Walkways and Supporting Facilities

Key Findings and Issues

The SPBP states that a 1993 study by the MAG found that a low level of pedestrian "friendliness" existed in low population density areas where pedestrian destinations were far from residential areas and where there were few or no pedestrian facilities.

Zegeer and Zegeer (1988) note that a before-after study in Japan found that pedestrian-vehicle crashes were reduced by 85 percent per grade-separated crossing structure. Moore and Older (1965) determined that the ratio of grade-separated crossing time to at-grade crossing time (R) needs to be 0.75 or less for there to be 100 percent usage of the facility. Richter and Fegan (1983) found that the elderly and disabled prefer to use the longer but safer route provided by a grade-separated pedestrian crossing.

According to the CARS data, 36.3 percent of the 818 pedestrians hit in 1999 were crossing at an unsignalized location. Of the 38 pedestrian fatalities in 1999, 47.4 percent were crossing at an unsignalized location. The CARS data do not distinguish between crossings at marked and unmarked crossings. By comparison, 15 percent of the pedestrians hit and 2.6 percent of the pedestrians killed were crossing at a signalized intersection. In addition, 10.5 percent of the pedestrians killed were crossing against a signal. The database does not indicate if any of the pedestrians hit who were crossing at a signalized intersection were disobeying the traffic controls.

Zegeer et al. (1982) found that fewer than 35 percent of pedestrians used pedestrian-actuated devices when such devices were available. The deficiencies noted in pushbutton signals included interference with coordinated signal timing systems, pedestrian impatience with the wait time, and unnecessary delays to vehicles caused by pedestrians who either cross without waiting or change their minds. Zegeer et al. (op. cit.) also found that there was no significant difference between the number of pedestrian-vehicle crashes at intersections with standard, concurrent pedestrian signal phases and intersections with no pedestrian signal indications. Abrams and Smith (1977) evaluated three types of pedestrian signal phasing. The late release of pedestrians with respect to vehicles was useful at locations with heavy right turn volumes. Scramble timing, in which there is an exclusive, pedestrian-only phase (in which pedestrians cross in all directions, including diagonally), increased delays to both vehicles and pedestrians. This type of phasing eliminates pedestrian-vehicle conflicts, however, and is the safest. The early release of pedestrians with respect to vehicles also increased pedestrian and vehicle delays, but may be effective in reducing conflicts with turning vehicles.

Zegeer et al. (1984) determined that the flashing DON'T WALK indication was not effective in warning pedestrians about turning vehicles, and the steady DON'T WALK was ineffective in controlling pedestrians during the clearance interval. Alternative indications, such as DON'T START, WALK WITH CARE, and animated "eyes" (which suggest a looking behavior to pedestrians), were all easily understood by pedestrians. Countdown pedestrian indicators (CPIs), several of which are in service in Salt Lake City, were determined to increase the functionality of pedestrian signals in Minneapolis-St. Paul from 67 percent to 75 percent. A total of 78 percent of all pedestrians preferred the CPIs to the traditional indicators. Illegal crossings dropped only slightly, from 15 percent to 13 percent, following installation of the CPIs (Farraher 2000).

For crossings at unsignalized locations, embedded, pedestrian-actuated crosswalk flashers have been observed to decrease pedestrian-vehicle conflicts. Observations of the flashers indicate that they are most effective when traffic volumes are moderate to high and pedestrian volumes are significant. One city, though, considered overhead flashers to be more effective than in-pavement flashers. One type of crosswalk flasher is activated by the presence of a pedestrian, rather than by pedestrian actuation. Pedestrians were observed to exercise greater caution when they were unaware that the crosswalk would flash (Van Winkle and Neal 2000).

In one study, pedestrian barriers led to a 20 percent reduction in pedestrian-vehicle crashes in Tokyo (Zegeer and Zegeer 1988). Barriers were considered to be most effective in downtown areas, near pedestrian overpasses, along high-speed roads with uncontrolled access, where little or no separation exists between the road and the walkway, near high-pedestrian generators, and on bridges that allow pedestrians. Barriers were considered to be minimally effective near high school and college campuses, on streets that allow curb parking, where the barriers cause sight restrictions, where pedestrian crossing points are not provided, and where motorists need access to a walkway.

In a study in London, pedestrian-vehicle crashes increased following the installation of pedestrian refuges, although vehicle-vehicle collisions decreased. To be effective in improving pedestrian safety, the refuges should be fitted with illuminated bollards, extend through the crosswalk, be "reinforced" with crosshatch markings, and have protective end treatments (Zegeer 1991). Pedestrian refuges should be avoided where the roadway alignment is poor, vehicles may be prone to hitting the island, the turning movement volume of trucks is high, and a lack of space forces the island to be narrow (Zegeer and Zegeer 1988).

