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.
68 S E C T I O N 6 This section of the report applies the existing traffic opera- tional and safety relationships from Section 2 and the new relationships from the research presented in Section 4 to pri- oritize the 13 controlling criteria for specific roadway types. The discussion begins by presenting the ranking of the 13 controlling criteria from a recent survey of highway agencies, followed by presentation of the priorities developed in this research. 6.1 Ranking of the 13 Controlling Criteria from NCHRP Synthesis 417 NCHRP Synthesis 417: Geometric Design Practices for Resur- facing, Restoration, and Rehabilitation (50) included the results of a survey in which state highway agencies were asked to rank the importance of the 13 controlling criteria for RRR projects from 1 (most important) to 13 (least important). The results of the survey are summarized in Table 68. While the survey did not specifically mention ranking the 13 controlling cri- teria based on their perceived relationship to safety, it is dif- ficult to imagine that perceived safety effects (and possibly perceived traffic operational effects) were not a key element in consideration of this question by the respondents. The survey did not specify a particular roadway type to be con- sidered by respondents in making these rankings, but given that the survey concerned RRR projects, it is likely that rural two-lane highways, and perhaps rural multilane highways or urban and suburban arterials, were key considerations for the respondents. It is also likely that freeways were not a key consideration. The average ranks for the design criteria in Table 68 show some clear distinctions in importance among the 13 control- ling criteria, with lane width and shoulder width being rated as having the highest importance and grade being rated as having the least importance. However, it is also evident that there is a broad range of opinion on the relative importance Prioritization of the 13 Controlling Criteria of the 13 controlling criteria. Every design criteria was ranked first or second in importance by at least one state and last or next to last by at least one state. Some difference in assess- ments between states is to be expected because different states face difference circumstances with respect to terrain, climate, and level of development. For example, it was noted that some states with high rainfall levels rated cross slope as par- ticularly important. Despite such differing circumstances, the survey results in Table 68 reflect a greater lack of consensus among highway agencies than seems desirable given the cur- rent state of knowledge about traffic operations and safety. The sensitivity analysis and prioritization results presented later in this chapter are intended to reduce the uncertainty in prioritizing the 13 controlling criteria by setting priorities based on known traffic operational and safety relationships. The inclusion of design speed, structural capacity, and verti- cal clearance in Table 68 may contribute to some of the uncer- tainty about the rankings. For example, design speed by itself does not directly affect traffic operations or safety, but has an indirect traffic operational or safety effect because choosing a design speed influences the design criteria for lane width, horizontal alignment, vertical alignment, and stopping sight distance. Structural capacity has no direct effect on traffic oper- ations or safety, but, obviously, a structural failure has traffic operational and safety effects. It is easy to see why some respon- dents to the survey might have rated structural capacity as hav- ing critical importanceâstructural failures can be catastrophic. Nonetheless, it is just as easy to see why other respondents may have rated it as having little importanceâstructural capacity has no direct influence on traffic operations or safety. There are similar concerns with vertical clearance. To eliminate these con- cerns, Table 69 presents the results of the NCHRP Synthesis 417 survey for the 10 remaining controlling criteria when design speed, structural capacity, and vertical clearance are omitted. Table 69 indicates that the 10 controlling criteria of great- est interest are ranked in the same order as the complete sur- vey results shown in Table 68, with lane width and shoulder
69 width ranked as having the greatest importance and horizon- tal clearance and grade ranked as having the least importance. 6.2 Sensitivity Analysis of Key Controlling Criteria for Specific Roadway Types The research included a sensitivity analysis for both traf- fic operations and safety of the 10 key controlling criteria (excluding design speed, structural capacity, and vertical clearance for the reasons cited earlier). The sensitivity analyses varied each of the controlling criteria over a specified range of interest to determine its traffic operational and safety effects. Whenever possible, the traffic operational and safety effects were determined with established relationships from the HCM (13), the HSM (12), and other previously published sources. Where no previously published sources are available, new relationships developed in Section 4 were used or engi- neering judgments were made by the research team based on the current state of knowledge. The roadway types considered in the sensitivity analysis were rural two-lane highways, rural multilane highways, urban and suburban arterials, and rural freeways. A key issue in planning