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. 69 · No. 1 > Preventing Roadway Departures|
Publication Number: FHWA-HRT-05-006
Preventing Roadway Departures
by Harry W. Taylor
Effective solutions are ready for deployment today, as these State examples show.
Of the 42,643 people who died on the Nation’s highways in 2003, more than 25,000 deaths (or 59 percent) occurred when vehicles left their lanes or ran off the road (ROR) and crashed. Although gradual improvements in infrastructure have helped keep the numbers from increasing, research by the Federal Highway Administration (FHWA) shows that higher traffic volumes have counteracted any real reductions in the number of deaths. The statistics lead to one conclusion: Strategies to prevent or lessen the effects of lane and roadway departures are important steps toward improving the safety of the Nation’s roads.
Vehicles depart from a lane or road for a variety of reasons, including the environment (weather or animal crossings), human factors (inattention or drowsiness), design factors (locations with substandard curves, unimproved shoulders, or travel lanes that are too narrow), or a combination of several factors. Lane and roadway departure countermeasures must be designed to keep the motorists in lanes and on the roads, enable the drivers to recover and safely return errant vehicles to the roadway, and keep vehicle occupants from greater harm if a vehicle does leave the roadway.
A critical look at the Nation’s roadways can suggest strategies for systematically dealing with crashes caused by known factors. Many strategies and low-cost countermeasures are available today or offer promise for the future. Six countermeasures with demonstrated benefits, or which show great promise, are signage/markings, rumble strips/stripes, median cable barriers, the safety edge, road widening, and raised median islands.
Signage and Markings
When compared with interstate highways, rural roads exhibit a higher ROR crash rate, especially on sections of roadway with horizontal curves. According to FHWA data, although horizontal curves make up only a small percentage of roadway length, nationally about 23 percent of ROR crashes occur on this kind of curve. Because many small secondary roads were constructed originally to follow property lines, their curves often are not up to current standards in terms of the degree of curvature. Some secondary roads on those old systems have no shoulders, no centerline stripes, and no road edgelines. Some have curves that start with one radius and end with another. Given these conditions, plus local economic constraints that limit reconstruction to change alignments, correctly placed and consistently employed signs may be the most effective strategy to save lives at a relatively low cost.
In Missouri in the past 3 years, 1,205 of the State’s 3,539 traffic fatalities occurred at horizontal curves. Of those, the majority—17 percent—took the lives of young drivers aged 15 to 20. To address this challenge, Missouri is looking at improving and increasing signage on rural roads.
“We’re using chevron signs extensively to mark horizontal curves with a high [crash] history,” says Kevin Irving, transportation engineer and safety programs engineer for FHWA’s Missouri Division. According to the Missouri Department of Transportation’s (MoDOT) Traffic Sign Manual, chevrons may be used in “locations where there is [a crash] history, evidence of drivers losing control on a curve or turn, or a sharp change in alignment not readily visible to the driver.”
For example, on Route F in Laclede County, MO, a curve in the northbound direction exhibited a high crash history. According to MoDOT, ROR crashes at this location appeared to be caused by motorists simply not realizing there was a curve ahead. Additionally, motorists negotiating this curve have just exited I-44, where they have potentially grown accustomed to the 113 kilometers per hour (kph) (70 miles per hour, mph) speed limit. Further complicating the location is the fact that the curve is concealed by a pronounced vertical curve, and there are about 30.5 meters (100 feet) of tangent past this crest before the point of curvature of the horizontal curve. To improve safety at this site, MoDOT installed signage including a “curve right” sign (visible while approaching the crest of the hill northbound) and chevrons.
For another State that has reduced its ROR crashes by improving signage on rural roads, see the successful, low-tech program established by Mendocino, CA. (“Signs Show the Way to Cost-Effective Rural Safety,” PUBLIC ROADS, January/February 2005). In addition, improvements to road signs to increase visibility have become common in recent years. MoDOT recently increased the size of the familiar black-on-yellow chevrons to make them more visible.
Rumble Strips and Stripes
For a number of years, FHWA has actively endorsed the use of rumble strips as a way to reduce ROR crashes. Several varieties of rumble strips (see “Rumbling Toward Safety,” PUBLIC ROADS, September/October 2003) have evolved over the past decade. The spacing between rumbles may vary, and rumble strips may be raised, formed, or—more commonly—milled, and can be used on the shoulder, travel lane, edgeline, or centerline. When used to supplement pavement markings as a low-cost delineation tool, rumble strips can present an effective deterrent to ROR incidents. This countermeasure relies on noise and, in the FHWA-recommended milled design, vibration to attract the attention of an unaware driver who has left the travel lane. In many cases, the rumble strips alert the driver in time to take corrective measures before leaving the road, potentially averting a road departure entirely.
