Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual (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.

P A R T 4 Case Studies This part presents three case studies illustrating the application of the HCM to typical plan- ning and preliminary engineering studies. Case Study 1: Freeway Master Plan • Overview • Example 1: focusing the study—screening for service volume problems • Example 2: forecasting v/c hot spots • Example 3: estimating speed and travel time • Example 4: predicting unacceptable motorized vehicle LOS hot spots • Example 5: estimating queues • Example 6: predicting reliability problems • Example 7: comparison of overcongested alternatives Case Study 2: Arterial BRT Analysis • Overview • Example 1: preliminary screening with service volume tables • Example 2: computing critical intersection v/c ratios • Example 3: calculation of intersection v/c ratio for permitted left turns • Example 4: estimating auto and BRT speeds • Example 5: predicting queue hot spots • Example 6: pedestrian, bicycle, and transit LOS Case Study 3: Long-Range Transportation Plan Analysis • Overview • Example 1: estimating free-flow speeds and capacities for model input • Example 2: HCM-based volume–delay functions for model input • Example 3: Predicting density, queues, and delay • Example 4: Predicting reliability

195 T. Case Study 1: Freeway Master Plan 1. Overview The case study site is the 70-mile-long stretch of U.S. 101 within San Luis Obispo County along the central California coast (Exhibit 133). Most of U.S. 101 within the county is a freeway, but there are also sections of multilane highway where access to the highway is provided by unsignalized intersections instead of by interchanges. A screening method is used to iden- tify focus sections for more detailed analysis, as HCM freeway and highway facility analyses should be limited in length to approximately 15 miles (the distance that can be traveled in about 15 minutes). Terminology This case study adopts the following terminology to carry the case study from the very high-level screening analysis to the more detailed HCM segment analysis: • Supersection: Subdivisions of U.S. 101 with consistent broad characteristics (e.g., freeway versus highway, urban versus rural) extending for a number of miles. • Section: Subdivisions of a supersection for more detailed planning application analysis. • Segments: HCM analysis segments (e.g., basic freeway, basic multilane highway, weaving, merge, diverge); each section consists of multiple segments. Planning Objective The agency’s planning objective is to develop a Corridor Mobility Master Plan to identify current and future mobility problems in the corridor, and to establish capital project priorities along the corridor. The corridor plan study area includes freeway interchanges, adjacent front- age roads, and access points for non-motorized transportation. Background U.S. 101 is a four-lane freeway throughout the region, with the exceptions of a six-lane segment with a 7% grade over the 1,522-foot-high Cuesta Grade, just north of the City of San Luis Obispo (California Department of Transportation 2002), and several rural multilane highway segments where unsignalized intersections rather than interchanges provide access to the highway. The majority of the freeway is located in rural areas; however, it passes through five urban areas within the county (Paso Robles, Atascadero, San Luis Obispo, Pismo Beach, and Arroyo Grande).

196 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual U.S. 101 carries between 20,000 and 74,000 AADT, depending on location. Truck traffic accounts for 8% to 10% of AADT. Trucks with five or more axles account for 50% to 55% of the observed truck volumes on the roadway. Recurring congestion occurs on short stretches that operate at LOS E or worse during the afternoon peak hour between San Luis Obispo and Pismo Beach. Example Problems Worked in this Case Study The planning problems illustrated in this case study focus on the identification of future auto mobility problem spots. The specific worked examples are: • Example 1: Focusing the Study—Screening for Service Volume Problems • Example 2: Forecasting v/c Hot Spots • Example 3: Estimating Speed and Travel Time • Example 4: Predicting Unacceptable Motorized Vehicle LOS Hot Spots • Example 5: Estimating Queues • Example 6: Predicting Reliability Problems • Example 7: Comparison of Overcongested Alternatives These planning problems illustrate: • The development, selection, and application of defaults for use in facility-level planning analyses of freeways and multilane highways; • The identification of capacity bottlenecks and the prediction of queues in the study corridor; • The computation of reliability for the freeway; and • A volume-to-capacity ratio approach for comparing the performance of two alternatives, neither of which is able to completely eliminate congestion. 2. Example 1: Focusing the Study—Screening for Service Volume Problems Approach The 70-mile facility will be split into supersections based on the facility’s general characteris- tics (e.g., freeway versus highway, urban versus rural). As described in Section H4 of the Guide for freeways and Section I4 for multilane highways, service volume tables will be used to evaluate each supersection. Exhibit 133. Case study 1: study area.

T. Case Study 1: Freeway Master Plan 197 In this case, it will turn out that the traffic flow and geometric characteristics of the super- sections generally correspond well with the defaults assumed in the construction of the HCM’s generalized daily service volume tables in HCM Chapter 12, Basic Freeway and Multilane High- way Segments (HCM Exhibits 12-39 through 12-42). To use these tables, one merely compares a supersection’s AADT to the value in the table for the appropriate K- (ratio of peak to daily traffic) and D- (directional) factors. However, this example problem will also illustrate how to adjust the HCM’s values for unique local circumstances. Therefore, the HCM’s daily service volumes will be converted to the equiva- lent peak hour, peak direction flow rates per lane for use in screening individual supersections. A flowchart for the analysis is shown in Exhibit 134. Step 1: Split Facility into Supersections The facility is split into supersections where the facility type (controlled-access freeway or multilane highway), the general development intensity of the area (urban or rural), and the general terrain (level, rolling, or mountainous) are fairly constant within the supersection. There is no length limit for a supersection. The terrain type is determined as follows: • If the supersection has short grades (under 1 mile) of 2% or less, it is considered as passing through “level” terrain. • If the supersection has short grades such that heavy vehicles are significantly slowed, but are not at their crawl speed (generally grades under 1 mile in length and 4% or less), then it can be considered as passing through “rolling” terrain. • Supersections with longer or steeper grades that cause heavy vehicles to operate at their crawl speeds are designated as “mountainous.” Exhibit 134. Case study 1: flowchart for example 1.

198 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual This process results in the 70-mile facility being split into nine supersections (A–I), as illus- trated in Exhibit 135. Step 2: Assemble Demand Data The minimum demand data required to use a freeway or multilane highway service volume table, as described in Guide Sections H4 and I4, respectively, are the bi-directional AADT and K- (ratio of peak hour to daily traffic) and D- (directional proportion) ratios. Additional data used to evaluate the suitability of a service volume table for a particular analysis and, if necessary, to adjust the table’s values are the percent heavy vehicles, peak hour factor (PHF), and a capacity adjustment for unfamiliar drivers. The bi-directional AADT is assembled for each supersection using data from the state DOT’s traffic monitoring program. If the AADT varies significantly (i.e., more than 25%) within a supersection, the analyst should consider splitting the supersection and evaluating each of its parts separately. K-factors are obtained from local data in this case, but if not available, default values of 0.09 (urban) and 0.10 (rural) could be used from Exhibit 19 (freeway service volume table) and Exhibit 30 (multilane highway service volume table). Similarly, D-factors are obtained from local data in this case, but these exhibits provide a default value of 0.60 for use when local data are not available. Except for supersection A, where another study was recently conducted, percent heavy vehicles and peak hour factors are not immediately available. Default values of 5% heavy vehicles (urban freeways and multilane highways), 10% heavy vehicles (rural multilane highways), and 12% heavy vehicles (rural freeways) are obtained instead from Exhibit 21 (freeway required data) and Exhibit 135. Case study 1: freeway supersections for screening.

T. Case Study 1: Freeway Master Plan 199 Exhibit 32 (multilane highway required data). Similarly, default PHFs of 0.95 (urban multilane highways), 0.94 (freeways), and 0.88 (rural multilane highways) are applied to the supersections lacking PHF data. Finally, a default capacity adjustment factor (CAF) for driver population (i.e., familiarity with the facility) of 1.00 is applied in urban areas and 0.85 in rural areas. The demand and other input data for this example problem are shown in rows 1–13 of Exhibit 136. Row Supersection A B C D E F G H I Input Data 1 Limits: From County L Arroyo G Avila Bc Los Osos SLO Ct N Atasc S Templ Paso S Paso N 2 Limits: To Arroyo G Avila Bch Los Osos SLO Ct N Atasca S Templ Paso S Paso N County L 3 Length (mi) 12.4 8.4 4.8 4.7 13.1 8.2 4.9 2.9 10.6 4 Through lanes, 2 directions 4 4 4 4 6 4 4 4 4 5 Facility type Highway Freeway Freeway Freeway Highway Freeway Freeway Freeway Highway 6 Area type Urban Urban Rural Urban Rural Urban Urban Urban Rural 7 Terrain type Level Level Level Level Mountain Level Level Level Level 8 AADT (2-Dir) 57,600 63,500 70,100 55,800 44,500 58,700 58,800 32,400 19,500 9 K-factor 0.09 0.08 0.08 0.09 0.09 0.09 0.09 0.09 0.08 10 D-factor 0.52 0.60 0.57 0.55 0.61 0.51 0.58 0.51 0.57 11 % heavy vehicles 10% 5% 12% 5% 12% 5% 5% 5% 12% 12 PHF 0.90 0.95 0.88 0.95 0.88 0.95 0.95 0.95 0.88 13 CAF 1.00 1.00 0.85 1.00 0.85 1.00 1.00 1.00 0.85 Demand 14 Peak direction, veh/h/ln 1,350 1,520 1,600 1,380 810 1,350 1,530 740 440 Initial HCM Service Volumes 15 HCM LOS C 1,360 1,550 1,460 1,550 1,220 1,550 1,550 1,550 1,220 16 HCM LOS D 1,700 1,890 1,770 1,890 1,520 1,890 1,890 1,890 1,520 17 HCM LOS E 1,940 2,150 2,010 2,150 1,730 2,150 2,150 2,150 1,730 Adjust for Heavy Vehicles 18 fHV, HCM 0.926 0.952 0.893 0.952 0.893 0.952 0.952 0.952 0.893 19 ET, local 2.00 2.00 2.00 2.00 5.00 2.00 2.00 2.00 2.00 20 fHV, local 0.909 0.952 0.893 0.952 0.676 0.952 0.952 0.952 0.893 21 Heavy vehicle adjustment 0.982 1.000 1.000 1.000 0.757 1.000 1.000 1.000 1.000 Adjust for PHF 22 HCM default PHF 0.95 0.95 0.88 0.95 0.88 0.95 0.95 0.95 0.88 23 Actual PHF 0.90 0.95 0.88 0.95 0.88 0.95 0.95 0.95 0.88 24 PHF adjustment 0.947 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 Adjust for CAF 25 HCM Default CAF 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 26 Actual CAF 1.00 1.00 0.85 1.00 0.85 1.00 1.00 1.00 0.85 27 CAF adjustment 1.050 1.000 0.850 1.000 0.850 1.000 1.000 1.000 0.850 28 Cumulative HCM service volume adjustment 0.930 1.000 0.850 1.000 0.643 1.000 1.000 1.000 0.850 29 Local LOS A service volume 0 0 0 0 0 0 0 0 0 30 Local LOS C service volume 1,260 1,550 1,240 1,550 780 1,550 1,550 1,550 1,040 31 Local LOS D service volume 1,580 1,890 1,500 1,890 980 1,890 1,890 1,890 1,290 32 Local LOS E service volume 1,800 2,150 1,710 2,150 1,110 2,150 2,150 2,150 1,470 33 LOS D A-C E A-C D A-C A-C A-C A-C Notes: AADT = average annual daily traffic volume in both directions. K-factor = proportion of daily traffic occurring in the peak hour of the day. D-factor = proportion of traffic in the peak direction during the peak hour of the day. CAF = capacity adjustment factor for unfamiliar driver population. PHF = peak hour factor. ET = passenger car equivalent of heavy vehicle traffic stream. fHV = adjustment factor for presence of heavy vehicles in traffic stream. Exhibit 136. Case study 1: screening analysis results.

