Guidelines for Integrating Safety and Cost-Effectiveness into Resurfacing, Restoration, and Rehabilitation (3R) Projects (2021)

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.

98 3R Project Design Guidelines for Specific Roadway Types This chapter presents design guidelines for 3R projects. The design guidelines are based on explicit consideration of estimates of the crash reduction effectiveness of design improvements, where available, and on the use of before–after benefit–cost analysis tools. The design guidelines are organized by roadway type and, within roadway type, by design element. 6.1 Rural Two-Lane Highways 6.1.1 Lane Widening Lane widening should be considered for each 3R project on a rural two-lane highway with existing lane widths of less than 12 ft. Decisions about lane widening for 3R projects should be based on benefit–cost analysis. Lane widening is a desirable investment only when the analysis indicates that the expected crash reduction benefits will exceed the costs of the improvement or when a crash analysis finds existing crash patterns that can potentially be reduced by lane widen- ing. Where there are no existing crash patterns that can potentially be reduced by lane widening and the expected crash reduction benefits are less than the costs of improvement, lane widening would be a poor investment, and available funds would be better invested at another location where the crash reduction benefits would be greater. Two approaches to benefit–cost analysis are applicable to lane widening: • Option 1. Conduct a site-specific benefit–cost analysis for each individual 3R project. • Option 2. Develop agency-specific guidelines for minimum traffic volumes that justify lane widening. 6.1.1.1 Option 1: Conduct a Site-Specific Benefit–Cost Analysis for Each Individual 3R Project The most desirable approach to cost-effective decisions about lane widening is to conduct a site-specific benefit–cost analysis for each individual 3R project. Site-specific benefit–cost analyses are desirable because even nominally similar sites may differ in roadway characteristics, traffic volumes, crash history, and improvement costs, such that lane widening may be cost- effective at one site and not at another. The crash reduction effectiveness of lane widening on rural two-lane highways is documented in Section 4.3.1.1. Procedures for site-specific benefit– cost analysis are illustrated in Section 5.2.2. The benefit–cost analysis tools provided with these guidelines enable site-specific benefit–cost analyses to be performed efficiently. Figure 53 illustrates the application of Spreadsheet Tools 1 and 2 for site-specific benefit–cost analysis of lane widening. The example shows a rural two-lane highway with existing 9-ft lanes for which the alternative with the maximum net benefit is to widen the lanes to 11 ft. It should C H A P T E R   6

3R Project Design Guidelines for Specific Roadway Types 99   Lane Widening on Rural Two-Lane Highways Section length 1.000 mi AADT 2,000 veh/day Terrain Level Pavement type Flexible Lane width 9.0 ft Shoulder width 2 ft Shoulder type Unpaved Roadside slope 1V:2H Centerline rumble strip No Shoulder rumble strip No Crash history No Number of curves in roadway section 2 Maximum superelevation rate 8% Design speed 55 mph Horizontal Curve Curve Length (mi) Radius (ft) Presence of Spiral Transitions Existing Superelevation Rate (%) 1 0.156 400 No 2.4 2 0.125 600 No 3.8 Results of Benefit–Cost Evaluation with Tool 1 for Widening to 11-ft Lanes Widening the existing 9-ft lanes to 11 ft produces the highest net benefit for a rural two-lane highway with these roadway characteristics. Thus, if lane widening was being considered for inclusion in a 3R project on this roadway section, the lanes should be widened to 11 ft to get the highest return on investment. Evaluation of a Full Range of Lane-Widening Scenarios with Tool 2 Tool 2 can be used to evaluate all lane-widening possibilities at one time. The results of the Tool 2 analysis shown below indicate that widening to 11-ft lanes provides the largest net benefit. Improved Lane Width (ft) Net Benefit ($) B/C Ratio Total Benefit ($) Total Cost ($) 11.0 100,040 2.147 187,256 87,216 11.5 94,628 1.918 197,668 103,041 12.0 89,215 1.751 208,081 118,865 10.5 63,803 1.894 135,194 71,391 10.0 27,566 1.496 83,132 55,567 9.5 1,824 1.046 41,566 39,742 Figure 53. Spreadsheet example: Lane widening on rural two-lane highways.

100 Guidelines for Integrating Safety and Cost-Effectiveness into Resurfacing, Restoration, and Rehabilitation (3R) Projects be noted that the HSM Chapter 10 crash prediction procedures show that the benefit of lane widening from 11 to 12 ft on rural two-lane highways is relatively small (2). Therefore, widening narrower lanes to 11 ft will often be the most cost-effective improvement for rural two-lane highways. Widening to 12 ft lanes may only be the most cost-effective improvement for roads with relatively high traffic volumes. 6.1.1.2 Option 2: Develop Agency-Specific Guidelines for Minimum Traffic Volumes That Justify Lane Widening A less-desirable, but still acceptable, approach to cost-effective lane-widening decisions in 3R projects is to develop agency-specific guidelines for minimum traffic volumes that justify lane widening analogous to those shown in Table 43. Each highway agency (or, in some cases, individual districts or regions within a highway agency) can develop guidelines for minimum traffic volumes as a basis for lane-widening decisions. Section 5.4 illustrates procedures for developing guidelines for minimum traffic volumes like those in Table 43. The results of the example benefit–cost analysis to develop minimum AADT guidelines presented in Table 43 show that lane widening to 12 ft is likely to be cost-effective for 3R projects on rural two-lane highways with existing lane widths of 10 ft or less and traffic volumes greater than approximately 3,000 to 4,000 veh/day, even for projects in level terrain, which generally cost less to implement than projects in rolling or mountainous terrain. On rural two-lane highways with lower traffic volumes, lane widening would not be cost-effective, and, therefore, should not be considered unless a site-specific crash pattern indicates a need for lane widening. Lane widening to 12 ft is likely to be cost-effective for 3R projects on rural two-lane highways with existing lane widths of 11 ft only for traffic volumes of approximately 14,000 veh/day or more, even in level terrain. Higher traffic volumes are needed to justify widening of existing 11-ft lanes because HSM Chapter 10 indicates that the differences in crash frequency between rural two-lane highways with 11- and 12-ft lanes are relatively small (2). The traffic volume thresholds presented above are merely examples, since the results of benefit–cost analyses can vary from site to site and from agency to agency on the basis of roadway characteristics, crash history, and improvement cost levels. Highway agencies also differ in the crash cost values used in benefit–cost analyses. For this reason, Table 43 is not intended for direct application, but individual highway agencies can develop tables analogous to Table 43 by following the procedures in Section 5.4. 6.1.2 Shoulder Widening and Paving Shoulder widening should be considered for each 3R project on a rural two-lane highway with existing shoulder widths of less than 6 ft. Decisions about shoulder widening for 3R projects should be based on benefit–cost analysis. Shoulder widening is a desirable investment only when the analysis indicates that the expected crash reduction benefits will exceed the costs of the improvement or when a crash analysis finds existing crash patterns that can poten- tially be reduced by shoulder widening. Where there are no existing crash patterns that can potentially be reduced by shoulder widening and the expected crash reduction benefits are less than the costs of improvement, shoulder widening would be a poor investment, and available funds would be better invested at another location where the crash reduction benefits would be larger. Two approaches to benefit–cost analysis are applicable to shoulder widening: • Option 1. Conduct a site-specific benefit–cost analysis for each individual 3R project. • Option 2. Develop agency-specific guidelines for minimum traffic volumes that justify shoulder widening.

3R Project Design Guidelines for Specific Roadway Types 101   6.1.2.1 Option 1: Conduct a Site-Specific Benefit–Cost Analysis for Each Individual 3R Project The most desirable approach to decisions about cost-effective shoulder widening is to conduct a site-specific benefit–cost analysis for each individual 3R project. Site-specific benefit–cost analyses are desirable because even nominally similar sites may differ in roadway characteristics, traffic volumes, crash history, and improvement costs, such that shoulder widening may be cost-effective at one site and not at another. The crash reduction effectiveness of shoulder widening on rural two-lane highways is documented in Section 4.3.1.2. Procedures for site- specific benefit–cost analysis are illustrated in Section 5.2.2. The benefit–cost analysis tools provided with these guidelines enable site-specific benefit–cost analyses to be performed efficiently. Figure 54 illustrates the application of Spreadsheet Tools 1 and 2 for site-specific benefit–cost analysis of shoulder widening. 6.1.2.2 Option 2: Develop Agency-Specific Guidelines for Minimum Traffic Volumes That Justify Shoulder Widening A less-desirable, but still acceptable, approach to cost-effective shoulder-widening decisions in 3R projects is to develop agency-specific guidelines for minimum traffic volumes that justify Shoulder Widening on Rural Two-Lane Highways Results of Benefit–Cost Evaluation with Tool 1 for Widening to 8-ft Shoulders For the same rural two-lane highway section as in the previous lane-widening example, Tool 1 can be used to find the width to which the shoulders should be widened in order to produce the highest net benefit. Widening the unpaved shoulders to 8 ft produces the highest net benefit, shown in the following results. Evaluation of a Full Range of Shoulder-Widening Scenarios with Tool 2 Tool 2 can be employed to analyze all shoulder-widening scenarios at one time. The results of the Tool 2 analysis, shown below, indicate that widening to 8-ft shoulders provides the largest net benefit. Improved Shoulder Width (ft) Net Benefit ($) B/C Ratio Total Benefit ($) Total Cost ($) 8 86,727 2.080 167,020 80,294 7 70,457 2.000 140,928 70,470 6 55,063 1.908 115,711 60,647 5 35,764 1.704 86,588 50,824 4 17,475 1.426 58,476 41,001 3 −2,274 0.927 28,904 31,178 Figure 54. Spreadsheet example: Shoulder widening on rural two-lane highways.

