Skip to contentUnited States Department of Transportation - Federal Highway Administration FHWA Home
Research Home
Public Roads
Featuring developments in Federal highway policies, programs, and research and technology.
This magazine is an archived publication and may contain dated technical, contact, and link information.
Federal Highway Administration > Publications > Public Roads > Vol. 73 · No. 1 > A New Left Turn

Jul/Aug 2009
Vol. 73 · No. 1

Publication Number: FHWA-HRT-09-005

A New Left Turn

by Ramanujan Jagannathan, Warren Hughes, and Joe G. Bared

These preliminary observations discuss the operational benefits and potential safety of displaced left-turn intersections.

This aerial view was taken approximately 12 months after this displaced left-turn (DLT) intersection in Baton Rouge, LA, opened to traffic.
This aerial view was taken approximately 12 months after this displaced left-turn (DLT) intersection in Baton Rouge, LA, opened to traffic.

Navigating intersections is one of the most complex actions that drivers perform. In 2008, 37,261 fatalities took place on the Nation's roadways. According to the National Highway Traffic Safety Administration's (NHTSA) Traffic Safety Facts 2006, about 23 percent of the total fatal crashes in 2006 occurred at intersections, and more than half of the combined number of fatal and injury crashes were intersection-related. The annual cost to society for intersection-related crashes is estimated to be approximately $40 billion. Despite improved design and traffic engineering measures, the annual human toll from motor vehicle crashes at or near intersections has not changed much in a quarter century, NHTSA notes.

In addition to the safety impacts, conventional intersections can impose costs on the public in terms of congestion and travel delay. Traffic congestion at signalized intersections is frequently worsened when left-turn phases are introduced. By eliminating left turns at the main intersection, transportation agencies could provide more green time to vehicles moving through the intersection instead of losing time for left turns that conflict with through traffic. The elimination of left turns using an approach known as a displaced left-turn lane (DLT) has been implemented primarily for congestion relief but could have safety benefits.

The DLT intersection, also known as a continuous flow intersection (CFI), eliminates potential conflicts between left-turning vehicles and oncoming traffic by adding a left-turn bay to the left of oncoming traffic prior to the main intersection. Vehicles access the left-turn bay upstream of the main signalized intersection and cross over the median and the opposing through segment and thereby eliminate left-turn crossing within the main intersection.

Transportation agencies have implemented DLTs to reduce congestion at several locations in the United States. The first DLT was built in Long Island, NY, in 1996. Another, built in Maryland in 2000, has a DLT lane only on one approach; it has reduced the average delay per vehicle through the intersection by about 60 percent of the major through movement and a comparable reduction on the major left-turn movement, according to simulated results from the Maryland State Highway Administration.

This photograph shows vehicles approaching a DLT and mainline intersection on Route 30 eastbound in St. Louis, MO.
This photograph shows vehicles approaching a DLT and mainline intersection on Route 30 eastbound in St. Louis, MO.

A 2004 traffic simulation analysis published by the Transportation Research Board, Design and Operational Performance of Crossover Displaced Left-Turn Intersections, estimated that DLTs helped reduce average delay 48 to 85 percent and cut queue lengths 62 to 88 percent, compared to conventional intersections. Researchers are still evaluating the effect of this design on intersection operations and safety, but many State departments of transportation are considering DLTs, based on these findings alone.

DLTs: What They Look Like

The main feature of this alternative intersection is relocation of the left-turn movement on an approach to the near side of the opposing roadway, which eliminates the left-turn phase for this approach at the main intersection. Traffic that would normally turn left at the main intersection would first cross the opposing through lanes at a signal-controlled intersection about 300-400 feet (91-122 meters) upstream of the main intersection. The left-turning vehicles then travel on a new roadway parallel to the opposing lanes and execute the left turn with a signal simultaneously with the through traffic at the main intersection. Traffic signals at the main intersection and the locations of the left-turn crossovers are timed so that all vehicles (including left-turning vehicles) have at most one red phase to wait for.

Diagram. This diagram depicts in plan view the crossover movements in a typical full DLT. A key indicates that red circles signify 'signal-controlled crossover' points, the blue oval in the center of the intersection denotes the 'signal-controlled main intersection,' orange arrows denote 'left-turn crossover movements,' and yellow arrows signify 'opposing through movements at signal-controlled crossover' points. The diagram consists of the intersection of two roadways, one in a north-south orientation and the other east-west. In each of the four legs (eastern, western, northern, southern) of the roadways, an orange arrow pointing toward the intersection indicates traffic flow. A little closer to the intersection, a red dot appears in each roadway, and immediately after this is an orange arrow pointed a little to the left; this is the beginning of each of four crossover displaced left-turn lanes (of the DLT movement). Also at these four points are yellow arrows pointing away from the intersection, indicating the opposing traffic and the lanes crossed over by the left-turn lanes. A blue oval covers the intersection, and within it are orange arrows indicating the left, 90 degree turns made from the crossover displaced left-turn lanes.

