, director of FHWA's Office of Safety Research and Development. "For years, the Federal Highway Administration has encouraged innovative and cost-effective intersection designs, and the double crossover diamond interchange is an excellent example."

This pedestrian walkway is located in the median of the DCD interchange in Springfield, MO.

Joe Bared, Ph.D., P.E., is team leader for Transportation Operations Concepts & Analysis in the FHWA Office of Operations Research and Development. He has worked at FHWA for more than 20 years and managed the program area on intersection/interchange safety and operational effects of design. He managed development of the first roundabout guide in the United States and has promoted intersection/interchange design innovations in a new FHWA publication, Alternative Intersections/Interchanges: Informational Report (FHWA-HRT-09-060).

Don Saiko, P.E., is a transportation project manager with MoDOT, where he has worked for 18 years. He supervised the Springfield DCD project. Saiko graduated from the University of Minnesota with a B.S. in civil engineering.

For more information, contact Joe Bared at 202-493-3314 or joe.bared@dot.gov, or Don Saiko at 417-895-7692 or donald.saiko@modot.mo.gov.