U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
202-366-4000


Skip to content
FacebookYouTubeTwitterFlickrLinkedIn

Safety

eSubscribe
eSubscribe Envelope

FHWA Home / Safety / Intersection / Advancing Innovative Intersection Safety Treatments for Two-Lane Rural Highways: Executive Summary

Advancing Innovative Intersection Safety Treatments for Two-Lane Rural Highways: Executive Summary

banner

 

Advancing Innovating Intersection Safety Treatments for Two-Lane Rural Highways

 

Submitted to

Federal Highway Administration

Contract No. DTFH61-12-C-00023

MRIGlobal Project No. 110818

 

Prepared by

MRIGlobal

 

December 2015

1. Report No. FHWA-SA-16-003 2. Government Accession No. 3. Recipient's Catalog No.
4. Title and Subtitle Advancing Innovative Intersection Safety Treatments for Two-Lane Rural Highways 5. Report Date December 2015
6. Performing Organization Code
7. Author(s) D.J. Torbic, D.J. Cook, J.M. Hutton, K.M. Bauer, and J.M. Sitzmann 8. Performing Organization Report No. 110818.01.004
9. Performing Organization Name and Address MRIGlobal
425 Volker Boulevard
Kansas City, MO 64110-2241
10. Work Unit No. (TRAIS)
11. Contract or Grant No. DTFH61-12-C-00023
12. Sponsoring Agency Name and Address Federal Highway Administration
Office of Safety
1200 New Jersey Avenue SE
Washington, DC 20590
13. Type of Report and Period Covered Final Report
September 2012 - December 2015
14. Sponsoring Agency Code
15. Supplementary Notes Project Manager: Jeffrey Shaw
16. Abstract Intersection safety is a national, state, and local priority. Approximately 26 percent of the fatal crashes that occur in the United States are intersection or intersection-related crashes. The objective of this guide is to advance efforts to improve safety at unsignalized intersections with minor-road stop control along rural two-lane roads, by focusing on strategies that are not yet widespread. The safety effectiveness of three low-cost safety treatments was evaluated to estimate their expected effectiveness in reducing crashes. The low-cost safety treatments included: (1) single luminaire intersection lighting, (2) transverse rumble strips in advance of stop-controlled approaches, and (3) supplementary pavement markings on intersection approaches. The effectiveness of each treatment in reducing crashes was estimated using the Empirical Bayes (EB) observational before-after safety evaluation analysis approach. Analyses were performed to estimate the effectiveness of each treatment in reducing crashes for different severity levels and crash types. An economic analysis of the treatments was also performed. Each of the treatments is effective in reducing crashes of different types and severity levels, and is economically justifiable for most traffic volumes. The information in this report can be combined with information on other strategies to reduce intersection or intersection-related crashes at unsignalized intersections with minor-road stop control along rural two-lane roads. With such information, agencies can make informed decisions about planning and programming safety improvements at intersections under their jurisdiction.
17. Key Words Highway safety, Countermeasure evaluation, Intersection safety, Crash modification factor, Lighting, Empirical Bayes before-after analysis, Rumble strips, Pavement markings 18. Distribution Statement No restrictions. This document is available to the public through the National Technical Information Service, Springfield, Virginia 22161
19. Security Classification (of this report) Unclassified 20. Security Classification (of this page) Unclassified 21. No of Pages 22. Price
Form DOT F 1700.7 (8-72) Reproduction of completed page authorized.

Title: SI (Modern Metric) Conversion Factors - Description: SI (Modern Metric) Conversion Factors

Figures

Figure 1. Equation. CMF for lighting.

Figure 2. Equation. Benefit-cost ratio for 3-leg intersections.

Figure 3. Equation. Benefit-cost ratio for 4-leg intersections.

Figure 4. Photo. Examples of single luminaire intersection lighting (Image credit: Google Earth™ Mapping Service).

Figure 5. Equation. Base model for predicted crashes per year.

Figure 6. Equation. Model for predicted crashes per year for specific crash types and severity levels, and accounting for local conditions.

Figure 7. Equation. CMF for total crashes.

Figure 8. Photo. Examples of transverse rumble strips placement (Image credit: Google Earth™ Mapping Service).

Figure 9. Equation. Base model for predicted crashes per year.

Figure 10. Equation. Model for predicted crashes per year for specific crash types and severity levels, and accounting for local conditions.

Figure 11. Diagram. MoDOT's design guidance for placement of transverse rumble strips.

Figure 12. Diagram. MoDOT's design detail for transverse rumble strips.

Figure 13. Diagram. North Dakota DOT's transverse rumble strip design detail.

Figure 14. Diagram. KDOT transverse rumble strip installation and design detail.

Figure 15. Photo. Aerial view and street view of supplementary pavement markings on stop-controlled approach (Image credit: Google Earth™ Mapping Service).

Figure 16. Photo. Supplementary pavement markings on uncontrolled approach. (Image credit: Pennsylvania DOT)

Figure 17. Equation. Base model for predicted crashes per year.

Figure 18. Equation. Model for predicted crashes per year for specific crash types and severity levels, and accounting for local conditions.

Figure 19. Diagram. Minnesota guidance for placement of STOP AHEAD pavement marking relative to W3-1a sign.

Figure 20. Diagram. Minnesota guidance for placement of W3-1a sign relative to stop bar.

Figure 21. Diagram. Placement guidance for supplementary pavement markings on uncontrolled approach.

Tables

Table 1. Crash reduction estimates for lighting from The Handbook of Road Safety Measures.

Table 2. States included in the safety evaluation of each treatment type.

Table 3. Crash costs by severity level assumed for economic analysis

Table 4. Major- and minor-road AADT ranges assumed for economic analysis.

Table 5. Single luminaire: number of sites by number of intersection legs.

Table 6. Single luminaire: summary of nighttime crash statistics for the before and after treatment periods for treatment intersections by number of intersection legs.

Table 7. Single luminaire: summary of nighttime crash statistics for the entire study period for nontreatment intersections by number of intersection legs.

Table 8. Single luminaire: SPF coefficients, target crash proportions, and calibration factors by number of intersection legs for EB analysis.

Table 9. Single luminaire: safety effectiveness on nighttime crashes for 3- and 4-leg intersections combined.

Table 10. Single luminaire: benefit-cost ratios for treatment at 3-leg rural stop-controlled intersection.

Table 11. Single luminaire: benefit-cost ratios for treatment at 4-leg rural stop-controlled intersection.

Table 12. Prioritization of lighting at rural intersections by traffic volume and functional class.

Table 13. Transverse rumble strips: number of sites by State, number of intersection legs, and number of treated approaches.

Table 14. Transverse rumble strips: summary crash statistics for the before and after treatment periods for treatment intersections by State and number of intersection legs.

Table 15. Transverse rumble strips: summary crash statistics for the entire study period for nontreatment intersections by State and number of intersection legs.

Table 16. Transverse rumble strips: SPF coefficients, target crash proportions, and calibration factors by State and number of intersection legs.

Table 17. Transverse rumble strips: safety effectiveness on target crashes by number of intersection legs.

Table 18. Transverse rumble strips: benefit-cost ratios for treatment on two minor-road approaches at 4-leg rural stop-controlled intersection for $10,000 installation cost and 20-yr service life.

Table 19. Transverse rumble strips: benefit-cost ratios for treatment on two minor-road approaches at 4-leg rural stop-controlled intersection for $2,000 installation cost and 5-yr service life.

Table 20. Supplementary pavement markings: number of sites by State, number of intersection legs, and number of treated approaches.

Table 21. Supplementary pavement markings installed on stop-controlled approaches: summary crash statistics for the before and after treatment periods for treatment intersections by State and number of intersection legs.

Table 22. Supplementary pavement markings installed on stop-controlled approaches: summary crash statistics for the entire study period for nontreatment intersections by State and number of intersection legs.

Table 23. Supplementary pavement markings installed on uncontrolled approaches: summary crash statistics for the before and after treatment periods for treatment intersections by State and number of intersection legs.

Table 24. Supplementary pavement markings installed on uncontrolled approaches: summary crash statistics for the entire study period for nontreatment intersections by State and number of intersection legs.

Table 25. Supplementary pavement markings installed on stop-controlled approaches: SPF coefficients, target crash proportions, and calibration factors by State and number of intersection legs.

Table 26. Supplementary pavement markings installed on uncontrolled approaches: SPF coefficients, target crash proportions, and calibration factors by State and number of intersection legs.

Table 27. Supplementary pavement markings installed on stop-controlled approaches: safety effectiveness on target crashes by number of intersection legs.

Table 28. Supplementary pavement markings installed on uncontrolled approaches: safety effectiveness on target crashes.

Table 29. Benefit-cost ratios for installing STOP AHEAD pavement marking on the stop-controlled approach of a 3-leg intersection for $750 installation cost and 1-yr service life.

Table 30. Benefit-cost ratios for installing STOP AHEAD pavement markings on the stop-controlled approaches of a 4-leg intersection for $1,500 installation cost and 1-yr service life.

Table 31. Benefit-cost ratios for installing STOP AHEAD pavement marking on the stop-controlled approach of a 3-leg intersection for $500 installation cost and 2-yr service life.

Table 32. Benefit-cost ratios for installing STOP AHEAD pavement markings on the stop-controlled approaches of a 4-leg intersection for $1,000 installation cost and 2-yr service Life.

Table 33. Benefit-cost ratios for installing supplementary pavement markings on uncontrolled approaches at 4-leg stop-controlled intersections.

Table 34. Placement of supplementary pavement markings relative to intersection by posted speed.

Table 35. Summary of treatment effectiveness by treatment and intersection type and crash type and crash severity level.

Chapter 1. Introduction

Background

Intersection safety is a national, state, and local priority. Approximately 26 percent of the fatal crashes that occur in the United States are intersection or intersection-related crashes.(1,2) In the period between 2009 and 2013, the average number of intersection-related fatal crashes was approximately 7,960 per year.(1)

Crashes in rural areas are often more severe than in urban areas because of higher vehicle speeds, and the outcome of crashes may be more severe, in part, due to longer emergency response times. In rural areas, more fatal and severe injury crashes occur at stop-controlled intersections than at signalized intersections.(3) At stop-controlled intersections, most crashes are caused by a failure to stop at a stop-controlled approach or an acceptance of an insufficient gap when entering the intersection.(4)

Objective

The objective of this guide is to advance efforts to improve safety at unsignalized intersections with minor-road stop control along rural two-lane roads, by focusing on strategies that are not yet widespread. The safety effectiveness of three low-cost safety treatments was evaluated to estimate their expected effectiveness in reducing crashes. The low-cost safety treatments included:

The information in this guide can be combined with information on other strategies to reduce intersection or intersection-related crashes at unsignalized intersections with minor-road stop control along rural two-lane roads. For example, the National Cooperative Highway Research Program (NCHRP) Report 500 Guidance for Implementation of the AASHTO Strategic Highway Safety Plan, Volume 5: A Guide for Addressing Unsignalized Intersection Collisions provides proven, tried, and experimental strategies for reducing crashes at unsignalized intersections.(5) The American Association of State Highway and Transportation Officials (AASHTO) Highway Safety Manual (HSM) provides crash prediction methods for estimating the predicted and/or expected crash frequency at unsignalized intersections on rural two-lane roads and related crash modification factors (CMFs).(6) Similarly, the online CMF Clearinghouse, sponsored by the Federal Highway Administration (FHWA), houses a web-based database of CMFs along with supporting documentation to help transportation engineers identify appropriate countermeasures for their safety needs.(7) With such information, agencies can make more informed decisions when planning and programming safety improvements at intersections under their jurisdiction.

The overall goal of this guide is to increase the deployment of treatments that reduce motor vehicle fatalities and injuries at unsignalized intersections with minor-road stop control along rural two-lane roads by providing practitioners with knowledge about their installation and expected safety benefits. Throughout this guide, the term "intersections" refers to unsignalized intersections with minor-road stop control along rural two-lane roads unless otherwise stated.

How to Use This Guide

This guide serves as both a technical report of the treatments evaluated in the research, and as a practical guide for practitioners who may be interested in implementing the treatments evaluated. Guidance includes, where appropriate, the conditions under which the treatment is expected to be effective and design and installation considerations.

Chapter 2 of this guide includes general information about how the research was conducted, including identification of intersections for inclusion in the research, the data that were collected, and the analysis approach.

Chapters 3 through 5 each discuss one of the three treatments evaluated as part of this research in depth: single luminaire intersection lighting, transverse rumble strips, and supplementary pavement markings. Each chapter begins with a summary table of key information about the treatment, its use, and its expected effectiveness. The chapter then includes a detailed discussion under the following subheadings:

Conclusions are presented in Chapter 6.

Chapter 2. Research Approach

This Chapter describes the general approach to the research, including identification of treatments to be evaluated, selection of treatment and nontreatment sites (i.e., intersections) for inclusion in the safety evaluations, a description of the data elements collected for the safety evaluations, and descriptions of the methodological approaches used to evaluate the safety effectiveness and economic benefits of the treatments.

Selection of Treatments for Evaluation

At the beginning of the project, the research team identified 32 low-cost safety treatments applicable to unsignalized intersections on rural two-lane roads for which limited knowledge about their safety effectiveness was available. The list of treatments was developed by gathering information through a combination of a literature review, a virtual/desktop scan of national and international experiences, and telephone discussions and email exchanges with highway agency staff and consultants. Examples of treatment types considered for evaluation as part of this research included, but were not limited to, the following:

For each potential treatment considered for evaluation, the following information was gathered:

After several iterations of prioritizing potential treatments, three were selected for evaluation:

Estimating the safety effectiveness of installing a single luminaire at a stop-controlled intersection on a rural two-lane highway was identified as a high priority because safety effectiveness information for this treatment available in the HSM is general in nature and was adapted from studies that were not specific to rural intersections. HSM Chapter 10, Predictive Method for Rural, Two-Lane, Two-Way Roads, includes the following CMF for intersection lighting at stop-controlled intersections:(6)

CMFLighting = 1 – 0.38 × pni

Figure 1. Equation. CMF for lighting.

where pni is the proportion of total crashes for unlighted intersections that occur at night.

The CMF is applicable to total crashes, and the base condition is the absence of intersection lighting. This CMF indicates that intersection lighting is expected to reduce nighttime crashes at intersections on rural two-lane roads by 38 percent. Chapter 10 of the HSM states that the CMF is adapted from crash reduction estimates presented in The Handbook of Road Safety Measures by Elvik and Vaa.(8) The Handbook provides the following crash reduction estimates for roadway lighting for crashes that occur in darkness based on a meta-analysis of 25 studies from several countries conducted between 1948 and 1993 (Table 1).

Table 1. Crash reduction estimates for lighting from The Handbook of Road Safety Measures.(8)
Crash severity Percent reduction in crashes occurring in darkness
Fatal 64
Injury 28
Property-damage-only 17

No information specific to intersections or rural roads is presented.

HSM Chapter 14, Intersections, which provides information about intersection treatments that goes beyond what is provided in the predictive methods chapters, includes a CMF for intersection lighting. The Chapter 14 CMF for intersection lighting on nighttime crashes is 0.62 (38 percent reduction, as shown in Chapter 10), but is specified for injury crashes only. In addition, a CMF for pedestrian nighttime injury crashes of 0.58 is provided. Chapter 14 of the HSM cites four sources (8,9,10,11) for these intersection lighting CMFs rather than the single reference (8) provided in HSM Chapter 10. However, in only one of the sources (11) is the safety effectiveness estimate for roadway lighting based solely on data from unsignalized intersections along rural two-lane roads. The study by Preston and Schoenecker (11) is also cited in NCHRP Report 500 (Vol. 5), A Guide for Addressing Unsignalized Intersection Collisions,(5) which lists "Improve Visibility of the Intersection by Providing Lighting" as a "proven" safety treatment, yet the results of the Preston and Schoenecker study are not incorporated into the HSM Chapter 10 predictive methodology for intersections on rural two-lane roads.

By conducting a detailed safety evaluation for the installation of a single luminaire at stop-controlled intersections on rural two-lane roads, the intention is to develop a more reliable estimate for intersection lighting specific to the condition of interest (i.e., unsignalized intersection with minor-road stop control on rural two-lane roads) than what is currently provided in the HSM. In addition, this study includes in the analysis only intersections where a single luminaire was installed, making the results specific to this type of lighting implementation, which is currently not specified in the HSM or other available literature.

Estimating the safety effectiveness of transverse rumble strips installed on minor-road intersection approaches was identified as a high priority because it is a relatively common treatment, but reliable safety estimates specific to stop-controlled intersections on rural two-lane roads are not available. HSM Chapter 14 provides a discussion of rumble strips installed on intersection approaches, and suggests that they are frequently used to inform drivers of an upcoming change in the roadway or as traffic calming devices. The discussion notes that in-lane rumble strips appear to reduce all crash types of all severities on urban roads, but that the magnitude of the effect is not known. No information is provided regarding rural intersections. In a recent before-after study using crash data from 3- and 4-leg stop-controlled intersections in rural areas, Srinivasan et al. reported that transverse rumble strips reduce KAB (fatal and injury) crashes by 21 percent and KA (fatal and severe injury) crashes by 39 percent.(12) The results of this study are included in the CMF Clearinghouse, with star ratings ranging from 3 to 4 stars. However, it is unknown whether the analysis results from this study are based on intersections located exclusively along rural two-lane roads or a combination of sites located along rural two-lane roads and rural multilane divided highways, and due to sample size issues, no definitive conclusions were provided regarding the impacts on specific crash types. Therefore, it is desirable to develop a crash reduction estimate for transverse rumble strips specifically for unsignalized intersections with minor-road stop control on rural two-lane roads and to estimate the safety effects of this treatment on specific target crashes (e.g., angle and rear-end).

