Factors Contributing to Median Encroachments and Cross-Median Crashes (2014)

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.

23 This section presents the combined results from three recent surveys of highway agencies related to median design practices: a 2003 survey conducted as part of NCHRP Proj- ect 17-14 (47); a 2006 survey conducted as part of NCHRP Project 22-21 (22); and a 2009 survey conducted as part of the current research. 3.1 Survey Method A survey of state highway agencies concerning their median design practices was conducted in 2003 as part of NCHRP Project 17-14. This previous survey was updated in 2006 as part of the research conducted for NCHRP Project 22-21. That survey was not identical to the earlier survey, but did contain several of the same questions relating to typical median cross sections and use of median barriers. In 2009, the survey was once again updated for the current research, containing some of the same questions found in previous surveys, but includ- ing new questions as well. Both the 2003 and 2006 surveys were sent to the design engineers of the 50 state highway agencies. The 2009 survey was sent to design and traffic engineers at the 50 state high- way agencies, as well as to engineers at 9 of the Canadian pro- vincial highway agencies, and 19 toll and turnpike agencies in the United States. The survey was distributed by email as a link to the survey website. To avoid duplication of effort, sur- vey respondents were first asked whether their median design policies had changed since 2006. If the agency had responded to the 2006 survey and their response to this first question in the 2009 survey indicated that the agency’s median design policies had not changed since 2006, many of the subsequent questions that had already been answered in the previous sur- vey were not repeated. The summary of survey results presented below is based on the combined results of the 2003, 2006, and 2009 surveys for all cases in which common questions were asked. 3.2 Response Rate The 2003 survey received responses from 37 of the 50 states, or 74 percent. The 2006 survey received responses from 34 of the 50 states, or 68 percent. Of the 78 agencies that were sent the 2009 survey, responses were received from 22 state highway agencies, 3 Canadian provincial agencies, and 6 turnpike or toll road authorities, for a response rate of 44 percent for the state highway agencies and 40 percent for all survey recipients. The combination of all three surveys includes responses from 47 states (or 94 percent). Table 3-1 lists the agencies that responded to each of the surveys. Table 3-2 summarizes the state agency responses concern- ing changes in median design policies from the 22 states that responded to the 2009 survey. No Canadian provincial agen- cies or toll authorities responded to this question. Half of the respondents reported that their median policy had not changed since 2006, and only three agencies reported a change in their policy between 2006 and 2009. Table 3-3 lists all of the state agencies that responded to any of the surveys and indi- cates whether there were median design policy changes from 2003 to 2006 or from 2006 to 2009. 3.3 Survey Summary 3.3.1 Median Design Criteria Respondents were asked whether their agency uses design criteria for highway medians that differ from AASHTO’s Pol- icy on Geometric Design of Highways and Streets, commonly known as the Green Book (48, 49, 50). There were 15 responses to this question. One state highway agency, one Canadian provincial agency, and one toll authority responded that they had median design criteria that differed from that provided in the AASHTO Green Book. All responses to this question are summarized in Table 3-4. S E C T I O N 3 Survey Results

24 Agencies responding to the 2003 survey Agencies responding to the 2006 survey Agencies responding to the 2009 survey Alabama Alabama Alaska Alaska Arkansas Arizona Arizona California Florida^ Arkansas Connecticut Georgia* California Delaware Indiana^ Colorado Florida Iowa^ Connecticut Idaho Kentucky Delaware Indiana Maryland^ Florida Iowa Massachusetts Hawaii Kentucky Michigan Indiana Maine Mississippi^ Iowa Maryland Montana^ Kansas Minnesota New Mexico Maine Mississippi North Carolina^ Maryland Missouri Ohio^ Massachusetts Montana Oregon Michigan Nebraska Rhode Island* Minnesota Nevada South Carolina^ Mississippi New Jersey South Dakota^ Missouri New Mexico Texas Montana New York Utah* Nebraska North Carolina Vermont* Nevada Ohio Canadian Provincial Agencies New Hampshire Oregon British Columbia New Jersey Pennsylvania Saskatchewan New York South Carolina New Brunswick North Carolina South Dakota Toll and Turnpike Authorities North Dakota Tennessee Florida's Turnpike Enterprise Ohio Texas Illinois State Toll Highway Authority Pennsylvania Virginia Kansas Turnpike Authority South Carolina Washington New York State Thruway Authority South Dakota West Virginia North Texas Tollway Authority Virginia Wisconsin Oklahoma Turnpike Authority Washington Wyoming West Virginia Wisconsin Wyoming ^ Indicates states that responded to all three surveys. * Indicates states that responded only to the 2009 survey. Table 3-1. State highway agencies that responded to each of the surveys. Response Number (percentage) of highway agencies No change in policy since 2006 11 (50.0) Policies have changed since 2006 3 (13.6) No response 8 (36.4) Total 22 (100.0) Table 3-2. Changes in median design policies between 2006 and 2009 indicated by state agencies that responded to the 2009 survey.

