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 - Sept/Oct 2007

Sept/Oct 2007
Issue No:
Vol. 71 No. 2
Publication Number:
Table of Contents

The Low-Cost Dropoff Solution

by Steve Moler

The safety edge, a relatively easy and inexpensive countermeasure to steep pavement edges, is reducing crashes on rural two-lane highways.

James Leben, an assistant district maintenance engineer with the Georgia Department of Transportation, measures a tapered safety edge applied in a 2004 demonstration project along a 21-kilometer (13-mile) section of State Route 88 near Augusta, GA. The safety edge is one means to help prevent crashes caused by pavementedge drop-offs.

Four teenage boys from a high school in Clayton County, GA, were driving to school on a rural two-lane highway on a March morning in 2003 when something went terribly wrong.

About a quarter-mile from campus, the car's right tires slipped off the pavement and dropped onto the sandy shoulder. While attempting to return to the pavement, the 16-year-old driver overcorrected and lost control. The compact sedan crossed the centerline and slammed head-on into a school bus coming in the opposite direction. The driver and one passenger were pronounced dead at the scene; another passenger died later at the hospital. The fourth teenager was seriously injured but eventually recovered. The driver of the school bus, which was carrying no passengers, suffered only minor injuries.

One cause of the crash, according to the police report, involved a condition known as pavement-edge dropoff (PEDO), the uneven edge or vertical dropoff between the paved travel lane and the unpaved shoulder. Highway safety experts consider a dropoff of 12.7 centimeters (5 inches) or more to be unsafe, especially if the edge is at a 90-degree angle to the shoulder surface. A dropoff of 5.1 centimeters (2 inches) or more is considered a potential driving hazard. The dropoff along the stretch of highway where the teens' car slipped off the pavement ranged from about 5.1 to 10.2 centimeters (2 to 4 inches), according to two safety engineers from the Federal Highway Administration (FHWA) who visited the site the day after the crash.

When a vehicle slips off the pavement and onto an unpaved shoulder, the steep edge can make it difficult for a driver to reenter the paved travel lane safely. Studies show that when a driver encounters a steep pavement edge, he or she attempts to return immediately to the paved travel lane but in doing so tends to oversteer, causing intense rubbing, or "scrubbing," of vehicle tires against the pavement edge, which initially prevents the vehicle from climbing back onto the pavement. This oversteering can cause loss of control at the moment when the right rear tire climbs back onto the pavement, causing the vehicle to fishtail or go into a broadside skid.

FHWA Safety Engineer Frank Julian uses his shoe for scale to demonstrate the extent of the pavement-edge dropoff along the section of roadway where three teenagers were killed in a crash.

"This is likely what happened in the Lovejoy High School crash," says FHWA Safety Engineer Frank Julian, who visited and took photographs of the crash site and learned that there was evidence of scrubbing on the inside edge of the vehicle's right tires. According to the police report, skid marks were found coming back onto the roadway, leading the inspectors to believe that "overcorrection played a role in the [crash]."

Although relatively rare compared with other crash types, PEDO-related crashes tend to be more severe, say the authors of a recent study sponsored by the AAA Foundation for Traffic Safety. In fact, these crashes are more likely than others on similar roadways to result in serious injuries and are two to three times more likely to be fatal, primarily because the vehicle often leaves the roadway, rolls over, hits a roadside object, or is involved in a head-on collision. According to FHWA, an estimated 11,000 people suffer injuries and roughly 160 die annually in crashes related to unsafe pavement edges.

"Pavement-edge dropoff has been around for a long time, and it will continue to be a serious problem unless we do something different," says former FHWA Chief Highway Safety Engineer Rudy Umbs, who now works in the agency's Resource Center. "We need to keep asking ourselves, What are we going to do differently tomorrow and next week to eliminate pavement-edge dropoffs and reduce the potential for lane departure crashes like run-off-the-road and head-on crashes? The safety edge can make that difference."

One solution is to install a 30- to 35-degree tapered asphalt wedge or fillet, known as a safety edge, along each side of the roadway during resurfacing projects. The safety edge not only provides an angled and compacted transition that eliminates the abrupt dropoff, but it also provides for a stronger and more stable pavement edge, which makes it easier for drivers to maneuver their vehicles safely back onto the roadway. By offering a tapered, rather than vertical, transition between the paved surface and the unpaved shoulder, the safety edge is a low-cost means of improving safety on paved two-lane highways.

At the time of the crash, the highway, shown here, was a typical two lane rural road with unpaved shoulders. Recently this section of road was upgraded to four lanes due to commercial and residential development in the area.