Engineers have designed various types of DLTs, including one with fully displaced left-turn lanes on all four approaches to the intersection, and another with displaced left-turn lanes on either the main road or crossroad of the intersection.

Traffic Operations

The major benefit of DLTs involves increased throughput (comparable to capacity) by reducing the number of phases in a signal cycle and eliminating the left-turn phases. A soon-to-be-published Federal Highway Administration (FHWA) study, Alternative Intersections/Interchanges: Informational Report (in preparation), used a traffic simulation model to compare throughput results for the following four DLT cases to those of conventional intersections:

Case 1. Three lanes on the major road intersecting three lanes on the crossroad.

Case 2. Three lanes intersecting two lanes.

Case 3. Two lanes intersecting two lanes.

Case 4. A T-intersection with three lanes intersecting two lanes.

For a full DLT, three of the four approach cases (#1, 2, and 3) yielded 30 percent increases in throughput when the flows are fully balanced (that is, when approximately equal numbers of cars are traveling in each opposing direction). Case 4 increased throughput by 16 percent. Case 4 reveals a smaller increase in throughput because the reduction of phases for a T-intersection is only from 3 to 2. The reduction in phases of cases 1 to 3 (four-leg intersections) is from 4 to 2 for a full DLT and from 4 to 3 for a partial DLT.

When mainline opposing flows are unbalanced with a 30/70 directional split (for example 30 percent traveling north, while 70 percent are traveling south), the increase in throughput from comparable conventional intersections is as follows: case 1, 25 percent; case 2, 25 percent; case 3, 25 percent; and case 4, 12 percent.

For a half or partial DLT, the increase in throughput of case 1 ranges from 14 to 20 percent from unbalanced to balanced flows. For case 2, the increase in throughput ranges from 10 to 20 percent from unbalanced to balanced flows. Cases 3 and 4 were not conducted for partial DLT.

Advantages and Disadvantages

Conversion to a DLT has some advantages over expanding capacity at a conventional intersection or changing to a grade-separated crossing. According to the FHWA study referred to above, Alternative Intersections/Interchanges: Informational Report, DLTs are much less expensive to construct compared to conventional grade-separated interchanges, and crews can build them in less time. At high-volume intersections, DLTs have the potential to reduce travel times considerably. Simultaneous movement of the left-turn and through traffic promotes improved progression of traffic platoons (moving of vehicles in groups where the vehicles are closely interspaced) on the arterial and increases vehicular throughput.

Diagram. This diagram depicts a plan view of the intersection of two six-lane highways, one in a north-south orientation and the other east-west. For the eastern and western legs, each has two left-turn lanes on the far left, having already crossed over the opposing lanes. For the northern and southern legs, each has a single conventional left-turn lane (between the opposing lanes of traffic). Flanking all the lanes on the right are curving exit lanes; for example, the western leg has a lane on the far right for cars exiting to get in the southbound lanes of the southern leg. At the intersection are eight pedestrian control devices. On the northwestern corner is a mast arm with two lights facing the two eastbound crossover displaced left-turn lanes and two lights facing the westbound lanes, across the intersection. Also on the northwestern corner is a pedestrian control device with two signals, one facing eastward and directing pedestrians looking to cross the north-south highway in a westerly direction, and the other facing southward and directing pedestrians looking to cross the east-west highway in a northerly direction. On the southeastern corner is a similar arrangement. A mast arm has two lights facing the two westbound left-turn lanes and two lights facing the eastbound lanes, across the intersection. Also on the southeastern corner is a pedestrian control device with two signals, one facing northward and directing pedestrians looking to cross the east-west highway in a southerly direction, and the other facing westward and directing pedestrians looking to cross the north-south highway in an easterly direction. On the northeastern corner is a mast arm with one light facing southward to direct the traffic in the single northbound, westerly turning left-turn lane, and two lights facing the northbound lanes, across the intersection. Also on the northeastern corner is a pedestrian control device with two signals, one facing westward and directing pedestrians looking to cross the north-south highway in an easterly direction, and the other facing southward and directing pedestrians looking to cross the east-west highway in a northerly direction. On the southwestern corner is a similar arrangement. A mast arm has one light facing northward to direct the traffic in the single southbound, easterly turning left-turn lane, and two lights facing the southbound lanes, across the intersection. Also on the southwestern corner is a pedestrian control device with two signals, one facing northward and directing pedestrians looking to cross the east-west highway in a southerly direction, and the other facing eastward and directing pedestrians looking to cross the north-south highway in a westerly direction.

