U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590

Skip to content
Facebook iconYouTube iconTwitter iconFlickr iconLinkedInInstagram

Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations

This report is an archived publication and may contain dated technical, contact, and link information
Publication Number: FHWA-HRT-07-059
Date: October 2007

Updates to Research on Recommended Minimum Levels for Pavement Marking Retroreflectivity to Meet Driver Night Visibility Needs

PDF Version (660 KB)

PDF files can be viewed with the Acrobat® Reader®


In order to develop recommended minimum in-service pavement marking retroreflectivity values, TARVIP models were constructed to determine driver needs under various scenarios. A matrix was constructed using factors that affect pavement marking visibility and could be adjusted in TARVIP. Reasonable values for each factor were chosen to test minimum required retroreflectivity sensitivity to those parameters. As not all factors affecting pavement marking visibility can be accounted for using TARVIP, care was taken to ensure that the modeled scenarios conservatively modeled those factors.

  • A list of the factors varied in the TARVIP analyses is shown below with the levels studied. Further justification is provided below.
  • Pavement surfaces (2 levels: old asphalt, old concrete).
  • Pavement marking configurations (3 levels: yellow dashed center line with white edge lines, yellow dashed center line, white left lane line).
  • Vehicle type (2 levels: passenger sedan, commercial truck).
  • Vehicle speed (3 levels: 64 km/h ( 40 mi/h ) , 88.5 km/h ( 55 mi/h ) , 112.7 km/h ( 70 mi/h ) ).

Factors held constant in the TARVIP analysis are

  • Pavement marking material (alkyd paint with standard glass beads).
  • Overhead lighting (none).
  • Line width (102 mm (4 inches)).
  • Required preview time (2.2 seconds).
  • Pavement wear (old surfaces).
  • Center line configuration ( 3.0-m (10 -ft ) skip with 9-m (30 -ft ) gaps).
  • Windshield transmission (0.7).
  • Atmospheric transmissivity (0.86 km -1).
  • Weather conditions (dry).
  • Oncoming vehicle glare (none).
  • Headlamp type (UMTRI 2004 50 percent low beam).
  • Driver age (62 years old).
  • Pavement marking degree of obliteration (none).
  • Lateral separation between double lines (no double lines were investigated).
  • Driver workload (not distracted/low workload).
  • Driver attention (full).
  • Horizon/sky luminance (none).

4.1 Selection of Pavement Surfaces

The pavement surfaces used in the TARVIP models were old concrete and old asphalt. The pavement surface retroreflectivity matrices used for these surfaces were developed by Schnell et al. by using a portable device that acted as a goniometer in recording luminance and illuminance over a range of entrance and observation angles for a variety of pavement surfaces.(16) Old concrete and old asphalt were chosen for analysis, as new concrete and new asphalt make up a small percentage of road surfaces in the United States. New concrete and new asphalt surfaces will also have new pavement markings, which are unlikely to be the focus of scrutiny in terms of minimum retroreflectivity.

4.2 Selection of Pavement marking Configurations

Three pavement marking configurations were included in the model: a single white dashed lane line to the left of the vehicle, a single yellow dashed center line to the left of the vehicle, and a single yellow dashed center line to the left of the vehicle with a solid white edge line to the right of the vehicle. For all dashed lines, a standard 3-m (10-ft) skip line was used with a 12-m (40-ft) cycle length. The two dashed line only scenarios provide the driver with minimal pavement marking surface and are located to the left of the vehicle (opposite the typical aiming direction of most U.S. headlamps, as shown in figure 4). The use of the two colors provided useful information on the difference in driver retroreflectivity needs between white and yellow lines. Additionally, the scenario including the edge lines serves to show the additional benefit drivers obtain from a fully marked roadway.(24) Twelve-foot lane widths were chosen, as the vast majority of travel lanes in the United States are no wider than 3.7 m (12 ft). With larger lanes, the pavement markings are farther away laterally from the vehicle headlamps. Therefore, using larger lanes as a default yields a conservative visibility scenario.