200 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual Step 3: Compute Peak Hour, Peak Direction Demands Each supersection’s bi-directional AADT is multiplied by the supersection’s K- and D-factors. The result is the supersection’s peak hour demand in the peak direction, shown in row 14 of Exhibit 136. Step 4: Look-Up HCM Service Volumes The analyst obtains the HCM maximum directional hourly service volumes LOS C, D, and E from Exhibit 19 (freeways) and Exhibit 30 (multilane highways). The unadjusted values for level terrain are shown in rows 15, 16, and 17, respectively, of Exhibit 136. The service volumes will be adjusted as needed for rolling or mountainous terrain in the next step. Step 5: Adjust for Local Heavy Vehicle Percentages The service volumes obtained in the previous step are adjusted for heavy vehicle percentage values that differ from those used to generate the service volume table (in this case, subsection A, which is a multilane highway). The heavy vehicle adjustment factor fHV for multilane highways is calculated using Equation 38. For supersection A, the percent heavy vehicles is 10% (0.10), the terrain is level, and the passenger car equivalency of one heavy vehicle on a multilane highway, according to Exhibit 31, is 2.0: f P E HV HV HV( ) ( )= + × − = + × − = 1 1 1 1 1 0.10 2.0 1 0.909 The fHV value used to create the service volume table, which assumed 5% heavy vehicles, is 0.952. Therefore, the HCM service volumes for supersection A will be multiplied by 0.909 / 0.952 = 0.955 in a later step to account for the differing heavy vehicle percentage. Rows 18–21 of Exhibit 136 show the calculation results for all supersections. Step 6: Adjust for Local Peak Hour Factors The ratio of the local PHF and the PHF value used by the service volume table is calculated and will be used in a later step as a local adjustment to the table’s service volumes. For supersection A, the local PHF is 0.90, the table’s PHF is 0.95 (from Exhibit 136), and the ratio of the two is 0.90/0.95 = 0.947. Rows 22–24 of Exhibit 136 show the calculation results for all supersections. Step 7: Adjust for Driver Population Similar to the two previous steps, the ratio of the local CAF to the CAF used by the service volume table is calculated and used later as a local adjustment to the table’s service volumes. In this case, a local CAF of 0.85 was applied to the rural sections (assuming non-regular drivers on this major route connecting northern and southern California) and 1.00 in the urban sections (assuming a high proportion of the traffic consists of commuters during the peak hour), while the service volume tables make no adjustment for driver population (i.e., CAF = 1.00). In the rural sections, the ratio is calculated as 0.85 / 1.00 = 0.850. Rows 25–27 of Exhibit 136 show the calcula- tion results for all supersections. Step 8: Compute Local Service Volumes and LOS This step multiplies the adjustment factors calculated by Steps 5–7 (Row 28 of Exhibit 136). It then applies the combined adjustment factor to the HCM service volumes to arrive at a local

T. Case Study 1: Freeway Master Plan 201 service volume for each supersection (Rows 29–32). Finally, the demands from Step 3 are com- pared to the local service volumes to obtain LOS (Row 33 Exhibit 136). For example, for supersection A, the local heavy vehicle adjustment 0.955 is multiplied by the local PHF adjustment of 0.947 and the local driver population adjustment of 1.000 to arrive at a combined adjustment of 0.904. The HCM service volumes of 1,360, 1,700, and 1,940 veh/h/ln for LOS C, D, and E, respectively, are multiplied by 0.904 and rounded down to the nearest 10 to obtain local service volumes of 1,220, 1,530, and 1,750 respectively. The peak hour, peak direc- tion demand in this supersection is 1,350 veh/h/ln, which is greater than 1,220 but less than or equal to 1,530 and therefore falls into the LOS D range. Comments This screening analysis finds that three supersections (A, C, and E) have an estimated LOS of D or E. Consequently, these supersections are recommended to be analyzed in greater detail. The remainder of this case study will focus on one of these supersections, C, a 4.8-mile stretch of rural freeway that operates at an estimated LOS E. 3. Example 2: Forecasting v/c Hot Spots Approach In Example 1, the 70-mile U.S. 101 in San Luis Obispo County, California, was screened for potential deficiencies that should be the focus of a more detailed planning analysis. This screen- ing found that the 4.8-mile supersection between Avila Beach Road and Los Osos Valley Road (Exhibit 137) operates at an estimated LOS E. The focus of Example 2, therefore, is to identify problem areas within this supersection using a volume-to-capacity hot spot analysis. This analysis focuses on the southbound direction of U.S.101 during the weekday p.m. peak hour, which the screening analysis found was the most critical. This example follows the simplified HCM method described in Section H6 of the Guide to gather the required data for a capacity analysis, divide the supersection into sections based on where traffic demands or capacity change, compute each section’s free-flow speed, and finally estimate capacity and the corresponding v/c ratio for each section. Step 1: Defining Freeway Sections Supersection C is split into freeway sections following the guidance in Section H6 of the Guide. Freeway section boundaries are defined to occur at points where either freeway demand or capac- ity changes (in other words, at all ramp merges, diverges, lane adds, and lane drops). Significant grade changes (to greater than 2% grades) should also be considered for separate sections. In this case, there are no lane drops or significant grade changes, so the supersection is divided into the seven sections shown in the upper half of Exhibit 138 on the basis of the location of on-ramps and off-ramps. The even-numbered sections are identified as “ramp” sections because they start with an on-ramp and end with an off-ramp and because no auxiliary lanes connect the ramps. The odd-numbered sections are identified as basic sections with no ramp merge or weave effects. Step 2: Determine Data Requirements The input data needed to evaluate v/c hot spots are shown in the left column of Exhibit 138. The global inputs include information for free-flow speed, peak hour factor (PHF), percent heavy vehicles, and K-factor. Future conditions analyses might also require a global growth factor.

Exhibit 137. Case study 1: map of supersection C. Section C-1 C-2 C-3 C-4 C-5 C-6 C-7 Section type Basic Ramps Basic Ramps Basic Ramps Basic Length (mi) 0.05 1.65 0.24 1.51 0.37 0.81 0.18 Lanes 2 2 2 2 2 2 2 Mainline AADT 41,700 On-ramp AADT 8,600 6,100 1,400 Off-ramp AADT 500 4,600 1,400 K-factor 0.08 % heavy vehicles 6% Free-flow speed 65 mph PHF 0.92 Exhibit 138. Case study 1: segmentation and input data for supersection C (southbound).

T. Case Study 1: Freeway Master Plan 203 Section-specific inputs consist of segment type, segment length, number of directional lanes, and directional demand AADT. The analyst must provide daily demands for the first mainline freeway section and for all on- and off-ramps. For this example, AADTs are obtained from the same traffic monitoring data source used for the screening analysis in Example 1. Because a smaller facility length is now being studied, it is also feasible to calculate a specific K-factor, heavy vehicle percentage, and peak hour factor, using data from a nearby permanent traffic counting station maintained by the state DOT. The free-flow speed is estimated on the basis of the speed limit. Because the freeway has a 65-mph speed limit for automobiles and a 55-mph speed limit for trucks and vehicles towing trailers, in the analyst’s judgment, a free-flow speed of 65 mph is most appropriate, rather than one greater than the automobile speed limit. Step 3: Estimate Section Capacities The capacity of each individual section is calculated using Equation 16. This equation uses the free-flow speed and percent heavy vehicles as inputs. In addition, a capacity adjustment factor (CAF) can be applied to account for unfamiliar driver populations and the generally lower capacities of ramp merges and diverges. From Example 1, familiar driver populations were assumed for urban freeway sections (i.e., no adjustment was made for driver population). Based on the guidance in Section H6 of the Guide, a CAF of 0.95 is recommended for merges (i.e., on-ramps), with 0.97 recommended for diverges (i.e., off-ramps). As planning sections, rather than HCM segments, are being evaluated in this example, the smaller of the two CAFs (0.95) will control the capacity of a ramps section. Then, for freeway section C-2: c S HV CAF c c C FFS C C ( ) ( ) ( ) ( ) ( ) ( ) = + × − + × = + × − + × = 2,200 10 min 70, 50 1 % 100 2,200 10 min 70,65 50 1 6 100 0.95 2,106 veh/h/ln -2 -2 2 The resulting capacity estimates for all sections are shown in Exhibit 139. Step 4: Convert AADTs to 15-Minute Flow Rates In this step, AADTs are converted to peak 15-minute flow rates by applying the K-factor, the PHF, and (for future conditions analyses) a growth factor, as shown in Equation 17. For freeway section C-1 during the peak 15 minutes (assumed to be the second 15-minute interval within the peak hour), the calculation is as follows: q AADT k PHF fC i gf= × ×     × = × ×     × = 1 41,700 0.08 1 0.92 1.00 3,626 veh/h-1,2 Section C-1 C-2 C-3 C-4 C-5 C-6 C-7 Section type Basic Ramps Basic Ramps Basic Ramps Basic CAF 1.00 0.95 1.00 0.95 1.00 0.95 1.00 Per-lane capacity (veh/h/ln) 2,217 2,106 2,217 2,106 2,217 2,106 2,217 Number of directional lanes 2 2 2 2 2 2 2 Section capacity (veh/h) 4,434 4,212 4,434 4,212 4,434 4,212 4,434 Exhibit 139. Case study 1: peak hour section capacities for supersection C (southbound).

204 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual Equation 17 also assumes that the first and third 15-minute intervals within the peak hour will have average flow rates for the peak hour, while the last 15-minute interval will have a lower- than-average flow rate such that the sum of the demands of the four intervals will equal the total hourly demand. The demands for other intervals within the peak hour need to be calculated when performance measures such as duration of congestion or queue length are to be com- puted, as demand that cannot be served in one 15-minute interval must be carried over to the next interval. For the purposes of this example—determining hot spots where demand exceeds capacity—evaluating only the peak 15 minutes of the peak hour is sufficient. However, Step 8 will demonstrate the calculations for the full peak hour. Exhibit 140 shows the results of the flow rate calculations for the mainline volume entering supersection C and for each of the ramps within the section, for the peak 15 minutes. Before proceeding, these flow rates should be checked for any potential capacity constraints, following the Guide’s recommendations in Section H6. The mainline demand flow rate enter- ing freeway section C-1, 3,626 veh/h, is less than the section capacity of 4,434 veh/h calculated in Step 3. The ramp flow rates are all less than the nominal capacity of 2,000 veh/h for a single- lane ramp stated in Section H6. Therefore, no constraints exist and these flow rates are carried forward to Step 5. Step 5: Assign Demands to Freeway Sections The demand in a given freeway section is computed as shown in Equation 18. Because uncon- strained demands have been provided, the discharge rate is temporarily set to zero in Equation 18 for the purposes of creating initial demand estimates. Therefore, the demand entering a section is the demand served by the preceding section plus the section’s on-ramp demand. The demand departing a section is the entering demand served minus the proportion of the off-ramp demand that can be served (i.e., can reach the off-ramp). If a section’s demand is less than or equal to the section’s capacity, then all of the off-ramp demand can be served. Otherwise, the off-ramp demand is reduced in proportion to the entering demand that is served. The unserved demand in a section is carried over to the next time period (a step not required for evaluating v/c hot spots, but which will be demonstrated in Step 8). For example, the demand entering section C-2 is the demand served by (i.e., departing) sec- tion C-1 (3,626 veh/h, from Exhibit 140) plus the on-ramp demand in section C-2 (748 veh/h), which totals 4,374 veh/h. Because section C-2’s capacity, as calculated in Step 5, is 4,212 veh/h, not all of this demand can be served, and the excess (162 veh/h) is carried over to the next time period. The demand served past the on-ramp is the section’s capacity 4,212 veh/h, and the pro- portion of the demand that is served is (4,212 / 4,374) or 0.963. Therefore, the off-ramp demand of 43 veh/h is reduced to (43 × 0.963) = 41 veh/h, as not all of the off-ramp demand can reach the ramp. The remaining demand, 4,212 – 41 = 4,171 veh/h is able to depart freeway section C-2 and become demand into section C-3. Section C-1 C-2 C-3 C-4 C-5 C-6 C-7 Section type Basic Ramps Basic Ramps Basic Ramps Basic Mainline demand (veh/h) 3,626 On-ramp demand (veh/h) 748 530 122 Off-ramp demand (veh/h) 43 400 122 Exhibit 140. Case study 1: peak flow rate calculations for sections in supersection C (southbound).

T. Case Study 1: Freeway Master Plan 205 Exhibit 141 presents the calculations for all sections. Section entering demands shown in bold represent demands that exceed a section’s capacity. Step 6: Compute d/c Ratios In this step, the analyst calculates the demand-to-capacity (d/c) ratio for each section, which is simply the section entering demand divided by the section capacity. Exhibit 142 presents the calculation results. Step 7: Interpret d/c Results In this step, potentially congested freeway sections are identified by examining which sections have d/c ratios greater than 1.00. The analysis indicates that sections C-2 and C-4 operate over capacity during the weekday p.m. peak hour. All other sections are expected to operate under capacity. Note, however, that if additional capacity was to be provided only in sections C-2 and C-4, other sections downstream of these sections might not be able to accommodate the addi- tional demand served by sections C-2 and C-4. Diagnosing these potential hidden bottlenecks is discussed in Example 7. Step 8: Peak Hour Analysis This step extends the analysis to the full p.m. peak hour. Although not needed to identify capacity problems, it is needed to calculate other performance measures that are demonstrated Section C-1 C-2 C-3 C-4 C-5 C-6 C-7 Section type Basic Ramps Basic Ramps Basic Ramps Basic Mainline demand (veh/h) 3,626 On-ramp demand (veh/h) 748 530 122 Off-ramp demand (veh/h) 43 400 122 Section entering demand (veh/h) 3,626 4,374 4,171 4,701 3,854 3,976 3,854 Section capacity (veh/h) 4,434 4,212 4,434 4,212 4,434 4,212 4,434 Proportion demand served 1.000 0.963 1.000 0.896 1.000 1.000 1.000 Off-ramp demand served (veh/h) 41 358 122 Mainline exiting demand served (veh/h) 3,626 4,171 4,171 3,854 3,854 3,854 3,854 Note: Values in bold indicate demands exceeding capacity, where downstream demand is constrained. Exhibit 141. Case study 1: section demands for supersection C (southbound). Section C-1 C-2 C-3 C-4 C-5 C-6 C-7 Section type Basic Ramps Basic Ramps Basic Ramps Basic Section entering demand (veh/h) 3,626 4,374 4,171 4,701 3,854 3,976 3,854 Section capacity (veh/h) 4,434 4,212 4,434 4,212 4,434 4,212 4,434 Demand-to-capacity ratio 0.82 1.04 0.94 1.12 0.87 0.94 0.87 Exhibit 142. Case study 1: demand-to-capacity ratios for supersection C (southbound).

