Crash Modification Factors in the Highway Safety Manual: Resources for Evaluation (2023)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Appendix I SPF Adjustment Factors for Segment SPFs from NCHRP Project 17-62 I-1 

Table of Contents 1. Introduction ............................................................................................................................... I-3 2. Process ...................................................................................................................................... I-3 3. Recommended CMFs................................................................................................................ I-3 Table 1. Modify Shoulder Width on Rural Two-Lane Undivided Segments ............................... I-4 Table 2. Modify Right Shoulder Width of Rural Non-Freeway Four-Lane Divided or Undivided Segments............................................................................................................... I-6 Table 3. Install Snowplowable, Permanent Raised Pavement Markers on Rural Two-Lane Undivided Segments............................................................................................................... I-8 Table 4. Change Median Width on Rural and Urban Four-Lane Non-Freeway Divided Segments .............................................................................................................................. I-10 Table 5. Change Vertical Grade of Road on Rural Two-Lane Undivided Segments ................. I-13 Table 6. Install Passing or Climbing Lane on Rural Two-Lane Segments ................................. I-15 Table 7. Install a Short Four-Lane Section on Rural Two-Lane Segments ................................ I-17 Table 8. Modify Lane Width on Rural Two-Lane Undivided Segments .................................... I-19 Table 9. Modify Horizontal Curvature on Rural Two-Lane Undivided Segments ..................... I-21 Table 10. Modify Driveway Density on Rural Two-Lane Undivided Segments........................ I-23 Table 11. Improve Roadside on Rural Two-Lane Undivided Segments .................................... I-25 Table 12. Two-Way Left-Turn Lanes on Rural Two-Lane Undivided Segments ...................... I-27 Table 13. Install Shoulder or Centerline Rumble Strips on Rural or Urban Two-Lane Undivided Segments............................................................................................................. I-29 I-2 

1. INTRODUCTION This document describes the findings from a review of the FHWA CMF Clearinghouse CMFs for the purpose of identifying CMFs that could be used with the SPFs developed in NCHRP Project 17-62 for segments. That project did not identify or produce any CMFs that can be applied with the SPFs developed. These findings were used to develop a set of recommended CMFs for use with the 17-62 SPFs. 2. PROCESS The process used to identify the CMFs was based on (1) the identification of the segment SPFs developed in Project 17-62 and (2) on a SPF-by-SPF basis, a review of the CMFs in the Clearinghouse to identify those CMFs that match the SPF base conditions and can be used with the crash types defined for the Project 17-62 SPFs. Note that the defined crash types for the Project 17-62 SPFs differ by HSM chapter site types. Most of the considered CMFs were required to have a quality rating that met or exceeded one (or both) of the following threshold values. • Star rating: 3, 4, or 5 stars (qualRating) • Numeric rating: ≥ 100 (overallRating) An exception was the existing CMFs provided in the 1st edition of the HSM which were also reviewed. For each treatment where a CMF has been recommended only one study was found to meet the criteria for consideration. 3. RECOMMENDED CMFs The following tables document the recommended CMFs. Each table includes the following: • A description of the treatment • Description of the application circumstance to which the CMF applies (area type, number of lanes, median type, treatment location, intersection vs non-intersection) • Details on the study that provides the CMF and the CMF values/formulae • Recommended CMF for each crash/severity type in the Project 17-62 SPFs to which the CMF may apply • Comments and assumptions related to applying or how the recommended CMF was derived in order to apply to the Project 17-62 SPFs • A reference to the original study Note that CMFs by crash/severity type could not be recommended for all base conditions. I-3 

Table 1. Modify Shoulder Width on Rural Two-Lane Undivided Segments TREATMENT: CMF APPLICATION: Modify shoulder width on rural two-lane undivided segments. • Area type: rural • Number of lanes: two-lane • Median type: undivided • Treatment location: shoulder • These CMFs are applicable to the “non-intersection crash frequency,” which includes all crashes not related to intersections. STUDIES: No. Reference Clearinghouse Jurisdiction Quality Area Number of AADT Crash Crash Severity CMF (std. err.) Number CRFID Data Source (stars/points) Type Lanes Type 1 1 5161-5172 Various Not rated/Not Rural 2 Unknown All non- ROR+HO+SOD+SSD CMF = (CMFra- rated intersection KABCO 1.0)Pra+1.0 (Unknown standard error) Pra is the proportion of Run-off- oad+Head- On+Sideswipe- Opposite- Direction+Sideswipe- Same-Direction COMMENTS AND ASSUMPTIONS: CRASH TYPE AND ESTIMATED EFFECT: 1. CMFs are applicable to HSM chapter 10. 2. Applies to a paved shoulder which is assumed for base condition SPFs. Crash Type Crash CMF (std. err.) 3. Assumes safety effect is the same for all severities where a CMF is recommended. Severity All KABCO (CMFra-1.0)xpra+1.0 KABC KAB KA Single vehicle KABCO Use CMFra KABC KAB I-4 