Considering potential safety benefits, a partial DLT has 30 conflict locations and a full DLT has 28, compared to a conventional intersection's 32 conflict locations. For both partial and full DLTs, the conflict points are more dispersed than at conventional intersections. In a conventional intersection, the traffic conflict points are concentrated in the middle of the intersection. However, in a DLT intersection, the conflict points are distributed at multiple intersections (that is, the central intersection and the crossover points). The only crash data comparison available is from a Baton Rouge, LA, site, which has been in operation since 2006. A raw, before-after data comparison shows a reduction in total annual crash frequencies by 24 percent. Annual fatal plus injury frequencies went down by 19 percent. Total crash rates per million entering vehicles were reduced by 24 percent for total crashes, and fatal plus injury rates went down by 22 percent.

On the downside, DLTs have larger footprints compared to conventional intersections. This might be a significant factor in urban areas where right-of-way is limited and costly. Access to land parcels in the quadrants of a DLT can be restricted, and agencies might need to eliminate U-turns at DLTs. In addition, pedestrians must walk across ramps to cross certain legs, and the intersection design can be challenging to those with visual impairments because the paths and some traffic movements are atypical.

Diagram. This diagram depicts a typical full DLT: the intersection of a north-south highway and an east-west highway. Each of the four legs (eastern, western, northern, southern) has a red area on its left side that broadens slightly as it nears and then meets the intersection. These are labeled 'Extra for DLT.' Each leg also has a thinner, gray area on its right side that tapers slightly as it nears and then meets the intersection. This is labeled 'Extra for Conventional.'

An FHWA study, Evaluation of Sign and Marking Alternatives for Displaced Left-Turn Lane Intersections (FHWA-HRT-08-071), using a driving simulator is finding economical and effective treatments that agencies can use at DLTs to help motorists recognize the left-turn crossovers upstream of major intersections. For instance, post-mounted signs on both sides of the road appear to work as well as mast-arm signs in guiding drivers to the correct paths.

The Baton Rouge Case

In recent years, traffic had become a serious problem for Baton Rouge, LA, given a backlog of State and Federal transportation projects awaiting funding. Fortunately, local engineers had been working on an innovative concept that would dramatically improve traffic flow at one of the city's busiest intersections.

Annual Average Collision Rates for Baton Rouge DLT
Year Number of Reported Collisions
By Severity
All Reported Crash Rate(per million entering vehicles on major road) Fatal + Injury Crash Rate(per million entering vehicles on major road)
Total Fatal + Injury
Annual Average "Before"(2002-2005) 147 37 5.09* 1.26*
Annual Average "After"(June 2006-May 2008) 111 30 3.87** 0.98**
Difference
"After"-"Before"(Percentage Difference)
-36
(-24.4%)
-7
(-18.9%)
-1.22
(-23.9%)
-0.28
(-22.2%)

Source: DOTD.

* AADTs for years 2003 and 2004 were interpolated.

** AADT for year 2006 was interpolated.

In 2002, an engineering firm had approached the Louisiana Department of Transportation and Development (DOTD) with a recommendation to consider a DLT for highly congested signalized intersections in the State. The firm explained that conventional improvements, such as overpasses, are costly and interrupt businesses, and a DLT would be a viable alternative for solving urban congestion.

DOTD identified the intersection of Airline Highway and Sherwood Forest Boulevard/Siegen Lane as having the most potential for a DLT solution. Baton Rouge jumped quickly from an affirmative decision to construction.

The St. Louis Case

The Missouri Department of Transportation's (MoDOT) St. Louis Area District was seeking to provide more mainline green time at the intersection of MO 30 and Summit Road, because of growing traffic volumes. MoDOT estimated a 25 percent traffic increase by 2030. The mainline is a four-lane divided route, carrying more than 50,000 vehicles per day.