206 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual in subsequent examples. Steps 4–6 are repeated for each 15-minute interval within the peak hour, with any unserved demand in a section carried over into the next interval. Exhibit 143 through Exhibit 146 show the calculation results. Note that the excess demand in freeway section C-4 cannot be cleared within the first hour of the analysis. A second hour of analysis should be performed using the same procedures (as demonstrated in Step 7) until no carryover demand remains. For simplicity’s sake, however, this case study will show the results only for the first hour of analysis. One potential way to present the facility results is to create a contour diagram similar to Exhibit 147 that shows the d/c ratio for each section for each time period, to visually detect potential bottleneck locations on the study facility. Section C-1 C-2 C-3 C-4 C-5 C-6 C-7 Section type Basic Ramps Basic Ramps Basic Ramps Basic Mainline demand (veh/h) 3,336 On-ramp demand (veh/h) 688 488 112 Off-ramp demand (veh/h) 40 368 112 Section entering demand (veh/h) 3,336 4,024 3,984 4,472 3,865 3,977 3,865 Section capacity (veh/h) 4,434 4,212 4,434 4,212 4,434 4,212 4,434 Proportion demand served 1.000 1.000 1.000 0.942 1.000 1.000 1.000 Off-ramp demand served (veh/h) 40 347 112 Mainline exiting demand served (veh/h) 3,336 3,984 3,984 3,865 3,865 3,865 3,865 Carryover demand to time period 2 (veh/h) 0 0 0 260 0 0 0 Note: Values in bold indicate demands exceeding capacity, where downstream demand is constrained. Exhibit 143. Case study 1: section demands for supersection C (southbound, time period 1). Section C-1 C-2 C-3 C-4 C-5 C-6 C-7 Section type Basic Ramps Basic Ramps Basic Ramps Basic Mainline demand (veh/h) 3,626 Carryover demand from time period 1 (veh/h) 0 0 0 260 0 0 0 On-ramp demand (veh/h) 748 530 122 Off-ramp demand (veh/h) 43 400 122 Section entering demand (veh/h) 3,626 4,374 4,171 4,961 3,872 3,994 3,872 Section capacity (veh/h) 4,434 4,212 4,434 4,212 4,434 4,212 4,434 Proportion demand served 1.000 0.963 1.000 0.849 1.000 1.000 1.000 Off-ramp demand served (veh/h) 41 340 122 Mainline exiting demand served (veh/h) 3,626 4,171 4,171 3,872 3,872 3,872 3,872 Carryover demand to time period 3 (veh/h) 0 162 0 749 0 0 0 Note: Values in bold indicate demands exceeding capacity, where downstream demand is constrained. Exhibit 144. Case study 1: section demands for supersection C (southbound, time period 2).

T. Case Study 1: Freeway Master Plan 207 4. Example 3: Estimating Speed and Travel Time Approach In the previous examples, the freeway was screened for potentially deficient facility supersections (Example 1). The identified critical supersection (C) was then further evaluated for capacity hot spots during the weekday p.m. peak period (Example 2). In this example problem, speed and travel time will be estimated for all individual sections within supersection C. Speeds will be estimated for each section for each 15-minute interval on the basis of delay rates, following the process described in Section H6 of the Guide. The estimated delay for a given Exhibit 145. Case study 1: section demands for supersection C (southbound, time period 3). Section C-1 C-2 C-3 C-4 C-5 C-6 C-7 Section type Basic Ramps Basic Ramps Basic Ramps Basic Mainline demand (veh/h) 3,336 Carryover demand from time period 2 (veh/h) 0 162 0 749 0 0 0 On-ramp demand (veh/h) 688 488 112 Off-ramp demand (veh/h) 40 368 112 Section entering demand (veh/h) 3,336 4,186 4,146 5,383 3,924 4,036 3,924 Section capacity (veh/h) 4,434 4,212 4,434 4,212 4,434 4,212 4,434 Proportion demand served 1.000 1.000 1.000 0.782 1.000 1.000 1.000 Off-ramp demand served (veh/h) 40 288 112 Mainline exiting demand served (veh/h) 3,336 4,146 4,146 3,924 3,924 3,924 3,924 Carryover demand to time period 4 (veh/h) 0 0 0 1,171 0 0 0 Note: Values in bold indicate demands exceeding capacity, where downstream demand is constrained. Exhibit 146. Case study 1: section demands for supersection C (southbound, time period 4). Section C-1 C-2 C-3 C-4 C-5 C-6 C-7 Section type Basic Ramps Basic Ramps Basic Ramps Basic Mainline demand (veh/h) 3,046 Carryover demand from time period 3 (veh/h) 0 0 0 1,171 0 0 0 On-ramp demand (veh/h) 628 446 102 Off-ramp demand (veh/h) 37 336 102 Section entering demand (veh/h) 3,046 3,674 3,637 5,254 3,943 4,045 3,943 Section capacity (veh/h) 4,434 4,212 4,434 4,212 4,434 4,212 4,434 Proportion demand served 1.000 1.000 1.000 0.802 1.000 1.000 1.000 Off-ramp demand served (veh/h) 37 269 102 Mainline exiting demand served (veh/h) 3,046 3,637 3,637 3,943 3,943 3,943 3,943 Carryover demand to time period 5 (veh/h) 0 0 0 1,042 0 0 0 Note: Values in bold indicate demands exceeding capacity, where downstream demand is constrained.

208 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual section will be added to the section’s travel time at free-flow speed to reflect congestion effects. Delay rates are computed separately for undersaturated (d/c ≤ 1.00) and oversaturated (d/c > 1.00) conditions. Travel times are computed from the delay rates. Finally, speeds are computed using each section’s travel time and length. Step 1: Estimate Section Speeds Freeway section delay rates are estimated using Equation 20 (for d/c ratios ≤ 1.05) or Equation 21 (otherwise) from Section H6 of the Guide. Both equations require only the d/c ratio as an input. For freeway section C-3 in time period 1, Exhibit 143 shows that the section demand is 3,984 veh/h, while the section capacity is 4,434 veh/h, which gives a d/c ratio of 0.899. As this ratio is less than 1, Equation 20 is used. 0 1.00 , , 3 , 2 , , , d c E A d c B d c C d c D E d c RU i t i i t i i t i i t i i t i i t∆ = <   +   +   + ≤ ≤      Values for the parameters A, B, C, D, and E are provided in Exhibit 25. For a freeway with a free-flow speed of 65 mph, these values are 92.45, –127.33, 56.34, –8.00, and 0.62, respectively. As the section’s d/c ratio is greater than 0.62, delay will occur in the section and the second part of the equation is applied as follows: RUC ( ) ( ) ( )∆ = − + − =92.45 0.899 127.33 0.899 56.34 0.899 8.00 6.9 s/mi3 2-3,1 As this freeway section is undersaturated, there is no additional oversaturated delay (i.e., ROi t∆ = 0, ). The average travel time for freeway section C-3 in time period 1 is then given by Equation 22: Exhibit 147. Case study 1: d/c contour diagram for supersection C (southbound, p.m. peak hour).

T. Case Study 1: Freeway Master Plan 209 3,600 3,600 0.24 65 0.24 6.9 0 14.9 s-3,1 -3 -3 3 -3,1 -3,1T L FFS LC C C C RU ROC C( ) ( )= + ∆ + ∆ = × + + = This travel time can then be converted into a speed as follows: (3,600 s/h) / (14.9 s) × (0.24 mi) = 58.0 mph. For freeway section C-4 in time period 1, Exhibit 143 shows that the section demand is 4,472 veh/h, while the section capacity is 4,212 veh/h, which gives a d/c ratio of 1.062. As this ratio is greater than 1, both Equation 20 and Equation 21 are used. First, Equation 20 is applied with a d/c ratio of 1.00: RUC ( ) ( ) ( )∆ = − + − =92.45 1 127.33 1 56.34 1 8.00 13.5 s/mi3 2-4,1 Note that this value results for any oversaturated freeway section with a free-flow speed of 65 mph. Next, Equation 21 is applied to determine the additional oversaturated delay: 900 2 1.51 1.062 1 18.5 s/mi-4,1ROC ( )∆ = × − = Equation 22 gives the average travel time for the section: 3,600 3,600 1.51 65 1.51 13.5 18.5 131.9 s-4,1 -4 -4 4 -4,1 -4,1T L FFS LC C C C RU ROC C( ) ( )= + ∆ + ∆ = × + + = Finally, the average travel time is converted into a speed as follows: (3,600 s/h) / (131.9 s) × (1.51 mi) = 41.2 mph. Exhibit 148 provides results for all sections and time periods. Section C-1 C-2 C-3 C-4 C-5 C-6 C7 Section type Basic Ramps Basic Ramps Basic Ramps Basic Length (mi) 0.05 1.65 0.24 1.51 0.37 0.81 0.18 Time Period 1 Undersat. delay rate (s/mi) 1.7 10.2 6.9 13.5 5.6 9.5 5.6 Oversat. delay rate (s/mi) 0 0 0 18.4 0 0 0 Travel time (s) 2.9 108.3 14.9 131.7 22.6 52.6 11.0 Speed (mph) 62.1 54.8 58.0 41.3 58.9 55.4 58.9 Time Period 2 Undersat. delay rate (s/mi) 3.5 13.5 9.3 13.5 5.7 9.8 5.7 Oversat. delay rate (s/mi) 0 10.4 0 53.0 0 0 0 Travel time (s) 2.9 130.9 15.5 184.0 22.6 52.8 11.0 Speed (mph) 62.1 45.4 55.7 29.5 58.9 55.2 58.9 Time Period 3 Undersat. delay rate (s/mi) 1.7 13.0 8.9 13.5 6.2 10.4 6.2 Oversat. delay rate (s/mi) 0 0 0 82.8 0 0 0 Travel time (s) 2.9 112.8 15.4 229.1 22.8 53.3 11.1 Speed (mph) 62.1 52.7 56.1 23.7 58.4 54.7 58.4 Time Period 4 Undersat. delay rate (s/mi) 0.6 5.6 3.6 13.5 6.4 10.6 6.4 Oversat. delay rate (s/mi) 0 0 0 73.7 0 0 0 Travel time (s) 2.8 100.7 14.1 215.3 22.9 53.4 11.1 Speed (mph) 64.3 59.0 61.3 25.2 58.2 54.6 58.4 Note: undersat. = undersaturated, oversat. = oversaturated. Exhibit 148. Case study 1: p.m. peak hour section speeds for supersection C (southbound).

210 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual Step 2: Interpreting Speed Results The computed average speeds can be used to generate a contour diagram similar to the one shown in Exhibit 149 for spotting a speed range that indicates section congestion. By the visual method, the analyst will be able to determine which sections experience low speeds, and for how long. Exhibit 149 indicates a bottleneck in freeway section C-4, as speeds are expected to drop to less than 30 mph during the analysis period. In addition, note that low speeds at the end of the hour indicate that queuing persists beyond the end of the peak hour in section C-4. This result suggests the need to continue the analysis for another hour to adequately capture the congestion occurring on the facility. For simplicity’s sake, however, this example only shows the first hour of analysis. 5. Example 4: Predicting Unacceptable Motorized Vehicle LOS Hot Spots Approach In the previous examples, the freeway was screened for potentially deficient facility supersec- tions (Example 1). The identified critical supersection (C) was then further evaluated for capac- ity hot spots during the weekday p.m. peak period (Example 2). Next, speed and travel time were estimated for all individual sections within supersection C (Example 3). In this example, vehicular density and motorized vehicle LOS will be determined for supersection C, following the process described in Section H6 of the Guide. Step 1: Compute Density The density of vehicles in each section, in vehicles per mile, is computed by dividing the sec- tion’s demand served by its average speed, as given by Equation 27. For example, for freeway Exhibit 149. Case study 1: speed contour diagram for supersection C (southbound, p.m. peak hour).

