Idaho 2006 Five Percent Report
This report is in response to the Federal requirement that each state describe at least 5 percent of its locations currently exhibiting the most severe highway safety needs, in accordance with Sections 148(c)(1)(D) and 148(g)(3)(A), of Title 23, United States Code. Each state’s report is to include potential remedies to the hazardous locations identified; estimated costs of the remedies; and impediments to implementation of the remedies other than costs. The reports included on this Web site represent a variety of methods utilized and various degrees of road coverage. Therefore, this report cannot be compared with the other reports included on this Web site.
Protection from Discovery and Admission into Evidence—Under 23 U.S.C. 148(g)(4) information collected or compiled for any purpose directly relating to this report shall not be subject to discovery or admitted into evidence in a Federal or State court proceeding or considered for other purposes in any action for damages arising from any occurrence at a location identified or addressed in the reports. |
Additional information, including the specific legislative requirements, can be found in the guidance provided by the Federal Highway Administration,
http://safety.fhwa.dot.gov/safetealu/fiveguidance.htm.
Idaho's Highway Segments and Intersections Exhibiting the Most Severe Safety Needs Idaho Transportation Department – 2006
Idaho's examination of the top five percent of locations with the most severe safety needs yielded 30 locations, including intersections, interstate segments, and non-interstate segments. A table of these locations with descriptive data is included attachment A. A map of these locations is included in attachment B.
The following includes additional information requested, as per FHWA HSIP "5% Report" Guidance, April 10, 2006.
Description of Methodology Used
The top five percent of the most severe of Idaho's highway segments and intersections was determined using an existing methodology that is used annually to generate reports of Idaho's high accident locations. This report uses the top 10 locations from the intersection, interstate segments, and non-interstate segments reports. Attachment C is a copy of this methodology.
Extent of Public Road Coverage
This list does not include any data for local public roads. Currently, the Idaho Transportation Department (ITD) has crash data for all public roads, including roads on and off the state highway system. However, ITD only maintains volume data for state highway system roads. Because the methodology relies on volume data for rates and calculating volume groups for roads, local roads are not included in this list.
ITD is taking steps toward obtaining data for local roads, including a local roads base map for crash locating and mapping purposes, and volume data. Initial contacts have been made with potential partners to develop or collect this data. Future steps must be taken to begin to schedule upgrading to full coverage for local road data.
Years Used in Data Analysis
The methodology uses three years of data. For this report data from 2002, 2003, and 2004 is used to create the 2004 rankings. See attachment C.
State Contact Person/Office
Brook Smith, Research Analyst Principal
Office of Highway Safety
(208) 334-8105
Appendices
Appendix A: Idaho's Highway Segments and Intersections Exhibiting the Most Severe Safety Needs.
| District | Type of Location |
Location | Segment Code & Milepost |
Description of Crash Problem |
Previous Behavioral Remedies |
Potential Remedies |
Estimated Costs |
Implementation Impediments |
|---|---|---|---|---|---|---|---|---|
| 1 | Interstate | 90 | 001660 MP 29.700 - 30.210 | There were 25 crashes between 2004-2002, with aggressive driving and inattention driver behaviors. The most frequently occurring most harmful events during this time were overturn, median barrier, and side swipe same. | - | |||
| 1 | Interstate | 90 | 001660 MP 12.790 - 13.030 | There were 13 crashes between 2004-2002, with aggressive driving and inattention driver behaviors. The most frequently occurring most harmful events during this time were rear end and median barrier. | FFY04 grant for impaired driving equipment purchase and enforcement. | |||
| 1 | Non-interstate | 95 | 001540 MP 406.808 - 407.128 | There were 16 crashes between 2004-2002, with aggressive driving driver behaviors. The most frequently occurring most harmful event was overturn. | - | |||
| 1 | Non-interstate | 95 | 001540 MP 410.880 - 411.200 | There were 11 crashes between 2004-2002, with inattention and aggressive driving behaviors. The most frequently occurring most harmful events during this time were head on and overturn. | - | |||
| 3 | Intersection | Eagle Rd @ Fairview Ave -- Meridian, Ada county | 002000 MP 71.593 @ 002005 MP 13.095 | There were 151 crashes between 2004-2002, with aggressive driving and inattention driver behaviors. The most frequently occurring most harmful events during this time were rear end and rear end turning. | Grant FFY04 for aggressive driving, including equipment purchases, enforcement, and education. | |||
| 3 | Intersection | Linder Rd @ US 20 -- Eagle, Ada county | 002070 MP 37.258 @ 002570 MP .000 | There were 42 crashes between 2004-2002, with aggressive driving and inattention driver behaviors. The most frequently occurring most harmful event during this time was rear end. | Grant FFY04 for aggressive driving, including equipment purchases, enforcement, and education. FFY04 grant for impaired driving, including enforcement and educational materials. | |||
| 3 | Intersection | Eagle Rd @ SH 44 -- Eagle, Ada county | 002005 MP 18.004 | There were 60 crashes between 2004-2002, aggressive driving and inattention driver behaviors. The most frequently occurring most harmful events during this time were rear end, rear end turning, and side swipe same. | Grant FFY04 for aggressive driving, including equipment purchases, enforcement, and education. FFY04 grant for impaired driving, including enforcement and educational materials. | |||
| 3 | Intersection | SH 44 @ State St -- Eagle, Ada county | 002130 MP 12.320 @ 016714 MP .000 | There were 29 crashes between 2004-2002, with aggressive driving and inattention driver behaviors. The most frequently occurring most harmful events during this time were rear end and angle turning. | Grant FFY04 for aggressive driving, including equipment purchases, enforcement, and education. | |||