KA Same direction KABCO Use CMFra KABC KAB KA Opposing KABCO Use CMFra direction KABC KAB KA Formulae for CMFra Shoulder AADT (veh/day) Width < 400 400 to 2000 >2000 0-ft 1.10 1.10+2.50x10- 1.50 4 (AADT-400) 2-ft 1.07 1.07+1.43x10- 1.30 4 (AADT-400) 4-ft 1.02 1.02+8.125x10- 1.15 5 (AADT-400) 6-ft 1.00 1.00 1.00 8-ft or more 0.98 0.98+6.875x10- 0.87 5 (AADT-400) REFERENCES: Harwood, D. W., Council, F. M., Hauer, E., Hughes, W. E., and Vogt, A., "Prediction of the Expected Safety Performance of Rural Two-Lane Highways." FHWA-RD-99-207, McLean, Va., Federal Highway Administration, 2000. I-5 

Table 2. Modify Right Shoulder Width of Rural Non-Freeway Four-Lane Divided or Undivided Segments TREATMENT: CMF APPLICATION: Modify right shoulder width of rural non-freeway 4 lane divided or undivided segments. • Area type: rural • Number of lanes: four-lane • Median type: undivided, divided • Treatment location: paved right shoulder • These CMFs are applicable to the “non-intersection crash frequency,” which includes all crashes not related to intersections. STUDIES: No. Reference Clearinghouse Jurisdiction Quality Area Number of AADT Crash Crash Severity CMF (std. err.) Number CRFID Data Source (stars/points) Type Lanes Type 1 1 5287-5293 CA, KY, MN 4/Not rated Rural 4 Unknown All non- R4D CMF = exp(a(shldwidth-BC)) intersection KABCO SV KABCO 𝑠𝑠. 𝑒𝑒. (𝐶𝐶𝐶𝐶𝐶𝐶) MV KABCO 1 = �𝑒𝑒𝑒𝑒𝑒𝑒[(𝑎𝑎−𝑏𝑏)(𝑠𝑠ℎ𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙ℎ−𝐵𝐵𝐵𝐵)] 2 R4U − 𝑒𝑒𝑒𝑒𝑒𝑒[(𝑎𝑎+𝑏𝑏)(𝑠𝑠ℎ𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙ℎ−𝐵𝐵𝐵𝐵)] � KABCO MVKABCO where shldwidth = paved right shoulder width in feet BC = Desired base condition of shoulder width in feet a = parameter estimate for shoulder width b = the standard error of the parameter estimate for shoulder width, assumed to be largest value giving giving statistical signficance at the 95% confidence level DIVIDED UNDIVIDED KABCO SV MV KABCO MV KABCO KABCO KABCO a -0.118 -0.053 -0.137 -0.067 -0.111 b 0.060 0.027 0.070 0.034 0.057 COMMENTS AND ASSUMPTIONS: CRASH TYPE AND ESTIMATED EFFECT: 1. CMFs are applicable to HSM chapter 11. I-6 

2. PSV and MMV are proportions of single- and and multi-vehicle crashes Crash Type Crash CMF (std. err.) 3. Assumes effect is the same for all severities where a CMF is recommended Severity 4. If formula for CMFSV is less than 1 use value of 0. All KABCO Use KABCO formula KABC KAB KA Single vehicle KABCO Divided – Use SV KABC KABCO formula KAB Undivided – Use KA KABCO and MV KABCO formulae to estimate SV as CMFSV = (CMFKABCO- CMFMV*PMV)/PSV *If CMFSV is predicted to be less than 0 used a value of 0 Same direction KABCO Use MV KABCO KABC formula KAB KA Opposing KABCO Use MV KABCO direction KABC formula KAB KA Intersecting KABCO 1.00 direction KABC 1.00 KAB 1.00 KA 1.00 REFERENCES: Stamatiadis, N., Pigman, J., Sacksteder, J, Ruff, W., and Lord, D. "Impact of Shoulder Width and Median Width on Safety". NCHRP Report 633, Transportation Research Board, Washington D.C., 2009. I-7 