25 3.3.2 Median Barrier Design Criteria Respondents were asked whether their agency uses design criteria for highway median barriers that differ from the AASHTO Roadside Design Guide (46). There were 18 responses to this question. One state highway agency and two Cana- dian provincial agencies responded that they had median bar rier design criteria that differed from that provided in the AASHTO Roadside Design Guide. All responses to this ques- tion are summarized in Table 3-5. 3.3.3 Median Barrier Warrant Criteria Highway agencies were asked in both the 2003 and 2006 surveys if they used the median barrier warrants in the 2002 AASHTO Roadside Design Guide. Table 3-6 indicates that 20 of the 30 states (66.7 percent) that responded to the ques- tion in at least one of the two surveys indicate that they use the AASHTO median barrier warrants. In the 2009 survey, the question was asked in an open-ended format. Agencies were asked simply to identify any median barrier warrants that their agency uses. Twelve agencies, including seven state highway agencies, responded to the 2009 question. Three state agencies and one toll agency indicated, in response to the open-ended question, that they use AASHTO median barrier warrants. All responses to this question in the 2009 survey are combined with the responses from the 10 states that indicated they did not use the AASHTO median bar- rier warrants in the 2003 and 2006 surveys. The responses are shown in Table 3-7 and include the most recent response Agency 2006 Response 2009 Response Alabama Alaska Arizona No change from 2003 Did not complete 2006 survey Did not complete 2006 survey Did not complete 2009 survey Did not respond to this question Did not respond to this question Arkansas Did not respond to this question Did not complete 2009 survey California No change from 2003 Did not complete 2009 survey Colorado Did not complete 2006 survey Did not complete 2009 survey Delaware New policies or practices since 2003 Did not complete 2009 survey Florida Did not respond to this question No change from 2006 Georgia Did not complete 2003 or 2006 survey Did not respond to this question Hawaii Did not complete 2006 survey Did not complete 2009 survey Indiana New policies or practices since 2003 No change from 2006 Iowa New policies or practices since 2003 No change from 2006 Kansas Did not complete 2006 survey Did not complete 2009 survey Maine No change from 2003 Did not respond to the 2003 survey Did not complete 2009 survey Maryland No change from 2006 Massachusetts Did not complete 2006 survey Did not respond to this question Michigan No change from 2003 Did not respond to this question Minnesota New policies or practices since 2003 Did not complete 2009 survey Mississippi No change from 2003 New policies or practices since 2006 Missouri New policies or practices since 2003 Did not complete 2009 survey Montana No change from 2003 No change from 2006 Nebraska New policies or practices since 2003 Did not complete 2009 survey Nevada No change from 2003 Did not complete 2009 survey New Jersey New policies or practices since 2003 Did not complete 2009 survey New Mexico No change from 2003 New policies or practices since 2006 New York New policies or practices since 2003 Did not complete 2009 survey North Carolina No change from 2003 No change from 2006 North Dakota Did not complete 2006 survey Did not complete 2009 survey Oregon Did not respond to the 2003 survey New policies or practices since 2006 Pennsylvania No change from 2003 Did not complete 2009 survey Rhode Island Did not complete 2003 or 2006 survey Did not respond to this question South Carolina No change from 2003 No change from 2006 South Dakota New policies or practices since 2003 No change from 2006 Tennessee No change from 2003 Did not complete 2009 survey Texas Did not respond to the 2003 survey No change from 2006 Utah Did not complete 2003 or 2006 survey Did not respond to this question Vermont Did not complete 2003 or 2006 survey Did not respond to this question Virginia Did not respond to the 2003 survey Did not complete 2009 survey Washington Did not respond to the 2003 survey Did not complete 2009 survey West Virginia New policies or practices since 2003 Did not complete 2009 survey Wisconsin New policies or practices since 2003 Did not complete 2009 survey Wyoming New policies or practices since 2003 Did not complete 2009 survey Connecticut No change from 2003 Did not complete 2009 survey Idaho Did not respond to the 2003 survey Did not complete 2009 survey Kentucky Did not respond to the 2003 survey No change from 2006 New Hampshire Did not complete 2006 survey Did not complete 2009 survey Ohio No change from 2003 No change from 2006 Table 3-3. Response from specific state agencies about changes in median cross-section design policies from 2003 to 2006 and from 2006 to 2009.

