A Performance-Based Highway Geometric Design Process (2016)

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.

154 7.1 Introduction A revised design process that is focused on problem solving using objective performance- based models and methods, and which recognizes inherent differences in project types and con- texts, offers considerable benefits. The value and benefits of a performance-based process would hence appear to be as follows: • There would be greater assurance that investments in infrastructure solutions requiring geo- metric design would produce actual performance enhancements commensurate with the implementation costs, such costs reflecting differences in the project context. Such assurance is particularly evident in the case of reconstruction projects, which form the increasing bulk of DOT project types. For such projects the revised process directly employs actual project data on operations and safety performance, as well as consensus science-based approaches to predicting future performance based on current knowledge and site-specific data. • Conversely, the process applied properly would tend to preclude investments in geometric design solutions for which there is no evidence of performance improvements, thus produc- ing savings to the owning agency. Geometric design as the means to an end (and not the end itself) is thus entirely focused on the specific nature and character of the site-specific problem. The more cost-effective solutions should maximize safety and capacity benefits for limited funding resources. • The process and design criteria employed, being based on research and data, would be self- reinforcing over time, thus ensuring the continued relevance and cost-effectiveness at a sys- tem level. • Agencies would be able to control the application of the process to their projects at both the project and program levels to reflect the current state of resources and competing priorities. Such control would occur through policy decisions on problem or needs definition and threshold cost-effectiveness ratios, all of which may be adjusted based on resources avail- able to the agency. • The process if applied in its purest form would end the current uncomfortable and unpro- ductive process of labeling a preferred solution as being or requiring a “design exception.” Rather, it would stress the importance of arriving at an optimal solution given exceptional circumstances or context. This would not only reduce unnecessary bureaucracy, but would enhance public acceptance of agency actions and decisions. • The process should reduce agency risk. The design exceptions process is inherently defensive in nature and in many cases is a hindrance to arriving at the optimal solution. Design risk that exists today can occur if or when a design exception is missed. Maintenance of good records and decision documents will be more important with the proposed process. A focus on the documentation in a positive manner of the reasons behind the design decision (including calculations of relevant safety and operational elements) to assure they are done properly is a Value and Benefits of Improved Process C h a p t e r 7

Value and Benefits of Improved process 155 positive element of the process and should be the focus of the agency rather than the exercise of explaining why something was NOT done. • The process engenders trust within the community that the owner has engaged them, has been transparent, thorough, and thus developed the best value solution. This will help agencies advance projects to implementation with less community resistance. Agencies should expect that the process would, over time, produce higher quality projects (as measured by their resultant performance outcomes) at lower overall aggregate costs. Agencies should expect to have to address important challenges. These are explained in detail in Chapter 8. However, one important challenge is worthy of note here when considering the above benefits to be expected. The process if applied properly treats every geometric design project as unique. As such it requires the full attention and independent thought process for each project by the project’s engineering team. Agency staff can no longer simply pull dimensions or values from tables or charts. The geometric design process, in becoming more performance-based in its approach, requires the application of more thorough thought and analytical approaches. The above represent cultural (and to an extent, educational) challenges to agencies and their staffs. They should not increase project time, or even project development costs; but they will require different skill sets and reallocation of agency resources to project development. In the end, the revised process should produce a more thoroughly professional and defensible outcome. 7.2 Case Studies In an effort to illustrate the application of the new process and to demonstrate the outcomes and benefits, the research team developed a number of “case studies.” These illustrative case studies are hypothetical data, however, they are the typical design scenarios that the designers face. 7.2.1 Case Study 1: Addition of a Bicycle Lane on an Urban Arterial Case Study 1 addresses an existing urban arterial street with one through lane in each direction of travel, a center two-way left-turn lane, and on-street parking. The proposed project involves removing the center two-way left-turn lane and providing a bicycle lane in each direction of travel. The traffic operational and safety performance of Alternative 1 (the existing condition) and Alternative 2 (the proposed future condition) are compared. Figure 50 shows typical cross sections for the existing and proposed conditions of a 2-mile tangent segment of this urban arterial with an AADT of 17,800 vehicles per day (vpd) for the design year with the same total cross-section dimension for both alternatives. For Alternative 1 (existing condition), a three-lane section with a center two-way left-turn lane is present, and a parking lane is provided on both sides of the road. Each travel lane is 12 feet wide, and the park- ing lanes are each 10 feet wide. For Alternative 2 (future condition), one through travel lane in each direction of travel is still provided, and bicycle lanes are also provided on each side of the road between the through travel lanes and the parking lanes. The through travel lanes are 11 feet wide, the bicycle lanes are 5 feet wide, and the parking lanes are 7 feet wide. The bicycle lanes are separated from the through travel lanes by a 3-foot buffer and from the parking lanes by a 2-foot buffer. The 2-mile urban arterial segment includes two signalized intersections at the endpoints of the segment, and three unsignalized (two-way stop-controlled) intersections within the segment, with a 0.5-mile spacing between intersections. The bicycle lanes will merge with the through

156 a performance-Based highway Geometric Design process travel lanes at the intersection approaches to allow for the provision of left-turn lanes. Therefore, the intersection design is the same for both existing and future conditions. The left-turn lanes at both signalized intersections have protected-permitted operation. The signals are uncoordinated; as a result, the optimal signal splits for each intersection are deter- mined separately. Signalized pedestrian crossings are present on all four approaches of the sig- nalized intersections. All of these intersection features apply to both the existing and proposed future conditions. The side-street geometrics at both signalized intersections are identical and will not change as a result of the proposed project. The total width of the shared through/right-turn lane on the main street in the proposed future condition is 12 feet in the existing condition and 11 feet in the proposed future condition. However, for capacity calculations, the lane width was assumed to be 16 feet with a 6-foot paved shoulder. This accounts for the width of the parking lane, which does not extend up to the intersection, as well as the merging of the bicycle lane into the through travel lanes prior to the intersection. These intersection cross sections are assumed to remain the same in both existing and proposed future conditions. Table 26 shows the traffic operational comparison between the existing and proposed future alternatives to the extent that it can be assessed with current HCM procedures. Table 27 shows the traffic safety comparison between the existing and proposed future alterna- tives to the extent that it can be assessed with current HSM procedures. The case study illustrates that there are substantial limitations in the ability of the current HCM and HSM procedures in assessing these alternatives. In lieu of the availability of the standard analyses procedures, other procedures including simulation, other research (NCHRP, NACTO, other), or engineering judg- ment could be considered for evaluation. ADT = 17,800 vpd Length = 2 miles Directional Distribution = 54/46 Trucks = 2% K = 10% DHV = 1,780 vph PHF = .90 Figure 50. Cross section without bike lanes (top) and cross section with bike lanes (bottom) (ADT = average daily traffic, K = percent of daily traffic in the DHV, PHF = peak-hour factor).