As with transverse rumble strips, supplementary pavement markings (such as STOP AHEAD) were identified as a high priority treatment for evaluation because they are relatively common treatment, but little reliable information is available regarding their impact on crashes. CMFs for STOP AHEAD pavement markings on various intersections types and for various crash types and severities are included in Chapter 14 of the HSM, but they are not incorporated into the HSM predictive chapters. Only one of the CMFs-for all crash types and severities at stop-controlled rural intersections-has a standard error low enough for it to be considered reliable. It indicates a 31 percent crash reduction when STOP AHEAD pavement markings are used. All the STOP AHEAD pavement marking CMFs in Chapter 14 of the HSM are based on a study by Gross et al.,(13) which used the Empirical Bayes before-after analysis approach. No CMF is provided specifically for unsignalized intersections with minor-road stop control on rural two-lane roads. The CMF Clearinghouse includes CMFs with ratings of 3 or 4 stars for this treatment based upon the study by Gross et al.(13) Therefore, it was desirable to develop a more reliable crash reduction estimate specifically for supplementary pavement markings (such as STOP AHEAD) for unsignalized intersections with minor-road stop control on rural two-lane roads for possible inclusion in HSM Chapter 10. It was also desirable to estimate the safety effects of this treatment on specific target crashes (e.g., angle and rear-end) for the conditions of interest.

In seeking potential study sites for this treatment, the research team did not limit the supplementary pavement marking message to STOP AHEAD only. While a majority of sites did have this message (installed on the minor approaches), several sites were identified at which a supplementary pavement marking treatment was installed on the major (uncontrolled) approaches to the intersection. Since no reliable safety effectiveness estimates were available for this specific application, a separate analysis of the safety effectiveness of this treatment was conducted as well.

Selection of Treatment and Nontreatment Sites

After selecting the three treatment types to be evaluated, the research team contacted state and county highway agencies to request location information (e.g., major-road name, minor-road name, county) for intersections at which a treatment of interest had been installed. The research team then gathered additional data for each site, including intersection geometrics and site charateristics data, traffic volumes, installation dates, and construction history to determine if each site was suitable for inclusion in the analysis. The following criteria were used to narrow the list of potential treatment sites to those included in the analysis:

Additional details about the data collection process are provided in the discussion on data collection later in this chapter.

Table 2 shows the states in which treatment sites included in the analysis were located.

Once the list of treatment sites for inclusion in the analysis was finalized, the research team identified nontreatment sites in each state where treatment sites were located. Nontreatment sites are used in the before-after analysis to account for crash trends over time. Potential nontreatment sites were identified by searching aerial and street-view images of rural roads in the vicinity of treatment intersections, and then expanding that search to similar, nearby routes. Nontreatment sites were selected from the potential sites using the following criteria:

Table 2. States included in the safety evaluation of each treatment type.
States Single luminaire intersection lighting Transverse rumble strips Supplementary pavement markings
Arkansas   Yes Yes
Kansas   Yes  
Minnesota Yes   Yes
Missouri   Yes  
Nebraska     Yes
North Dakota   Yes  
Oregon   Yes  
Pennsylvania     Yes
Vermont     Yes

Approximately 30 nontreatment sites were identified in each state for inclusion in the analysis. All nontreatment sites in each state were used in the analysis of all treatments in that state.

The discussion of the statistical analysis in Chapters 3, 4, and 5 includes descriptive statistics for the treatment and nontreatment sites included in the analysis.

Data Collection

The following data are critical for performing a safety evaluation of a treatment:

Final selection of intersections for inclusion in the analyses was determined, in part, by the availability of such data.

Treatment Installation Date

For each potential treatment site, the research team obtained the treatment installation year from state highway agencies. This information was critical for defining the appropriate analysis period for evaluation of the treatment at the respective intersection. The goal was to include sites at which treatments were installed between 2005 and 2008 so that four to five years of crash data would be available after installation of the treatment, and before-period data would be from within the last 15 years. In cases where only a limited number of treatment sites met this criterion, exceptions were made to allow sites with fewer years of before or after data in the analysis or include "before" period crash data prior to 2000 in the analysis. In cases where the installation date of an intersection treatment was unknown, that intersection could not be included in the Empirical Bayes (EB) before-after evaluation. The EB before-after analysis methodology is discussed below, under "Analysis Approach." .

The research team also requested information from state highway agencies as to whether substantial improvements or changes other than the installation of the treatment being evaluated were made during the study period at any of the intersections considered for inclusion in the evaluation. Intersections where other substantial improvements or changes were made were either excluded from the evaluation or the study period was adjusted to only include the years prior to or after the changes were made. Where possible, the research team used Google Earth to review images over a number of years to assess the history of changes to intersections. These images were used to gather construction history data when the highway agency was not able to provide it or to validate construction history information when it was provided. The history of changes to nontreatment intersections was assessed in the same manner.

Crash Data

The research team obtained crash data from state highway agencies for treatment and nontreatment sites in electronic form. As indicated above, the goal was to obtain crash data for four to five years before and four to five years after installation of the treatment at the treatment sites. Data were requested up to the most recent year of available data, even if this resulted in more than five years of after-period data. For a given treatment and state, crash data for nontreatment sites were requested for years covering the full range of before and after periods of the treatment sites included in the analysis. For Minnesota, some crash data were obtained from the Highway Safety Information System (HSIS) for the analysis.

Data for individual crashes were obtained at the crash and vehicle levels. The primary crash data elements of interest for the safety evaluations included:

Traffic Volume Data

The research team obtained traffic volumes [annual average daily traffic (AADT)] for the major- and minor-road approaches of each treatment and nontreatment intersection. AADTs were obtained for as many years as available for the study period. If the AADTs differed for both major-road approaches to an intersection, the greater of the two AADT values was used in the analysis; AADTs for the minor-road approaches were treated in a similar fashion. If an AADT for a particular year was missing, and AADTs for years before and after that year were known, then the AADT for the missing year was estimated through interpolation. If an AADT was missing at the beginning or end of a study period, then the closest AADT was simply used for that year. If AADTs were missing for most or all study years, the intersection was excluded from the analysis.

Intersection Characteristics Data

Treatment and nontreatment site characteristics were obtained using Google Earth and Google Street View. Data for the following intersection characteristics were collected:

These data were used to 1) choose the appropriate safety performance function (SPF) and CMFs to predict the expected number of crashes at the intersection, and 2) to better define the specific implementation of the treatment.

Analysis Approach

The effectiveness of each treatment in reducing crashes was estimated using the Empirical Bayes (EB) observational before-after safety evaluation analysis approach. An economic analysis of the treatments was also performed using the safety effectiveness information developed in this research. The general approach to these analysis methodologies is described below, while specific details of how each approach was applied for the individual treatments is provided in the respective chapters.

EB Before-After Safety Evaluation Method

The collected data lent themselves to an Empirical Bayes (EB) observational before-after safety evaluation with reference sites. The EB method is now a standard approach to safety evaluations.(14,15) The EB method has been applied by members of the research team in a number of recent projects (16,17) and by others (13,18).

The EB approach overcomes the difficulties associated with conventional before-after comparisons by:

The specific EB approach used for this evaluation follows the steps presented in Appendix 9A in Chapter 9 of the HSM.(6) The analysis approach is comprised of four basic steps.

STEP 1: Calibrate the appropriate HSM SPFs within each state using the reference site data within that State. The research team also included the "before" period data for the treatment sites within that State in this step.

STEP 2: Estimate the expected number of crashes in the before period by taking a weighted average of the observed crash count and the predicted crash frequency calculated from the calibrated HSM SPF to estimate the EB-adjusted expected crash frequency in the before period.

STEP 3: Estimate the expected number of crashes in the "after" period had the treatment not been installed. This estimate is obtained by adjusting the EB-adjusted expected crash frequency from the before period (as calculated in Step 2) for the difference between before and after AADTs and between before and after number of years.

STEP 4: Estimate the effectiveness of the treatment by comparing the observed number of crashes in the after period to the expected number of crashes in the after period, had the treatment not been installed (as calculated in Step 3).

The effectiveness of each treatment in reducing crashes was estimated separately for the following crash severity levels and crash types:

Severity level:

Crash type:

The injury scale used in this analysis can be translated to the KABCO crash severity scale as follows:

where:

K = fatal crash

A = disabling injury crash

B = evident injury crash

C = possible injury crash

Economic Analysis

An economic analysis was performed for each of the three intersection treatment types to estimate the benefit-cost ratio of each treatment. The benefit-cost ratio is the ratio of the present benefit of a treatment, measured in monetary terms of the number of crashes reduced due to installation of the treatment, to its construction costs. For a countermeasure to be economically justified, its benefit-cost ratio should be greater than 1.0. The most desirable countermeasures are those with the highest benefit-cost ratios. The safety effectiveness results of the three treatments were incorporated in the economic evaluations.

Implementation Cost and Service Life

The implementation costs and service lives of the safety treatments were determined based on input from State highway agencies. Costs and service lives can vary by geographic regions. For example, a cold-weather State may want to assume a shorter service life than a warm-weather State for pavement markings due to snowplows damaging the treatment. Care should be taken to select a service life appropriate to the region in which the treatment is being installed.

Crash Reduction Benefit

The annual benefits for the safety treatments were estimated as follows:

  1. Estimate crash frequency by crash severity level for the existing 3-leg and 4-leg stop-controlled intersections on rural two-lane roads prior to installation of the treatment, based on the HSM Part C predictive methods.
  2. Estimate the reduction in crash frequency by severity due to the treatment implementation, based on CMFs derived from this research.
  3. Calculate annual crash cost savings.

Table 3 presents the crash costs used in the analysis, which are the values provided in HSM Chapter 7. Transportation agencies may use different values for crash costs that may be more current or specific to a location than the values shown in Table 3. A 2015 NHTSA report (19) estimates the economic and societal impact of a motor vehicle crash fatality to be $9,146,000, while a June 2015 memo from the Office of the Secretary of Transportation (20) provides guidance for transportation analyses to use $9.4 million as the economic value of a statistical life. Substituting higher values, such as these, for the values shown in Table 3 will only increase the calculated benefit of a given treatment. The crash cost values presented in Table 3 were used in the economic analyses for consistency with the HSM.

Table 3. Crash costs by severity level assumed for economic analysis.(6)
Crash Severity Level Comprehensive Societal Crash Costs
Fatal (K) $4,008,900
Disabling injury (A) $216,000
Evident injury (B) $79,000
Possible injury (C) $44,900

Property damage only (O)

$7,400

The SPFs in the HSM Part C predictive methods require major- and minor-road AADTs to compute predicted crash frequencies. The intersections used to produce CMFs in this research consisted of a wide range of traffic volumes on both the major- and minor-road approaches. Annual crash reduction benefits were calculated for each treatment type by varying the major- road AADT in 1,000-veh/day increments. The minor-road AADT was varied as a percentage of the major-road AADT from 5 to 95 percent, in increments of 5 percent. Table 4 presents the major-road and minor-road AADT ranges used in the economic analysis.

Table 4. Major- and minor-road AADT ranges assumed for economic analysis.
Treatment Type Intersection Type Major-Road AADT Range (veh/day) Minor-Road AADT Range (veh/day)
Single luminaire 3 legs 220 to 5,900 30 to 2,000
4 legs 315 to 2,100 210 to 1,750
Transverse rumble strips 4 legs 200 to 5,000 90 to 3,185
Supplementary pavement markings installed on stop-controlled approach (i.e., STOP AHEAD) 3 legs 115 to 5,200 40 to 2,940
4 legs 155 to 3,650 40 to 1,500
Supplementary pavement markings installed on uncontrolled approaches legs 2,310 to 14,000 330 to 2,730

Benefit-Cost Ratio

To estimate the benefit-cost ratio of a treatment, it is necessary to convert the annual benefit of a treatment to a net present value. To do so, a discount rate or minimum attractive rate of return of 7 percent was assumed, in accordance with current Federal guidelines. The benefit-cost ratio is computed by dividing the net present value of the crash reduction benefit by the implementation cost. Results of the economic analysis for each treatment are provided in Chapters 3 through 5.

The benefit-cost ratio for a given treatment installed at 3-leg stop-controlled intersections on rural two-lane roads is calculated using the following equation:

B over C equals open squiggly bracket open parenthesis 1 minus CMF close parenthesis times C subscript r times exponent open parenthesis minus 9.86 plus 0.79 times natural logarithm of AADT subscript maj plus 0.49 times natural logarithm of AADT subscript min closed parenthesis closed parenthesis times quotient of open squiggly bracket open bracket open parenthesis 1 + i closed parenthesis to the power n closed bracket minus 1 closed squiggly bracket over i times open parenthesis 1 + i closed parenthesis to the power n all over C  subscript Imp plus C  subscript Ann times quotient of open squiggly bracket open bracket open parenthesis 1 + i closed parenthesis to the power n closed bracket minus 1 closed squiggly bracket over i times open parenthesis 1 + i closed parenthesis to the power n all times open parenthesis C  subscript K times P  subscript K plus C  subscript A times P  subscript A plus C  subscript B times P  subscript B plus C  subscript C times P  subscript C plus C  subscript O times P  subscript O closed parenthesis

Figure 2. Equation. Benefit-cost ratio for 3-leg intersections

The benefit-cost ratio for a given treatment installed at 4-leg stop-controlled intersections on rural two-lane roads is calculated using the following equation:

B over C equals open squiggly bracket open parenthesis 1 minus CMF close parenthesis times C subscript r times exponent open parenthesis minus 8.56 plus 0.60 times natural logarithm of AADT subscript maj plus 0.61 times natural logarithm of AADT subscript min closed parenthesis closed parenthesis times quotient of open squiggly bracket open bracket open parenthesis 1 + i closed parenthesis to the power n closed bracket minus 1 closed squiggly bracket over i times open parenthesis 1 + i closed parenthesis to the power n all over C  subscript Imp plus C  subscript Ann times quotient of open squiggly bracket open bracket open parenthesis 1 + i closed parenthesis to the power n closed bracket minus 1 closed squiggly bracket over i times open parenthesis 1 + i closed parenthesis to the power n all times open parenthesis C  subscript K times P  subscript K plus C  subscript A times P  subscript A plus C  subscript B times P  subscript B plus C  subscript C times P  subscript C plus C  subscript O times P  subscript O closed parenthesis

Figure 3. Equation. Benefit-cost ratio for 4-leg intersections

where:

B/C = benefit-cost ratio

CMF = crash modification factor for total intersection crashes for a given treatment or combination of treatments

Cr = calibration factor for SPF

AADTmaj = major-road AADT

AADTmin = minor-road AADT

i = discount rate (assumed 7 percent)

n = service life

CImp = installation cost ($)

CAnn = annual operational cost of treatment ($)

CK = fatal crash cost ($ per fatal crash)

PK = proportion of total intersection crashes that are fatal crashes

CA = disabling injury crash cost ($ per disabling injury crash)

PA = proportion of total intersection crashes that are disabling injury crashes

CB = evident injury crash cost ($ per evident injury crash)

PB = proportion of total intersection crashes that are evident injury crashes

CC = possible injury crash cost ($ per possible injury crash)

PC = proportion of total intersection crashes that are possible injury crashes

CO = property damage only crash cost ($ per property damage only crash)

PO = proportion of total intersection crashes that are property damage only crashes

Chapter 3. Single Luminaire

Treatment Name Single Luminaire
Description A single luminaire is installed at an intersection to make drivers aware of the presence of the intersection at night.
States Data from one state-Minnesota–were included in the safety evaluation of this treatment.
Safety Effectiveness Reduction by severity level:

The EB analysis showed a 71-percent reduction in nighttime crashes (SE = 29 percent) for all severity levels combined.
Cost and Economic Benefits

Benefit-cost ratios ranged from 0.5 to 35.0, assuming an $8,000 installation cost and $300 annual energy cost over a 20-year life. These costs assume light fixtures that used traditional wired power and the availability of a nearby power source. However, street lighting that uses solid-state LED bulbs and solar power generally have no need for a wired power source near the intersection and have no annual energy cost.

Where to Implement

This treatment should be considered at intersections with a high proportion of crashes occurring during hours of darkness, or simply at intersections with a moderate to high frequency of nighttime crashes. Intersections with patterns of nighttime crashes that suggest drivers are unaware of the presence of the intersections (such as near horizontal curves or locations with skewed approaches) may particularly benefit from this treatment.

Additional Factors for Consideration

If there is vegetation near the intersection, foliage should be trimmed and maintained on a regular basis so it does not cause shadows or reduce the visibility generated from the luminaire. Luminaire poles should have a breakaway design and should be located so as to minimize the risk of being struck by a vehicle.

Treatment Description

A single luminaire is a safety treatment used to reduce nighttime crashes by making drivers aware of the presence of an intersection that may otherwise be difficult to see at night. For drivers on a stop-controlled approach, the lighting may provide additional time for the approaching driver to perceive the need to stop by increasing the visibility of features located at the site such as pavement markings and signs. For drivers on the uncontrolled approach, the lighting may provide an indication that drivers may be entering the roadway at that location and can improve the visibility of vehicles, bicyclists, and pedestrians located near the intersection. The luminaire may be pole-mounted near one corner of the intersection, or may be wire-mounted over the intersection. Example installations of the single luminaire treatment at rural intersections are shown in Figure 4.

This group of three photos shows three examples of single luminaire lighting at rural intersections.  In the upper photo, the single luminaire is wire-mounted above the center of a 3-leg intersection.  In the lower left photo, a single luminaire with a solar power panel is pole mounted at the corner of an intersection.  In the lower right photo, a single luminaire is pole mounted with a traditional power source in the corner of an intersection.
Figure 4. Photo. Examples of single luminaire intersection lighting (Image credit: Google Earth™ Mapping Service).(21)

Safety Evaluation

The safety effectiveness of installing a single luminaire at an intersection with minor-road stop control on a rural two-lane road was estimated using the EB before-after analysis approach as discussed in Chapter 2 (Analysis Approach). The illuminance or luminance level generated from the single luminaire was not considered in the safety evaluation as neither measure was recorded in the field at the

treatment sites. The descriptive statistics, research methodology, and analysis results of the safety evaluation are presented below.

Descriptive Statistics

A total of 27 treatment and 61 nontreatment sites in Minnesota were available for analysis of the safety effectiveness of a single luminaire installation. Their breakdown by number of intersection legs is shown in Table 5.

Table 5. Single luminaire: number of sites by number of intersection legs.
State Number of Intersection Legs Number of Sites
Treatment Nontreatment
MN 3 21 21
4 6 40
All legs 27 61

Crash and traffic volume data were available for varying periods before and after treatment installation, depending on treatment installation date at the individual sites. Only nighttime crashes, at four crash severity levels-total, fatal and severe injury (FS), fatal and all injury (FI), and property damage only (PDO)-were considered in the analysis of this treatment. Table 6 (treatment intersections) and Table 7 (nontreatment intersections) summarize the crash data used in the analysis. They present nighttime crash data summed across all intersections of a given configuration.

