National Academies Press: OpenBook

Evaluation of the 13 Controlling Criteria for Geometric Design (2014)

Chapter: Section 7 - Interpretation of Results

« Previous: Section 6 - Prioritization of the 13 Controlling Criteria
Page 81
Suggested Citation:"Section 7 - Interpretation of Results." National Academies of Sciences, Engineering, and Medicine. 2014. Evaluation of the 13 Controlling Criteria for Geometric Design. Washington, DC: The National Academies Press. doi: 10.17226/22291.
×
Page 81
Page 82
Suggested Citation:"Section 7 - Interpretation of Results." National Academies of Sciences, Engineering, and Medicine. 2014. Evaluation of the 13 Controlling Criteria for Geometric Design. Washington, DC: The National Academies Press. doi: 10.17226/22291.
×
Page 82
Page 83
Suggested Citation:"Section 7 - Interpretation of Results." National Academies of Sciences, Engineering, and Medicine. 2014. Evaluation of the 13 Controlling Criteria for Geometric Design. Washington, DC: The National Academies Press. doi: 10.17226/22291.
×
Page 83
Page 84
Suggested Citation:"Section 7 - Interpretation of Results." National Academies of Sciences, Engineering, and Medicine. 2014. Evaluation of the 13 Controlling Criteria for Geometric Design. Washington, DC: The National Academies Press. doi: 10.17226/22291.
×
Page 84
Page 85
Suggested Citation:"Section 7 - Interpretation of Results." National Academies of Sciences, Engineering, and Medicine. 2014. Evaluation of the 13 Controlling Criteria for Geometric Design. Washington, DC: The National Academies Press. doi: 10.17226/22291.
×
Page 85

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.

81 S E C T I O N 7 This section addresses the interpretation of the research results presented in Sections 2 through 6 and the formulation of the recommendations that will be presented in Section 8 of the report. The interpretation focuses mainly on applica- tion of controlling criteria in reconstruction projects, which includes most projects that are designed and constructed by highway agencies today. New construction projects have fewer constraints than reconstruction projects, and it is assumed that most new construction projects will be designed to full AASHTO design criteria, whether there are controlling crite- ria in place or not. The 13 controlling criteria were implemented in 1985 as an administrative control on the design process, to ensure that cer- tain design decisions with implications for traffic operations or for crash frequencies and severities were referred to manage- ment levels above the project design team within a highway agency or within FHWA. When the project team sees a need to design one or more particular geometric features of a project to criteria less than full AASHTO design criteria, a design excep- tion document is prepared and submitted for approval, as appropriate, to higher management levels within the highway agency or to FHWA. There was a clear rationale for this pro- cess in 1985 because the traffic operational and safety implica- tions of the 13 controlling criteria were largely unknown. Since knowledge was lacking, judgment was required, and the design exception process was created as an administrative control to elevate that judgment to a higher management level than the project design team. The design process in 1985 and before was very much a standards-based process, in which compliance with the design criteria in the Green Book or highway agency design manuals was presumed to result in design of a safe and efficient road- way. This approach is referred to in the HSM (12) as achieving nominal safety for a project. Today, as this report demon- strates, much more is known about the traffic operational and safety effects of the 13 controlling criteria. There is, thus, the potential for more of a performance-based design process Interpretation of Results that has the potential to achieve what the HSM refers to as substantive safety, rather than mere nominal safety. Adminis- trative controls like the 13 controlling criteria and the design exception process, using a standards-based process, should be less necessary in today’s more knowledgeable environment, when designers can explicitly consider the traffic operational or safety implications of their decisions. At present, the state of knowledge may simply support changes to the con- trolling criteria. Given the current and likely future state of knowledge, it may be reasonable in the future to change to a performance-based design process in which highways are designed toward target levels of crash frequency and sever- ity, and designers have flexibility about what combinations of geometric elements are used to achieve those levels on specific roadway sections. The appropriate administrative controls to be incorporated in a performance-based process will need to be determined at a later date. The scope of this research is limited to the 13 specific design elements for roadway geometrics that have been selected as controlling criteria by FHWA. Neither FHWA’s controlling cri- teria nor the scope of this research address other design fea- tures such as intersections, access control, or roadside features. Ultimately, retaining, modifying, or dropping any of the 13 controlling criteria is a policy decision, and the portion of that decision that involves federal policy is beyond the scope of this research. However, documentation of knowledge about the traffic operational and safety effects of the 13 control- ling criteria, establishment of priorities for the 13 controlling criteria, and recommendations concerning modification of the controlling criteria for application to non-federal proj- ects is within the scope of the research. All recommendations given below or in Section 8 concerning modification of the 13 controlling criteria should be read as referring to projects to which the controlling criteria are applied based on state policy, rather than federal policy. The interpretation discussion presented below focuses on traffic operational issues in terms of the effect of specific