4.3 Selection of Vehicle Types

Two vehicle types were included in the model: a passenger sedan and a large commercial vehicle. The dimensions used were those of the 1998 Chevrolet Lumina and the 1986 Freightliner from the TTI study.(15) Many of the vehicles in the United States are either similar to one of these vehicles or somewhere in between them in terms of driver and headlamp locations.

4.4 Selection of Operating Speeds

The TARVIP model was evaluated at three vehicle speeds: 64.4 km/h (40 mi/h), 88.5 km/h (55 mi/h) and 112.7 km/h (70 mi/h). This loosely follows the recommendations of the Turner research, which recommended one minimum retroreflectivity level for all speeds below 64.4 km/h (40 mi/h), as drivers will always need close-proximity vehicle placement information, no matter how slow they are traveling. Turner also showed that nearly 68 percent of the rural two-lane highways in the United States have a speed limit of 88.5 km/h (55 mi/h), creating another natural investigation speed. (7) Finally, 112.7 km/h (70 mi/h) was chosen as the final investigation speed as 38 of 50 U.S. states have a maximum rural interstate speed limit of 112.7 km/h (70 mi/h) or less (see figure 8). The 12 states with a speed limit of 120.7 km/h (75 mi/h) accounted for only 9.6 percent of annual vehicle miles traveled in the United States in 2003.(37, 38)

Figure 8. Bar graph/line graph. Maximum speed limits in U.S. states and associated VMT. From Insurance Institute for Highway Safety’s 2006 statistics and Federal Highway Administration’s 2003 statistics. The x-axis is labeled “Maximum Rural Interstate Speed Limit (mph)”, the y-axis for the bar graph is labeled “Number of States”, and the y-axis for the line graph is labeled “2003 VMT (Billions)”. 1 mi/h equals 1.61 km/h. There are four bars in the bar graph that represent the number of states, one at x = 60 with a y value of 1, one at x = 65 with a y value of 19, one at x = 70 with a y value of 18, and one at x = 75 with a y value of 12. There are four points in the line graph that represent vehicle miles traveled (VMT) in 2003, one at around (60, 0), one at around (65, 1,150), one at around (70, 1,400), and one at around (75, 280). This graph shows that 38 of 50 U.S. states have a maximum rural interstate speed limit of 112.7 km/h (70 mi/h) or less, and the 12 states with a speed limit of 120.75 km/h (75 mi/h) accounted for only 9.6 percent of annual vehicle miles traveled in the United States in 2003.
1 mi/h = 1.61 km/h

Figure 8. Maximum speed limits in U.S. states and associated VMT.

4.5 Consideration of Roadway Lighting

Determining the effect of roadway lighting on minimum pavement marking retroreflectivity values is desirable. TARVIP contains an ambient luminance option that allows the user to account for luminance that originates from nonheadlamp sources. However, as described in the previous section, TARVIP works by calculating the luminance contrast between the pavement marking material and the road surface. The ambient luminance option adds an equal amount of luminance to the pavement marking material and the road surface, which serves to reduce the luminance contrast ratio. Increasing the ambient luminance in TARVIP decreases pavement marking visibility when research has shown that the opposite is true.(10) Therefore, a dark roadway was used exclusively in the TARVIP model.

4.6 Selection of Pavement marking Materials

The TARVIP pavement marking described as alkyd paint and beads was selected for analysis, as Turner found that markings composed of paint with beads make up the vast majority of pavement marking material used by state and local agencies in his survey of those agencies.(7) Pennsylvania Transportation Institute researchers also surveyed nine state agencies and found that 98 percent of the lane-miles of pavement markings in those states are either water-based or epoxy-based paints.(39)

4.7 Selection of Vehicle Headlamp Performance

The headlamp used for the TARVIP analysis was the 2004 UMTRI 50 th percentile market weighted headlamp. This headlamp file comprises a luminous intensity matrix that represents the latest available market-weighted average representation of the U.S. vehicle fleet. It is a conglomerate of the 20 best-selling 2004 model year passenger vehicles in the United States, representing 39 percent of all vehicles sold in the United States. The photometric information for each headlamp was weighted according to how many vehicles of that type were sold.(23)

4.8 Establishment of Required Preview Time

COST 331 states that the absolute minimum driver preview time is 1.8 seconds and established a recommended preview time of 2.2 seconds.(34) Other research has used preview times ranging from 2.0 to as high as 3.65 seconds in recommending minimum pavement marking retroreflectivity, the latter producing relatively high RL recommendations.(1) For the purpose of this research, a preview time of 2.2 seconds was used, aligning with the value recommended and used in COST 331.