Table 6. Single luminaire: summary of nighttime crash statistics for the before and after treatment periods for treatment intersections by number of intersection legs.
State Number
of Legs
Before Period After Period
Years of Data in State Number of Sites Number of
Site-Years
Nighttime Crash Counts Range of Years of Data in State Number of Sites Number of Site-Years Nighttime Crash Counts
Total FS FI PDO Total FS FI PDO
MN 3 5 21 105 18 0 7 11 1 to 3 21 34 1 1 1 0
4 5 6 30 2 0 0 2 1 to 2 6 9 0 0 0 0
 
Table 7. Single luminaire: summary of nighttime crash statistics for the entire study period for nontreatment intersections by number of intersection legs.

State

Number of Legs

Entire Study Period

Years of Data in State

Number of Sites

Number of Site-Years

Nighttime Crash Counts

Total

FS

FI

PDO

MN

3

9

21

189

16

0

6

10

4

8

40

320

11

0

3

8

Methodology

The safety effectiveness of installing a single luminaire was evaluated using an EB before-after method. To implement this method, the following steps were taken:

  1. Select appropriate SPF: The SPFs for intersections on rural two-lane roads from Chapter 10 of the HSM were selected. These are given for total crashes only. The coefficients of these SPFs vary by number of intersection legs. Use of the intersection SPFs from Chapter 10 of the HSM provides an estimate of the intersection-related predicted crash frequency for sites included in the analysis in the absence of the treatment.
  2. Obtain the proportions of target crashes (PR1) relevant to the evaluation: The proportion of nighttime crashes to all crashes for each severity level (total, FS, FI, and PDO) were calculated using all crashes from nontreatment sites and from the before-period years of treatment sites in Minnesota. These proportions were calculated separately for 3- and 4-leg intersections. These proportions scale the total crash predictions (i.e., all crash types) to predictions for the target crashes (i.e., nighttime crashes).
  3. Obtain the proportions of FS, FI, and PDO crashes (PR2): These proportions were calculated as the ratio of all (daytime plus nighttime) FS, FI, or PDO crashes over total crashes using all crashes from the nontreatment sites and from the before-period years of treatment sites in Minnesota. These proportions were calculated separately for 3- and 4-leg intersections. These proportions scale the total crash predictions (i.e., all severity levels combined) to predicted crashes for specific severity level crashes (i.e., FS, FI, and PDO).
  4. Calibrate the SPFs to the local jurisdiction: Calibration was performed using all crashes (total daytime plus nighttime crashes), separately for each intersection configuration, again using all nontreatment intersections and before-period years from treatment intersections combined. Total crash counts rather than target crashes were used due to the scarcity of nighttime crashes, especially FS and FI crashes. The calibration factor adjusts the HSM SPFs for varying conditions in the local jurisdiction such as crash reporting thresholds, environmental factors, etc.

The SPFs presented in the HSM for intersections on rural two-lane roads for total severity level (i.e., all severity levels combined) have the general form:

Predicted crashes/yr = exp[a + b(lnAADTmaj) + c(lnAADTmin)]

Figure 5. Equation. Base model for predicted crashes per year.

where a, b, and c are the regression coefficients shown in Table 8. These coefficients apply to base conditions and vary by number of intersection legs. For the intersection SPFs in Chapter 10 of the HSM, the base conditions are:

Crash modification factors (CMFs), calibration factor (Cr), proportions of nighttime crashes (PR1), and proportions of FS, FI, and PDO crashes (PR2) were then used to adjust for local conditions as follows:

Predicted crashes/yr = {exp[a + b(lnAADTmaj) + c(lnAADTmin)]} × PR1 × PR2 × CMFCombined × Cr

Figure 6. Equation. Model for predicted crashes per year for specific crash types and severity levels, and accounting for local conditions.

The CMFCombined is the product of the CMFs from Chapter 10 of the HSM for skew angle (CMF1i), number of major-road left-turn lanes (CMF2i), and number of major-road right-turn lanes (CMF3i), for a particular intersection configuration (note that the CMF for intersection lighting, CMF4i , equals 1 in all cases for this treatment evaluation).

SPF coefficients (a, b, and c and overdispersion parameter), nighttime crash proportions (PR1), proportions of FS, FI, and PDO crashes (PR2), and calibration factors (Cr) are shown for each intersection configuration in Table 8. The table also shows the default proportions of PR1 and PR2 presented in Chapter 10 of the HSM (see HSM Tables 10-15 and 10-5, respectively). Note that PR2 is always equal to 1 for total crashes. The decision of which proportions to use-those calculated from the data or those provided by the HSM-was based on whether calculated proportions of nighttime crashes (PR1) were nonzero for all severity levels. If PR1 was equal to zero for any severity level, then the default HSM proportions (both PR1 and PR2) were used in the EB before-after analysis. The selection of which proportions were used is indicated in Table 16 by an asterisk.

Table 8. Single luminaire: SPF coefficients, target crash proportions, and calibration factors by number of intersection legs for EB analysis.
Number of Legs Number of Site- Years Severity Level Number of Crashes, All Number of Crashes, Nighttime Only Intercept (a) lnAADTmaj Coefficient (b) lnAADTmin Coefficient (c) Overdispersion Parameter Proportion of Nighttime Crashes (PR1) PR1 HSM Proportion of FS, FI, or PDO of Total Crashes (PR2) PR2 HSM Calibration Factor (Cr)
3 315 Total 69 34 -9.86 0.79 0.49 0.54 0.49 0.26* 1.00 1.00* 0.60
FS 1 0 0.00 0.26* 0.01 0.06*
FI 24 13 0.54 0.26* 0.35 0.42*
PDO 45 21 0.47 0.26* 0.65 0.59*
4 390 Total 63 13 -8.56 0.60 0.61 0.24* 0.21 0.24* 1.00 1.00* 0.23
FS 3 0 0.00 0.24* 0.05 0.06*
FI 28 3 0.11 0.24* 0.44 0.43*
PDO 35 10 0.29 0.24* 0.56 0.57*
NOTE 1: Asterisked proportions were those used in the analyses.

Safety Effectiveness

The EB analysis was based on before and after traffic volumes and crash data from 27 treatment and 61 nontreatment intersections in Minnesota, and HSM SPFs for 3- and 4-leg intersections on rural two-lane roads. Traffic volumes at the treatment sites ranged from 200 to 7,300 veh/day (3-leg intersections) and from 315 to 2,100 (4-leg intersections) on the major-road approaches and from 30 to 2,000 veh/day (3-leg intersections) and from 210 to 1,750 (4-leg intersections) on the minor-road approaches. The EB method was first applied separately to 3-leg and 4-leg intersections. However, only six 4-leg intersections were available, and none experienced any crashes in the after period. It was, therefore, decided to pool 3- and 4-leg intersections to estimate the safety effectiveness of installing a single luminaire across both 3- and 4-leg intersections combined. The EB before-after analysis results are shown in Table 9 for the following crash types and severity levels:

Although the analyses for FS injury crashes were performed, the analysis results were not considered reliable and are therefore not shown. The occurrence of FS crashes was too rare across all intersections in the study (both treatment and nontreatment sites). The statistics shown in Table 9 for each crash severity are: A negative percent safety effectiveness indicates that crash frequencies decreased due to the treatment.
Table 9. Single luminaire: safety effectiveness on nighttime crashes for 3- and 4-leg intersections combined.
Crash Severity Number of Treatment Sites Safety Effectiveness (%) SE of Treatment Effect (%) Significance
3- and 4-Leg Intersections Combined
Total 27 -71 29 Significant at 95% CL
FI -21 79 Not significant at 90% CL
PDO -100 a NC NC
a Crashes recorded in before period; none in after period. NC=Not Calculated; standard error and significance could not be estimated.

The following general observations can be made based on Table 6, Table 8, and Table 9:

The EB analysis results from Minnesota sites indicate that installing a single luminaire reduced total nighttime crashes by 71 percent; this safety effect is statistically significant at the 95-percent confidence level. Although the analysis shows a 21-percent reduction in FI nighttime crashes, the result is not significant at the 90-percent confidence level due to the large standard error of the estimate.

The EB analysis made use of AADT data for the entire day. Traffic volume information for nighttime hours was not incorporated into the analyses, so interpretation of the analysis results should be taken with this analysis approach in mind.

Economic Analysis

An economic analysis was conducted to calculate benefit-cost ratios to estimate the economic benefits of installing a single luminaire at 3-leg and 4-leg stop-controlled intersections on rural two-lane roads. The Economic Analysis discussion in Chapter 2 describes the procedure used to calculate the benefit-cost ratios.

The estimated crash reduction benefit is calculated using the CMF for total nighttime crashes produced in this study for installing a single luminaire at an intersection, which is 0.29 for 3- and 4-leg stop-controlled intersections combined. The following equation uses the proportion of total nighttime crashes at 3-leg and 4-leg rural stop-controlled intersections to translate the single luminaire CMF, that applies to nighttime crashes, to a CMF that applies to total crashes during the entire 24-hour day:

CMFTotal = (CMFLighting - 1) × PR1 + 1

Figure 7. Equation. CMF for total crashes.

where PR1 is the proportion of total crashes that occur at night. The HSM provides default values for PR1 (shown as pni). The value used in this economic evaluation is the HSM default provided in Table 10-15 in Chapter 10, which is 0.26 for 3-leg stop-controlled intersections and 0.24 for 4-leg stop-controlled intersections.

To calculate the treatment cost for the economic analysis, both installation and maintenance costs were included. The calculations assume an $8,000 installation cost for a single luminaire, an annual energy cost of $300, and a service life of 20 years. These estimates were obtained from the state DOTs that provided data for the safety analysis.

Table 10 displays the benefit-cost ratios for installing a single luminaire at a 3-leg stop-controlled intersection on rural two-lane roads. The AADTs in the table cover the range of AADTs of the study sites used for the estimation of the CMF. At very low intersection volumes, the treatment does not always produce a benefit-cost ratio above 1.0. However, single luminaire installation is economically justified at intersections with a major-road AADT at or above 1,000 veh/day, regardless of the minor-road AADT.

Table 11 presents the benefit-cost ratios for installing a single luminaire at a 4-leg stop-controlled intersection on rural two-lane roads. The AADTs in the table cover the range of AADTs of the study sites used for the estimation of the CMF. The single luminaire is economically justified at all AADT levels represented in the study, with all the benefit-cost ratios exceeding 3.0.

>Table 10. Single luminaire: benefit-cost ratios for treatment at 3-leg rural stop-controlled intersection.

Major- Road AADT

% of Major-Road AADT on Minor Road

5%

10%

15%

20%

25%

30%

35%

40%

45%

50%

60%

70%

80%

90%

300

N/A

0.5

0.5

0.6

0.7

0.8

0.8

0.9

0.9

1.0

1.1

1.2

1.2

1.3

1,000

1.5

2.1

2.6

3.0

3.3

3.6

3.9

4.1

4.4

4.6

5.1

5.5

5.8

6.2

2,000

3.6

5.1

6.2

7.2

8.0

8.8

9.4

10.1

10.7

11.2

12.3

13.3

14.2

15.0

3,000

6.1

8.6

10.5

12.1

13.4

14.7

15.9

16.9

17.9

18.9

20.7

N/A

N/A

N/A

4,000

8.8

12.4

15.1

17.4

19.4

21.3

22.9

24.5

25.9

27.3

N/A

N/A

N/A

N/A

5,000

11.8

16.5

20.1

23.2

25.9

28.3

30.5

32.6

N/A

N/A

N/A

N/A

N/A

N/A

5,900

14.5

20.4

24.9

28.7

32.0

35.0

N/A

N/A

N/A

N/A

N/A

N/A

N/A

N/A

CMF = 0.82 (total crashes, entire day), CMFLighting = 0.29 (total night crashes), PR1 = 0.26 (proportion of nighttime crashes to total crashes)

Installation cost = $8,000

Annual energy cost = $300

Service life = 20 yrs

N/A indicates the combination of major- and minor-road AADTs was not represented in the study.

 

Table 11. Single luminaire: benefit-cost ratios for treatment at 4-leg rural stop-controlled intersection.
Major- Road AADT % of Major-Road AADT on Minor Road
10% 15% 20% 25% 30% 40% 50% 55% 60% 70% 80% 85% 90% 95%
400 N/A; N/A N/A N/A N/A N/A N/A 3.3 3.4 3.8 4.1 4.3 4.4 4.6
1,000 N/A N/A N/A 6.1 6.8 8.1 9.3 9.9 10.4 11.4 12.4 12.9 13.3 13.8
2,000 N/A 10.3 12.3 14.1 15.8 18.8 21.6 22.9 24.1 26.5 28.7 29.8 N/A N/A
2,100 8.6 11.0 13.1 15.0 16.8 20.0 22.9 24.2 25.6 28.1 30.5 N/A N/A N/A

CMF = 0.83 (total crashes, entire day), CMFLighting = 0.29 (total night crashes), PR1 = 0.24 (proportion of nighttime crashes to total crashes)

Installation cost = $8,000

Annual energy cost = $300

Service life = 20 yrs

N/A indicates the combination of major- and minor-road AADTs was not represented in the study.

Implementation

As indicated in Volume 5 of the NCHRP Report 500 Series, A Guide for Addressing Unsignalized Intersection Collisions (5), the primary purposes of providing lighting at an intersection to reduce nighttime crashes are to:

Agencies should consider installation of a single luminaire at intersections with minor-road stop control on rural two-lane roads where the proportion of nighttime crashes to total crashes is above the statewide average for this intersection type, or simply at intersections of this type that have a moderate to high frequency of nighttime crashes. Except at intersections with very low major-road AADTs (e.g., 3-leg intersections with minor-road AADTs less than 300 veh/day), this treatment is economically justifiable. In particular, agencies may want to consider installation of a single luminaire at intersection locations where view of the intersection may be obstructed by horizontal curves or significant skew angles. (This assumes that the skew angle for an intersection is defined as the absolute value of the deviation from an intersection angle of 90 degrees.) At these locations, a single luminaire may increase driver awareness of the intersection and available sight distance.

For agencies considering installation of a single luminaire at intersections with minor-road stop control on rural two-lane roads, the FHWA Lighting Handbook may be a useful reference.(22) The handbook provides guidance to designers and state, city, and town officials concerning the application of roadway lighting. In particular, it includes several examples of warrants others have developed to prioritize the need for intersection lighting. For example, Table 12 illustrates a simple approach developed by Preston and Schoenecker for establishing priorities for installing lighting at rural intersections based on traffic volumes and major-road functional classification.(11)

Table 12. Prioritization of lighting at rural intersections by traffic volume and functional class.(11)
Priority Major-Road Functional Classification
Principal Arterial Minor Arterial Collector Local
Major-Road Volumes in veh/day
(% of Major-Street Volume that is Recommended on the Minor-Street)
Low 0-2,000 (10%) 0-1,000 (10%) 0-5,000 (10%) 0-250 (10%)
Moderate 2,000-5,000 (15%) 1,000-2,000 (15%) 500-1,000 (15%) 250-500 (15%)
High >5,000 (20%) >2,000 (20%) >1,000 (20%) >500 (20%)

Additionally, the AASHTO Roadway Lighting Design Guide may serve as a useful reference.(23) While it does not provide specific information regarding the need for and design of intersection lighting, it reflects current practices in roadway lighting.

Chapter 4. Transverse Rumble Strips

Treatment Name

Transverse Rumble Strips

Description Grooves or elevated strips placed in the travel lane perpendicular to the direction of travel, generally used in sets in advance of a stop sign to bring drivers' attention to the stop ahead condition.
States

Data from the following states were included in the safety evaluation of this treatment: Arkansas, Kansas, Missouri, North Dakota, and Oregon.

Safety Effectiveness

Reduction by severity level:

The estimated safety effects of this treatment in reducing total crashes (all crash types combined) at 3-leg and 4-leg intersections are as follows:

3-Leg Intersections:

· 37% reduction in FI crashes
(SE = 20%)

4-Leg Intersections:

· 13% reduction in total crashes

(SE = 7%)

· 29% reduction in FI crashes

(SE = 8%)

Reduction by crash type:

The estimated safety effects of this treatment in reducing angle and rear-end crashes at 3-leg and 4-leg intersections are as follows:

3-Leg Intersections:

Angle crashes

· 61% reduction in PDO crashes

(SE = 28%)

Rear-end crashes

· 60% reduction in FI crashes

(SE = 29%)

4-Leg Intersections:

Angle crashes

· 25% reduction in FI crashes

(SE = 10%)

Rear-end crashes

· 56% reduction in total crashes

(SE = 8%)

· 78% reduction in FI crashes

(SE = 8%)

· 54% reduction in PDO crashes

(SE = 10%)

Cost and Economic Benefits

Benefit-cost ratios calculated for 4-leg intersections ranged from 1.1 to 241.1 when assuming installation costs between $1,000 and $5,000 per intersection approach.

Where to Implement

The treatment should be considered for stop-controlled approaches to intersections where crash patterns indicate that drivers fail to recognize the stop condition (e.g., angle crashes related to stop sign violations). Rumble strips may be particularly effective on the stop-controlled approach to an intersection that is hidden from view due to horizontal or vertical curvature. Intersections following a long tangent section may also benefit from this treatment.

The economic analysis indicates this treatment is economically justifiable, even at intersections with low traffic volumes.

The proximity of the intersection to nearby residents or businesses should be considered prior to selecting this treatment for implementation, as noise generated from the rumble strips may result in complaints from nearby residents.

Additional Factors for Consideration

Typically, transverse rumble strips are considered for implementation after less intrusive measures have been tried and failed to improve the crash experience at an intersection.

Several unintended consequences of this treatment may occur, and should be considered prior to any decision to implement this treatment, including: (a) potential loss-of-control problems for motorcyclists and bicyclists, (b) difficulties associated with snowplow operations, and (c) inappropriate driver responses such as using the opposing travel lanes to drive around the rumble strips (5).