this sensitivity analysis was decid- ing what variations in each of the controlling criteria to assess. Since each of the controlling criteria have different magnitudes and different units of measure, and some measures are unit- less (e.g., cross slope and percent grade), a sensitivity analysis could not be performed simply by varying each criterion by one unit. Therefore, to conduct this assessment, the research team identified a typical range of design values in current use for each of the controlling criteria on each roadway type and then assessed the traffic operational and safety effect of reduc- ing each of the controlling criteria, in turn, from the upper end of that range to the midpoint of the range. The range of interest for each of the controlling criteria was defined as follows: ⢠The upper end of the range of interest was set equal to an ideal or, at least, typical high value for each of the control- ling criteria; for example, the upper end of the range of interest for lane width was set equal to 12 ft. Design criterion Average rank Highest rank Lowest rank Lane width 3.8 1 12 Shoulder width 4.6 1 13 Design speed 4.6 1 13 Stopping sight distance 5.8 1 13 Horizontal alignment 6.4 1 13 Structural capacity 7.0 1 13 Superelevation 7.1 1 13 Bridge width 7.4 2 12 Vertical alignment 7.7 2 13 Cross slope 8.1 1 13 Horizontal clearance/lateral offset 9.3 1 13 Vertical clearance 9.3 1 13 Grade 9.9 2 13 NOTE: Based on ranking of the importance of each design criterion by 46 highway agencies on a scale from 1 (most important) to 13 (least important) Table 68. Ranking of the 13 controlling criteria from NCHRP Synthesis 417 (50). Design criterion Average rank Highest rank Lowest rank Lane width 2.8 1 9 Shoulder width 3.7 1 10 Stopping sight distance 4.4 1 10 Horizontal alignment 4.9 1 10 Superelevation 5.7 1 10 Bridge width 5.8 1 10 Vertical alignment 6.2 1 10 Cross slope 6.3 1 10 Horizontal clearance/lateral offset 7.2 1 10 Grade 8.0 1 10 NOTE: Based on ranking of the importance of each design criterion by 46 highway agencies on a scale from 1 (most important) to 10 (least important); adapted from results presented in Table 68 by omitting rankings for design speed, structural capacity, and vertical clearance. Table 69. Ranking of the 13 controlling criteria omitting design speed, structural capacity, and vertical clearance (adapted from NCHRP Synthesis 417 [50]).
70 ⢠The lower end of the range of interest, representing restric- tive design, was set equal to the lowest value generally used in practice or a typical low value for each of the controlling criteria; for example, the lower end of the range of interest for lane width was set equal to 9 ft. Using the example of lane width, the sensitivity of traffic opera- tions to lane width and the sensitivity of safety to lane width were each determined by reducing lane width from the upper end value of 12 ft to the midpoint value of 10.5 ft. For each roadway type, a typical roadway section was spec- ified as the base condition for sensitivity analyses. For exam- ple, the base condition for rural two-lane highway analyses was a roadway section 5 mi in length with two intersections, five driveways, and one horizontal curve per mile. Three sce- narios for sensitivity analyses were developed considering a range of AADT levels and variations of one to three selected key variables. For example, the sensitivity analysis scenarios for rural two-lane highways included a roadway with AADT of 2,000 veh/day and gravel shoulders, a roadway with AADT of 5,000 veh/day and paved shoulders, and a roadway with AADT of 10,000 veh/day and paved shoulders. All other char- acteristics of the base condition remained unchanged in all analyses except for the values of the controlling criteria and the three scenarios described above. All 10 relevant controlling criteria (i.e., the 13 controlling criteria excluding design speed, structural capacity, and verti- cal clearance) were considered in each sensitivity analysis. For stopping sight distance, two scenarios were considered in each sensitivity analysis: scenarios with and without the presence of horizontal curves, intersections, or driveways hidden from the view of approaching drivers by limited sight distance. The sensitivity analysis for each roadway type is presented below. 6.2.1 Rural Two-Lane Highways Table 70 presents the plan for the sensitivity analysis of rural two-lane highways. The table shows, for rural two-lane highways, the range over which the controlling criteria were varied (ideal or typical value to midpoint of the range of interest), the site characteristics that varied among the three scenarios considered, and the values of the base condition characteristics that remained constant in all of the scenarios. Table 71 identifies the estimation methods that were used to quantify the traffic operational and safety relationships for the scenarios considered for rural two-lane highways. The results of the traffic operational and safety sensitivity analyses for rural two-lane highways are presented in Tables 72 and 73, respectively. Table 72 presents the traffic operational effect of each design criterion in terms of the reduction in average travel speed due to the change in the value of that criterion, as described above. Table 73 presents comparable values for the safety effect of each design criterion in terms of the change in fatal-and-injury crash frequency per mile per year. Tables 72 and 73 also present a rank order (from 1 to 11) for the magnitude of the