Value and Benefits of Improved process 157 The LOS reported in Table 26 is the facility LOS. Note that the NB approach at the south inter- section and SB approach at the north intersection experience spillback due to volume exceeding capacity. This causes the facility travel time to increase and facility speed to decrease for both alternatives. Traffic Operational Analysis Results Table 27 shows a comparison of the traffic operational performance or throughput for these designs, based on HCM methods and assumptions. The facility travel times and speeds are virtu- ally identical between the existing and proposed future conditions and the levels of service are identical. The small increases in facility travel time are attributable to the decrease in width of the through travel lanes from 12 to 11 feet. The traffic operational analysis procedures for urban streets in the HCM do not address mid-block delays due to removal of the center two-way left- turn lane. The HCM procedures permit analysts to incorporate mid-block delays manually, but only if estimates are derived from external sources, such as field studies. The HCM procedures also do not quantify the reduction in delay to motor vehicles by removing bicycles from the through travel lanes or the resulting improvement in traffic operational conditions for the bicyclists. Traffic Safety Analysis Results Table 28 shows a comparison of the predicted safety performance (annual crashes) for the roadway segments between intersections for the design year, 2035. The following inputs were used in the HSM roadway segment analysis: Note that all inputs to the segment worksheet are identical for both alternatives, except that the roadway type changes from three-lane with a center two-way left-turn lane (3T) to two- lane undivided (2U). The small increase in predicted crashes for the proposed future condi- tion shown in Table 27 results from the change in roadway type from 3T to 2U and the 1-foot decrease in through travel lane width. The HSM does not include procedures to quantify the effect on crash frequency due to the addition of the bicycle lanes, the narrowing of the parking lanes, or the addition of the buffer areas. The Intersection Worksheet for HSM Chapter 12 was not used because the intersection design did not change between the existing condition to the proposed future condition, and therefore the predicted crash frequencies at each intersection would remain unchanged. Alternative Predicted Crashes per mile per year Total Fatal and Injury PDO 1 9.5 2.7 6.8 2 9.8 2.8 7.0 Predicted crashes were determined using HSM 1st Edition, Chapter 12 Worksheets (uncalibrated model without crash history data). Note: PDO = property damage only. Table 27. Case Study 1, crash study results. Alternative Traffic Operational Analysis Results Facility LOS Facility Travel Time (s) Facility Travel Speed (mph) 1 NB C 267.02 26.96 SB B 225.41 31.94 2 NB C 267.03 26.96 SB B 225.43 31.94 LOS was determined using Highway Capacity Software (HCS) 2010 Streets Version 6.30. Note: A = incapacitating injury; B = non-incapacitating injury; C = reported injury. Table 26. Case Study 1, traffic analyses results.

158 a performance-Based highway Geometric Design process Conclusions The results of Case Study 1 provide an incomplete picture of the traffic operational and safety effects of the proposed project: • HCM traffic operational analysis procedures indicate that narrowing of the through travel lanes from 12 to 11 feet has no substantive effect on the LOS for motor vehicles in the corridor. • HCM analysis procedures cannot quantify the traffic operational effects of removing the cen- ter two-way left-turn lane or providing the bicycle lanes. • HSM traffic safety analysis procedures indicate that removing the center two-way left-turn lane and narrowing the through travel lanes from 12 to 11 feet would increase roadway seg- ment crashes in the corridor from 9.5 to 9.8 crashes per mile per year. • HSM analysis procedures cannot quantify the effects on crashes of providing the bicycle lanes in each direction of travel, of narrowing the parking lanes from 10 to 7 feet, or of providing buffer areas between the through travel lanes and the bicycle lanes and between the bicycle lanes and the parking lanes. Case Study 1 illustrates that substantial additions to the HCM and HSM procedures will be needed before a fully performance-based design process can be applied to a project like that considered here. 7.2.2 Case Study 2: Addition of a Freeway Entrance Loop Ramp to a Diamond Interchange A suburban interchange is being reconstructed to enable major development. Figure 51 shows the existing interchange alignment, and Figure 52 shows the proposed design alternative. The proposed design alternative includes a new entrance loop ramp in the southwest quadrant of the interchange, serving southbound traffic on the cross road entering eastbound on the free- way. This reconfiguration of the interchange removes the heavy left-turn movement on the Input Data Alternative 1 (existing) Alternative 2 (proposed future) Roadway type (2U, 3T, 4U, 4D, 5T) 3T 2U Length of segment, L (mile) 2.0 2.0 AADT (veh/day) 17,800 17,800 Type of on-street parking (none/parallel/angle) Parallel Parallel Proportion of curb length with on-street parking 0.75 0.75 Median width (feet) - for divided only Not Present Not Present Lighting (present / not present) Present Present Auto speed enforcement (present / not present) Not Present Not Present Major commercial driveways (number) 2 2 Minor commercial driveways (number) 20 20 Major industrial / institutional driveways (number) 2 2 Minor industrial / institutional driveways (number) 20 20 Major residential driveways (number) 0 0 Minor residential driveways (number) 2 2 Other driveways (number) 0 0 Speed Category Posted Speed 35 mph or higher Posted Speed 35 mph or higher Roadside fixed-object density (fixed objects / mile) 10 10 Offset to roadside fixed objects (feet) [If greater than 30 or Not Present, input 30] 6 6 Note: 2U = two-lane undivided; 3T = three-lane two-way left-turn lane; 4U = four-lane undivided; 4D = four-lane divided; 5T = five-lane with center turn lane. Table 28. Crash analyses input values.

Value and Benefits of Improved process 159 southbound approach to the southern ramp terminal. The existing eastbound exit ramp and eastbound entrance ramp will be realigned to accommodate the new loop ramp. The predicted safety and operational performance of the existing interchange configuration and the proposed design alternative are compared for the design year. Free-flow speed on the mainline freeway is 65 mph. The design speed of the first curve off the Interstate on the diamond exit ramps and the last curve on the entrance ramps is 60 mph. The design speed of the tightest curve on the loop ramp is 25 mph. The freeway is on level terrain, and the traffic contains 7% trucks and no RVs. The driver population factor is 1.00 since most of the drivers are regular users. The freeway cross section includes three 12-foot lanes in each direction of travel, separated by a 28-foot median with a concrete barrier in the center. Rumble strips are present on the left and right shoulders except along speed change lanes for entrance and exit ramps. The ramp terminals are signalized intersections. Left-turning maneuvers are assumed to be protected. In the existing condition, the southern ramp terminal operates as a 4-leg inter- section, including the crossroad on the north and south side, the exit ramp on the west side, and the entrance ramp on the east side. In the proposed condition, both the southbound-to- eastbound loop ramp and the northbound-to-eastbound diamond ramp leave the crossroad as a speed change lane upstream of the intersection. Only the exit ramp meets the crossroad at Figure 51. Existing interchange alignment. Figure 52. Proposed redesign of interchange, including new entrance loop ramp in southwest quadrant.