82 design features on average travel speed for roadway traffic and focuses on safety issues in terms of average fatal-and-injury crash frequency. Whenever the following discussion men- tions crash frequency, it is referring specifically to fatal-and- injury crash frequency. The interpretations presented below tend to be based more on safety effects than on traffic opera- tional effects, because safety concerns tend to be the focus of most discussions about design criteria. The sensitivity analyses presented in Section 6 addressed rural two-lane highways, rural multilane highways, and rural freeways. While the sensitivity analyses addressed rural free- ways, the results of these analyses appear applicable to urban freeways as well, so the interpretations below are applicable to freeways of all types. Interpretation of the research results for each type of con- trolling criteria is presented below, followed by two brief summary sections. 7.1 Shoulder Width The research results indicate that shoulder width should remain as a controlling criterion for rural two-lane highways, rural multilane highways, and rural and urban freeways. Shoulder width has the largest effect on crash frequency of any of the controlling criteria for rural highways. Shoulder width also has the largest effect on traffic speed of any of the controlling criteria for rural two-lane highways. Thus, it is reasonable that, if part of the current design exception pro- cess is to remain in place, highway agencies should require design exceptions for shoulder width on rural highways. Shoulder width is less appropriate as a controlling criterion for urban and suburban arterials. There are no documented effects of shoulder width on traffic speed or crash frequency for urban and suburban arterials. Furthermore, it is acceptable to design urban and suburban arterials meant to be lower speed roads with curb-and-gutter sections, rather than with shoulders. Therefore, there does not appear to be a strong need to retain shoulder width as a controlling criterion for urban and suburban arterials. 7.2 Lane Width Lane width appears to be the second most important design criterion with respect to crash frequency on rural highways, and generally the first or second most important design cri- terion with respect to traffic speed on rural highways. Thus, lane width appears very appropriate to retain as a controlling criterion for rural highways. It should be noted that the HSM shows very limited differences in crash frequency between 11- and 12-ft lanes on rural two-lane and multilane highways (nonfreeways). It appears reasonable that designers should be provided with great flexibility to choose between 11- and 12-ft lanes for rural two-lane and multilane highways (non- freeways) and that the controlling criterion for lane width, and thus the need for design exceptions, should apply to lane widths less than 11 ft on rural two-lane and multilane high- ways (nonfreeways). There are no documented relationships that indicate an effect of lane width on crash frequency for urban and subur- ban arterials, and research under NCHRP Project 17-53 found no effect of lane width on traffic speed for urban and suburban arterials. Recent research found no effect of lane width on safety for urban and suburban arterials, with only limited exceptions that may possibly represent random effects (23, 24). The Green Book provides substantial flexibility in choosing among 10-, 11-, and 12-ft lanes for urban and suburban arterials (4, 5). Using narrower lanes on urban and suburban arterials can provide space for incorporation of other features that are pos- itive for operations and safety including medians, turn lanes, bicycle lanes, parking lanes, and shorter pedestrian crossings. It appears reasonable that designers should be provided with substantial flexibility to choose among 10-, 11-, and 12-ft lanes on urban and suburban arterials and that the controlling criterion for lane width, if retained, should apply only to lane widths less than 10 ft on urban and suburban arterials. This is not intended to imply that lane widths are not an important consideration in the design of urban and suburban arterials, or that any lane width can be used at any location, but rather that lane widths should be selected on a location-by-location basis to complement the other selected features of the road- way cross section within the available cross-section width. The Green Book should include clear design guidance that the needs of bicycles, trucks, and buses should be considered in any decision to use lane widths less than 12 ft across all lanes on an urban or suburban arterial. Where substantial volumes of bicycles, trucks, or buses are present, consideration should be given to maintaining wider curb lanes to accommodate them, even where other lanes are narrowed to less than 12 ft. Future research on lane width effects on urban and suburban arterials is planned under NCHRP Project 03-112. 7.3 Horizontal Curve Radius Horizontal curve radius has a documented relationship to crash frequency, either the third or fourth largest effect of any design criterion, for rural highways of all types. The effect of horizontal curve radius on crash frequency is quite substantial on horizontal curves themselves, but is limited to third or fourth place overall because horizontal curves typi- cally constitute only a portion of the length of any extended roadway section (e.g., 20 percent of the total roadway length in the sensitivity analyses reported in Section 6). It appears appropriate to retain horizontal curve radius as a controlling criterion for rural highways.