| 3 | Intersection | Deer Flat Rd @ SH 69 -- Kuna, Ada county | 002150 MP 3.168 | There were 26 crashes between 2004-2002, with aggressive driving and inattention driver behaviors. The most frequently occurring most harmful events during this time were angle and head-on turning. | Grants FFY04 for aggressive driving, including equipment purchases, enforcement, and education. Grant FFY04 for safety restraint education and enforcement. | |||
| 3 | Intersection | Chinden Blvd @ Eagle Rd -- Boise, Ada county | 002005 MP 16.093 @ 002070 MP 40.229 | There were 66 crashes between 2004-2002, with aggressive driving and inattention driver behaviors. The most frequently occurring most harmful event was rear end. | Grant FFY04 for aggressive driving, including equipment purchases, enforcement, and education. | |||
| 3 | Intersection | Linder Rd @ SH 44 -- Eagle, Ada county | 002130 MP 14.564 @ 002570 MP 2.170 | There were 26 crashes between 2004-2002, with aggressive driving and inattention driver behaviors. The most frequently occurring most harmful event during this time was rear end. | Grant FFY04 for aggressive driving, including equipment purchases, enforcement, and education. FFY04 grant for impaired driving, including enforcement and educational materials. | |||
| 3 | Intersection | 11th South Ave @ 2nd South St -- Nampa, Canyon county | 002040 MP 58.746 @ 002042 MP 58.670 @ 002160 MP 27.725 | There were 113 crashes between 2004-2002, with aggressive driving and inattention driver behaviors. The most frequently occurring most harmful event during this time was angle, same direction turning, and head-on turning. | - | |||
| 3 | Intersection | SH 16 @ Substation Rd -- Emmett, Gem county | 001390 MP 12.561 @ 005590 MP 1.780 | There were 22 crashes between 2004-2002, with aggressive driving and inattention driver behaviors. The most frequently occurring most harmful event during this time was angle turning. | - | |||
| 3 | Interstate | 84 | 001010 MP 44.300 - 45.190 | There were 73 crashes between 2004-2002, with aggressive driving and inattention driver behaviors. The most frequently occurring most harmful event during this time was rear end. | Grant FFY04 for aggressive driving, including equipment purchases, enforcement, and education. FFY04 grant for impaired driving, including enforcement and educational materials. | |||
| 3 | Interstate | 84 | 001010 MP 38.095 - 38.645 | There were 39 crashes between 2004-2002, with aggressive driving, inattention, and impaired driver behaviors. The most frequently occurring most harmful event during this time was rear end. | - | |||
| 3 | Interstate | 84 | 001010 MP 118.895 - 119.095 | There were 7 crashes between 2004-2002, with aggressive driving driver behaviors. The most frequently occurring most harmful event during this time was overturn. | - | |||
| 3 | Non-interstate | 20 | 002070 MP 40.165 - 40.409 | There were 45 crashes between 2004-2002, with aggressive driving and inattention driver behaviors. The most frequently occurring most harmful events during this time were head-on turning and angle turning. | Grants FFY04 for aggressive driving, including equipment purchases, enforcement, and education. | |||
| 3 | Non-interstate | 55 | 002005 MP 15.884 - 16.093 | There were 35 crashes between 2004-2002, with aggressive driving and inattention driver behaviors. The most frequently occurring most harmful event during this time was angle turning. | Grants FFY04 for aggressive driving, including equipment purchases, enforcement, and education. | |||
| 3 | Non-interstate | 84 B | 002040 MP 55.943 - 56.163 | There were 44 crashes between 2004-2002, with aggressive driving and sleepy/drowsy driver behaviors. The most frequently occurring most harmful events during this time were angle turning, head-on turning, and rear end. | - | |||
| 3 | Non-interstate | 55 | 001990 MP 15.628 - 15.818 | There were 27 crashes between 2004-2002, with aggressive driving and distracted driving behaviors. The most frequently occurring most harmful events during this time were rear end and angle turning. | - | |||
| 3 | Non-interstate | 55 | 001990 MP 16.009 - 16.180 | There were 23 crashes between 2004-2002, with aggressive driving and sleepy/drowsy driver behaviors. The most frequently occurring most harmful events were head-on turning and rear end. | - | |||
| 3 | Non-interstate | 55 | 001990 MP 86.520 - 87.100 | There were 19 crashes between 2004-2002, with aggressive driving and inattention driving behaviors. The most frequently occurring most harmful event during this time was overturn. | - | |||
| 3 | Non-interstate | 55 | 001990 MP 93.900 - 94.250 | There were 11 crashes between 2004-2002, with aggressive driving and inattention driver behaviors. The most frequently occurring most harmful events during this time were head on and over turn. | - | |||
| 4 | Interstate | 84 | 001010 MP 256.495 - 257.625 | There were 33 crashes between 2004-2002, with aggressive driving driver behaviors. The most frequently occurring most harmful event during this time was overturn. | FFY04 grant for bicycle and pedestrian safety, through data analysis, observation surveys, development of safe routes, and public education. | |||
| 5 | Interstate | 15 | 001330 MP 79.000 - 79.130 | There were 5 crashes between 2004-2002, with aggressive driving and sleepy/drowsy driver behaviors. The most frequently occurring most harmful event during this time was overturn. | - | |||
| 5 | Interstate | 15 | 001330 MP 73.700 - 73.960 | There were 9 crashes between 2004-2002, with aggressive driving driver behaviors. The most frequently occurring most harmful event during this time period was overturn. | - | |||
| 5 | Interstate | 15 | 001330 MP 94.381 - 94.631 | There were 10 crashes between 2004-2002, with aggressive driving driver behaviors. The most frequently occurring most harmful events during this time were rear end and side swipe. | - | |||
| 5 | Interstate | 15 | 001330 MP 88.693 - 88.893 | There were 6 crashes between 2004-2002, with aggressive driving driver behaviors. The most frequently occurring most harmful events during this time were rear end, side swipe same, and angle. | - | |||
| 5 | Non-interstate | 15 B | 001360 MP 3.590 - 4.630 | There were 172 crashes between 2004-2002, with aggressive driving, inattention, alcohol impaired, and sleepy/drowsy driver behaviors. The most frequently occurring most harmful event during this time was rear-end. | - | |||
| 6 | Intersection | Hitt Rd @ US 20 -- Idaho Falls, Bonneville county | 002070 MP 313.391 | There were 34 crashes between 2004-2002, with aggressive driving and sleepy/drowsy driver behaviors. The most frequently occurring most harmful events during this time were rear end, angle, and head-on turning. | - |
Appendix B: Map of Idaho's Highway Segments and Intersections Exhibiting the Most Severe Safety Needs.