160 a performance-Based highway Geometric Design process the signal. The existing three southbound through lanes between the northern and southern ramp terminal remain in the proposed design, but the rightmost lane will be an exit only onto the new loop ramp, while the other two through lanes will proceed through the southern ramp terminal. In the existing condition, entrance ramps are two lanes at the crossroad ramp terminal that merge to a single lane prior to meeting the Interstate. In the proposed interchange alignment, the eastbound entrance ramp becomes a single lane for its entirety. The proposed loop ramp is a single lane. Barrier is present along the outside of a portion of the loop ramp and along a portion of the eastbound exit ramp. In both the existing and proposed conditions, the westbound exit ramp leaves the Interstate as a single lane and widens to two lanes before splitting to dual left turns and a right-turn lane at the crossroad terminal intersection. AADT and peak-hour turning movement projections are shown for the existing condition in Figure 53 and for the proposed design in Figure 54. For the operational analysis, it was assumed that the freeway ramps upstream and downstream have no effect on lane utilization at the subject interchange. For the operational analysis, the eastbound upstream entrance ramp DHV is assumed to be 1,000 veh/h, and the downstream exit ramp DHV is assumed to be 400 veh/h. For the safety analysis, all ramps within one-half mile upstream or downstream of the freeway analysis seg- ment (not including ramps included in the analysis) have an AADT of 4,000. Traffic Operational Analysis Results The HCS 2010 Freeway Facilities procedure was used to assess the LOS for the eastbound freeway mainline, as well as its individual components in the existing and proposed condi- tions for the design year. The analysis area started 2,000 feet upstream of the nearest entrance ramp to the east of the study interchange and ended at the nearest exit ramp west of the study interchange. The LOS results are shown in Table 29. LOS, average travel time, and average travel speed for each analysis segment on the eastbound mainline freeway facility in the design year, 2025. Figure 53. Design year traffic volumes for existing interchange configuration.

Value and Benefits of Improved process 161 Figure 54. Design year peak-hour turning movement counts for planned interchange reconfiguration. Traffic operations at the ramp terminals were assessed with the HCM 2010 Chapter 18 and 22 procedures. The Chapter 22 procedures were used to calculate adjusted saturation flow rates for all lane groups at both ramp terminals, while the Chapter 18 procedures were used to opti- mize the signal timings. Common cycle lengths were chosen based on a target flow ratio of 0.85 for both ramp terminals so that the signals could be coordinated. Adjusted flow ratios and control delays were computed for each lane group. Then origin-destination control delays were computed for the interchange in order to determine LOS values. Alternative Segment LOS Avg. density (pc/mi/ln) Avg. travel time (min/veh) Avg. travel speed (mph) Existing 1: Begin to upstream entrance ramp F 73.2 1.07 21.3 2: Upstream entrance ramp to Ramp C E 47.9 0.33 44.8 3: Ramp C to Ramp D D 27.4 0.57 63.4 4: Ramp D ramp beg to ramp end E 35.7 0.27 56.1 5: Ramp D ramp end to downstream exit D 34.5 0.16 58.2 Proposed 1: Begin to upstream entrance ramp F 73.2 1.07 21.3 2: Upstream entrance ramp to Ramp C E 47.9 0.33 44.8 3a: Ramp C to Ramp E D 27.5 0.32 63.3 3b: Ramp E ramp begin to Ramp D entrance ramp begin D 33.4 0.37 56.9 4: Ramp D entrance ramp begin to ramp end E 35.5 0.17 56.5 5: Ramp D ramp end to downstream exit D 33.8 0.16 59.3 Table 29. LOS, average travel time, and average travel speed for each analysis segment on the eastbound mainline freeway facility in design year 2025.

162 a performance-Based highway Geometric Design process The LOS assessment was based solely on volume-capacity ratios and control delays. It was assumed that the queue ratios would not cause any movements to operate at LOS F. Table 30 presents the control delay and LOS results for the ramp terminals for each movement through the interchange for both the existing and proposed conditions in the design year, 2025. The construction of the loop ramp will result in lower control delays for all movements at the ramp terminals, with one exception. (There was a very slight increase in control delay for vehicles traveling south along the arterial and not entering the freeway.) LOS improved for some movements—most notably the southbound arterial to the eastbound freeway, which is served by the new ramp and improved from LOS C in the existing condition to LOS A in the proposed condition. No movement experienced a reduced LOS in the proposed condition. However, the addition of the loop ramp did not improve LOS scores for the freeway segments. Speeds down- stream of the new ramp entrance were slightly higher, but not enough to impact LOS. Traffic Safety Analysis Results The safety analysis used procedures presented in Chapters 18 (Predictive Method for Freeways) and 19 (Predictive Method for Ramps) of the HSM. The elements included in the analysis were: • A 1,300-foot tangent freeway section extending from the location of the gore point of the proposed loop ramp on the west side to the end of taper of the proposed loop ramp on the east side (the length of the speed change lane for the planned loop ramp). The HSM pro- cedure does not consider eastbound and westbound lanes separately—both directions of travel are considered together. • The southern ramp terminal at the interchange: – In the existing condition, a type D4 signalized intersection (four-leg ramp terminal with diagonal ramps). – In the proposed condition, a type A4 signalized intersection (four-leg ramp terminal at four-quadrant partial cloverleaf A). Note that the intersection actually operates as a three-leg intersection, and that both the southbound loop ramp and the northbound entrance ramp exit the crossroad using a speed change lane upstream of the intersection. However, the type A4 intersection was most appropriate for the analysis because it rec- ognizes that there are no northbound or southbound left-turn movements at the ramp terminal with this configuration. The two entrance ramps are coded as having channel- ized right turns at the intersection. Traffic Movement through the Interchange Existing Control Delay (s) Proposed Control Delay (s) LOS Existing LOS Proposed EB Freeway to NB Arterial 32.3 23.6 C B EB Freeway to SB Arterial *** *** *** *** WB Freeway to NB Arterial 25.4 22.3 B B WB Freeway to SB Arterial 31.4 30.9 C C SB Arterial to EB Freeway 40.5 14.2 C A SB Arterial to WB Freeway *** *** *** *** NB Arterial to EB Freeway 18.2 +++ B +++ NB Arterial to WB Freeway 46.2 31.9 C C SB Arterial to SB Arterial 20.4 21.8 B B NB Arterial to NB Arterial 25.4 14.1 B A NOTE: all movements have v/c < 1.0. ***These movements are yield controlled. +++This movement is a free-flow ramp. Table 30. Control delay and LOS for each movement at the freeway ramp terminals for existing and proposed conditions in design year 2025.