26 Agency Agency type Response Design criteria different from AASHTO Green Book Agency design criteria Alaska State Highway Agency No Arizona State Highway Agency No Georgia State Highway Agency Yes From the GDOT Design Policy Manual available through the following link: http://www.dot.ga.gov/doingbusiness/PoliciesManu als/roads/Pages/DesignPolicies.aspx 6.8. Medians–Several factors will be considered when determining the applicable median treatments, such as classification of roadway, number of lanes, base year traffic, design year traffic, posted speed limit, design speed limit, and accident/crash data. Below are the roadway classifications and the median guidelines for those classifications. 6.8.1. Interstate Medians–All Interstates shall require a depressed median, as specified in the AASHTO Green Book (2004), or positive barrier separation in areas of right-of-way restrictions. Positive barrier separation is required for all median widths 52-ft or where mutually exclusive clear zone for each direction of traffic cannot be obtained. Positive barrier separation will not be required for median widths > 64-ft. Median barrier is optional for median widths between 52-ft and 64- ft. Positive barrier separation should be considered for all existing medians where there is a history of cross-median type accidents. 6.8.2. Arterial (Non- GRIP) Medians with Posted Speeds or Design Speeds < 45 mph Median ADT (Base Year) ADT (Design Year) 5-lane section (paved median) < 18,000 < 24,000 5-lane section (paved median) (1) < 18,000 > 24,000 20-ft raised median (2) > 18,000 > 24,000 NOTES: (1) The project shall be designed to incorporate a future 20-ft raised median or preferably a 24-ft raised median depending on impacts. Right-of-way shall be purchased for footprint determined by raised 20-ft or 24-ft median typical section. The need and implementation of a raised median section shall be determined by monitoring of accidents and traffic volumes on a 5- year cycle by the Safety Engineer in the GDOT Office of Traffic Operations. (2) GDOT prefers the use of a 24-ft raised median if there are minimal impacts associated with a wider median. Raised medians shall be constructed on multilane facilities at intersections that exhibit one of the following characteristics: high turning volumes relating to 18,000 ADT (base year) and 24,000 ADT (design year) accident rate greater than the state average for its classification excessive queue lengths (as determined by District Traffic Engineer) in conjunction with excessive number of driveways. Table 3-4. Survey responses concerning use of AASHTO Green Book for median design criteria.

27 Agency Agency type Response Design criteria different from AASHTO Green Book Agency design criteria more lanes in each direction shall include positive separation of opposing traffic using a median. The type of median required shall depend on guidelines stated above. All rural multilane roadways interchanging with an Interstate highway shall have a raised median for a minimum distance of 1,000 ft from the ramp termini or the first major intersection. A median break may be provided in accordance with GDOT’s access guidelines. Massachusetts State Highway Agency No Mississippi State Highway Agency No Oregon State Highway Agency No Rhode Island State Highway Agency No Vermont State Highway Agency No Illinois State Toll Highway Authority No Kansas Turnpike Authority No New York Thruway Authority No North Texas TollwayAuthority No Oklahoma Turnpike Authority Yes In the past, medians were 15 ft but since 1991 we use AASHTO. British Columbia Canadian Provincial Agency Yes http://www.th.gov.bc.ca/publications/eng_publicatio ns/geomet/TAC/TAC_2007_Supplement/Ch400- 2007.pdf. Please refer to Fig 440.B - 440.D. Saskatchewan Canadian Provincial Agency No All arterials with design speeds greater than 45 mph will require: a 24-ft raised median with a sloped curb (Type 7 curb-face), which will require a 2-ft additional paved shoulder offset from the edge of travel to the edge of the gutter (4-ft inside shoulder width from the edge of travel to the face of the curb). A 44-ft depressed median or a positive barrier system depending upon functional classification, the type of development along the corridor, type of access management and right-of- way impacts. All multilane facilities with three or Table 3-4. (Continued).