Factors in PEDO Crashes

Numerous studies over the past three decades have analyzed the various combinations of conditions and circumstances that lead to PEDO-related crashes, such as the one involving the Georgia teens. Virtually all of the studies found that whether a driver regains control or crashes after slipping off the pavement edge depends on a variety of circumstances, including vehicle speed, steer angle, the vehicle's departure and return angle, vehicle size, dropoff severity, driver skills, roadside obstacles, and whether another vehicle is coming in the opposite direction. In the Georgia incident, as in many PEDO-related crashes, several dangerous circumstances converged to create the conditions conducive to a crash.

Red paint, as well as FHWA Safety Engineer Frank Julian's foot, marks the spot where the car reentered the paved travel lane and went into a broadside skid, colliding head on with a school bus coming in the opposite direction.

First, according to police findings and eyewitness accounts, the driver was likely exceeding — perhaps far exceeding — the 72 kilometer-per-hour, km/h (45 mile-per-hour, mi/h), speed limit. The car was a small sedan, which studies, including one conducted by researchers at the University of Michigan Transportation Research Institute, show have more difficulty recovering from pavement-edge dropoffs. The dropoff at the crash site was within the range that experts consider unsafe, and the driver was inexperienced, having had his driver's license for just 13 days. And finally, a vehicle — indeed, a fairly large one — was coming in the opposite direction. To complicate the situation, none of the teenagers was wearing a safety belt.

Magnitude of the Problem

A variety of conditions in the roadway environment can contribute to PEDO, including pavement-edge breaking, erosion, wear of the unpaved shoulders, or inadequate maintenance. A Transportation Research Board (TRB) report, Construction of a Safe Pavement Edge: Minimizing the Effects of Shoulder Dropoff, indicates that edge dropoff most commonly is encountered around mailboxes, on the insides of curves, on steep grades, at turnarounds, and along shaded areas. A combination of shoulder erosion and edge rutting caused by harsh weather and vehicles repeatedly leaving the paved travel lane typically is found at these locations.

Another circumstance that can aggravate PEDO is failure to bring the shoulder flush with the pavement following a resurfacing project. Problems develop when the pavement edge begins to crumble quickly from the lack of compaction, creating a vertical drop. Edge rutting and soil erosion from repeated vehicle impacts and the weather soon follow.

Just how serious is the PEDO problem overall? In 2006 the AAA Foundation for Traffic Safety sponsored a comprehensive study that attempted to answer that question. Conducted by the Iowa State University Center for Transportation Research and Education (CTRE) and the Midwest Research Institute, the study, Safety Impacts of Pavement Edge Drop-offs, found that most States routinely sample edge dropoffs for maintenance purposes. But that information is either not available or only indicates that a sampled section exceeds a certain threshold — in a sense, passes or fails, without giving a more descriptive assessment.

To gain a clearer understanding of the magnitude of the PEDO problem, the AAA Foundation studied randomly selected sections of paved rural two-lane highways with unpaved shoulders in two Midwestern States. The study found that 12 percent of dropoffs sampled in one State were 5.1 centimeters (2 inches) or more, 1 percent were 7.6 centimeters (3 inches) or more, and less than 1 percent were 10.2 centimeters (4 inches) or more. In the other, the situation was slightly worse. Almost 19 percent of the dropoffs sampled there measured 5.1 centimeters (2 inches) or more, 3 percent were 7.6 centimeters (3 inches) or more, 1 percent were 10.2 centimeters (4 inches) or more, and less than 1 percent were 12.7 centimeters (5 inches) or more.

The researchers found that in most cases, States are aware of edge dropoffs as an issue and, in many cases, have aggressive maintenance policies in place. However, edge dropoff persists due to harsh weather and traffic conditions. State and local agencies also have significant highway mileage to maintain and might not always be aware of all the locations where dropoffs have formed, according to Shauna Hallmark, a transportation engineer and professor in the Department of Civil, Construction, and Environmental Engineering at Iowa State's CTRE and the lead researcher on the AAA Foundation's study.

Another question transportation researchers are trying to answer is how often do PEDO-related crashes occur? Various studies in recent years have examined this question. One study, known as the Southeast United States Fatal Crash Study and headed by Karen Dixon, a civil engineering associate professor at the Georgia Institute of Technology, evaluated 150 randomly selected fatal crashes on rural two-lane State and nonstate highways in Georgia in 1997. Dixon and her colleagues estimated that in 21 of the 69 nonState-system fatal crashes in Georgia, or about 30 percent, edge rutting or PEDO was a causal factor.