Value and Benefits of Improved process 163 • The ramps on the southern half of the interchange: – In the existing condition, the eastbound diamond exit and entrance ramps. – In the proposed condition, the eastbound diamond exit and entrance ramps as well as the eastbound loop entrance ramp. The safety analysis gave the following results (Tables 31 and 32): Despite providing a slight improvement in operations, the proposed interchange design results in an estimated 43 percent increase in total crashes (from 17.1 to 24.4). Most of the increase occurs in less severe crashes. The results can be explained as follows: • Expected crashes on the freeway segment increase slightly as this segment changes from a 6-lane segment with no lane additions, drops, or speed change lanes, to a segment that includes an entrance ramp speed change lane along the entire segment in the eastbound direction. In the existing condition, no merging maneuvers are required through the segment, while they are required for traffic entering from the loop ramp in the proposed condition. • Expected crashes on the ramps more than double, but the proposed interchange design adds a third ramp with tight, but variable, curvature. • Expected crashes on the ramp terminal increase 62 percent (from 8.4 to 13.6). At first, this finding is counterintuitive, given that a heavy southbound left-turn movement is removed from the intersection. However, in the existing condition, this left-turn maneuver is pro- tected, reducing left-turning collisions. In the proposed condition, the entrance ramps are removed from the intersection by substantial channelization. Right-turn channelization has a CMF larger than 1.0 in the HSM, indicating that it leads to more crashes than the intersection would otherwise experience. Limitations The safety analysis was simplified for the sake of this case study. It would have been more appropriate to consider the freeway segments upstream and downstream of the study segment to encompass the entire interchange area. This is especially true for the segment including the speed change lane on the northbound-to-eastbound entrance ramp. Because the volume on this ramp decreases with the addition of the loop ramp, it is likely that crashes along the freeway Crashes by Facility Component No. of Sites Total K A B C PDO Freeway segments, crashes 1 7.5 0.0 0.1 0.5 1.4 5.4 Ramps, crashes 2 1.2 0.0 0.0 0.2 0.3 0.8 Crossroa Note: K = fatal, A = incapacitating injury, B = non-incapacitating injury, C = reported injury, PDO = property damage only. d ramp terminals, crashes 1 8.4 0.0 0.1 0.4 2.4 5.5 Total 4 17.1 0.1 0.2 1.2 4.1 11.6 Table 31. Crashes by segment type and severity level for existing conditions in design year 2025. Crashes by Facility Component No. of Sites Total K A B C PDO Freeway segments, crashes 1 8.0 0.0 0.1 0.6 1.6 5.7 Ramps, crashes 3 2.8 0.0 0.1 0.4 0.6 1.6 Crossroad ramp terminals, crashes 1 13.6 0.0 0.2 1.1 4.4 7.8 Total 5 24.4 0.1 0.4 2.2 6.7 15.1 Table 32. Crashes by segment type and severity level for proposed realignment in design year 2025.

164 a performance-Based highway Geometric Design process segment along the speed change lane would decrease and offset some of the increase seen in the segment included in this analysis (which included the speed change lane for the loop ramp in the proposed condition). The proposed design of the southern ramp terminal didn’t exactly fit any of the ramp terminal types available for analysis in the HSM. Specifically, both entrance ramps exit the crossroad via speed change lanes 100 feet or more upstream of the intersection and do not interact with any other intersection movements. Therefore the SPF and CMFs used in the analysis for a type A4 ramp terminal may not accurately describe what would be expected at this terminal. While the configuration of the ramp terminal on the north side of the intersection didn’t change, it’s likely that it would operate differently once the southbound left-turn lanes were removed from the southern ramp terminal intersection. In addition, it is unknown how the interchange reconfiguration might impact the overall volumes served at the interchange due to drivers currently using adjacent interchanges rerouting for improved convenience. Conclusions The traffic operational analysis of the project found that the addition of the loop ramp would have very little effect of the LOS on the mainline freeway. This result was expected because the project adds a new mainline freeway ramp terminal, but the total volumes served by the ramps remain unchanged. The LOS for traffic movements at the individual ramp terminals either improved slightly or remained the same, but in several cases there were marked reduc- tions in control delay at the ramp terminals. Thus, overall, the project has a positive effect on traffic operations. The traffic safety analysis indicates that the project would increase crashes in the analysis area by 43 percent, from 17.1 to 24.4 crashes per year. A small portion of this increase in crash frequency results from the addition of a new mainline freeway ramp terminal. Ramps crashes more than doubled (from 1.2 to 2.8 crashes per year) with the addition of the new loop ramp with a curved alignment. In addition, cross road ramp terminal crashes increased from 8.4 to 13.6 crashes per year with the addition of a new crossroad ramp terminal. The project provides modest traffic operational benefits, but will result in a substantial increase in crashes. This result suggests that investigation of other alternatives for improvement of this inter- change would be desirable. Case Study 2 shows that the available HCM and HSM tools are well suited to evaluation of interchange improvement projects of this sort. Thus, performance-based approaches for evalu- ation and comparison of interchange improvement alternatives are a practical tool within the current state of the art. 7.2.3 Case Study 3: Reconstruction (or 3R) Project on a Rural Two-Lane Highway Case Study 3 involves the assessment of a proposed reconstruction or 3R project on a rural two-lane highway. Figure 55 shows typical cross sections for the existing and proposed future designs for a 2-mile rural two-lane highway segment carrying 11,000 vpd. Specifically, the existing 2-foot paved shoulders will be widened to 6-foot paved shoulders, and both centerline and shoulder rumble strips will be added. The 2-mile segment also contains three horizontal curves that will be flat- tened and combined into two curves with larger radii, as shown in Figure 56.

Value and Benefits of Improved process 165 Traffic Operational Analysis Results Table 33 shows a comparison of the traffic operational LOS for these alternatives, based on HCM methods and assumptions. The full reports from the HCS software are attached. In general, the proposed project would provide a slightly higher travel speed (increased by about 2 mph in both directions of travel), and a higher LOS in the northbound direction of travel. The improve- ments would not impact percent time spent following. In this case, because of the lower speed on the road, LOS is determined primarily by average travel speed, rather than by percent time spent following. Traffic Safety Analysis Results Table 34 shows a comparison of the predicted safety performance (annual crashes) for the section for the design year. The following assumptions were made in the HSM analysis for both the existing and proposed scenarios: • Zero-percent grade, • No superelevation variance on horizontal curves, • No spiral transitions on horizontal curves, • No roadway lighting, • No automated speed enforcement, • Roadside hazard rating = 2, and • 10 access points per mile along the section. Chapter 10 of the HSM at this time does not include a CMF for shoulder rumble strips, so the safety benefit these may provide is not accounted for in this analysis; however, the CMF for centerline rumble strips was applicable and did provide safety benefit. The Intersection Worksheet for HSM Chapter 10 was not used because the section did not include any major intersections. ADT = 11,000 vpd Length = 2 mi DHV = 700 SB/400 NB Trucks = 2% PHF = .90 Figure 55. Cross section before 3R improvement (top) and cross section with improvement (bottom).

166 a performance-Based highway Geometric Design process Figure 56. Two horizontal curves with existing and proposed alignments.