T. Case Study 1: Freeway Master Plan 211 section C-4 in time period 1, the demand served is 4,212 veh/h (Exhibit 143), the average speed is 41.3 mph (Exhibit 148), and the vehicular density is therefore D d S C C C = = = 4,212 41.3 102.0 veh/mi-4,1 -4,1 -4,1 As the section has two lanes in the study direction, the density is 51.0 veh/mi/ln when expressed on a per-lane basis. Because the HCM expresses density in units of passenger cars per mile per lane for the purpose of determining motorized vehicle LOS, Equation 28 is used to make the conversion from vehicles to passenger cars. The section’s peak hour factor is the same as its parent supersection C, which was found in Example 2 (0.92). Also from Example 2, supersection C is level (i.e., the heavy vehi- cle equivalency factor is 2.0 from Exhibit 20) and the percentage of heavy vehicles is 6%. Then: f P E HV HV HV( ) ( )= + × − = + × − = 1 1 1 1 1 0.06 2.0 1 0.943 D D PHF f PC HV = × = × = 51.0 0.92 0.943 58.8 pc/mi/ln Step 2: Determine LOS The section’s computed density is used to look-up the LOS by facility type (freeway or high- way) and area type (urban or rural). As supersection C is a rural freeway, the right-hand column of Exhibit 26 is used. In the case of section C-4 during the first time period, the section’s volume- to-capacity exceeds 1.00, so the LOS is automatically F regardless of the density. In the case of section C-2 during the third time period, the section’s volume-to-capacity ratio is less than 1.00, but the density of 45.8 pc/mi/ln exceeds the LOS F threshold of 39 pc/mi/ln for rural freeways, so this section is assigned LOS F during this time period. Exhibit 150 summarizes the results for all sections and time periods. Step 3: Interpreting LOS Results At this stage, an analyst would have all basic performance measures identified for all indi- vidual sections of supersection C. It is estimated that freeway sections C-2 and C-4 would expe- rience congested conditions during the weekday p.m. peak hour, based on their LOS F results. The sections with worse LOS operations are typically indicated by high d/c ratios, low speeds, and high travel time delay, as illustrated in the previous examples. As was the case for d/c ratios and speeds, a contour diagram similar to Exhibit 151 can be developed to visually illustrate the extent and duration of poor LOS conditions. The diagram indicates LOS problems in freeway sections C-2 and C-4. LOS F conditions in section C-2 are contained within the peak hour, but the LOS F conditions in section C-4 per- sist for the entire peak hour. In an actual study, it would be recommended that the analysis be extended earlier and later in the afternoon to better account for all of the congestion associated with the bottleneck in section C-4. 6. Example 5: Estimating Queues Approach In the previous examples, the freeway was screened for potentially deficient facility supersections (Example 1). The identified critical supersection (C) was then further evaluated for capacity hot

212 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual Section C-1 C-2 C-3 C-4 C-5 C-6 C-7 Section type Basic Ramps Basic Ramps Basic Ramps Basic Number of lanes 2 2 2 2 2 2 2 Time Period 1 Served demand (veh/h) 3,336 4,024 3,984 4,212 3,865 3,977 3,865 Speed (mph) 62.1 54.8 58.0 41.3 58.9 55.4 58.9 Density (veh/mi/ln) 26.9 36.7 34.4 51.0 32.8 35.9 32.8 Density (pc/mi/ln) 31.0 42.3 39.6 58.8 37.8 41.3 37.8 LOS D E E F E E E Time Period 2 Served demand (veh/h) 3,626 4,212 4,171 4,212 3,872 3,994 3,872 Speed (mph) 62.1 45.4 55.7 29.5 58.9 55.2 58.9 Density (veh/mi/ln) 29.2 46.4 37.4 71.3 32.8 36.2 32.9 Density (pc/mi/ln) 33.7 53.5 43.1 82.2 37.9 41.7 37.9 LOS D F E F E E E Time Period 3 Served demand (veh/h) 3,336 4,186 4,146 4,212 3,924 4,036 3,924 Speed (mph) 62.1 52.7 56.1 23.7 58.4 54.7 58.4 Density (veh/mi/ln) 26.9 39.7 36.9 88.8 33.6 36.9 33.6 Density (pc/mi/ln) 31.0 45.8 42.6 102.3 38.7 42.5 38.7 LOS D F E F E E E Time Period 4 Served demand (veh/h) 3,046 3,674 3,637 4,212 3,943 4,045 3,943 Speed (mph) 64.3 59.0 61.3 25.2 58.2 54.6 58.4 Density (veh/mi/ln) 23.7 31.1 29.7 83.4 33.9 37.0 33.8 Density (pc/mi/ln) 27.3 35.9 34.2 96.1 39.1 42.7 38.9 LOS D E D F E E E Exhibit 150. Case study 1: p.m. peak hour section densities and LOS for supersection C (southbound). Exhibit 151. Case study 1: LOS contour diagram for supersection C (southbound, p.m. peak hour).

T. Case Study 1: Freeway Master Plan 213 spots during the weekday p.m. peak period (Example 2). Next, speed and travel time were esti- mated for all individual sections within supersection C (Example 3). Most recently, vehicular density and motorized vehicle LOS were determined for supersection C (Example 4). In this example, queue lengths will be estimated for individual freeway sections experiencing queuing during the weekday p.m. peak hour, following the approach described in Section H6 of the Guide. A queue typically occurs on sections when the demand is greater than the freeway section capacity. A section is considered to be 100% queued if the queue length (at the estimated queue density) exceeds the available lane-miles of storage in that section. Step 1: Queue Estimation Equation 31 is used to estimate queuing. The freeway sections where demand exceeds capacity during at least one time period during the weekday p.m. peak hour were determined in Example 2 to be sections C-2 and C-4. All of the information needed to estimate queue length—demand, capacity, and density—have been determined in previous examples. For example, for section C-4 during the first time period, the demand is 4,472 veh/h while the capacity is 4,212 veh/h (Example 2), and the section density is 51.0 veh/mi/ln (Example 4). Then: max , 0 max 4,472 4,212, 0 51.0 51.0 mi-4,1 4,1 4 4,1 QL d c D C C C C ( ) ( ) = − = − = − − As section C-4 has two lanes, the queue is 2.55 miles long. As the section is only 1.51 miles long, an additional 1.04 miles of queue is unserved, and the section is considered to be 100% in queue. Exhibit 152 provides queuing results for all of supersection C during the weekday p.m. peak hour. Section C-1 C-2 C-3 C-4 C-5 C-6 C-7 Section type Basic Ramps Basic Ramps Basic Ramps Basic Number of lanes 2 2 2 2 2 2 2 Section length (mi) 0.05 1.65 0.24 1.51 0.37 0.81 0.18 Capacity (veh/h) 4,434 4,212 4,434 4,212 4,434 4,212 4,434 Time Period 1 Demand (veh/h) 3,336 4,024 3,984 4,472 3,865 3,977 3,865 Density (veh/mi/ln) 26.9 36.7 34.4 51.0 32.8 35.9 32.8 Queue length (mi) 2.55 Percent queue 100% Time Period 2 Demand (veh/h) 3,626 4,374 4,171 4,961 3,872 3,994 3,872 Density (veh/mi/ln) 29.2 46.4 37.4 71.3 32.8 36.2 32.9 Queue length (mi) 1.74 5.26 Percent queue 100% 100% Time Period 3 Demand (veh/h) 3,336 4,186 4,146 5,383 3,924 4,036 3,924 Density (veh/mi/ln) 26.9 39.7 36.9 88.8 33.6 36.9 33.6 Queue length (mi) 6.60 Percent queue 100% Time Period 4 Demand (veh/h) 3,046 3,674 3,637 5,254 3,943 4,045 3,943 Density (veh/mi/ln) 23.7 31.1 29.7 83.4 33.9 37.0 33.8 Queue length (mi) 6.25 Percent queue 100% Exhibit 152. Case study 1: p.m. peak hour queue lengths for supersection C (southbound).

214 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual Step 2: Interpreting the Results The planning-level freeway facility method does not contain a queue propagation algorithm, which explains why freeway sections C-2 and C-4 have queue lengths exceeding the section length. As such, the measure of “percent segment queued” is only meaningful up to 100%. The analyst is encouraged to perform a more detailed operational analysis for facilities with signifi- cant queuing impacts. 7. Example 6: Predicting Reliability Problems Approach Previously, the freeway was screened for potentially deficient facility supersections (Exam- ple 1). The identified critical supersection (C) was then further evaluated for capacity hot spots during the weekday p.m. peak period (Example 2). Later, speed and travel time (Example 3), vehicular density and motorized vehicle LOS (Example 4), and queue lengths (Example 5) were estimated for individual sections within supersection C. In this example, freeway reliability will be evaluated as described in Section H7 of the Guide. Two performance measures will be calculated: the 95th percentile travel time index and the percent of trips traveling under 45 mph. The method requires the following data, which were assembled or calculated in previous examples: section lengths, number of through lanes, demand served, capacities, and travel times. Step 1: Compute Vehicle-Miles Traveled For each section and 15-minute time period, the mainline demand served is obtained from Example 2 and multiplied by the section length and the time period length to obtain the VMT. For example, for freeway section C-1 in time period 1, the mainline demand served is 3,336 veh/h, the section length is 0.05 miles, and the time period length is 0.25 hours. The VMT is then 3,336 × 0.05 × 0.25 = 42 veh-mi. Exhibit 153 shows the calculation results for all sections and time periods. Step 2: Compute Vehicle-Hours Traveled For each section and 15-minute time period, the section travel time is obtained from Exam- ple 3 and multiplied by the mainline demand served and time period length to obtain the VHT. For example, for freeway section C-1 in time period 1, the mainline demand served is 3,336 veh/h, the section travel time is 2.9 seconds, and the time period length is 0.25 hours. The VHT is 3,336 × (2.9 s / 3,600 s/h) × 0.25 h = 0.67 veh-h. Exhibit 153 shows the calculation results for all sections and time periods. Step 3: Compute Average Facility Speed for the Peak Hour The section VMTs and VHTs are summed over the sections and the 15-minute time periods to obtain the grand totals for the facility (i.e., supersection C) for the weekday p.m. peak hour: 19,519 VMT and 464.3 VHT. Dividing the VHT into the VMT yields an average facility speed for the hour of 42.0 mph. Step 4: Compute Maximum Facility Demand-to-Capacity Ratio The maximum d/c ratio observed over all sections and 15-minute time periods is obtained from Example 2. It occurs in section C-4 during time period 3, where the demand is 5,383 veh/h, compared to a capacity of 4,212 veh/h, corresponding to a d/c ratio of 1.28.

T. Case Study 1: Freeway Master Plan 215 Step 5: Compute the Recurring Delay Rate The recurring delay rate for the facility is computed using Equation 33, found in Section H7 of the Guide. RDR S FFS = − = − = 1 1 1 42.0 1 65 0.0084 h/mi Although this equation is likely to be most accurate at the facility level, it is also applied at the section level in this example to identify which sections are the biggest contributors to the facil- ity’s reliability problems on the facility. The results are shown in Exhibit 154. Step 6: Compute the Incident Delay Rate The incident-caused delay rate for the facility is computed using Equation 34. Note that the volume-to-capacity ratio X used in the equation is capped at 1.00 when demand on the critical Section C-1 C-2 C-3 C-4 C-5 C-6 C-7 Section type Basic Ramps Basic Ramps Basic Ramps Basic Number of lanes 2 2 2 2 2 2 2 Section length (mi) 0.05 1.65 0.24 1.51 0.37 0.81 0.18 Time Period 1 Served demand (veh/h) 3,336 4,024 3,984 4,212 3,865 3,977 3,865 VMT 42 1,660 239 1,590 358 805 174 Travel time (s) 2.9 108.3 14.9 131.7 22.6 52.6 11.0 VHT 0.67 30.26 4.12 38.52 6.07 14.53 2.95 Time Period 2 Served demand (veh/h) 3,626 4,212 4,171 4,212 3,872 3,994 3,872 VMT 45 1,737 250 1,590 358 809 174 Travel time (s) 2.9 130.9 15.5 183.8 22.4 52.0 10.9 VHT 0.73 38.29 4.49 53.76 6.02 14.42 2.93 Time Period 3 Served demand (veh/h) 3,336 4,186 4,146 4,212 3,924 4,036 3,924 VMT 42 1,727 249 1,590 363 817 177 Travel time (s) 2.9 112.8 15.4 228.8 22.5 52.3 10.9 VHT 0.67 32.79 4.43 66.92 6.13 14.66 2.97 Time Period 4 Served demand (veh/h) 3,046 3,674 3,637 4,212 3,943 4,045 3,943 VMT 38 1,516 218 1,590 365 819 177 Travel time (s) 2.8 100.7 14.1 215.1 22.6 52.6 11.0 VHT 0.59 25.69 3.56 62.92 6.19 14.78 3.01 Exhibit 153. Case study 1: p.m. peak hour VMT and VHT for supersection C (southbound). Section C-1 C-2 C-3 C-4 C-5 C-6 C-7 Section type Basic Ramps Basic Ramps Basic Ramps Basic Recurring delay rate (h/mi) 0.0006 0.0035 0.0019 0.0181 0.0017 0.0028 0.0017 Exhibit 154. Case study 1: recurring delay rates for supersection C (southbound, p.m. peak hour).

216 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual section within the facility exceeds capacity. The number of lanes is limited to 2, 3, or 4. For the facility: IDR N X[ ] [ ]( ) ( )= − − × × = − − × × =0.020 2 0.003 0.020 2 2 0.003 1 0.020 h/mi12 12 Although the equation is likely to be most accurate at the facility level, it is also applied at the section level in this example to identify which sections are the biggest contributors to the reli- ability problems on the facility. The results are shown in Exhibit 155. Step 7: Compute the Mean Travel Time Index The mean travel time index for the facility is computed using Equation 32. It is the ratio of the mean annual peak hour travel time (with incidents) to the travel time under free-flow condi- tions. The calculation for the facility is as follows: TTI FFS RDR IDRm ( ) ( )= + × + = + × + =1 1 65 0.0084 0.0200 2.85 This result indicates that, on average, travel through supersection C during the weekday p.m. hour takes 2.85 times as long as under free-flow conditions, implying an average annual peak hour speed of 65 / 2.85 = 22.8 mph. Although the mean travel time equation is likely to be most accurate at the facility level, it is also applied at the section level in this example to identify which sections are the biggest con- tributors to the reliability problems on the facility. The results are shown in Exhibit 156. Freeway sections C-2 and C-4 appear to be the greatest contributors to the peak hour reliability problems on the facility. Step 8: Compute the 95th Percentile Travel Time Index The 95th percentile travel time index for the facility is computed using Equation 35. TTI TTIm( ) ( )= + × = + × =1 3.67 ln 1 3.67 ln 2.85 4.8495 The result of 4.84 implies that travel speeds through subsection C fall below 13.4 mph during 5% of the peak hours during the year. Section C-1 C-2 C-3 C-4 C-5 C-6 C-7 Section type Basic Ramps Basic Ramps Basic Ramps Basic Maximum d/c ratio 0.82 1.04 0.94 1.28 0.89 0.96 0.89 Incident delay rate (h/mi) 0.0018 0.0200 0.0095 0.0200 0.0049 0.0123 0.0049 Exhibit 155. Case study 1: incident delay rates for supersection C (southbound, p.m. peak hour). Section C-1 C-2 C-3 C-4 C-5 C-6 C-7 Section type Basic Ramps Basic Ramps Basic Ramps Basic Recurring delay rate (h/mi) 0.0006 0.0035 0.0019 0.0181 0.0017 0.0028 0.0017 Incident delay rate (h/mi) 0.0018 0.0200 0.0095 0.0200 0.0049 0.0123 0.0049 Mean travel time index 1.16 2.55 1.75 3.57 1.43 1.98 1.43 Exhibit 156. Case study 1: mean travel time index values for supersection C (southbound).