83 Horizontal curve radius does influence speeds on urban and suburban arterials. There is no definitive relationship of horizontal curve radius to crash frequency for urban and suburban arterials. Hauer at al. (32) found on-road crash fre- quencies for horizontal curves on urban four-lane undivided arterials to be lower than for tangent sections in the same corridors; the opposite was found to be the case for off-road crashes. Since on-road crashes are predominant on urban arterials, this suggests that horizontal curves do not have a role in increasing crash frequencies. Based on the available evidence, consideration might be given to dropping hori- zontal curve radius as a controlling criterion for urban and suburban arterials, at least for arterials with design speeds of 45 mph or less. 7.4 Superelevation The sensitivity analysis in Section 6 found that the effect of superelevation on crash frequency for rural two-lane highways is similar in magnitude to, but slightly smaller than, the effect on crash frequency of horizontal curve radius. Among the documented effects of design criteria on crash frequency for rural two-lane highways, superelevation ranks fifth in magnitude. There has been no research on the effects of superelevation on safety for rural multilane highways or freeways, but there is no reason to presume that this effect is not similar to the effect for rural two-lane highways. It seems reasonable to retain superelevation as a controlling criterion for rural highways and freeways, as long as horizontal curve radius is also retained. Because of generally lower speeds, superelevation of hori- zontal curves is likely to have a much less important influence on crash frequency on urban and suburban arterials than on rural highways or freeways. Curves on urban and suburban arterials that have limited radii typically have lower speeds such that normal cross slope can be retained throughout the curve rather than using a superelevated cross section. Based on the available evidence, consideration might be given to dropping superelevation as a controlling criterion for urban and suburban arterials, at least for arterials with design speeds of 45 mph or less. 7.5 Grade Grade has a documented effect on crash frequency for rural two-lane highways that is slightly larger than the effect of horizontal curves discussed above. This result is based on sensitivity analyses in which the grades of interest extended throughout the length of the roadway section of interest, in rolling terrain consisting of alternating upgrades and down- grades. Grade may have an effect on other rural roadway types, but this is difficult to document because very few rural multilane highways and freeways have steep grades, except in mountainous terrain where only steeper grades are practical. It seems reasonable to retain grade as a controlling criterion for rural highways, as long as the design criteria recognized that steeper grades are needed in mountainous terrain. Grade has no documented effect on traffic speed or crash frequency on urban and suburban arterials. Steep grades on urban and suburban arterials are rare except in locations where steep terrain gives designers little choice. There does not appear to be a strong need to retain grade as a controlling criterion for urban and suburban arterials. 7.6 Stopping Sight Distance The results of research conducted in this project (see Sec- tion 4.7) provide an important perspective on the role of stopping sight distance in safety. These results indicate that stopping sight distance has no effect on safety at crest vertical curves except when the presence of a crest vertical curve hides a horizontal curve, intersection, or driveway from the view of approaching drivers. When no hidden curve, intersection, or driveway is present, the situations in which drivers might be called upon to make a stop on a rural highway are rare. Thus, the research results indicate that stopping sight distance is much more important in some locations than in others. Our current approach to design appears to treat stopping sight distance as if it were equally important at all locations on the highway system. New construction projects generally can and should be designed to provide the full stopping sight distances presented in the AASHTO Green Book. However, in improve- ment projects on existing roadways, where stopping sight dis- tances less than (especially just less than) AASHTO criteria are present, consideration should be given to any history of sight-distance-related crashes at the site and to the presence of hidden features that might lead to future crashes as part of any decision to invest in sight distance improvements. Table 73 shows that even where a hidden feature is pres- ent at each crest vertical curve, the effect on crash frequency for a 5-mi section of rural two-lane highway, as a whole, is extremely small (0.02 or 0.03 percent); it is likely, however, that some hidden features could influence crash frequency more substantially depending on the frequency of slowing, turning, or stopping vehicles. There is no reason to suppose that this research finding for vertical sight restrictions would not also apply to horizontal sight restrictions caused by sight obstructions on the inside of horizontal curves. Similarly, only hidden curves, intersections, or driveways on rural multi- lane highways or hidden curves or ramps on freeways appear likely to increase crash frequencies. It might be argued that queues of stalled traffic could be hidden by limited sight dis- tance on freeways, but it should be kept in mind that vehicles are substantially taller than the 2-ft object height used in