ze Figure 6.1 Agency Agency type Response Design criteria different from AASHTO Green Book Agency design criteria Alaska State Highway Agency No Arizona State Highway Agency Yes Our guidelines are located at: http://www.azdot.gov/highways/Roadway_Engineering/ Roadway_Design/Guidelines/Manuals/PDF/Roadway DesignGuidelines.pdf Median barriers are covered in Sections 304.4, 305, 305.9, and 305.11. Median barriers shall be installed on: a) Rural high-speed controlled-access highways with a median width 30 ft and less. From > 30 ft to < 50 ft, utili of the AASHTO Roadside Design Guide; b) urban freeway sections with median widths 50 ft and less; c) all freeway sections with median widths 75 ft and less when there are three or more through lanes in each direction and natural barriers are not present. Georgia State Highway Agency No Massachusetts State Highway Agency No Michigan State Highway Agency No Mississippi State Highway Agency No New Mexico State Highway Agency No Oregon State Highway Agency No Rhode Island State Highway Agency No Vermont State Highway Agency No British Columbia Canadian Provincial Agency Yes 680 mm for roadside barrier; 810 mm for median barrier New Brunswick Canadian Yes TAC Geometric Design Guide for Canadian Roads Provincial Agency Saskatchewan Canadian Provincial Agency No Illinois State Toll Highway Authority No Kansas Turnpike Authority No New York Thruway Authority No North Texas TollwayAuthority No Oklahoma Turnpike Authority No Table 3-5. Survey responses concerning use of AASHTO Roadside Design Guide for median barrier design criteria.

29 Agency Does your agency use the 2002 AASHTO median barrier warrants? Alabama (2003) Yes California (2003) No Delaware (2006) Yes Florida (2006) No Idaho (2006) Yes Indiana (2006) Yes Iowa (2006) Yes Kentucky (2006) Yes Maine (2003) No Maryland (2006) No Minnesota (2006) Yes Mississippi (2003) Yes Missouri (2006) Yes Montana (2003) Yes Nebraska (2006) Yes Nevada (2003) Yes New Jersey (2006) No New York (2006) No North Carolina (2003) No Ohio (2003) Yes Oregon (2006) No Pennsylvania (2006) Yes South Carolina (2006) Yes South Dakota (2006) Yes Texas (2006) Yes Virginia (2006) Yes Washington (2006) No West Virginia (2006) Yes Wisconsin (2006) No Wyoming (2006) Yes The remaining survey questions, discussed in the following sections, were unique to the 2009 survey and, therefore, can- not be compared to previous responses. 3.3.4 Factors Contributing to Median Encroachments and Cross-Median Crashes In the 2009 survey only, respondents were presented a list of factors related to the design of the traveled way and the roadside that may contribute to median encroachments or cross-median crashes. They were asked for their profes- sional opinions regarding which of the factors they believe contribute most to these types of crashes in their states. Responses were marked by indicating how much each factor contributed to these crash types and using a five-point scale. Although the positions on the scale from least contribution to greatest contribution were not given numerical values on the survey page, they were later assigned ratings 1 through 5 so that an “average rating” could be developed and the fac- tors could be compared. Those factors that were ranked by more agencies as being greater contributors received a higher rating average than did those that were ranked as lower con- tributors by a majority of respondents. The list of factors, the frequency of responses for each factor, and the average rating for each factor are presented in Table 3-8. Respondents also were given an opportunity to rank a fac- tor labeled other and invited to share what they considered to be contributing factors to median encroachments and cross-median crashes in their states. Other contributing fac- tors named by respondents are listed below; the number in parenthesis indicates the number of agencies identifying that factor: • Median width (7), • Excessive speed (5), • Inattentive/distracted driving (8), • Driver error (3), • Weather/road conditions (5), • Glare/lighting—dawn, dusk, headlights (1), • Fatigue (3), • Impaired drivers (3), • ADT (2), and • Presence of median barrier (1). After respondents identified the factors that contribute to median encroachments and cross-median crashes, they were asked if their agencies had taken any measures to address those factors and to share what those measures were. Twenty- four agencies responded to this question and those responses are presented in Table 3-9. Table 3-6. Response from specific agencies on whether they use the 2002 AASHTO median barrier warrants from 2003 and 2006 surveys. received from each agency that provided an answer to the question in any of the surveys. The AASHTO median barrier warrants were revised in 2006; however, only one agency that responded in 2006 also responded in 2009, so it is difficult to know if the changes in the median barrier warrants led to changes in the various agencies’ warrants. Table 3-7 summarizes the median barrier warrant criteria of the 10 states that indicated that they do not use the 2002 AASHTO criteria in the 2003 and 2006 surveys, as well as the 12 agencies that responded to the question in the 2009 survey. Where more than one criterion is used by an agency, mul- tiple columns appear in Table 3-7. Factors other than median width that were considered in these criteria included ADT, posted speed limit, cross-median crash rates, location within 1 mile of entrance/exit ramp gore areas, and roadway type (freeway vs. nonfreeway). For criteria based on median width alone, minimum median widths where barriers are not required ranged from 18 to 64 feet. One state (Maryland) specified that they do not install barrier if the median is more than 75 feet wide.