The aforementioned study by the AAA Foundation also evaluated the number of crashes where characteristics indicated that edge dropoff might have had an impact. The researchers found that 17.7 percent of crashes on rural two-lane roadways in one State and 24.7 percent in another were probably or possibly related to edge dropoffs. They also found indications that PEDO crashes are run-off-the-road crashes, which in general are more likely to be severe than other crash types.

Evolution of the Safety Edge

In numerous studies over the years, researchers have sought to understand the conditions that lead to PEDO crashes. In 1982, the Texas Transportation Institute conducted one of the first studies on the advantages of using an angled wedge along the pavement edge to minimize dropoff severity. The theory behind a tapered pavement edge was that it could help drivers make a smoother, more controlled reentry onto the paved travel lane than if there was a more abrupt or vertical edge.

A researcher is using a ruler and level to measure pavement-edge dropoff where rutting and erosion have occurred on a rural road in Iowa.

In a 1986 study, researchers Don L. Ivey, a civil engineering professor emeritus at Texas A&M University, and Dean L. Sicking, a civil engineering professor at the University of Nebraska-Lincoln, analyzed the steer angle needed to remount dropoffs with different heights and edge shapes at 80 km/h (50 mi/h). A 10.2-centimeter (4-inch) vertical edge generally caused loss of control, but as the edge shape became flatter, fewer impacts were felt. The researchers then evaluated 5.1-, 10.2-, and 15.2-centimeter (2-, 4-, and 6-inch) dropoffs with a 45-degree wedge and found that drivers could recover within the 3.7-meter (12-foot) travel lane even with as much as a 15.2-centimeter (6-inch) dropoff.

That same year, researchers at the University of Michigan Transportation Research Institute compared vertical and 45-degree wedge dropoffs with hard and soft shoulders, various passenger vehicle sizes, and front- and rear-wheel-drive vehicles using nonprofessional drivers. The results showed that none of the nonprofessional drivers could negotiate a vertical dropoff of 11.4 centimeters (4.5 inches) or higher at any speed. Dropoffs near 7.6 centimeters (3 inches) could be negotiated at speeds of 48 km/h (30 mi/h) in large passenger cars, but smaller cars needed lower speeds to recover.

But with a 45-degree edge, drivers always were able to recover within their own lane when traveling at speeds up to 89 km/h (55 mi/h). The researchers also evaluated soft shoulders using a professional driver and concluded that dropoff height, not shoulder material, was the determining factor in being able to recovery safely.

A 2005 study by the Center for Intelligent Systems Research used computer modeling to analyze vehicle recovery for various vehicle types, dropoff heights, and wedge angles. Pierre Delaigue, a research scientist with the center, presented the results from the study, Safety of Excessive Pavement Wedge Due to Overlays, at the 2005 TRB meeting in Washington, DC.

The Georgia wedge, shown here, is bolted onto the screed end gate.

The study concluded that flatter wedges were always safer than steeper ones, regardless of dropoff height. Tractor semitrailers were most sensitive to PEDO, while pickup trucks were least sensitive. Passenger cars recovered from dropoffs of up to 12.7 centimeters (5 inches), provided a pavement wedge of 45 degrees or flatter was present. However, a 10.2-centimeter (4-inch) dropoff with a 45-degree wedge was too severe for the tractor semitrailer. But a 30-degree wedge allowed all vehicle types to recover safely.

Creating the Taper

Although the idea of using a tapered wedge had been around for years, determining a method to lay the wedge along a road shoulder during a resurfacing project posed an ongoing challenge.

In early 2003, Resource Center Safety Engineer Frank Julian and Pavement and Materials Engineer Chris Wagner began developing some basic concepts for how to create a tapered edge along the roadway shoulder. First, they built on Wagner's experience at the National Center for Asphalt Technology, where he conducted research in the late 1990s on creating tapered wedges at the longitudinal joints of asphalt pavement. Julian and Wagner next conceived of developing a device that could be attached to the end of an asphalt paver screed. The device needed to be able to create a smooth and compacted wedge along the edge of the roadway.

As shown here, the safety edge is measured 30 to 35 degrees up from the horizontal at the toe of the wedge surface.