Appendix C: High Accident Location Report Methodology, 2001.
SECTION I. INTRODUCTION
HISTORY
Time Line
High Accident Location (HAL) identification and analysis has traditionally been an important tool in managing safety on Idaho's highways. Beginning in 1986, various combinations of computer and manual calculations have been used to identify and rank locations for HAL listings. Annual HAL listings were used to identify unusual collision patterns and evaluate potential construction projects.
The original HAL computer program became operational in 1986 and functioned until 1991. The HAL was a software program, which retrieved information from several databases to create a Roadway/Accident/Environment (RAE) file. The program then used the RAE file to identify and rank high accident sections and spot locations.
In 1991, one of the databases used to create the RAE file changed format, making updated roadway information inaccessible to HAL. The Traffic section continued to produce HAL listings using updated collision data combined with 1991 roadway data until 1994 when the collision records format was changed – making collision data inaccessible to the HAL program.
From 1994 through 1998, the Traffic section manually calculated intersection HAL listings using data from the Crash Analysis Reporting System (CARS) and the MACS/ROSE database. Manual calculations proved to be very time consuming, leading to efforts to restore a functional HAL program as early as 1995. Financial and time constraints delayed the production of a new HAL program until 1999.
Methodological Problems
The old HAL program was never beta tested so several problems encountered in running and utilizing the program were never addressed, corrected or updated. Furthermore, a lack of resources prevented any attempt to make major corrections or modifications. During the design of the new HAL program several problems with the old HAL were addressed. The two most prominent problems that were considered are listed below and discussed in the following paragraphs.
- Locations with fatalities were over represented through economic loss ranking.
- Road type sections varied greatly in length; favoring shorter sections.
Perhaps the largest problem experienced with the old HAL program was that locations with fatalities were over represented through economic loss ranking. The program used National Safety Council cost estimates to quantify the economic loss incurred from collisions. Each fatality, A injury, B injury, and C Injury were assigned a set dollar value. The value for each of fatalities and injuries were summed with the amount of property damage to obtain a total cost for each collision. Finally, the economic losses of all the collisions occurring at a location were summed to obtain the total economic loss at a location – locations were then ranked by economic loss.
This approach was sensitive to the number and type of injuries that occurred in collisions, especially fatalities because the cost of a fatality was quite high relative to the cost of other injuries. The appearance of a single fatality or multiple fatalities in one accident often gave a location more priority than others – essentially making the high accident listing a list of locations that had experienced fatalities. Since fatalities are such rare events, and are often times unexplainable or unrelated to roadway deficiencies the influence on the HAL rankings was a disadvantage.
The second difficulty with the old HAL program involved the number and length of road type categories. The old HAL program used multiple physical characteristics (width, # of lanes, traffic volume, etc...) to classify roadways into road types. Using multiple physical characteristics to classify road types created several problems in the old HAL program and its output. First, using multiple criteria created numerous road types (78) some of which only contained a few miles of roads. The collision rate ranking favored these smaller road type categories. Second, the collision rate listing only compared roads within the same type. For example, narrow, high volume, two lane roads were only compared to other narrow, high volume, two lane roads so only the worst of these roads ranked highly – when in reality all of the roads of this road type are dangerous relative to other road types. The opposite was also true – low volume, access controlled, divided highways were only compared to other roads with the same characteristics. The HAL program highlighted the worst of these roads and caused elevated rankings, when in reality these are some the safest roads in the state.
The new HAL program methodology was developed with these problems in mind. In part, the rest of this document describes how these problems were addressed.
THE NEW HAL PROGRAM
Objectives
- To identify locations on the State Highway System with potential safety deficiencies.
- To systematically compare problem locations on a statewide basis.
- To minimize the probability of identifying spurious problem areas.
HAL Listings and Their Use
The HAL program will produce several reports annually. The main reports are the interstate segment report, the interstate-interchange report, the non-interstate segment report, and the non-interstate-intersection report.
Figure 1. The Hal Reports
The appearance of a location on a HAL listing does not conclusively define the location as a problem. It should merely be considered a pointer to possible problem locations. Only a technical safety analysis of each section will determine whether it actually is a problem location and what priority the location should be given for improvements.
The remainder of this document is dedicated to describing the ranking of locations on the State Highway System. First, in Section II the methods the new HAL program uses to select collisions for each report is discussed. Next, in section III, the ranking criteria and their application are described.
SECTION II.SELECTION OF COLLISIONS
The HAL program uses two separate methodologies: one to identify problem intersections and interchanges, and another to identify problem road sections. The HAL program uses collisions designated as intersection related to analyze intersections. To identify dangerous roadway sections the HAL program uses non-intersection related collisions in a clustering process to identify locations that are prone to non-intersection related collisions.
The HAL program analyses all collisions in which an injury, or property damage greater that $750 occurs. The HAL program only includes collisions which occur on roads with segment codes and mile posts on the State Highway System. Roads under local jurisdiction may be considered by the HAL program in the future as segment codes, mileposts, and traffic volumes become more universally available on those roads.
Each of the four HAL reports is created using separate subsets of collision data. The following paragraphs describe how collisions are selected for use in the creation of HAL reports.
INTERSECTIONS
When law enforcement officers complete the Idaho Collision Report Form they are required to determine the relationship of the collision to the nearest junction for each vehicle involved. If any of the vehicles involved in a collision have a designation of intersection or intersection related the collision is considered an intersection collision and is selected by HAL for inclusion in the intersection report calculations. Once collisions are used in the Intersection Report, these crashes are not used in any of the other reports.
Figure 2 displays the section of the Idaho Collision Report Form where law enforcement officers indicate the relationship to junction. Appendix A is the Idaho Collision Report Form, the relationship to junction box is on page three
Figure 2. Idaho Collision Report Form – Relationship to Junction Box.