Table 3. Install Snowplowable, Permanent Raised Pavement Markers on Rural Two-Lane Undivided Segments TREATMENT: CMF APPLICATION: Install snowplowable, permanent raised pavement markers on rural two-lane undivided • Area type: rural segments. • Number of lanes: 2 • Median type: undivided • Treatment location: centerline • These CMFs are applicable to the “non-intersection crash frequency,” which includes all crashes not related to intersections. STUDIES: No. Reference Clearinghouse Jurisdiction Quality Area Number of Crash Crash AADT Degree of CMF (std. Number CRFID Data Source (stars/points) Type Lanes Type Severity Curvature err.) 1 1 103 IL, NJ, NY, PA 5/130 Rural 2 Night KABCO <= 5,000 <= 3.5 1.16 (0.03) 2 1 104 IL, NJ, NY, PA 4/130 Rural 2 Night KABCO <= 5,000 > 3.5 1.43 (0.09) 3 1 105 IL, NJ, NY, PA 3/110 Rural 2 Night KABCO 5,001 – <= 3.5 0.99 (0.05) 15,000 4 1 106 IL, NJ, NY, PA 4/130 Rural 2 Night KABCO 5,001 – > 3.5 1.26 (0.10) 15,000 5 1 107 IL, NJ, NY, PA 4/130 Rural 2 Night KABCO 15,001- <= 3.5 0.76 (0.07) 20,000 6 1 108 IL, NJ, NY, PA 3/110 Rural 2 Night KABCO 15,001- > 3.5 1.03 (0.11) 20,000 COMMENTS AND ASSUMPTIONS: CRASH TYPE AND ESTIMATED EFFECT: 1. CMFs are applicable to HSM chapter 10. 2. The AADTs for the sites used to develop the CMFs ranged from 2,850 to 10,944. CMFseverity = (1-PNight) + CMFseverity,night*PNight 3. It is assumed that sites used to develop HSM SPFs did not have PRPM installed. PNight is the proportion of crashes that occur at night 4. Assumes there is no effect on daytime crashes. 5. Assumes effect on night crashes is the same for all crash types Crash Type Crash CMF (std. err.) 6. Assumes effect is same across crash severities for night crashes Severity All KABCO For all categories use the KABC appropriate CMF for night provided and the formula above with the KAB proportion of night crashes in the KA application data Single vehicle KABCO I-8 

KABC KAB KA Same direction KABCO KABC KAB KA Opposing KABCO direction KABC KAB KA REFERENCES: Bahar, G., Mollett, C., Persaud, B., Lyon, C., Smiley, A., Smahel, T., and McGee, H., "NCHRP Report 518: Safety Evaluation of Permanent Raised Pavement Markers." Washington, D.C., Transportation Research Board, National Research Council, (2004). I-9 

Table 4. Change Median Width on Rural and Urban Four-Lane Non-Freeway Divided Segments TREATMENT: CMF APPLICATION: Change median width on rural and urban 4 lane non-freeway divided segments. • Area type: rural, urban • Number of lanes: 4 • Median type: undivided • Treatment location: median • These CMFs are applicable to the “non-intersection crash frequency,” which includes all crashes not related to intersections. STUDIES: No. Reference Clearinghouse Jurisdiction Quality Area Number Crash Crash AADT CMF (std. err.) Number CRFID Data (stars/points) Type of Type Severity Source Lanes 1 1 4523-1371 CA Not rated/Not Rural, 4 All non- KABCO Unknown rated Urban intersection Total Crashes Cross-Median Crashes Site Type a b a b R4D -0.00461 0.00080 -0.01695 0.00200 U4D -0.00533 0.00090 -0.01340 0.00205 CMF = exp(a(MedianWidth-BC)) 𝑠𝑠. 𝑒𝑒. (𝐶𝐶𝐶𝐶𝐶𝐶) 1 = �𝑒𝑒𝑒𝑒𝑒𝑒[(𝑎𝑎−𝑏𝑏)(𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀ℎ−𝐵𝐵𝐵𝐵)] 2 − 𝑒𝑒𝑒𝑒𝑒𝑒[(𝑎𝑎+𝑏𝑏)(𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀ℎ−𝐵𝐵𝐵𝐵)] � where BC is the base condition median width COMMENTS AND ASSUMPTIONS: CRASH TYPE AND ESTIMATED EFFECT: 1. CMFs are applicable to HSM chapter 11. 2. 17-62 SPF calibration data had average of 68 feet median for rural roads and 33 feet Total Crashes Cross-Median Crashes Site Type a b a b for urban roads R4D -0.00461 0.00080 -0.01695 0.00200 3. Assumes effect is the same for all severities where CMF is provided. U4D -0.00533 0.00090 -0.01340 0.00205 CMF = exp(a(MedianWidth-BC)) I-10 