States that have implemented extensive rumble strip programs include Kansas, Michigan, Minnesota, Mississippi, Oklahoma, and Pennsylvania, among others. Michigan, for example, has a successful track record with installing rumble strips. Research by the Michigan Department of Transportation (MDOT) revealed that the milled-in rumble strip demonstrates a design advantage by allowing vehicle tires to partially drop into them, providing a vibration to the vehicle that translates up to the steering wheel. Whereas rolled and concrete intermittent designs can provide some outside noise to alert a drifting driver, the milled design produces a louder noise and adds a vehicle vibration that most certainly increases the potential for alerting a drowsy or distracted driver.
MDOT reports that milled rumble strips installed on Michigan roadways have reduced drift-off-the-road crashes by 40 percent, through the entire range of traffic volumes studied. For more information on the Michigan study, see “Rumbling Toward Safety,” PUBLIC ROADS, September/October 2003.
Minnesota also began exploring rumble strips as a potential solution to high crash rates on the State’s rural roads. Today, the State has instituted a comprehensive policy that mandates placing edgeline rumble strips on all rural multilane and two-lane highway projects where shoulders are constructed, reconstructed, or overlaid, and where the posted speed limit is 80 kph (50 mph) or greater and shoulders are 1.8 meters (6 feet) or greater in width.
According to Gary Dirlam, District 3 traffic engineer for the Minnesota Department of Transportation (Mn/DOT), the department reviewed several reports, including the 1999 FHWA summary report, Safety Evaluation of Rolled-In Continuous Shoulder Rumble Strips Installed on Freeways (FHWA-RD-00-032), which estimated that approximately one single-vehicle, run-off-the-road incident (at an average cost of $62,200) could be prevented every 3 years based on an investment of $217 to install continuous shoulder rumble strips for 1 kilometer of roadway.
“Our crash statistics revealed a significant run-off-the-road crash problem that followed national trends,” Dirlam says. “Shoulder rumble strips on two-lane highways were slow to be implemented due to concerns that a driver could overcompensate and veer into oncoming traffic. Eventually, the success of shoulder rumble strips in reducing run-off-the-road crashes overcame the fear of a one-car crash becoming a two-car crash.”
In the late 1990s, an average of 17 head-on and sideswipe crashes occurred annually on rural roads in Mn/DOT District 3. “But the centerline rumble strip was considered experimental at best,” says Dirlam. Since then, Mn/DOT conducted further research geared toward finalizing a draft policy on implementing centerline strips throughout the State. In October 2003, District 3 installed multiple sections of centerline strips, totaling 274 kilometers (170 miles), across 13 counties. The project doubled the number of miles of in-State centerline strips, which now totals more than 483 kilometers (300 miles).
With national statistics showing that about half of all ROR incidents occur at night, many States, including Mississippi and Pennsylvania, also are applying paint to rumble strips to increase their visibility after dark. A painted rumble strip is referred to as a “rumble stripe.” Because the vertical edges of the strips are painted, the paint line is more visible at nighttime and during wet conditions.
Rumble strips are not without their challenges. During the winter months, sand and salt can get caught in the strips, and because of difficulties navigating around them, roadway sanders may have to apply more material to achieve a thorough application of sand on icy roads. Painting crews as well have noted difficulties with their guide wheels bouncing over centerline strips. And some members of the bicycling public have criticized edgeline rumble strips for making cycling more difficult. Bicyclists prefer to ride along the edgeline—safely out of the travel lane and generally free of the debris that may collect farther back on the shoulder. As a compromise, some State highway departments are now placing breaks in continuous edgeline rumble strips to make an intermittent rumble strip that is more bicycle friendly.
In the future, Minnesota plans to research midlane rumble strips, which run down the center of each lane and take effect when a vehicle’s wheels veer to the right or left from the travel lane. Already installed at the Minnesota Highway Safety & Research Center at St. Cloud State University, midlane strips hold the promise of eliminating concerns for snowplowers, striping crews, and bicyclists.
All current rumble strip installations have met largely positive response from Mn/DOT’s customers. “We hear a number of compliments on the new strip installations, especially from people who have had to drive in whiteout conditions,” Dirlam says.
