U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
To gain more insight into the part that pavement markings play in roadway safety, the researchers conducted a literature review at the start of Phase II of the study. This chapter details previous findings regarding adverse weather crashes, work zone safety, and pavement marking effectiveness. This information provided valuable direction for the conduct of Phase II.
Rainfall frequency across the United States varies greatly, and it creates more opportunities for difficult driving conditions for certain areas. Higher rainfall means an increased exposure to wet driving surfaces. In 2008, Pisano, Goodwin, and Rossetti gathered adverse weather crash data and related it to weather conditions across the country.(1) Their report included a nationwide rainfall map, shown in figure 1. From a rainfall frequency perspective, it is clear that drivers in the northwestern and southeastern United States are at a higher risk due to higher rainfall.
Figure 1. Chart. 2008 mean total precipitation.(1)
Nighttime driving conditions also vary over parts of the U.S. In the northern parts of the continental U.S., nighttime can last over 15 hours, and drivers are forced to operate their vehicles in dark conditions, including during peak travel periods. This is the case in such cities as Seattle, Washington; Minneapolis, Minnesota; and Rochester, New York. Even in the southernmost urban areas such as Miami, Florida, and Corpus Christi, Texas, wintertime nights are around 13.5 hours.(2)
Driving in rainy, snowy, sleety, and/or windy conditions carries additional risk above that of driving in dry conditions. Both a driver’s perception of the road ahead and the wheel’s ability to grip the road are reduced, making it more difficult to maneuver safely on highways. Between 1995 and 2005, 24 percent of all traffic fatalities were due to adverse weather, about 7,400 each year. The number injured annually in adverse weather crashes is a couple of magnitudes larger, totaling 673,000.(1)When further examining the statistics, 47 percent of these crashes occurred during rainfall, and 75 percent occurred on wet pavement. The remainder are caused by snowy and slushy roads, fog, and high winds.
The Midwest has the highest rate of average weather-related crashes because of high amounts of snowfall. In fact, 40 percent of all nationwide weather-related crashes occur in this area, which only accounts for 22 percent of the U.S. population. The South, however, is not far behind, with 32 percent of total weather-related crashes and 36 percent of the total U.S. population. The ratio of rainfall-related crashes in the South is the highest in the nation—over 64 percent of all weather-related crashes, which likely corresponds to the high levels of rainfall.
Compared to the growth in vehicle miles traveled, adverse weather-related crashes decreased during the study period. Regardless, the average number of crashes is over 1.5 million annually. The risk of driving in adverse weather will remain high due to low visibility during precipitation and poor tire traction on wet pavement, which consequently reduces vehicle control and maneuverability.
Qiu and Nixon conducted a study to compile the research done on the effects of adverse weather on crash rates, injury crash rates, and fatal crash rates. Through meta-analysis, they compared daily cash rates during adverse weather with those during non-adverse weather. At a confidence interval of 95 percent, the study concluded that rain could increase the crash rate by 71 percent and the injury rate by 49 percent.(3) Although the percentage of crash rates related to snow events has decreased over time, the study found that this was not the case for crash rates related to rain events. The increased risk of crashes, injuries, and fatalities during rainy and/or wet conditions demonstrates the need for more informative, noticeable pavement markings.
Hummer et al. analyzed data from the Highway Safety Information System to document the characteristics of crashes in North Carolina.(4) They found that, from 2003 to 2005, approximately one-third of crashes occurred in reduced lighting conditions (dusk, dawn, or dark). Additionally, 23 percent of collisions happened during adverse surface conditions (19 percent when wet and 4 percent with ice, snow, or slush).