4.9 Selection of Driver Age and Visual Performance

Research sponsored by the FHWA to establish minimum in–service retroreflectivity levels for traffic signs has shown that about 90 percent of the nighttime driving population is 62 years of age or less.(21) The human factors study to support the minimum retroreflectivity levels for traffic signs used drivers aged 55 and older, with an average age of 62 years. Therefore, in order to maintain consistency with previous FHWA-sponsored work on minimum retroreflectivity, a driver age of 62 years was used in the TARVIP models. By selecting age in TARVIP, the visual abilities of the drivers are also set.

4.10 Consideration of RRPMs

Accounting for RRPMs in minimum pavement marking retroreflectivity recommendations is desirable because their superior retroreflective performance during wet night conditions can reduce the luminance required of pavement markings by drivers. However, TARVIP does not currently have a module built in to directly model RRPMs or for determining the relative efficiency of two different pavement marking materials deployed under the same scenario. Zwahlen and Schnell were the only previous researchers who have made minimum pavement marking retroreflectivity recommendations and attempted to account for RRPM presence. Their methodology involved reducing the required preview time from 3.65 seconds without RRPMs to 2.0 seconds with RRPMs. While there is little support for this value, it is still greater than the absolute minimum driver preview time established by COST 331. It was the first attempt to establish a “discount” factor for pavement marking retroreflectivity in the presence of RRPMs.(1)

The methodology developed for determining minimum pavement marking retroreflectivity when RRPMs are present does not generate minimum retroreflectivity levels for the RRPMs themselves and assumes that the RRPMs are in adequate working condition. This approach is based on the driver being able to receive enough information from the pavement markings to identify the nature of curves in the roadway and the configuration of the pavement markings. As most pavement markings become unreliable sources of driver information under wet night conditions, RRPMs are designed to aid drivers under these conditions. However, they are not continuous linear devices like pavement markings; the information provided by RRPMs is intermittent. Therefore, understanding their ability to provide advanced roadway alignment information is critical. A literature review identified work by Zwahlen and Park, which shows that drivers need a minimum of three cues to detect changes in the horizontal alignment. (40) Using this information, the researchers developed a criterion such that the recommended minimum retroreflectivity levels for pavement markings when RRPMs exist will be based on a requirement that drivers be able to detect at least three RRPMs. Note, these RRPMs do not have to be continuously spaced, but three should be visible. In other words, there is allowance for missing or damaged RRPMs as long as there are still three within view to the nighttime driver.

Because drivers also need to receive close-proximity information from pavement markings (for peripheral vision tasks such as lane keeping and regarding passing zone information on two-lane highways), a second criterion was also developed. This second criterion was established to supplement the initial criterion of having a preview time of at least 2.2 seconds. With this additional criterion, it was felt that the recommended minimum maintained pavement marking retroreflectivity levels could accommodate both the near range visibility needs of nighttime drivers as well as their long range visibility needs. The near range criterion was based on a 24.4-m (80-ft) detection distance. This distance was chosen based on the close-proximity information markings provide while considering the occluded distance caused by the hood of typical vehicles. It provides about 1 second of preview time traveling at 88.5 km/h (55 mi/h).

4.11 Determining Minimum Pavement marking Retroreflectivity

Based on the information outlined above, 48 scenarios were developed for use with TARVIP to produce an array of retroreflectivity levels that could be used in conjunction with previous research and previous recommendations to put forth an updated set of recommendations that incorporate the best-known scientific findings and expertise currently available. Thirty-six of the 48 scenarios were used to compute the required retroreflectivity of pavement markings without RRPMs. The remaining 12 scenarios included RRPMs. There were an additional 18 scenarios developed to evaluate the sensitivity of required RL to preview time.

Previous | Table of Contents | Next

Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000
Turner-Fairbank Highway Research Center | 6300 Georgetown Pike | McLean, VA | 22101