30 Agency Response Median barrier warrant criteria Arizona (2009) Our guidelines are located at: http://www.azdot.gov/hig hways/Roadway_Engine ering/Roadway_Design/ Guidelines/Manuals/PD F/RoadwayDesignGuide lines.pdf Section 304.4 covers median barrier warrants. For items not covered here, we use AASHTO RDG. California (2003) Conduct study if median width is 0 to 20 ft and ADT exceeds 20,000 vehicles per day. Conduct study if median width is less than 75 ft and ADT exceeds 60,000 vehicles per day. Study any median with 0.5 cross- median crashes per mile per year or 0.12 fatal crashes per mile per year. Florida (2006) On Interstate, install barrier if median width less than 64 ft; 50 ft on other freeways. On Interstates and expressways, median barrier is required within 1 mile of exit/entrance gore with one or more cross-median crashes within 5 years. Georgia (2009) For Interstate highways or other grade- separated facilities: less than or equal to 52 ft requires barrier; barrier is optional for median widths greater than 52 ft and less than 64 ft; barrier not required for median widths greater than 64 ft. Cable barrier is sometimes installed if crash experience indicates the need for median protection. Depends on AADT and speed for other types of routes. Refer to Design Policy Manual Section 6.8.2. Maine (2003) Install barrier if the median width is < 20 ft and ADT > 20,000. Install barrier if median width is < 30 ft and ADT > 30,000 vehicles per day. Barrier optional if width is < 20 ft and ADT is 5,000 to 20,000 vehicles per day. Barrier optional if median width is 30 ft to 50 ft and ADT > 40,000 vehicles per day. Maryland (2006) Install median barrier if width <= 30 ft. Install median barrier if width > 30 ft but < 50 ft and ADT > 40,000 vehicles per day. Install median barrier if width > 50 ft but < 75 ft and ADT > 80,000 vehicles per day. Do not install barrier if median width > 75 ft. Massachusetts (2009) Use AASHTO RDG median warrant criteria. Michigan (2009) Michigan DOT uses the median Barrier Warrant specified in the AASHTO Roadside Design Guide. In addition, Michigan DOT utilizes cost- benefit analyses to justify median barrier installations. New Mexico (2009) We do not have any in place but are in the process of developing a policy for median barrier placement. New York (2006) Install barrier if median width < 36 ft and ADT > 20,000 vehicles per day. Barrier encouraged if median width < 72 ft. Barrier is optional if median width is < 45 ft and ADT > 10,000 vehicles per day. Table 3-7. Median barrier warrant criteria.

31 Respondents were then asked what factors related to driver behavior and weather conditions played a role in median encroachments and cross-median crashes. Response frequency and rating average for each factor are shown in Table 3-10. This question was presented and analyzed in the same manner as the previous question presented in Table 3-8. Respondents also were given an opportunity to rank a factor labeled other and invited to share what they consider to be contributing factors to median encroachments and cross-median crashes in their states. Other contributing fac- tors named by respondents are listed below; the number in parenthesis indicates the number of agencies identifying that factor: • Driving too fast for conditions/speeding (4), • Alcohol/substance abuse (3), • Horizontal curvature (1), • Poor or inadequate delineation (signs and striping) (1), • Roadside obstacles (1), and • Unexpected roadway geometry (1). Agency Response Median barrier warrant criteria North Carolina (2003) Install barrier if median width < 70 ft. Oregon (2009) Install barrier if median width less than or equal to 60 ft. Over 60 ft base warrant on cross-median collision statistics. Vermont (2009) AASHTO Roadside Design Guide Virginia (2006) 18 ft Washington (2006) Provide median barrier on multilane highways with full access control with median widths of 50 ft or less and posted speeds of 45 mph or more. Consider median barrier on highways with wider medians or lower posted speeds when there is a history of cross- median accidents. Median barrier is not normally placed on collectors or other state highways that do not have limited- access control. Wisconsin (2006) On new freeway construction: range (median width, ADT) from (< = 20 ft, > = 20,000 vehicles per day) to (< 60-ft, > = 50,000 vehicles per day) No retro-fit warrant Illinois State Toll Authority (2009) Use AASHTO barrier warrants. New York State Thruway Authority (2009) In a programmed manner, the Thruway installs median barrier in medians up to 72 ft in width, pursuant to most current New York State DOT criteria. North Texas Tollway Authority (2009) NTTA specifies concrete barrier on all high-speed divided facilities. Barrier is placed on one direction of roadway only. Oklahoma Turnpike Authority (2009) AASHTO and accident data British Columbia (2009) http://www.th.gov.bc.ca/ publications/eng_publica tions/geomet/TAC/TAC_ 2007_Supplement/Ch60 0-2007.pdf Please refer to Section 630. Table 3-7. (Continued).