1 𝑠𝑠. 𝑒𝑒. (𝐶𝐶𝐶𝐶𝐶𝐶) = �𝑒𝑒𝑒𝑒𝑒𝑒[(𝑎𝑎−𝑏𝑏)(𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀ℎ−𝐵𝐵𝐵𝐵)] − 𝑒𝑒𝑒𝑒𝑒𝑒[(𝑎𝑎+𝑏𝑏)(𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀ℎ−𝐵𝐵𝐵𝐵)] � 2 R4D SPFs Crash Type Crash CMF (std. err.) Severity All KABCO Use formula KABC KAB KA Single vehicle KABCO 1.00 KABC 1.00 KAB 1.00 KA 1.00 Same direction KABCO 1.00 KABC 1.00 KAB 1.00 KA 1.00 Opposing KABCO Use formula direction KABC KAB KA U4D SPFs Crash Type Crash CMF (std. err.) Severity All KABCO Use formula KABC KAB MVN KABCO No recommendation RE KABCO 1.00 SSD KABCO 1.00 HO+SOD KABCO Use cross-median formula MVNOTHER KABCO 1.00 I-11 

MVD KABCO 1.00 SV KABCO 1.00 NIGHT KABCO No recommendation PEDESTRIAN KABCO 1.00 REFERENCES: Harkey, D.L., R. Srinivasan, J. Baek, B. Persaud, C. Lyon, F.M. Council, K. Eccles, N. Lefler, F. Gross, E. Hauer, J. Bonneson, "Crash Reduction Factors for Traffic Engineering and ITS Improvements", NCHRP Project 17-25 Final Report, Washington, D.C., National Cooperative Highway Research Program, Transportation Research Board, (2008). I-12 

Table 5. Change Vertical Grade of Road on Rural Two-Lane Undivided Segments TREATMENT: CMF APPLICATION: Change vertical grade of road on rural two-lane undivided segments. • Area type: rural • Number of lanes: 2 • Median type: undivided • Treatment location: roadway • These CMFs are applicable to the “non-intersection crash frequency,” which includes all crashes not related to intersections. STUDIES: No. Reference Clearinghouse Jurisdiction Quality Area Number Crash Crash AADT CMF (std. err.) Number CRFID Data Source (stars/points) Type of Lanes Type Severity 1 1 4642 UT Not rated/Not Rural 2 All non- KABCO Unknown KABCO CMF = 1.016G rated intersection where G=gradient of road as a percent (Unknown standard error) COMMENTS AND ASSUMPTIONS: CRASH TYPE AND ESTIMATED EFFECT: 1. CMFs are applicable to HSM chapter 10. 2. Assumes effect is the same for all crash types and severities. Crash Type Crash CMF (std. err.) 3. Assumes base condition of 0% grade in calibration data for 17-62 SPFs. Severity All KABCO Use formula KABC KAB KA Single vehicle KABCO Use all crash KABC KABCO CMF KAB KA Same direction KABCO Use all crash KABC KABCO CMF KAB I-13 

KA Opposing KABCO Use all crash direction KABCO CMF KABC KAB KA REFERENCES: Harwood, D. W., Council, F. M., Hauer, E., Hughes, W. E., and Vogt, A., "Prediction of the Expected Safety Performance of Rural Two-Lane Highways." FHWA-RD-99-207, McLean, Va., Federal Highway Administration, 2000. I-14 