Value and Benefits of Improved process 167 The safety comparison of the two designs reflects the following: • Widening the paved shoulder, including centerline rumble strips, and flattening the curves results is predicted to result in a substantial (62 percent) reduction in crash frequency (from 8.1 to 3.1 crashes per mile per year). • The HSM Chapter 10 worksheets allow the analyst to separate the impact of each of the three improvements if such information were desired. • A CMF for shoulder rumble strips could be taken from HSM Part D or the CMF Clearing- house to adjust these results from the Chapter 10 worksheets to account for their impact on safety if the analyst chose to do so. This would increase the safety benefits of the proposed project even further. • Calibration to local conditions and the incorporation of historical crash data into the Empiri- cal Bayes analysis would provide results more specific to this specific location. Conclusions The proposed project will improve the LOS on the northbound direction of travel from LOS D to C, while having no effect on the LOS in the southbound direction of travel, which will remain at LOS E. Thus, the project will have a positive effect on traffic operations, but further consideration of alternatives to improve traffic operations in the northbound direction of travel would be desirable. The project is predicted to result in a substantial (62 percent) reduction in crash frequency, from 8.1 to 3.1 crashes per mi per year. The predicted reduction in crashes would be even larger if a CMF for shoulder rumble strips were included in the HSM Chapter 10 procedure. This case study shows that existing HCM and HSM procedures are well suited to evaluation of this type of project, since effects of all of the proposed changes are accounted for, with the single exception of the effect of shoulder rumble strips on rural two-lane highways. This indicates that performance-based design is a practical approach to assessment of reconstruction or 3R projects on rural two-lane highways. 7.2.4 Case Study 4: Proposed Bypass Around a Small Community A two-lane highway roughly serves as a border between this small community of fewer than 200 people, and a city of nearly 15,000. Projected traffic volumes for design year 2040 are Alternative Traffic Operational Analysis Results LOS Average Travel Speed (mph) Percent Time Spent Following (%) Existing NB D 44.2 61.7 SB E 43.0 82.3 Proposed NB C 46.8 61.7 SB E 45.6 82.3 LOS was determined using HCS 2010 Version 6.30 Table 33. Case Study 3 traffic analyses results. Alternative Predicted Crashes per mile per year Total Fatal and Injury PDO Existing 8.1 2.6 5.5 Proposed 3.1 1.0 2.1 Predicted crashes were determined using HSM 1st Edition, Chapter 10 Worksheets (uncalibrated model without crash history input). Table 34. Case Study 3 crash analyses results.

168 a performance-Based highway Geometric Design process substantially higher than what the current facility can accommodate, so the proposed bypass, which will rebuild the highway along a path to the west of its current alignment and bring it within the city limits, will be a divided four-lane facility. The existing highway alignment will essentially become a local road to serve the homes and farms of this small community. The exist- ing highway will turn off of the proposed new alignment at a stop-controlled intersection on the south side of the bypass. The original road will dead-end just prior to where it would rejoin the proposed new alignment. Figure 57 provides a map of the existing highway overlaid with a drawing of the proposed bypass and other improvements to the surrounding facilities. The purpose of the proposed project is to provide mobility to through traffic and to reduce traffic (especially truck traffic) through the small community. The existing highway is not described as a high crash corridor. This case study compares the operational and safety performance of the existing facility with projected volumes for the 2040 design year with the proposed realignment, including both the new highway segments and intersections and the segments and intersections that will remain along the old highway. The analysis runs from intersection EI1/PI1 on the south side of the cor- ridor to EI5/PI5 on the north side of the corridor. Performance of the dead-end segment of the old highway in the proposed condition was not considered. In addition, performance on the side streets is not considered in this case study, even though some realignments and improvements to side streets were included as part of this project. Figure 58 shows the 2009 volumes for the existing highway alignment and Figure 59 shows the projected 2040 volumes for the proposed project that were used in this evaluation. Figure 60 shows the typical cross section for the proposed alignment of the new highway. Assumptions Detailed inputs used in the safety and operational analysis came from existing data for the actual project site. Where existing data were not available, reasonable assumptions were made by the analysts. The following assumptions were made: • Existing and proposed routes were evaluated as an urban/suburban arterial. The proposed “old highway” will likely function more as a local road than an arterial, but the anticipated design year volumes are within the range of acceptable volumes for an urban/suburban arterial. • Traffic growth rate is the same for both the “no-build” scenario and the “bypass build” sce- nario. (Expected volumes along the existing highway alignment for design year 2040 were not provided, so the analysts made this assumption.) • Conversion of this highway from a two-lane to a four-lane facility extends beyond the north and south borders of the bypass project in design year 2040; however, only the area imme- diately surrounding the proposed bypass is evaluated in the safety and operational analysis. This assumption was needed for the operational analysis, which required a consistent cross section in the segment to the south of the bypass area. The safety and operational impacts of conversion to a four-lane highway outside of the bypass area were not evaluated. • Improvements to side streets were not considered in the analysis; only segments on the new and old alignment were included. • A tree line exists along much of the existing highway right-of-way line (about 25 to 30 feet from the roadway) making the roadside fixed-object density fairly high; however, trees are proposed along the new facility as well and they are closer (see proposed typical cross section for bypass in Figure 60). An assumption was made that the roadside fixed-object density along the new facility would be slightly lower than existing. • Access points along existing highway in design year 2040 are the same as existing. While there is space for development along the east side of the corridor, the analysts chose not to speculate on what accesses may be necessary for potential future growth.

Value and Benefits of Improved process 169 EE EE EE Figure 57. Map of existing conditions overlaid with drawing of proposed highway realignment.

Figure 58. 2009 peak-hour volumes and ADT at intersections along existing highway.

Figure 59. Design year 2040 peak-hour turning movement volumes and AADT for intersections along both the proposed highway alignment and the existing alignment.

172 a performance-Based highway Geometric Design process • Access points along existing highway in design year 2040 are the same regardless of whether bypass is built. • Access points along the new highway bypass in design year 2040 will serve existing adjacent properties with no access to other routes only; no access points to serve possible future devel- opment were considered. • New bypass segments and intersections will be lighted. • Peak-hour factor is 0.95. • Percent heavy vehicles is 2 percent. • Terrain is level. • Pedestrian and bicycle activity is insignificant, thus not considered in the operational analysis. Traffic Safety Analysis Methodology. The traffic safety prediction methodology for urban and suburban arteri- als presented in HSM Chapter 12 was used for the safety analysis of all roadway segments and intersections within the project boundaries considered in this case study. The calculations were performed using the associated Excel spreadsheets that were developed to implement that pro- cedure (which are available on the HSM website). Safety estimates are based on the safety performance functions presented in HSM Chapter 12. A calibration factor of 1.0 was used. Crash history at this site was not considered in the analysis, so all estimates are based on the safety performance of similar sites used to develop the SPFs in the HSM. Results. The expected safety performance of the segments and intersections along the existing alignment of the highway in the design year 2040 are shown in Table 35. Results are presented by crash severity level [total, fatal and injury (FI), and PDO] and crash type (single- vehicle, multivehicle nondriveway, and driveway) for individual segments and intersections. Total predicted annual crashes by severity level for all segments and intersections combined are shown in the final row of the table. The results of the safety analysis show that if left unchanged, the highway through the analysis section is predicted to experience 33 crashes per year in 2040, including 10 fatal and injury crashes. Table 36 presents similar information for the proposed project, including the new realign- ment and the remaining “old highway” segments and intersections. In the proposed condi- tions, old and new highway segments and intersections combined are predicted to experience 22 crashes per year in 2040, including seven fatal and injury crashes. This represents 33 percent reduction in crashes compared to the expected condition if no project were constructed. Figure 60. Proposed typical cross section for proposed alignment of highway (bypass).