102 Guidelines for Integrating Safety and Cost-Effectiveness into Resurfacing, Restoration, and Rehabilitation (3R) Projects shoulder widening analogous to those shown in Table 44. Each highway agency (or, in some cases, individual districts or regions within a highway agency) can develop guidelines for minimum traffic volumes as a basis for shoulder-widening decisions. Section 5.4 illustrates procedures for developing guidelines for minimum traffic volumes like those in Table 44. Benefit–cost analyses can also be applied to rural two-lane highway sites with unpaved shoulders to consider whether the shoulder width should be fully or partially paved. The most desirable approach is to consider paving alternatives for each shoulder width alternative being evaluated. A less-desirable but still acceptable approach is to determine the optimal shoulder width for the project and then evaluate whether the shoulder should be paved. Many highway agencies have design policies that utilize paved shoulders only for specific roadway types or above specific traffic volume levels and otherwise use unpaved or composite shoulders. Nothing in these guidelines requires a highway agency to provide paved shoulders on projects where the agency’s design policy calls for unpaved or composite shoulders. 6.1.3 Horizontal Curve Improvements Improvements to horizontal curves on rural two-lane highways may be considered 3R projects in some cases. Where the superelevation of an existing curve is less than the design super- elevation value in the AASHTO Green Book by more than 1% (4), HSM Chapter 10 indicates that there will be a safety benefit from restoring the superelevation to the Green Book value (2). Spreadsheet Tool 1 can be used to assess the cost-effectiveness of such superelevation improve- ments. Figure 55 gives an example of benefit–cost analysis with Tool 1 for a superelevation improvement. Realignment of an isolated horizontal curve on a rural two-lane highway can also be consid- ered as part of a 3R project. Typically, such an improvement would involve increasing the radius and, therefore, lengthening the curve. Spreadsheet Tool 1 includes a benefit–cost evaluation procedure for assessing realignment of a single horizontal curve as part of a 3R project on a rural two-lane highway. While Spreadsheet Tool 1 has the capability to evaluate curve realignment Superelevation Improvement on Rural Two-Lane Highway This example considers the same rural two-lane highway segment considered in the lane-widening example in Section 6.1.1. This segment contains two horizontal curves. These two curves have existing superelevation rates that are less than the superelevation rates given in the Green Book. Tool 1 is used to determine whether improving the superelevation rates to the Green Book’s proposed rates yields a safety benefit and whether the benefit outweighs the implementation cost. For both curves, the improved superelevation rate will be 8.0%. The Tool 1 results indicate that the superelevation improvement is cost effective, with a benefit–cost ratio exceeding 8 and a net benefit of $33,007. Figure 55. Spreadsheet example: Superelevation improvement on rural two-lane highway.

3R Project Design Guidelines for Specific Roadway Types 103   projects, sensitivity analyses with the tool have shown that, because of their high cost, curve realignment improvements are seldom cost-effective. Realignment of multiple horizontal curves on a roadway section would generally be considered as reconstruction and, therefore, out of the scope of 3R improvements. At horizontal curve sites with sharp curves where curve realignment is found not to be cost- effective, chevron markers can be installed as part of a 3R project to better delineate the curved path for drivers. There are no accepted CMFs for installation of chevron markers to improve delineation on a horizontal curve, so evaluation of chevron markers is not included in the spreadsheet tools. 6.1.4 Sight Distance Improvements Stopping sight distance (SSD) is provided along roadways to assist drivers in detecting and responding to situations on the roadway ahead that may require them to slow or stop. Sight distance limitations may include crest vertical curves or objects on the inside of horizontal curves. The SSD design criteria used by most states are based on the SSD values in the AASHTO Green Book (4). Recent research has shown that limited SSD on rural two-lane highways is unlikely to lead to crashes unless the portion of the roadway hidden from the driver’s view by the sight distance limitation includes a roadway feature at which drivers need to take steering or braking action, such as an intersection, a driveway, or a horizontal curve (5). In an area on a rural two-lane highway whose SSD is less than the Green Book design criteria but has no hidden features such as intersections, driveways, or horizontal curves and no history of crashes related to limited SSD, sight distance improvements are unlikely to have any effect on crash frequency or severity. Therefore, such improvements are unlikely to be cost-effective and need not be considered in 3R projects. In contrast, where the SSD of a portion of the highway is less than the Green Book design criteria and an intersection, driveway, or horizontal curve is present or the crash history shows a pattern of crashes potentially related to limited SSD, an SSD improvement would be desirable. SSD improvements can include realigning a crest vertical curve or removing or relocating objects on the inside of a horizontal curve. Where the cost of the SSD improve- ment is substantial, consideration may be given to mitigation measures such as providing wider shoulders and advance warning signs. 6.1.5 Bridge Width Recent research has shown that narrow bridges on rural two-lane highways are not typically associated with increases in crash frequency (5). [Narrow bridges are defined as bridges where the curb-to-curb width of the bridge roadway is less than the width of the approach roadway (lanes and shoulders combined).] Widening or replacing bridges on rural two-lane highways is not likely to result in crash reduction, even in the case of narrow bridges. Bridges on two-lane highways should remain in place in 3R projects unless there is either (a) a structural need to strengthen or replace the bridge or (b) a documented pattern of crashes at the bridge that can potentially be reduced by widening or replacing the bridge. 6.1.6 Normal Pavement Cross Slope Pavement cross slope is needed for drainage so that water flows off the traveled way during and after precipitation. Each highway agency has its own design criteria for normal pavement cross slope; these criteria are typically selected as suitable for local climate conditions. Where the cross section of the traveled way within a 3R project does not have sufficient cross slope

104 Guidelines for Integrating Safety and Cost-Effectiveness into Resurfacing, Restoration, and Rehabilitation (3R) Projects for drainage, the pavement cross slope should be restored to meet the highway agency’s appli- cable design criteria. Given the importance of pavement cross slope to drainage, no benefit–cost analysis is needed to justify restoration of normal pavement cross slope in a 3R project on a rural two-lane highway. 6.1.7 Rumble Strip Improvements Centerline rumble strips are provided between the two directions of travel on rural two-lane highways to provide an audible and tactile warning to drivers when their vehicle leaves its intended travel lane and begins to cross the roadway centerline. Similarly, shoulder rumble strips are provided on a paved shoulder or at the edge of the traveled way on rural two-lane highways to provide an audible and tactile warning to drivers when their vehicle is leaving the roadway and begins to encroach on the shoulder. These warnings alert drivers to take corrective action— steering and, where appropriate, braking—to return to their intended travel lane. The crash reduction effectiveness of installing rumble strips on rural two-lane highways is documented in Sections 4.3.1.5 and 4.3.1.6. An example benefit–cost analysis has shown that installation of centerline or shoulder rumble strips in conjunction with a resurfacing project can become cost-effective on rural two-lane highways at AADTs as low as 400 veh/day (6). This AADT threshold is merely an example, and agency-specific minimum AADT guidelines can be developed with Spreadsheet Tool 1. However, such agency-specific guidelines are likely to show that installing centerline or shoulder rumble strips is cost-effective for most rural two-lane highways, other than very-low-volume roads. The examples of benefit–cost analysis presented in Figures 56 and 57 show that both center- line and shoulder rumble strips can be highly cost-effective on rural two-lane highways, even at a modest traffic volume of 2,000 veh/day. For any level of traffic volume on rural two-lane highways where installing rumble strips would be cost-effective, highway agencies are encouraged to provide centerline or shoulder rumble strips as part of 3R projects, where they are not already present, and to restore existing rumble strips when the pavement or shoulder is resurfaced. Rumble strips should be designed Install Centerline Rumble Strip on Rural Two-Lane Highway This example considers the same rural two-lane highway segment considered in the lane-widening example in Section 6.1.1. The economic impact of installing a centerline rumble strip on the rural two-lane highway segment is analyzed with Tool 1. The benefit–cost ratio is exceptionally high, and the net benefit of the treatment installation is $53,138. Due to low installation cost and high safety benefit, installing centerline rumble strips will almost always be economically justified. Figure 56. Spreadsheet example: Installation of centerline rumble strip on rural two-lane highway.