Table 6. Install Passing or Climbing Lane on Rural Two-Lane Segments TREATMENT: CMF APPLICATION: Install passing or climbing lane on rural two-lane segments. • Area type: rural • Number of lanes: 2 • Median type: undivided • Treatment location: n/a • These CMFs are applicable to the “non-intersection crash frequency,” which includes all crashes not related to intersections. STUDIES: No. Reference Clearinghouse Jurisdiction Quality Area Number Crash Crash AADT CMF (std. err.) Number CRFID Data Source (stars/points) Type of Lanes Type Severity 1 1 4654 Unknown Not rated/Not Rural 2 All non- KABCO Unknown 0.75 rated intersection (Unknown standard error) COMMENTS AND ASSUMPTIONS: CRASH TYPE AND ESTIMATED EFFECT: 1. CMFs are applicable to HSM chapter 10. 2. CMF assumes the passing or climbing lane is operationally warranted and of adequate Crash Type Crash CMF (std. err.) length. Severity 3. Assumes safety effect is the same for all severities where CMF is recommended. All KABCO 0.75 4. CMF for same direction and opposite direction crashes derived using percentage of KABC 0.75 same direction and opposite direction crashes in the 17-62 calibration data (38%) and KAB 0.75 assuming treatment affects both crash types equally but no effect on single-vehicle KA 0.75 crashes. Single vehicle KABCO 1.00 KABC 1.00 KAB 1.00 KA 1.00 Same direction KABCO 0.34 KABC 0.34 KAB 0.34 KA 0.34 Opposing KABCO 0.34 direction KABC 0.34 I-15 

KAB 0.34 KA 0.34 REFERENCES: Harwood, D. W., Council, F. M., Hauer, E., Hughes, W. E., and Vogt, A., "Prediction of the Expected Safety Performance of Rural Two-Lane Highways." FHWA-RD-99-207, McLean, Va., Federal Highway Administration, 2000. I-16 

Table 7. Install a Short Four-Lane Section on Rural Two-Lane Segments TREATMENT: CMF APPLICATION: Install a short four-lane section on rural two-lane segments. • Area type: rural • Number of lanes: 2 • Median type: undivided • Treatment location: n/a • These CMFs are applicable to the “non-intersection crash frequency,” which includes all crashes not related to intersections. STUDIES: No. Reference Clearinghouse Jurisdiction Quality Area Number Crash Crash AADT CMF (std. err.) Number CRFID Data Source (stars/points) Type of Lanes Type Severity 1 1 4655 NY, OR, WA Not rated/Not Rural 2 All non- KABCO Unknown 0.65 rated intersection (Unknown standard error) COMMENTS AND ASSUMPTIONS: CRASH TYPE AND ESTIMATED EFFECT: 1. CMFs are applicable to HSM chapter 10. 2. Assumes effect is the same for all severities where CMF is recommended. Crash Type Crash CMF (std. err.) 3. CMF for same direction and opposite direction crashes derived using percentage of Severity same direction and opposite direction crashes in the 17-62 calibration data (38%) and All KABCO 0.65 assuming treatment affects both crash types equally but no effect on single-vehicle KABC 0.65 crashes. KAB 0.65 KA 0.65 Single vehicle KABCO 1.00 KABC 1.00 KAB 1.00 KA 1.00 Same direction KABCO 0.08 KABC 0.08 KAB 0.08 KA 0.08 Opposing KABCO 0.08 direction KABC 0.08 I-17 

KAB 0.08 KA 0.08 REFERENCES: Harwood, D. W., Council, F. M., Hauer, E., Hughes, W. E., and Vogt, A., "Prediction of the Expected Safety Performance of Rural Two-Lane Highways." FHWA-RD-99-207, McLean, Va., Federal Highway Administration, 2000. I-18 

Table 8. Modify Lane Width on Rural Two-Lane Undivided Segments TREATMENT: CMF APPLICATION: Modify lane width on rural two-lane undivided segments. • Area type: rural • Number of lanes: 2 • Median type: undivided • Treatment location: Roadway • These CMFs are applicable to the “non-intersection crash frequency,” which includes all crashes not related to intersections. STUDIES: No. Reference Clearinghouse Jurisdiction Quality Area Number Crash Crash AADT CMF (std. err.) Number CRFID Data Source (stars/points) Type of Lanes Type Severity 1 1 5173-5181 Various Not rated/Not Rural 2 All non- KABCO Unknown CMF formulae provided in rated intersection table below (Unknown standard error) Pra is the proportion of Run- off-oad+Head- On+Sideswipe-Opposite- Direction+Sideswipe-Same- Direction Formulae for CMF ra Lane Width AADT (veh/day) < 400 400 to 2000 >2000 9-ft or less 1.05 1.05+2.81x10- 1.50 4 (AADT-400) 10-ft 1.02 1.02+1.75x10- 1.30 4 (AADT-400) 11-ft 1.01 1.01+2.50x10- 1.05 5 (AADT-400) 12-ft or more 1.00 1.00 1.00 COMMENTS AND ASSUMPTIONS: CRASH TYPE AND ESTIMATED EFFECT: 1. CMFs are applicable to HSM chapter 10. I-19 

