Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments (2017)

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.

48 Study Objectives and Results There continues to be a problem in the United States related to the safety for pedestrians who attempt to cross streets, particularly on high-speed, high-volume, multi-lane roads. Furthermore, there is a need to better understand the safety effects of some of the more promising treatments on pedestrian crashes. The objective of this study was to develop CMFs for several different types of pedestrian treatments at unsignalized pedestrian crossings. After considering numerous treat- ment options, the four unique treatment types selected for evaluation included RRFBs, PHBs, pedestrian refuge islands, and advanced YIELD or STOP markings and signs. A total of approxi- mately 1,000 treatment and comparison sites were selected from 14 different cities throughout the United States. Most of the study sites were selected at intersections on multi-lane streets in urban and suburban areas, since these are sites with a high risk for pedestrian crashes and where counter- measures are typically most needed. For each study site, relevant data were collected regarding the treatment characteristics; traffic, geometric, and roadway variables; and the pedestrian crashes and other crash types that occurred at each site. Cross-sectional models and before/after EB analy- sis techniques were used to determine the crash effects of each treatment type. Wherever pos- sible, study sites were selected on multi-lane, high-volume roads since pedestrians are at greater risk when crossing such roads, and, therefore, the need for pedestrian crossing enhancements is greater on such roads compared to lower-volume two-lane roads. All four of the treatment types were found to be associated with reductions in pedestrian crash risk compared to the untreated sites. Pedestrian hybrid beacons (CMF of 0.453) and PHBs with advanced YIELD or STOP markings and signs (CMF of 0.432) were associated with the greatest benefit to pedestrian crash risk, followed by RRFBs (CMF of 0.526), pedestrian refuge islands (CMF of 0.685), and advanced YIELD or STOP markings and signs (0.75). CMFs for some other crash types (e.g., rear-end, sideswipe, and total crashes) were also found for some of the four pedes- trian treatments, as given in Table 4-23. As a general caution, in the application of the CMFs, users should consider the summary statistics in Chapter 4 to see how closely the site under consideration compares to sites in the research study when estimating the expected CMF for their new treatment. Specifically, most of the CMFs in this study were based on intersections (and some midblock loca- tions) in urban and suburban areas. Although most of the study sites were on multi-lane (three-, four-, and five-lane) roads, some of the treatment sites were on two-lane roads. Data Limitations and Recommendations Although several CMFs were estimated, it is important to note that the expected safety effects of applying the treatments may vary across specific applications. A great majority of the treat- ment sites selected for inclusion in this study were in urban (and suburban) areas on multi-lane Summary and Conclusions C H A P T E R 5

Summary and Conclusions 49 roads. While most of the treatments are at intersections, a substantial number of midblock loca- tions are also represented in the database. For example, it is likely that factors such as operating speed, AADT, roadway width, lane width, and others can be expected to influence the efficacy of any of these interventions. The limited data available did not allow for such a disaggregation of expected effects. Future work is needed to make more precise predictions about how much of a crash reduction should be expected at locations with specific characteristics. Therefore, this investigation should be replicated in several years after more of these treatments have been deployed in order to obtain more specific predictions about crash reductions across different types of roadway, volumes, and speeds. To this end, and given the challenges of acquiring large samples of pedestrian crash and exposure data, it is recommended that the data from this project remain accessible for future research that may also include the enhanced development of safety performance functions for pedestrians. It is also likely that these treatments may have different effects on number of crashes and the severity of crashes. For example, treatments that slow down drivers and reduce the probability of a crash may also have an effect on the severity of a crash when one occurs. Hence, it may be possible to have a CMF for total crashes and a different CMF for incapacitating and fatal crashes. Another factor that could not be fully assessed is how the use of several of these treat- ments together would influence crashes. In some cases, two CMFs may sum algebraically and in other cases they may not yield a larger effect than the more effective of the two CMFs. A much larger dataset would be required to address these questions. It should also be remembered that the CMF for the RRFB was based on a very limited sample (i.e., 50 treatment sites), and hence should be used with caution. One of the data limitations for this study and other pedestrian- and bicycle-related studies is the lack of available exposure data related to walking (and bicycling). As a result of the lack of such data from most of the 14 agencies selected for this study, it was necessary to conduct short- term (i.e., 1- or 2-hour) counts and then to extrapolate those counts to obtain an estimate of the average annual daily pedestrian traffic (i.e., pedestrian AADTs). Developing pedestrian AADTs based on the actual hourly count, the time of day of the count (e.g., 4:00 to 5:00 p.m. was the usual count time), and the type of area within the city required a substantial effort. Pedestrian counts were conducted at a great majority of the selected sites, which was quite costly. Estimat- ing pedestrian daily volume was necessary for use as a control variable in the analysis; however, the accuracy of such estimates (i.e., extrapolating short-term counts) would have been greatly improved if a larger data sample (e.g., 8- or 10-hour counts) was available from city transporta- tion agencies. Unfortunately, only two of the cities (Charlotte and St. Petersburg) had existing pedestrian counts at any of the sites selected for study. Certainly, if more city and state DOTs routinely collected pedestrian (and bicycle) counts, such count data would have value not just for research studies, but also to help determine the need and justification for all types of pedes- trian improvements on a routine basis. It is also important to remember that all four treatment types evaluated in this study did not appear in all (or even most) of the 14 test cities. For example, approximately 90 percent of the PHBs under study were found in Tucson, Arizona, a city that was found to have installed the most PHBs nationwide. Likewise, approximately 75 percent of the RRFB sites in this study were from St. Petersburg, Florida, due to the frequent use of these devices in that city. Therefore, readers should not necessarily assume that the CMFs found in this study for a given treatment will necessarily apply to usage of that treatment in all cities. Individual cities and state jurisdic- tions have differences in driver and pedestrian behaviors, terrain, laws, weather patterns, and many other factors that can affect the way that any countermeasure will perform. Therefore, it is recommended that agencies select among countermeasures with caution and try to identify and address the specific safety problem(s) at a location and then the countermeasure(s) most