Policies, Plans and Actions

The MAG's non-motorized transportation plan included urban design and development guidelines, traffic calming information, and recommendations for improvements in "walkability." The SPBP provides a number of statements advocating the development of walkways. The plan offers the assistance of UDOT to communities in assessing their walkability needs and in designing facilities. Ten UDOT policy issues for walkways are discussed in the SPBP. The plan notes that these "are to be evaluated over time for the possible adoption of policies within UDOT or for use in developing standard procedures for planning, identification of facility needs, project concept development, environmental review, design, construction, and maintenance of State transportation facilities."

As discussed earlier, the Salt Lake City Transportation Division has deployed several pedestrian safety enhancements, including pedestrian crossing flags and CPIs. The flags are available for pedestrians to carry while crossing a street at an unsignalized, midblock, marked location. Although no formal studies have been conducted, observations of the crossing flags are: motorists tend to be alert to the potential for a pedestrian upon seeing the flags posted at the curb; business and property owners are interested in maintaining the flags as part of the Adopt-a-Crosswalk Program; and the number of flag sites rapidly increased from four to 49 between January and June 2001. Flag theft has been a problem, and only 14 percent of all pedestrians crossing have been observed to use the flags. The city of Spanish Fork has started to install crossing flags (Bergenthal 2001).

Recommendations

Policy Issues: Overall

The evaluation of the ten UDOT policy issues on walkways, discussed in the SPBP, should be moved forward and, ultimately, policies should be adopted. Guidance should be provided on each issue. Where appropriate, minimum design standards and planning thresholds should be given.

Policy Issue: Grade-Separated Crossings

In the SPBP, policy issue A.02 mentions that projects should consider the potential for increasing pedestrian travel by providing grade-separated crossings. There is currently, however, no UDOT policy on grade-separated crossings. It is recommended that such a policy be developed.

The MUTCD does not offer specific criteria on grade-separated pedestrian crossings. Based on UDOT's survey of state DOTs, summarized in Cottrell (2001), grade-separated crossing guidelines vary by state; some states do not appear to have any formal guidelines. Louisiana's guidelines offer a compromise between data-intensive warrants and vague, planning-related factors; the guidelines are based on those offered in Bowman et al. (1989). The criteria are:

• Freeways (high-speed roadways): 100 pedestrians and bicycles crossing during the peak 4 hours, 7,500 through vehicles during the same 4 hours, and an AADT of 25,000.
• Arterials: 300 pedestrians and bicycles crossing during the peak 4 hours, 10,000 through vehicles during the same 4 hours, and an AADT of 35,000.
• Caveat: The volumes apply to the completion year of a grade-separated project.
• Exceptions: One of the three criteria is greatly exceeded, a high percentage of pedestrians or bicyclists are children, a nearby, attractive crossing opportunity negates the need for a grade-separated facility.

To ensure that a grade-separated pedestrian facility is used, and that pedestrians do not continue to cross at-grade, it is important to plan the crossing such that R, defined above, is less than or equal to 0.75. A pedestrian origin-destination study may be needed to determine the optimal location of the crossing. If the R threshold cannot be attained, pedestrian barriers may be needed to prevent or at least discourage at-grade crossings. The needs of the elderly, disabled, and visually impaired must be considered when designing grade-separated pedestrian crossings given these groups' preference for such facilities. Among the considerations are the usage of stairs, ramp gradients, ramp configurations, the slip resistance of walkway surfaces, surface textures, and the locations of landings and rest areas (along the way). Based on the survey of state DOTs, the costs of grade-separated pedestrian crossings are variable, based on the specific site conditions, length, width, height above the road, materials, and so forth. The mean cost of a pedestrian bridge appears to be about $1.2 million.

Policy Issue: Crosswalks at Unsignalized Locations

UDOT policy issue A.02 also mentions the placement of crosswalks at locations at which there is no traffic signal. It is indicated that such placement "should consider pedestrian safety and convenience." It is recommended that the placement of midblock crossings be given greater, more detailed attention. A policy should include firm guidance on when and where midblock crossings should be used, and whether they should be at-grade or grade-separated. Crosswalk flashers, either overhead or embedded, appear to be associated with a high rate of driver compliance. Pedestrian compliance with the flashers - that is, the extent to which a pedestrian will use the crosswalk - depends on crosswalk placement and traffic volume levels. Pedestrian warning flashers appear to be most effective when they are supplemented with bright, luminescent signs both at and in advance of the crossing. Zebra crosswalks (crosswalks enhanced with approach markings, hatched striping, and signs) appear to be effective in improving pedestrian safety. The literature offers conflicting findings on the effect of marked crosswalks. Some authors suggest that pedestrians are less vigilant when "protected" by a marked crosswalk. Other authors have found that pedestrians increase their looking behavior when in a crosswalk. The resolution of these conflicting findings might come through study of the behavior of pedestrians in Utah. The results might indicate how effective marked crossings are, as well as how various enhancements can improve the safety of the crossings. There are also conflicting findings in the literature on the effect of marked crosswalks on drivers. Some drivers tend to slow upon seeing a crosswalk - regardless of the presence of a pedestrian - while others react only upon seeing a pedestrian. It may be useful to study the speeds of drivers in the vicinity of marked crossings. Such studies may be most beneficial in areas where there are pedestrian safety problems.