3R Project Design Guidelines for Specific Roadway Types 105   and located in accordance with each agency’s current practices. Rumble strips are appropriate in most rural locations, except where the noise created by the rumble strips may disturb nearby residents. Rumble strips on paved shoulders need to be located sufficiently close to the traveled way that the effectiveness of the paved shoulder as a travel path for bicyclists is not reduced. Where rumble strips are cost-effective, either centerline or shoulder rumble strips, or both, may be used. Use of both centerline and shoulder rumble strips together is very effective but may not be desirable on two-lane highways if the distance between the centerline and shoulder rumble strips is less than 12 ft. Limited separation between the centerline and shoulder rumble strips is undesirable because normal variations in lateral positioning of vehicles within a lane may lead drivers to strike one rumble strip or the other with enough frequency to become annoying. 6.1.8 Striping and Delineation Improvements Nearly every 3R project includes pavement resurfacing, so the cost of restoring pavement markings after resurfacing is automatically part of most 3R projects. These guidelines assume that, as a default, highway agencies will restore the pavement markings with the equiva- lent of the existing markings. Conventional paint is the least-expensive pavement marking material and typically has a service life of 1 to 2 years, depending on traffic volume and climate conditions. Spreadsheet Tool 1 provides the capability to assess the cost-effectiveness of striping and delineation packages for inclusion in 3R projects. The primary element of a striping and delin- eation package that may be evaluated with the spreadsheet tool as part of a 3R project is the provision of more durable pavement markings with longer life and higher retroreflectivity than conventional paint markings. Striping and delineation packages may also include the addition of wider edgelines and post-mounted roadside delineators. The implementation cost for the striping and delineation package is the full cost of striping and delineation minus the cost of the pavement markings that would have been implemented as a default. The crash reduction effectiveness of improved striping and delineation on rural two-lane highways is documented in Section 4.3.1.7. Figure 58 presents results of the benefit–cost analysis for adding enhanced striping and delineation on a rural two-lane highway. Install Shoulder Rumble Strip on Rural Two-Lane Highway This example considers the same rural two-lane highway segment considered in the lane-widening example in Section 6.1.1. Tool 1 is used to analyze the economic impact of installing shoulder rumble strips on the rural two-lane highway segment. As with the results for the centerline rumble strip, the benefit–cost ratio is exceptionally high and the net benefit is $68,355. Figure 57. Spreadsheet example: Installation of shoulder rumble strip on rural two-lane highway.

106 Guidelines for Integrating Safety and Cost-Effectiveness into Resurfacing, Restoration, and Rehabilitation (3R) Projects 6.1.9 Roadside Slope Flattening Roadside slope flattening should be considered for each 3R project on a rural two-lane highway with roadside slopes steeper than 1V:4H where sufficient right-of-way for slope flattening is available (or could be acquired) and where slope flattening would not adversely affect adjacent properties, structures, or environmentally sensitive areas such as wetlands. Decisions about roadside slope flattening for 3R projects should be based on benefit–cost analysis, and roadside slope flattening is a desirable investment only when the analysis indicates that the expected crash reduction benefits will exceed the improvement costs or when a crash analysis finds existing crash patterns (e.g., run-off-the-road crashes) that can potentially be reduced by roadside slope flattening. In many cases, roadside slope flattening will not be cost-effective because of the high costs of project implementation. Where there are no existing crash patterns that can potentially be reduced by roadside slope flattening and the expected crash reduction benefits from road- side slope flattening are less than the costs of improvement, roadside slope flattening would be a poor investment, and available funds would be better invested at another location where the crash reduction benefits would be larger. Roadside slope flattening should be considered only where a benefit–cost analysis with the site-specific benefit–cost analysis tool (Spreadsheet Tool 1) or with the RSAP model (23–25) indicates that the present value of the benefits of the slope flattening project would exceed the cost (i.e., the net benefits would exceed zero). The crash reduction effectiveness of roadside slope flattening on rural two-lane highways is documented in Section 4.3.1.8. Procedures for site-specific benefit–cost analysis are illustrated in Section 5.4. The benefit–cost analysis tools provided with these guidelines enable site-specific benefit–cost analyses to be performed efficiently. Adding Enhanced Striping and Delineation on a Rural Two-Lane Highway This example considers the same rural two-lane highway segment considered in the lane-widening example in Section 6.1.1. Tool 1 is used to examine the economic impact of adding enhanced striping and delineation. The following table presents additional inputs needed to perform the analysis. Percentage of roadway section with dashed centerline striping 20.00% Percentage of roadway section with solid–dash centerline striping 25.00% Percentage of roadway section with double solid centerline striping 55.00% Total length of roadway section with delineator posts (includes each side separately) 2.000 mi On the basis of the results from Tool 1, the net benefit of adding this treatment is $133,209 with a benefit–cost ratio of 2.466. Figure 58. Spreadsheet example: Adding enhanced striping and delineation on rural two-lane highway.

3R Project Design Guidelines for Specific Roadway Types 107   There is no option for developing minimum AADT guidelines for roadside slope flattening projects because the costs of such projects can vary widely from site to site, making site-specific cost estimates necessary. Figure 59 presents the results of benefit–cost analysis for roadside slope flattening on a rural two-lane highway. Where roadside slope flattening is considered, but not implemented, in a 3R project, con- sideration may also be given to mitigation measures such as removing roadside objects (see Section 6.1.10) or providing traffic barriers (see Section 6.1.11), where use of the RSAP model finds such improvements to be cost-effective. 6.1.10 Removal of Roadside Objects Where roadside objects are present within the clear zone width on a rural two-lane highway, their removal should be considered on the basis of a cost-effectiveness analysis with the RSAP model (23–25). Generally, only objects greater than 4 inches in diameter and not of breakaway design need to be considered. Where appropriate, the design of culvert ends should be improved [see Chapter 3 of the AASHTO Roadside Design Guide (30)]. Where roadside objects are present continuously or at regular intervals throughout all or part of the length of a 3R project, removal of those objects is not likely to be cost-effective, and a formal RSAP analysis is not needed. As an alternative to removing roadside objects, consideration may also be given to replacing such objects with a similar object of breakaway design, relocating the object behind an existing guardrail or traffic barrier, or installing a new guardrail or traffic barrier. Decisions concerning the cost-effectiveness of new guardrail or other traffic barriers should be made with the RSAP model (see Section 6.1.11). Figure 59. Spreadsheet example: Roadside slope flattening on a rural two-lane highway. Roadside Slope Flattening on a Rural Two-Lane Highway This example considers the same rural two-lane highway segment considered in the lane-widening example in Section 6.1.1. Tool 1 is used to determine whether roadside slope flattening is cost- effective. Analyzing each possible slope-flattening scenario individually can determine which improved roadside slope generates the highest net benefit. The results shown below are for flattening the roadside slope to a 1V:6H slope, which produces the highest net benefit. Tool 2 can be used to examine all possible slope-flattening scenarios at one time. The results of the Tool 2 analysis are shown below. Improved Slope Net Benefit ($) B/C Ratio Total Benefit ($) Total Cost ($) 1V:6H 46,868 1.732 110,932 64,064 1V:4H 12,757 1.299 55,466 42,709 1V:3H −22,788 0.289 9,244 32,032

108 Guidelines for Integrating Safety and Cost-Effectiveness into Resurfacing, Restoration, and Rehabilitation (3R) Projects 6.1.11 Installation/Rehabilitation of Guardrail and Other Traffic Barriers Existing guardrail or other traffic barriers that have reached the end of their useful life or are of obsolete design should generally be rehabilitated or replaced as part of a 3R project. The decision to rehabilitate or replace a guardrail or traffic barrier does not require a formal economic analysis, as long as the highway agency believes that the guardrail or traffic barrier is needed and warranted at its location. If the highway agency has reason to believe that the existing guardrail or traffic barrier may not be warranted, the warrants and length of need for the barrier can be assessed with the AASHTO Roadside Design Guide (30). Installation of new guardrail or other traffic barriers may be considered by highway agencies as part of 3R projects on rural two-lane highways. The RSAP model should be used to evaluate the cost-effectiveness of any new guardrail or traffic barrier installations that are considered. 6.1.12 Passing Lanes Passing lanes may be added as part of 3R projects on rural two-lane highways to increase passing opportunities. Passing lanes can be added in one or both directions of travel. Where a passing lane is added on an upgrade to allow vehicles to pass trucks or other heavy vehicles slowed by the upgrade, the added lane is known as a “climbing lane.” The traffic operational analysis procedures in the HCM can be used to assess the effect of added passing or climbing lanes on the traffic operational LOS for a rural two-lane highway (11). The AASHTO Green Book presents criteria for where climbing lanes should be considered on rural two-lane highways (4). Passing and climbing lanes should generally be installed only where an HCM analysis has established that there is a traffic operational need for the added lane. For climbing lanes, the traffic operational criteria in the AASHTO Green Book should also be met. In most cases in which an added passing or climbing lane is needed for traffic operational reasons, it will also have sufficient safety benefit to be cost-effective as a safety improvement. This can be verified with Spreadsheet Tool 1. 6.1.13 Intersection Turn Lane Improvements The crash reduction effectiveness of intersection turn lane improvements on rural two-lane highways is documented in Section 4.3.1.10. Benefit–cost analyses have shown that inter- section turn-lane improvements, including installation of left- and right-turn lanes, are likely to be cost-effective at any intersection where installation of such improvements makes traffic operational sense (7). Highway agencies should, therefore, assess the traffic operational need for intersection turn lanes according to established access management policies or traffic operational analysis tools. If the highway agency concludes that installation of the turn lane is justified on a traffic operational basis, there is little doubt that the turn lane will be cost-effective on a safety basis as well. 6.1.14 Other Intersection Improvements A 3R project may provide an opportunity for implementing other intersection improvements involving traffic control, signage, delineation, marking, or sight distance. Highway agencies should implement such improvements if a need is identified by a crash history review.

