Disclaimer

The contents of this report reflect the views of the authors, who are responsible for the facts and the accuracy of the information presented. This document is disseminated under the sponsorship of the Department of Transportation, University Transportation Centers Program, in the interest of information exchange. The U.S. Government assumes no liability for the contents or use thereof.

Abstract

A study of collisions occurring at intersections along state routes in Utah was conducted. The number of crashes, the crash severity score and, for selected locations, the crash rate were determined and "ranked" for intersections within the Utah Department of Transportation's Region 1, Region 2, Region 3, Region 4's three districts, and the entire state. Study periods of ten years (1994-2003) and three years (2001-2003) were used. The Crash Data Delivery System (CDDS) was used to identify intersections, determine crash frequencies and severities, and find other collision-related statistics. The intersection of Bangerter Highway and 5400 South in Taylorsville experienced the largest number of crashes between 1994 and 2003 (949) and between 2001 and 2003 (296). The intersection of 700 East and 3300 South in South Salt Lake had the state's highest crash severity score between 1994 and 2003 (9,524.8); the score assigns 1,000 points to a fatal crash, 100 points to an incapacitating injury crash, and so forth. This intersection also witnessed the largest number of fatal crashes between 1994 and 2003 (5). The intersection of 31st and Wall Avenue in Ogden had the highest severity score between 2001 and 2003 (3,394.5). The intersection of Main Street and 800 East in Hyrum had the highest crash rate among intersections between two state routes, at 4.54 per million entering vehicles, based on 2001-2003 data. University Avenue and 900 North in Provo experienced the greatest number of crashes among non-signalized intersections between 1994 and 2003 (623), while Riverdale Road and Pacific Avenue in Riverdale had the highest number between 2001 and 2003 (156). The intersection of U.S. 89 and an unnamed road in Sanpete County, at milepoint 216.54, had the highest severity scores among non-signalized intersections for both the ten- and three-year study periods. Detailed examinations of individual crash sites were not performed in this study, although the report contains some general recommendations. For example, signalization may be a mitigating strategy at certain non-signalized intersections. Red light cameras, pedestrian facilities, bicycling facilities, and the removal of on-street parking are among the numerous interventions that could be considered for reducing and "calming" crashes. Additional analysis should investigate collision types by crash severity at a variety of intersections, crash rates at intersections between state and non-state routes, and variable functional influence areas. The research team considered the CDDS to be a useful tool, and encourage its continued development, including the customization of certain database search procedures.

Executive Summary

This report summarizes the findings of a study of intersection collisions along state routes in Utah. The Utah Department of Transportation's (UDOT) Crash Data Delivery System was used to identify intersections, crash frequencies and severities at the intersections, and other collision-related information. Intersections were "ranked" according to their respective number of crashes and crash severity score, the latter of which is explained in Section 2. Study periods of ten years (1994-2003) and three years (2001-2003) were used in the development of the rankings. Crash rates were ranked and computed for selected intersections, including those between two state routes, and some additional ones discussed in Section 6. Intersections experiencing recurring fatal crashes - at least two fatal crashes in the ten-year period and at least one fatal crash in the more recent three-year period - were also identified. Separate rankings were developed for Utah (i.e., the entire state), UDOT Region 1, Region 2, Region 3, and the Region 4 districts of Cedar City, Price and Richfield. Also, statewide rankings (only) were developed for non-signalized intersections.

Statewide intersections lists are found in Section 2, while regional and district lists are found in Section 3. Non-signalized intersections are discussed in Section 4. The following intersections were "ranked at the top" of their respective lists (4CC = Region 4, Cedar City District; 4P = Region 4, Price District, and 4R = Region 4, Richfield District):

The crash statistics are discussed in greater detail in the body of the document. On the statewide lists, the ranked intersections were concentrated in Regions 1, 2 and 3. Only two intersections from Region 4 appear on any of the six statewide lists. As indicated in Table 2, about 82 percent of the state's highway travel in 2003 was in Regions 1, 2 and 3. The development of separate lists for the Region 4 districts was useful, therefore, in identifying recurring intersection crash sites in those areas, and toward understanding the magnitudes of any intersection crash problems (relative to those in Regions 1-3). The appearance of an intersection on both the 10-year and 3-year list for a given study area suggested that the location had not been improved, improvements did not have the desired impacts, or growth in traffic had offset any improvements. The research team speculated that these "repeat appearance" intersections needed special attention. For non-signalized intersections with recurrent or severe crashes, a proposed mitigation was signalization. The research team cautioned, though, that other factors needed to be considered at each location as part of a traffic signal warrant study.

Intersections that appeared on all four statewide lists, all four statewide non-signalized lists, all four of the lists for any of the regions or districts, or had two or more fatal crashes between 1994 and 2003, one or more fatal crashes between 2001 and 2003 and were on two statewide lists were selected for further study. A total of 35 intersections met these criteria. The spatial dispersion of the 35 intersections was good, with six in Region 1, 14 in Region 2, six in Region 3, and nine in Region 4. Crash types (i.e., vehicle-vehicle, vehicle-pedestrian, vehicle-bicycle, etc.), crash severities (no injury, possible injury, bruises and abrasions, broken bones and bleeding wounds, fatal), vehicular involvement by intersection approach, crash rates (per million entering vehicles), collision types (i.e., head-on, rear-end, right-angle, etc.), and radius of influence were investigated for each intersection. The associated crash statistics were obtained from the CDDS for the 1994-2003 study period. None of the intersections were examined in detail, but a few general observations were made, with suggestions for further study. For example, a disproportionately large number of bicycle-motor vehicle collisions occurred at 800 North and State Street in Orem; similarly, a comparatively large number of pedestrian-vehicle collisions were observed at 700 East and 3300 South in South Salt Lake. Mitigating strategies for these and other intersections would need to consider the prevalent crash types. Collectively, about 94 percent of the crashes at the 35 study intersections involved two or more motor vehicles; the other 6 percent were single-vehicle incidents.

Perhaps the most useful next analytical step would be to tabulate crash statistics by collision type. For example, a tabulation of collision type by crash severity would isolate fatal and incapacitating injury crashes by type. It is likely that right-angle, head-on, and pedestrian-vehicle collisions would be among the most severe. It would also be useful to compute crash rates for a larger number of intersections than in this study. The Iowa DOT, for example, was using a composite score to prioritize its intersections for further analysis or mitigation. The composite represented a combination of rankings based on the number of crashes, crash "losses" (analogous to the crash severity score), and crash rates. Such composite scores could be computed for Utah's intersections. Finally, it would also be useful to consider the functional areas of influence of intersections in greater detail. The research team applied a 500-ft radius of influence to all intersections. An examination of the 35 study intersections, however, showed that a 100-ft radius was applicable to about 25 intersections, and only two of the intersections appeared to have a 500-ft radius of influence. Field investigations would be useful for confirming the locations of conflict points relative to intersection stop lines.

The research team considered the CDDS to be a useful tool in this study. Although the compilations were quite time consuming, it was estimated that a similar study without the CDDS would have accomplished about one-third as many summaries. The CDDS was particularly valuable for its most fundamental elements: intersection names and milepoints, intersection controls, crash totals, and crash severities. A few state routes were not in the CDDS intersection tools, and the research team had to "match" the milepoints of crashes with "estimated" intersection milepoints. These exercises, which typically involved scaling distances off of street maps, were quite inefficient without the CDDS. The research team's opinion is that the CDDS is a worthwhile investment and it should continue to be developed. Customization of some of the crash and intersection data search procedures, as discussed in Section 7, may be useful in further research.