Treatment Description

Transverse rumble strips are grooves in the roadway surface or elevated strips placed in the travel lane perpendicular to the direction of travel. They are designed to generate noise and vibration in the vehicle as the driver crosses over them to alert the driver to a condition that may require attention or action. The specific application of transverse rumble strips evaluated in this research is their placement on the stop-controlled approach to intersections on rural two-lane roads. The rumble strips are placed at a distance in advance of the stop sign sufficient to allow the driver time to perceive the need to stop, react to that need, and brake appropriately.

Transverse rumble strips are generally placed in sets of several closely-placed strips to form a set, and sometimes more than one set is used on a given approach. The strips may be located only in the wheel path or across the full lane width. They are often used in conjunction with stop ahead signing.

Transverse rumble strips may be rolled or grooved into asphalt, formed into fresh concrete, or created as epoxy strips on the surface of the pavement. Figure 8 shows pictures of rumble strips milled into asphalt.

Safety Evaluation

The safety effectiveness of transverse rumble strips installed on the stop-controlled approaches of intersections on rural two-lane roads was estimated using the EB before-after analysis approach as discussed in Chapter 2 (Analysis Approach). The descriptive statistics, research methodology, and analysis results are presented below.

Descriptive Statistics

A total of 72 treatment and 126 nontreatment sites in five states-Arkansas, Kansas, Missouri, North Dakota, and Oregon-were available for analysis of the safety effectiveness of transverse rumble strips installed on stop-controlled approaches of intersections on rural two-lane roads. Their breakdown by state, number of intersection legs, and number of treated approaches (1 or 2; always 1 for 3-leg intersections) is shown in Table 13. Traffic volumes at the treatment sites ranged from 245 to 11,700 veh/day (3-leg intersections) and from 165 to 6,700 veh/day (4-leg intersections) on the major-road approaches and from 110 to 7,000 veh/day (3-leg intersections) and from 65 to 4,120 veh/day (4-leg intersections) on the minor-road approaches.

Crash and traffic volume data were obtained for varying before- and after-treatment installation periods, depending on treatment installation date at the individual sites. Three crash types were considered in this analysis: all collision types combined, angle crashes, and rear-end crashes.

Table 14 (treatment intersections) and Table 15 (nontreatment intersections) summarize the crash data used in the analysis. They present total, angle, and rear-end crash data summed across all intersections of a given configuration (3-leg or 4-leg) within each state.

This group of four photos show transverse rumble strips from various perspectives. All examples are milled into asphalt pavement.  The upper photo shows a close up of a set of rumble strips just upstream of a STOP AHEAD sign. The middle photo shows an aerial image of two sets of rumble strips at an intersection approach—one approximately 200 ft upstream of the stop sign, and the other approximately 500 ft upstream of the stop sign. The lower two photos show additional examples similar to those shown in the photos already described.
Figure 8. Photo. Examples of transverse rumble strips placement (Image credit: Google Earth™ Mapping Service).(21)

Table 13. Transverse rumble strips: number of sites by State, number of intersection legs, and number of treated approaches.

State

Number of
Intersection
Legs

Number of Treated Approaches

Number of Sites

m

Treatment

Nontreatment

AR

4

1

1

19

2

1

KS

4

1

2

18

MO

3

1

5

6

4

1

5

31

2

10

NDa

3

1

17

13

4

1

20

17

2

6

OR

3

1

3

8

4

1

1

14

2

1

All sites

72

126

a No control intersections were available in North Dakota; nontreatment sites from Nebraska were used in the analysis.

 

Table 14. Transverse rumble strips: summary crash statistics for the before and after treatment periods for treatment intersections by State and number of intersection legs.

Number
of Legs

State

Before Period

After Period

Range of Years of Data in State

Number
of Sites

Number of Site-Years

Crash Counts

Range of Years of Data in State

Number of Sites

Number of Site-Years

Crash Counts

Total

FS

FI

PDO

Total

FS

FI

PDO

All Crashes

3

MO

4 to 5

5

23

12

0

7

5

3 to 9

5

35

19

2

7

12

ND

5

17

85

9

0

2

7

2 to 3

17

41

4

0

2

2

OR

5

3

15

9

0

3

6

3 to 8

3

16

8

0

2

6

4

AR

5

2

10

14

6

10

4

9 to 10

2

19

14

2

7

7

KS

5

2

10

4

2

2

2

1 to 2

2

3

0

0

0

0

MO

3 to 5

15

67

75

12

42

33

4 to 10

15

111

174

28

77

97

ND

5

26

130

14

0

7

7

2 to 3

26

62

11

0

2

9

OR

5

2

10

14

0

7

7

8 to 9

2

17

28

3

19

9

Angle Crashes

3

MO

4 to 5

5

23

3

0

3

0

3 to 9

5

35

5

0

3

2

ND

5

17

85

0

0

0

0

2 to 3

17

41

0

0

0

0

OR

5

3

15

0

0

0

0

3 to 8

3

16

0

0

0

0

4

AR

5

2

10

9

3

6

3

9 to 10

2

19

7

2

4

3

KS

5

2

10

3

2

2

1

1 to 2

2

3

0

0

0

0

MO

3 to 5

15

67

53

11

36

17

4 to 10

15

111

115

23

56

59

ND

5

26

130

2

0

1

1

2 to 3

26

62

4

0

1

3

OR

5

2

10

6

0

2

4

8 to 9

2

17

9

2

9

0

Rear-End Crashes

3

MO

4 to 5

5

23

3

0

2

1

3 to 9

5

35

6

1

1

5

ND

5

17

85

0

0

0

0

2 to 3

17

41

1

0

1

0

OR

5

3

15

1

0

0

1

3 to 8

3

16

2

0

0

2

4

AR

5

2

10

2

2

2

0

9 to 10

2

19

5

0

2

3

KS

5

2

10

0

0

0

0

1 to 2

2

3

0

0

0

0

MO

3 to 5

15

67

9

1

3

6

4 to 10

15

111

17

0

3

14

ND

5

26

130

5

0

2

3

2 to 3

26

62

3

0

0

3

OR

5

2

10

0

0

0

0

8 to 9

2

17

5

0

3

2

 

Table 15. Transverse rumble strips: summary crash statistics for the entire study period for nontreatment intersections by State and number of intersection legs.

Number of Intersection Legs

State

Entire Study Period

Range of Years of Data in State

Number of Sites

Number of Site-Years

Crash Counts

Total

FS

FI

PDO

All Crashes

3

MO

14

6

84

8

0

2

6

ND

9

13

117

24

3

7

17

OR

14

8

112

71

1

42

29

4

AR

16

19

304

160

35

96

64

KS

8

18

144

25

1

11

14

MO

14

31

434

189

30

100

89

NDa

9

17

153

41

8

26

15

OR

15

14

210

88

0

52

36

Angle Crashes

3

MO

14

6

84

2

0

0

2

ND

9

13

117

4

2

3

1

OR

14

8

112

0

0

0

0

4

AR

16

19

304

98

24

63

35

KS

8

18

144

10

1

7

3

MO

14

31

434

102

24

65

37

NDa

9

17

153

21

6

16

5

OR

15

14

210

29

0

20

9

Rear-End Crashes

3

MO

14

6

84

1

0

0

1

ND

9

13

117

2

0

0

2

OR

14

8

112

18

0

9

9

4

AR

16

19

304

15

3

8

7

KS

8

18

144

3

0

1

2

MO

14

31

434

24

1

8

16

NDa

9

17

153

5

2

3

2

OR

15

14

210

15

0

8

7

a No control intersections were available in North Dakota; nontreatment sites from Nebraska were used in the analysis.

Methodology

The safety effectiveness of installing transverse rumble strips was evaluated using an EB method similar to that discussed in Chapter 3 for the EB evaluation of a single luminaire installation. Prior to implementing the EB method, the following points were addressed:

  1. Select appropriate SPF: The SPFs for intersections on rural two-lane roads from Chapter 10 of the HSM were selected. These are given for total crashes only. The coefficients of these SPFs vary by number of intersection legs. Use of the intersection SPFs from Chapter 10 of the HSM provide an estimate of the intersection-related predicted crash frequency for sites included in the analysis in the absence of the treatment.
  2. Obtain the proportion of target crashes for total, FS, FI, and PDO crashes (PR1): The proportions of target crashes (angle and rear end) to all crashes for each severity level (total, FS, FI, and PDO) were calculated using all crashes from nontreatment sites and from the before-period years of treatment sites. These proportions were calculated separately for 3- and 4-leg intersections. These proportions scale the total crash predictions (i.e., all crash types) to predictions for the target crashes (i.e., angle and rear-end crashes).
  3. Obtain the proportions of FS, FI, and PDO crashes (PR2): These proportions were calculated as the ratio of all FS, FI, or PDO crashes over total crashes using all crashes from the nontreatment sites and from the before-period years of treatment sites. These proportions were calculated separately for 3- and 4-leg intersections. These proportions scale the total crash predictions (i.e., all severity levels combined) to predicted crashes for specific severity level crashes (i.e., FS, FI, and PDO).
  4. Calibrate the SPFs to the local jurisdiction: Calibration was performed using all crashes (all collision types combined), separately for each intersection configuration within a given state, again using all nontreatment intersections and before treatment intersections combined. Total crash counts were used rather than target crashes due to the scarcity of target crashes, especially FS and FI angle and rear-end crashes. The calibration factor adjusts the HSM SPFs for varying conditions in the local jurisdiction such as crash reporting thresholds, environmental, etc.

The SPFs presented in the HSM for intersections on rural two-lane roads for total severity level (i.e., all severity levels combined) have the general form:

Predicted crashes/yr = exp[a + b(lnAADTmaj) + c(lnAADTmin)]

Figure 9. Equation. Base model for predicted crashes per year.

where a, b, and c are the regression coefficients shown in Table 16. These coefficients apply to base conditions and vary by number of intersection legs. For the intersection SPFs in Chapter 10 of the HSM, the base conditions are:

Crash modification factors (CMFs), calibration factors (Cr), proportions of angle and rear-end crashes (PR1), and proportions of FS, FI, and PDO crashes (PR2) were then used to adjust for local conditions as follows:

Predicted crashes/yr = {exp[a + b(lnAADTmaj) + c(lnAADTmin)]} × PR1 × PR2 × CMFCombined × Cr

Figure 10. Equation. Model for predicted crashes per year for specific crash types and severity levels, and accounting for local conditions.

The CMFCombined is the product of the CMFs from Chapter 10 of the HSM for skew angle (CMF1i), number of major-road left-turn lanes (CMF2i), and number of major-road right-turn lanes (CMF3i), for a particular intersection configuration.

SPF coefficients (a, b, and c and overdispersion parameter), target crash proportions (PR1), proportions of FS, FI, and PDO out of total crashes (PR2), and calibration factors (Cr) are shown for each intersection configuration in Table 16. Number of site-years, total crash counts (all severity levels), and target crash counts are also displayed. The table also shows the default proportions of PR1 and PR2 presented in Chapter 10 of the HSM (see Tables 10-6 and 10-5, respectively). Note that PR2 is always equal to 1 for total crashes. The decision of which proportions to use-those calculated from the data or those provided by the HSM-was based on whether calculated proportions of target crashes (PR1) were nonzero for all severity levels. If they were not, then the HSM proportions (both PR1 and PR2) were used in the EB before-after analysis. The selection of which proportions were used is indicated in Table 16 by an asterisk.

The calibration factors for the SPFs shown in Table 16 ranged from 0.53 to 6.98, meaning that several of the sites used in this analysis had substantially different crash experience than the sites used to develop the SPFs. For example, 4-leg intersections in Arkansas had a calibration factor of 6.98, indicating that those intersections experienced approximately 7 times more crashes than the 4-leg intersections used to develop the SPFs in the HSM.

Safety Effectiveness

The EB before-after method was applied to estimate the safety effectiveness of installing transverse rumble strips on stop-controlled approaches to intersections on rural two-lane roads. The analysis included treatment and nontreatment sites in Arkansas, Kansas, Missouri, Nebraska (nontreatment sites only), North Dakota (treatment sites only), and Oregon, and used HSM SPFs for 3- and 4-leg intersections on rural two-lane roads.

The EB before-after analysis results are shown in Table 17 for the following crash types and severity levels:

Table 16. Transverse rumble strips: SPF coefficients, target crash proportions, and calibration factors by State and number of intersection legs.

State

Number of Legs

Number of Site-Years

Severity Level

Number of Crashes, All

Target Crash Type

Number of Target Crashes

Intercept (a)

lnAADTmaj Coefficient (b)

lnAADTmin Coefficient (c)

Overdispersion Parameter

Proportion of Target Crashes (PR1)

PR1 HSM

Proportion of FS, FI, or PDO of Total Crashes (PR2)

PR2 HSM

Calibration Factor (Cr)

AR

4

314

Total

174

Angle

107

 

 

 

-8.56

 

 

 

 

 

 

 

0.60

 

 

 

 

 

 

 

0.61

 

 

 

 

 

 

 

0.24

 

 

 

 

0.62*

0.43

1.00*

1.00

 

 

 

6.98

 

 

 

 

Rear End

17

0.10*

0.24

1.00*

1.00

FS

41

Angle

27

0.66*

0.53

0.24*

0.06

Rear End

5

0.12*

0.21

0.24*

0.06

FI

106

Angle

69

0.65*

0.53

0.61*

0.43

Rear End

10

0.09*

0.21

0.61*

0.43

PDO

68

Angle

38

0.56*

0.35

0.39*

0.57

Rear End

7

0.10

0.27

0.39

0.57

KS

4

154

Total

29

Angle

13

 

 

 

-8.56

 

 

 

 

 

 

 

0.60

 

 

 

 

 

 

 

0.61

 

 

 

 

 

 

 

0.24

 

 

 

 

0.45

0.43*

1.00

1.00*

 

 

 

1.86

 

 

 

 

Rear End

3

0.10

0.24*

1.00

1.00*

FS

3

Angle

3

1.00

0.53*

0.10

0.06*

Rear End

0

0.00

0.2*

0.10

0.06*

FI

13

Angle

9

0.69

0.53*

0.45

0.43*

Rear End

1

0.08

0.21*

0.45

0.43*

PDO

16

Angle

4

0.25

0.35*

0.55

0.57*

Rear End

2

0.13

0.27*

0.55

0.57*

MO

3

107

Total

20

Angle

5

 

 

 

-9.86

 

 

 

 

 

 

 

0.79

 

 

 

 

 

 

 

0.49

 

 

 

 

 

 

 

0.54

 

 

 

 

0.25

0.24*

1.00

1.00*

 

 

 

0.53

 

 

 

 

Rear End

4

0.20

0.28*

1.00

1.00*

FS

0

Angle

0

0.00

0.28*

0.00

0.06*

Rear End

0

0.00

0.26*

0.00

0.06*

FI

9

Angle

3

0.33

0.28*

0.45

0.42*

Rear End

2

0.22

0.26*

0.45

0.42*

PDO

11

Angle

2

0.18

0.21*

0.55

0.59*

Rear End

2

0.18

0.29*

0.55

0.59*

4

501

Total

264

Angle

155

 

 

 

-8.56

 

 

 

 

 

 

 

0.60

 

 

 

 

 

 

 

0.61

 

 

 

 

 

 

 

0.24

 

 

 

 

0.59*

0.43

1.00*

1.00

 

 

 

1.01

 

 

 

 

Rear End

33

0.13*

0.24

1.00*

1.00

FS

42

Angle

35

0.83*

0.53

0.16*

0.06

Rear End

2

0.05*

0.21

0.16*

0.06

FI

142

Angle

101

0.71*

0.53

0.54*

0.43

Rear End

11

0.08*

0.21

0.54*

0.43

PDO

122

Angle

54

0.44*

0.35

0.46*

0.57

Rear End

22

0.18*

0.27

0.46*

0.57

ND

3

 

 

 

 

 

 

 

202

 

 

 

 

 

 

 

Total

33

Angle

4

 

 

 

-9.86

 

 

 

 

 

 

 

0.79

 

 

 

 

 

 

 

0.49

 

 

 

 

 

 

 

0.54

 

 

 

 

0.12

0.24*

1.00

1.00*

 

 

 

0.97

 

 

 

 

Rear End

2

0.06

0.28*

1.00

1.00*

FS

3

Angle

2

0.67

0.28*

0.09

0.06*

Rear End

0

0.00

0.26*

0.09

0.06*

FI

9

Angle

3

0.33

0.28*

0.27

0.42*

Rear End

0

0.00

0.26*

0.27

0.42*

PDO

24

Angle

1

0.04

0.21*

0.73

0.59*

Rear End

2

0.08

0.29*

0.73

0.59*

4

 

 

 

 

 

 

 

 

 

 

 

 

283

 

 

 

 

 

 

 

Total

55

Angle

23

 

 

 

-8.56

 

 

 

 

 

 

 

0.60

 

 

 

 

 

 

 

0.61

 

 

 

 

 

 

 

0.24

 

 

 

 

0.42*

0.43

1.00*

1.00

 

 

 

0.55

 

 

 

 

Rear End

10

0.18*

0.24

1.00*

1.00

FS

8

Angle

6

0.75*

0.53

0.15*

0.06

Rear End

2

0.25*

0.21

0.15*

0.06

FI

33

Angle

17

0.52*

0.53

0.60*

0.43

Rear End

5

0.15*

0.21

0.60*

0.43

PDO

22

Angle

6

0.27*

0.35

0.40*

0.57

Rear End

5

0.23*

0.27

0.40*

0.57

OR

3

 

 

 

 

 

 

 

127

 

 

 

 

 

 

 

Total

80

Angle

0

 

 

 

-9.86

 

 

 

 

 

 

 

0.79

 

 

 

 

 

 

 

0.49

 

 

 

 

 

 

 

0.54

 

 

 

 

0.00

0.24*

1.00

1.00*

 

 

 

1.17

 

 

 

 

Rear End

19

0.24

0.28*

1.00

1.00*

FS

1

Angle

0

0.00

0.28*

0.01

0.06*

Rear End

0

0.00

0.26*

0.01

0.06*

FI

45

Angle

0

0.00

0.28*

0.56

0.42*

Rear End

9

0.20

0.26*

0.56

0.42*

PDO

35

Angle

0

0.00

0.21*

0.44

0.59*

Rear End

10

0.29

0.29*

0.44

0.59*

4

 

 

 

 

 

 

 

220

 

 

 

 

 

 

 

Total

102

Angle

35

 

 

 

-8.56

 

 

 

 

 

 

 

0.60

 

 

 

 

 

 

 

0.61

 

 

 

 

 

 

 

0.24

 

 

 

 

0.34

0.43*

1.00

1.00*

 

 

 

1.79

 

 

 

 

Rear End

15

0.15

0.24*

1.00

1.00*

FS

0

Angle

0

0.00

0.53*

0.00

0.06*

Rear End

0

0.00

0.21*

0.00

0.06*

FI

59

Angle

22

0.37

0.53*

0.58

0.43*

Rear End

8

0.14

0.21*

0.58

0.43*

PDO

43

Angle

13

0.30

0.35*

0.42

0.57*

Rear End

7

0.16

0.27*

0.42

0.57*

NOTE: Asterisked proportions were those used in the analyses.