3R Project Design Guidelines for Specific Roadway Types 109   6.2 Rural Multilane Undivided Highways 6.2.1 Lane Widening Lane widening should be considered for each 3R project on a rural multilane undivided high- way with existing lane widths of less than 12 ft. Decisions about lane widening for 3R projects should be based on benefit–cost analysis. Lane widening is a desirable investment only when the analysis indicates that the expected crash reduction benefits will exceed the costs of the improvement or when a crash analysis finds existing crash patterns that can potentially be reduced by lane widening. Where there are no existing crash patterns that can potentially be reduced by lane widening and the expected crash reduction benefits are less than the costs of improvement, lane widening would be a poor investment, and available funds would be better invested at another location where the crash reduction benefits would be greater. Two approaches to benefit–cost analysis are applicable to lane widening: • Option 1. Conduct a site-specific benefit–cost analysis for each individual 3R project. • Option 2. Develop agency-specific guidelines for minimum traffic volumes that justify lane widening. 6.2.1.1 Option 1: Conduct a Site-Specific Benefit–Cost Analysis for Each Individual 3R Project The most desirable approach to cost-effective decisions about lane widening is to conduct a site-specific benefit–cost analysis for each individual 3R project. Site-specific benefit–cost analyses are desirable because even nominally similar sites may differ in roadway characteristics, traffic volumes, crash history, and improvement costs, such that lane widening may be cost- effective at one site and not at another. The crash reduction effectiveness of lane widening on rural multilane undivided highways is documented in Section 4.3.2.1. Procedures for site-specific benefit–cost analysis are illustrated in Section 5.2.2. The benefit–cost analysis tools provided with these guidelines enable site-specific benefit–cost analyses to be performed efficiently. An example of the application of Spreadsheet Tool 1 to a lane-widening improvement on a rural two-lane highway is presented in Section 6.1.1. Analysis of lane widening for rural multi- lane undivided highways can be performed in the same manner, with the only difference in input data being the roadway type. 6.2.1.2 Option 2: Develop Agency-Specific Guidelines for Minimum Traffic Volumes That Justify Lane Widening A less-desirable, but still acceptable, approach to cost-effective lane-widening decisions in 3R projects is to develop agency-specific guidelines for minimum traffic volumes that justify lane widening analogous to those shown in Table 43. Each highway agency (or, in some cases, individual districts or regions within a highway agency) can develop guidelines for minimum traffic volumes as a basis for lane-widening decisions. Section 5.4 illustrates procedures for developing guidelines for minimum traffic volumes like those in Table 43. An example of a benefit–cost analysis to develop minimum AADT guidelines for rural two-lane highways is presented in Table 43. The same analysis approach can be applied to lane widening on rural multilane undivided highways. 6.2.2 Shoulder Widening and Paving Shoulder widening should be considered for each 3R project on a rural multilane undivided highway with existing shoulder widths of less than 6 ft. Decisions about shoulder widening for

110 Guidelines for Integrating Safety and Cost-Effectiveness into Resurfacing, Restoration, and Rehabilitation (3R) Projects 3R projects should be based on benefit–cost analysis. Shoulder widening is a desirable invest- ment only when the analysis indicates that the expected crash reduction benefits will exceed the costs of the improvement or when a crash analysis finds existing crash patterns that can potentially be reduced by shoulder widening. Where there are no existing crash patterns that can potentially be reduced by shoulder widening and the expected crash reduction benefits are less than the costs of improvement, shoulder widening would be a poor investment, and avail- able funds would be better invested at another location where the crash reduction benefits would be larger. Two approaches to benefit–cost analysis are applicable to shoulder widening: • Option 1. Conduct a site-specific benefit–cost analysis for each individual 3R project. • Option 2. Develop agency-specific guidelines for minimum traffic volumes that justify shoulder widening. 6.2.2.1 Option 1: Conduct a Site-Specific Benefit–Cost Analysis for Each Individual 3R Project The most desirable approach to cost-effective decisions about shoulder widening is to con- duct a site-specific benefit–cost analysis for each individual 3R project. Site-specific benefit–cost analyses are desirable because even nominally similar sites may differ in roadway characteristics, traffic volumes, crash history, and improvement costs, such that shoulder widening may be cost- effective at one site and not at another. The crash reduction effectiveness of shoulder widening on rural multilane undivided highways is documented in Section 4.3.2.2. Procedures for site- specific benefit–cost analysis are illustrated in Section 5.2.2. The benefit–cost analysis tools pro- vided with these guidelines enable site-specific benefit–cost analyses to be performed efficiently. An example of the application of Spreadsheet Tool 1 to a shoulder widening improvement on a rural two-lane highway is presented in Section 6.1.2. Analysis of shoulder widening for rural multilane undivided highways can be performed in the same manner, with the only difference in input data being the roadway type. 6.2.2.2 Option 2: Develop Agency-Specific Guidelines for Minimum Traffic Volumes That Justify Shoulder Widening A less-desirable, but still acceptable, approach to cost-effective shoulder-widening deci- sions in 3R projects is to develop agency-specific guidelines for minimum traffic volumes that justify shoulder widening analogous to those shown in Table 44. Each highway agency (or, in some cases, individual districts or regions within a highway agency) can develop guide- lines for minimum traffic volumes as a basis for shoulder-widening decisions. Section 5.4 illustrates procedures for developing guidelines for minimum traffic volumes like those in Table 44. Benefit–cost analyses can also be applied to sites on rural multilane undivided highways with unpaved shoulders to consider whether the shoulder width should be fully or partially paved. The most desirable approach is to consider shoulder paving alternatives for each shoulder width alternative evaluated. A less desirable, but still acceptable, approach is to determine the optimal shoulder width for the project and then evaluate whether the shoulder should be paved. Many highway agencies have design policies that utilize paved shoulders only for specific roadway types or above specific traffic volume levels and otherwise use unpaved or composite shoulders. Nothing in these guidelines requires a highway agency to provide paved shoulders on projects where the agency’s design policy calls for unpaved or composite shoulders.

3R Project Design Guidelines for Specific Roadway Types 111   6.2.3 Horizontal Curve Improvements Horizontal curve improvements on rural multilane undivided highways may be considered 3R projects in some cases. Variances in superelevation of greater than 1% on horizontal curves on rural multilane undivided highways should be restored in 3R projects. Realignment of an isolated horizontal curve on a rural multilane undivided highway can also be considered as part of a 3R project. Typically, such an improvement would involve flattening the radius and, therefore, lengthening the curve. The spreadsheet tools address improvements of this type for rural two-lane highways (see Section 6.1.3), but not for rural multilane undivided highways, so their safety benefits would need to be assessed by other means. The HSM Chapter 11 procedures do not address the effects of horizontal curve radius and length for rural multilane highways, so some other assessment would need to be made. The HSM Chapter 10 procedures could provide a conservative approach to such an analysis. Realignment of multiple horizontal curves on a roadway section would generally be considered as reconstruction and, therefore, out of the scope of 3R improvements. Chevron markers can be installed at sharp horizontal curves as part of a 3R project to better delineate the curved path for drivers. There are no accepted CMFs for installation of chevron markers to improve delineation on a horizontal curve, so evaluation of chevron markers is not included in the spreadsheet tools. 6.2.4 Sight Distance Improvements SSD is provided along roadways to assist drivers in detecting and responding to situations on the roadway ahead that may require them to slow or stop. Sight distance limitations may include crest vertical curves or objects on the inside of horizontal curves. The SSD design criteria used by most states are based on the SSD values in the AASHTO Green Book (4). Recent research has shown that limited SSD on rural two-lane highways is unlikely to lead to crashes unless the portion of the roadway hidden from the driver’s view by the limitation in sight distance includes a roadway feature that may require drivers to take steering or braking action, such as an intersection, a driveway, or a horizontal curve (5). There has been no similar research on rural multilane undivided highways, but the research findings for rural two-lane highways can logically be extended to rural multilane highways. Where an area with SSD that is less than the Green Book SSD design criteria is present on a rural multilane undivided highway, but there are no hidden features such as intersections, driveways, or horizontal curves and no history of crashes related to limited SSD, sight distance improvements are unlikely to be cost-effective and need not be considered in 3R projects. In contrast, where a portion of the highway has SSD that is less than the Green Book SSD design criteria and an intersection, driveway, or horizontal curve is present in the area with limited SSD, or when the crash history shows a pattern of crashes potentially related to limited SSD, an SSD improvement is desirable. SSD improvements can include realigning a crest vertical curve or removing or relocating objects on the inside of a horizontal curve. Where the cost of the SSD improvement is substantial, consideration may be given to mitigation measures such as providing wider shoulders and/or advance warning signs. 6.2.5 Bridge Width Recent research has shown that narrow bridges on rural two-lane highways are not typically associated with increases in crash frequency (5). [Narrow bridges are defined as bridges where the curb-to-curb width of the bridge roadway is less than the width of the approach roadway