32 3.3.5 Availability of Safety Evaluations and Maintenance Records Respondents were asked whether their agencies had con- ducted any published or unpublished safety evaluations of the factors listed above, or whether their agencies had any data to support the professional opinions given above regarding factors that contribute to median encroachment and cross- median crashes. Of the 27 agencies that responded to the question, 10 answered yes. Of those 10 respondents, 4 state highway agencies, 1 toll agency, and 2 Canadian provincial agencies provided further explanation. Their responses are shown in Table 3-11. Respondents were asked whether their agencies had mainte- nance records in a form that could be used to identify locations in highway medians with high frequencies of encroachments. Of the 26 agencies that responded, 9 indicated that those types of records were available, and 13 agencies provided further explanation. The responses to this question can be found in Table 3-12. Many highway agencies use rumble strips on the median shoulder to reduce the frequency of median encroachments and use median barriers, where warranted, to reduce the severity of crashes resulting from such encroachments. Sur- vey respondents were asked whether their agencies used countermeasures other than rumble strips and median barri- ers intended specifically to reduce the frequency and severity of crashes resulting from median encroachments. Out of the 25 responding agencies, six replied that they did use coun- termeasures other than rumble strips and median barriers to reduce the frequency and severity of median encroachments. The responses are shown in Table 3-13. After responding to the question asking about the use of countermeasures other than rumble strips and median bar- riers to reduce the rate or severity of median encroachments and cross-median crashes, agencies were asked if they had conducted any formal evaluations of the countermeasures they identified. The responses of the six agencies that identified alternative countermeasures above are shown in Table 3-14. The agencies were asked to provide information about, or a copy of, their evaluations, and the comments received in response to this request also are shown in the table. Only one agency indicated that formal evaluations had been conducted of these countermeasures. Finally, survey respondents were asked to provide any additional information that they believe might be relevant Answer options Number of agencies responding Rating average Response count Greatest contribution (5) (4) (3) (2) Least contribution (1) Absence of rumble strips 3 10 5 0 5 3.26 23 Horizontal curves 3 9 6 3 3 3.25 24 Peak-period volumes 4 5 7 4 3 3.13 23 Other unexpected roadway geometry 2 6 8 3 3 3.05 22 Poor roadway delineation 1 8 5 3 4 2.95 21 Left shoulder width 1 6 2 7 5 2.57 21 Posted speed 1 5 5 5 7 2.48 23 Mixture of vehicle types 1 3 4 12 3 2.43 23 Presence of interchange ramps 1 3 5 7 6 2.36 22 Peak-period duration 0 2 9 3 7 2.29 21 Lane drops 1 2 4 6 8 2.14 21 Access control 0 1 6 6 8 2.00 21 Interchange spacing 0 1 5 7 8 1.95 21 Presence of speed-transition zones 0 1 3 8 9 1.81 21 Land use 0 3 2 3 14 1.73 22 Presence of rumble strips 0 0 1 2 19 1.18 22 Other (please specify) 12 3 0 0 0 4.80 15 answered question 26 skipped question 5 Table 3-8. Traveled way and roadside factors contributing to median encroachments and cross-median crashes.