Simulation analysis estimated a typical traditional intersection design (including dual left-turn lanes on each approach) would have resulted in a delay in the peak hour of 110.1 seconds per vehicle, a level of service (LOS) of F (on a scale of A to F, where A corresponds to best service and F means worst), based on simulation analysis. The DLT, on the other hand, is expected to result in an average delay of 29.5 seconds per vehicle, with a LOS of C. Further, the traditional solution would have provided 54 of 115 seconds, or 47 percent, mainline green time, but the DLT completed in 2007 provides 74 of 115 seconds, or 64 percent, mainline green time. Another reason MoDOT ruled out a traditional interchange was that the site is too close to another interchange to allow for adequate weaving distances. (Weaving distance is the length of the roadway section where the vehicles from the expressway section change lanes to take the off-ramp, and vehicles entering the expressway section from the on-ramp change lanes to merge with the expressway traffic.)

This screen grab from FHWA's driving simulator shows a left-turn crossover movement on a four-lane highway, with two leftturn lanes branching off the northbound lanes and crossing the two southbound lanes before turning left (westward). Signs posted downstream of the left-turn lanes give motorists advance warning of the crossover displaced left-turn lanes.
This screen grab from FHWA's driving simulator shows a left-turn crossover movement on a four-lane highway, with two leftturn lanes branching off the northbound lanes and crossing the two southbound lanes before turning left (westward). Signs posted downstream of the left-turn lanes give motorists advance warning of the crossover displaced left-turn lanes.

The DLT approach is not perfect, of course. "We get maybe four to six snowstorms every year with accumulation that needs to be plowed," says Jeanne Olubogun, P.E., traffic operations engineer with MoDOT's St. Louis District. "Our operations staff dislikes the islands created by the DLT because they complicate plowing, but they do understand the benefits." Notwithstanding these issues, the State decided that a DLT was the best option and proceeded to install it.

Overall construction of the DLT lasted about 8 months and cost $4.5 million, compared with $3 million for a traditional intersection. Construction was not without difficulty: The intersection is in a large rock cut, and much rock had to be blasted to build the new road connection and the DLT. "We had very few problems in traffic management during the transition from existing condition to the DLT," notes Olubogun. The increased cost for the project included the increased quantity of pavement for a DLT versus a standard intersection, additional islands constructed to separate conflicting movements, and the additional rock cuts to facilitate the larger footprint of the DLT intersection.

Applications

In addition to the Baton Rouge and St. Louis sites, several other DLTs now are operating in the United States:

  • Crews built a DLT prototype at a T-intersection at the entrance of the Dowling College National Aviation Technology Center in Long Island, NY, in 1996.
  • The meeting of Route 210 (Indian Head Highway) and Route 228 (Berry Road) in Accokeek, MD, is a T-intersection that operates under a full signal. Built in 2000, the left-turn movement is on the side-street approach only, rather than on the major road approach.
  • A DLT at 3500 South and Bangerter Highway in Salt Lake City, UT, opened in September 2007.
Photo. This aerial photo shows a DLT at the T-intersection in Shirley, NY. In a mostly rural area, a four-lane highway (CR 46) runs in a roughly north-south direction (the northern part is tilted a little west of the southern part). Another four-lane highway (Brookhaven Airport Road), running in a roughly east-west direction (the eastern end is tilted a bit north of the western junction with the other highway), intersects the former from the east at a T-intersection. Near the top of the photo, two southbound lanes branch off to the left to become the left-turn lanes, flow south, then cross over the northbound lanes and continue south until reaching the intersection. Then the lanes turn left and become part of the eastbound lanes on the other (east-west) highway. At the intersection, two lanes branch off the westbound lanes and flow north to join the northbound lanes. Thus, the left-turn lanes, once they have crossed the northbound lanes, are between the northbound lanes and the north-flowing lanes coming off the east-west highway.
At the three-legged half DLT in Shirley, NY, shown in this aerial photograph, the DLT movement is on the side-street approach to the intersection, rather than on the major road approach.

Promising Outlook For Navigation

Some transportation professionals feared DLTs would require an educational process for drivers, and experience shows that some acclimation may be needed. Large signs guide drivers to the displaced left-turn lane, which is itself intuitive and easy to follow, according to the latest summary report, Evaluation of Sign and Marking Alternatives for Displaced Left-Turn Lane Intersections, (FHWA-HRT-08-071). An educational process may be helpful to improve adaptation to this new design.

"Studying the map is more confusing than driving the intersection," says Michael G. Bruce, an engineer involved in the Baton Rouge project. "This is one of the reasons that safety has improved."

In fact, DOTD officials were concerned early on that the DLT concept would confuse Baton Rouge drivers during power outages and result in dangerous conditions. "The power went out for about 3 hours about a month after the CFI [DLT] opened," Bruce says. "The intersection happens to be near a State traffic camera, so we have 3 hours of video of drivers operating very safely. There was no confusion and no dangerous situation. The drivers knew exactly how to handle the intersection."