The United States has a transportation system that is reaching its capacity limits. The infrastructure is aging and needs to be replaced. Therefore, to maintain mobility for drivers, roadway maintenance and construction must occur concurrently with traffic operations. Safety in work zones is essential, especially as the number of highway rehabilitation and construction projects continues to increase. Since 1983, the first year of data available, the National Work Zone Safety Information Clearinghouse reports a general rise in work zone fatalities as a percentage of the total number of highway fatalities (figure 2).(5)
Over the 10-year period from 1997 to 2006, work zone fatalities in crashes increased by 45 percent.(6) Although there was a marked drop in 2007 to levels not seen since the 1990s, work zone fatalities are increasing as a percentage of total fatalities. In the 1980s, work zone fatalities accounted for 1.5 percent of all highway fatalities, but this increased to 1.8 percent in the 1990s, and then to 2.4 percent since 2000, peaking at near 2.8 percent in 2002.(5) This equates to 717 fatalities in 1996 and 1,095 in 2003, or an increase of 35 percent.(7)
Figure 2. Graph. Safety in work zones.(5)
Work zone safety has become a larger issue over the years. Even as total annual fatalities have slowly dropped, the percentage of work zone fatalities (as a share of total fatalities) has increased. The increase in work zone fatalities is likely a result of increased amounts of maintenance activity on the nation’s highways. Between 1997 and 2004, an increased share of construction costs was spent on existing facilities, growing from 47.6 to 51.8 percent of capital funds.(6) The FHWA also estimates that at any given time, 20 percent of some portion of each roadway in the National Highway System is under construction.(8) Additionally, the crash rate of any particular stretch of highway has been shown to increase by almost 30 percent during a construction period.(9) Although the percentages vary, another study also found that work zones greatly increase crash risks.(10) Lastly, according to the FHWA, approximately half of all fatal work zone crashes occurred at night.(6) This stresses the importance of visible pavement markings to help drivers navigate through work zones.
The most common method of marking pavement is to spray pavement marking on the roadway surface and then drop glass beads with an index of refraction of 1.5 to make the markings more visible at night. The pavement marking, typically made of waterborne pavement marking, thermoplastics, epoxy, or solvent pavement marking, acts as a binder for the glass beads. Eighty-nine percent of State transportation agencies use waterborne pavement markings, which are the least expensive among these options.(11)
The retroreflective performance of the glass beads under given conditions depends on their refractive index (RI). Generally speaking, under dry conditions beads with an RI of 1.9 provide the highest retroreflectivity. Under wet conditions the RI needs to be greater, around 2.4, for high retroreflectivity because a film of water covering the beads changes the optical conditions that provide retroreflectivity.
An issue with conventional pavement marking systems is their poor performance in wet conditions. While dry, headlight illumination on pavement markings with 1.5 index beads typically provides sufficient guidance information to the driver. When covered by a layer of water, such as during rainfall, their retroreflective properties diminish. To a driver, it appears that the pavement markings have disappeared. The reason for this drop in performance is that the water reduces the amount of light that is retroreflected from the markings to the oncoming driver, due to a loss of effectiveness of the optical system. It becomes very difficult to see the pavement markings, and drivers have increased difficulty navigating a safe route.
To improve visibility of pavement markings in wet weather, 3M developed the AWP, which maintains retroreflective properties while covered with a film of water. The key to this technology is specially developed elements that provide retroreflection in both dry and wet conditions. After the AWP is sprayed on the road surface, 3M bonded core elements are dropped, followed by a second drop of conventional glass beads. Figure 3 shows a comparison of standard pavement marking and the AWP during a typical rain event in Henderson, North Carolina. The AWP, when applied correctly, is visibly more retroreflective.
Figure 3. Photos. Wet weather comparison of standard pavement marking (left) and the AWP (right) pavement markings.
The AWP Marking
Each 3M element is composed of a structural core surrounded by microcrystalline ceramic beads that have a range of higher refractive indices.(12) The refractive indices of the small attached beads are 1.9 and 2.4, compared to the typical 1.5 used in most standard pavement markings. The beads with a refractive index of 1.9 perform well in dry conditions, while the beads with the higher RI of 2.4 perform well in wet conditions, thereby improving overall visibility for the driver in a variety of weather conditions. Figure 4 displays a side-by-side comparison between standard pavement marking and the AWP marking.