Value and Benefits of Improved process 173 Limitations • Intersection of Old Highway and Independence Boulevard in the proposed condition (PI9 in Figure 57) is an all-way stop. This intersection type is not currently included in the HSM; therefore, no safety analysis was performed at this intersection. However, traffic volumes are expected to be very low at this intersection, so it is anticipated that there would be very few crashes here. • It is unclear if “Old Highway” will function as an urban arterial in 2040 as it currently serves very few properties and will no longer be a through street; however, this methodology was the best available option. • Inclusion of pedestrian and bike facilities are not accounted for in this analysis. The HSM provides procedures for estimating bike and pedestrian crashes at signalized intersections, but those procedures were not used here. The HSM segment procedure for urban and suburban arterials does not provided a methodology for assessing the safety impact of pedestrian and bike facilities along the roadway. Traffic Operational Analysis Methodology. The traffic operational analysis for this project was conducted with the urban street procedures of the 2010 HCM. Procedures from HCM Chapters 16 (Urban Street Site Type Predicted Average Crash Frequency (crashes/year) N predicted (TOTAL) N predicted (FI) N predicted (PDO) ROADWAY SEGMENTS Multiple-Vehicle Nondriveway Segment 1 6.388 1.831 4.557 Segment 2 1.346 0.386 0.960 Segment 3 0.954 0.272 0.682 Segment 4 0.567 0.164 0.403 Single-vehicle Segment 1 0.966 0.148 0.818 Segment 2 0.204 0.031 0.172 Segment 3 0.118 0.017 0.101 Segment 4 0.143 0.026 0.117 Multiple-Vehicle Driveway-Related Segment 1 0.433 0.140 0.293 Segment 2 0.182 0.059 0.123 Segment 3 0.678 0.219 0.459 Segment 4 0.156 0.050 0.106 INTERSECTIONS Multiple-Vehicle Intersection 1 2.326 0.786 1.540 Intersection 2 1.084 0.460 0.624 Intersection 3 7.100 2.127 4.973 Intersection 4 7.779 2.257 5.523 Intersection 5 1.211 0.431 0.780 Single-Vehicle Intersection 1 0.175 0.052 0.123 Intersection 2 0.071 0.022 0.049 Intersection 3 0.412 0.123 0.289 Intersection 4 0.475 0.148 0.327 Intersection 5 0.128 0.039 0.089 COMBINED (sum of column) 32.9 9.8 23.1 Table 35. Predicted crashes by severity level for segments and intersections along the existing highway alignment in design year 2040.

174 a performance-Based highway Geometric Design process Site Type Predicted Average Crash Frequency (crashes/year) N predicted (TOTAL) N predicted (FI) N predicted (PDO) ROADWAY SEGMENTS Multiple-Vehicle Nondriveway Segment 1 1.762 0.476 1.286 Segment 2 1.834 0.501 1.333 Segment 3 0.672 0.184 0.488 Segment 4 1.082 0.303 0.779 Segment 5 0.156 0.047 0.110 Segment 6 0.058 0.017 0.040 Segment 7 0.007 0.002 0.005 Segment 8 0.031 0.009 0.022 Single-Vehicle Segment 1 0.239 0.044 0.195 Segment 2 0.287 0.052 0.235 Segment 3 0.105 0.019 0.086 Segment 4 0.226 0.039 0.187 Segment 5 0.193 0.054 0.139 Segment 6 0.071 0.020 0.051 Segment 7 0.022 0.008 0.014 Segment 8 0.107 0.039 0.068 Multiple-Vehicle Driveway-Related Segment 1 0.048 0.014 0.035 Segment 2 0.036 0.010 0.026 Segment 3 0.160 0.045 0.114 Segment 4 0.010 0.003 0.007 Segment 5 0.054 0.017 0.036 Segment 6 0.028 0.009 0.019 Segment 7 0.035 0.011 0.024 Segment 8 0.015 0.005 0.010 INTERSECTIONS Multiple-Vehicle Intersection 1 2.116 0.715 1.401 Intersection 2 2.909 0.880 2.029 Intersection 3 0.866 0.369 0.497 Intersection 4 4.638 1.568 3.070 Intersection 5 1.375 0.548 0.827 Intersection 6 0.058 0.032 0.026 Intersection 7 1.526 0.599 0.927 Intersection 8 0.025 0.015 0.010 Single-Vehicle Intersection 1 0.159 0.047 0.112 Intersection 2 0.236 0.068 0.168 Intersection 3 0.064 0.020 0.044 Intersection 4 0.272 0.067 0.205 Intersection 5 0.135 0.041 0.094 Intersection 6 0.018 0.007 0.012 Intersection 7 0.167 0.046 0.120 Intersection 8 0.017 0.007 0.010 COMBINED (sum of column) 21.8 7.0 14.9 Table 36. Predicted crashes by severity level for segments and intersections along the proposed highway alignment (bypass) and “old highway” in design year 2040.