112 Guidelines for Integrating Safety and Cost-Effectiveness into Resurfacing, Restoration, and Rehabilitation (3R) Projects (lanes and shoulders combined).] Widening or replacing bridges on two-lane highways is not likely to result in crash reduction, even in the case of narrow bridges. There has been no similar research on rural multilane undivided highways, but the research findings for rural two-lane highways can logically be extended to rural multilane highways. Bridges on rural multilane undivided highways should remain in place in 3R projects unless there is either (a) a structural need to strengthen or replace the bridge or (b) a documented pattern of crashes at the bridge that can potentially be reduced by widening or replacing the bridge. 6.2.6 Normal Pavement Cross Slope Pavement cross slope is needed for drainage so that water flows off the pavement during and after precipitation. Each highway agency has its own design criteria for normal pavement cross slope; these criteria are typically selected as suitable for local climate conditions. Where the pavement cross section within a 3R project does not have sufficient cross slope for drainage, the pavement cross slope should be restored to meet the highway agency’s applicable design criteria. Given the importance of pavement cross slope to drainage, no benefit–cost analysis is needed to justify restoration of normal pavement cross slope in a 3R project on a rural multilane undivided highway. 6.2.7 Rumble Strip Improvements Centerline rumble strips are provided between the two directions of travel on rural multilane undivided highways to provide an audible and tactile warning to drivers when their vehicle leaves its intended travel lane and begins to cross the roadway centerline. Similarly, shoulder rumble strips are provided on a paved shoulder or at the edge of the traveled way on rural multi- lane undivided highways to provide an audible and tactile warning to drivers when their vehicle is leaving the roadway and begins to encroach on the shoulder. These warnings alert drivers to take corrective action—steering and, where appropriate, braking—to return to their intended travel lane. An example benefit–cost analysis has shown that installation of centerline or shoulder rumble strips in conjunction with a resurfacing project can become cost-effective on rural two-lane highways at AADTs as low as 300 veh/day (5). This AADT threshold is merely an example, and agency-specific minimum AADT guidelines can be developed with Spreadsheet Tool 1. However, there are essentially no rural multilane undivided highways with AADTs as low as 300 veh/day, so any agency-specific guidelines are likely to show that installation of centerline or shoulder rumble strips is cost-effective for all rural multilane undivided highways. Highway agencies are encouraged to provide centerline or shoulder rumble strips as part of 3R projects, where they are not already present, and to restore existing rumble strips when the pavement or shoulder is resurfaced. Rumble strips should be designed and located in accordance with each agency’s current practices. Rumble strips are appropriate in most rural locations, except where the noise created by the rumble strips may disturb nearby residents. Rumble strips on paved shoulders need to be located sufficiently close to the traveled way that the effectiveness of the paved shoulder as a travel path for bicyclists is not reduced. Where rumble strips are cost- effective, either centerline or shoulder rumble strips, or both, may be used. Use of both center- line and shoulder rumble strips together is very effective and should be considered, consistent with each agency’s current practices. 6.2.8 Striping and Delineation Improvements Nearly every 3R project includes pavement resurfacing, so the cost of restoring pavement markings after resurfacing is automatically part of most 3R projects. These guidelines assume

3R Project Design Guidelines for Specific Roadway Types 113   that, as a default, highway agencies will restore the pavement markings with the equivalent of the existing pavement markings. Conventional paint is the least-expensive pavement marking material and typically has a service life of 1 to 2 years, depending on traffic volume and climate conditions. Spreadsheet Tool 1 provides the capability to assess the cost-effectiveness of striping and delineation packages for inclusion in 3R projects. The primary element of a striping and delin- eation package that may be evaluated with the spreadsheet tool as part of a 3R project is the provision of more durable pavement markings with longer life and higher retroreflectivity than conventional paint markings. Striping and delineation packages may also include the addition of wider edgelines and post-mounted roadside delineators. The implementation cost for the striping and delineation package is the full striping and delineation cost minus the cost of the pavement markings that would have been implemented as a default. The crash reduction effectiveness of improved striping and delineation on rural multilane undivided highways is documented in Section 4.3.2.7. 6.2.9 Roadside Slope Flattening Roadside slope flattening should be considered for each 3R project on a rural multilane undivided highway with roadside slopes steeper than 1V:4H where sufficient right-of-way for slope flattening is available (or could be acquired) and where slope flattening would not adversely affect adjacent properties, structures, or environmentally sensitive areas, such as wetlands. Decisions about roadside slope flattening for 3R projects should be based on benefit– cost analysis, and roadside slope flattening is a desirable investment only when the analysis indicates that the expected crash reduction benefits from roadside slope flattening will exceed the costs of the improvement or when a crash analysis finds existing crash patterns (e.g., run-off- the-road crashes) that can potentially be reduced by roadside slope flattening. In many cases, roadside slope flattening will not be cost-effective because of high project implementation costs. Where there are no existing crash patterns that can potentially be reduced by roadside slope flattening and the expected crash reduction benefits are less than the costs of improvement, roadside slope flattening would be a poor investment, and available funds would be better invested at another location where the crash reduction benefits would be larger. Roadside slope flattening should be considered only where a benefit–cost analysis with the site-specific benefit–cost analysis tool (Spreadsheet Tool 1) or with the RSAP model (23–25) indicates that the present value of the benefits of the slope flattening project would exceed the cost. The crash reduction effectiveness of roadside slope flattening on rural multilane undivided highways is documented in Section 4.3.2.8. Procedures for site-specific benefit–cost analysis are illustrated in Section 5.2.2. The benefit–cost analysis tools provided with these guidelines enable site-specific benefit–cost analyses to be performed efficiently. There is no option for developing minimum AADT guidelines for roadside slope flattening projects because the costs of such projects can vary widely from site to site, so site-specific cost estimates are needed. Where roadside slope flattening is considered, but not implemented, in a 3R project, consideration may also be given to mitigation measures such as removing road- side objects (see Section 6.2.10) or providing traffic barriers (see Section 6.2.11), where such improvements are found to be cost-effective with the RSAP model. 6.2.10 Removal of Roadside Objects Where roadside objects are present within the clear zone width on a rural multilane undivided highway, their removal should be considered on the basis of a cost-effectiveness analysis with the RSAP model (23–25). Generally, only objects greater than 4 inches in diameter and not of

114 Guidelines for Integrating Safety and Cost-Effectiveness into Resurfacing, Restoration, and Rehabilitation (3R) Projects breakaway design need to be considered. Where appropriate, the design of culvert ends should be improved [see Chapter 3 of the AASHTO Roadside Design Guide (30)]. Where roadside objects are present continuously or at regular intervals throughout all or part of the length of a 3R project, removal of those objects is not likely to be cost-effective, and a formal RSAP analysis is not needed. As an alternative to removing roadside objects, consideration may also be given to replacing the object with a similar object of breakaway design, relocating the object behind an existing guardrail or traffic barrier, or installing a new guardrail or traffic barrier. Decisions concern- ing the cost-effectiveness of new guardrail or other traffic barriers should be made with the RSAP model (see Section 6.2.11). 6.2.11 Installation/Rehabilitation of Guardrail and Other Traffic Barriers Existing guardrail or other traffic barriers that have reached the end of their useful life or are of obsolete design should generally be rehabilitated or replaced as part of a 3R project. The decision to rehabilitate or replace a guardrail or traffic barrier does not require a formal economic analysis, as long as the highway agency believes that the guardrail or traffic barrier is needed and warranted at its location. If the highway agency has reason to believe that the existing guardrail or traffic barrier may not be warranted, the warrants and length of need for the barrier can be assessed with the AASHTO Roadside Design Guide (30). Installation of new guardrail or other traffic barriers may be considered by highway agencies as part of 3R projects on rural multilane undivided highways. The RSAP model should be used to evaluate the cost-effectiveness of any new guardrail or traffic barrier installations that are considered. 6.2.12 Intersection Turn Lane Improvements The crash reduction effectiveness of intersection turn lane improvements on rural multilane undivided highways is documented in Section 4.3.2.10. Benefit–cost analyses have shown that intersection turn-lane improvements, including installation of left- and right-turn lanes, are likely to be cost-effective at any intersection where installation of such improvements makes traffic operational sense (5). Highway agencies should, therefore, assess the traffic operational need for intersection turn lanes according to established access management policies or traffic operational analysis tools. If the highway agency concludes that installation of the turn lane is justified on a traffic operational basis, there is little doubt that the turn lane will be cost-effective on a safety basis as well. 6.2.13 Other Intersection Improvements A 3R project may provide an opportunity for implementing other intersection improvements involving traffic control, signage, delineation, marking, or sight distance. Highway agencies should implement such improvements if a need is identified on the basis of a crash history review. 6.3 Rural Multilane Divided Highways (Nonfreeways) 6.3.1 Lane Widening Lane widening should be considered for each 3R project on a rural multilane divided nonfreeway with existing lane widths of less than 12 ft. Decisions about lane widening for 3R projects should be based on benefit–cost analysis. Lane widening is a desirable investment