INTERCHANGES
The HAL program follows the American National Standard (ANSI D16.1-1996) as a guideline to determine which collisions are interchange related. According to the American National Standard:
2.7.7 Interchange Accident
An interchange accident is a traffic accident in which the first harmful event occurs within the boundaries which include all ramps or auxiliary roadways and include each roadway entering or leaving the interchange to a point 30 meters (100 feet) beyond the gore curb or curb return at the outermost ramp connection. Interchange accidents may include at-intersection accidents, intersection related accidents, driveway access accidents or nonjunction accidents. See Figure 3.
Figure 3. Collisions which occur within the shaded area are interchange collisions.
The HAL program uses all of the collisions which occur in the shaded area of figure 3 to create the interchange report. Collisions used by the HAL program to create the interchange report are not used in any of the other reports.
ROAD SEGMENT REPORTS (Clusters)
Collisions that are not related to intersections or interchanges are used to create the interstate and non-interstate cluster reports. Collisions occurring on the interstate highway system are utilized by the interstate cluster report, whereas collisions occurring on all other roads on the State Highway System are included in the non-interstate cluster report.
While the HAL program applies ranking criteria to all the intersections on the state road system (where collisions occurred), ranking criteria are only applied to selected roadway sections called clusters. A cluster is a section of roadway defined by a high frequency of collisions per mile relative to the surrounding roadway. Clusters are not automatically high accident locations, they are simply sites that will be analyzed by ranking criteria.
The HAL clustering process is easiest to describe and understand using a graphical approach. While the following steps describe the clustering process graphically, the actual HAL program processes data in a mathematical fashion. The HAL program user will not be able to view the clustering process as it is described here, rather the output will be in report form. That said, the following steps describe the clustering process:
| Step One – | All collision locations are assigned a tolerance factor. | |||||||||||||||
| Law Enforcement Officers record the location of collisions on the Idaho Collision Report form. For collisions occurring outside of intersections officers estimate the distance from the nearest intersection or milepost marker. Since the distance estimates made by officers are not always accurate they are assigned tolerances. Tolerances judge the accuracy of distance estimates and assign a ‘cushion' to the estimate, designating an area within which the collision is believed to have occurred. | ||||||||||||||||
| Data technicians at the Office of Highway Safety assign tolerances based on the following guidelines: | ||||||||||||||||
| ||||||||||||||||
| Step Two – | Collision locations are described using a point distribution. | |||||||||||||||
| The HAL program assigns points to each collision occurring during the study period. Each collision receives 100 points that are assigned to an area of road. The size of the area of road depends on the tolerance assigned to the collision location. For example, if the tolerance assigned to a collision location was .05 then the 100 points for that collision would be distributed evenly over the .05 miles on either side of that location. The result is a block of points over a stretch of road. Collisions that are assigned small tolerances have taller blocks. If points from more than one collision overlap, the points stack on top of each other, creating taller columns. The point assigning process results in ‘bar charts´ representing all of the roadways on the State Highway System. Figure four is an example of a road section to which points have been assigned. | ||||||||||||||||
Figure 4. Roadway ‘bar chart´ showing average point values and cluster locations.  | ||||||||||||||||
| Step Three – | The average point value is calculated for each volume range. | |||||||||||||||
| The HAL program computes the average point value for each of the volume range categories (explained in section III) on the state system. The green line represents the average point value in figure 4. | ||||||||||||||||
| Step Four – | The clustering process begins. The HAL program identifies locations with the highest point values. | |||||||||||||||
| The Hal Program uses a ‘window´ through which it views the roadway ‘bar chart´ 1/10 mile at a time. The window slides along the roadway analyzing the quantity of points as it goes. As the window slides along the chart it calculates the average point value for the 1/10 mile section of roadway currently displayed in the window. The average value of the points displayed in the window is plotted in a linear fashion as the window moves along the roadway chart. The red line in figure four represents the average number of points displayed in the window as it slides down the roadway. | ||||||||||||||||
| Step Five – | The clusters are identified – and grown. | |||||||||||||||
| The HAL program identifies the location of clusters by analyzing the relationship between the average number of points in a window (the red line) and the average number of points in a volume range (the green line). When the average number of points in the window exceeds the average number of points in a volume range by two standard deviations a cluster is started. | ||||||||||||||||
| After a cluster is started, the window moves by 1/100 mile in alternating directions. As the window moves in a direction, the HAL program re-evaluates the relationship between the average number of points in the window, and the average number of points for that volume range. If the difference between these two averages drops below 1.5 standard deviations the growth of the cluster stops. On the other hand, if the difference stays above 1.5 standard deviations the window moves a 1/100 mile in the opposite direction. Again, the HAL program re-evaluates the relationship between the red and green lines – as described above. The window moves back and forth growing the cluster until the ratio between the lines drops below the 1.5 standard deviations. | ||||||||||||||||
Clusters are not automatically high accident locations, they are simply sites that will be analyzed by ranking criteria. To be considered for analysis a cluster must average more than one collision per year over the three year study period. By using a minimum crash threshold the HAL program seeks to eliminate clusters created by one or two fluke collisions.
SECTION III.RANKING OF LOCATIONS
CRITERIA
All of the HAL reports employ the same ranking criteria with only minor variations. The following concepts have been selected for inclusion in the HAL program as the result of extensive research, peer review and technical consulting.
The position of a location in the HAL listing is determined by it's statewide ranking in three categories:
- Collision frequency – Locations that experience more crashes are ranked higher than locations that experience fewer crashes.
- Severity (Economic Loss) – Locations characterized by crashes of greater injury severity and cost to society are ranked higher than locations with less crash severity.
- Collision rate – Locations which have a tendency to experience more collisions than expected for the amount of vehicle travel are ranked higher than locations which do not.
The HAL listing combines the results of the collision frequency, severity, and collision rate rankings into a single report. Each of the three rankings described above is weighted, and for each location the weighted rankings are summed giving each location a weighted score. Locations are then listed in ascending order by the weighted scores.
RANKING
Locations are ranked within each criteria based on the values assigned by the HAL program. The highest rankings (1, 2, 3…) are assigned to locations with the highest values in the criteria. Correspondingly, locations with the lowest values receive low rankings (200, 201, 202…). If two or more locations have the same value for a given criteria the HAL program scores the locations as tied. Locations that tie are given the same numerical ranking based on the following formula:
where r is the rank the locations would receive if ties had not been used and n is the number of locations in the tie.