33 Agency Description of Measures Arizona Relocated median cable barrier from center of median to edge of shoulder. Florida 1. Require shoulder rumble strips on center of all freeways and more recently require audible-vibratory edge lines on non-limited-access rural roads. 2. Require median barrier on limited-access roadways with less than 64-ft median width. 3. Require median barrier within 1 mile of interchanges where there is any history of cross- median crashes. Georgia Inclusion of rumble strips on all shoulders 4 ft or wider. Addition of cable barrier in areas with higher crash experience. Indiana We are deploying approximately 360 mi of median cable barrier to segments of Interstate highway that have a significant history of cross-median crashes. We are also reviewing our rumble strip policy. Iowa Added shoulder rumble strips and have begun to add median cable Kentucky Speed blitzes, installed cable median barrier Maryland Since 2005, we have established a special funding category to address traffic barriers throughout the state. The funding is to install new and/or upgrade existing barrier to meet today's standards and criteria. We have been focusing our efforts to install median barrier on divided highways with narrow medians, high speeds, and high volumes of traffic. Massachusetts During resurfacing projects, we try to reclaim shoulder width. We try to delineate the horizontal curves. Michigan Michigan DOT has taken steps to reduce median crossover crashes by installing centerline rumble strips, enhancing roadway delineation, and installing median barrier where deemed appropriate. Mississippi We have started placing cable barrier in locations as needed. Montana We install rumble strips on both the outside and median shoulder on all Interstate projects. We also are developing criteria to determine where the installation of median rail would be beneficial. New Mexico In determining locations for test sections for median barrier we identified locations with a history of median crossover crashes. Many of these locations occur on either horizontal curves or on stretches of Interstate highways where we have had a history of lane departures, many attributed to driver fatigue. The state's Highway Safety Improvement Program has placed a priority on projects addressing lane departures. Ohio We are installing cable guardrail in our high crossover sections. Oregon Mandatory closure for any median 18 m (60 ft) in width or less (fogline to fogline) policy statement (see attached Q No. 6). South Carolina SCDOT has developed a project to install rumble strips along all suitable four-lane divided highways. We also are conducting corridor studies on four-lane divided roads to systematically install offset left turn lanes Texas Developed a rumble strip policy. Revised the median barrier policy. Funding towards installation of median barriers/cable barrier. Vermont Increased delineation, shoulder rumble strips from engineering perspective. General increase in safety awareness through Strategic Highway Safety Plan, none directly related to median crashes. Illinois State Toll Highway Authority The Tollway has always had full roadway delineation using crystal amber reflectors on the left and white on the right. Skip-dash striping is 6" in width with a 25 ft stripe/25 ft space pattern. On all pavement, raised pavement markers are installed in conjunction with skip dash striping. The Tollway employs continuous shoulder edge rumble strips on both right and left shoulders, both directions on all mainline and directional, high- speed ramps. In urban/suburban segments of our system with ADT greater than 70,000 vehicles per day, the Tollway has installed continuous concrete barrier median protection. In rural segments, cable median barrier has been installed over the past 3 to 4 years. The Tollway currently has 100 percent median protection. Kansas Turnpike Authority Use of rumble strips and roadway delineation New York State Thruway Authority Rumble strips North Texas Tollway Authority NTTA added concrete barrier to the entire length of the President George Bush Turnpike in 2004. NTTA implements a snow and ice mitigation plan to address roadway conditions. Oklahoma Turnpike Authority Using AASHTO British Columbia Installed median barriers. Installed rumble strips on median shoulders. Improved delineation. Saskatchewan The Ministry has established geometric standards for medians. The Ministry has also established access control standards and policies. All are in an effort to provide a safer road corridor. Table 3-9. Measures taken to address factors that contribute to median encroachments and cross-median crashes by highway agency.

34 to their agency’s efforts to reduce the frequency and severity of median encroachments and cross-median crashes. Com- ments were received from eight agencies as follows: • Arizona: Implemented speed-enforcement cameras on many of our urban freeways. • Georgia: Addressing lane departure crashes using a sys- temwide approach continues to be an integral part of our Highway Safety Improvement Plan. High crash corridors are identified and reviewed annually to determine appro- priate countermeasures. • Indiana: Continuing crash reviews will be conducted on divided highways to determine the need for median barrier and rumble strip deployments. • Massachusetts: We have identified the top median cross- over crash locations and then conducted RSAs at these loca- tions. We have developed site-specific counter measures for these locations including prioritizing (in process) median barrier installation. We have not taken a systemic approach to evaluating median barrier locations. • Oregon: In today’s uncertain fiscal climate it is becom- ing very difficult to push a policy upgrade that could have fiscal impact without proving where funding comes from. Even then, there is turf protection everywhere and I fear that safety upgrades will take the back seat to preservation. • South Dakota: SD Department of Public Safety has been providing a good campaign regarding that drivers need to “Buckle Up.” Majority of the fatalities in median encroach- ments involve rollovers and occupants are ejected due to not using seat belts. Our medians are relatively wide with 1V:6H median inslopes with relatively flat median ditch bottoms. Answer options Number of agencies responding Rating average Response count Most important (5) (4) (3) (2) Least important (1) Drive fatigue/drowsiness 8 9 5 3 1 3.77 26 Wet pavement, snow- and ice-covered pavement, or other adverse weather condition 6 6 8 4 2 3.38 26 Previous collision or driver maneuver to avoid a collision 1 3 4 12 4 2.38 24 Other factors [to be specified] 3 3 1 1 2 3.40 10 answered question 26 skipped question 5 Table 3-10. Driver behavior and weather factors contributing to median encroachments and cross-median crashes. Agency Information provided about the availability of safety evaluations Georgia Field investigations of areas where wet weather crashes appear to be overrepresented. Format of reports is not conducive to inclusion in this text box, but a contact for obtaining these reports is provided. Iowa Increase in cross-median crashes in adverse winter weather can be documented. New Mexico Our data is in the form of safety and/or engineering analysis related to applications submitted by our districts for HSIP funding. A copy can be provided; we will need to scan into an email-able file. Ohio OH-1 reports (crash reports) from the Ohio Highway Patrol. Illinois State Toll Highway Authority Since the Illinois Tollway does not enjoy tort immunity or any other form of protective liability legislation, all such studies and reports are considered draft in nature and hence are privileged and not publicly available. British Columbia Attached are the contributing factors for the three regions that cover the province of British Columbia. Saskatchewan The Ministry investigates all fatal collisions to determine if the roadway and/or the roadside contributed to the collision. These investigations are not published. Table 3-11. Agency responses regarding the availability of safety evaluations related to factors contributing to median encroachment and cross-median crashes.