50 Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments likely to address the specific problem(s). Also, after implementation of any countermeasures, it is advisable to routinely monitor crashes and behaviors at treatment sites to ensure that the countermeasure is operating effectively. Criteria for Treatment Installation The four treatments of interest are typically installed according to the following recommen- dations and criteria given below. Please note that reference to the MUTCD is based on the 2009 version, which was in effect when most of the countermeasures included in this study were installed (62). Any more current versions of the MUTCD should be considered when these criteria are updated. Rectangular Rapid Flashing Beacon Criteria for Countermeasure Installation 1. If the device is side mounted it should be located above the Pedestrian W11-2 sign placed adjacent to the crosswalk location (according to MUTCD “Interim Approval for Optional Use of Rectangular Rapid Flashing Beacons – IA-11)” (56). 2. The beacon must meet the J595 Class 1 Beacon luminosity standard. It should also meet the flash rate specified in the interim approval (56). 3. The devices evaluated in the FHWA study, Effects of Yellow Rectangular Rapid-Flashing Beacons on Yielding at Multilane Uncontrolled Crosswalks, were visible from both directions (57, 58) so consideration should be given to installing them this way on roads that serve two-way traffic. This requirement is not specified in the Notice of Proposed Amendments (for the MUTCD), although it may have influenced the outcome results of Shurbutt et al. (57) and Shurbutt and Van Houten (58) because it produced a gateway effect for approaching vehicles whose drivers could see the beacons on the opposite side of the street, increasing the visibility of the device. 4. There is also evidence that the device produces higher levels of yielding, particularly at night, when installed on the median or refuge island on roads with multiple lanes in each direction that are separated by a median or refuge island at the crosswalk (58). 5. The installation of advanced YIELD or STOP markings and signs at sites with crosswalks has been shown to decrease conflicts related to possible multiple-threat crashes on multi-lane roads (59–61). This countermeasure is designed to decrease the risk of a multiple-threat crash. These are more common when the AADT is relatively high. When used, the R1-5 or R1-5a “YIELD HERE FOR PEDESTRIANS” sign or the R1-5b or R1-5c “STOP HERE FOR PEDESTRIANS” sign should be installed adjacent to the advanced YIELD or STOP markings and signs (59–61). 6. It is also recommended that high-visibility crosswalk markings be installed at these sites (59–61). 7. Signs or foliage should not obstruct the view of the sign or beacons by drivers approaching the crosswalk. Considerations 1. Number of lanes. The more lanes that a pedestrian has to cross, the more likely it is that other vehicles may visually screen the device. Although a device mounted on a mast arm could improve visibility under this condition, data are not available on the efficacy of this type of configuration. The RRFB is ideally suited for four-lane roads with a median or refuge islands because it increases yielding right-of-way to pedestrians and is more effective with beacons on the sides and on the refuge island (58). 2. AADT. Data show that median and refuge islands are highly desirable when AADT is over 10,000. High AADT is a surrogate measure for the risk of a multiple-threat crash on multi-lane