84 stopping sight distance design, and vehicles are typically 6-ft wide and, therefore, extend 3 ft both to the left and the right of the nominal sight line along the center of a lane. Based on the available research findings, there does not appear to be a strong case for retaining stopping sight dis- tance as a controlling criterion for rural highways and free- ways. Many millions of dollars have been spent by highway agencies in improving crest vertical curves on existing rural highways and freeways to full AASHTO design criteria while providing little or no reduction in crash frequency. Funds available for safety improvement are too scarce to be spent in ways that provide little or no safety benefit. Better design guid- ance would be to improve stopping sight distance on existing rural highways or freeways only where specific crash patterns are present that indicate a need for such improvements or where an approaching curve, intersection, ramp, or driveway is hidden from the driver’s view by the stopping sight distance limitation. This guidance is not meant to suggest that stopping sight distance is unimportant, but rather to suggest that its importance varies substantially from location to location and is best assessed on a location-by-location basis. There is no research on the effect of stopping sight distance on crash frequency on urban and suburban arterials. However, there is no reason to suppose that the effect of stopping sight distance on urban and suburban arterials is much different than the effect of stopping sight distance on rural highways. Indeed, given the nature of urban and suburban arterials, it is likely that intersection sight distance, which is outside the scope of this research and outside the scope of the 13 control- ling criteria, has a larger effect on crash frequency on urban and suburban arterials than stopping sight distance. There- fore, consideration should be given to dropping stopping sight distance as a controlling criterion for urban and suburban arterials, while emphasizing the importance of considering stopping sight distance where specific crash patterns are pres- ent that indicate a need for such improvements or where an approaching curve, intersection, ramp, or driveway is hidden by the stopping sight distance limitation. 7.7 Bridge Width Research conducted in this project found no relationship between bridge width and crash frequency on rural two-lane highways (see Section 4.3). Current design guidance is to maintain the full roadway width of the approach to the bridge (lane width plus shoulder width) across the bridge. However, many existing bridges have roadway widths that are narrower than the approach roadway width. The analysis reported in Section 4.3 found no evidence that such bridges, on aver- age, experience more crashes than bridges on which the full roadway width is carried across the bridge; in fact, narrower bridges in many cases appeared to have fewer crashes than bridges on which the full roadway width is carried across the bridge, although the differences were not statistically signifi- cant. The research did not address one-lane bridges. The research results do not indicate any need to retain bridge width as a controlling criterion for rural two-lane highways. The logical interpretation of the research results is that if an existing bridge on a rural two-lane highway has a roadway narrower than the approach roadway, is in good structural condition (i.e., does not need replacement for structural reasons), and has no accompanying pattern of crashes (e.g., fixed-object, sideswipe, or head-on collisions) indicating a con- cern related to bridge width, the existing bridge may remain in place. Since funds for safety improvements are limited, it would likely be preferable to find a better investment of those funds than widening a bridge that is performing satisfacto- rily. There is no logical reason to believe that this same design approach is not applicable to bridges on other roadway types as well. 7.8 Cross Slope There is no research that indicates a relationship between the normal cross slope of roadway pavements and crash frequency. In fact, such research has never been conducted because existing roadway inventory data sets do not gener- ally include the cross slopes used on roadways. As a practi- cal matter, the normal cross slope likely does not vary much from the recommended Green Book (4, 5) design value of 1.5 to 2 percent, with appropriate adjustments for multi- lane pavements and areas that experience intense rainfall. Nevertheless, pavement cross slope is important to drainage, and improper drainage could contribute to potential vehicle loss of control under some circumstances. Given the lack of research (and, indeed, the lack of data for research) on this issue and the potential consequences of poor drainage, it makes logical sense to retain cross slope as a controlling criterion. While advances in knowledge have reduced the need for some other design elements to serve as controlling criteria, there have been no advances in knowledge about the traffic operational and safety effects of cross slope, so retaining it as a controlling criterion until better knowledge is available makes sense. The Green Book establishes a maximum design value of 8 percent for the cross-slope break between the outside edge of a superelevated pavement on a horizontal curve and a shoulder that slopes away from the roadway. As discussed in Section 2.10, the NTSB (37) has requested that FHWA and AASHTO investigate this design criterion, and research to address this issue is being conducted under NCHRP Project 03-105. Pending completion of that research, no change in the inclusion of pavement cross-slope breaks in the control- ling criterion for cross slope is recommended.