 

Table 17. Transverse rumble strips: safety effectiveness on target crashes by number of intersection legs.

Crash Severity

Number of Treatment Sites

Safety Effectiveness (%)

SE of Treatment Effect (%)

Significance

All States Combined 3-Leg Intersections

All Crashes

Total

25

-18

16

Not significant at 90% CL

FI

-37

20

Significant at 90% CL

PDO

-13

21

Not significant at 90% CL

Angle Crashes

Total

25

-43

26

Not significant at 90% CL

FI

-44

33

Not significant at 90% CL

PDO

-61

28

Significant at 95% CL

Rear-End Crashes

Total

25

-16

29

Not significant at 90% CL

FI

-60

29

Significant at 95% CL

PDO

-6

37

Not significant at 90% CL

All States Combined; 4-Leg Intersections

All Crashes

Total

47

-13

7

Significant at 90% CL

FI

-29

8

Significant at 95% CL

PDO

-14

9

Not significant at 90% CL

Angle Crashes

Total

47

-13

8

Not significant at 90% CL

FI

-25

10

Significant at 95% CL

PDO

-13

12

Not significant at 90% CL

Rear-End Crashes

Total

47

-56

8

Significant at 95% CL

FI

-78

8

Significant at 95% CL

PDO

-54

10

Significant at 95% CL

Although the analyses for FS injury crashes were performed, the analysis results were not considered reliable and are therefore not shown. The occurrence of FS crashes was too rare across all intersections in the study (both treatment and nontreatment sites).

The EB method was applied to all states combined based on the following reasoning: (1) only a small number of treatment sites were available in some states-individually, they could not have been used for evaluation; and (2) all EB intermediate calculations (up to the final effectiveness calculations) are performed on a state/site basis, thus using SPFs, proportions (PR1 and PR2), combined CMFs, and calibration factors (Cr) specific to that state. This approach, to some extent, takes differences among states into account, while increasing site and crash sample sizes.

The EB method was applied separately to 3-leg and 4-leg intersections, but at 4-leg intersections, sites with treatment installation on a single approach and on both approaches were pooled to ensure a sufficient sample size for evaluation.

The statistics shown in Table 17 for each crash severity are:

A negative percent safety effectiveness indicates that crash frequencies decreased due to the treatment.

The following general observations can be made when looking at Table 14, Table 16, and Table 17 together:

Economic Analysis

An economic analysis was conducted to estimate the economic benefits of installing transverse rumble strips at stop-controlled intersections on rural two-lane roads. The economic analysis was based on calculations assuming installation of transverse rumble strips on both stop-controlled approaches of 4-leg intersections. The Economic Analysis discussion in Chapter 2 describes the procedure used for estimating the benefit-cost ratios of the treatment. The economic benefits of this treatment were estimated using the safety effectiveness estimates developed in this research.

The estimated crash reduction benefit of the treatment was calculated using the CMF for total crashes (all severity levels and crash types combined) for 4-leg intersections (CMF = 0.87), which was statistically significant at the 90-percent level. The CMF for total crashes for 3-leg intersections was not significant at the 90-percent level.

Cost and service life of transverse rumble strips may vary by geographic region and installation type. Two scenarios were used to calculate the installation costs of transverse rumble strips in this evaluation. In the first scenario, a $5,000 installation cost per approach and a 20-year life cycle were assumed. These figures were reported by the Oregon Department of Transportation.

Table 18 presents the benefit-cost ratios as a function of major- and minor-road AADTs for a 4-leg stop-controlled intersection on a rural two-lane highway where transverse rumble strips are installed on both stop-controlled approaches. The AADTs in the table cover the range of AADTs of the study sites used for the estimation of the CMF. All the benefit-cost ratios shown are equal to or greater than 1.0, meaning that the installation of transverse rumble strips is economically justified at all AADT ranges included in this study.

Table 18. Transverse rumble strips: benefit-cost ratios for treatment on two minor-road approaches at 4-leg rural stop-controlled intersection for $10,000 installation cost and 20-yr service life.

Major- Road AADT

% of Major-Road AADT on Minor Road

5%

10%

20%

30%

40%

45%

50%

60%

65%

70%

75%

80%

90%

95%

200

N/A

N/A

N/A

N/A

N/A

1.1

1.1

1.3

1.3

1.4

1.5

1.5

1.6

1.7

1,000

N/A

3.0

4.5

5.8

6.9

7.5

8.0

8.9

9.3

9.8

10.2

10.6

11.4

11.8

2,000

4.5

6.9

10.5

13.5

16.1

17.2

18.4

20.6

21.6

22.6

23.6

24.5

26.3

27.2

3,000

7.4

11.3

17.2

22.0

26.2

28.2

30.0

33.6

35.3

36.9

38.5

40.0

43.0

44.4

4,000

10.4

15.9

24.3

31.2

37.1

39.9

42.6

47.6

49.9

52.2

54.5

N/A

N/A

N/A

5,000

13.7

20.9

31.9

40.8

48.6

52.3

55.7

62.3

N/A

N/A

N/A

N/A

N/A

N/A

CMF = 0.87 (total crashes)

Installation cost = $10,000 (i.e., $5,000 per approach)

Service life = 20 yrs

N/A indicates the combination of major- and minor-road AADTs was not represented in the study

In the second scenario, a $1,000 installation cost per approach and a 5-yr service life were assumed. This installation cost was reported by the Kansas Department of Transportation for milled transverse rumble strips. Kansas did not report a service life for this treatment; however, for the sake of the analysis, a shorter service life was assumed. Assuming that a longer service life will only increase the benefit of the treatment, selecting a short service life represents a conservative approach for a benefit-cost analysis.

Table 19 displays the benefit-cost ratios as a function of major- and minor-road AADTs for a 4-leg stop-controlled intersection on a rural two-lane highway where transverse rumble strips are installed on both stop-controlled approaches. The AADTs in the table cover the range of AADTs of the study sites used for the estimation of the CMF. The installation of transverse rumble strips again are economically justified at all AADT levels, all benefit-cost ratios are greater than 4.0.

Table 19. Transverse rumble strips: benefit-cost ratios for treatment on two minor-road approaches at 4-leg rural stop-controlled intersection for $2,000 installation cost
and 5-yr service life.

Major- Road AADT

% of Major-Road AADT on Minor Road

5%

10%

20%

30%

40%

45%

50%

60%

65%

70%

75%

80%

200

N/A

N/A

N/A

N/A

N/A

4.1

4.4

4.9

5.2

5.4

5.6

5.8

1,000

N/A

11.5

17.6

22.5

26.9

28.9

30.8

34.4

36.1

37.8

39.4

41.0

2,000

17.5

26.7

40.7

52.1

62.1

66.8

71.2

79.6

83.5

87.4

91.2

94.8

3,000

28.5

43.6

66.5

85.1

101.5

109.0

116.3

130.0

136.5

142.8

148.9

154.9

4,000

40.4

61.7

94.2

120.6

143.7

154.4

164.7

184.1

193.3

202.2

210.9

N/A

5,000

53.0

80.8

123.4

158.0

188.3

202.3

215.7

241.1

N/A

N/A

N/A

N/A

CMF = 0.87 (total crashes)

Installation cost = $2,000 (i.e., $1,000 per approach)

Service life = 5 yrs

N/A indicates the combination of major- and minor-road AADTs was not represented in the study

Implementation

The primary purpose of installing transverse rumble strips on stop-controlled approaches to intersections on rural two-lane roads is to alert drivers to a condition that requires attention and action. In this case, the treatment is intended to increase drivers' awareness of the intersection and to the traffic control, as drivers are required to stop along the controlled approach before proceeding into the intersection. Transverse rumble strips should be considered for installation on stop-controlled approaches to intersections where a pattern of crashes is present related to a lack of driver recognition of the presence of a stop sign (e.g., angle crashes related to stop sign violations).(5) In particular, transverse rumble strips should be considered along the stop-controlled approach to an intersection that is hidden from view due to horizontal or vertical curvature.(24) In addition, transverse rumble strips should be considered along the stop-controlled approach to an intersection following a long tangent section as drivers may become less attentive to their environment or may underestimate their approach speed to the intersection having driven for a long time at a significantly higher speed. Typically, transverse rumble strips are considered for implementation after less intrusive measures have been tried and failed to improve the crash experience at an intersection. The economic analysis provided above shows that this treatment is economically justifiable, even along roads with low traffic volumes.

Noise generated from transverse rumble strips may disturb nearby residents or businesses. The proximity of the intersection to nearby residents or businesses should be considered prior to selecting this treatment for implementation.

There is not a single recommended design for installing transverse rumble strips on stop-controlled approaches to intersections on rural two-lane roads. Design details are provided below from three states to serve as examples of current implementation practice. Agencies may choose to adopt or modify one of these example designs when developing their own rumble strip policy. The examples illustrate that while the dimensions of the actual rumble strips themselves are fairly consistent, there is a great deal of variation in the placement of the strips and how many individual sets of rumble strips are used in advance of a stop condition.

Figure 11 illustrates the transverse rumble strip placement guidance used by Missouri Department of Transportation (MoDOT). Design details of the rumble strips themselves are shown in Figure 12.

Figure 13 illustrates North Dakota Department of Transportation's design details for transverse rumble strips, which includes six sets of rumble strips encountered by a driver approaching a stop condition. The dimensions of the milled rumble strips are as follows:

Figure 14 illustrates design guidance for transverse rumble strips provided by the Kansas Department of Transportation (KDOT). KDOT's policy is to include three sets of 25 lateral grooves provided in advance of the warning area. KDOT's policy recommends application of transverse rumble strips at intersections where three or more right-angle collisions have occurred in a 12-month period that involve stop sign violations, where the crash rate is higher than 15 crashes per 10 million entering vehicles, and where the previous intersection on that road requiring a stop or vehicle maneuver is more than 15.5 miles prior to the intersection.

This diagram shows MoDOT’s guidance for placement of transverse rumble strips.  The first panel should be 175 ft in advance of the STOP AHEAD sign.  The second panel should be placed 250 ft in advance of the stop sign for a 50 mph speed limit, 300 ft in advance of the stop sign for a 55 mph speed limit, 350 ft in advance of the stop sign for a 60 mph speed limit, and 425 ft in advance of the stop sign for a 65 mph speed limit.  For speeds below 50 mph, the second panel of rumble strips is not used. The location of the STOP AHEAD sign is placed in accordance with guidance from the MUTCD.
Figure 11. Diagram. MoDOT's design guidance for placement of transverse rumble strips.(24)

This diagram shows MoDOT’s design detail for transverse rumble strips.  Each panel is to consist of 25 grooves per 24-ft panel, with the center of each groove spaced 1-ft apart.  Grooves are 4 inches wide and 3/8 inches deep.  Grooves are placed at a 10-degree skew to a line perpendicular to the direction of travel.  Grooves begin 18 inches from the edge of pavement and end 12 inches from the centerline of the roadway.
Figure 12. Diagram. MoDOT's design detail for transverse rumble strips.(24)

This diagram shows North Dakota DOT’s design guidance for transverse rumble strips. Six panels of rumble strips are placed along the stop-controlled approach. Moving upstream from the stop sign, the first panel begins at 250 ft and the second panel begins at 315 ft. A junction sign is located between 600 and 1200 ft upstream of the stop sign.  The third panel of rumble strips begins between 70 and 170 ft upstream of the junction sign, and the fourth panel begins 65 ft upstream of that point.  These four panels are each 15 ft long with 16 grooves. The fifth panel of rumble strips begins 215 ft upstream of the fourth panel (200 ft between them panels) and the sixth panel begins 100 ft upstream of the second panel.  The fifth and sixth panels are each 25 ft long with 26 grooves. For all panels, the grooves are centered with one-foot spacing.  Grooves are 4 inches wide and ½ to 5/8 inches deep.  Grooves are placed perpendicular to the direction of travel.  Grooves begin 12 inches from the edge of pavement and end 12 inches from the centerline of the roadway.
Figure 13. Diagram. North Dakota DOT's transverse rumble strip design detail.(25)

 

This diagram shows KDOT’s design guidance for transverse rumble strips. Three panels of rumble strips are placed along the stop-controlled approach. The first panel begins 100 ft in advance of the STOP AHEAD warning sign. The second and third panels are each spaced 100 ft upstream of the previous panel.  These three panels are each 24 ft long with 25 grooves at 1-ft spacings. Grooves are 4 inches wide and 3/8 inches deep.  Grooves are placed at a 10-degree skew to a line perpendicular to the direction of travel.  Grooves begin 18 inches from the shoulder edge and end 6 inches from the centerline.
Figure 14. Diagram. KDOT transverse rumble strip installation and design detail.(26)

Chapter 5. Supplementary Pavement Markings

Treatment Name

Supplementary Pavement Markings

Description

This evaluation considered two applications of supplementary pavement markings. In the first application, STOP AHEAD pavement markings were installed on the stop-controlled approach to an intersection. These markings are used to alert drivers of the presence of the intersection and the need to stop before proceeding into the intersection. These markings are intended to reduce crashes related to a failure to stop at the intersection.

In the second application, supplementary pavement markings were installed on uncontrolled approaches to an intersection. These pavement markings include a graphical depiction of the intersection with one of the following symbols: ┼, ├, or ┤, and are used to alert drivers of the presence of the intersection and of the potential for vehicles turning onto or off the roadway at that location.

States

Data from the following states were included in the safety evaluation of this treatment: Arkansas, Minnesota, Nebraska, Pennsylvania, and Vermont.

Safety Effectiveness

 

Reduction by severity level:

The estimated safety effects of this treatment in reducing total crashes (all crash types combined) at 3-leg and 4-leg intersections are as follows:

Markings on stop-controlled approaches

3-Leg Intersections:

· 67% reduction of total crashes
(SE = 7%)

· 76% reduction in FI crashes
(SE = 7%)

· 72% reduction in PDO crashes
(SE = 7%)

4-Leg Intersections:

· 66% reduction in total crashes (SE = 4%)

· 69% reduction in FI crashes

(SE = 5%)

· 77% reduction in PDO crashes

(SE = 4%)

Reduction by crash type:

The estimated safety effects of this treatment in reducing angle and rear-end crashes at 3-leg and 4-leg intersections are as follows:

Markings on stop-controlled approaches

3-Leg Intersections:

Angle Crashes

· 92% reduction in total crashes
(SE = 5%)

· 88% reduction in FI crashes
(SE = 7%)

Rear-End Crashes

· 95% reduction in total crashes
(SE = 4%)

· 96% reduction in FI crashes
(SE = 5%)

97% reduction in PDO crashes
(SE = 3%)

4-Leg Intersections:

Angle Crashes

· 74% reduction in total crashes
(SE = 4%)

· 71% reduction in FI crashes
(SE = 5%)

· 88% reduction in PDO crashes
(SE = 3%)

Rear-End Crashes

· 89% reduction in total crashes
(SE = 3%)

· 86% reduction in FI crashes
(SE = 5%)

· 95% reduction in PDO crashes
(SE = 2%)

Safety Effectiveness

Reduction by severity level:

Markings on uncontrolled approaches

3-Leg and 4-Leg Intersections Combined

· 46% reduction in total crashes
(SE = 5%)

· 49% reduction in FI crashes
(SE = 7%)

· 50% reduction in PDO crashes
(SE = 6%)

 

Reduction by crash type:

Markings on uncontrolled approaches

3-Leg and 4-Leg Intersections Combined

Angle Crashes

· 38% reduction in total crashes
(SE = 7%)

· 42% reduction in FI crashes
(SE = 8%)

· 35% reduction in PDO crashes
(SE = 10%)

Rear-End Crashes

· 69% reduction in total crashes
(SE = 7%)

· 76% reduction in FI crashes
(SE = 9%)

· 75% reduction in PDO crashes
(SE = 8%)

Cost and Economic Benefits

To estimate benefit-cost ratios, assumed installation costs for painting supplementary pavement markings (i.e., STOP AHEAD) on a stop-controlled approach to an intersection ranged from $300-$750. The assumed cost for installing supplementary pavement markings on both uncontrolled approaches to an intersection with thermoplastic markings was $10,000 (cost per intersection treatment). Benefit-cost ratios calculated for installing supplementary pavement markings (i.e., STOP AHEAD) on stop-controlled approaches to intersections ranged from 1.8 to 528.9, and benefit-cost ratios for installing supplementary pavement markings on uncontrolled approaches to intersections ranged from 15.1 to 138.7.

Where to Implement

The treatment should be considered at intersections with a crash pattern related to a lack of driver awareness of the presence of the intersection.

Additional Factors for Consideration

Supplementary pavement markings may not be visible during winter conditions with snow and ice. Supplementary pavement markings may have a lower coefficient of friction compared to the rest of the intersection approach, especially during wet conditions (5).

Treatment Description

Supplementary pavement markings may be used as a safety treatment at unsignalized intersections with minor-road stop control on rural two-lane roads. This study considered two distinct types of supplementary pavement markings: markings on stop-controlled approaches and markings on uncontrolled approaches. These two types of markings are discussed in greater detail below.

Supplementary pavement markings installed on stop-controlled approaches alert drivers of the presence of the intersection and the need to stop before proceeding into the intersection. At all of the study locations included in the evaluation of supplementary pavement markings installed on stop-controlled approaches to an intersection, the supplementary pavement markings read "STOP AHEAD." Figure 15 shows two images of supplementary pavement markings on stop-controlled approaches.