3R Project Design Guidelines for Specific Roadway Types 115   only when the analysis indicates that the expected crash reduction benefits will exceed the costs of the improvement or when a crash analysis finds existing crash patterns that can potentially be reduced by lane widening. Where there are no existing crash patterns that can potentially be reduced by lane widening and the expected crash reduction benefits are less than the costs of improvement, lane widening would be a poor investment, and available funds would be better invested at another location where the crash reduction benefits would be greater. Two approaches to benefit–cost analysis are applicable to lane-widening: • Option 1. Conduct a site-specific benefit–cost analysis for each individual 3R project. • Option 2. Develop agency-specific guidelines for minimum traffic volumes that justify lane widening. 6.3.1.1 Option 1: Conduct a Site-Specific Benefit–Cost Analysis for Each Individual 3R Project The most desirable approach to cost-effective decisions about lane-widening is to conduct a site-specific benefit–cost analysis for each individual 3R project. Site-specific benefit–cost analyses are desirable because even nominally similar sites may differ in roadway character- istics, traffic volumes, crash history, and improvement costs, such that lane widening may be cost-effective at one site and not at another. The crash reduction effectiveness of lane widening on rural multilane divided nonfreeways is documented in Section 4.3.3.1. Procedures for site-specific benefit–cost analysis are illustrated in Section 5.2.2. The benefit–cost analysis tools provided with these guidelines enable site-specific benefit–cost analyses to be performed efficiently. An example of the application of Spreadsheet Tool 1 to a lane-widening improvement on a rural two-lane highway is presented in Section 6.1.1. Analysis of lane widening for rural multi- lane divided nonfreeways can be performed in the same manner, with the only differences in input data being the roadway type and the addition of a value for the existing median width. 6.3.1.2 Option 2: Develop Agency-Specific Guidelines for Minimum Traffic Volumes That Justify Lane Widening A less-desirable, but still acceptable, approach to cost-effective lane-widening decisions in 3R projects is to develop agency-specific guidelines for minimum traffic volumes that justify lane widening analogous to those shown in Table 43. Each highway agency (or, in some cases, individual districts or regions within a highway agency) can develop guidelines for minimum traffic volumes as a basis for lane-widening decisions. Section 5.4 illustrates procedures for developing guidelines for minimum traffic volumes like those in Table 43. An example benefit–cost analysis to develop minimum AADT guidelines for rural two-lane highways is presented in Table 43. The same analysis approach can be applied to lane widening on rural multilane divided highways. 6.3.2 Shoulder Widening Shoulder widening should be considered for each 3R project on a rural multilane divided nonfreeway with existing shoulder widths of less than 8 ft. Decisions about shoulder widen- ing for 3R projects should be based on benefit–cost analysis. Shoulder widening is a desirable investment only when the analysis indicates that the expected crash reduction benefits will exceed the improvement costs or when a crash analysis finds existing crash patterns that can potentially be reduced by shoulder widening. Where there are no existing crash patterns that can potentially be reduced by shoulder widening and the expected crash reduction benefits are less than the costs of improvement, shoulder widening would be a poor investment, and

116 Guidelines for Integrating Safety and Cost-Effectiveness into Resurfacing, Restoration, and Rehabilitation (3R) Projects available funds would be better invested at another location where the crash reduction benefits would be larger. Two approaches to benefit–cost analysis are applicable to shoulder widening: • Option 1. Conduct a site-specific benefit–cost analysis for each individual 3R project. • Option 2. Develop agency-specific guidelines for minimum traffic volumes that justify shoulder widening. 6.3.2.1 Option 1: Conduct a Site-Specific Benefit–Cost Analysis for Each Individual 3R Project The most desirable approach to cost-effective decisions about shoulder widening is to conduct a site-specific benefit–cost analysis for each individual 3R project. Site-specific benefit– cost analyses are desirable because even nominally similar sites may differ in roadway character- istics, traffic volumes, crash history, and improvement costs, such that shoulder widening may be cost-effective at one site and not at another. The crash reduction effectiveness of shoulder widening on rural multilane divided nonfreeways is documented in Section 4.3.3.2. Procedures for site-specific benefit–cost analysis are illustrated in Section 5.2.2. The benefit–cost analysis tools provided with these guidelines enable site-specific benefit–cost analyses to be performed efficiently. An example of the application of Spreadsheet Tool 1 to a shoulder-widening improvement on a rural two-lane highway is presented in Section 6.1.2. Analysis of shoulder widening for rural multilane divided nonfreeways can be performed in the same manner, with the only difference in input data being the roadway type. 6.3.2.2 Option 2: Develop Agency-Specific Guidelines for Minimum Traffic Volumes That Justify Shoulder Widening A less-desirable, but still acceptable, approach to cost-effective shoulder-widening decisions in 3R projects is to develop agency-specific guidelines for minimum traffic volumes that justify shoulder widening analogous to those shown in Table 44. Each highway agency (or, in some cases, individual districts or regions within a highway agency) can develop guidelines for minimum traffic volumes as a basis for shoulder-widening decisions. Section 5.4 illustrates procedures for developing guidelines for minimum traffic volumes like those in Table 44. 6.3.3 Horizontal Curve Improvements Realignment of horizontal curves on divided highways, including rural multilane divided highways, is sufficiently complex that such projects are considered reconstruction and are, therefore, outside the scope of a 3R project as defined in these guidelines. Variances in super- elevation of greater than 1% on horizontal curves on rural multilane divided highways should be restored in 3R projects. Spreadsheet Tool 1 can be used to evaluate the cost-effectiveness of such improvements. Chevron markers can be installed at sharp horizontal curves as part of a 3R project to better delineate the curved path for drivers. There are no accepted CMFs for installation of chevron markers to improve delineation on a horizontal curve, so evaluation of chevron markers is not included in the spreadsheet tools. 6.3.4 Sight Distance Improvements SSD is provided along roadways to assist drivers in detecting and responding to situations on the roadway ahead that may require them to slow or stop. Sight distance limitations may include

3R Project Design Guidelines for Specific Roadway Types 117   crest vertical curves or objects on the inside of horizontal curves. The SSD design criteria used by most states are based on the SSD values in the AASHTO Green Book (4). Recent research has shown that limited SSD on rural two-lane highways is unlikely to lead to crashes unless the portion of the roadway hidden from the driver’s view by the limitation in sight distance includes a roadway feature that may require drivers to take steering or braking action, such as an intersection, a driveway, or a horizontal curve (5). There has been no similar research on rural multilane divided nonfreeways, but the research findings for rural two-lane highways can logically be extended to rural multilane highways. Where an area with SSD less than the Green Book SSD design criteria is present on a rural multilane divided nonfreeway but there are no hidden features such as intersections, driveways, ramp terminals, or horizontal curves and no history of crashes related to limited SSD, sight distance improvements are unlikely to be cost-effective and need not be considered in 3R projects. In contrast, when a portion of the highway has SSD less than the Green Book SSD design criteria and an intersection, driveway, or horizontal curve is present in the area with limited SSD, or when the crash history shows a pattern of crashes potentially related to limited SSD, an SSD improvement is desirable. SSD improvements can include realigning a crest vertical curve or removing or relocating objects on the inside of a horizontal curve. Where the cost of the SSD improvement is substantial, consideration may be given to mitigation measures such as providing wider shoulders and/or advance warning signs. 6.3.5 Bridge Width Recent research has shown that narrow bridges on rural two-lane highways are not typically associated with increases in crash frequency (5). [Narrow bridges are defined as bridges where the curb-to-curb width of the bridge roadway is less than the width of the approach roadway (lanes and shoulders combined).] Widening or replacing bridges on two-lane highways is not likely to result in crash reduction, even in the case of narrow bridges. There has been no similar research on rural multilane divided nonfreeways, but the research findings for rural two-lane highways can logically be extended to rural multilane nonfreeways. Bridges on rural multilane divided nonfreeways should remain in place in 3R projects unless there is either (a) a structural need to strengthen or replace the bridge or (b) a documented pattern of crashes at the bridge that can potentially be reduced by widening or replacing the bridge. 6.3.6 Normal Pavement Cross Slope Pavement cross slope is needed for drainage so that water flows off the pavement during and after precipitation. Each highway agency has its own design criteria for normal pavement cross slope; these criteria are typically selected as suitable for local climate conditions. Where the pavement cross section within a 3R project does not have sufficient cross slope for drainage, the pavement cross slope should be restored to meet the highway agency’s applicable design criteria. Given the importance of pavement cross slope to drainage, no benefit–cost analysis is needed to justify restoration of normal pavement cross slope in a 3R project on a rural multilane divided nonfreeway. 6.3.7 Rumble Strip Improvements Shoulder rumble strips are provided on a paved shoulder or at the edge of the traveled way on rural multilane divided nonfreeways to provide an audible and tactile warning to drivers when their vehicle is leaving the roadway and begins to encroach on the shoulder. These warnings alert drivers to take corrective action—steering and, where appropriate, braking—to return to their