Median Cable Barriers
Median cable barriers constitute yet another countermeasure that is slated for increased and expanded use in many States. A median cable barrier is a system in which three (or four in some designs) strands of tensioned heavy-duty wire are strung between steel posts and installed in the median to keep vehicles from crossing into oncoming traffic and causing head-on crashes. Cable barriers and rumble strips are two of FHWA’s priority technologies, which are a compilation of innovations that have proven benefits and are ready for deployment. (See www.fhwa.dot.gov/crt/lifecycle/ptisafety.cfm for more information.)
The North Carolina Department of Transportation (NCDOT), for example, conducted research studies in 1993 and 1998 and subsequently installed protective median barriers that dramatically reduced cross-median crashes. By installing median cable barriers on divided roads, and especially those with narrow medians, the department nearly cut in half the number of fatalities from cross-median crashes from January 1999 through December 2003, saving hundreds of millions of dollars in fatal crash costs alone. Many of the cross-median crashes that occurred during this time were on stretches of highway that were slated to receive median barriers but had not yet had them installed. To date, the installation of median barriers in North Carolina has resulted in an estimated 90-percent reduction in freeway cross-median crashes, approximately 25 to 30 lives saved each year, and hundreds of injuries prevented or reduced in severity.
South Carolina also installed cable barriers to reduce crossover crashes. The South Carolina Department of Transportation (SCDOT) recently began implementing this countermeasure along with establishing a reduced speed limit for urban interstate sections and launching a 1-year program to restrict truck traffic on six-lane highways to the outside two lanes.
“The engineers were seriously concerned about reducing crashes and were studying the potential causes [in addition to] the effectiveness of using median barriers,” says South Carolina Department of Transportation (SCDOT) Executive Director Elizabeth S. Mabry. “When the State Infrastructure Bank made funding possible, [FHWA South Carolina Division Administrator] Bob Lee and I announced that we would immediately begin placing barriers. Not only were the barriers installed, but [also] when the barriers are hit, we get them put back within a matter of hours. We’ve had well over 6,000 hits on the median barriers in the past 3 years.”
South Carolina has installed more than 644 kilometers (400 miles) of median cable barrier, which Mabry and Lee believe has significantly reduced this type of fatal crash. The statewide program in place today has become a centerpiece of the South Carolina safety program and the State’s long-term negative fatality trend is beginning to reverse. Total fatalities have dropped each year since 2000, resulting in more than 300 lives saved thus far.
In Missouri, where head-on collisions claimed 480 lives and caused 2,433 disabling injuries from 2001 through 2003, MoDOT is nearing completion of cable barriers or other median barrier installations on its entire I-70 corridor, about 403 kilometers (250 miles). And Missouri plans to begin installations on other interstate routes in the near future.
The Safety Edge
A pavement edge where there is a dropoff of more than 10 centimeters (4 inches) and the angle of the road to the shoulder is 90 degrees is considered unsafe, according to R.A. Zimmer and D. L. Ivey in a paper, “Pavement Edges and Vehicle Stability—A Basis for Maintenance Guidelines,” in Transportation Research Record 946. An estimated 11,000 injuries and 200 deaths per year may be attributed to unsafe dropoffs. Once a vehicle has crossed from a paved surface onto an unimproved shoulder, the driver’s reaction often is to overcorrect to get back on the road. In the process, the rear wheel may catch on the shoulder edge and spin the vehicle around. In many instances, drivers attempting to return to the road often veer into the adjacent lane, cross into opposing traffic, or leave the opposite side of the roadway and become an ROR statistic.
“An easily traversable transition provides a countermeasure to help an errant vehicle reenter the travel lane from the unpaved shoulder,” says Chris Wagner, pavement and materials engineer with the FHWA Resource Center. The safety edge is an angled asphalt fillet of 30 to 45 degrees that creates a tapered edge between the paved travel way and the unpaved shoulder.
In 2003, the Georgia Department of Transportation (GDOT) conducted a pilot project incorporating the safety edge into a 21-kilometer (13-mile) asphalt overlay on Georgia State Route 88, a rural, two-lane, undivided highway. Adding the safety edge was “a concept that appealed to Georgia as a simple way to provide safe roadway conditions during and after a roadway resurfacing project,” says Bryant Poole, who served as GDOT State maintenance engineer during the test project and is currently Atlanta metro district engineer.