Value and Benefits of Improved process 175 Facilities) and 17 (Urban Street Segments) were used in conjunction with the Streets, two-way stop-controlled (TWSC) intersection, and all-way stop-controlled (AWSC) intersection mod- ules of HCS 2010 to develop LOS assessments for the existing highway alignment and the pro- posed alignments of the new and old highway in design year 2040 during the AM and PM peak hours. The Streets module does not allow the analysis of street segments with boundaries at locations other than signalized intersections, so procedures from HCM Chapters 16 and 17 were necessarily used in order to combine the control delay outputs from the HCS modules and assess LOS scores for roadway segments and for the facility as a whole. Existing Highway Alignment in Design Year 2040. This urban street facility is composed of Segments ES1 through ES4. As mentioned above, this facility could not be completely ana- lyzed in the Streets module of the HCS due to uncontrolled segment boundaries. As defined in the HCM, an urban street facility is divided into segments by intersections that impose a stop or yield condition on through vehicles, although theoretically a roadway can be divided at an uncontrolled location. So, this urban street facility was logically broken down into four seg- ments: ES1; ES2; ES3; and ES4. While ES1 and ES2 could be combined into one segment, the speed limit changes between these two segments, so they were kept separate in the analysis. Seg- ment ES3 is bound by two signalized intersections, so ideally this segment can be analyzed using the HCS Streets module. These two signalized intersections (EI3 and EI4) are separated by just over 300 ft. The HCM suggests using the ramp terminal procedures to analyze this situation due to a high probability of queue spillback into the upstream signal, however the HCS version 6.3 does not have a functioning interchange module. Segment ES3 was analyzed using the Streets module instead. Despite optimal signal timing strategies, several intersection approaches in both AM and PM peak periods including mainline approaches have v/c ratios exceeding 1.0. The Urban Street Facilities procedure indicates that any segment whose boundary has a v/c greater than 1.0 is an automatic LOS F for the urban street facility as a whole. HCM Chapter 16 and 17 procedures were used to tie together the HCS outputs into LOS scores for both AM and PM peak periods. Proposed Alignment (Bypass) in Design Year 2040. This urban street facility is composed of Segments PS1 through PS4. Again, this facility could not be completely analyzed in the Streets module of the HCS. This facility was divided into two segments: PS1–PS3; and PS4. The Streets module was used to optimize the PI4 signal and generate control delays for both directions. It was assumed that the TWSC intersections that bounded both segments (PI1 and PI5) will not impose any control delay on the uncontrolled through movements. The TWSC module was used to verify this assumption. HCM Chapter 16 and 17 procedures were used to tie together the HCS outputs into LOS scores for both AM and PM peak periods. Old Highway Alignment in Design Year 2040 (with Highway on Bypass). This urban street facility is composed of Segments PS5 through PS8. This facility was divided into two urban street segments: PS5–PS6; and PS7–PS8. Both segments are bounded by stop control. TWSC and AWSC modules were used to compute control delays for the through movements at the segment boundaries, and HCM Chapter 16 and 17 procedures were used to tie together the HCS outputs into LOS scores for both AM and PM peak periods. Results Table 37 summarizes the results of the traffic operational analysis for year 2040 conditions. The table shows that the existing highway alignment, if not improved, would operate at LOS B in the northbound direction of travel during the AM peak period, but would operate at LOS F in the southbound direction during the AM peak period and at LOS F in both directions of travel during the PM peak period. This clearly documents the need for the project.

176 a performance-Based highway Geometric Design process The table indicates that the proposed new highway alignment (bypass) would operate at LOS B or C for both directions of travel during both peak periods in 2040. This indicates that the new alignment will provide much improved traffic service than the existing facility in 2040. The table indicates that the old highway alignment, carrying reduced traffic volumes, will operate at LOS B or C in both directions of travel during the PM peak period. However, the old highway alignment will operate at LOS D in the northbound direction and LOS F in the SB direc- tion during 2040. These poor LOSs occur because the left turns at existing intersection labeled PI7 in Figure 57 have substantial delay. This intersection is currently signalized, but would oper- ate as a TWSC intersection in the proposed future condition. The SB left-turn movements at this intersection, while very low in volume, create substantial control delay for the SB approach. Some change in traffic control at this intersection may be needed. Conclusions The proposed project is expected to reduce crashes by 33 percent under design year condi- tions. The new alignment (bypass) operates effectively at LOS B or C in the design year, while the existing alignment, if not improved, would operate at LOS F under most peak conditions. The old alignment, even with much reduced traffic volumes, operates poorly as an urban arterial segment due to substantial control delay at one intersection. Further investigation of one inte- rsection on the old alignment is recommended. Case Study 4 shows that the available traffic operational and safety analysis tools are very appropriate to address this type of project and a performance-based design approach would be very appropriate. 7.2.5 Case Study 5 Current design policy per AASHTO and FHWA places strict limitations on lane width for free- ways, with no allowance under policy for lane widths less than 12 feet. Of course, there are sufficient examples of freeway projects in which lesser lane widths have been constructed or reconstructed such that the relative safety performance of lane widths less than 12 feet can be assessed. Existing Alignment in Design Year 2040 AM Peak 2040 PM Peak 2040 NB SB NB SB Travel Speed as a Percentage of Base Free- Flow Speed (%) 76% 20% 15% 24% LOS B F F F Proposed Alignment (Bypass) in Design Year 2040 AM Peak 2040 PM Peak 2040 NB SB NB SB Travel Speed as a Percentage of Base Free- Flow Speed (%) 67% 62% 71% 66% LOS C C B C Old Highway Alignment in Design Year 2040 AM Peak 2040 PM Peak 2040 NB SB NB SB Travel Speed as a Percentage of Base Free- Flow Speed (%) 50% 25% 76% 52% LOS D F B C Table 37. LOS assessment for roadway segments in the existing and proposed alternatives for the design year 2040.

Value and Benefits of Improved process 177 The design process for high-volume urban freeways should allow for the evaluation of alter- native cross-section designs that include lane widths of less than 12 feet. There are quantifiable benefits associated with narrower lanes that include the ability to provide additional lanes within limited space, and the cost to reconstruct what are the most costly facilities in the highway sys- tem. (See Figures 61 and 62.) The ISATe HSM and 2010 HCM procedures reveal the relative value of different freeway cross sections using the same total width (lanes and shoulders) but allocating the dimensions such that one additional lane of travel is provided. Table 38 shows a comparison of the capacity or throughput for these designs, based on HCM methods and assumptions. The traffic-carrying capability of the five-lane segment is substan- tially greater than that of the four-lane segment, even with narrower lanes and left shoulder. Table 39 shows a comparison of the predicted safety performance (annual crashes) for a 1-mile tangent freeway segment with 150,000 vpd for two design alternatives that use the same total cross-section dimension. A 4-lane section with 12-foot lanes and full shoulders left and right produce more crashes, although less severe crashes, than a 5-lane section with 11-foot lanes, full right shoulder but narrower left shoulder. The substantive safety comparison of the two designs may appear counterintuitive. It reflects the following: • Safety performance of higher-volume freeways reflects multivehicle crashes, typically rear- end, which are little affected by marginal differences in lane width. • Crash frequency increases with density and volume to capacity; both of which are a function of the throughput capacity, which is defined by the number of lanes. The benefits of providing one more lane of traffic, even with lesser width dimensions and reduced left shoulder width, exceed the marginal adverse effects of the narrower lane and left shoulder widths. ADT = 150,000 vpd Length = 1 mile Directional Distribution = 60/40 Trucks = 8% K = 10% DHV = 9000 vph PHF = .94 Figure 61. Case Study 5, alternative 1 typical section. Figure 62. Case Study 5, alternative 2 typical section.