From this conceptual stage, Julian and Wagner developed a partnership with the Georgia Department of Transportation (GDOT) to design and plan a demonstration project to study the constructability of what they termed the safety edge on a resurfacing project. GDOT began the demonstration project in 2004 along a 21-kilometer (13-mile) section of State Route 88 just south of the town of Augusta. GDOT's maintenance department developed its own in-house device known as the Georgia wedge. Conceived by GDOT Maintenance Project Manager Lynn Bean, the wedge is essentially a modified strike-off bolted onto the screed end gate. The shoe of the end gate rides on the pavement shoulder and moves freely vertically, allowing it to adjust to height changes. A rounded leading edge produces the smooth appearance.

The safety edge was successfully installed with little impact on production and a project cost increase of less than 1 percent. After 1 year, the Georgia demonstration project found no visible signs of deterioration and reported no expectations for any long-term degeneration along the safety edge sections.

The sections of roadway paved without the safety edge during the demonstration project had degraded to a near-vertical edge during that same time, with cracking developing near the edge. The Georgia study concluded that the safety edge showed "promise as a low-cost solution to mitigate pavement shoulder dropoff . . . The implementation of the safety edge design would be most applicable to asphalt resurfacing projects on two-lane undivided roadways with limited paved shoulders."

According to highway safety experts at FHWA, the safety edge is not intended to be an alternative to a flush shoulder, but rather can serve as a safety feature used in conjunction with current shoulder specifications. The recommended pavement wedge, measured 30 to 35 degrees from the horizontal, helps prevent drivers from overcorrecting if they drift onto the shoulder, thereby decreasing the likelihood of the vehicle crossing into opposing traffic or leaving the roadway.

"The safety edge is an ideal solution since the impact of a vehicle encountering a vertical difference between the edge of the roadway and shoulder can be lessened when dropoffs form and agencies aren't immediately aware and able to address the problem," says CTRE's Hallmark.

Without the safety edge, this section of highway in a 2004 Georgia demonstration project developed a near-vertical drop just 1 year after a pavement overlay.

For these reasons and others, the safety edge is now a standard feature of resurfacing projects in Georgia.

Other Demonstration Projects

While the Georgia demonstration project got underway, FHWA asked TransTech Systems, Inc., to develop a device to manufacture the safety edge commercially. The company adapted its Notch Wedge Joint MakerTM, which creates a tapered edge at the longitudinal joint on asphalt resurfacing projects, to produce the Shoulder Wedge Maker, which would create the safety edge. The device attaches to the screed face instead of the end gate and has a self-adjusting internal spring that allows it to follow the roadside surface independently of other paver components. The device has an angled surface that precompacts the asphalt as it enters the device, while another fixed-angle surface forms the tapered edge. As the asphalt is placed beneath the wedge-forming surface, it is smoothed to a finished surface on the tapered edge.

A recent demonstration project in New York's Schenectady County using the Shoulder Wedge Maker showed positive results. After the safety edge was installed in 2004 along two rural roads, annual inspections revealed that "the shoulder wedge has held up exceptionally well, with no degradation of the edge," says Dave Clements, an associate director in the Office of Operations Management for the New York State Department of Transportation, who supervised the inspections. Additional analysis has shown no cracking or breaking away of the wedge from the main rolled mat area.

The sections with the safety edge in the same 2004 Georgia demonstration project showed no visible signs of deterioration.

The Indiana Department of Transportation (INDOT) currently is involved in an FHWA-sponsored study under the Transportation Pooled Fund Program to evaluate the safety edge's effectiveness in helping prevent and mitigate PEDO-related crashes. Eight companies were awarded contracts that called for installing the safety edge in 2004 and 2005 along nine rural, two-lane highways with minimal shoulders, using either the Shoulder Wedge Maker or a similar device developed in-house at INDOT. Four contractors used the existing device, while the others developed their own versions loosely based on the Georgia wedge.

The contractors have installed the safety edge successfully on seven of the nine projects to date, at minimal additional cost. "In fact, most of the contractors didn't even factor the safety edge into their bids," says Elizabeth Pastuszka, an INDOT pavement and materials engineer who was involved in construction of the demonstration projects. "In the two unsuccessful projects, the problems we had were totally unrelated to the safety edge itself."

Reducing Tort Liability

Another benefit of the safety edge is the potential to reduce tort liability. In the 2004 FHWA report Construction of a Safe Pavement Edge: Minimizing the Effects of Shoulder Dropoff, FHWA's Wagner and GDOT researcher Yeonsoo Stanley Kim noted that PEDO is a common source of tort claims against highway agencies. The authors cite court cases in Louisiana, Minnesota, and South Carolina where monetary judgments were awarded to motorists involved in PEDO crashes. In these cases, the transportation agencies were found liable for creating unsafe conditions and not warning about them.