The rank of the next non-tied location would be:
Example. If three locations experience the same frequency of collisions and a tie system is not used they would be ranked as follows:
| Location | # Collisions | Rank |
|---|---|---|
| A | 25 | 1 |
| B | 23 | 2 |
| C | 23 | 3 |
| D | 23 | 4 |
| E | 22 | 5 |
Using the HAL tie system the locations are ranked as follows:
| Location | # Collisions | Rank |
|---|---|---|
| A | 25 | 1 |
| B | 23 | 2 |
| C | 23 | 3 |
| D | 23 | 4 |
| E | 22 | 5 |
RANKING METHODS
Collision Frequency
The collision frequency ranking uses records from the state collision database to list locations in a "worst-first" format. Locations with a greater number of collisions rank higher than those with a lower number of collisions.
The collision frequency listing does not take into account the differing volumes of traffic at each location. Therefore, this listing tends to rank high volume urban sites as high accident locations, even if those locations have a relatively low number of collisions for the traffic volume. To avoid bias towards urban locations collision frequency rankings are not considered alone by the HAL program, but combined with severity (event cost) and collision rate rankings to identify High Accident Locations.
Severity (Event Cost)
The new methodology is designed to better predict where severe crashes will happen in the future by identifying what causes these crashes. For example, the speed of the vehicles involved has a tremendous impact on the severity of the injuries, as does the most harmful event, the most harmful event is the event that caused the most severe injury or most property damage to each individual vehicle, such as an overturn or a rear end collision. The new severity methodology incorporates the posted speed at a location and the types of most harmful events that are happening to determine the severity of collisions at that location.
To assess the gravity of potentially hazardous locations the HAL program assigns costs to the collision types described in the previous paragraph. Specifically, the new HAL methodology calculates statewide average collision costs based on different most harmful events and posted speed, then applies these costs to collision data from the current study period.
The following paragraphs represent the steps that the HAL program takes in ranking locations by severity. The steps represent a logical thought process to help the user understand methodology, and may not reflect the actual sequence of steps in the computer programming.
| Step One – | Historical collision data are categorized by Most Harmful Event (completed for each vehicle). |
||||||||||||||||||||
| Since particular types of collisions are prone to different injury severity, historical collision data (1994 – 1999) are first categorized by most harmful event. The most harmful event is the event that caused the most severe injury or most property damage to each individual vehicle. | |||||||||||||||||||||
| The HAL program breaks historical collision records down into categories based on the most harmful event of each vehicle in a collision. Each individual vehicle is assigned to a most harmful event category. | |||||||||||||||||||||
| Step Two – | Historical collision data are further categorized into posted speed ranges. | ||||||||||||||||||||
| Within most harmful event categories, speed effects severity uniquely. Historical collision data are further categorized into three posted speed ranges. | |||||||||||||||||||||
| After historical collision data are categorized by most harmful event, they are further broken down into posted speed ranges using the highest posted speed for intersections, and the posted speed of the road designated as the primary road on the collision report for road segments. | |||||||||||||||||||||
| Posted speed ranges were set after an analysis of empirical data revealed natural separations in severity. The ranges used by the HAL program are: | |||||||||||||||||||||
| 25 mph and lower 30 mph to 45 mph 50 mph and higher | |||||||||||||||||||||
| Step Three – | Current Federal Highway Administration (FHWA) injury cost estimates are applied to historical collision data. | ||||||||||||||||||||
| Current FHWA injury cost estimates are determined by multiplying the 1994 FHWA injury cost estimates by the Gross Domestic Product (GDP) implicit price deflator (an economic index which takes national productivity and inflation into account): | |||||||||||||||||||||
|
|||||||||||||||||||||
| A total cost is calculated for each vehicle in the historical database by adding the FHWA injury cost estimates for every individual injured in the vehicle. Average vehicle costs are calculated for each of the most harmful event categories, broken down by speed range and signal presence as discussed in steps 1 through 3. | |||||||||||||||||||||
| The result of steps 1 through 3 is the most harmful event cost table that lists the statewide average cost of collisions based on their most harmful event, and posted speed (Appendix B). The HAL program recalculates the table every year. By using five years of historical collision data the HAL program minimizes the impact of extreme events on the ranking process and establishes reasonable average vehicle costs that can be used to evaluate the severity of collisions. | |||||||||||||||||||||
| Step Four – | Average vehicle costs are applied to vehicles involved in collisions during the study period. |
||||||||||||||||||||
| The statewide average cost per vehicle for each most harmful event as calculated in step 3 are applied to each vehicle involved in a collision occurring during the study period. Next, the costs assigned to the vehicles involved in collisions at a location are summed to come up with a combined cost for all the collisions occurring at the given location | |||||||||||||||||||||
| The severity listing is created by ranking the combined cost for each location, from the highest cost to the lowest cost. | |||||||||||||||||||||
Collision Rate
The collision rate listing for the new HAL program uses road type categories that are fundamentally different from those traditionally used. The HAL program uses new road type classifications based solely on traffic volume rather than multiple physical roadway characteristics. The change in road type methodology reflects the revelation that certain roadway characteristics are inherently unsafe, and by grouping by this characteristic we are only highlighting the problem in the worst of the worst locations. Furthermore, while traffic volume is not a physical characteristic of the road, it does reflect the physical characteristics - volumes tell us what the public is demanding of roadways. In the remainder of this document 'road types' are referred to as volume ranges or volume groups.