T. Case Study 1: Freeway Master Plan 217 Step 9: Compute Percent Trips Under 45 mph The percent of peak hour trips over the year that travel at average speeds below 45 mph for the facility is computed using Equation 36. PT TTIm[ ] [ ]( ) ( )= − − × − = − − × − =1 exp 1.5115 1 1 exp 1.5115 2.85 1 0.9445 Approximately 94% of the peak hour trips in supersection C are completed at average speeds below 45 mph, which is to be expected when the average annual speed is 42.1 mph for recurring congestion (i.e., without incidents). 8. Example 7: Comparison of Overcongested Alternatives Approach Previous examples demonstrated planning-level calculations of a variety of performance mea- sures for supersection C, which showed that two sections of this facility, C-2 and C-4, operated poorly. This example illustrates how one might compare the performance effects of alternative mitigation measures, none of which completely eliminate congestion. In this case, the agency is choosing between two alternatives—doing nothing or adding a lane—neither of which is able to change LOS F conditions to any better than LOS F. This example shows how to numeri- cally compare the improvement to the before condition and determine whether there is a net improvement in freeway operations, even though the facility will still operate at LOS F. This example also demonstrates that addressing capacity problems at one bottleneck may reveal other hidden bottlenecks downstream. To analyze and compare the effects of the two alternatives, this example returns to Exam- ple 2, which identified v/c hot spots along the facility. The data and analysis results for the “Do Nothing” alternative are retrieved directly from Example 2, while the Example 2 calculations are repeated for the “Add Lane” alternative, but adding an auxiliary lane in freeway section C-4 to relieve the bottleneck. Analysis The results for the “Do Nothing” alternative are obtained from Exhibit 143 through Exhibit 146 and summarized in Exhibit 157. The results for the “Add Lane” alternative are calculated as described herein and summarized in Exhibit 158. In the “Add Lane” alternative, freeway section C-4 is modified by adding an auxiliary lane between the on-ramp and off-ramp. This transforms the section from a ramps section into a weave section, which requires the calculation of a new capacity adjustment factor. Equation 23, located in Section H6 of the Guide, is used to calculate this factor. As no informa- tion is available about specific movements (e.g., ramp-to-ramp demands) within the weaving section, the guidance from Section H6 of the Guide is followed to assume zero ramp-to-ramp demand, so that the volume ratio Vr is the sum of the on- and off-ramp volumes, divided by the sum of the mainline entering volume and the on-ramp volume. For freeway section C-4 during time period 1, using data from Exhibit 143 in Example 2, the on-ramp demand flow rate is 488 veh/h, the off-ramp demand flow rate is 368 veh/h, and the mainline entering flow rate is 3,984 veh/h. Then: Vr = + + = 488 368 3,984 488 0.191

218 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual Section C-1 C-2 C-3 C-4 C-5 C-6 C-7 Section type Basic Ramps Basic Ramps Basic Ramps Basic Section capacity (veh/h) 4,434 4,212 4,434 4,212 4,434 4,212 4,434 Time Period 1 Mainline demand (veh/h) 3,336 On-ramp demand (veh/h) 688 488 112 Off-ramp demand (veh/h) 40 368 112 Entering demand (veh/h) 3,336 4,024 3,984 4,472 3,865 3,977 3,865 d/c ratio 0.75 0.96 0.90 1.06 0.87 0.94 0.87 Mainline vol. served (veh/h) 3,336 3,984 3,984 3,865 3,865 3,865 3,865 Time Period 2 Mainline demand (veh/h) 3,626 Carryover demand (veh/h) 0 0 0 260 0 0 0 On-ramp demand (veh/h) 748 530 122 Off-ramp demand (veh/h) 43 400 122 Entering demand (veh/h) 3,626 4,374 4,171 4,961 3,872 3,994 3,872 d/c ratio 0.82 1.04 0.94 1.18 0.87 0.95 0.87 Mainline vol. served (veh/h) 3,626 4,171 4,171 3,872 3,872 3,872 3,872 Time Period 3 Mainline demand (veh/h) 3,336 Carryover demand (veh/h) 0 162 0 749 0 0 0 On-ramp demand (veh/h) 688 488 112 Off-ramp demand (veh/h) 40 368 112 Entering demand (veh/h) 3,336 4,186 4,146 5,383 3,924 4,036 3,924 d/c ratio 0.75 0.99 0.94 1.28 0.88 0.96 0.88 Mainline vol. served (veh/h) 3,336 4,146 4,146 3,924 3,924 4,036 3,924 Time Period 4 Mainline demand (veh/h) 3,046 Carryover demand (veh/h) 0 0 0 1,169 0 0 0 On-ramp demand (veh/h) 628 446 102 Off-ramp demand (veh/h) 37 336 102 Entering demand (veh/h) 3,046 3,674 3,637 5,252 3,943 4,045 3,943 d/c ratio 0.69 0.87 0.82 1.25 0.89 0.96 0.89 Mainline vol. served (veh/h) 3,046 3,637 3,637 3,943 3,943 4,045 3,943 Full Hour Section demand (veh/h) 3,336 4,024 3,984 4,472 3,865 3,977 3,865 Demand served (veh/h) 3,336 4,024 3,984 4,212 3,865 3,977 3,865 Average d/c ratio 0.75 0.96 0.90 1.19 0.88 0.95 0.88 Maximum d/c ratio 0.82 1.04 0.94 1.28 0.89 0.96 0.89 Note: vol. = volume. Exhibit 157. Case study 1: d/c results for the “do nothing” alternative.

T. Case Study 1: Freeway Master Plan 219 Section C-1 C-2 C-3 C-4 C-5 C-6 C-7 Section type Basic Ramps Basic Weave Basic Ramps Basic Section capacity (veh/h) 4,434 4,212 4,434 6,651 4,434 4,212 4,434 Time Period 1 Mainline demand (veh/h) 3,336 On-ramp demand (veh/h) 688 488 112 Off-ramp demand (veh/h) 40 368 112 Entering demand (veh/h) 3,336 4,024 3,984 4,472 4,104 4,216 4,100 d/c ratio 0.75 0.96 0.90 0.71 0.93 1.00 0.92 Mainline vol. served (veh/h) 3,336 3,984 3,984 4,104 4,104 4,100 4,100 Time Period 2 Mainline demand (veh/h) 3,626 Carryover demand (veh/h) 0 0 0 0 0 4 0 On-ramp demand (veh/h) 748 530 122 Off-ramp demand (veh/h) 43 400 122 Entering demand (veh/h) 3,626 4,374 4,171 4,701 4,301 4,427 4,096 d/c ratio 0.82 1.04 0.94 0.74 0.97 1.05 0.92 Mainline vol. served (veh/h) 3,626 4,171 4,171 4,701 4,301 4,096 4,096 Time Period 3 Mainline demand (veh/h) 3,336 Carryover demand (veh/h) 0 162 0 0 0 213 0 On-ramp demand (veh/h) 688 488 112 Off-ramp demand (veh/h) 40 368 112 Entering demand (veh/h) 3,336 4,186 4,146 4,634 4,266 4,593 4,109 d/c ratio 0.75 0.99 0.94 0.73 0.96 1.09 0.93 Mainline vol. served (veh/h) 3,336 4,146 4,146 4,266 4,266 4,109 4,109 Time Period 4 Mainline demand (veh/h) 3,046 Carryover demand (veh/h) 0 0 0 0 0 379 0 On-ramp demand (veh/h) 628 446 102 Off-ramp demand (veh/h) 37 336 102 Entering demand (veh/h) 3,046 3,674 3,638 4,083 3,747 4,230 4,110 d/c ratio 0.69 0.87 0.82 0.65 0.85 1.00 0.93 Mainline vol. served (veh/h) 3,046 3,638 3,638 3,747 3,747 4,110 4,110 Full Hour Section demand (veh/h) 3,336 4,024 3,984 4,472 4,104 4,216 4,100 Demand served (veh/h) 3,336 4,024 3,984 4,472 4,104 4,212 4,100 Average d/c ratio 0.75 0.96 0.90 0.71 0.93 1.04 0.93 Maximum d/c ratio 0.82 1.04 0.94 0.74 0.97 1.09 0.93 Note: vol. = volume. Exhibit 158. Case study 1: d/c results for the “add lane” alternative.

220 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual CAF V Lr s= − + ≤0.884 0.0752 0.0000243 1.00weave CAF ( )( )= − + × = ⇒0.884 0.0752 0.191 0.0000243 1.51mi 5,280 ft/mi 1.06 1.00weave The capacity of freeway section C-4 during time period 1 is determined from Equation 16: c S HV CAF c c C FFS C C ( ) ( ) ( ) ( ) ( ) ( ) = + × − + × = + × − + × = 2,200 10 min 70, 50 1 % 100 2,200 10 min 70,65 50 1 6 100 1.00 2,217 veh/h/ln -4 -4 -4 Because the relative proportions of weaving and non-weaving volumes may vary during the peak hour, the volume ratio, the CAF, and ultimately the section capacity may also vary in each time period. However, in this case, the weaving section is long enough that CAFweave exceeds 1.00 in all time periods and therefore is constrained to 1.00 in all time periods, with the result that section C-4’s capacity is the same in all time periods. Interpretation Exhibit 159 provides a side-by-side comparison of the average peak hour d/c ratios for each section for the two alternatives. Adding a lane to freeway section C-4 has no effect on the upstream sections C-1, C-2, and C-3; significantly improves the d/c ratio for Section C-4; and significantly worsens the d/c ratios for the downstream sections C-5, C-6, and C-7. It is hard to say which alternative is better until one computes the volume- and distance- weighted average d/c ratio for the facility under each alternative. The section demands are mul- tiplied by the section lengths to get VMT demanded. The section capacities are multiplied by the section lengths to get VMT of capacity. The result is an average d/c ratio of 1.01 for the Do Nothing alternative, and 0.86 for the Add Lane alternative. Therefore, Add Lane is the more- effective alternative. Using an average d/c ratio can give the mistaken impression that somehow the Add Lane alternative has solved the freeway’s congestion problems on the freeway. Examination of Exhibit 159, however, shows that this is clearly not the case. Congestion on a freeway facility is driven by its critical bottleneck; therefore, it is better to compare the worst case d/c ratios on the facility under each alternative. Exhibit 160 shows such a comparison. Scenario C-1 C-2 C-3 C-4 C-5 C-6 C-7 Do Nothing 0.75 0.96 0.90 1.19 0.88 0.95 0.88 Add Lane 0.75 0.96 0.90 0.71 0.93 1.04 0.93 Exhibit 159. Case study 1: comparison of average d/c ratios. Scenario C-1 C-2 C-3 C-4 C-5 C-6 C-7 Do Nothing 0.82 1.04 0.94 1.28 0.89 0.96 0.89 Add Lane 0.82 1.04 0.94 0.74 0.97 1.09 0.93 Exhibit 160. Case study 1: comparison of maximum d/c ratios.

T. Case Study 1: Freeway Master Plan 221 Again, the results are ambivalent until one compares the worst case d/c ratio under each alter- native. The worst case d/c ratio for the Do Nothing alternative is 1.28 in freeway section C-4. The worst case d/c ratio for the Add Lane alternative is 1.09 in section C-6. Again, Add Lane is the more-effective alternative. These conclusions are confirmed by computing and comparing the freeway performance measures under each alternative following the procedures illustrated in the previous example problems. The results are shown in Exhibit 161. 9. Reference California Department of Transportation. Caltrans Journal, Vol. 2, Issue 6, May–June 2002. Time Period Full 1 2 3 4 Hour Do Nothing Alternative Facility travel time (min) 5.7 7.0 7.5 7.0 6.9 Space mean speed (mph) 50.3 41.3 38.7 41.2 42.0 Facility density (veh/mi/ln) 53.2 66.3 70.7 64.1 63.6 Total queue length (mi) 2.6 7.0 6.6 6.3 5.6 Facility LOS F F F F F Maximum d/c ratio on facility 1.06 1.18 1.28 1.25 1.19 Add Lane Alternative Facility travel time (min) 5.0 5.8 5.8 4.9 5.4 Space mean speed (mph) 57.3 49.5 49.6 59.1 53.6 Facility density (veh/mi/ln) 48.1 57.6 57.1 43.6 51.6 Total queue length (mi) 0.1 3.6 2.7 0.2 1.7 Facility LOS F F F F F Maximum d/c ratio on facility 1.00 1.05 1.09 1.00 1.09 Notes: The total travel time for the hour is the volume-weighted average of the 15-minute time periods. The space mean speed for the hour is the inverse of the average total travel time. The density and queue length for the hour are simple averages of the 15-minute values. The facility LOS and the maximum d/c ratio for the hour are the worst of the 15-minute periods. Exhibit 161. Case study 1: comparison of freeway performance for the two alternatives.