The study by the AAA Foundation in Iowa and Missouri found that crashes in which PEDO was the major cause resulted in major tort liability suits. Claims filed between 2000 and 2005 in Iowa, for example, in which "pavement/shoulder edge" or "shoulder conditions" were cited as the major cause of the crash were the highest ranking tort liability claims in terms of total dollar value. In fiscal years 2000-2003, these claims accounted for 38 percent of the total dollar value of claims filed against the Iowa Department of Transportation.

In a typical pavement resurfacing project, the asphalt at the edge remains soft because there is little or no compaction, as seen here.

From 2000 to 2005, 23 PEDO- related tort liability claims were filed against Iowa, the study reported. Of those, however, compensation was awarded to the plaintiff in only two cases. Based on these findings, the authors concluded: "The [Iowa] DOT believes [the State's] demonstrably strong maintenance policy has contributed to [its] success in defending against tort liability claims related to pavement-edge dropoff. When States and other highway agencies are not able to defend themselves successfully, edge dropoff can result in significant liability."

From 1988 to 2003, Louisiana had 388 claims filed against it for alleged roadway shoulder defects, including PEDO, according to former Assistant Attorney General James R. Dawson. The State paid out an average of $62,144 per claim, totaling more than $241 million, regardless of whether it won or lost the case.

A 30- to 35-degree tapered edge or fillet, as demonstrated here by an engineer using a ruler and level, provides a compacted, and therefore more stable, pavement edge.

Citing these statistics at a 2004 workshop on managing pavement-edge dropoffs in Atlanta, GA, Dawson concluded that "shoulder defects, including dropoff problems, are a major factor in how we are able to spend our dollars on improving highway safety."

The safety edge is indeed a promising technology that safety experts say can help eliminate these shoulder defects and prevent tragedies like the one near Lovejoy High School from happening again.

Guidelines and Recommendations on PEDO

Although no national standards currently exist for pavement-edge dropoffs, several government and industry organizations provide some guidance:

A 2006 AAA Foundation for Traffic Safety study recommends that highway agencies require routine comprehensive sampling of PEDO on their roads, suggesting that any dropoffs of 5.1 centimeters (2 inches) or more should be corrected. The report also recommends that agencies adopt a policy of providing paved shoulders with a minimum width of 0.6 meter (2 feet) wherever possible and incorporate the safety edge in all roadway resurfacing projects to prevent severe PEDO.

The AAA Foundation further recommends that highway agencies review their databases to assess how PEDO might have contributed to crashes and then conduct additional research on crash occurrences and pavement-edge hazards specifically for rural roads. In addition, the study encourages highway agencies to train maintenance and construction staff, including private contractors, on the potential hazards of PEDO.

In its Roadside Design Guide, the American Association of State Highway and Transportation Officials (AASHTO) states that no vertical dropoff greater than 5 centimeters (2 inches) should occur between adjacent lanes, and pavement-edge dropoff greater than 7.6 centimeters (3 inches) should not be left overnight.

The Manual on Uniform Traffic Control Devices provides recommendations for signs used to warn motorists of unexpected conditions. For example, if the pavement-edge dropoff is less than 7.6 centimeters (3 inches), a "Low Shoulder" sign should be used. If the PEDO exceeds 7.6 centimeters (3 inches), a "Shoulder Drop Off" sign is recommended.

The AASHTO book A Policy on Geometric Design of Highways and Streets (also known as the "Green Book"), states that regular maintenance should provide for a shoulder that is flush with the pavement surface. Unstable shoulders generally undergo consolidation over time, and the elevation of the shoulder tends to sink below the paved travel lane. The resulting dropoff can adversely affect drivers when they slip onto the shoulder.

FHWA's Standard Specifications for Construction of Roads and Bridges on Federal Highway Projects offers guidance on PEDO in work zones. The document states that when shoulder dropoffs exceed 5 centimeters (2 inches) a "Low Shoulder" warning sign should be placed during construction. With dropoffs greater than 10 centimeters (4 inches), a 1:3 (18-degree) fillet with "Low Shoulder" warning signs should be provided.

Steve Moler is a public affairs specialist at FHWA's Resource Center. He has been with FHWA since 2001, providing the agency's field offices and partners with support in media relations, public relations, and public involvement communications. He earned a bachelor's degree in journalism from the University of Colorado at Boulder. He can be reached at 415-744-3103 or

For more information about road departure issues and effective countermeasures, please visit the FHWA Office of Safety's "Road Departure Safety" Web site at, or contact Chris Wagner at 404-562-3693 or, or Frank Julian at 404-562-3689 or