The following paragraphs represent the steps that the HAL program takes in ranking locations by collision rate. The steps represent a logical thought process to help the user understand methodology, and may not reflect the actual sequence of steps in program operation.
| Step One – | Roadways and Intersections are divided into volume ranges, based on the average daily traffic. |
|||||||||||||||||||||||||||||||||||||||||||||||
| For all roadways the average ADT over a three year period is calculated. Intersections and road segments are categorized into one of the following volume ranges according to their three year average ADT. | ||||||||||||||||||||||||||||||||||||||||||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||
| Assigning volumes to intersections on the State Highway System is a difficult task complicated by jurisdictional boundaries and missing traffic volumes for legs of some intersections. With these difficulties in mind, the HAL program assigns volumes to intersections in the following manner: | ||||||||||||||||||||||||||||||||||||||||||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||
| For intersections with unknown volumes the HAL program tends to underestimate ADT. Underestimating the ADT for an intersection may lead to elevated critical rate rankings. For this reason, intersections with missing volumes should be evaluated cautiously – ideally volumes for these locations should be measured and submitted to improve future HAL listings. | ||||||||||||||||||||||||||||||||||||||||||||||||
| Step Two – | The HAL program calculates an average collision rate for each of the volume ranges. | |||||||||||||||||||||||||||||||||||||||||||||||
| For each volume range, the HAL program computes the average collision rate for roadways with volumes in that range. Collision rates are measured in collisions per million vehicles at intersections and collisions per million vehicle miles traveled for road segments. | ||||||||||||||||||||||||||||||||||||||||||||||||
| Step Three – | The collision rates of individual locations are mathematically compared to the average collision rate for that volume range. | |||||||||||||||||||||||||||||||||||||||||||||||
| Collision rates at all the individual locations are compared to the average collision rate for that volume range in order to generate a 'rate multiplier': | ||||||||||||||||||||||||||||||||||||||||||||||||
 | ||||||||||||||||||||||||||||||||||||||||||||||||
| Step Four – | The HAL program ranks locations in descending order by their collision rate multipliers. | |||||||||||||||||||||||||||||||||||||||||||||||
| The rate multipliers are listed in descending order to create the collision rate ranking. Locations with higher than expected collision rates tend to gravitate towards the top of the list while locations with fewer than expected collisions rank near the bottom. | ||||||||||||||||||||||||||||||||||||||||||||||||
Combined Ranking
The final HAL listing combines the results of the frequency, severity, and collision rate rankings into a single listing. Each of the three rankings described above is assigned a weighing factor, and for each location the combined report computes three weighted values by multiplying the location's position in a list by the weight factor associated with the list. These three weighted values are then summed and this becomes the location's weighted score.
The formula for the combined ranking is:
Weighted Score = Frequency rank(.25) + Severity rank (.50) + Collision rate rank (.25)
After all the locations are scored, the HAL program ranks locations in ascending order by the weighted scores.
The weight of individual rankings was determined through consultation with transportation industry professionals and university researchers. The selected weights were proposed by Dr. Layton of the Transportation Research Institute at Oregon State University. The weights provide a balance between the number of collisions (which is reflected in both frequency and rate) and their severity.
Ranking Criteria – Special Cases
After clusters are identified, they are subjected to the same ranking criteria as intersections. However, since clusters can be various sizes (rather than a point –like an intersection) the HAL program uses the number of collisions per mile in the frequency and critical rate rankings, and the economic loss per mile for the severity ranking, both are explained below.
Clusters are ranked according to the frequency of collisions in the road section. Unlike intersections, the size of clusters may vary greatly. Since the size of clusters vary the number of collisions per mile is used for ranking in order to avoid bias towards larger clusters. The number of collisions per mile is calculated by dividing the number of collisions occurring in a cluster by the length of the cluster. The resulting collision rate is used to rank clusters in descending order.