This figure includes two photos of the supplemental pavement marking treatment used in advance of a stop sign.  Both photos show the marking “STOP AHEAD�. The photo on the left presents an aerial view of the marking while the photo on the right shows the marking from the view of a driver just upstream of the pavement marking.
Figure 15. Photo. Aerial view and street view of supplementary pavement markings on stop-controlled approach (Image credit: Google Earth™ Mapping Service).(21)

At uncontrolled approaches, the Pennsylvania Department of Transportation deployed a unique set of supplementary pavement markings on the major-road approaches (i.e., uncontrolled approaches) at intersections with minor-road stop control. These markings are intended to alert drivers of the presence of the intersection and of the potential for vehicles and other road users entering or exiting the roadway at that location. Figure 16 shows the pavement markings on the uncontrolled approach to an intersection that is located over the crest of the vertical curve. The treatment consists of two sets of pavement markings. In the direction of travel, the first set of markings warns drivers to slow to the speed limit of the roadway (e.g., "SLOW, XX MPH,"). The second set of markings illustrates the configuration of the upcoming intersection as either 4 legs or 3 legs with one of the following symbols: ┼, ├, or ┤. For 3-leg intersections, the symbol indicates the side of the roadway from which the minor road intersects the major road. In Figure 16, the supplementary pavement markings warn drivers to slow down to a speed of 30 mph as they are approaching a 4-leg intersection.

This photo shows a supplemental pavement marking treatment used on the uncontrolled approach of a two-way stop-controlled intersection.  The message shown is “SLOW, 30 MPH� followed by two “plus signs� that indicate the presence of an upcoming 4-leg intersection.
Figure 16. Photo. Supplementary pavement markings on uncontrolled approach.
(Image credit: Pennsylvania DOT)

Safety Evaluation

The safety effectiveness of supplementary pavement markings was evaluated separately for markings installed on stop-controlled approaches of intersections and those installed on uncontrolled approaches, because the two types of installations are very different in their mechanism for reducing crashes. In the first case, drivers are being reminded that they must take action at the intersection-the treatment is aimed at reducing crashes related to a failure to stop at the intersection. In the second case, drivers are being warned to use caution while approaching the intersection as other vehicles or road users may be entering the roadway, slowing to exit the roadway, or making a left-turn in front of the driver; however, no specific action is necessarily required. The EB before-after analysis approach was used in both cases, as discussed in Chapter 2 (Analysis Approach). The descriptive statistics, research methodology, and analysis results of the safety evaluations are presented below.

Descriptive Statistics

A total of 76 treatment and 140 nontreatment sites in four states-Arkansas, Minnesota, Nebraska, and Vermont-were available for analysis of the safety effectiveness of supplementary pavement markings on stop-controlled intersection approaches. Their breakdown by state, number of intersection legs, and number of treated approaches is shown in Table 20. The treatment was always installed on one minor-road approach at 3-leg intersections and either one or two minor-road approaches at 4-leg intersections. Traffic volumes at the treatment sites ranged from 90 to 5,700 veh/day (3-leg intersections) and from 105 to 4,800 veh/day (4-leg intersections) on the major-road approaches and from 40 to 3,000 veh/day (3-leg intersections) and from 25 to 1,770 veh/day (4-leg intersections) on the minor-road approaches.

In Pennsylvania, 11 treatment sites and 28 nontreatment sites were available for analysis of supplementary pavement markings on uncontrolled intersection approaches. Supplementary pavement markings were installed on both major-road approaches at all intersections included in the evaluation. Their breakdown by number of intersection legs and number of treated approaches is also shown in Table 20. Traffic volumes at the treatment sites ranged from 1,890 to 14,020 veh/day (3-leg intersections) and from 2,310 to 16,270 veh/day (4-leg intersections) on the major-road approaches and from 180 to 5,380 veh/day (3-leg intersections) and from 330 to 3,480 veh/day (4-leg intersections) on the minor-road approaches.

Table 20. Supplementary pavement markings: number of sites by State, number of intersection legs, and number of treated approaches.

State

Number of Intersection Legs

Number of Treated Approaches

Number of Sites

Treatment

Nontreatment

Treatment on Stop-Controlled Approach(es)

AR

3

1

1

11

4

1

1

19

2

1

MN

3

1

29

21

4

1

21

40

2

9

NE

3

1

3

13

4

1

4

17

2

2

VT

3

1

2

9

4

2

3

10

All sites

76

140

Treatment on Uncontrolled Approaches

PA

3

2

3

8

4

8

20

All sites

11

28

Crash and traffic volume data were obtained for varying periods before and after treatment installation, depending on the treatment installation date at the individual sites and crash data availability. Three crash types were considered in the analyses: all collision types combined, angle crashes, and rear-end crashes.

Table 21 (treatment intersections) and Table 22 (nontreatment intersections) summarize the crash data incorporated into the analysis of supplementary pavement markings installed on stop-controlled approaches to intersections on rural two-lane roads. They present total, angle, and rear-end crash data summed across all intersections of a given configuration (number of intersection legs) within each state.

Table 23 (treatment intersections) and Table 24 (nontreatment intersections) summarize the corresponding data incorporated into the analysis of supplementary pavement markings installed on uncontrolled approaches to intersections on rural two-lane roads.

Methodology

The safety effectiveness of installing supplementary pavement markings was evaluated using an EB method similar to that discussed in the EB evaluation of installing transverse rumble strips in Chapter 4. The safety effectiveness of supplementary pavement markings was evaluated separately for markings installed on stop-controlled approaches of intersections and those installed on uncontrolled approaches. Prior to implementing the EB method, the following points were addressed:

  1. Select appropriate SPF: The SPFs for intersections on rural two-lane roads from Chapter 10 of the HSM were selected. These are given for total crashes only. The coefficients of these SPFs vary by number of intersection legs. Use of the intersection SPFs from Chapter 10 of the HSM provide an estimate of the intersection-related predicted crash frequency for sites included in the analysis in the absence of the treatment.
  2. Obtain the proportion of target crashes for total, FS, FI, and PDO crashes (PR1): The proportions of target crashes (angle and rear end) to all crashes for each severity level (total, FS, FI, and PDO) were calculated using all crashes from nontreatment sites and from the before-period years of treatment sites. These proportions were calculated separately for 3- and 4-leg intersections. These proportions scale the total crash predictions (i.e., all crash types) to predictions for the target crashes (i.e., angle and rear-end crashes).
  3. Obtain the proportions of FS, FI, and PDO crashes (PR2): These proportions were calculated as the ratio of all FS, FI, or PDO crashes over total crashes using all crashes from the nontreatment sites and from the before-period years of treatment sites. These proportions were calculated separately for 3- and 4-leg intersections. These proportions scale the total crash predictions (i.e., all severity levels combined) to predicted crashes for specific severity level crashes (i.e., FS, FI, and PDO).
  4. Calibrate the SPFs to the local jurisdiction: Calibration was performed using all crashes (all collision types combined), separately for each intersection configuration within a given state, again using all nontreatment intersections and before treatment intersections combined. Total crash counts were used rather than target crashes due to the scarcity of target crashes, especially FS and FI angle and rear-end crashes. The calibration factor adjusts the HSM SPFs for varying conditions in the local jurisdiction such as crash reporting thresholds, environmental, etc.
Table 21. Supplementary pavement markings installed on stop-controlled approaches: summary crash statistics for the before and after treatment periods for treatment intersections by State and number of intersection legs.

Number of Intersection Legs

State

Before Period

After Period

Range of Years of Data in State

Number of Sites

Number of Site-Years

Crash Counts

Range of Years of Data in State

Number of Sites

Number of Site-Years

Crash Counts

Total

FS

FI

PDO

Total

FS

FI

PDO

All Crashes

3

AR

5

1

5

1

1

1

0

9

1

9

0

0

0

0

MN

5

29

145

14

0

7

6

9

29

261

17

0

10

7

NE

3

3

9

2

1

2

0

16

3

48

13

2

5

8

VT

5

2

10

8

1

1

7

1 to 2

2

3

2

0

0

2

4

AR

4 to 5

2

9

4

1

1

3

9 to 12

2

21

27

6

16

11

MN

5

30

150

24

6

17

4

9

30

270

26

1

14

12

NE

3

6

18

10

1

7

3

16

6

96

40

7

24

16

VT

5

3

15

11

0

3

8

2 to 7

3

11

7

0

3

4

Angle Crashes

3

AR

5

1

5

1

1

1

0

9

1

9

0

0

0

0

MN

5

29

145

1

0

0

1

9

29

261

1

0

1

0

NE

3

3

9

0

0

0

0

16

3

48

2

1

2

0

VT

5

2

10

1

1

1

0

1 to 2

2

3

0

0

0

0

4

AR

4 to 5

2

9

0

0

0

0

9 to 12

2

21

14

5

12

2

MN

5

30

150

11

0

8

2

9

30

270

7

0

5

2

NE

3

6

18

7

0

4

3

16

6

96

22

6

17

5

VT

5

3

15

5

0

1

4

2 to 7

3

11

3

0

0

3

Rear-End Crashes

3

AR

5

1

5

0

0

0

0

9

1

9

0

0

0

0

MN

5

29

145

1

0

0

1

9

29

261

0

0

0

0

NE

3

3

9

0

0

0

0

16

3

48

2

0

1

1

VT

5

2

10

0

0

0

0

1 to 2

2

3

0

0

0

0

4

AR

4 to 5

2

9

2

0

0

2

9 to 12

2

21

4

0

2

2

MN

5

30

150

2

2

2

0

9

30

270

4

0

3

1

NE

3

6

18

0

0

0

0

16

6

96

3

0

2

1

VT

5

3

15

3

0

0

3

2 to 7

3

11

1

0

1

0

 

Table 22. Supplementary pavement markings installed on stop-controlled approaches: summary crash statistics for the entire study period for nontreatment intersections by State and number of intersection legs.

Number of
Intersection
Legs

State

Entire Study Period

Range of Years of Data in State

Number of Sites

Number of Site-Years

Crash Counts

Total

FS

FI

PDO

All Crashes

3

AR

15

11

165

119

14

64

55

MN

15

21

315

56

2

24

30

NE

20

13

260

71

5

23

48

VT

8

9

72

22

0

6

16

4

AR

17

19

323

161

35

96

65

MN

15

40

600

106

4

47

55

NE

20

17

340

123

22

65

58

VT

13

10

130

58

1

27

31

Angle Crashes

3

AR

15

11

165

28

5

17

11

MN

15

21

315

4

0

3

1

NE

20

13

260

10

3

5

5

VT

8

9

72

2

0

0

2

4

AR

17

19

323

98

24

63

35

MN

15

40

600

31

2

17

11

NE

20

17

340

70

18

44

26

VT

13

10

130

35

0

18

17

Rear-End Crashes

3

AR

15

11

165

36

2

21

15

MN

15

21

315

4

0

2

2

NE

20

13

260

5

0

2

3

VT

8

9

72

4

0

1

3

4

AR

17

19

323

16

3

8

8

MN

15

40

600

21

1

8

13

NE

20

17

340

14

2

7

7

VT

13

10

130

5

0

0

5

 

Table 23. Supplementary pavement markings installed on uncontrolled approaches: summary crash statistics for the before and after treatment periods for treatment intersections by State and number of intersection legs.

Number of Intersection Legs

State

Before Period

After Period

Range of Years of Data in State

Number
of Sites

Number of Site-Years

Crash Counts

Range of Years of Data in State

Number
of Sites

Number of Site-Years

Crash Counts

Total

FS

FI

PDO

Total

FS

FI

PDO

All Crashes

3

PA

5

3

15

14

3

11

3

2

3

6

6

0

5

1

4

PA

4 to 5

8

37

94

8

50

44

2 to 12

8

52

156

5

74

82

Angle Crashes

3

PA

5

3

15

7

1

5

2

2

3

6

3

0

2

1

4

PA

4 to 5

8

37

62

7

38

24

2 to 12

8

52

122

5

62

60

Rear-End Crashes

3

PA

5

3

15

1

0

1

0

2

3

6

1

0

1

0

4

PA

4 to 5

8

37

13

0

5

8

2 to 12

8

52

19

0

7

12

 

Table 24. Supplementary pavement markings installed on uncontrolled approaches: summary crash statistics for the entire study period for nontreatment intersections by State and number of intersection legs.

Number of
Intersection
Legs

State

Entire Study Period

Range of
Years of
Data in
State

Number
of Sites

Number of Site-Years

Crash Counts

Total

FS

FI

PDO

All Crashes

3

PA

8

8

64

28

4

20

8

4

PA

17

20

340

324

24

188

136

Angle Crashes

3

PA

8

8

64

5

1

3

2

4

PA

17

20

340

201

20

121

80

Rear-End Crashes

3

PA

8

8

64

5

0

4

1

4

PA

17

20

340

39

0

22

17

The SPFs presented in the HSM for intersections on rural two-lane roads for total severity level (i.e., all severity levels combined) have the general form:

Predicted crashes/yr = exp[a + b(lnAADTmaj) + c(lnAADTmin)]

Figure 17. Equation. Base model for predicted crashes per year.

where a, b, and c are the regression coefficients shown in Table 25 for the analysis associated with markings installed on stop-controlled approaches of intersections and in Table 27 for the analysis associated with markings installed on uncontrolled approaches. These coefficients apply to base conditions and vary by number of intersection legs. For the intersection SPFs in Chapter 10 of the HSM, the base conditions are:

Intersection skew angle: 0 degrees
Number of intersection left-turn lanes: None on approaches without stop control
Number of intersection right-turn lanes: None on approaches without stop control
Presence of lighting: None

Crash modification factors (CMFs), calibration factors (Cr), proportions of angle and rear-end crashes (PR1), and proportions of FS, FI, and PDO crashes (PR2) were then used to adjust for local conditions as follows:

Predicted crashes/yr = {exp[a + b(lnAADTmaj) + c(lnAADTmin)]} × PR1 × PR2 × CMFCombined × Cr

Figure 18. Equation. Model for predicted crashes per year for specific crash types and severity levels, and accounting for local conditions.

The CMFCombined is the product of the CMFs from Chapter 10 of the HSM for skew angle (CMF1i), number of major-road left-turn lanes (CMF2i), and number of major-road right-turn lanes (CMF3i), for a particular intersection configuration.

SPF coefficients (a, b, and c and overdispersion parameter), target crash proportions (PR1), proportions of FS, FI, and PDO out of total crashes (PR2), and calibration factors (Cr) are shown for each intersection configuration in Table 25 and Table 26. Number of site-years, total crash counts (all severity levels), and target crash counts are also displayed. The tables also show the default proportions of PR1 and PR2 presented in Chapter 10 of the HSM (see Tables 10-6 and 10-5, respectively). Note that PR2 is always equal to 1 for total crashes. The decision of which proportions to use-those calculated from the data or those provided by the HSM-was based on whether calculated proportions of target crashes (PR1) were nonzero for all severity levels. If they were not, then the HSM proportions (both PR1 and PR2) were used in the EB before-after analysis. The selections of which proportions were used are indicated in Table 25 and Table 26 by an askterisk.

Safety Effectiveness

The EB before-after method was applied to estimate the safety effectiveness of installing supplementary pavement markings at rural stop-controlled intersections. The analysis of the safety effectiveness of supplementary pavement markings installed on stop-controlled approaches to intersections included treatment and nontreatment sites in Arkansas, Minnesota, Nebraska, and Vermont, and used HSM SPFs for 3- and 4-leg intersections on rural two-lane roads. The analysis of the safety effectiveness of supplementary pavement markings installed on uncontrolled approaches to intersections was based on treatment and nontreatment intersections in Pennsylvania.

Table 25. Supplementary pavement markings installed on stop-controlled approaches: SPF coefficients, target crash proportions, and calibration factors by State and number of intersection legs.