222 U. Case Study 2: Arterial BRT Analysis 1. Overview This case study addresses the implementation of a 14.4-mile bus rapid transit (BRT) line connecting Berkeley, Oakland, and San Leandro, California, on Telegraph Avenue and Inter- national Blvd. The proposed route and study area are shown in Exhibit 162. There are a total of 110 signalized and 19 stop- controlled intersections along the proposed BRT route. A screening method will be used to identify the focus sections for more-intense analysis. Planning Objective The agency’s planning objective is to identify the traffic, tran- sit, pedestrian, and bicycle impacts of the proposed BRT project. Background The BRT route will run primarily on Telegraph Avenue and International Blvd.: • Telegraph Avenue is a 4-lane divided and undivided arterial, with posted speed limits of 25 to 30 mph, with curbside parallel parking and continuous sidewalks on both sides. • International Blvd. has similar geometric, speed limit, parking, and sidewalk characteristics as Telegraph Avenue (AC Transit 2012). • Peak hour directional through volumes on both streets range from 400 to 1,000 vehicles per hour. • Eight intersections on the project alignment regularly experience peak hour LOS E/F condi- tions (before project implementation). The BRT project includes bus priority at traffic signals, exclusive bus-only lanes, and elimina- tion of some through lanes and left turns. Curbside parking is generally retained between BRT stations. Curbside parking is lost at most BRT stations. Example Problems Worked in this Case Study The planning problems to be illustrated by worked examples are: • Example 1: Preliminary Screening with Service Volume Tables • Example 2: Computing Critical Intersection Volume-to-Capacity Ratios • Example 3: Calculating Intersection v/c Ratios with Permitted Left Turns

U. Case Study 2: Arterial BRT Analysis 223 • Example 4: Estimating Auto and BRT Speeds • Example 5: Predicting Queue Hot Spots • Example 6: Pedestrian, Bicycle, and Transit LOS The HCM does not currently address the analysis of truck LOS for urban streets, so truck LOS analysis is excluded from this case study. 2. Example 1: Preliminary Screening with Service Volume Tables Approach To reduce the resources required to evaluate the environmental impacts of the proposed BRT project, the 14-mile route length will be split into supersections where traffic demands, roadway geometry, and signal timing are relatively similar. HCM service volume tables will then be used to screen the average LOS for these supersections, as described in Section K4 of the Guide. Next, motorized vehicle LOS will be assessed after removing one through lane in each direction and dedicating the space exclusively to BRT. Supersections with LOS in the E/F range will be selected for more detailed analyses in subsequent example problems within this case study. Step 1: Divide the BRT Route into Supersections The BRT route is divided into supersections on the basis of a significant change in: • Posted speed limit, • Number of through lanes, • Median presence, or • Traffic demand (such as a major trip generator). The resulting 10 supersections are listed in Exhibit 163. Exhibit 162. Case study 2: study area.

224 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual Step 2: Obtain AADT Estimates for Supersections In this example, AADT volumes are not available for each supersection, but peak hour counts are available, as are data from a few permanent traffic recorders located on other arterial streets. These traffic recorders indicate that typical peak hour traffic in the area is approximately 10% of the daily traffic and that 55% of peak hour traffic travels in the peak direction. Therefore, super- section AADTs are estimated by dividing the peak hour volume of the most heavily traveled portion of each supersection by 0.10 and 0.55. The resulting AADTs are shown in Exhibit 164. Step 3: Select Service Volumes for Supersections Because this is a preliminary screening process to identify which sections of the BRT route require additional analysis, it is recommended that the analyst select a conservatively good Exhibit 163. Case study 2: analysis supersections. Super- section Street Limits Length (mi) Peak Volume (1-dir.) Speed Limit (mph) No. of Lanes (1-dir.) Median A Telegraph Ave. Dwight to Woolsey 0.84 910 25 2 No B Telegraph Ave. Woolsey to SR 24 0.80 1,010 30 2 No C Telegraph Ave. SR 24 to 45th 0.60 910 30 2 TWLTL D Telegraph Ave. 45th to Broadway 2.01 890 25 2 TWLTL E International Blvd. Lake Merritt to 23rd 1.58 420 30 2 No F International Blvd. 23rd to 35th 0.87 550 25 2 No G International Blvd. 35th to High 0.51 660 25 2 TWLTL H International Blvd. High to Hegenberger 1.78 560 30 2 TWLTL I International Blvd. Hegenberger to 98th 1.37 610 30 2 TWLTL J International Blvd. 98th to Dutton 1.06 460 30 2 TWLTL Notes: TWLTL = two-way left-turn lane, dir. = direction. Several sections of other streets along the BRT route have been left out of the example for the sake of clarity. The same screening approach would be applied to these sections as well. Several sections of other streets along the BRT route have been left out of the example for clarity’s sake. The same screening approach would be applied to these sections as well. Exhibit 164. Case study 2: service volume screening results. Before BRT After BRT Super- section Street Limits AADT Through Lanes (2-dir.) LOS D Service Volume Through Lanes (2-dir.) LOS D Service Volume A Telegraph Ave. Dwight to Woolsey 16,550 4 22,300 2 10,700 B Telegraph Ave. Woolsey to SR 24 18,360 4 22,300 2 10,700 C Telegraph Ave. SR 24 to 45th 16,550 4 22,300 2 10,700 D Telegraph Ave. 45th to Broadway 16,180 4 22,300 2 10,700 E International Blvd. Lake Merritt to 23rd 7,640 4 22,300 2 10,700 F International Blvd. 23rd to 35th 10,000 4 22,300 2 10,700 G International Blvd. 35th to High 12,000 4 22,300 2 10,700 H International Blvd. High to Hegenberger 10,180 4 22,300 2 10,700 I International Blvd. Hegenberger to 98th 11,090 4 22,300 2 10,700 J International Blvd. 98th to Dutton 8,360 4 22,300 2 10,700 Notes: AADT = annual average daily traffic, dir. = direction. Service volumes in bold indicate supersections where operations are estimated to be worse than LOS D.

U. Case Study 2: Arterial BRT Analysis 225 LOS for the screening. In this case the agency’s policy is to maintain LOS E for motorized vehicles, so for screening purposes, the service volume threshold for motorized vehicle LOS D will be used. Exhibit 45 in Section K4 of the Guide is used for the screening. As this service volume table only provides two choices of posted speeds (30 and 45 mph), the section for a 30-mph posted speed is used for supersections with posted speeds of both 25 and 30 mph, as 30 mph is closer to 25 mph than 45 mph. The LOS D service volume for “Before BRT” (i.e., existing) conditions is then read from the LOS D column for four-lane streets (i.e., two through lanes in each direction), using the previously determined K-factor of 0.10 and D-factor of 0.55. The process is repeated for the “After BRT” conditions, except that only one through lane will be provided for general traffic once the exclusive BRT lane is constructed. Therefore, the LOS D column for two-lane streets (i.e., one through lane in each direction) is used to determine the “After BRT” service volume. The results of the evaluation are shown in Exhibit 164. Step 4: Identify Supersections for Further Analysis The AADTs are compared to the service volumes to identify those supersections requiring further analysis. Six supersections (A, B, C, D, G, and I) are retained for further analysis, and four are dropped from further analysis. Interpretation of Results The screening analysis shows that all supersections currently have spare capacity, and that for four of them, the dedication of one through lane in each direction to BRT would not reduce the auto LOS below D. For the six other supersections, the screening analysis suggests that there may be some capacity problems requiring more-detailed evaluation. The remaining examples in this case study will focus on just one of the supersections: the stretch of Telegraph Avenue between State Route 24 and 45th Street, supersection C. 3. Example 2: Computing Critical Intersection Volume-to-Capacity Ratios Approach Example 1 divided the study corridor into 10 supersections for screening and found that six supersections were likely to experience LOS E or F operations following the dedication of one travel lane in each direction to BRT. This example will focus on one of these supersections and will evaluate which intersections within the supersection will likely experience LOS problems if the number of through lanes is reduced, following the process described in Section L4 of the Guide. Usually, the signalized intersections on an urban street will be its critical capacity choke points. Therefore, this example will examine peak hour intersection v/c ratios at these intersections to determine which ones should be evaluated in more detail. Should there be some other capacity choke point (such as a lane drop, a narrow bridge or tunnel, a major shopping center driveway, or a major grade change) then those choke points should be checked as well. Exhibit 165 pro- vides lane configurations and peak hour turning-movement volumes for the major signalized intersections within supersection C.

226 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual Step 0: Assemble Data The turning-movement volumes and lane configurations for the six major intersections in supersection C were shown in Exhibit 165. According to Exhibit 60 in Section L3 of the Guide, the following additional data are required to evaluate capacity: • Peak hour factor, • Percent heavy vehicles, • Parking activity, and • Pedestrian activity. For the intersection of Telegraph and 51st Street, the traffic counts used to develop the turning- movement volumes are also used to identify the values for peak hour factor (0.92) and percent heavy vehicles (5%), and to characterize the pedestrian activity at the intersection as “Medium.” The analyst’s knowledge is used to determine that parking is allowed on 51st Street, but will not be allowed on Telegraph Avenue following construction of the exclusive BRT lane. The remain- der of this example will show the computations for this intersection. Computations for the other intersections are similar. Step 1: Determine Left-Turn Phasing The future left-turn phasing is not known. Therefore, the process for selecting left-turn phasing described in Section L4 of the Guide will be used. Protected left-turn phasing is selected if any of the following three conditions are met; otherwise, permitted left-turn phasing is selected: Exhibit 165. Case study 2: intersection lane configurations and turning-movement volumes.

U. Case Study 2: Arterial BRT Analysis 227 • Left-turn volume exceeds 240 veh/h; • The product of the left-turn volume and the opposing through volume exceeds a given thresh- old (50,000 if there is one opposing through lane, 90,000 if there are two opposing through lanes, and 110,000 if there are three or more opposing through lanes); or • The number of left-turn lanes exceeds one. If an approach has an exclusive left-turn lane, and its opposing approach meets at least one condition for protected left-turn phasing, then it will also be assumed to have protected left-turn phasing. Exhibit 166 shows the results of these checks. In this case, the southbound and eastbound approaches met one or more of the conditions for protected left-turn phasing. Because their opposing approaches (northbound and westbound, respectively) have exclusive left-turn lanes, the opposing approaches are also assumed to have protected left turns. Step 2: Identify Lane Groups The turn volumes are assigned to lane groups according the criteria given in Section L4 of the Guide: 1. When a traffic movement uses only an exclusive lane(s), it is analyzed as an exclusive lane group. 2. When two or more traffic movements share a lane, all lanes which convey those traffic move- ments are analyzed as a mixed lane group. By these criteria, all left-turn movements at this intersection are assigned to exclusive lane groups, while all through and right-turn movements are assigned to mixed lane groups. Multiple-lane mixed lane groups also need to be examined to determine whether a de facto turn lane exists, due to a high volume of turning traffic relative to through traffic. As shown in Exhibit 167, only the westbound and eastbound approaches have mixed lane groups with two or more lanes. The right-turn volumes on these approaches are small relative to the through volumes; no de facto turn lanes exist, and the original assignment of a mixed lane group is retained. Exhibit 166. Case study 2: protected left-turn checks for Telegraph Avenue/51st Street. Approach NB SB WB EB Check 1 Left-turn volume (veh/h) 83 283 89 261 Is the left-turn volume > 240 veh/h? No Yes No Yes Check 2 Opposing through volume (veh/h) 531 676 670 474 Left-turn volume × opposing volume 44,073 191,308 59,630 123,714 Number of opposing through lanes 1 1 2 2 Threshold for Check 2 50,000 50,000 90,000 90,000 Is product > threshold? No Yes No Yes Check 3 Left-turn lanes 1 1 1 2 Is there more than 1 left-turn lane? No No No Yes Check 4 Is there an exclusive left-turn lane? Yes Yes Yes Yes Does the opposite approach meet Check 1, 2, or 3? Yes No Yes No Result Protected left-turn phase? Yes Yes Yes Yes Note: NB = northbound, SB = southbound, WB = westbound, EB = eastbound.