178 a performance-Based highway Geometric Design process The above analysis, while hypothetical, demonstrates both the value of considering more flexible lane and shoulder width design values, as well as the concept of optimizing a design for a given available total width dimension. 7.2.6 Case Study 6 A state DOT is undertaking the development of a long range plan for reconstruction of a sub- urban freeway system, including its interchanges. The current freeway is six lanes (three in each direction) with full shoulders on each side and a 40-foot median. One interchange within the corridor between the freeway and a county trunk highway (CTH) will require reconstruction. The CTH is a two-lane road. Development along it has led to the county’s need to widen the CTH to a four-lane divided arterial. The existing diamond inter- change, including a crossroad bridge over the freeway and ramp terminal intersections, has insufficient capacity to accommodate expected future traffic demands. In addition to the need to reconstruct the CTH interchange, there is an additional road struc- ture over the freeway about ¼ mile to the east of the CTH bridge. This structure carries local road traffic. It is at a 45 degree skew over the freeway. Although there are no traffic operational or safety performance issues associated with this bridge, its regular condition reports indicate a near-term need for substantial structural repair investments. (There are no recent traffic counts available for the bridge, but county staff estimate it carries no more than a few hundred vehicles a day.) Although the bridge carries local (county) traffic, the bridge itself is owned by and the responsibility of the state DOT. The DOT has indicated that it may need to post or even close the bridge to traffic should further deterioration occur. Figure 63 shows CTH interchange and the local bridge. A summary of land use conditions is as follows: to the north, commercial and office development zoning exists, with some development already occurring in the northeast quadrant of the interchange. The northwest quadrant, and lands further north abutting the CTH are planned for extensive development. The southwest quadrant includes a creek and wetlands. Some developable property exists, but local officials would prefer it remain open. Moreover, access to the southwest quadrant would be difficult given proximity of driveways to the CTH interchange. In the southeast quadrant, land use is limited to single family residences with low density. Alternative Capacity Analysis Results LOS Density (pc/mi/ln) Speed (mph) 1 F 61.3 43.7 2 E 35.5 60.5 LOS was determined using HCS 2010 Freeways Version 6.60 Table 38. Case Study 5, traffic analyses results. Predicted crashes were determined using ISATe (Build6.10) (uncalibrated model without crash data input) K = Fatal A = Incapacitating Injury B = Non-incapacitating Injury C = Reported Injury PDO = Property Damage Only Total K A B C PDO 1 46.8 0.2 0.6 3.2 9.7 33.2 2 40.1 0.3 0.6 3.5 8.1 27.7 Alternative Predicted Crashes per mile per year Table 39. Case Study 5, crash analyses results.

Value and Benefits of Improved process 179 A design team has been assigned the challenge of developing a geometric design solution for the CTH interchange with the freeway, and to address the issue of the local road bridge. Traffic demand studies and agency policies regarding traffic operations have brought the following conclusions: 1. The freeway should be widened to eight basic lanes (four each direction). 2. At least two configurations for the CTH/freeway interchange appear reasonable. Both of these would require widening of the CTH bridge crossing over the freeway from two lanes to either six or eight lanes, depending on the configuration selected. 3. Exiting demand from the east is such that a two-lane exit, with auxiliary lanes from westbound, are considered necessary. Similarly, an auxiliary lane for entering traffic eastbound is needed. 4. Points 1. and 3. above mean that the existing local bridge over the freeway (in poor struc- tural condition) cannot remain in place, as the available openings between the abutments are insufficient to accommodate the additional two lanes plus shoulders, even considering the potential use of design exceptions for lane and shoulder width along the freeway. 5. Given 4. above, the design team concludes that the local road bridge must be replaced, i.e., structural repairs make no sense given it would not fit within the context of the plans for the freeway and new CTH interchange. The design “problems” or “needs” have thus been defined as selecting an interchange con- figuration and designing a new local road bridge to replace the one in poor condition. An imple- mentation challenge is the sequencing of the improvements. The new CTH interchange may take 4 years to complete considering all environmental and public consultation processes. It seems evident that the replacement of the local road bridge will need to be included in the CTH interchange project. Engaging Stakeholders in Project Development Internal project stakeholders include the traffic operations engineer, geometric design engi- neer, travel demand forecasters, state bridge engineer, environmental coordinator, and public involvement specialists. External agency stakeholders include the county engineer, county land use and zoning staff, and local town planning staff. All are supportive of and engaging in devel- oping solutions to the new interchange and local bridge replacement effort. Source: Google Earth Figure 63. Case Study 6 location map.

180 a performance-Based highway Geometric Design process External stakeholders include existing business owners to the north, developers, and home- owners who live south of the freeway. An important concern is the potential for detours or traffic shifts during construction. Dia- logue with stakeholders reveals interesting insights that shape the project: • Both county staff and local developers agree on the importance of widening the CTH in a manner that preserves traffic flow during construction, so access to their properties is not adversely affected. During such discussions, project staff learn that the businesses rely solely on the CTH for access and that they do not consider the local road as being important to serving their properties. • Homeowners living south of the freeway fear the influx of traffic from development as it occurs and especially during reconstruction of the CTH interchange. They suspect that “spill- over” or detouring traffic from the north will emerge on the local road. They express concerns over this expected problem. Interestingly, few of them use the local road. • Engagement with the local emergency service providers reveals that they do rely on the local road for access to the homeowners (e.g., the fire station serving them is north of the project area along the CTH). However, project staff learn that the reason the providers use the local road is that the CTH is often congested. They would prefer to use the CTH and stated an intent to use it once it is widened and the congestion eliminated. When all of the above information is processed by the project team, they come to a conclusion that was not apparent at all at the project outset. They are responsible for a local road bridge in a condition of disrepair, which conflicts with the needs of the freeway passing beneath it, but which few use and none consider essential. The DOT project manager suggests that perhaps they have been mischaracterizing the prob- lem all along. The local road bridge should not be referred to as “in need of replacement,” but rather as being in a state of disrepair. With this broader problem definition, the following solu- tions may be considered: 1. Repair the bridge now to buy time while the new CTH interchange is being reconstructed; then replace it sometime later, removing the cost of this replacement from the CTH inter- change project and providing more flexibility in funding the overall corridor work. 2. Replace the bridge now as originally envisioned. 3. Remove the bridge and do not replace it. Option 3 is fully discussed with all stakeholders. The DOT project team confirms that no one considers it essential, no one would miss it; and indeed the local homeowners would be delighted if the bridge is removed as this would preclude the possibility of any spillover traffic. Plans for widening the CTH cross section and interchange had to maintain traffic along the CTH, and hence in no way relied on the use of the local road crossing. Conclusion The design solution decision was thus Option 3. Minor repairs were made to enable a few more years of service (primarily to address emergency service provider needs during CTH con- struction) with the agreement of all that once the CTH interchange was completed, the DOT would remove the local road bridge forever. This solution, endorsed by all, saved the DOT over $2 million in initial cost and additional perpetual maintenance costs associated with a bridge that in fact no one needed or wanted.