“Six areas where dropoffs tend to occur were identified during the preconstruction investigation of the research test sections,” says FHWA’s Wagner, who worked closely with GDOT. “These six areas were at horizontal curves, near mailboxes, turnarounds, shaded areas, eroded areas, and asphalt pavement overlays.”
The test section’s 3.8-centimeter (1.5-inch) asphalt overlay was paved with two hot-mix asphalt designs that are typically specified in resurfacing low-volume roadways in Georgia. Two devices—a steel wedge fabricated by GDOT and the Safety Edge MakerTM from TransTech Systems, Inc.—were used to create a 30-degree safety edge. Both devices produced a durable edge with each of the hot-mix designs. An added benefit noted by GDOT: The safety edge is a useful temporary fix for dropoffs that occur after a pavement overlay is placed, but before earth shoulders can be reconstructed flush with the travel way.
According to Wagner, “Field observations after 1 year of service show that the safety edge has no visible signs of deterioration.” GDOT’s experience showed three important facts: the safety edge can be placed on any type of roadway facility as an integral part of the asphalt preventative maintenance paving process, it has no impact on production, and it represents less than 1 percent of additional material costs.
“We’ve been very pleased with the success of the test project,” says GDOT’s Poole, pointing out that the project met with success “from the beginning to final application of the safety edge, including long-term results.” The test project was so well received, Poole adds, that “in Georgia, starting in March 2005, the safety edge is now required for any preventative maintenance resurfacing projects let to construction by GDOT.” Since the test project’s conclusion, GDOT has worked with representatives of the Georgia Highway Contractors Association, Inc., to ensure that the new safety edge requirement can be put into practice easily. “As a road departure countermeasure, the safety edge provides an easy-to-implement, low-cost measure,” Poole says.
To enhance safety, the Texas Department of Transportation (TxDOT) is widening roads with a paved surface of less than 7 meters (24 feet) in width that have an average daily traffic volume of more than 400 vehicles. Although today’s roadway design criteria place the minimum width of new roadways in Texas at 8 meters (26 feet), “Texas has more than [51,520 kilometers] 32,000 miles of State-maintained roads that are classified as rural two-lane roads with a paved surface of less than [7 meters] 24 feet wide,” says David Bartz, technical assistance team leader and safety coordinator for FHWA’s Texas Division. “Some of these roads are so narrow,” adds Bartz, “that there isn’t even room for an edge stripe. Crash statistics show that the crash rate for similar categories of incidents is three times greater on [5.5- to 5.8-meter] 18- to 19-foot-wide roadways than on [7-meter] 24-foot-wide roadways.”
Bartz also points out that the benefit-cost ratio of shoulder widening is an impressive 4.5:1. Using 2001 data from the National Safety Council’s (NSC) report, Estimating the Costs of Unintentional Injuries, 2003, this figure is calculated by taking the annual safety benefits (estimated lives saved and injuries prevented over the 20-year service life of this type of improvement, using the NSC societal costs) and dividing them by the estimated pavement widening and edge striping cost. “Bringing these roadways up to [7 meters] 24 feet wide makes a huge difference and really does pay off quite a bit,” says Bartz.
William D. Lawson, P.E., Ph.D., a TxDOT researcher who is also a senior research associate and faculty member at Texas Tech University’s Center for Multidisciplinary Research in Transportation (TechMRT), recently presented findings on best practices for pavement edge maintenance at TRB’s 2005 annual conference.
“Among the more rigorous of approaches,” Lawson says, “road widening is perhaps the best long-term solution to the pavement edge dropoff problem because it takes care of some of the fundamental issues that cause dropoffs in the first place. Dropoffs are very strongly correlated with narrow roads and the lack of a shoulder. If you add a shoulder, you widen the road.”
From a systems perspective, Lawson points out that policymakers need to realize that “increases in vehicle size, speed, and traffic level reveal that older, narrow roads in the highway system are being overloaded. Narrow roads without shoulders consistently have the most serious and most prevalent edge problems, and road widening is a reasonable solution to this problem. Widening may not always be the immediate answer because it’s no small thing,” he says, pointing to cost, right-of-way issues, environmental regulations, and a host of other potential policy, design, resource allocation, and methodology issues, “but it is helpful to think in terms of approaching the ideal.”