228 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual Step 3: Convert Turning Movements to Through Passenger Car Equivalents This step converts turning movements to through passenger car equivalents, consider- ing the effect of heavy vehicles, variations in traffic flow during the hour, the impact of opposing through vehicles on permitted left-turning vehicles, the impact of pedestrians on right-turning vehicles, lane utilization, and the impact of parking maneuvers on through and right-turning vehicles. Step 3a: Heavy Vehicle Adjustment The adjustment for heavy vehicles EHVadj is calculated using Equation 75. E P EHVadj HV HV( ) ( )= + − = + − =1 1 1 0.05 2 1 1.05 Step 3b: Peak Hour Factor Adjustment The adjustment for variation in flow during the peak hour is calculated using Equation 76. E PHF PHF = = = 1 1 0.92 1.09 Step 3c: Turn Impedance Adjustment The turn impedance adjustment factors ELT and ERT adjust for impedances experienced by left- and right-turning vehicles, respectively. For protected left turns (the situation on all four intersection approaches), ELT = 1.05 regardless of volume. For permitted right turns (the typical situation), Exhibit 63 is used to determine the value of ERT. For a “Medium” level of pedestrian activity, ERT = 1.30. Step 3d: Parking Adjustment Factor The parking adjustment factor Ep is determined from Exhibit 64. For exclusive left-turn lanes and all movements on Telegraph Avenue, where no adjacent on-street parking is provided, Ep = 1.00. For the eastbound and westbound through lane groups on 51st Street, which each have two lanes and adjacent parking, Ep = 1.10. Step 3e: Lane Utilization Factor The lane utilization factor ELU is determined from Exhibit 65. For the northbound, south- bound, and westbound exclusive left-turn lanes, ELU = 1.00, as only one left-turn lane is pro- vided. Two exclusive lanes are provided for the eastbound left turn; therefore, its ELU = 1.03. In the northbound and southbound directions, one shared through–right lane is provided, with Exhibit 167. Case study 2: lane group determination for Telegraph Avenue/51st street. Northbound Southbound Westbound Eastbound L T R L T R L T R L T R Peak hour volume (veh/h) 83 676 59 283 531 22 89 474 105 261 670 84 Number of lanes 1 1 1 1 1 2 2 2 De facto exclusive lane? No No Lane group type Ex. Mixed Ex. Mixed Ex. Mixed Ex. Mixed Note: L = left, T = through, R = right, Ex. = Exclusive.

U. Case Study 2: Arterial BRT Analysis 229 ELU = 1.00. In the eastbound and westbound directions, two through or shared lanes are pro- vided, with ELU = 1.05. Step 3f: Adjustment Factor for Other Effects In the absence of information on other effects, Eother is set to the default 1.00. Step 3g: Through Passenger Car Equivalent Flow Rate The through passenger car equivalent flow rate vadj is calculated using Equation 78. For the northbound left-turn, the calculation is as follows: v VE E E E E E E v adj NBLT HVadj PHF LT RT p LU adj NBLT ( )( )( )( )( )( )( )( ) = = =83 1.05 1.09 1.05 1.00 1.00 1.00 1.00 100 tpc/h , other , Step 3h: Equivalent Per-Lane Flow Rate Finally, the equivalent per-lane flow rate vi for a given lane group i is calculated using Equa- tion 79. For the northbound left-turn, which operates in a single lane, the calculation is: 100 1 100 tpc/h/ln , v v N NBLT adj NBLT NBLT = = = Exhibit 168 shows the computation results for all lane groups. Step 4: Calculate Critical Lane Group Volumes Step 4a: Identify Critical Movements When opposing approaches use protected left-turn phasing, as is the case at this inter- section, Equation 80 and Equation 81 are used to determine the critical lane group volumes in the east–west and north–south directions, respectively. v v v v v v v c EW EBLT WBTH WBRT WBLT EBTH EBRT ( ) ( )= + +  = + +  =  =max max , max , max 323 404 107 515 max 727 622 727 tpc/h/ln, Exhibit 168. Case study 2: through passenger car equivalents for Telegraph Avenue/51st street. Northbound Southbound Westbound Eastbound L T R L T R L T R L T R Movement volume (veh/h) 83 676 59 283 531 22 89 474 105 261 670 84 Heavy vehicle adj., EHVadj 1.05 1.05 1.05 1.05 1.05 1.05 1.05 1.05 1.05 1.05 1.05 1.05 PHF adj., EPHF 1.09 1.09 1.09 1.09 1.09 1.09 1.09 1.09 1.09 1.09 1.09 1.09 Left-turn impedance adj., ELT 1.05 1.00 1.00 1.05 1.00 1.00 1.05 1.00 1.00 1.05 1.00 1.00 Right-turn impedance adj., ERT 1.00 1.00 1.30 1.00 1.00 1.30 1.00 1.00 1.30 1.00 1.00 1.30 Parking adj., Ep 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.10 1.10 1.00 1.10 1.10 Lane utilization adj., ELU 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.05 1.05 1.03 1.05 1.05 Other effects adj., Eother 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 Equivalent flow rate (tpc/h) 100 774 88 340 608 33 107 627 180 323 886 144 Number of lanes 1 1 1 1 1 2 2 2 Lane group type Ex. Mixed Ex. Mixed Ex. Mixed Ex. Mixed Lane group flow rate (tpc/h) 100 862 340 641 107 807 323 1,030 Equivalent flow rate (tpc/h/ln) 100 862 340 641 107 404 323 515 Note: L = left, T = through, R = right, adj. = adjustment, PHF = peak hour factor, Ex. = Exclusive.

230 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual Similarly for the north–south approaches, the critical volume Vc,NS is calculated using Equation 81. max max , max , max 100 641 340 862 max 741 1,202 1,202 tpc/h/ln,v v v v v v v c NS NBLT SBTH SBRH SBLT NBTH NBRH ( ) ( )= + +  = + +  =  = The critical lane group volumes for the intersection of Telegraph Avenue and 51st Street are: northbound through (862 tpc/h/ln); southbound left (340 tpc/h/ln); eastbound left (323 tpc/h/ln); and westbound through (404 tpc/h/ln). Step 4b: Calculate the Sum of the Critical Lane Group Volumes The sum of the critical lane group volumes is calculated using Equation 87. 727 1,202 1,929 tpc h ln, ,V v vc c EW c NS= + = + = Step 5: Compute Intersection Volume-to-Capacity Ratio The intersection v/c ratio is computed using Equation 88. The capacity is assumed to be the default 1,650 tpc/h/ln suggested in Section L4 of the Guide. 1,929 1,650 1.17X V c c c i = = = Applying the guidance provided in Exhibit 66, it is determined that the intersection will likely operate over capacity with the proposed lane geometry. The demands are likely to exceed the intersection’s capacity under most feasible signal timing plans. Interpretation of v/c Results for the Entire Facility An analysis of all six major intersections within supersection C found that Telegraph Avenue/51st Street is the only intersection where the predicted v/c ratio is expected to be over capacity following implementation of the BRT project (results for other intersections are pre- sented in Example 4). It is concluded that the Telegraph Avenue/51st Street intersection will likely have insufficient auto capacity under the proposed lane geometry with the BRT project in place. Consequently, a more detailed analysis using more site-specific data and fewer default values is recommended for the intersection, using actual or planned signal timings to compute delays and queues with the BRT project in place. 4. Example 3: Calculation of Intersection v/c Ratio for Permitted Left Turns Approach Example 2 showed the v/c ratio computations for an intersection with all protected left- turn phases. This example shows the computations for an intersection with all permitted left-turn phases, the intersection of Telegraph Avenue and Claremont Avenue. This exam- ple follows the process described in Section L4 of the Guide and uses the same steps as in Example 2.

U. Case Study 2: Arterial BRT Analysis 231 Step 0: Assemble Data The turning-movement volumes and lane configurations for this intersection were shown in Exhibit 165 in Example 2. According to Exhibit 60 in Section L3 of the Guide, the following additional data are required to evaluate capacity: • Peak hour factor, • Percent heavy vehicles, • Parking activity, and • Pedestrian activity. For the intersection of Telegraph Avenue and Claremont Avenue, the traffic counts used to develop the turning-movement volumes are also used to identify the values for peak hour factor (0.92) and percent heavy vehicles (5%), and to characterize the pedestrian activity at the inter- section as “Medium.” The analyst’s knowledge is used to determine that parking is allowed on Claremont Avenue, but will not be allowed on Telegraph Avenue following construction of the exclusive BRT lane. Step 1: Determine Left-Turn Phasing The future left-turn phasing is not known. Therefore, the process for selecting left-turn phasing described in Section L4 of the Guide will be used. Protected left-turn phasing is selected if any of the following three conditions are met; otherwise, permitted left-turn phasing is selected: • Left-turn volume exceeds 240 veh/h; • The product of the left-turn volume and the opposing through volume exceeds a given thresh- old (50,000 if there is one opposing through lane, 90,000 if there are two opposing through lanes, and 110,000 if there are three or more opposing through lanes); or • The number of left-turn lanes exceeds one. If an approach has an exclusive left-turn lane, and its opposing approach meets at least one condition for protected left-turn phasing, then it will also be assumed to have protected left-turn phasing. Exhibit 169 shows the results of these checks. In this case, none of the approaches met one or more of the conditions for protected left-turn phasing. Therefore, permitted left-turn phasing will be assumed for all approaches. Step 2: Identify Lane Groups The turn volumes are assigned to lane groups according the criteria given in Section L4 of the Guide: 1. When a traffic movement uses only an exclusive lane(s), it is analyzed as an exclusive lane group. 2. When two or more traffic movements share a lane, all lanes which convey those traffic move- ments are analyzed as a mixed lane group. By these criteria, the southbound and westbound left-turn movements at this intersection are assigned to exclusive lanes. All other movements are assigned to mixed lane groups. Multiple- lane mixed lane groups also need to be examined to determine whether a de facto turn lane exists, due to a high volume of turning traffic relative to through traffic. As shown in Exhibit 170, only the northbound approach has a mixed lane group with multiple lanes. In this case, both the

232 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual right-turn and left-turn volumes on the approach are small relative to the through volume, so no de facto turn lanes exist and the original assignment of a mixed lane group is retained. Step 3: Convert Turning Movements to Passenger Car Equivalents This step converts turning movements to through passenger car equivalents, considering the effect of heavy vehicles, variations in traffic flow during the hour, the impact of opposing through vehicles on permitted left-turning vehicles, the impact of pedestrians on right-turning vehicles, lane utilization, and the impact of parking maneuvers on through and right-turning vehicles. Step 3a: Heavy Vehicle Adjustment The adjustment for heavy vehicles EHVadj is calculated using Equation 75. 1 1 1 0.05 2 1 1.05E P EHVadj HV HV( ) ( )= + − = + − = Step 3b: Peak Hour Factor Adjustment The adjustment for variation in flow during the peak hour is calculated using Equation 76. 1 1 0.92 1.09E PHF PHF = = = Approach NB SB WB EB Check 1 Left-turn volume (veh/h) 8 61 114 12 Is the left-turn volume > 240 veh/h? No No No No Check 2 Opposing through volume (veh/h) 717 864 5 61 Left-turn volume × opposing volume 5,736 52,704 570 732 Number of opposing through lanes 1 2 1 1 Threshold for Check 2 50,000 90,000 50,000 50,000 Is product > threshold? No No No No Check 3 Left-turn lanes 1 1 1 1 Is there more than 1 left-turn lane? No No No No Check 4 Is there an exclusive left-turn lane? No Yes Yes No Does the opposite approach meet Check 1, 2, or 3? No No No No Result Protected left-turn phase? No No No No Note: NB = northbound, SB = southbound, WB = westbound, EB = eastbound. Exhibit 169. Case study 2: protected left-turn checks for Telegraph Avenue/Claremont Avenue. Peak hour volume (veh/h) 8 864 170 61 717 69 114 61 77 12 5 5 Number of lanes 2 1 1 1 1 1 De facto exclusive lane? No Lane group type Mixed Ex. Mixed Ex. Mixed Mixed Note: L = left, T = through, R = right, Ex. = Exclusive. Northbound Southbound Westbound Eastbound L T R L T R L T R L T R Exhibit 170. Case study 2: lane group determination for Telegraph Avenue/Claremont Avenue.

U. Case Study 2: Arterial BRT Analysis 233 Step 3c: Turn Impedance Adjustment The turn impedance adjustment factors ELT and ERT adjust for impedances experienced by left- and right-turning vehicles, respectively. All left turns are permitted movements and therefore the left-turn adjustment factor will take on significantly higher values than in Example 2. Exhibit 62 is used to determine the value for permitted left turns, using the sum of the opposing through and right-turn volumes. For the northbound left turn, the sum of the southbound through and right-turn volumes is 786 veh/h and therefore ELT = 3.00 for the northbound left-turn. Similarly, ELT values for the southbound, westbound, and eastbound left turns are 5.00, 1.10, and 1.10, respectively. For permitted right turns (the typical situation), Exhibit 63 is used to determine the value of ERT. For a “Medium” level of pedestrian activity, ERT = 1.30. Step 3d: Parking Adjustment Factor The parking adjustment factor Ep is determined from Exhibit 64. For exclusive left-turn lanes and all movements on Telegraph Avenue, where no adjacent on-street parking is provided, Ep = 1.00. For the eastbound and westbound through lane groups on Claremont Avenue, which each have one lane and adjacent parking, Ep = 1.20. Step 3e: Lane Utilization Factor The lane utilization factor ELU is determined from Exhibit 65. For the southbound and west- bound exclusive left-turn lanes, ELU = 1.00, as only one left-turn lane is provided. The north- bound approach provides two shared lanes, with ELU = 1.05. All other movements occur in single shared lanes, with ELU = 1.00. Step 3f: Adjustment Factor for Other Effects In the absence of information on other effects, Eother is set to the default 1.00. Step 3g: Through Passenger Car Equivalent Flow Rate The through passenger car equivalent flow rate vadj is calculated using Equation 78. For the southbound left turn, the calculation is as follows: 61 1.05 1.09 5.00 1.00 1.00 1.00 1.00 349 tpc h , other , v VE E E E E E E v adj SBLT HVadj PHF LT RT p LU adj SBLT ( )( )( )( )( )( )( )( ) = = = Step 3h: Equivalent Per-Lane Flow Rate Finally, the equivalent per-lane flow rate vi for a given lane group i is calculated using Equa- tion 79. For the southbound left turn, which operates in a single lane, the calculation is: 349 1 349 tpc h ln , v v N SBLT adj SBLT SBLT = = = Exhibit 171 shows the computation results for all lane groups. Step 4: Calculate Critical Lane Group Volumes Step 4a: Identify Critical Movements When opposing approaches use permitted phasing, as is the case at this intersection, the critical lane volume will be the highest lane volume of all lane groups for a pair of approaches. For the