35 Agency Are maintenance records available? Information provided about the availability of maintenance records Arizona Yes There is a spreadsheet that could be used to identify some locations with high encroachments. Data includes route, MP, direction, date. The missing data is the begin date of install and the duration that a location has been in place. Florida No Georgia Yes Highway Maintenance Management System (HMMS) contains location information. Encroachments can be located if they resulted in damages that require repair to signs, guardrail, cable barrier, etc. Indiana No Iowa Yes Already done for 2001 to 2008 crashes. Kentucky Yes Crash through the KY State Police Maryland No Even though statistics are “reported,” the data used to develop statistics is questionable. To my knowledge, the only way to determine if a crash is truly a cross-median fatality or non- fatality is to read through the accident reports that are completed by the State Police and to read the description of what happened. This is a very tedious process and to my knowledge, is not being done. Massachusetts No Michigan No Mississippi No Montana No New Mexico No Ohio Yes All state highways are inspected yearly with GPS to record maintenance problems. Oregon No "No" because some police report location to the nearest mile only. South Carolina Yes SCDOT has installed cable guardrail along the Interstate system where criteria such as median width is met. Our maintenance office keeps records of every median hit along these corridors. South Dakota No Texas No Vermont No Only crash reports Illinois State Toll Highway Authority Yes Since we have full median protection, computerized maintenance records can be used to determine number of hits on cable barrier system and costs to restore. Crash data can be cross-referenced with this repair to further analyze the nature and cause of the encroachment. On concrete median barrier protected areas of the system, crash data can be obtained indicating vehicle vs. barrier median wall. Kansas Turnpike Authority No New York State Thruway Authority No North Texas Tollway Authority No Oklahoma Turnpike Authority Yes Spreadsheet and histories British Columbia No New Brunswick Yes Accident reports Saskatchewan No Our provincial insurance agency, Saskatchewan Government Insurance, maintains a Traffic Accident Information System that contains all reported roadway collisions. Table 3-12. Availability of maintenance records that would help identify locations of median encroachments and cross-median crashes.

36 • Illinois State Toll Highway Authority: The Tollway has installed a network of 39 over-the-road dynamic message signs prior to all system (Interstate to Interstate) inter- changes. These signs are used 24/7 to provide motorists information regarding road conditions, travel times, and incidents—with a small percentage of information for safety messages. Regular relevant messaging to the motor- ists can enhance their attention to driving and potentially changing conditions that we feel can enhance driving safety and reduce encroachments or run-off-road accidents. • Kansas Turnpike Authority: We have installed concrete median barrier the entire length of our roadway. Agency Are countermeasure records available? Information provided about the availability of countermeasure records Arizona Yes We have enhanced the delineation of the posts to provide increased awareness for nighttime driving. Florida Yes Audible-vibratory pavement markings Georgia Yes Replacement of select median drainage structures, vegetation removal. In some cases, median barriers were removed from wide medians where absence of barrier would result in less serious crashes. Indiana No Iowa Yes For the most part, Iowa four-lane divided rural roadways were built with wider than “normal” (at the time) medians to mitigate cross-median multi-vehicle crashes. Kentucky No Maryland No Massachusetts No Michigan No Mississippi Yes Cable barrier Montana No New Mexico No Ohio No Oregon Yes Raised median berm, but it is too low and the slopes are too flat to be effective. It is slated to be replaced with barrier. South Carolina No South Dakota No Texas No Vermont No Kansas Turnpike Authority No New York State Thruway Authority No North Texas Tollway Authority No Oklahoma Turnpike Authority No British Columbia No New Brunswick No Saskatchewan No Table 3-13. Countermeasures used to reduce the frequency and severity of median encroachments. Agency Were evaluations conducted? Information provided about evaluations Arizona No Florida No response Georgia Yes Benefits of vegetation removal projects were monitored. Iowa No Mississippi No Oregon No FHWA local office has expressed desire to have the berm removed. Table 3-14. Formal evaluations of countermeasures identified in previous question.