Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Research Results Digest 305 August 2006 The objective of NCHRP Project 17- 28 was to develop guidelines for use of pavement marking materials and markers based on their safety impact and cost- effectiveness. A key component of the re- search was the attempt to correlate the safety impact of pavement markings and markers with their performance, princi- pally measured in terms of their retrore- flectivity. If such a correlation were found, it would be possible to estimate cost- effectiveness and then develop the desired guidelines since retroreflectivity is a direct function of the costs of marking materials and their application. Thus, the project sought to test the null hypothesis that the safety impact of pave- ment materials and markers and their level of retroreflectivity are not significantly cor- related. If the null hypothesis were rejected, it would be possible to conclude that greater retroreflectivity translates into greater safety, thereby justifying the costs of maintaining higher levels of retroreflectivity. The project was hindered by a lack of datasets of sufficient depth and breadth to test this hypothesis. The research team per- formed a pilot analysis with the most com- prehensive dataset available (the dataset from California), and, in this instance, it found that the null hypothesis was accepted: no statistically significant relationship was found between safety and the retroreflec- tivity of pavement markings and markers. This preliminary result, if sustained, could substantially alter the frequency of application, and therefore the cost, of pave- ment markings and markers. Thus, the cost- effectiveness equation for pavement mark- ings and markers could also be altered. The remainder of this digest summa- rizes the final report of the project. It re- views the methodology used for the analy- sis of the California data, its limitations, and key findings. Though preliminary, these findings offer âfood for thoughtâ for state highway agency personnel tasked with selecting and maintaining pavement markings and markers to provide safe op- erating conditions. The full final report (available online as NCHRP Web-Only Document 92 at http://trb.org/news/blurb_ detail.asp?id=6475) offers guidance to PAVEMENT MARKING MATERIALS AND MARKERS: TESTING THE RELATIONSHIP BETWEEN RETROREFLECTIVITY AND SAFETY This digest summarizes key findings from NCHRP Project 17-28, âPavement Marking Materials and Markers: Safety Impact and Cost-Effectiveness,â conducted by iTRANS Consulting, Ltd. The digest is an abridgement of portions of the project final report authored by the principal investigator, Geni Bahar, as well as Maurice Masliah, Tara Erwin, and Errol Tan, all of iTRANS Consulting, Ltd., and Ezra Hauer. The full final report is available online as NCHRP Web-Only Document 92. Subject Area: IVB Safety and Human Performance Responsible Senior Program Officer: Edward T. Harrigan NATIONAL COOPERATIVE HIGHWAY RESEARCH PROGRAM
states that may wish to carry out similar analyses for their own conditions. BACKGROUND Longitudinal pavement markings are found on nearly all freeways and highways in the United States. Previous research has emphasized the im- portance of quantifying the impact of different pave- ment marking material types on safety, but no such quantification has yet been achieved. This study takes a different approach from previous research by focusing on quantifying the relationship be- tween retroreflectivity and safety over time, indepen- dent of the marking or marker material type. METHODOLOGY This study examined the safety effect of retrore- flectivity of longitudinal pavement markings and markers over time on non-intersection locations during non-daylight conditions. For this study, safety is defined as the number of crashes by severity per unit of time and distance. The National Transportation Product Evalua- tion Program (NTPEP), a service provided by the American Association of State Highway and Trans- portation Officials (AASHTO), collects data and evaluates pavement markings and markers (among other products) using a formal and detailed work plan. For this study, NTPEP data were assembled into a database and used to derive mathematical models of retroreflectivity performance as a function of age, color, marking material or marker type, cli- matic region, and level of snow removal. As a re- sult of this modeling, a significant contribution of this study is the generation of retroreflectivity per- formance models as a function of various factors. These models have not previously been achieved using other datasets. The models were used to es- timate the retroreflectivity of pavement markings and markers on state-maintained freeways and highways in California for 1992â1994 and 1997â2002, covering more than 5,000 miles of road segments. The innovative study approach solved for multi- pliers that represented the change in the expected number of crashes as a function of retroreflectivity. Safety effect multipliers were solved for yellow and white pavement markings separately and in com- bination, and for pavement markers for different road types and crash severity, using the retroreflec- tivity models and Californiaâs data of over 118,000 non-intersection, non-daylight (night, dawn, and dusk) recorded crashes. DISCUSSION OF RESULTS A review of the literature on the safety effect of the retroreflectivity of markings and markers leads to the following conclusions: 1. The safety effect of pavement marking and marker retroreflectivity is very hard to detect. 