Since cluster sizes vary, and the number of collisions differs between clusters, the economic cost per mile is used to assign severity rankings. Economic cost per mile is calculated by dividing the total cost of all the collisions occurring in a cluster by the length of the cluster in miles. The total cost of all the collisions occurring in a cluster is calculated using the same methods described previously.
APPENDIX A. COLLISION REPORT FORM
APPENDIX B. HAL EVENT COST TABLES
| Event | Speed | Cost Sum | Units | Average |
|---|---|---|---|---|
| 01 Overturn | 1 – 25 | $109,995,908.00 | 883 | $124,570.68 |
| 26 – 46 | $414,082,430.00 | 3705 | $111,763.14 | |
| 46 – 100 | $2,130,483,505.00 | 12619 | $168,831.41 | |
| 02 Seperation of Units | 1– 25 | $15,932.00 | 8 | $1,991.50 |
| 26 – 45 | $110,374.00 | 24 | $4,598.92 | |
| 46 – 100 | $1,067,294.00 | 80 | $13,341.18 | |
| 03 Cargo Loss/Shift | 1–26 | $11,380.00 | 5 | $2,276.00 |
| 26 – 45 | $164,991.00 | 29 | $5,689.34 | |
| 46 – 100 | $1,770,478.00 | 133 | $13,311.86 | |
| 04 Jackknifed | 1 – 25 | $18,208.00 | 8 | $2,276.00 |
| 26 – 45 | $599,637.00 | 47 | $12,758.23 | |
| 46 – 100 | $2,973,247.00 | 466 | $6,380.36 | |
| 05 Ran Off Road | 1 – 25 | $2,276.00 | 1 | $2,276.00 |
| 26 – 45 | $28,446.00 | 4 | $7,111.50 | |
| 46 – 100 | $125,159.00 | 3 | $41,719.67 | |
| 06 Down Hill Runaway | 1 – 25 | $226,425.00 | 2 | $113,212.50 |
| 26 – 45 | $6,828.00 | 3 | $2,276.00 | |
| 46 – 100 | $26,170.00 | 3 | $8,723.33 | |
| 07 Fire/Explosion | 1 – 25 | $3,280,322.00 | 10 | $328,032.20 |
| 26 – 45 | $682,694.00 | 15 | $45,512.93 | |
| 46 – 100 | $33,834,092.00 | 136 | $248,780.09 | |
| 08 Gas/Inhalation | 26 – 45 | $2,276.00 | 1 | $2,276.00 |
| 09 Other Non-Collision | 1 – 25 | $3,961,867.00 | 30 | $132,062.23 |
| 26 – 45 | $3,159,712.00 | 73 | $43,283.73 | |
| 46 – 100 | $6,013,379.00 | 167 | $36,008.26 | |
| 10 Loss of Control | 46 – 100 | $286,729.00 | 1 | $286,729.00 |
| 11 Fell/Pushed/Jumped | 1 – 25 | $22,033,784.00 | 67 | $328,862.45 |
| 26 – 45 | $7,593,768.00 | 54 | $140,625.33 | |
| 46 – 100 | $24,018,131.00 | 40 | $600,453.28 | |
| 12 Non-Collision Injury | 1 – 25 | $3,204,087.00 | 3 | $1,068,029.00 |
| 26 – 45 | $815,812.00 | 19 | $42,937.47 | |
| 46 – 100 | $1,153,739.00 | 15 | $76,915.93 | |
| 13 Immersion | 1 – 25 | $15,457,216.00 | 25 | $618,288.64 |
| 26 – 45 | $25,983,146.00 | 55 | $472,420.84 | |
| 46 – 100 | $51,511,163.00 | 96 | $536,574.61 | |
| 14 Pedestrian | 1 – 25 | $95,191,752.00 | 1080 | $88,140.51 |
| 26 – 45 | $149,743,182.00 | 1163 | $128,755.96 | |
| 46 – 100 | $85,564,807.00 | 265 | $322,886.06 | |
| 15 Pedacycle | 1 – 25 | $43,917,178.00 | 1324 | $33,170.07 |
| 26 – 45 | $92,773,696.00 | 2313 | $40,109.68 | |
| 46 – 100 | $18,597,570.00 | 166 | $112,033.55 | |
| 16 Railroad Train | 1 – 25 | $5,004,127.00 | 80 | $62,551.59 |
| 26 – 45 | $45,658,261.00 | 127 | $359,513.87 | |
| 46 – 100 | $27,490,752.00 | 69 | $398,416.70 | |
| 17 Domestic Animal | 1 – 25 | $359,553.00 | 18 | $19,975.17 |
| 26 – 45 | $14,889,514.00 | 276 | $53,947.51 | |
| 46 – 100 | $32,701,091.00 | 1401 | $23,341.25 | |
| 18 Wild Animal | 1 – 25 | $251,461.00 | 13 | $19,343.15 |
| 26 – 45 | $2,565,850.00 | 282 | $9,098.76 | |
| 46 – 100 | $24,081,450.00 | 2610 | $9,226.61 | |
| 19 Other Object Not Fixed | 1 – 25 | $2,492,982.00 | 162 | $15,388.78 |
| 26 – 45 | $6,807,594.00 | 240 | $28,364.98 | |
| 46 – 100 | $19,931,208.00 | 599 | $33,274.14 | |
| 20 Parked on Private Property | 1 – 25 | $6,074,848.00 | 253 | $24,011.26 |
| 26 – 45 | $7,263,836.00 | 202 | $35,959.58 | |
| 46 – 100 | $987,629.00 | 46 | $21,470.20 | |
| 21 Impact Attenuator | 1 – 25 | $252,595.00 | 5 | $50,519.00 |
| 26 – 45 | $281,040.00 | 13 | $21,618.46 | |
| 46 – 100 | $1,055,896.00 | 40 | $26,397.40 | |
| 22 Bridge/Pier/Abutment | 1 – 25 | $672,451.00 | 22 | $30,565.95 |
| 26 – 45 | $2,543,019.00 | 70 | $36,328.84 | |
| 46 – 100 | $14,742,666.00 | 136 | $108,401.96 | |
| 23 Bridge/Parapet End | 1 – 25 | $254,872.00 | 5 | $50,974.40 |
| 26 – 45 | $475,611.00 | 13 | $36,585.46 | |
| 46 – 100 | $5,291,976.00 | 44 | $120,272.18 | |
| 24 Bridge Rail | 1 – 25 | $210,498.00 | 16 | $13,156.13 |
| 26 – 45 | $2,227,851.00 | 83 | $26,841.58 | |
| 46 – 100 | $10,643,139.00 | 222 | $47,942.07 | |
| 25 Overpass | 1 – 25 | $127,441.00 | 21 | $6,068.62 |
| 26 – 45 | $436,930.00 | 35 | $12,483.71 | |
| 46 – 100 | $13,178,179.00 | 45 | $292,848.42 | |
| 26 Guardrail Face | 1 – 25 | $4,354,426.00 | 59 | $73,803.83 |
| 26 – 45 | $11,858,332.00 | 214 | $55,412.77 | |
| 46 – 100 | $52,358,979.00 | 1336 | $39,190.85 | |
| 27 Guardrail End | 1 – 25 | $119,472.00 | 10 | $11,947.20 |