2. If lane widths differ by direction CMFs are determined for each direction and then averaged. Crash Type Crash CMF (std. err.) 3. Assumes effect is the same for all crash types and severities Severity All KABCO Use formula KABC KAB KA Single vehicle KABCO KABC KAB KA Same direction KABCO KABC KAB KA Opposing KABCO direction KABC KAB KA REFERENCES: Harwood, D. W., Council, F. M., Hauer, E., Hughes, W. E., and Vogt, A., "Prediction of the Expected Safety Performance of Rural Two-Lane Highways." FHWA-RD-99-207, McLean, Va., Federal Highway Administration, 2000. I-20 

Table 9. Modify Horizontal Curvature on Rural Two-Lane Undivided Segments TREATMENT: CMF APPLICATION: Modify horizontal curvature on rural two-lane undivided segments. • Area type: rural • Number of lanes: 2 • Median type: undivided • Treatment location: n/a • These CMFs are applicable to the “non-intersection crash frequency,” which includes all crashes not related to intersections. STUDIES: No. Reference Clearinghouse Jurisdiction Quality Area Number of Crash Crash AADT Degree of CMF (std. err.) Number CRFID Data Source (stars/points) Type Lanes Type Severity Curvature 1 1 5182 WA Not rated/130 Rural 2 All non- KABCO Unknown Unknown 𝐶𝐶𝐶𝐶𝐶𝐶 intersection 80.2 1.55𝐿𝐿𝑐𝑐 + − 0.012𝑆𝑆 𝑅𝑅 = 1.55𝐿𝐿𝑐𝑐 (Unknown standard error) COMMENTS AND ASSUMPTIONS: CRASH TYPE AND ESTIMATED EFFECT: 1. CMFs are applicable to HSM chapter 10. 2. Assumes effect is the same for all crash types and severities. 80.2 1.55𝐿𝐿𝑐𝑐 + − 0.012𝑆𝑆 𝑅𝑅 𝐶𝐶𝐶𝐶𝐶𝐶 = 1.55𝐿𝐿𝑐𝑐 Lc = length of horizontal curve in miles S = 1 is spiral transition present and 0 if not R = radius of curvature in feet Crash Type Crash CMF (std. Severity err.) All KABCO Use formula KABC above KAB KA Single vehicle KABCO I-21 

KABC (Unknown KAB standard error) KA Same direction KABCO KABC KAB KA Opposing KABCO direction KABC KAB KA REFERENCES: Harwood, D. W., Council, F. M., Hauer, E., Hughes, W. E., and Vogt, A., "Prediction of the Expected Safety Performance of Rural Two-Lane Highways." FHWA-RD-99-207, McLean, Va., Federal Highway Administration, 2000. I-22 

Table 10. Modify Driveway Density on Rural Two-Lane Undivided Segments TREATMENT: CMF APPLICATION: Modify driveway density on rural two-lane undivided segments. • Area type: rural • Number of lanes: 2 • Median type: undivided • Treatment location: roadside • These CMFs are applicable to the “non-intersection crash frequency,” which includes all crashes not related to intersections. STUDIES: No. Reference Clearinghouse Jurisdiction Quality Area Number Crash Crash AADT CMF (std. err.) Number CRFID Data Source (stars/points) Type of Lanes Type Severity 1 1 5185 International Not rated/Not Rural 2 All non- KABCO Unknown 𝐶𝐶𝐶𝐶𝐶𝐶 = rated intersection 0.2+[0.05−0.005 ln(𝐴𝐴𝐴𝐴𝐴𝐴)]𝐷𝐷𝐷𝐷 0.2+[0.05−0.005 ln(𝐴𝐴𝐴𝐴𝐴𝐴)]5 (Unknown standard error) COMMENTS AND ASSUMPTIONS: CRASH TYPE AND ESTIMATED EFFECT: 1. CMFs are applicable to HSM chapter 10. 2. Assumes that driveway density is equal to 5 per mile in the 17-62 SPF calibration data. 3. Assumes effect is the same for all severities where CMF is recommended. 0.2 + [0.05 − 0.005 ln(𝐴𝐴𝐴𝐴𝐴𝐴)]𝐷𝐷𝐷𝐷 𝐶𝐶𝐶𝐶𝐶𝐶 = 0.2 + [0.05 − 0.005 ln(𝐴𝐴𝐴𝐴𝐴𝐴)]5 ADT = annual average daily traffic volume DD = driveway density (driveways per mile) Crash Type Crash CMF (std. err.) Severity All KABCO Use formula KABC KAB (Unknown standard error) KA Single vehicle KABCO 1.00 KABC 1.00 I-23 