traffic operational and safety effects of each design criterion, so the design criteria with the largest effects are ranked highest. Table 72 shows that three of the design criteria have quan- titative, nonzero effects on traffic speeds on rural two-lane highways. These three design criteria are, in descending order of the magnitude of the effect on traffic speed: shoulder width, lane width, and horizontal curve radius. There is no evidence that any of the remaining design criteria have effects on traffic speed sufficiently large to be quantified. It is pos- sible that some of these design criteria have effects on traffic speed that are too small to be quantified, and it is likely that some design criteria, in fact, have no effect on traffic speed. The design criteria shown in Table 72 as having no quanti- fied effect on traffic speed have been ranked in descending rank order based on literature review results and judgment. Grade, for example, clearly affects vehicle speeds to some extent (especially for heavy vehicles), but such effects are quantified in HCM procedures only if the terrain category changes or if an individual grade is over 0.5 mi in length. Bridge width has no quantified effect on average travel speed; the lane- and shoulder-width effects suggest that a bridge that is narrower than the approach roadway may slow traffic, but such effects on the average travel speed over an extended roadway section are likely to be small because of the limited length of a bridge in relation to the section length as a whole. Other design criteria likely have zero or essentially zero effect on traffic speed. In particular, lateral offset likely has no effect on traffic speed on rural two-lane highways because the pres- ence of shoulders in the sensitivity analysis scenarios ensures that appropriate lateral offset should always be present. Table 73 shows that six of the design criteria have quan- titative, nonzero effects on fatal-and-injury crash frequen- cies on rural two-lane highways. These six design criteria are, in descending order of the magnitude of the effect on crash frequency are shoulder width, lane width, grade, horizontal curve radius, superelevation, and stopping sight distance (in the situation where a hidden curve, intersection, or driveway is present). There is no evidence that any of the remaining design criteria have effects on crash frequency sufficiently large to be quantified. It is possible that some of these design criteria have effects on crash frequency that are too small to be quantified, and it is likely that some design criteria, in fact, have no effect on crash frequency. The design criteria shown in Table 73 as having no quan- tified effect on fatal-and-injury crash frequency have been ranked in descending rank order based on literature review results and judgment. For example, bridge width has no
Levels for Specific Controlling Criteria Midpoint between Typical or typical and Design criterion ideal design Restrictive design restrictive design Comment Lane width (ft) 12 9 10.5 Range from ideal to most restrictive lane width Shoulder width (ft) 6 0 3 Range from ideal to most restrictive shoulder width Bridge width difference (ft) 0 10 5 Range from ideal 36-ft bridge width to 26-ft bridge width Horizontal curve radius (ft) 3000 1000 2000 Range from well above minimum radius to slightly below minimum Sag vertical curve length (ft) above AASHTO criteria below AASHTO criteria below AASHTO criteria Range from above AASHTO minimum to below AASHTO minimum for A=6% Grade (%) 3 6 4.5 Range from steepest rural arterial grade in level terrain to steepest grade in mountainous terrain Stopping sight distance category meets AASHTO criteria below AASHTO criteria below AASHTO criteria Presence of a hidden curve, intersection, or driveway is addressed in a separate analysis Cross slope (%) 2 0 1 Range from normal cross slope to no cross slope Superelevation (%) 5 2 3.5 Range from normal superelevation for 3,000-ft radius curve with emax=8% to normal cross slope Lateral offset (ft) 1.5 0 0.75 Range from AASHTO minimum lateral offset to no lateral offset Input Parameters That Vary by Scenario Input parameter Scenario #1 Scenario #2 Scenario #3 AADT for major road (veh/day) 2000 5000 10000 Shoulder type gravel paved paved Input Parameters Held Constant for All Scenarios Value for all Input parameter scenarios Analysis section length (mi) 5 Design speed (mph) 60 Base free-flow speed (mph) 60 emax (%) 8 Roadside rating 3 Intersections per mi 2 AADT for minor road (veh/day) 1000 Intersection type 4-leg, minor road stop Major-road left-turn lanes per intersection 2 Major-road right-turn lanes per intersection 2 Presence of skew angle at intersections not present Presence of lighting at intersections not present Driveways per mi 5 No. of curves per mi 1 Length of curve (mi) 0.2 Distance between crests and sags (mi) 0.5 Presence of spiral transitions on curves none present Presence of centerline rumble strips none present Presence of passing lanes none present Presence of two-way left-turn lanes none present Presence of lighting between intersections none present Use of automated speed enforcement not used Design hour factor (K) 0.1 Directional factor (D) 0.5 Peak-hour factor (PHF) 0.95 Percent trucks in traffic flow 5 Percent RVs in traffic flow 2 Percent no-passing zones 80 Highway class Class I Calibration factor for roadway segments 1 Calibration factor for intersections 1 Table 70. Base case and analysis scenarios for rural two-lane highways.