The pedestrian crossing flag installations in Salt Lake City are popular, although their impact must be assessed. Despite the lack of an assessment, it is recommended that such installations continue with possible expansion into other cities. An expansion would be sensible because the flags are a low-cost safety improvement, and private support from adjacent property owners can be sought. An evaluation of the impact of the flags on pedestrian safety should be conducted. Issues that need to be addressed include the effect of the flags on pedestrian-vehicle collisions, choosing the best sites for the flags, the optimal size, weight and coloring of the flags, how to reduce flag theft, and instructions for proper use of the flags. The evaluation should also determine if the flags induce a change in pedestrian vigilance.

Policy Issue: Sidewalk Provision

UDOT policy issues A.04, A.05, A.08 and A.09 all refer to the provision of sidewalks. An additional concern is the maintenance and upkeep of sidewalks. Research has found, for example, that sidewalks placed on a non-uniformly compacted subgrade are subject to "pop-up" (where the center of the sidewalk rises), "sagging" (where the center of the sidewalk falls), cracking, and spalling. Freezing and thawing and tree roots can also produce such effects. Sidewalks in such condition can be somewhat hazardous to pedestrians. Corrective maintenance strategies, such as replacing sidewalk sections that are in poor condition, should be established. Routine maintenance strategies, such as cleaning and snow removal, should also be in place.

Other Issues: Accommodating Pedestrians at Traffic Signals

The following recommendations are offered in response to the preceding findings and issues regarding pedestrian signals:

• Exclusive pedestrian signal phasing seems to be of greatest use only where there are high pedestrian and traffic volumes. Otherwise, the special phasing does not have an effect on pedestrian safety. This finding should be considered during the development of traffic control-related pedestrian safety strategies.
• An alternative to concurrent pedestrian-vehicle phasing might be needed in locations where traffic is heavy but pedestrian volumes are light. Studies have shown that such locations are particularly dangerous for pedestrians. Traffic signal phasing improvements may be needed for these locations. A pedestrian-actuated early-release phase would be one approach.
• Pedestrian conflicts with left-turning vehicles at "T" intersections are typically high during the early part of the green phase, and low during the remainder of the green. At four-legged intersections, the conflicts are low during the first part of the green, but are high for the latter half. Possible signalization strategies include the late release of pedestrians for "T" intersections, and early release or pedestrian-only phases for four-legged intersections.
• The MUTCD offers criteria for pedestrian signals. Based on an article by Robertson (1984) and others, the criteria can be disputed. The argument opposing widespread usage of pedestrian signals is that there is no strong evidence that such signals improve pedestrian safety. It is clear, however, that pedestrian signals are needed in many situations.
• Pedestrian startup times and volumes should be considered in pedestrian signal timing. Design startup times range from 3 seconds for the young to 3.75 seconds for the elderly.
• Signal timing at locations with heavy pedestrian volumes should be carefully developed. The size of a pedestrian "herd" can affect the amount of time required for a pack of walkers to cross a street. The "herd size" can also affect the amount of mid-crossing refuge space needed by pedestrians.
• Overall pedestrian and vehicle delays are minimized with standard, concurrent pedestrian-vehicle signal timing. Alternative phasing schemes may be needed, to serve certain pedestrian and vehicle movements. Overall delay should not serve as the sole measure of performance.
• If there is a heavy right-turn volume, the late release of pedestrians is a useful strategy for reducing the right-turn queue length. Signing is needed to inform both drivers and pedestrians of the signal phasing scheme.
• The safest signalization strategy for pedestrians is scramble timing, which provides pedestrians with an exclusive phase. Pedestrians can cross in all directions, including diagonally - during the scramble phase. Delays to both pedestrians and vehicles are high with scramble timing, but pedestrian safety is enhanced.
• To increase pedestrian compliance with pedestrian pushbuttons, response time should be quick - preferably immediate.
• Certain pedestrian signal display enhancements seem to work well, including DON'T START, WALK WITH CARE, countdown pedestrian indications, and animated eyes.
• Some of the traditional pedestrian signal indications, including both the steady and flashing DON'T WALK displays, offer little information and are frequently confusing to pedestrians.
• To facilitate the placement and timing of pedestrian signals and other pedestrian facilities, the following need to be determined: the portion of pedestrians who are elderly, the portion of pedestrians who are children (under age 16 and under age 10), primary pedestrian origin-destination pairs (particularly for the placement of midblock and grade-separated crossings), and pedestrian volumes.
• On wide streets, which are quite commonplace in Utah, pedestrian signals that enable full, nonstop crossing should be considered. On busy streets, the red intervals of the crossing approaches can be offset to allow one approach to continue to flow while the pedestrian is crossing the other approach (Urbanik et al. 2000).
• If an adequate green time for full crossing of a wide street cannot be provided, then a pedestrian refuge island should be installed. The island should both meet minimum design standards and deflect vehicle impacts.