KAB 1.00 KA 1.00 Same direction KABCO No recommendation KABC KAB KA Opposing KABCO No recommendation direction KABC KAB KA REFERENCES: Harwood, D. W., Council, F. M., Hauer, E., Hughes, W. E., and Vogt, A., "Prediction of the Expected Safety Performance of Rural Two-Lane Highways." FHWA-RD-99-207, McLean, Va., Federal Highway Administration, 2000. I-24 

Table 11. Improve Roadside on Rural Two-Lane Undivided Segments TREATMENT: CMF APPLICATION: Improve roadside on rural two-lane undivided segments. • Area type: rural • Number of lanes: 2 • Median type: undivided • Treatment location: roadside • These CMFs are applicable to the “non-intersection crash frequency,” which includes all crashes not related to intersections. STUDIES: No. Reference Clearinghouse Jurisdiction Quality Area Number Crash Crash AADT CMF (std. err.) Number CRFID Data Source (stars/points) Type of Lanes Type Severity 1 1 5187 MN, WA Not rated/Not Rural 2 All non- KABCO 159 to 𝑒𝑒𝑒𝑒𝑒𝑒(−0.6869+0.0668×𝑅𝑅𝑅𝑅𝑅𝑅) KABCO 𝐶𝐶𝐶𝐶𝐶𝐶 = rated intersection 17,766 exp (−0.4865) 𝑠𝑠. 𝑒𝑒. (𝐶𝐶𝐶𝐶𝐶𝐶) 𝑒𝑒𝑒𝑒𝑒𝑒(−0.6869+0.879×𝑅𝑅𝑅𝑅𝑅𝑅) 𝑒𝑒𝑒𝑒𝑒𝑒(−0.6869+0.0457×𝑅𝑅𝑅𝑅𝑅𝑅) � − � exp (−0.4865) exp (−0.4865) = 2 RHR = roadside hazard rating COMMENTS AND ASSUMPTIONS: CRASH TYPE AND ESTIMATED EFFECT: 1. CMFs are applicable to HSM chapter 10. 2. Assumes that there is no effect on same-direction or opposing-direction crashes. PSV = Proportion of crashes that are single-vehicle 3. Assumes effect is the same for all severities where CMF is recommended. PNON-SV = Proportion of crashes that are not single-vehicle Crash Type Crash CMF (std. err.) Severity All KABCO KABC Use formula KAB KA Single vehicle KABCO (CMFALL-PNON-SV)/PSV KABC I-25 

KAB KA Same direction KABCO 1.00 KABC 1.00 KAB 1.00 KA 1.00 Opposing KABCO 1.00 direction KABC 1.00 KAB 1.00 KA 1.00 REFERENCES: Harwood, D. W., Council, F. M., Hauer, E., Hughes, W. E., and Vogt, A., "Prediction of the Expected Safety Performance of Rural Two-Lane Highways." FHWA-RD-99-207, McLean, Va., Federal Highway Administration, 2000. I-26 

Table 12. Two-Way Left-Turn Lanes on Rural Two-Lane Undivided Segments TREATMENT: CMF APPLICATION: Two-way left-turn lanes on rural two-lane undivided segments. • Area type: rural • Number of lanes: 2 • Median type: undivided • Treatment location: median\ • These CMFs are applicable to the “non-intersection crash frequency,” which includes all crashes not related to intersections. STUDIES: No. Reference Clearinghouse Jurisdiction Quality Area Number Crash Crash AADT CMF (std. err.) Number CRFID Data Source (stars/points) Type of Lanes Type Severity 1 1 583 AR, CA, IL, 5/140 Rural 2 All non- KABCO Unknown 0.64 (0.04) NC intersection 2 1 584 AR, CA, IL, 4/135 Rural 2 All non- KAB Unknown 0.65 (0.08) NC intersection 3 1 585 AR, CA, IL, 5/140 Rural 2 Rear-End KABCO Unknown 0.53 (0.05) NC COMMENTS AND ASSUMPTIONS: CRASH TYPE AND ESTIMATED EFFECT: 1. CMFs are applicable to HSM chapter 10. 2. Assumes that there is no safety effect for single-vehicle crashes. Crash Type Crash CMF (std. err.) 3. CMF for KAB is only off by 0.01 from KABCO CMF so use KABCO CMF. Severity 4. Assumes all same direction crashes are affected in the same way as rear-end crashes. All KABCO 0.64 KABC 0.64 KAB 0.64 KA 0.64 Single vehicle KABCO 1.00 KABC 1.00 KAB 1.00 KA 1.00 Same direction KABCO 0.53 KABC 0.53 KAB 0.53 KA 0.53 I-27 