72 Controlling criterion Traffic operational effect Traffic safety effect Lane width Table 5 and Equation 1 Table 6 and Equation 2 Shoulder width Table 5 and Equation 1 Tables 13 and 14 and Equation 7 Bridge width Table 5 and Equation 1a No known effect based on Section 4.3 Horizontal curve radius Table 21 Equation 15 Sag vertical curve length No known effect based on Section 2.7 Equations 30 through 33 Grade Tables 23 through 31 and Equation 34 Table 32 and Equation 35 Stopping sight distance No known effect based on Section 2.9 Effect based on Section 4.7 if a hidden feature is present Cross slope No known effect based on Section 2.10 No known effect based on Section 2.10 Superelevation No known effect based on Section 2.11 Equations 36 through 38 Lateral offset Not applicable where shoulders are present Not applicable where shoulders are present a No additional effect beyond the effect of a narrower lane or shoulder, if present on bridge. Table 71. Estimation methods for traffic operational and safety effects for rural two-lane highways. Design criterion Traffic operational effect: Change in average travel speed (mph)a in comparison to base condition Rank order of traffic operational effectb Scenarioc No. 1 Scenario No. 2 Scenario No. 3 Scenario No. 1 Scenario No. 2 Scenario No. 3 Combined Lane width â0.75 â0.75 â0.75 2 2 2 2 Shoulder width â1.14 â1.22 â1.22 1 1 1 1 Bridge width 0.00 0.00 0.00 5 5 5 5 Horizontal curve radius â0.24 â0.24 â0.24 3 3 3 3 Sag vertical curve length 0.00 0.00 0.00 9 9 9 9 Grade 0.00 0.00 0.00 4 4 4 4 Stopping sight distanced 0.00 0.00 0.00 8 8 8 8 Stopping sight distancee 0.00 0.00 0.00 7 7 7 7 Cross slope 0.00 0.00 0.00 10 10 10 10 Superelevation 0.00 0.00 0.00 6 6 6 6 Lateral offset 0.00 0.00 0.00 11 11 11 11 a Traffic operational effects are in comparison to average travel speed of 52.7 mph for the base condition in Scenario #1, 49.8 mph for base condition in Scenario #2, and 47.4 mph for base condition in Scenario #3. b Estimated based on literature review results and judgment for situations with zero effects. c Scenarios No. 1 through 3 are defined in Table 70; methods for estimating traffic operational effects are defined in Table 71. d With no hidden features. e With hidden curve, intersection, or driveway. Table 72. Traffic operational effects of the controlling criteria for selected scenarios on rural two-lane highways.
73 Design criterion Traffic safety effect: Percent change in fatal-and-injury crashes/mi/yeara in comparison to base condition Rank order of traffic safety effectb Scenarioc #1 Scenario #2 Scenario #3 Scenario #1 Scenario #2 Scenario #3 Combined Lane width 4.27 5.15 5.84 2 2 2 2 Shoulder width 5.24 6.62 7.51 1 1 1 1 Bridge width 0.00 0.00 0.00 7 7 7 7 Horizontal curve radius 0.88 1.06 1.20 4 4 4 4 Sag vertical curve length 0.00 0.00 0.00 9 9 9 9 Grade 0.97 1.17 1.33 3 3 3 3 Stopping sight distanced 0.00 0.00 0.00 10 10 10 10 Stopping sight distancee 0.03 0.03 0.02 6 6 6 6 Cross slope 0.00 0.00 0.00 8 8 8 8 Superelevation 0.66 0.80 0.91 5 5 5 5 Lateral offset 0.00 0.00 0.00 11 11 11 11 a Traffic safety effects are in comparison to crash frequency of 0.71 fatal-and-injury crashes/mi/year for the base condition in Scenario #1, 1.46 fatal-and-injury crashes/mi/year for base condition in Scenario #2, and 2.58 fatal-and-injury crashes/mi/year for base condition in Scenario #3. b Estimated based on literature review results and judgment for situations with zero effects. c Scenarios #1 through #3 are defined in Table 70; methods for estimating traffic safety effects are defined in Table 71. d With no hidden features. e With hidden curve, intersection, or driveway. Table 73. Traffic safety effects of the controlling criteria in comparison to base condition for selected scenarios on rural two-lane highways. quantified effect on crash frequency; the lane- and shoulder- width effects suggest that a bridge that is narrower than the approach roadway might increase crash risk, but such effects on crash frequency are likely to be very small over an extended roadway section because of the limited length of a bridge in relation to the section length as a whole. Other design criteria likely have zero or essentially zero effect on crash frequency. In particular, lateral offset likely has no effect on crash fre- quency on rural two-lane highways because the presence of shoulders in the sensitivity analysis scenarios ensures that appropriate lateral offset should always be present. 6.2.2 Rural Multilane Highways Table 74 presents the plan for the sensitivity analysis of rural multilane highways. The table shows, for rural multi- lane highways, the range over which the controlling criteria were varied (ideal or typical value to midpoint of the range of interest), the site characteristics that varied among the three scenarios considered, and the values of the base condition characteristics that remained constant in all of the scenarios. Table 75 identifies the estimation methods that were used to quantify the traffic operational and safety relationships for the scenarios considered for rural multilane highways. The results of the traffic operational and safety sensitiv- ity analyses for rural multilane highways are presented in Tables 76 and 77, respectively. Tables 76 and 77 also present a rank order (from 1 to 11) for the magnitude of the traffic operational and safety effects of each design criterion, so the design criteria with the largest effects are ranked highest. Table 76 shows that only two of the design criteria have quantitative, nonzero effects on traffic speeds on rural multilane highways. These two design criteria are, in descending order of the magnitude of the effect on traffic speed: lane width and horizontal curve radius. There is no evidence that any of the remaining design criteria have effects on traffic speed suffi- ciently large to be quantified. It is possible that some of these design criteria have effects on traffic speed that