Lawson’s TxDOT research effort looked at 3 years of available data from the Texas Maintenance Assessment Program (TxMAP), which evaluates a statistical sample of all State roads for pavement condition, traffic operations, and roadside features. Among the observations that emerged: The maintenance district with the best statistics has two-lane roads that are 9.8 meters (32 feet) wide, as opposed to a State average of about 6.7 meters (22 feet). “The TxMAP data showed a strong correlation between those districts with good edge conditions and good roadway performance in general,” Lawson says. “This suggests that a good edge maintenance strategy is necessary for good roads and is perhaps the key element of a successful highway maintenance program.”
Raised Median Islands
Sometimes too much of a good thing can be a liability. Trees can be an aesthetic delight, but when growing too close to a road at locations where run-off-the-road crashes are likely to occur, they often are involved in driver and passenger deaths. In 2003, 452 fatalities involving trees occurred on urban principal arterials, minor arterials, and collectors. A desirable balance between a community’s need for both safety and aesthetics is to plant trees in the median and then shield motorists from them. FHWA, through the National Crash Analysis Center, has funded computer modeling of various protective barrier shapes, and State DOTs have developed, crash-tested, and installed the technology.
“In Washington State, we’ve installed low-profile barriers of 18 to 20 inches [46 to 51 centimeters] tall in a handful of urban locations where the speed limit is less than 45 mph [72 kph],” says Richard Albin, assistant State design engineer for the Northwest Region of the Washington State Department of Transportation (WSDOT). These raised median islands consist of an elevated area in the middle of a roadway with a low-profile traffic barrier that surrounds landscaping elements such as shrubs and trees. Several low-profile traffic barriers have been approved that meet NCHRP Report 350: Recommended Procedures for the Safety Performance Evaluation of Highway Features Test Level 2 criteria of impacts at speeds of 70 kph (45 mph) and 25 degrees.
WSDOT officials are evaluating the safety and effectiveness of the raised median island in cities like Des Moines, WA, and SeaTac, WA, where two-way left-turn lanes have been replaced with a planting strip on portions of State Route 99. WSDOT worked with these cities to develop median solutions that would limit left-turn access and reduce the potential for head-on collisions between vehicles in a two-way left-turn lane approaching from opposite directions. The project also will revitalize the area through the use of attractive landscaping and minimize the potential hazards posed by fixed objects.
Raised median treatments often are associated with traffic calming and speed reduction. But Albin notes, “The research on the effectiveness of these treatments for reducing speeds is inconclusive.” WSDOT’s inservice evaluation will monitor these design elements over the next several years. According to Albin, “We’re going to compare data to see if there’s any speed reduction, reduction in pedestrian collisions, or an increase in fixed-object impacts. From that, we’re hoping to develop better design criteria and identify roadside safety issues for urban areas like these.”
The raised median islands along Washington’s State Route 99 are an attempt to provide a context-sensitive solution that simultaneously meets or enhances safety and capacity needs, environmental goals, and community revitalization desires. Albin explains, “We see this as an optimization of tradeoffs.” By monitoring raised median islands and watching for associated crash rates, WSDOT should be able to determine the effectiveness of more widespread use of the median design throughout the State or alter the solution, if needed.
“In the interim, if a correlation is noted between increased collisions and fixed objects like newly planted trees in median islands where there is no barrier, the trees will be removed or mitigated in some other way,” says Albin. “The sections with barriers will be studied for comparison purposes.” While monitoring this group of context-sensitive solutions, Washington State aims to develop new median designs for urban areas. WSDOT will incorporate identified solutions, where appropriate, into a new design manual.
Roadside Safety Opportunities
Every opportunity for intervention presents an opportunity to save lives. Safety measures designed to keep the driver on the road or to minimize the consequences of leaving the road are the goal. To attain this vision, the vehicle and roadside must be designed to work together to protect vehicle occupants and pedestrians from serious harm. Which of the safety measures available today can help save lives on your roads?
Harry W. Taylor is the road departure safety team leader in the FHWA Office of Safety Design. He has been involved in highway and roadside safety work for more than 25 years. He has participated on numerous NCHRP panels and industry-government roadside safety groups. Taylor is also one of two U.S. observers to the European Committee for Standardization (CEN), Technical Committee 226, Working Group 1, Road Equipment-Safety Barriers. He has a bachelor’s degree in civil engineering from Tennessee State University and a master’s in engineering administration from The George Washington University.
For more information, contact Harry Taylor at harry.taylor @fhwa.dot.gov.
Page Owner: Office of Corporate Research, Technology, and Innovation Management
Scheduled Update: Archive - No Update
Technical Issues: TFHRC.WebMaster@dot.gov