Summary and Conclusions 51 roads, the use of advanced YIELD or STOP markings and signs is strongly recommended when vehicle AADT is high (59–61). 3. Posted speed limit. The RRFB has typically been installed on roads with speed limits between 25 and 35 mph. With higher speeds (e.g., 40 mph or higher) the use of a PHB should be considered, particularly when AADT is high (e.g., above approximately 15,000). 4. Outreach. Consider public outreach to increase driver understanding of RRFBs when RRFB devices are not in common use in an area. 5. Direction of vehicle traffic. Whether the road carries one-way or two-way traffic can influ- ence how and where the RRFB devices are implemented. Roads with one-way traffic typically are easier to cross, but multiple-threat crashes still may be present if there are multiple vehicle lanes. Pedestrian Hybrid Beacon Criteria for Countermeasure Installation 1. This device can be installed only at a marked crosswalk. The device should be mounted on a mast arm or, in the case of a four-lane road with a median island, side mounted on both sides of the approach to the crosswalk. If there is no median island or more than two lanes in each direction, the device must be mounted overhead. If the device is mounted on a span wire, an incandescent signal head should be used rather than an LED signal head. The MUTCD specifies that a “CROSSWALK STOP ON RED” (symbolic circular red, R10-23) sign shall be mounted adjacent to a PHB face on each major street approach. If an overhead PHB face is provided, the sign should be mounted adjacent to the overhead signal face (62). 2. Stop lines must be installed on each approach to a PHB. Although a specific distance is not recommended in the MUTCD, the MUTCD specifies that stop lines at midblock signalized locations should be placed at least 40 feet in advance of the nearest signal. Although no speci- fication is given for the PHB, it is recommended that the stop line be placed at a similar dis- tance to decrease the multiple-threat crash risk, based on previous research by Van Houten for Michigan DOT (70). Considerations 1. Warning signs. Whether to install a Pedestrian (W11-2) warning sign with an AHEAD (W16-9P) supplemental plaque in advance of the PHB. 2. Experimental signs. Some jurisdictions have installed an experimental “STOP ON FLASH- ING RED THEN PROCEED WHEN CLEAR” sign because of poor compliance with the flashing signal. 3. Outreach. When PHB devices are not in common use in an area, consider conducting out- reach or educational efforts to explain how to respond to the device. 4. Warning beacon. Whether a warning beacon was installed to supplement the W11-2. 5. Supporting signs. Whether the R1-6 or R1-6a sign is used with the PHB. 6. Refuge island. Because the PHB only protects pedestrians when the device is activated, consider installing a refuge island when AADT is high and pedestrian push-button compliance is low. 7. Crosswalk type. Consider high-visibility crosswalk markings to increase visibility to approaching drivers. Advanced YIELD or STOP Markings and Signs Criterion for Countermeasure Installation Advanced YIELD or STOP markings and signs may be used to indicate the point behind which vehicles are required to stop or yield in compliance with a “STOP HERE FOR PEDESTRIANS”

52 Development of Crash Modification Factors for Uncontrolled Pedestrian Crossing Treatments (R1-5b or R1-5c) sign, or “YIELD HERE TO PEDESTRIANS” (R1-5 or R1-5a) sign (62). Research on these markings at crosswalks with a multi-lane approach indicated that they worked well when placed 30 to 50 feet in advance of the crosswalk (59–61). The MUTCD specifies that these markings used at crosswalks on an uncontrolled multi-lane approach should be placed 20 to 50 feet in advance of the nearest crosswalk line, and parking should be prohibited in the area between the yield or stop line and the crosswalk (62). Considerations 1. State law stop versus yield. States or communities with laws requiring motorists to stop for pedestrians typically use stop rather than yield markings. States with laws requiring motor- ists to yield to pedestrians in crosswalks typically use yield markings. However, some states with laws requiring motorists to yield permit the use of stop bars because the statute requires motorists to “yield or stop.” Check which is the practice in your jurisdiction. 2. AADT. If AADT is high (e.g., 15,000 or above) and a median is not present, consider installing a pedestrian refuge island. 3. Crosswalk type. Consider using high-visibility crosswalk markings at sites with high AADT to increase visibility to approaching drivers. 4. Signs. Use of R1-6 or R1-6a signs should be considered if AADT is high. Pedestrian Refuge Island Criteria for Countermeasure Installation The refuge island should be raised and of sufficient width that it can serve as a place of ref- uge for pedestrians who are attempting to cross at a midblock or intersection location. Center islands allow pedestrians to find an adequate gap in one direction of traffic at a time as they are able to stop, if necessary, in the center island or median area and wait for an adequate gap in the other direction of traffic before crossing the second half of the street or highway. The minimum widths for accessible refuge islands and for design and placement of detectable warning surfaces are provided in the “Americans with Disabilities Act Accessibility Guidelines for Buildings and Facilities (ADAAG)”. Considerations 1. Length and width of the island. If pedestrians are expected to cross in large groups, consider providing adequate pedestrian storage. 2. Driver yielding behavior. If yielding levels are low, consider placing R1-6 signs on top of the island curb face in each direction (63). 3. Pedestrian line of sight. If space exists, consider placing a diagonal refuge or median crossing that orients pedestrians to face traffic as they approach the second half of the roadway. 4. Stop or yield markings. If a stop or yield marking is installed, specify the distance in advance of the crosswalk (59–61). 5. Signs. Consider whether or not to install the R1-5 or R1-5a, “YIELD HERE FOR PEDESTRIANS” sign or the R1-5b or R1-5c “STOP HERE FOR PEDESTRIANS” sign and whether or not to install the R11-2 warning sign with the advance markings (62). 6. AADT. Raised medians are particularly appropriate for streets where AADT is moderate to high (e.g., above approximately 5,000). 7. Posted speed limit. Refuge islands may be particularly effective on multi-lane roads, particu- larly where vehicle volumes exceed approximately 1,000 vehicles per day.