are too small to be quantified. For example, shoulder width likely has at least a small effect on traffic speed. In addition, grade clearly has some effect on traffic speed, although that effect is minimal unless the terrain category changes or an individual grade is more than 0.5 mi in length. It is also likely that some design criteria, in fact, have no effect on traffic speed. Table 77 shows that three of the design criteria have quan- titative, nonzero effects on fatal-and-injury crash frequencies on rural multilane highways. These three design criteria are, in descending order of the magnitude of the effect on crash frequency: shoulder width, lane width, and stopping sight distance (in the situation where a hidden curve, intersection, or driveway is present). The stopping sight distance effect is based on an analogy to the documented effect for rural two- lane highways. There is no evidence that any of the remain- ing design criteria have effects on crash frequency sufficiently
Levels for Specific Controlling Criteria Midpoint between Typical or typical and Design criterion ideal design Restrictive design restrictive design Comment Lane width (ft) 12 9 10.5 Range from ideal to most restrictive lane width Outside shoulder width (ft) 6 0 3 Range from ideal to most restrictive shoulder width Bridge width difference (ft) 0 10 5 Range from ideal 36-ft bridge width to 26-ft bridge width Horizontal curve radius (ft) 3000 1000 2000 Range from well above minimum radius to slightly below minimum Sag vertical curve length (ft) above AASHTO criteria below AASHTO criteria below AASHTO criteria Range from above AASHTO minimum to below AASHTO minimum for A=6% Grade (%) 3 6 4.5 Range from steepest rural arterial grade in level terrain to steepest grade in mountainous terrain Stopping sight distance meets AASHTO criteria below AASHTO criteria below AASHTO criteria Presence of a hidden curve, intersection, or driveway is addressed in a separate analysis Cross slope (%) 2 0 1 Range from normal cross slope to no cross slope Superelevation (%) 5 2 3.5 Range from normal superelevation for 3,000-ft radius curve with emax=8% to normal cross slope Lateral offset (ft) 1.5 0 0.75 Range from AASHTO minimum lateral offset to no lateral offset Input Parameters That Vary by Scenario Input parameter Scenario #1 Scenario #2 Scenario #3 AADT for major road (veh/day) 10000 20000 30000 Divided/undivided undivided divided divided Median width (ft) N/A 20 40 Presence of median barrier N/A Yes Yes Outside shoulder type gravel paved paved Input Parameters Held Constant for All Scenarios Value for all Input parameter scenarios Analysis section length (mi) 5 Design speed (mph) 60 Base free-flow speed (mph) 60 emax (%) 8 Roadside slopes 1V:4H Intersections per mi 2 AADT for minor road (veh/day) 1000 Intersection type 4-leg, minor road stop Major-road left-turn lanes per intersection 2 Major-road right-turn lanes per intersection 2 Presence of skew angle at intersections not present Presence of lighting at intersections not present No. of curves per mi 1 Length of curve (mi) 0.2 Distance between crests and sags (mi) 0.5 Presence of lighting between intersections not present Use of automated speed enforcement not used Design hour factor (K) 0.1 Directional factor (D) 0.5 Peak-hour factor (PHF) 0.95 Driver population factor (fp) 1 Percent trucks in traffic flow 5 Percent RVs in traffic flow 2 Table 74. Base case and analysis scenarios for rural multilane highways.
75 Controlling criterion Traffic operational effect Traffic safety effect Lane width Table 7 and Equation 3 Table 8 or 9 and Equation 2 Shoulder width Table 16 and Equation 3 Tables 13 and 14 and Equation 7 or Table 17 Bridge width Table 16 and Equation 3a No known effect based on Section 2.4 Horizontal curve radius Equation 18 Equations 19 and 20 Sag vertical curve length No known effect based on Section 2.7 No known effect based on Section 2.7 Grade Tables 33 through 36 and HCM Equation 14-4 No known effect based on Section 2.8 Stopping sight distance No known effect based on Section 2.9 Effect estimated as equivalent to the rural two-lane highway effect in Section 4.7 if a hidden feature is present Cross slope No known effect based on Section 2.10 No known effect based on Section 2.10 Superelevation No known effect based on Section 2.11 No known effect based on Section 2.11 Lateral offset Not applicable where shoulders are present Not applicable where shoulders are present a No additional effect beyond the effect of a narrower lane or shoulder, if present on bridge Table 75. Estimation methods for traffic operational and safety effects for rural multilane highways. Design criterion Traffic operational effect: Change in average travel speed (mph)a in comparison to base condition Rank order of traffic operational effectb Scenario #1 Scenario #2 Scenario #3 Scenario #1 Scenario #2 Scenario #3 Combined Lane width â2.50 â7.50 â7.50 1 1 1 1 Shoulder width 0.00 0.00 0.00 3 3 3 3 Bridge width 0.00 0.00 0.00 5 5 5 5 Horizontal curve radius â0.10 â0.10 â0.10 2 2 2 2 Sag vertical curve length 0.00 0.00 0.00 9 9 9 9 Grade 0.00 0.00 0.00 4 4 4 4 Stopping sight distancec 0.00 0.00 0.00 8 8 8 8 Stopping sight distanced 0.00 0.00 0.00 7 7 7 7 Cross slope 0.00 0.00 0.00 10 10 10 10 Superelevation 0.00 0.00 0.00 6 6 6 6 Lateral offset 0.00 0.00 0.00 11 11 11 11 a Traffic operational effects are in comparison to average travel speed of 54.8 mph for the base condition in Scenario #1, 59.8 mph for base condition in Scenario #2, and 59.8 mph for base condition in Scenario #3. b Estimated based on literature review results and judgment for situations with zero effects. c With no hidden features. d With hidden curve, intersection, or driveway. NOTE: Scenarios #1 through #3 are defined in Table 74; methods for estimating traffic operational effects are defined in Table 75. Table 76. Traffic operational effects of the controlling criteria for selected scenarios on rural multilane highways.