Opposing KABCO No recommended direction CMF KABC No recommended CMF KAB No recommended CMF KA No recommended CMF REFERENCES: Persaud, B., Lyon, C., Eccles, K., Lefler, N., Carter, D., and Amjadi, R., "Safety Evaluation of Installing Center Two-Way Left-Turn Lanes on Two-Lane Roads", Report No. FHWA-HRT-08-042, Federal Highway Administration, 2007. I-28 

Table 13. Install Shoulder or Centerline Rumble Strips on Rural or Urban Two-Lane Undivided Segments TREATMENT: CMF APPLICATION: Install shoulder or centerline rumble strips on rural or urban two-lane undivided segments. • Area type: rural, urban • Number of lanes: 2 • Median type: undivided • Treatment location: centerline • These CMFs are applicable to the “non-intersection crash frequency,” which includes all crashes not related to intersections. STUDIES: No. Reference Treatment Clearinghouse Jurisdiction Quality Area Number Crash Crash AADT CMF (std. err.) Number CRFID Data Source (stars/points) Type of Lanes Type Severity 1 1 Shoulder n/a MN, MO, PA, not rated/not Rural 2 Single- KABCO Unknown 0.85 (0.07) rumble strip MN rated vehicle 2 1 Shoulder n/a MN, MO, PA, not rated/not Rural 2 Single- KABC Unknown 0.71 (0.09) rumble strip MN rated vehicle 3 1 Centerline n/a MN, PA, WA. not rated/not Rural 2 All KABCO Unknown 0.91 (0.02) rumble strip MN rated 4 1 Centerline n/a MN, PA, WA, not rated/not Rural 2 All KABC Unknown 0.88 (0.03) rumble strip MN rated 5 1 Centerline n/a MN, PA, WA, not rated/not Rural 2 Opposite- KABCO Unknown 0.70 (0.05) rumble strip MN rated direction 6 1 Centerline n/a MN, PA, WA, not rated/not Rural 2 Opposite- KABC Unknown 0.56 (0.06) rumble strip MN rated direction 7 1 Centerline 3613 PA/? 3/not rated Urban 2 Opposite- KABCO Unknown 0.60 (0.17) rumble strip direction COMMENTS AND ASSUMPTIONS: CRASH TYPE AND ESTIMATED EFFECT: 1. CMFs are applicable to HSM chapters 10 and 12. Rural Two-Lane Undivided 2. Note that recommended CMFs from reference 1 were a combination of the CMFs developed in reference 1 and previous studies. Crash Type Crash Shoulder Rumble Centerline Rumble 3. Assumes all single-vehicle crashes are run-off-road. Severity Strips CMF (std. Strips CMF (std. 4. Assumes centerline rumble strip target crashes are all opposite-direction crashes. err.) err.) All KABCO 0.91 0.91 KABC 0.82 0.88 KAB 0.82 0.88 I-29 

KA 0.82 0.88 Single vehicle KABCO 0.85 1.00 KABC 0.71 1.00 KAB 0.71 1.00 KA 0.71 1.00 Same direction KABCO 1.00 1.00 KABC 1.00 1.00 KAB 1.00 1.00 KA 1.00 1.00 Opposing KABCO 1.00 0.70 direction KABC 1.00 0.56 KAB 1.00 0.56 KA 1.00 0.56 Urban Two-Lane Undivided Crash Type Crash Centerline Rumble Severity Strip CMF (std. err.) All KABCO 0.91 KABC 0.88 KAB 0.88 MVN KABCO No CMF recommended RE KABCO 1.00 SSD KABCO 1.00 HO+SOD KABCO 0.60 MVNOTHER KABCO 1.00 MVD KABCO 1.00 SV KABCO 1.00 NIGHT KABCO No CMF recommended PEDESTRIAN KABCO 1.00 REFERENCES: 1. NCHRP Report 641: Guidance for the Design and Application of Shoulder and Centerline Rumble Strips, 2009. I-30