85 7.9 Sag Vertical Curve Length Sag vertical curves by their nature appear to be less related to crash frequency than crest vertical curves. The entire length of a sag vertical curve is visible to drivers under daylight con- ditions except in the rare cases where an overpass structure is present. At night, vehicle headlights illuminate only a portion of a sag vertical curve. However, it is known that headlight illumination distance is less than stopping sight distance even on level tangent roadways, so drivers “outdrive their head- lights” in many roadway situations, not just on sag vertical curves. The recent change in design criteria for crest verti- cal curves to use a 2-ft object height indicates that the small objects implied by the headlight sight distance model for sag vertical curve design may not represent an appropriate design approach. There does not appear to be justification for treating sag vertical curve length as a controlling criterion for design. 7.10 Horizontal Clearance/Lateral Offset Sections 2.13 and 5.1.8 document the controlling criterion for horizontal clearance and the change in terminology to lat- eral offset that occurred in the 2011 Green Book (5) and the 2011 RDG (40). Lateral offset is essentially irrelevant as a con- trolling criterion for roadway types other than urban and sub- urban arterials, because the controlling criterion for shoulder width ensures that there will be a lateral offset to roadside objects of at least 18 in. On urban and suburban arterials, any effect on traffic speed due to roadside objects less than 18 in behind the curb would be minimal. The primary function of the lateral offset design criterion is to ensure that mirrors or other appurtenances of heavy vehicles do not strike roadside objects and that passengers in parked cars are able to open their doors. While these considerations are important, they do not appear to rise to the level of importance that attaches to other design criteria that may address the likelihood of fatal- and-injury crashes and, therefore, horizontal clearance/lateral offset does not appear to need administrative control as a con- trolling criterion for design. 7.11 Summary of Results for Rural Two-Lane Highways, Rural Multilane Highways, and Rural and Urban Freeways It is recommended that the following design criteria should be retained as controlling criteria for rural two-lane highways, rural multilane highways, and rural and urban freeways: shoulder width, lane width (for lane width less than 11 ft), horizontal curve radius, superelevation, grade, stopping sight distance (for locations where a hidden curve, intersection, ramp, or driveway is present), and cross slope. There does not appear to be any need, based on their traffic operational and safety effects, for the following design criteria to be retained as controlling criteria: bridge width, sag vertical curve length, and horizontal clearance/lateral offset. This does not imply that bridge width, sag vertical curve length, and horizontal clear- ance/lateral offset should not continue to be important design considerations; clearly, they should continue to be addressed in the Green Book, in highway agency design manuals, and during the design process. Rather, it means that the traffic operational and safety effects of these design criteria do not appear to rise to the level that requires an administrative con- trol like the controlling criteria. The priority rankings of the 13 controlling criteria in Table 82 and the quantitative sensi- tivity analysis results presented in Section 6 provide support to this recommendation. 7.12 Summary of Results for Urban and Suburban Arterials The research results for urban and suburban arterials presented in Sections 2 through 6 do not indicate that the roadway design features represented by the 13 controlling criteria are critical factors in the design of urban and sub- urban arterials. More than other roadway types, the traffic operational and safety performance of urban and subur- ban arterials appears to depend on factors outside the scope of the 13 controlling criteria and outside the scope of this research, such as intersection design and access management. Well-reasoned and well-explained geometric design criteria, with flexibility to adapt roadway cross sections to the spe- cific needs of each corridor, along with appropriate intersec- tion design and access management criteria, would appear to be of greater importance to design of urban and suburban arterials than the administrative controls provided by the 13 controlling criteria and the design exception process. There- fore, it is recommended that consideration be given to drop- ping application of the 13 controlling criteria to urban and suburban arterials or restricting the controlling criteria to a minimal set, including lane width (for lane widths less than 10 ft), stopping sight distance (for locations with a hidden curve, intersection, or driveway), and cross slope. A possible exception to this recommendation is for urban and suburban arterials with design speeds over 45 mph; such arterials are designed more like rural highways and the same controlling criteria as for rural two-lane highways, rural multilane high- ways, and rural and urban freeways might be applied.

Next: Section 8 - Conclusions and Recommendations »
Evaluation of the 13 Controlling Criteria for Geometric Design Get This Book
×
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

TRB’s National Cooperative Highway Research Program (NCHRP) Report 783: Evaluation of the 13 Controlling Criteria for Geometric Design describes the impact of the controlling roadway design criteria on safety and operations for urban and rural roads.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!