118 Guidelines for Integrating Safety and Cost-Effectiveness into Resurfacing, Restoration, and Rehabilitation (3R) Projects intended travel lane. Shoulder rumble strips are typically used on both the right (outside) and left (inside) shoulders of rural multilane divided nonfreeways. An example benefit–cost analysis has shown that installation of shoulder rumble strips in conjunction with a resurfacing project can become cost-effective on rural multilane divided nonfreeways at AADTs as low as 400 veh/day (7). This AADT threshold is merely an example, and agency-specific minimum AADT guidelines can be developed with Spreadsheet Tool 1. However, there are essentially no rural multilane divided nonfreeways with AADTs as low as 400 veh/day, so any agency-specific guidelines are likely to show that installation of shoulder rumble strips is cost-effective for all rural multilane divided nonfreeways. Highway agencies are encouraged to provide shoulder rumble strips as part of 3R projects, where they are not already present, and to restore existing rumble strips when the pavement or shoulder is resurfaced. Rumble strips should be designed and located in accordance with each agency’s current practices. Rumble strips are appropriate at most rural locations, except where the noise created by the rumble strips may disturb nearby residents. Rumble strips on paved shoulders need to be located sufficiently close to the traveled way that the effectiveness of the paved shoulder as a travel path for bicyclists is not reduced. 6.3.8 Striping and Delineation Improvements Nearly every 3R project includes pavement resurfacing, so the cost of restoring pavement markings after resurfacing is automatically part of most 3R projects. These guidelines assume that, as a default, highway agencies will restore the pavement markings with the equivalent of the existing pavement markings. Conventional paint is the least-expensive pavement marking material and typically has a service life of 1 to 2 years, depending on traffic volume and climate conditions. Spreadsheet Tool 1 provides the capability to assess the cost-effectiveness of striping and delineation packages for inclusion in 3R projects. The primary element of a striping and delin- eation package that may be evaluated with the spreadsheet tool as part of a 3R project is the provision of more durable pavement markings with longer life and higher retroreflectivity than conventional paint markings. Striping and delineation packages may also include the addition of wider edgelines and post-mounted roadside delineators. The implementation cost for the striping and delineation package is the full cost of striping and delineation minus the cost of the pavement markings that would have been implemented as a default. The crash reduction effectiveness of improved striping and delineation on rural multilane nonfreeways is docu- mented in Section 4.3.3.6. 6.3.9 Roadside Slope Flattening Roadside slope flattening should be considered for each 3R project on a rural multilane divided nonfreeway with roadside slopes steeper than 1V:4H where sufficient right-of-way for slope flattening is available (or could be acquired) and where slope flattening would not adversely affect adjacent properties, structures, or environmentally sensitive areas such as wet- lands. Decisions about roadside slope flattening for 3R projects should be based on benefit–cost analysis. Roadside slope flattening is a desirable investment only when the analysis indicates that the expected crash reduction benefits will exceed the costs of the improvement or when a crash analysis finds existing crash patterns (e.g., run-off-the-road crashes) that can potentially be reduced by roadside slope flattening. In many cases, roadside slope flattening will not be cost-effective because of high project implementation costs. Where there are no existing crash patterns that can potentially be reduced by roadside slope flattening and the expected crash

3R Project Design Guidelines for Specific Roadway Types 119   reduction benefits are less than the costs of improvement, roadside slope flattening would be a poor investment, and available funds would be better invested at another location where the crash reduction benefits would be larger. Roadside slope flattening should be considered only where a benefit–cost analysis with the site-specific benefit–cost analysis tool (Spreadsheet Tool 1) or with the RSAP model (23–25) indicates that the present value of the benefits of the slope flattening project would exceed the cost. The crash reduction effectiveness of roadside slope flattening is documented for rural multilane undivided highways in Section 4.3.2.8 and for multilane divided nonfreeways in Section 4.3.3.7. Procedures for site-specific benefit–cost analysis are illustrated in Section 5.2.2. The benefit–cost analysis tools provided with these guidelines enable site-specific benefit–cost analyses to be performed efficiently. There is no option for developing minimum AADT guidelines for roadside slope flattening projects because the costs of such projects can vary widely from site to site, so site-specific cost estimates are needed. Where roadside slope flattening is considered, but not implemented, in a 3R project, consideration may also be given to mitigation measures such as removing roadside objects (see Section 6.3.10) or providing traffic barriers (see Section 6.3.11), where such improvements are found to be cost-effective with the RSAP model. 6.3.10 Removal of Roadside Objects Where roadside objects are present within the clear zone width on a rural multilane divided nonfreeway, their removal should be considered on the basis of a cost-effectiveness analysis with the RSAP model (23–25). Generally, only objects greater than 4 inches in diameter and not of breakaway design need to be considered. Where appropriate, the design of culvert ends should be improved [see Chapter 3 of the AASHTO Roadside Design Guide (30)]. Where roadside objects are present continuously or at regular intervals throughout all or part of the length of a 3R project, removal of those objects is not likely to be cost-effective and a formal RSAP analysis is not needed. As an alternative to removing roadside objects, consideration may also be given to replacing the object with a similar object of breakaway design, relocating the object behind an existing guardrail or traffic barrier, or installing a new guardrail or traffic barrier. Decisions concerning the cost-effectiveness of new guardrail or other traffic barriers should be made with the RSAP model (see Section 6.3.11). 6.3.11 Installation/Rehabilitation of Guardrail and Other Traffic Barriers Existing guardrail or other traffic barriers that have reached the end of their useful life or are of obsolete design should generally be rehabilitated or replaced as part of a 3R project. The decision to rehabilitate or replace a guardrail or traffic barrier does not require a formal economic analysis, as long as the highway agency believes that the guardrail or traffic barrier is needed and warranted at its location. If the highway agency has reason to believe that the existing guardrail or traffic barrier may not be warranted, the warrants and length of need for the barrier can be assessed with the AASHTO Roadside Design Guide (30). Installation of new guardrail or other traffic barriers may be considered by highway agencies as part of 3R projects on rural multilane divided nonfreeways. The RSAP model should be used to evaluate the cost-effectiveness of any new guardrail or traffic barrier installations that are considered.

120 Guidelines for Integrating Safety and Cost-Effectiveness into Resurfacing, Restoration, and Rehabilitation (3R) Projects 6.3.12 Intersection Turn Lane Improvements The crash reduction effectiveness of intersection turn lane improvements on rural multilane divided nonfreeways is documented in Section 4.3.3.9. Benefit–cost analyses have shown that intersection turn-lane improvements, including installation of left- and right-turn lanes, are likely to be cost-effective at any intersection where installation of the turn lane makes traffic operational sense (5). Highway agencies should, therefore, assess the traffic operational need for intersection turn lanes according to established access management policies or traffic operational analysis tools. If the highway agency concludes that installation of the turn lane is justified on a traffic operational basis, there is little doubt that the turn lane will be cost-effective on a safety basis as well. 6.3.13 Other Intersection Improvements A 3R project may provide an opportunity for implementing other intersection improvements involving traffic control, signage, delineation, marking, or sight distance. Highway agencies should implement such improvements if a need is identified on the basis of a crash history review. 6.4 Urban and Suburban Arterials 6.4.1 Lane Widening Benefit–cost analyses cannot be applied to lane widening on urban and suburban arterials at this time, because there are no documented CMFs for lane widening in HSM Chapter 12. However, the AASHTO Green Book provides broad flexibility for use of 10-, 11-, and 12-ft lanes on urban and suburban arterials (4). Recent research also suggests that, in most cases, there are no substantial differences in safety performance between 10-, 11-, and 12-ft lanes on urban and suburban arterials (see Section 4.3.4.1). In addition, narrower through travel lanes can have substantive advantages on urban and suburban arterials by providing space in the cross section for turn lanes, median treatments, bicycle lanes, and shorter pedestrian crossings, all of which can themselves reduce crashes. Therefore, lane widening is not a desirable investment in 3R projects on urban and suburban arterials with existing lane widths of 10 ft or more unless there is a documented crash pattern that can potentially be mitigated with wider lanes or there is a documented traffic operational need for wider lanes. 6.4.2 Shoulder Widening Benefit–cost analyses cannot be applied to shoulder widening on urban and suburban arterials at this time, because there are no documented CMFs for shoulder widening in HSM Chapter 12. However, the AASHTO Green Book provides broad flexibility for use of shoulders with a range of widths on urban and suburban arterials, including, in appropriate speed ranges, the use of curb-and-gutter sections with no shoulder (4). For many low- and intermediate-speed roads, curb-and-gutter sections without shoulders are appropriate because they enhance drainage. Therefore, shoulder widening is not a desirable investment in 3R projects on urban and sub- urban arterials unless there is a documented crash pattern that can potentially be mitigated with wider shoulders or there is a documented traffic operational need for wider shoulders. 6.4.3 Horizontal Curve Improvements Realignment of horizontal curves on urban and suburban arterials is sufficiently complex that such projects are considered reconstruction and are, therefore, outside the scope of a 3R project as defined in these guidelines.