| 26 – 45 | $787,375.00 | 51 | $15,438.73 | |
| 46 – 100 | $20,402,174.00 | 229 | $89,092.46 | |
| 28 Median Barrier | 1 – 25 | $96,719.00 | 17 | $5,689.35 |
| 26 – 45 | $5,551,405.00 | 74 | $75,018.99 | |
| 46 – 100 | $13,369,393.00 | 655 | $20,411.29 | |
| 30 Highway Traffic Signpost | 1 – 25 | $881,828.00 | 81 | $10,886.77 |
| 26 – 45 | $2,871,883.00 | 173 | $16,600.48 | |
| 46 – 100 | $6,697,256.00 | 359 | $18,655.31 | |
| 31 Overhead Sign Support | 1 – 25 | $52,341.00 | 6 | $8,723.50 |
| 26 – 45 | $527,949.00 | 22 | $23,997.68 | |
| 46 – 100 | $3,373,624.00 | 17 | $198,448.47 | |
| 32 Luminaire/Light Support | 1 – 25 | $3,865,173.00 | 81 | $47,718.19 |
| 26 – 45 | $8,120,602.00 | 149 | $54,500.68 | |
| 46 – 100 | $1,485,997.00 | 74 | $20,081.04 | |
| 33 Utility Pole | 1 – 25 | $16,524,519.00 | 494 | $33,450.44 |
| 26 – 45 | $41,504,143.00 | 985 | $42,136.19 | |
| 46 – 100 | $72,890,765.00 | 1132 | $64,391.14 | |
| 39 Other Pole | 1 – 25 | $1,797,755.00 | 77 | $23,347.47 |
| 26 – 45 | $9,340,330.00 | 128 | $72,971.33 | |
| 46 – 100 | $5,188,449.00 | 101 | $51,370.78 | |
| 40 Delineator Post | 1 – 25 | $4,552.00 | 2 | $2,276.00 |
| 26 – 45 | $71,683.00 | 6 | $11,947.17 | |
| 46 – 100 | $7,495,958.00 | 166 | $45,156.37 | |
| 41 Culvert | 1 – 25 | $501,778.00 | 19 | $26,409.37 |
| 26 – 45 | $2,101,541.00 | 75 | $28,020.55 | |
| 46 – 100 | $6,048,625.00 | 136 | $44,475.18 | |
| 42 Curb | 1 – 25 | $2,548,728.00 | 139 | $18,336.17 |
| 26 – 45 | $10,978,799.00 | 179 | $61,334.07 | |
| 46 – 100 | $502,918.00 | 29 | $17,342.00 | |
| 43 Ditch | 1 – 25 | $3,169,950.00 | 119 | $26,638.24 |
| 26 – 45 | $28,284,950.00 | 671 | $42,153.43 | |
| 46 – 100 | $77,779,950.00 | 1326 | $58,657.58 | |
| 44 Embankment | 1 – 25 | $13,108,763.00 | 187 | $70,100.34 |
| 26 – 45 | $49,675,854.00 | 708 | $70,163.64 | |
| 46 – 100 | $166,745,659.00 | 1713 | $97,341.31 | |
| 45 Fence | 1 – 25 | $7,242,245.00 | 320 | $22,632.02 |
| 26 – 45 | $19,832,237.00 | 682 | $29,079.53 | |
| 46 – 100 | $47,105,733.00 | 1239 | $38,019.15 | |
| 46 Mailbox | 1 – 25 | $780,557.00 | 58 | $13,457.88 |
| 26 – 45 | $3,984,648.00 | 86 | $46,333.12 | |
| 46 – 100 | $1,495,115.00 | 94 | $15,905.48 | |
| 47 Tree | 1 – 25 | $57,395,944.00 | 491 | $116,896.02 |
| 26 – 45 | $104,410,469.00 | 950 | $109,905.76 | |
| 46 – 100 | $137,541,361.00 | 825 | $166,716.80 | |
| 48 Building/Wall | 1 – 25 | $8,320,873.00 | 240 | $34,670.30 |
| 26 – 45 | $7,764,450.00 | 160 | $48,527.81 | |
| 46 – 100 | $5,430,789.00 | 50 | $108,615.78 | |
| 49 Other Fixed Object | 1 – 25 | $19,569,327.00 | 403 | $48,559.12 |
| 26 – 45 | $23,767,884.00 | 538 | $44,178.22 | |
| 46 – 100 | $26,813,775.00 | 436 | $61,499.48 | |
| 50 Head On | 1 – 25 | $28,898,273.00 | 765 | $37,775.52 |
| 26 – 45 | $152,413,737.00 | 1917 | $79,506.38 | |
| 46 – 100 | $822,557,132.00 | 1596 | $515,386.67 | |
| 51 Rear End | 1 – 25 | $77,432,653.00 | 8076 | $9,588.00 |
| 26 – 45 | $518,492,136.00 | 43698 | $11,865.35 | |
| 46 – 100 | $289,161,695.00 | 12154 | $23,791.48 | |
| 52 Side Swipe Same | 1 – 25 | $14,446,477.00 | 2570 | $5,621.20 |
| 26 – 45 | $52,456,203.00 | 6779 | $7,738.04 | |
| 46 – 100 | $75,034,108.00 | 4441 | $16,895.77 | |
| 53 Side Swipe Opposite | 1 – 25 | $14,787,310.00 | 1230 | $12,022.20 |
| 26 – 45 | $55,152,552.00 | 2368 | $23,290.77 | |
| 46 – 100 | $184,215,856.00 | 2165 | $85,088.16 | |
| 54 Head On Turning | 1 – 25 | $45,774,924.00 | 3728 | $12,278.68 |
| 26 – 45 | $253,235,743.00 | 12912 | $19,612.43 | |
| 46 – 100 | $155,320,864.00 | 2611 | $59,487.12 | |
| 56 Rear End Turning | 1 – 25 | $6,032,802.00 | 667 | $9,044.68 |
| 26 – 45 | $30,227,609.00 | 2958 | $10,218.93 | |
| 46 – 100 | $45,320,423.00 | 1634 | $27,735.88 | |
| 58 Angle | 1 – 25 | $174,162,504.00 | 13245 | $13,149.30 |
| 26 – 45 | $384,114,471.00 | 18517 | $20,743.88 | |
| 46 – 100 | $602,461,536.00 | 6029 | $99,927.27 | |
| 59 Angle Turning | 1 – 25 | $33,814,655.00 | 4399 | $7,686.90 |
| 26 – 45 | $214,614,940.00 | 14663 | $14,636.50 | |
| 46 – 100 | $158,200,918.00 | 2939 | $53,828.14 | |
| 60 Backed Into | 1 – 25 | $14,025,786.00 | 3440 | $4,077.26 |
| 26 – 45 | $8,064,204.00 | 1726 | $4,672.19 | |
| 46 – 100 | $3,057,416.00 | 402 | $7,605.51 | |
| 61 Parked Car | 1 – 25 | $46,777,324.00 | 9506 | $4,920.82 |
| 26 – 45 | $21,394,919.00 | 2463 | $8,686.53 | |
| 46 – 100 | $35,920,919.00 | 873 | $41,146.53 | |
| 62 Same Direction Turning | 1 – 25 | $6,350,541.00 | 1397 | $4,545.84 |
| 26 – 45 | $23,560,478.00 | 3467 | $6,795.64 | |
| 46 – 100 | $76,166,787.00 | 2078 | $36,653.89 | |
| 72 Drove L/R Of Center | 46 – 100 | $2,276.00 | 2 | $1,138.00 |
| 99 Other | 1 – 25 | $4,089,326.00 | 118 | $34,655.31 |
| 26 – 45 | $3,482,910.00 | 327 | $10,651.10 | |
| 46 – 100 | $22,608,513.00 | 602 | $37,555.67 |