76 large to be quantified. It is possible that some of these design criteria have effects on crash frequency that are too small to be quantified, and it is likely that some design criteria, in fact, have no effect on crash frequency. A rank order for the remaining design criteria has been established based on lit- erature review results and judgment. 6.2.3 Rural Freeways The sensitivity analysis for freeways was based on rural free- ways, rather than urban freeways, because it was expected that traffic operational and safety effects would be more critical on higher speed roadways. The results may also be applicable to urban freeways, although urban freeways present their own unique issues with higher volumes and more frequent entrance and exit ramps than rural freeways. Table 78 presents the plan for the sensitivity analysis of rural freeways. The table shows, for rural freeways, the range over which the controlling criteria were varied (ideal or typical value to midpoint of the range of interest), the site characteristics that varied among the three scenarios consid- ered, and the values of the base condition characteristics that remained constant in all of the scenarios. Table 79 identifies the estimation methods that were used to quantify the traffic operational and safety relationships for the scenarios considered for rural multilane highways. The results of the traffic operational and safety sensitiv- ity analyses for rural multilane highways are presented in Tables 80 and 81, respectively. Tables 80 and 81 also present a rank order (from 1 to 11) for the magnitude of the traffic operational and safety effects of each design criterion, so the design criteria with the largest effects are ranked highest. The only variable shown in Table 80 with a quantitative, nonzero effect on average travel speeds on rural freeways is lane width. The likely range in variation of shoulder widths on rural freeways was too small to have any quantifiable effect. It is possible that some of these design criteria have effects on traffic speed that are too small to be quantified. For example, grade clearly has some effect on traffic speed, although that effect is minimal unless the terrain category changes or an individual grade is more than 0.5 mi in length. It is also likely that some design criteria, in fact, have no effect on traffic speed. Table 81 shows that four of the design criteria have quanti- tative, nonzero effects on fatal-and-injury crash frequencies on rural freeways. These four design criteria are, in descend- ing order of the magnitude of the effect on crash frequency: shoulder width, lane width, horizontal curve radius, and stopping sight distance (in the situation where a hidden curve or ramp is present). The stopping sight distance effect is based on an analogy to the documented effect for rural two- lane highways. There is no evidence that any of the remain- ing design criteria have effects on crash frequency sufficiently large to be quantified. It is possible that some of these design criteria have effects on crash frequency that are too small to be quantified, and it is likely that some design criteria, in Design criterion Traffic safety effect: Percent change in fatal-and-injury crashes/mi/yeara in comparison to base condition Rank order of traffic safety effectb Scenario #1 Scenario #2 Scenario #3 Scenario #1 Scenario #2 Scenario #3 Combined Lane width 2.82 2.80 2.81 2 2 2 2 Shoulder width 4.52 3.78 3.79 1 1 1 1 Bridge width 0.00 0.00 0.00 7 7 7 7 Horizontal curve radius 0.00 0.00 0.00 6 6 6 6 Sag vertical curve length 0.00 0.00 0.00 9 9 9 9 Grade 0.00 0.00 0.00 5 5 5 5 Stopping sight distancec 0.00 0.00 0.00 10 10 10 10 Stopping sight distanced 0.03 0.03 0.02 3 3 3 3 Cross slope 0.00 0.00 0.00 8 8 8 8 Superelevation 0.00 0.00 0.00 4 4 4 4 Lateral offset 0.00 0.00 0.00 11 11 11 11 a Traffic safety effects are in comparison to crash frequency of 2.83 fatal-and-injury crashes/mi/year for the base condition in Scenario #1, 3.13 fatal-and-injury crashes/mi/year for base condition in Scenario #2, and 4.49 fatal-and-injury crashes/mi/year for base condition in Scenario #3. b Estimated based on literature review results and judgment for situations with zero effects. c With no hidden features. d With hidden curve, intersection, or driveway. NOTE: Scenarios #1 through #3 are defined in Table 74; methods for estimating traffic safety effects are defined in Table 75. Table 77. Traffic safety effects of the controlling criteria in comparison to base condition for selected scenarios on rural multilane highways.