2. The safety effect of pavement marking and marker retroreflectivity is most likely very small. This study addressed the difficulty of measuring a hard-to-quantify safety effect by applying an inno- vative time series approach that allowed the use of historical data covering more than 118,000 crashes and 5,000 miles of highways and freeways with 8 years of known marking installation data. Safety effect multipliers were computed for different retroreflec- tivity ranges that represent markings of different brightness. The scope of this study is believed to be larger than any previous work on the safety effect of the retroreflectivity of pavement markings and markers. The size of the present study, combined with the innovative time series methodology, was used to look for a hard-to-find, overall average safety effect of retroreflectivity. The retroreflectivity safety effects of the follow- ing factors were estimated: marking and marker com- binations, road type, and crash severity. This study concludes that the difference in safety between new markings and old markings during non-daylight con- ditions on non-intersection locations is approximately zero. No measurable safety effect was ascertained on multilane freeways, multilane highways, or two-lane highways as a function of the relative retroreflectivity of either white or yellow pavement markings, or for pavement markers. The sample for pavement markers available for California was too small to be conclusive to examine combinations of markers and markings. This study did not identify any change in safety with low marking or marker retroreflectivity, nor did it identify any change in safety with bright mark- ing or marker retroreflectivity, with respect to non- daylight, non-intersection locations. The safety of pavement markings during non-daylight conditions 2
for non-intersection locations appears to be indepen- dent of whether the markings are new or deteriorated to the average level found on roads in California. According to the findings in the literature, the presence of lane lines is important. In addition, the literature clearly identifies that there is a strong driver preference for brighter pavement markings. But do brighter markingsâthat is, markings with brighter retroreflectivity than that of old markings in Californiaâlead to increases in safety? Accord- ing to the current study, they do not. Although drivers prefer higher retroreflectivity markings and markersâwhich may therefore allow them to drive more confidentlyâthe overall safety difference in the number of crashes when compared with driving with less bright markings is approximately zero. As established by several studies, when sight detection distance is reduced, as it is during non- daylight hours and adverse weather, lane control becomes more difficult and driver work load increases, causing drivers to compensate by reducing their speed. The increase in sight detection distance due to higher retroreflectivity of pavement markings and markers may cause drivers to maintain higher speeds, thereby increasing the possibility of a crash under certain geometric conditions. In other words, driver adapta- tion to road conditions may be minimizing any im- provement in safety due to greater sight detection distances from retroreflectivity markings and mark- ers. According to extensive analysis of pavement marking and marker data, roadway inventory data, and crash data, the best estimate of the joint effect of retroreflectivity and driver adaptation is approximately zero for non-intersection road segments during non- daylight hours. Questions about the validity of the study and its limitations are discussed in the following sections. HOW DO WE KNOW THAT THE METHODOLOGY IS CORRECT? The methodology for estimating maximum like- lihood simultaneously estimates the seasonal effects and the pavement marking and marker effects. âSea- sonal effectsâ refers to the variations in number of crashes that fluctuate from month to month but that repeat from year to year. The seasonal effect must be estimated in order to separate the safety effect due to markings and the effect due to the season. The seasonal effect parameters obtained in this study are very similar, even across road types. In every esti- mation of the safety effect of markings, the estimate of the seasonal effect is very reasonable: there are more crashes during the winter months and fewer crashes during the summer months. The seasonal effects of this study are reasonable because the values are very similar to other published seasonal effects. For example, the seasonal effects of this study are similar to those found by Hauer et al. (1) for data from New York State (see Table 1). How- ever, the seasonal factors by Hauer et al. are 24-hour seasonal factors, whereas this study focuses on non- daylight crashes only. The number of non-daylight hours per month in this study changes from a high 3 Table 1 New York State seasonal factors from Hauer et al. (1) and California non-daylight hours Seasonal Effect Jan. Feb. Mar. Apr. May Jun. Jul. Aug. Sep. Oct. Nov. Dec. New York seasonal factors 1.25 0.97 0.97 0.79 0.84 0.89 0.96 0.98 0.85 0.95 1.15 1.40 California non- daylight hours 14.8 13.9 12.4 11.2 10.0 9.1 9.0 9.8 11.0 13.0 13.3 14.7 California non- daylight hours/ California average non-daylight hours (CMF) 1.2 1.2 1.0 0.9 0.8 0.8 0.8 0.8 0.9 1.1 1.1 1.2 New York seasonal à California CMF 1.6 1.1 1.0 0.7 0.7 0.7 0.7 0.8 0.8 1.0 1.3 1.7 CMF = Crash modification factor.