The AWP binds the larger glass beads and the 3M elements to the pavement. According to 3M’s website, “3M All Weather Paint includes a ‘high-build’ resin to enable thicker application for lines that last up to twice as long as conventional traffic pavement marking,” and is manufactured with a high-build polymer emulsion.(13) The pavement marking is ideal in locations where other existing waterborne equipment is available.(12) It is applied at 25 mil (.635 mm) wet thickness, providing a very thick, wear-resistant layer of pavement marking.
Figure 4. Photos. Comparison between standard pavement marking (left) and the AWP marking with microcrystalline ceramic beads (right).
To optimize the construction, 3M tested various mixtures of beads and elements, drop rates, and pavement marking thicknesses. Due to the variations possible, there is an opportunity for customization and optimization. The mixtures for this test were selected from 24 candidate samples of pavement markings tested on a New Orleans, Louisiana, test track in November 2007.(14) These candidates were measured for their retroreflectivity under dry and wet conditions over a period of 6 weeks. The degradation of the different mixes varied—elements and glass beads became loosened from the binding pavement marking and scattered away due to repeated traffic passes—and only those that performed well were recommended for future tests.
Later, the Texas Transportation Institute, using the field research facility at the Texas A&M University Riverside Campus in Bryan, Texas, performed a human factors study to determine drivers’ ability to see the AWP.(14) The driving course, outfitted with a rain range able to simulate rainy conditions, was sprayed with the pavement marking mixes that had performed well on the New Orleans test track. Two test segments were set up, once under the rain range and once elsewhere on the course for a dry weather comparison. Researchers first subjected drivers to the dry weather nighttime drive-through. Participants would announce when they were first able to see the end of the pavement marking, and the researcher would note the distance the vehicle was at that time. Next, the drivers were instructed to drive through the rain range where the wet conditions were simulated and similarly announced when they could view the end of each pavement marking. From these tests, one pavement marking mix was chosen to be implemented in the field due to its better visibility. The specifications provided are:
3M medium-sized high refractive index dual-optics drop-on elements at a drop rate of 8g/lineal ft in combination with MODOT Type P (or AASHTO M247 Type I).1.5 index glass beads at a drop rate of 12g/lineal ft applied in a double-drop onto a high-build waterborne pavement marking applied at a 20 mil [.508 mm] wet film thickness.
Lane markings influence driver behavior continuously, and they are one of the most common ways drivers receive information about the roadway alignment ahead. Their presence helps influence the driver’s position on the road, specifically within a travel lane. The position of a vehicle in a lane has a high correlation with safety, with correlation coefficients typically between 0.7 and 0.8, showing that the existence of pavement markings to guide the driver is extremely important. To ensure drivers have adequate guidance at night, the pavement marking’s retroreflective properties must provide sufficient visual detection distance. Rumar and Marsh concluded that drivers need between 3 and 5 seconds of “preview time” to guide their vehicle.(15) They describe that, “in night driving, the single-vehicle crash risk increases proportionately with the decrease of average geometric sight distance” and that “this is an indication of the importance of longer preview times.” Pavement markings that do not have high retroreflective properties put drivers at risk at night.
Pavement markings are also influential on a driver’s speed, but the precise nature of that influence is unclear. Improved lane markings have reportedly increased speeds, but speeds have also dropped in some locations.(15) Speed increases are likely to happen because drivers have an increased amount of information being provided to them about the roadway ahead. This increases driver confidence to navigate the road, allowing an increase in speed. However, it is still to be determined if this increase in speed creates a safety risk.
Rumar and Marsh focused their study on visibility to evaluate pavement markings, as opposed to the crash rate, which they rejected as being inconclusive. They also hesitated to use speed, lateral placement, and the number of overtakings as measures of effectiveness.(15) Detection distance was identified as the most important aspect to the researchers. Emphasizing their case, the authors discussed several studies on the visibility of lane markings and retroreflectivity. Their paper presents examples of where lane markings give great benefit, such as at night, when a driver’s point of fixation moves down and to the right, to avoid the bright lights of oncoming vehicles. Here, edge lines become critical for the driver to navigate the roadway safely.