Levels for Specific Controlling Criteria Midpoint between Typical or typical and Design criterion ideal design Restrictive design restrictive design Comment Lane width (ft) 12 9 10.5 Range from ideal to most restrictive lane width Outside shoulder width (ft) 10 6 8 Range from ideal to most restrictive typical shoulder width Bridge width difference (ft) 0 10 5 Range from ideal 36- bridge width to 26- bridge width Horizontal curve radius (ft) 4500 2000 3750 Range from well above minimum radius to slightly below minimum Sag vertical curve length (ft) above AASHTO criteria below AASHTO criteria below AASHTO criteria Range from above AASHTO minimum to below AASHTO minimum for A=6% Grade (%) 3 6 4.5 Range from steepest rural arterial grade in level terrain to steepest grade in mountainous terrain Stopping sight distance meets AASHTO criteria below AASHTO criteria below AASHTO criteria Presence of a hidden curve or ramp is addressed in a separate analysis Cross slope (%) 2 0 1 Range from normal cross slope to no cross slope SuperelevaÂon (%) 5 2 3.5 Range from normal superelevaÂon for 3,000- radius curve with emax=8% to normal cross slope Lateral offset (ft) N/A N/A N/A Not applicable to freeways Input Parameters That Vary by Scenario Input parameter Scenario #1 Scenario #2 Scenario #3 AADT for major road (veh/day) 20000 40000 60000 ProporÂon of AADT under congested condiÂons 0 0 0.2 Median width (ft) 40 60 60 Input Parameters Held Constant for All Scenarios Value for all Input parameter scenarios Analysis section length (mi) 5 Design speed (mph) 75 Base free-flow speed (mph) 75 emax (%) 8 Number of lanes per direction of travel 2 Inside shoulder width (ft) 4 Le -side clearance (ft) 6 Interchanges per mi 0.4 Entrance ramps per mi 0.4 Exit ramps per mi 0.4 IntersecÂons per mi 2 Ramp AADT (veh/day) 1000 No. of curves per mi 1 Length of curve (mi) 0.2 Distance between crests and sags (mi) 0.5 Design hour factor (K) 0.1 Directional factor (D) 0.5 Peak-hour factor (PHF) 0.95 Percent trucks in traffic flow 10 Percent RVs in traffic flow 2 fp 1 Roadside slopes 1V:6H Clear zone width (ft) 30 ProporÂon of length with barrier beyond outside shoulder 0.05 Distance from edge of outside shoulder to barrier face (ft) 10 ProporÂon of length with barrier beyond inside shoulder 0.05 Distance from edge of inside shoulder to barrier face (ft) 10 ProporÂon of segment length with rumble strip on outside shoulder 1 ProporÂon of segment length with rumble strip on inside shoulder 1 Len, ramp entrance length (i.e., speed-change lane length) (mi) 0.5 Le -side ramp indicator for entranance ramps 0 Length of entrance-ramp speed-change lanes adjacent to through lanes ( 0.2 Number of lanes on entrance-ramp speed-change lane adjacent to freeway 1 ProporÂon of entrance-ramp speed-change lane length on curve 0 ProporÂon of entrance-ramp speed-change lane length with barrier present 0 Lex, ramp exit length (i.e., speed-change lane length) (mi) 0.5 Le -side ramp indicator for entrance ramps 0 Length of exit-ramp speed-change lanes adjacent to through lanes (mi) 0.2 Number of lanes on exit-ramp speed-change lane adjacent to freeways 1 ProporÂon of exit-ramp speed-change lane length on curve 0 ProporÂon of exit-ramp speed-change lane length with barrier present Presence of weaving section in analysis section not present Table 78. Base case and analysis scenarios for rural freeways.
78 Controlling criterion Traffic operational effect Traffic safety effect Lane width Table 10 and Equation 4 Equations 5 and 6 Shoulder width Table 19 and Equation 4 Equations 8 through 13 Bridge width Table 19 and Equation 4a No known effect based on Section 2.4 Horizontal curve radius No known effect based on Section 2.6 Equations 22 through 25 Sag vertical curve length No known effect based on Section 2.7 No known effect based on Section 2.7 Grade Tables 39 through 42 and HCM Equations 11-2 through 11-4 No known effect based on Section 2.8 Stopping sight distance No known effect based on Section 2.9 Effect estimated as equivalent to the rural two-lane highway effect in Section 4.7 if a hidden feature is present Cross slope No known effect based on Section 2.10 No known effect based on Section 2.10 Superelevation No known effect based on Section 2.11 No known effect based on Section 2.11 Lateral offset Not applicable where shoulders are present Not applicable where shoulders are present a No additional effect beyond the effect of a narrower lane or shoulder, if present on bridge Table 79. Estimation methods for traffic operational and safety effects for rural freeways. Design criterion Traffic operational effect: Change in average travel speed (mph)a in comparison to base condition Rank order of traffic operational effectb Scenario #1 Scenario #2 Scenario #3 Scenario #1 Scenario #2 Scenario #3 Combined Lane width â7.50 â6.91 â3.40 1 1 1 1 Shoulder width 0.00 0.00 0.00 2 2 2 2 Bridge width 0.00 0.00 0.00 5 5 5 5 Horizontal curve radius 0.00 0.00 0.00 3 3 3 3 Sag vertical curve length 0.00 0.00 0.00 9 9 9 9 Grade 0.00 0.00 0.00 4 4 4 4 Stopping sight distancec 0.00 0.00 0.00 8 8 8 8 Stopping sight distanced 0.00 0.00 0.00 7 7 7 7 Cross slope 0.00 0.00 0.00 10 10 10 10 Superelevation 0.00 0.00 0.00 6 6 6 6 Lateral offset 0.00 0.00 0.00 11 11 11 11 a Traffic operational effects are in comparison to average travel speed of 75.0 mph for the base condition in Scenario #1, 74.4 mph for base condition in Scenario #2, and 67.1 mph for base condition in Scenario #3. b Estimated based on literature review results and judgment for situations with zero effects. c With no hidden features. d With hidden curve or ramp junction. NOTE: Scenarios #1 through #3 are defined in Table 78; methods for estimating traffic operational effects are defined in Table 79. Table 80. Traffic operational effects of the controlling criteria for selected scenarios on rural freeways.