of 14.8 hours in January to a low of 9.0 hours in July (based on sunrise and sunset times for Redding, California), as shown in the third row of Table 1. A non-daylight crash modification factor (CMF) for California may be estimated by dividing the number of non-daylight hours by the average number of non- daylight hours (11.9 hours) for Redding, California, which is shown in the fourth row of Table 1. The product of the New York seasonal factors and the California non-daylight CMF is shown in the fifth row of Table 1, and they range from a high of 1.7 to a low of 0.7. This last row of Table 1 can now be compared with the seasonal factors estimated in this study. In both cases, the magnitude of the seasonal effects are very similar and the highest number of non- daylight crashes occur during January, November, and December. WHAT IF THE RETROREFLECTIVITY MODELS ARE INACCURATE? In the second month of installation (which is the first full month of new markings, the month where the retroreflectivity effect is supposed to be greatest), there was no measurable effect on safety. Therefore, independent of what the on-the-road retroreflectivity may actually be, the safety difference between old and new markings is essentially zero. This result remains valid regardless of how accurate the retroreflectivity models may be. HOW SMALL MIGHT THE SAFETY EFFECT OF PAVEMENT MARKINGS BE? When crashes are artificially added to the data during the first full month of installation, the safety effect of markings is less than 300 crashes spread over 8 years and 1,388 miles of road on multilane free- ways. This conclusion may be drawn from Figure 55 in the full report, which shows that an additional 300 crashes occurring on roads with markings with the same retroreflectivity (markings within the same bin range) would have resulted in a safety effect of 1.05. A safety effect of 1.05 is large enough to have been detected as significant. In other words, the cur- rent study is sensitive to a difference of about 300 out of approximately 90,000 total crashes, or 0.3% sensitivity. The purpose of this sensitivity test is to demonstrate that the scope and design of this study are sufficiently large and robust for the researchers to confidently conclude that the safety effect of retro- reflectivity during non-daylight conditions on non- intersection locations is approximately zero. LIMITATIONS OF THE STUDY This study does not address the safety effect of pavement markings or markers themselves; rather, the focus has been on the safety effect of retroreflectivity. This study cannot be used to quantify the safety effect of the presence or absence of pavement markings and markers. This study also cannot be used to quan- tify the safety effect of retroreflectivity greater or less than the ranges modeled for California. There are associated limitations in defining pre- cisely what the true retroreflectivity ranges are for Cal- ifornia. The retroreflectivity values used in this study are estimates from model data based upon NTPEP test deck retroreflectivity measurements. The modeled retroreflectivity estimates have not been calibrated for California. This means that the true retroreflectivity of markings and markers in California may differ from the modeled NTPEP retroreflectivity. The applications of pavement markings and markers at NTPEP test decks may be more carefully applied than average highway installations, in which state department of transportation crews have tighter deadlines and bud- gets. However, state departments of transportation may be choosing marking and marker materials and types that perform above the average NTPEP perfor- mance. The average retroreflectivity found on NTPEP test decks may differ from the average retroreflectivity found on state roads. It is not known if the retro- reflectivity found on California highways and free- ways is higher or lower than the retroreflectivity found on NTPEP test decks. Therefore, while there is cer- tainty that the difference in safety between new mark- ings and old markings during non-daylight hours at non-intersection locations is approximately zero, there is uncertainty regarding the value of retro- reflectivity of new and old markings in California. Another limitation of this study is that very few states maintain a pavement marking management system like the one currently used in California. It may be that the very existence of a marking management system leads to an improved marking and marker program, thus causing very few roads to have relatively low levels of retroreflectivity (below the proposed FHWA minimum of â¼100 mcd/m2/lux). A pavement marking management system may be a leading factor 4
in having better-than-average pavement markings on the road. Therefore, it is possible that California is not a representative state if the condition of its markings are better than average. It may be that the absolute brightness level does not have a major effect on safety if the agency has a management system and the roads are maintained above a âminimum.