In this Baton Rouge, LA, intersection, the left-turn lane upstream of the main intersection crosses the median and the opposing side of through traffic to the displaced left-turn lane.
In this Baton Rouge, LA, intersection, the left-turn lane upstream of the main intersection crosses the median and the opposing side of through traffic to the displaced left-turn lane.

The DLT solution is also a much less expensive approach than overpasses. The Baton Rouge DLT is expected to cost about $5 million, compared to an overpass that would have cost about $30 million to build. "When you do need to acquire land, the DLT costs much less than a conventional overpass," Bruce says. A conventional widening could provide comparable capacity improvement. However, the additional lanes in both directions of travel would have to be extended for a certain distance beyond the intersection to provide continuity. When more right-of-way is necessary, additional costs are difficult to estimate in comparison to the DLT unless the estimate is site-specific.

Getting Traffic Moving

The Baton Rouge and St. Louis cases are among several current success stories, with more likely to come. A two-legged DLT that opened in 2007 on the Bangerter Highway in Utah, similar to the Baton Rouge intersection, is proving successful, according to the Utah Department of Transportation (UDOT). Also, according to Lisa Wilson, P.E., UDOT Region 2 traffic operations engineer, "Several other cities and the traveling public are wondering when CFIs will be built at their intersections along the Bangerter Highway."

In addition, crews are building a DLT in Mississippi, and designers are drawing up two in Ohio.

This photograph shows a head-on view of traffic entering the DLT intersection on Route 30 in St. Louis, MO.
This photograph shows a head-on view of traffic entering the DLT intersection on Route 30 in St. Louis, MO.

 

This aerial shot shows the St. Louis intersection of Gravois Bluffs Boulevard and Summit Drive, with the DLT crossover upstream of the main intersection.
This aerial shot shows the St. Louis intersection of Gravois Bluffs Boulevard and Summit Drive, with the DLT crossover upstream of the main intersection.

Applicability

Replacing a conventional intersection with a full DLT intersection can produce a 50 to 85 percent reduction in average intersection delays and a 10 to 25 percent increase in intersection throughput, according to traffic simulation results. Replacing a conventional intersection with a half DLT intersection can produce a 30 to 40 percent reduction in average intersection delays and a 10 to 20 percent increase in intersection throughput. Some of the situations where a DLT intersection may be suitable are the following:

  • If the volume/capacity ratio is greater than 0.8 on two opposing intersection approaches.
  • If the product of left-turn and opposing through vehicles is greater than 150,000 on two opposing intersection approaches.
  • If the left-turning volume is greater than 250 vehicles per lane and the opposing through volume is greater than 500 vehicles per lane during peak hours on two opposing intersection approaches.
  • If an intersection is heavily congested with many failures of signal phases to handle peak traffic volumes.
  • If left-turn queues at an intersection spill beyond the left-turn storage bays.

For more information, see FHWA's Alternative Intersections/Interchanges: Informational Report (in draft).


Ramanujan Jagannathan is a transportation engineer in the Vienna, VA, office of Vanasse Hangen Brustlin, Inc. (VHB). Jagannathan has 6 years of experience with operational analysis and pedestrian safety analysis. He has a thorough understanding of transportation planning, traffic engineering, geometric design, and traffic safety, and a broad knowledge of current transportation-related software packages.

Warren Hughes, P.E., P.T.O.E, is the regional manager for the Capital District of VHB, where he oversees and participates in projects relevant to traffic engineering and operations, transportation planning, highway and roadway design, safety, ITS, and airport landside transportation systems. Hughes has more than 29 years of diverse transportation engineering experience.

Joe G. Bared, Ph.D., P.E., has been a highway research engineer in FHWA's Office of Safety Research and Development for the past 18 years. He currently manages the research program area on intersection safety and design, and conducts research in the areas of safety and operational effects of highway design. He initiated and managed the projects that published FHWA's Roundabouts: An Informational Guide (FHWA-RD-00-67) and Signalized Intersections: Informational Guide (FHWA-HRT-04-091).

For more information, contact Joe Bared at 202-493-3314 or joe.bared@dot.gov, Ramanujan Jagannathan at 703-847-3071 or rjagannathan@vhb.com, or Warren Hughes at 703-847-3071 or whughes@vhb.com.

ResearchFHWA
FHWA
United States Department of Transportation - Federal Highway Administration