State

Number of Legs

Number of Site-Years

Severity Level

Number of Crashes, All

Target Crash Type

Number of Target Crashes

Intercept (a)

lnAADTmaj Coefficient (b)

lnAADTmin Coefficient (c)

Overdispersion Parameter

Proportion of Target Crashes (PR1)

PR1 HSM

Proportion of FS, FI, or PDO of Total Crashes (PR2)

PR2 HSM

Calibration Factor (Cr)

AR

3

170

Total

120

Angle

29

N/A

N/A

N/A

N/A

0.24

0.24*

1.00*

1.00

N/A

Rear End

36

N/A

N/A

N/A

N/A

0.30*

0.28

1.00*

1.00

N/A

FS

15

Angle

6

N/A

N/A

N/A

N/A

0.40*

0.28

0.13*

0.06

N/A

Rear End

2

-9.86

0.79

0.49

0.54

0.13*

0.26

0.13*

0.06

13.68

FI

65

Angle

18

N/A

N/A

N/A

N/A

0.28*

0.28

0.54*

0.42

N/A

Rear End

21

N/A

N/A

N/A

N/A

0.32*

0.26

0.54*

0.42

N/A

PDO

55

Angle

11

N/A

N/A

N/A

N/A

0.20*

0.21

0.46*

0.59

N/A

Rear End

15

N/A

N/A

N/A

N/A

0.27*

0.29

0.46*

0.59

N/A

4

332

Total

165

Angle

98

N/A

N/A

N/A

N/A

0.59*

0.43

1.00*

1.00

N/A

Rear End

18

N/A

N/A

N/A

N/A

0.11*

0.24

1.00*

1.00

N/A

FS

36

Angle

24

N/A

N/A

N/A

N/A

0.67*

0.53

0.22*

0.06

N/A

Rear End

3

-8.56

0.60

0.61

0.24

0.08*

0.21

0.22*

0.06

9.18

FI

97

Angle

63

N/A

N/A

N/A

N/A

0.65*

0.53

0.59*

0.43

N/A

Rear End

8

N/A

N/A

N/A

N/A

0.08*

0.21

0.59*

0.43

N/A

PDO

68

Angle

35

N/A

N/A

N/A

N/A

0.52*

0.35

0.41*

0.57

N/A

Rear End

10

N/A

N/A

N/A

N/A

0.15*

0.27

0.41*

0.57

N/A

MN

3

460

Total

70

Angle

5

N/A

N/A

N/A

N/A

0.07

0.24*

1.00

1.00*

N/A

Rear End

5

N/A

N/A

N/A

N/A

0.07

0.28*

1.00

1.00*

N/A

FS

2

Angle

0

N/A

N/A

N/A

N/A

0.00

0.28*

0.03

0.06*

N/A

Rear End

0

-9.86

0.79

0.49

0.54

0.00

0.26*

0.03

0.06*

1.35

FI

31

Angle

3

N/A

N/A

N/A

N/A

0.10

0.28*

0.44

0.42*

N/A

Rear End

2

N/A

N/A

N/A

N/A

0.07

0.26*

0.44

0.42*

N/A

PDO

36

Angle

2

N/A

N/A

N/A

N/A

0.06

0.21*

0.51

0.59*

N/A

Rear End

3

N/A

N/A

N/A

N/A

0.08

0.29*

0.51

0.59*

N/A

4

750

Total

130

Angle

42

N/A

N/A

N/A

N/A

0.32*

0.43

1.00*

1.00

N/A

Rear End

23

N/A

N/A

N/A

N/A

0.18*

0.24

1.00*

1.00

N/A

FS

10

Angle

2

N/A

N/A

N/A

N/A

0.20*

0.53

0.08*

0.06

N/A

Rear End

3

-8.56

0.60

0.61

0.24

0.30*

0.21

0.08*

0.06

1.24

FI

64

Angle

25

N/A

N/A

N/A

N/A

0.39*

0.53

0.49*

0.43

N/A

Rear End

10

N/A

N/A

N/A

N/A

0.16*

0.21

0.49*

0.43

N/A

PDO

59

Angle

13

N/A

N/A

N/A

N/A

0.22*

0.35

0.45*

0.57

N/A

Rear End

13

N/A

N/A

N/A

N/A

0.22*

0.27

0.45*

0.57

N/A

NE

3

269

Total

73

Angle

10

N/A

N/A

N/A

N/A

0.14

0.24*

1.00

1.00*

N/A

Rear End

5

N/A

N/A

N/A

N/A

0.07

0.28*

1.00

1.00*

N/A

FS

6

Angle

3

N/A

N/A

N/A

N/A

0.50

0.28*

0.08

0.06*

N/A

Rear End

0

-9.86

0.79

0.49

0.54

0.00

0.26*

0.08

0.06*

8.80

FI

25

Angle

5

N/A

N/A

N/A

N/A

0.20

0.28*

0.34

0.42*

N/A

Rear End

2

N/A

N/A

N/A

N/A

0.08

0.26*

0.34

0.42*

N/A

PDO

48

Angle

5

N/A

N/A

N/A

N/A

0.10

0.21*

0.66

0.59*

N/A

Rear End

3

N/A

N/A

N/A

N/A

0.06

0.29*

0.66

0.59*

N/A

4

358

Total

133

Angle

77

N/A

N/A

N/A

N/A

0.58*

0.43

1.00*

1.00

N/A

Rear End

14

N/A

N/A

N/A

N/A

0.11*

0.24

1.00*

1.00

N/A

FS

23

Angle

18

N/A

N/A

N/A

N/A

0.78*

0.53

0.17*

0.06

N/A

Rear End

2

-8.56

0.60

0.61

0.24

0.09*

0.21

0.17*

0.06

7.57

FI

72

Angle

48

N/A

N/A

N/A

N/A

0.67*

0.53

0.54*

0.43

N/A

Rear End

7

N/A

N/A

N/A

N/A

0.10*

0.21

0.54*

0.43

N/A

PDO

61

Angle

29

N/A

N/A

N/A

N/A

0.48*

0.35

0.46*

0.57

N/A

Rear End

7

N/A

N/A

N/A

N/A

0.12*

0.27

0.46*

0.57

N/A

VT

3

82

Total

30

Angle

3

N/A

N/A

N/A

N/A

0.10

0.24*

1.00

1.00*

N/A

Rear End

4

N/A

N/A

N/A

N/A

0.13

0.28*

1.00

1.00*

N/A

FS

1

Angle

1

N/A

N/A

N/A

N/A

1.00

0.28*

0.03

0.06*

N/A

Rear End

0

-9.86

0.79

0.49

0.54

0.00

0.26*

0.03

0.06*

1.82

FI

7

Angle

1

N/A

N/A

N/A

N/A

0.14

0.28*

0.23

0.42*

N/A

Rear End

1

N/A

N/A

N/A

N/A

0.14

0.26*

0.23

0.42*

N/A

PDO

23

Angle

2

N/A

N/A

N/A

N/A

0.09

0.21*

0.77

0.59*

N/A

Rear End

3

N/A

N/A

N/A

N/A

0.13

0.29*

0.77

0.59*

N/A

4

145

Total

69

Angle

40

N/A

N/A

N/A

N/A

0.58

0.43*

1.00

1.00*

N/A

Rear End

8

N/A

N/A

N/A

N/A

0.12

0.24*

1.00

1.00*

N/A

FS

1

Angle

0

N/A

N/A

N/A

N/A

0.00

0.53*

0.01

0.06*

N/A

Rear End

0

-8.56

0.60

0.61

0.24

0.00

0.21*

0.01

0.06*

2.44

FI

30

Angle

19

N/A

N/A

N/A

N/A

0.63

0.53*

0.44

0.43*

N/A

Rear End

0

N/A

N/A

N/A

N/A

0.00

0.21*

0.44

0.43*

N/A

PDO

39

Angle

21

N/A

N/A

N/A

N/A

0.54

0.35*

0.57

0.57*

N/A

Rear End

8

N/A

N/A

N/A

N/A

0.21

0.27*

0.57

0.57*

N/A

NOTE: Asterisked proportions were those used in the analyses.

 

Table 26. Supplementary pavement markings installed on uncontrolled approaches: SPF coefficients, target crash proportions, and calibration factors by State and number of intersection legs.

State

Number of Legs

Number of Site-Years

Severity Level

Number of Crashes, All

Target Crash Type

Number of Target Crashes

Intercept (a)

lnAADTmaj Coefficient (b)

lnAADTmin Coefficient (c)

Overdispersion Parameter

Proportion of Target Crashes (PR1)

PR1 HSM

Proportion of FS, FI, or PDO of Total Crashes (PR2)

PR2 HSM

Calibration Factor (Cr)

PA

3

79

Total

42

Angle

12

N/A

N/A

N/A

N/A

0.29

0.24*

1.00

1.00*

N/A

Rear End

6

N/A

N/A

N/A

N/A

0.14

0.28*

1.00

1.00*

N/A

FS

7

Angle

2

N/A

N/A

N/A

N/A

0.29

0.28*

0.17

0.06*

N/A

Rear End

0

-9.86

0.79

0.49

0.54

0.00

0.26*

0.17

0.06*

1.05

FI

31

Angle

8

N/A

N/A

N/A

N/A

0.26

0.28*

0.74

0.42*

N/A

Rear End

5

N/A

N/A

N/A

N/A

0.16

0.26*

0.74

0.42*

N/A

PDO

11

Angle

4

N/A

N/A

N/A

N/A

0.36

0.21*

0.26

0.59*

N/A

Rear End

1

N/A

N/A

N/A

N/A

0.09

0.29*

0.26

0.59*

N/A

4

377

Total

418

Angle

263

N/A

N/A

N/A

N/A

0.63

0.43*

1.00

1.00*

N/A

Rear End

52

N/A

N/A

N/A

N/A

0.12

0.24*

1.00

1.00*

N/A

FS

32

Angle

27

N/A

N/A

N/A

N/A

0.84

0.53*

0.08

0.06*

N/A

Rear End

0

-8.56

0.60

0.61

0.24

0.00

0.21*

0.08

0.06*

2.82

FI

238

Angle

159

N/A

N/A

N/A

N/A

0.67

0.53*

0.57

0.43*

N/A

Rear End

27

N/A

N/A

N/A

N/A

0.11

0.21*

0.57

0.43*

N/A

PDO

180

Angle

104

N/A

N/A

N/A

N/A

0.58

0.35*

0.43

0.57*

N/A

Rear End

25

N/A

N/A

N/A

N/A

0.14

0.27*

0.43

0.57*

N/A

NOTE: Asterisked proportions were those used in the analyses.

The EB before-after analysis results are shown for the following crash types and severity levels:

Although the analyses for FS injury crashes were performed, the analysis results were not considered reliable and are therefore not shown. The occurrence of FS crashes was too rare across all intersections in the study (both treatment and nontreatment sites).

For the analysis of the safety effectiveness of supplementary pavement markings installed on stop-controlled approaches to intersections, the EB method was applied to all states combined based on the following reasoning: (1) a small number of treatment sites were available in Arkansas and Vermont-individually, they could not have been used for evaluation; and (2) all EB intermediate calculations (up to the final effectiveness calculations) are performed on a state/site basis, thus using SPFs, proportions (PR1 and PR2), combined CMFs, and calibration factors (Cr) specific to that state/site. This approach, to some extent, took state-to-state variability into account while increasing site and crash sample sizes. Additionally, the EB method was initially applied separately to 3-leg and 4-leg intersections and whether the treatment was installed on one or two stop-controlled approaches to the intersection. However, due to sample size issues, it was decided to pool across the number of treated approaches and perform the EB analysis separately for 3- and 4-leg intersections for the pooled data from the four states.

For the analysis of the safety effectiveness of supplementary pavement markings installed on uncontrolled approaches to intersections, the data for 3- and 4-leg intersections were combined and analyzed together to have a sufficient sample size for analysis.

The analysis results of the safety effectiveness estimates for supplementary pavement markings installed on stop-controlled approaches to intersections are shown in Table 27. The statistics shown for each crash severity are:

A negative percent safety effectiveness indicates that crash frequencies decreased due to the treatment.

The analysis results of the safety effectiveness estimates for supplementary pavement markings installed on uncontrolled approaches to intersections are shown in Table 28.

Table 27. Supplementary pavement markings installed on stop-controlled approaches: safety effectiveness on target crashes by number of intersection legs.

Crash Severity

Number of Treatment Sites

Safety Effectiveness (%)

SE of
Treatment
Effect (%)

Significance

All States Combined; 3-Leg Intersections

All Crashes

Total

35

-67

7

Significant at 95% CL

FI

-76

7

Significant at 95% CL

PDO

-72

7

Significant at 95% CL

Angle Crashes

Total

35

-92

5

Significant at 95% CL

FI

-88

7

Significant at 95% CL

PDO

-100a

NC

NC

Rear-End Crashes

Total

35

-95

4

Significant at 95% CL

FI

-96

5

Significant at 95% CL

PDO

-97

3

Significant at 95% CL

All States Combined;  4-Leg Intersections

All Crashes

Total

41

-66

4

Significant at 95% CL

FI

-69

5

Significant at 95% CL

PDO

-77

4

Significant at 95% CL

Angle Crashes

Total

41

-74

4

Significant at 95% CL

FI

-71

5

Significant at 95% CL

PDO

-88

3

Significant at 95% CL

Rear-End Crashes

Total

41

-89

3

Significant at 95% CL

FI

-86

5

Significant at 95% CL

PDO

-95

2

Significant at 95% CL

a Crashes recorded in before period; none in after period.

NC=Not Calculated; standard error and significance could not be estimated.

 

Table 28. Supplementary pavement markings installed on uncontrolled approaches: safety effectiveness on target crashes.

Crash Severity

Number of Treatment Sites

Safety Effectiveness
(%)

SE of Treatment Effect (%)

Significance

3- and 4-Leg Intersections Combined

All Crashes

Total

 

-46

5

Significant at 95% CL

FI

11

-49

7

Significant at 95% CL

PDO

 

-50

6

Significant at 95% CL

Angle Crashes

Total

 

-38

7

Significant at 95% CL

FI

11

-42

8

Significant at 95% CL

PDO

 

-35

10

Significant at 95% CL

Rear-End Crashes

Total

 

-69

7

Significant at 95% CL

FI

11

-76

9

Significant at 95% CL

PDO

 

-75

8

Significant at 95% CL

In general, the EB analysis results for supplementary pavement markings installed on stop-controlled approaches to intersections yielded statistically significant results for most of the crash types and severity levels analyzed for both 3- and 4-leg intersections. The results show that the overall effectiveness of this treatment (for all crash types and severities combined) is nearly identical for 3-leg and 4-leg intersections. Also, greater reductions were estimated for target crashes (i.e., angle and rear-end crashes) than for total crashes (i.e., all crash types combined).

The EB analysis results for supplementary pavement markings installed on uncontrolled approaches to intersections also show statistically significant reductions in crashes for all of the crash types and severity levels analyzed. The analysis results also show greater reductions in rear-end crashes compared to total (i.e., all crash types combined) and angle crashes.

Economic Analysis

An economic analysis was conducted to estimate the economic benefits of installing supplementary pavement markings at unsignalized intersections with minor-road stop control on rural two-lane roads. Separate economic analyses were conducted for supplementary pavement markings installed on stop-controlled approaches and uncontrolled approaches. The Economic Analysis discussion in Chapter 2 describes the procedure for estimating the benefit-cost ratios of the treatments. The economic benefits of these treatments were estimated using the safety effectiveness estimates developed in this research.

The economic analysis produced two separate benefit-cost ratio tables for installing supplementary pavement markings on stop-controlled approaches: one for 3-leg intersections and the second for 4-leg intersections on rural two-lane roads. The safety evaluation produced CMFs for total crashes on both 3-leg and 4-leg intersections, and these CMFs were used to determine the expected annual benefit. This particular analysis assumes that one STOP AHEAD supplementary pavement marking is installed on each stop-controlled approach.

Kansas Department of Transportation reports the average cost of a STOP AHEAD pavement marking to be $750 per approach. For the sake of the analysis, a 1-yr service life was assumed. Longer service lives will only increase the benefit of a treatment, so assuming a shorter service life is conservative for a benefit-cost analysis. Table 29 and Table 30 present the benefit-cost ratios for installing a STOP AHEAD pavement marking on the stop-controlled approaches of a 3-leg and 4-leg rural intersection, respectively. The AADTs in the table cover the range of AADTs of the study sites used for the estimation of the CMF. For both intersection types, the benefit-cost ratio exceeds 1.0 for all AADT combinations, indicating that installation of a STOP AHEAD pavement marking on stop-controlled approaches is economically justified at all AADT levels. At higher-volume intersections, the benefit can be 100 or more times the cost.

Table 29. Benefit-cost ratios for installing STOP AHEAD pavement marking on the stop-controlled approach of a 3-leg intersection for $750 installation cost and 1-yr service life.

Major- Road AADT

% of Major-Road AADT on Minor Road

5%

10%

15%

20%

25%

30%

40%

50%

60%

70%

80%

90%

200

N/A

N/A

N/A

1.8

2.0

2.2

2.5

2.8

3.1

3.3

3.5

3.7

1,000

7.1

10.0

12.2

14.0

15.6

17.1

19.7

22.0

24.0

25.9

27.6

29.3

2,000

17.3

24.2

29.6

34.0

38.0

41.5

47.8

53.3

58.3

62.9

67.1

71.1

3,000

29.0

40.7

49.7

57.2

63.8

69.8

80.3

89.6

98.0

105.7

112.8

119.5

4,000

41.9

58.9

71.8

82.7

92.2

100.8

116.1

129.5

141.6

152.7

N/A

N/A

5,000

55.8

78.3

95.5

110.0

122.7

134.1

154.5

172.3

N/A

N/A

N/A

N/A

5,200

58.6

82.3

100.4

115.6

129.0

141.1

162.4

181.2

N/A

N/A

N/A

N/A

CMF = 0.33 (total crashes)

Installation cost = $750 (i.e., $750 per approach)

Service life = 1 yr

N/A indicates the combination of major- and minor-road AADTs was not represented in the study

 

Table 30. Benefit-cost ratios for installing STOP AHEAD pavement markings on the stop-controlled approaches of a 4-leg intersection for $1,500 installation cost and 1-yr service life.

Major- Road AADT

% of Major-Road AADT on Minor Road

5%

10%

15%

20%

25%

30%

40%

50%

60%

70%

80%

90%

200

N/A

N/A

N/A

1.9

2.2

2.5

3.0

3.4

3.8

4.2

4.5

4.8

1,000

5.8

8.9

11.4

13.6

15.6

17.4

20.7

23.7

26.5

29.1

31.6

34.0

2,000

13.5

20.6

26.3

31.4

36.0

40.2

47.9

54.9

61.4

67.4

N/A

N/A

3,000

22.0

33.6

43.0

51.3

58.8

65.7

78.3

89.7

N/A

N/A

N/A

N/A

3,600

27.5

41.9

53.7

64.0

73.3

81.9

97.6

N/A

N/A

N/A

N/A

N/A

CMF = 0.34 (total crashes)

Installation cost = $1,500 (i.e., $750 per approach)

Service life = 1 yr

N/A indicates the combination of major- and minor-road AADTs was not represented in the study

Vermont Department of Transportation reports the average cost of a STOP AHEAD pavement marking to be between $300 and $500 per approach with a service life of 2 years. Table 31 and Table 32 present the benefit-cost ratios for installing a STOP AHEAD pavement marking on the stop-controlled approaches of a 3-leg and 4-leg intersection, respectively. The AADTs in the table cover the range of AADTs of the study sites used for the estimation of the CMF. For both intersection types, the benefit-cost ratio exceeds 5.0 for all AADT combinations.

Table 31. Benefit-cost ratios for installing STOP AHEAD pavement marking on the stop-controlled approach of a 3-leg intersection for $500 installation cost and 2-yr service life.

Major- Road AADT

% of Major-Road AADT on Minor Road

5%

10%

15%

20%

25%

30%

40%

50%

60%

70%

80%

90%

200

N/A

N/A

N/A

5.2

5.8

6.4

7.3

8.2

8.9

9.6

10.3

10.9

1,000

20.7

29.1

35.5

40.9

45.6

49.9

57.5

64.1

70.1

75.6

80.7

85.5

2,000

50.4

70.7

86.3

99.4

110.8

121.2

139.5

155.7

170.2

183.6

196.0

207.6

3,000

84.6

118.9

145.0

167.0

186.2

203.7

234.5

261.6

286.0

308.5

329.3

348.9

4,000

122.3

171.8

209.6

241.3

269.2

294.3

338.9

378.0

413.4

445.8

N/A

N/A

5,000

162.8

228.6

278.8

321.1

358.1

391.6

450.9

503.0

N/A

N/A

N/A

N/A

5,200

171.1

240.4

293.2

337.6

376.6

411.8

474.1

528.9

N/A

N/A

N/A

N/A

CMF = 0.33 (total crashes)

Installation cost = $500 (i.e., $500 per approach)

Service life = 2 yrs

N/A indicates the combination of major- and minor-road AADTs was not represented in the study

 

Table 32. Benefit-cost ratios for installing STOP AHEAD pavement markings on the stop-controlled approaches of a 4-leg intersection for $1,000 installation cost and 2-yr service Life.