â The only way to test this possibility would be to compare the results for California with the results for an agency that does a poor job of maintaining its system. Un- fortunately, such an agency, if existent, would not have the data records to conduct the current study. FINDINGS This research study investigated the safety effect of the retroreflectivity of pavement markings and markers on state-maintained multilane freeways, multilane highways, and two-lane highways in Cal- ifornia. An innovative approach was developed that analyzed historical pavement marking and marker installation data over time, thereby making use of large quantities of data that otherwise could not be analyzed using traditional before-after methods. By converting the age of pavement markings into their corresponding retroreflectivity, the study could com- pare different marking material types with one another using retroreflectivity as a common metric. This approach is based on the assumption that different pavement marking material types at the same retro- reflectivityâfor example, waterborne and thermo- plastic both at 150 mcd/m2/luxâhave the same level of safety. Safety was examined as a function of dif- ferent ranges of retroreflectivity brightness. Retroreflectivity performance of pavement mark- ings and markers was based on NTPEP data. A database was built using published NTPEP retro- reflectivity measurements, and mathematical models were built that computed retroreflectivity as a function of age, color, material type or marker type, climate region, and amount of snow removal. These retro- reflectivity models provided the average retroreflectiv- ity performance for pavement markings and markers (tables are provided in Appendix A of NCHRP Web- Only Document 92, which is available online at http:// trb.org/news/blurb_detail.asp?id=6475). These mod- els may be useful to jurisdictions seeking estimates of their pavement marking and marker retroreflectivity or for comparing the performance of new products with the average performance of a particular material type. The retroreflectivity models were applied to convert California installation date data into retro- reflectivity data. The safety effect of retroreflectivity of pavement markings (which deteriorates over time) was studied by examining the change in the number of non-intersection, non-daylight (nighttime, dawn, and dusk) crashes over a period of 8 years and over 118,000 crashes. The analysis methodology used in this study solved for multipliers representing the safety effect for different retroreflectivity ranges (i.e., bin ranges). Because a time-series approach was used, it was nec- essary to separate out the monthly seasonal effect from the cyclic pattern of pavement marking and marker installation. Multipliers for the seasonal effect show- ing higher crash counts in January, November, and December provide support for the validity of the analysis methodology. No conclusions were drawn regarding the safety effect of the retroreflectivity of pavement markers because the sample size was too small. For pavement markings or markers, the differ- ence in safety (measured during non-daylight hours at non-intersection locations) between time periods with high-retroreflectivity markings and time periods with low-retroreflectivity markings is approximately zero for roads that are maintained at the level implemented by California. Californiaâs level of maintenance appears to be frequent: pavement markings are in- stalled on high-volume highways up to three times per year with waterborne paint or every 2 years with thermoplastic markings. What appears to be important is that markings are present and visible to drivers, but what is less impor- tant with respect to safety is whether the markings are ânew marking brightâ or âold marking bright.â One hypothesis is that drivers compensate for lower visibility by reducing their speed and take advantage of higher visibility by maintaining higher speeds. Therefore, any effect of the level of brightness of pave- ment markings may be minimized by driver adaptation to road conditions. In other words, the best estimate of the joint effect of retroreflectivity and driver adapta- tion is approximately zero for non-intersection road segments during non-daylight hours. The approach used in this study was found to be reliable and straightforward to implement and is rec- ommended for safety treatments that change one way or another over time. The approach allows for max- imum inclusion of historical data and does not have the same sampling problems of traditional before-after studies. 5
REFERENCES 1. Hauer, E., Terry, D., and Griffith, M. S., âEffect of Re- surfacing on Safety of Two-Lane Rural Roads of New York State.â Transportation Research Record 1467, Transportation Research Board (1994), pp. 30â37. 2. Musick, J. V., âEffect of Pavement Edge Marking on Two-Lane Rural State Highways in Ohio.â Highway Research Board Bulletin 266, Transportation Research Board (1960), pp. 1â7. 3. Bahar, G., Mollett, C., Persaud, B., Lyon, C., Smiley, A., Smahel, T., and McGee, H., NCHRP Re- port 518: Safety Evaluation of Permanent Raised Pavement Markers, Transportation Research Board (2004). 6
Transportation Research Board 500 Fifth Street, NW Washington, DC 20001 These digests are issued in order to increase awareness of research results emanating from projects in the Cooperative Research Programs (CRP). Persons wanting to pursue the project subject matter in greater depth should contact the CRP Staff, Transportation Research Board of the National Academies, 500 Fifth Street, NW, Washington, DC 20001. COPYRIGHT PERMISSION Authors herein are responsible for the authenticity of their materials and for obtaining written permissions from publishers or persons who own the copyright to any previously published or copyrighted material used herein. Cooperative Research Programs (CRP) grants permission to reproduce material in this publication for classroom and not-for-profit purposes. Permission is given with the understanding that none of the material will be used to imply TRB, AASHTO, FAA, FHWA, FMCSA, FTA, or Transit Development Corporation endorsement of a particular product, method, or practice. It is expected that those reproducing the material in this document for educational and not-for-profit uses will give appropriate acknowledgment of the source of any reprinted or reproduced material. For other uses of the material, request permission from CRP.