79 fact, have no effect on crash frequency. A rank order for the remaining design criteria has been established based on lit- erature review results and judgment. 6.2.4 Urban and Suburban Arterials No formal sensitivity analyses have been conducted for urban and suburban arterials because there are very few design criteria for which quantitative traffic operational or safety relationships are available. In fact, design consid- erations that are outside the scope of this research, such as intersection design and access management, appear to have a much greater effect on traffic operations and safety for urban and suburban arterials than roadway design criteria. 6.3 Priorities for the Controlling Criteria Based on the sensitivity analyses presented above, Table 82 presents the recommended priorities for the controlling criteria. Design criterion Traffic safety effect: Percent change in fatal-and-injury crashes/mi/yeara in comparison to base condition Rank order of traffic safety effectb Scenario #1 Scenario #2 Scenario #3 Scenario #1 Scenario #2 Scenario #3 Combined Lane width 5.74 5.73 5.73 2 2 2 2 Shoulder width 9.74 8.39 7.05 1 1 1 1 Bridge width 0.00 0.00 0.00 7 7 7 7 Horizontal curve radius 0.75 0.68 0.60 3 3 3 3 Sag vertical curve length 0.00 0.00 0.00 9 9 9 9 Grade 0.00 0.00 0.00 6 6 6 6 Stopping sight distancec 0.00 0.00 0.00 10 10 10 10 Stopping sight distanced 0.03 0.03 0.02 4 4 4 4 Cross slope 0.00 0.00 0.00 8 8 8 8 Superelevation 0.00 0.00 0.00 5 5 5 5 Lateral offset 0.00 0.00 0.00 11 11 11 11 a Traffic safety effects are in comparison to crash frequency of 0.9 fatal-and-injury crashes/mi/year for the base condition in Scenario #1, 1.7 fatal-and-injury crashes/mi/year for base condition in Scenario #2, and 2.6 fatal-and-injury crashes/mi/year for base condition in Scenario #3. b Estimated based on literature review results and judgment for situations with zero effects. c With no hidden features. d With hidden curve or ramp junction. NOTE: Scenarios #1 through #3 are defined in Table 78; methods for estimating traffic safety effects are defined in Table 79. Table 81. Traffic safety effects of the controlling criteria in comparison to base condition for selected scenarios on rural freeways.
80 Priority rank Roadway type Rural two-lane highways Rural multilane highways Rural freeways TRAFFIC OPERATIONS 1 (highest priority) Shoulder width Lane width Lane width 2 Lane width Shoulder width Shoulder width 3 Horizontal curve radius Horizontal curve radius Horizontal curve radius 4 Grade Grade Grade 5 Bridge width Bridge width Bridge width 6 Superelevation Superelevation Superelevation 7 Stopping sight distancea Stopping sight distancea Stopping sight distanceb 8 Stopping sight distancec Stopping sight distancec Stopping sight distancec 9 Sag vertical curve length Sag vertical curve length Sag vertical curve length 10 Cross slope Cross slope Cross slope 11 (lowest priority) Lateral offset Lateral offset Lateral offset TRAFFIC SAFETY 1 (highest priority) Shoulder width Shoulder width Shoulder width 2 Lane width Lane width Lane width 3 Grade Stopping sight distancea Horizontal curve radius 4 Horizontal curve radius Superelevation Stopping sight distanceb 5 Superelevation Grade Superelevation 6 Stopping sight distancea Horizontal curve radius Grade 7 Bridge width Bridge width Bridge width 8 Cross slope Cross slope Cross slope 9 Sag vertical curve length Sag vertical curve length Sag vertical curve length 10 Stopping sight distancec Stopping sight distancec Stopping sight distancec 11 (lowest priority) Lateral offset Lateral offset Lateral offset a With hidden curve, intersection, or driveway. b With hidden curve or ramp junction. c With no hidden features. Table 82. Priorities for the 13 controlling criteria based on sensitivity analysis.