Major- Road AADT

% of Major-Road AADT on Minor Road

5%

10%

15%

20%

25%

30%

40%

50%

60%

70%

80%

90%

200

N/A

N/A

N/A

5.6

6.4

7.2

8.6

9.8

11.0

12.1

13.1

14.1

1,000

16.9

25.8

33.1

39.4

45.1

50.4

60.1

68.9

77.0

84.6

91.8

98.6

2,000

39.1

59.7

76.5

91.1

104.4

116.7

139.1

159.4

178.1

195.7

N/A

N/A

3,000

63.9

97.5

124.9

148.8

170.5

190.6

227.2

260.3

N/A

N/A

N/A

N/A

3,600

79.7

121.6

155.7

185.6

212.6

237.7

283.3

N/A

N/A

N/A

N/A

N/A

CMF = 0.34 (total crashes)

Installation cost = $1,000 (i.e., $500 per approach)

Service life = 2 yrs

N/A indicates the combination of major- and minor-road AADTs was not represented in the study

The Pennsylvania Department of Transportation was the only state in the study that installed supplementary pavement markings on uncontrolled approaches to intersections. The Pennsylvania Department of Transportation reports the average cost of installation to be $10,000 with a service life of 5 years. This cost is based on the use of thermoplastic pavement marking material and cost of the treatment for the entire intersection. Table 33 shows the benefit-cost ratios for the supplementary pavement markings on uncontrolled approaches at a 4-leg stop-controlled intersection on rural two-lane roads. The ranges of the AADTs in the table are within the minimum and maximum AADTs of the study sites used in the safety evaluation. All the benefit-cost ratios exceed 15, meaning the installation of supplementary pavement markings on uncontrolled approaches is economically justified across the entire AADT range evaluated in this study.

Table 33. Benefit-cost ratios for installing supplementary pavement markings on uncontrolled approaches at 4-leg stop-controlled intersections.

Major- Road AADT

% of Major-Road AADT on Minor Road

5%

10%

15%

20%

30%

40%

50%

60%

70%

80%

90%

2,400

N/A

N/A

15.1

18.0

23.0

27.4

31.4

35.1

38.6

41.8

45.0

3,000

N/A

N/A

19.7

23.5

30.1

35.9

41.1

46.0

50.5

54.8

58.9

4,000

N/A

21.8

28.0

33.3

42.7

50.9

58.3

65.1

N/A

N/A

N/A

5,000

N/A

28.6

36.6

43.7

55.9

66.6

76.3

N/A

N/A

N/A

N/A

6,000

N/A

35.7

45.7

54.4

69.7

83.1

N/A

N/A

N/A

N/A

N/A

7,000

28.2

43.0

55.0

65.6

84.0

N/A

N/A

N/A

N/A

N/A

N/A

8,000

33.1

50.5

64.7

77.1

98.7

N/A

N/A

N/A

N/A

N/A

N/A

9,000

38.2

58.2

74.6

88.9

113.8

N/A

N/A

N/A

N/A

N/A

N/A

10,000

43.3

66.2

84.7

101.0

N/A

N/A

N/A

N/A

N/A

N/A

N/A

11,000

48.6

74.2

95.1

113.3

N/A

N/A

N/A

N/A

N/A

N/A

N/A

12,000

54.0

82.5

105.6

125.9

N/A

N/A

N/A

N/A

N/A

N/A

N/A

13,000

59.5

90.9

116.4

138.7

N/A

N/A

N/A

N/A

N/A

N/A

N/A

14,000

65.1

99.4

127.3

N/A

N/A

N/A

N/A

N/A

N/A

N/A

N/A

CMF = 0.54 (total crashes)

Installation cost = $10,000

Service life = 5 yrs

N/A indicates the combination of major- and minor-road AADTs was not represented in the study

Implementation

The primary purpose of installing supplementary pavement markings on stop-controlled approaches to intersections on rural two-lane roads is to alert drivers of a stop ahead. The treatment is intended to increase drivers' awareness of the intersection and to the traffic control. Supplementary pavement markings should be considered for installation on stop-controlled approaches to intersections with a pattern of crashes related to a lack of driver recognition of the presence of the intersection (e.g., angle crashes related to stop sign violations).(5) In particular, supplementary pavement markings should be considered for installation along the stop-controlled approach to an intersection that is hidden from view due to horizontal or vertical curvature or where the traffic control is hidden from view as the driver approaches the intersection.

The primary purpose of supplementary pavement markings on uncontrolled approaches to intersections on rural two-lane roads is to increase driver awareness of the intersection and reduce speeds of vehicles on the major road near the intersection. Supplementary pavement markings on uncontrolled approaches to intersections should be considered on approaches to intersections where it is difficult to recognize the presence of the crossroad(s) and/or at intersections where speeds on the major-road approaches are higher than desired for the conditions.

The economic analyses provided above show that both types of supplementary pavement markings are economically justifiable, even along roads with low traffic volumes.

Both types of supplementary pavement markings should be implemented in accordance with the Manual on Uniform Traffic Control Devices (MUTCD). The MUTCD gives little guidance about the design and placement of supplementary pavement markings. Section 3B.20 of the MUTCD gives support for the use of markings, stating, "These pavement markings can be helpful to road users in some locations by supplementing signs and providing additional emphasis for important regulatory, warning, or guidance messages, because the markings do not require diversion of the road user's attention from the roadway surface.(27)" Section 3B.20 also includes recommended sizes for words on the pavement and states that "STOP AHEAD" may be used to supplement signs.

The following paragraphs present several example design details on the use of supplementary pavement markings. The information was provided by several state agencies.

As indicated, the MUTCD permits the use of the phrase "STOP AHEAD" on pavement to warn drivers of an upcoming stop. Figure 19 and Figure 20 are from the Minnesota Traffic Engineering Manual and provide guidance on the placement of the supplementary pavement marking. Figure 19 shows that the placement of the pavement marking is dependent on the placement of the W3-1a advance warning sign. Figure 20 indicates the placement of the W3-1a sign relative to the location of the stop bar.

Vermont Department of Transportation recommends installing the STOP AHEAD pavement marking at the location of the W3-1a sign (stop ahead sign). The distance to the stop bar is usually 200-250 ft; however, engineering judgment in the field is used to make sure sight obstructions such as crest vertical curves do not warrant longer distances.

Figure 21 shows design details for the placement and size of the supplementary pavement markings on the uncontrolled approach. It is recommended that this treatment not be used in areas with significant grade differences. The placement of the markings relative to the intersection is dependent on the posted speed of the roadway (see Table 34 for distances). These markings are part of a comprehensive intersection treatment that includes unique signage. See Appendix A for additional guidance from the Pennsylvania Department of Transportation regarding this treatment.

This figure shows Minnesota’s guidance for the placement of STOP AHEAD pavement markings.  The diagram shows that the message letters are 8 ft tall and that there is 40 ft of space between the words “stop� and “ahead�.  The upstream side of the word “stop� is located 100 ft downstream of the STOP AHEAD sign.  The notes with the diagram state 1) Do not install message if intersection has adequate lighting, and 2) Install only one STOP AHEAD pavement message. If a stop ahead sign needs to be installed more than 1,000 ft from the stop sign, contact the district traffic engineer to determine if and where a second set of STOP AHEAD pavement messages should be installed.
Figure 19. Diagram. Minnesota guidance for placement of STOP AHEAD pavement marking relative to W3-1a sign.(28)

This figure provides Minnesota’s guidance for placement of a W3-1a sign (STOP AHEAD symbol sign) relative to the STOP sign.  The route marker is placed 300 ft upstream of the STOP sign, and the STOP AHEAD sign is placed 300 ft upstream of the route marker signs
Figure 20. Diagram. Minnesota guidance for placement of W3-1a sign relative to stop bar.(28)

This figure shows guidance for the placement of supplemental pavement markings on uncontrolled intersection approaches.  The distance between the side street and the most upstream element of the marking (the bottom of the word “SLOW�) is labeled as lower-case d.  The distance between the side street and the location of the W2-1 sign (intersection warning sign) is labeled as capital Y.  The values for d (pattern length) and Y (distance to sign) are provided in Table 34 by speed limit. The figure shows that the plus signs are each 12 ft tall with 50 ft of space between them.  The total height of the two lines of text “XX MPH� and “SLOW� is 96 ft. There is 50 ft of space between the text and the first plus sign marking. Figure 21. Diagram. Placement guidance for supplementary pavement markings on uncontrolled approach.(29)

Table 34. Placement of supplementary pavement markings relative to intersection by posted speed.(29)

Posted Speed
(mph)

d, Pattern Length
(ft)

Y, Distance to Sign
(ft)

25

265

340

30

300

380

35

340

450

40

375

500

45

410

550

50

450

600

55

485

650

Chapter 6. Conclusions

The objective of this research is to advance efforts to improve safety at intersections with minor-road stop control along rural two-lane roads. The safety effectiveness of three low-cost safety treatments was evaluated to estimate their expected effectiveness in reducing crashes. The low-cost safety treatments included:

The effectiveness of each treatment in reducing crashes was estimated using the Empirical Bayes (EB) observational before-after safety evaluation analysis approach. In addition, economic analyses were performed to estimate the benefit-cost ratio of each treatment, incorporating safety effectiveness results from the EB analyses.

The safety and economic analyses show that the treatments are effective in reducing crashes of different types and severity levels and are economically justifiable for installation at intersections with patterns of crashes that suggest drivers are unaware of the presence of the intersection. The expected safety effectiveness (and standard error) of each treatment in reducing crashes of different types and severities are shown in Table 35.

Table 35. Summary of treatment effectiveness by treatment and intersection type and crash type and crash severity level.

Intersection
Type

Crash Type

Severity
Level

Crash Reduction (%)
(SE (%))

Single Luminaire

3 or 4 Legs

Total nighttime

Total

71 (29)

Transverse Rumble Strips

3 Legs

All

FI

37 (20)

Angle

PDO

61 (28)

Rear End

FI

60 (29)

4 Legs

All

Total

13 (7)

FI

29 (8)

Angle

FI

25 (10)

Rear End

Total

56 (8)

FI

78 (8)

PDO

54 (10)

Supplementary Pavement Markings on Stop-Controlled Approaches (i.e., STOP AHEAD)

3 Legs

All

Total

67 (7)

FI

76 (7)

PDO

72 (7)

Angle

Total

92 (5)

FI

88 (7)

Rear End

Total

95 (4)

FI

96 (5)

PDO

97 (3)

4 Legs

All

Total

66 (4)

FI

69 (5)

PDO

77 (4)

Angle

Total

74 (4)

FI

71 (5)

PDO

88 (3)

Rear End

Total

89 (3)

FI

86 (5)

PDO

95 (2)

Supplementary Pavement Markings on Uncontrolled Approaches

3 or 4 Legs

All

Total

46 (5)

FI

49 (7)

PDO

50 (6)

Angle

Total

38 (7)

FI

42 (8)

PDO

35 (10)

Rear End

Total

69 (7)

FI

76 (9)

PDO

75 (8)

The information in this research can be combined with information on other strategies to reduce crashes at intersections with minor-road stop control along rural two-lane roads. With this information, agencies can make informed decisions about planning and programming safety improvements at intersections under their jurisdiction

Chapter 7. References

  1. National Highway Traffic Safety Administration (NHTSA). Fatality Analysis Reporting System (FARS). [cited September 2015]
  2. Federal Highway Administration (FHWA), Focused Approach to Safety. http://safety.fhwa.dot.gov/fas/. [cited September 2015]
  3. Federal Highway Administration (FHWA), Example Intersection Safety Implementation Plan, US Department of Transportation, 2009.
  4. Preston, H., R. Storm, M. Donath, and C. Shankwitz, Review of Minnesota's Rural Intersection Crashes: Methodology for Identifying Intersections for Intersection Decision Support (IDS), Report No. MN/RC-2004-31, Minnesota Department of Transportation, 2004.
  5. Neuman, T. R., R. Pfefer, K. L. Slack, K. K. Hardy, D. W. Harwood, I. B. Potts, D. J. Torbic, and E. R. Kohlman Rabbani, Guidance for Implementation of the AASHTO Strategic Highway Safety Plan, Volume 5: A Guide for Addressing Unsignalized Intersection Collisions, NCHRP Report 500, Volume 5, Transportation Research Board, 2003.
  6. American Association of State Highway and Transportation Officials (AASHTO), Highway Safety Manual, Washington, DC, 2010.
  7. Federal Highway Administration (FHWA), Crash Modification Factors Clearinghouse. http://www.cmfclearinghouse.org/. [cited July 2015]
  8. Elvik, R., and T. Vaa. The Handbook of Road Safety Measures. Elsevier Science, Burlington, MA, 2004.
  9. Elvik, R. Meta-Analysis of Evaluations of Public Lighting as Accident Countermeasure. In Transportation Research Record 1485. TRB, National Research Council, Washington, DC, 1995.
  10. Griffith, M.S. Comparison of the Safety of Lighting Options on Urban Freeways. Public Roads, Vol. 58, No. 2, 1994.
  11. Preston, H., and T. Schoenecker. Safety Impacts of Street Lighting at Rural Intersections. Minnesota Department of Transportation, 1999.
  12. Srinivasan, R., J. Baek, and F. Council, Safety Evaluation of Transverse Rumble Strips on Approaches to Stop Controlled Intersections in Rural Areas, Presented at the 89th Annual Meeting of the Transportation Research Board, Washington, DC, 2010.
  13. Gross, F., R. Jagannathan, B. Persaud, C. Lyon, K. Eccles, N. Lefler, and R. Amjadi, Safety Evaluation of STOP Ahead Pavement Markings, Report No. FHWA-HRT-08-043, Federal Highway Administration, 2007.
  14. Hauer, E. Observational Before-After Studies in Road Safety, Pergamon/Elsevier Science, Inc., Tarrytown, New York, 1997.
  15. Hauer, E., D. W. Harwood, F. M. Council, and M. S. Griffith, "Estimating Safety by the Empirical Bayed Method: A Tutorial," presented at the 81st annual meeting of the Transportation Research Board, January 2002.
  16. Torbic, D. J., J. M. Hutton, C. D. Bokenkroger, K. M. Bauer, D. W. Harwood, D. K. Gilmore, J. M. Dunn, J. J. Ronchetto, E. T. Donnell, H. J. Sommer III, P. Garvey, B. Persaud, and C. Lyon, Guidance for the Design and Application of Shoulder and Centerline Rumble Strips, NCHRP Report 641, Transportation Research Board, Washington DC, 2009.
  17. Torbic, D. J., K. M. Bauer, and J. M. Hutton, "Delta Region Transportation Development Program: Rural Safety Innovation Program Evaluation," Report FHWA-SA-14-029, Federal Highway Administration, 2014.
  18. Srinivasan, R., D. Carter, K. Eccles, B. Persaud, N. Lefler, C. Lyon, and R. Amjadi, Safety Evaluation of Flashing Beacons at STOP-Controlled Intersections, Report No. FWHA-HRT-08-044, Federal Highway Administration, 2007.
  19. Blincoe, L. J., Miller, T. R., Zaloshnja, E., & Lawrence, B. A. The economic and societal impact of motor vehicle crashes, 2010. (Revised). Report No. DOT HS 812 013. Washington, DC: National Highway Traffic Safety Administration, 2015.
  20. United States Department of Transportation (US DOT). "Guidance on Treatment of the Economic Value of a Statistical Life (VSL) in U.S. Department of Transportation Analyses- 2015 Adjustment." Office of the Secretary of Transportation Memorandum, June 17, 2015. Accessed September 11, 2015.
  21. Google Earth™ Mapping Service.
  22. Lutkevich, P., D. McLean, and J. Cheung. FHWA Lighting Handbook, Federal Highway Administration, Washington, DC, 2012.
  23. American Association of State Highway and Transportation Officials (AASHTO), Roadway Lighting Design Guide, Washington, DC, 2005.
  24. Missouri Department of Transportation. MoDOT Engineering Policy Manual. Section 626.4 Transverse Rumble Strips. http://epg.modot.org/index.php?title=626.4_Transverse_Rumble_Strips [cited March 2015]
  25. North Dakota Department of Transportation. CADD Standard Drawings. Standard No. D760-05, Saw Slotted Rumble Strips at Intersections. https://www.dot.nd.gov/divisions/design/docs/standards/D760-05.pdf [cited February 2015]
  26. Kansas Department of Transportation. Highway Sign Manual. 2007.
  27. Federal Highway Administration (FHWA). Manual on Uniform Traffic Control Devices. U.S. Department of Transportation. Washington, DC, 2009A.
  28. Minnesota Department of Transportation. Traffic Engineering Manual. 2015 http://www.dot.state.mn.us/trafficeng/publ/tem/
  29. Pennsylvania Department of Transportation. District Highway Safety Guidance Manual. 2014.
  30. SAS Institute Inc. 2011. SAS/STAT® 9.3 User's Guide. Cary, NC:SAS Institute Inc.

Appendix A. Pennsylvania Design Guidelines for Installation of Pavement Markings

This appendix shows a reproduction of Appendix D, Section 3 of the Pennsylvania District Highway Safety Guidance Manual.

This appendix shows a reproduction of Appendix D, Section 3 of the Pennsylvania District Highway Safety Guidance Manual.

This appendix shows a reproduction of Appendix D, Section 3 of the Pennsylvania District Highway Safety Guidance Manual.

This appendix shows a reproduction of Appendix D, Section 3 of the Pennsylvania District Highway Safety Guidance Manual.

This appendix shows a reproduction of Appendix D, Section 3 of the Pennsylvania District Highway Safety Guidance Manual.

This appendix shows a reproduction of Appendix D, Section 3 of the Pennsylvania District Highway Safety Guidance Manual.

This appendix shows a reproduction of Appendix D, Section 3 of the Pennsylvania District Highway Safety Guidance Manual.

Return to top

Page last modified on February 1, 2017
Safe Roads for a Safer Future - Investment in roadway safety saves lives
Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000