Emphasizing visibility at critical times, Jacobs et al. discussed pavement markings in rainy conditions.(16) The study looked at several pavement marking products that differed in performance properties during wet conditions. Drivers were asked to drive a section of road where the pavement markings changed from traditional to all-weather performance. The drivers were to indicate when they noticed the change in the pavement marking products. The distance from where drivers indicated that they noticed the change to where the pavement marking actually changed was the primary measure of effectiveness and was obtained during dry and wet conditions. Table 1 displays the comparison table from that study; however, it should be noted that the samples taken in wet and dry conditions were not comparable and must be read independently. The authors strongly articulated that, in these situations, lane markings are most necessary, but they are prone to fail more often at these times. Such concern is what the AWP pavement markings are intended to address, especially in temporary application such as work zones.
|Condition of markings||New||Old|
|Condition of road||Dry||Wet||Dry||Wet|
No oncoming glare
Oncoming low-beam glare
Carlson et al. chose two conflict MOEs to study the effect of pavement markings in rural high-speed maintenance work zones.(17) They considered 14 different conflict types and found only 2 applicable to work zones. The first was the slow vehicle conflict, which involves a faster vehicle overtaking a slower vehicle. The second was the lane change conflict, where a vehicle needs to brake to switch travel lanes because of a lack of gaps in the other lane, or a vehicle changing lanes and forcing vehicles behind it to brake or swerve when there was not a large enough gap.
In Phase I of this study, the researchers identified three additional MOEs.(18) The Phase I report stresses the importance of these MOEs and describes the following as necessary for evaluation of the new pavement markings: mean of the lateral placement of the vehicle in the travel lane, variance of the lateral placement of the vehicle in the travel lane, and number of times the vehicle touches the edge markings. The first two MOEs are correlated with crash frequency. There is a long tradition in the field that vouches that edge markings can be related to crash frequency as well. Increases in the inadvertent contact of the edge markings is a result of drivers varying their position in a lane away from the center, and a reduction would indicate increased safety benefit.
The retroreflectivity of pavement markings is defined by the coefficient of retroreflected luminance, usually expressed in units of millicandela per meter squared per lux (mcd/m2/lux). Higher values indicate that the marking will appear brighter when illuminated by a set of headlights. Thus, a higher value of retroreflectivity should provide a longer detection distance and enable drivers to make earlier navigation decisions. As speed increases and drivers have less time to make judgments, it is critical that the pavement markings are still detected at distances that adequately inform users of the road geometry ahead. In Schieber’s study on aging drivers, he showed how retroreflectivity needs increase at a higher rate than speed, as seen in figure 5.(19) Schieber also described how higher retroreflectivity is critical for aging drivers whose vision has decreased.
Figure 5. Graph. Minimum pavement marking retroreflectivity requirements as a function of speed.(19)
Opposing arguments for high retroreflective pavement markings exist. Bahar et al. found that increased retroreflectivity in longitudinal pavement markings does not have any safety benefit.(11) Their claim is that the most important aspect of pavement markings is that they be present and visible for drivers; the level of brightness of the pavement markings is of less importance to safety. The argument is that drivers adapt to higher retroreflective markings and increase speeds due to increased confidence of the road ahead, thereby negating the safety benefit. Higher speeds place the driver at risk for more frequent and higher severity crashes.
Summarizing past literature, Bahar et al. described how the relationship between retroreflectivity and nighttime crashes has not been sufficiently proved. Likewise, Lee, Maleck, and Taylor were unable to identify any relationship between nighttime crashes and the retroreflectivity of pavement markings(20) Another study, by Abboud and Bowman, compared long-term crash rates to the average crash rate of crashes influenced by line visibility.(21) Although this study recommended a minimum value of retroreflectivity, Bahar et al. disputed the data analysis methods used. Bahar et al. studied the safety effect of pavement markings themselves and found that no pattern of improved safety exists as retroreflectivity is increased. Ultimately, they concluded that any safety effect is too small to detect, and the brightness of lane markers shows no increase in safety. They also reported that driver adaptation minimizes road safety improvements even in adverse weather conditions.