3R Project Design Guidelines for Specific Roadway Types 121   Where the superelevation of an existing curve is less than the design superelevation value in the AASHTO Green Book (4) by more than 1%, restoration of superelevation to the Green Book value should be considered as part of any 3R project conducted. There is no formal benefit–cost tool for superelevation improvement in this situation, so decisions should be based on engi- neering judgment. Chevron markers can be installed at sharp horizontal curves as part of a 3R project to better delineate the curved path for drivers. There are no accepted CMFs for installation of chevron markers to improve delineation on a horizontal curve, so evaluation of chevron markers is not included in the spreadsheet tools. 6.4.4 Striping and Delineation Improvements Nearly every 3R project includes pavement resurfacing, so the cost of restoring pavement markings after resurfacing is automatically part of most 3R projects. These guidelines assume that, as a default, highway agencies will restore the pavement markings with the equivalent of the existing pavement markings. Conventional paint is the least-expensive pavement marking material and typically has a service life of 1 to 2 years, depending on traffic volume and climate conditions. Highway agencies may choose to consider more durable pavement markings with longer life and higher retroreflectivity, as well as an increased implementation cost, as part of some 3R projects. Other delineation improvements may also be considered as part of an overall striping and delineation package. There are no documented measures of the crash reduction effectiveness of striping and delineation packages on urban and suburban arterials, so highway agencies should exercise judgement based on their experience with striping and delineation packages on other roadway types to assess when striping and delineation improvements are appropriate on urban and suburban arterials. 6.4.5 Rumble Strip Improvements Centerline and shoulder rumble strips are effective in reducing crashes on urban and sub- urban arterials, although there are no documented estimates of the crash reduction effectiveness of rumble strips in HSM Chapter 12. The literature includes the crash reduction effectiveness of one application on urban and suburban arterials—centerline rumble strips on urban two-lane arterials (see Section 4.3.4.5). Nevertheless, it is expected that most urban and suburban arterials have traffic volumes sufficiently high to make rumble strips cost-effective for use on urban and suburban arterials. Noise that can potentially disturb nearby residents is an elevated concern in urban and suburban areas. Highway agencies should exercise judgment based on current agency policies about locations where installation of centerline and shoulder rumble strips is or is not desirable on urban and suburban arterials. 6.4.6 Removal of Roadside Objects Where roadside objects are present within the clear zone width on an urban or suburban arterial, their removal may be considered on the basis of a cost-effectiveness analysis with the RSAP model (23–25). Generally, only objects greater than 4 inches in diameter and not of breakaway design are considered. Where appropriate, the design of culvert ends should be improved [see Chapter 3 of the AASHTO Roadside Design Guide (30)]. Where roadside objects are present continuously or at regular intervals throughout all or part of the length of a 3R project, as is the case on many urban and suburban arterials, removal of those objects is not likely to be cost-effective and a formal RSAP analysis is not needed.

122 Guidelines for Integrating Safety and Cost-Effectiveness into Resurfacing, Restoration, and Rehabilitation (3R) Projects As an alternative to removing roadside objects, consideration may also be given to replacing the object with a similar object of breakaway design, relocating the object behind an existing guardrail or traffic barrier, or installing a new guardrail or traffic barrier. Decisions concerning the cost-effectiveness of new guardrail or other traffic barriers should be made with the RSAP model (see Section 6.4.7). 6.4.7 Installation/Rehabilitation of Guardrail and Other Traffic Barriers Existing guardrail or other traffic barriers that have reached the end of their useful life or are of obsolete design should generally be rehabilitated or replaced as part of a 3R project. The decision to rehabilitate or replace a guardrail or traffic barrier does not require a formal economic analysis, as long as the highway agency believes that the guardrail or traffic barrier is needed and warranted at its location. If the highway agency has reason to believe that the existing guardrail or traffic barrier may not be warranted, the warrants and length of need for the barrier can be assessed with the AASHTO Roadside Design Guide (30). Installation of new guardrail or other traffic barriers may be considered by highway agencies as part of 3R projects on urban and suburban arterials. The RSAP model should be used to evaluate the cost-effectiveness of any new guardrail or traffic barrier installations that are considered. 6.4.8 Intersection Turn Lane Improvements Benefit–cost analyses have shown that intersection turn-lane improvements, including installation of left- and right-turn lanes, are likely to be cost-effective at any intersection where installation of the turn lane makes traffic operational sense (5). Highway agencies should, there- fore, assess the traffic operational need for intersection turn lanes by using established access management policies or traffic operational analysis tools. If the highway agency concludes that installation of the turn lane is justified on a traffic operational basis, there is little doubt that the turn lane will be cost-effective on a safety basis as well. 6.4.9 Other Intersection Improvements A 3R project may provide an opportunity for implementing other intersection improvements involving traffic control, signage, delineation, marking, or sight distance. Highway agencies should implement such improvements if a need is identified on the basis of a crash history review. 6.5 Rural and Urban Freeways 6.5.1 Lane Widening Most urban and rural freeways have existing 12-ft lanes, except at locations where a high- way agency has made a previous decision to use narrower lanes to provide space for additional through lanes. At locations that were built with lanes less than 12 ft in width, Spreadsheet Tool 1 can be used for benefit–cost evaluation of lane-widening alternatives. In such cases, widening to 12-ft lanes is likely to be cost-effective over the typical range of freeway volumes. Site-specific analyses with Tool 1 are suggested, however, to justify such projects. 6.5.2 Outside Shoulder Widening Existing freeways may have a range of outside shoulder widths. Most existing freeways have outside shoulder widths of at least 10 ft, although a few sites may have narrower shoulders.

3R Project Design Guidelines for Specific Roadway Types 123   At locations that were built with outside shoulders less than 12 ft in width, Spreadsheet Tool 1 can be used for benefit–cost evaluation of shoulder-widening alternatives. In such cases, widening to 12-ft shoulders is likely to be cost-effective over the typical range of freeway volumes. Site-specific analyses with Tool 1 are suggested, however, to justify such projects. 6.5.3 Inside Shoulder Widening Existing freeways may have a range of inside (median) shoulder widths. Most existing freeways have inside shoulder widths of at least 4 ft, although a few sites may have narrower shoulders. At locations that were built with inside shoulders less than 12 ft in width, Spreadsheet Tool 1 can be used for benefit–cost evaluation of shoulder-widening alternatives. In such cases, widening to 12-ft shoulders is likely to be cost-effective over the typical range of freeway volumes. Site- specific analyses with Tool 1 are suggested, however, to justify such projects. Many highway agencies prefer to limit inside shoulder widths to a maximum of 4 ft to encourage motorists to stop on the outside shoulder, rather than the inside shoulder, when the need to stop arises. Nothing in these guidelines suggests that highway agencies should change such policies if highway agency experience indicates that vehicles stopping on the inside shoulder are undesirable. 6.5.4 Installation of Median Barriers While Spreadsheet Tools 1 and 2 can be used to assess the cost-effectiveness of installing median barriers, these spreadsheet tools for 3R projects based on HSM Chapter 18 (3) are not well suited to assessing the shifts in the distribution of crash severity that may result from installing median barriers. Research generally indicates that installation of median barriers reduces the frequency of fatal and serious injury crashes but may increase the frequency of less- severe crashes. In general, the AASHTO Roadside Design Guide (30) is better suited to assessing the need for median barriers than the predictive models in HSM Chapter 18. Therefore, the Roadside Design Guide is preferred to Spreadsheet Tools 1 and 2 for such analyses. 6.5.5 Installation/Restoration of Guardrail or Other Barrier Types on the Roadside Outside the Traveled Way While Spreadsheet Tools 1 and 2 can be used to assess the cost-effectiveness of installing guardrail or other roadside barriers, these spreadsheet tools for 3R projects based on HSM Chapter 18 (3) are not well suited to assessing the shifts in the distribution of crash severity that may result from installing guardrail and other roadside barriers. Research generally indicates that barrier installation reduces fatal and serious injury crashes but may increase less-severe crashes. In general, the AASHTO Roadside Design Guide (30) is better suited to assessing the need for guardrail or other roadside barriers than the predictive models in HSM Chapter 18. Therefore, the Roadside Design Guide is preferred to Spreadsheet Tools 1 and 2 for such analyses. 6.5.6 Shoulder Rumble Strips Spreadsheet Tools 1 and 2 are well suited to estimating the anticipated cost-effectiveness of installing centerline and shoulder rumble strips. Calculations with these spreadsheet tools indicate that shoulder rumble strips on either or both sides of the roadway are generally cost- effective on rural and urban freeways over the full range of traffic volume levels typically present on those freeways.

124 Guidelines for Integrating Safety and Cost-Effectiveness into Resurfacing, Restoration, and Rehabilitation (3R) Projects 6.5.7 Horizontal Curve Improvements Where the superelevation of an existing curve is less than the design superelevation value in the AASHTO Green Book (4) by more than 1%, restoration of superelevation to the Green Book value should be considered as part of any 3R project. There is no formal benefit–cost tool for superelevation improvement in this situation, so decisions should be based on engineering judgment. Realignment of one or more horizontal curves on a freeway is sufficiently complex that such projects are considered reconstruction and are, therefore, outside the scope of 3R work as defined in these guidelines. Chevron markers can be installed at sharp horizontal curves as part of a 3R project to better delineate the curved path for drivers. There are no accepted CMFs for installation of chevron markers to improve delineation on a horizontal curve, so evaluation of chevron markers is not included in the spreadsheet tools. 6.5.8 Sight Distance Improvements As on rural two-lane highways, SSD improvements on freeways are only likely to be cost- effective where the sight distance limitation hides key roadway features from the driver’s view. On freeways, the roadway features whose presence may indicate the need for a sight distance improvement or mitigation measures include horizontal curves, ramp terminals, and locations where standing queues are frequently present (e.g., on a daily basis). Where such features are present in an area with limited sight distance, alignment improvements to provide more sight distance should be considered. However, projects to realign freeways are generally considered reconstruction rather than 3R work. 6.5.9 Normal Pavement Cross Slope Pavement cross slope is needed for drainage so that water flows off the pavement during and after precipitation. Each highway agency has its own design criteria for normal pavement cross slope; these criteria are typically selected as suitable for local climate conditions. Where the pavement cross section within a 3R project does not have sufficient cross slope for drainage, the pavement cross slope should be restored to meet the highway agency’s applicable design criteria. Given the importance of pavement cross slope to drainage, no benefit–cost analysis is needed to justify restoration of normal pavement cross slope in a 3R project on a freeway.