234 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual east–west approaches, the critical volume Vc,EW is calculated using Equation 82, while the critical volume for the north–south approaches Vc,NS is calculated using Equation 83. max , , , , , max 0,31,0,144,221,0 221 tpc h ln,V v v v v v vc EW EBLT EBTH EBRT WBLT WBTH WBRT( ) ( )= = = max , , , , , max 0,667,0,349,924,0 924 tpc h ln,V v v v v v vc NS NBLT NBTH NBRT SBLT SBTH SBRT( ) ( )= = = The critical lane group volumes for the intersection of Telegraph Avenue and Claremont Ave- nue are southbound through (924 tpc/h/ln) and westbound through (221 tpc/h/ln). Step 4b: Calculate the Sum of the Critical Lane Group Volumes The sum of the critical lane group volumes is calculated using Equation 87. 221 924 1,145 tpc h ln, ,V v vc c EW c NS= + = + = Step 5: Compute Intersection Volume-to-Capacity Ratio The intersection v/c ratio is computed using Equation 88. The capacity is assumed to be the default 1,650 tpc/h/ln suggested in Section L4 of the Guide. 1,145 1,650 0.69X V c c c i = = = Applying the guidance provided in Equation 66, it is determined that the intersection will likely operate under capacity with the proposed lane geometry. 5. Example 4: Estimating Auto and BRT Speeds Approach This example determines auto delay, auto travel times, and auto and bus speeds for super- section C along Telegraph Avenue. First, control delay is calculated for the northbound and southbound through movements on Telegraph Avenue at each intersection, following the guid- ance in Section L5 of the Guide. Next, these control delays are used as inputs to the simplified HCM urban street analysis method described in Section K6 to determine auto speeds and travel Exhibit 171. Case study 2: through passenger car equivalent calculations. Movement volume (veh/h) 8 864 170 61 717 69 114 61 77 12 5 5 Heavy vehicle adj., EHVadj 1.05 1.05 1.05 1.05 1.05 1.05 1.05 1.05 1.05 1.05 1.05 1.05 PHF adj., EPHF 1.09 1.09 1.09 1.09 1.09 1.09 1.09 1.09 1.09 1.09 1.09 1.09 Left-turn impedance adj., ELT 3.00 1.00 1.00 5.00 1.00 1.00 1.10 1.00 1.00 1.10 1.00 1.00 Right-turn impedance adj., ERT 1.00 1.00 1.30 1.00 1.00 1.30 1.00 1.00 1.30 1.00 1.00 1.30 Parking adj., Ep 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.20 1.20 1.00 1.20 1.20 Lane utilization adj., ELU 1.05 1.05 1.05 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 Other effects adj., Eother 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 Equivalent flow rate (tpc/h) 29 1,038 266 349 821 103 144 84 137 15 7 9 Number of lanes 2 1 1 1 1 1 Lane group type Mixed Ex. Mixed Ex. Mixed Mixed Lane group flow rate (tpc/h) 1,333 349 924 144 221 31 Equivalent flow rate (tpc/h/ln) 667 349 924 144 221 31 Note: L = left, T = through, R = right, adj. = adjustment, PHF = peak hour factor, Ex. = Exclusive. Northbound Southbound Westbound Eastbound L T R L T R L T R L T R

U. Case Study 2: Arterial BRT Analysis 235 times within each section and for supersection C as a whole. Finally, the transit speed method for urban streets described in Section O4 is used to estimate BRT speeds. Signalized Intersection Control Delay This portion of the example continues the calculations started in Example 3 for the inter- section of Telegraph Avenue and Claremont Avenue. Summary results for the other five signal- ized intersections within supersection C are presented at the end of this step. Step 6: Calculate Capacity Step 6a: Calculate Cycle Length The traffic signal cycle length C is assumed to be 30 seconds per critical phase, per Equation 89. As permitted left turns are used at the Telegraph Avenue/Claremont Avenue intersection, there are two critical phases, and C would be calculated as 60 seconds. However, at an intersection with protected left turns on all approaches, such as Telegraph Avenue/51st Street, there would be four critical phases, and C would be 120 seconds. Similarly, three-legged intersection with a protected left turn on the main street, such as Telegraph Avenue/48th Street, would have three critical phases and C would be calculated as 90 seconds. As urban street facilities are typically timed to provide a common cycle length at all intersections within the facility, the largest calcu- lated cycle length (120 seconds) will be assumed to be the cycle length used at all intersections along the facility. Step 6b: Calculate the Total Effective Green Time The total effective green time gTOT available during the cycle is calculated using Equation 90. In the absence of other information, the default value of 4 seconds per critical phase for lost time per cycle is used. For the Telegraph Avenue/Claremont Avenue intersection, gTOT is then: 120 8 112 sg C LTOT = − = − = The total effective green time is then allocated to each critical phase in proportion to the critical lane group volume for that movement using Equation 91. From Example 3, Step 4a, the critical lane group volumes are southbound through (924 tpc/h/ln) and westbound through (221 tpc/h/ln). From Example 3, Step 4b, the critical intersection volume is 1,145 tpc/h/ln. Then: 112 924 1,145 90.4 sg g V V SBTH TOT cSBTH c =   =   = 112 221 1,145 21.6 sg g V V WBTH TOT cWBTH c =   =   = The effective green time for the non-critical phases (and the movements served by those phases) is set equal to the green time for the phase on the opposing approach that serves the same move- ments. In this case, with only two critical phases, all northbound and southbound movements are initially assigned an effective green time of 90.4 seconds, while all westbound and eastbound movements are initially assigned an effective green time of 21.6 seconds. The green time for each phase should be reviewed against considerations such as the minimum green time and the time required for pedestrians to cross the approach, as stated in local policy and standards such as the Manual on Uniform Traffic Control Devices (FHWA 2009). In this case, on the basis of the length of the Telegraph Avenue crosswalks and local policy on minimum pedestrian Walk time, a minimum of 27.0 seconds for the combined pedestrian Walk and flashing Don’t Walk intervals is required to cross Telegraph Avenue, which translates into a minimum effective green time

236 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual of 23.0 seconds for the parallel through movements. Therefore, the effective green time assigned to the westbound and eastbound movements is increased to 23.0 seconds and the effective green time assigned to the northbound and southbound movements is decreased to 89.0 seconds. Step 6c: Calculate Capacity and Volume-to-Capacity Ratio The capacity ci and volume-to-capacity ratio Xi for each critical lane group i are calculated using Equation 92 and Equation 93. As the study area is located in a large metropolitan area with millions of residents, a base saturation flow rate of 1,900 pc/h/ln is used in Equation 92. For the southbound through lane group at Telegraph Avenue and Claremont Avenue, the capacity and v/c ratio are: 1,900 89.0 120 1,409 tpc h lnc BaseSat g C SBTH SBTH =     =     = 924 1,409 0.66X v c SBTH SBTH SBTH = = = For the westbound through lane group, the capacity and v/c ratio are: 1,900 23.0 120 364 tpc h lnc BaseSat g C WBTH WBTH =     =     = 221 364 0.61X v c WBTH WBTH WBTH = = = For the intersection as a whole, the critical degree of saturation Xc is calculated using Equation 94 and Equation 95. 1,900 1,900 89.0 23.0 120 1,773 tpc hSUM 1c g C cii n∑ =     = +    == 924 221 1,773 0.65c 1X v c cii n SUM ∑ = = + = = The initial finding in Example 3 that this intersection will operate below capacity is confirmed. Summary Capacity Results for Supersection C Exhibit 172 summarizes the results of the capacity calculations for all of the signalized intersec- tions within supersection C. As can be seen from the table, only the Telegraph Avenue/51st Street intersection is expected to operate over capacity. Step 7: Estimate Delay The control delay for di for the northbound and southbound lane groups on Telegraph Avenue is calculated using Equation 96. This equation requires first computing the uniform delay d1 with Equation 97 and the incremental delay d2 with Equation 98. There are no unsignalized turning movements at any intersection within supersection C, so unsignalized movement delay dunsig is zero. The following presents the calculation details for the northbound and southbound through lane groups at the Claremont Avenue intersection; summary results for the northbound and southbound lane groups at all intersections in supersection C follow. Uniform delay for the southbound through lane group is determined using the southbound movement volume from Step 3 (Exhibit 171) and the critical north–south effective green time

U. Case Study 2: Arterial BRT Analysis 237 from Step 6b (Exhibit 172). The volume-to-capacity ratio X for the southbound direction is 924 / 1,409 = 0.66. The calculation proceeds as follows: 0.5 1 1 min 1, 0.5 120 1 89 120 1 min 1, 924 1,409 89 120 7.8 s1 2 2 d C g C X g C[ ] [ ][ ] ( ) ( ) ( )( ) ( )( )( ) ( )= − − = − − = Incremental delay for the southbound through lane group is calculated as follows: 225 1 1 16 225 0.66 1 0.66 1 16 0.66 1,409 2.4 s2 2 2d x x X c ( ) ( ) ( ) ( )= − + − +   = − + − + ×    = In the absence of other information, average signal progression is assumed, and a value of 1.00 is obtained from Exhibit 68 for the progression factor PF. The control delay for the southbound lane group is then: 7.8 1.00 2.4 0.0 10.2 s1 2d d PF d di unsig ( )( )= + + = + + = Similarly, for the northbound lane group, the v/c ratio is 0.47; the uniform delay is 6.1 s; the incremental delay is 1.1 s; and the control delay is 7.2 s. The calculations for the other intersections are conducted similarly. However, the 3-leg intersec- tions with protected southbound left turns on Telegraph Avenue (48th and 49th Streets) are an excep- tion to the general rule that non-critical phases (the southbound through in these cases) are assigned the same effective green time as their counterpart critical phase (i.e., the northbound through). As Exhibit 172. Case study 2: capacity calculations for supersection C. Cross Street 45th 48th 49th 51st Claremont 55th Number of critical phases 2 3 3 4 2 3 Cycle length, C (s) 120 120 120 120 120 120 Total effective green time, gTOT (s) 112 108 108 104 112 108 Critical EW TH volume, VEWTH,C (tpc/h) 188 77 170 404 221 518 Critical EW LT volume, VEWLT,C (tpc/h) 0 0 0 323 0 0 Critical NS TH volume, VNSTH,C (tpc/h) 759 651 992 862 924 972 Critical NS LT volume, VNSLT,C (tpc/h) 0 125 95 340 0 117 Sum of critical volumes, VC (tpc/h) 947 853 1,257 1,929 1,145 1,607 Critical EW TH effective green, gEWTH,C (s) 22.2 9.7 14.6 21.8 21.6 34.8 Adj. critical EW TH effective green, gEWTH,C (s) 23.0 23.0 23.0 23.0 23.0 34.8 Critical EW LT effective green, gEWLT,C (s) 0.0 0.0 0.0 17.4 0.0 0.0 Critical NS LT effective green, gNSLT,C (s) 0.0 15.8 8.2 18.3 0.0 7.9 Critical NS TH effective green, gNSTH,C (s) 89.8 82.5 85.2 46.5 90.4 65.3 Adj. critical NS TH effective green, gNSTH,C (s) 89.0 69.2 76.8 45.3 89.0 65.3 Critical EW TH capacity, cEWTH,C (tpc/h) 364 364 364 364 364 551 Critical EW LT capacity, cEWLT,C (tpc/h) 276 Critical NS TH capacity, cNSTH,C (tpc/h) 1,409 1,096 1,216 717 1,409 1,034 Critical NS LT capacity, cNSLT,C (tpc/h) 250 130 290 125 Intersection capacity, cSUM (tpc/h) 1,773 1,710 1,710 1,647 1,773 1,710 EW TH volume-to-capacity ratio, XEWTH 0.52 0.21 0.47 1.11 0.61 0.94 EW LT volume-to-capacity ratio, XEWLT 1.17 NS TH volume-to-capacity ratio, XNSTH 0.54 0.59 0.82 1.20 0.66 0.94 NS LT volume-to-capacity ratio, XNSLT 0.50 0.73 1.17 0.94 Intersection volume-to-capacity ratio, Xc 0.53 0.50 0.74 1.17 0.65 0.94 Note: EW = east–west, NS = north–south, TH = through, LT = left turn, adj. = adjusted.

238 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual there is no northbound left turn at these intersections, the southbound through movement can also move while the southbound left turn is being served; therefore, the effective green time used for the southbound through is the sum of the southbound left turn and northbound through effective green. Exhibit 173 provides the northbound and southbound delay calculation results for all signal- ized intersections in supersection C. Step 8: Level of Service Comparing the calculated control delays from Step 7 to the values given in Exhibit 69, it is found that the northbound through movement would operate at LOS A, while the southbound through movement would operate at LOS B. Exhibit 173 provides the northbound and south- bound LOS results for all signalized intersections in supersection C. It can be seen that through movements on Telegraph Avenue will operate at LOS D or better at all intersections following implementation of BRT, except at 51st Street, where the northbound direction will operate at LOS F. In addition, the northbound movement at 55th Street will operate close to capacity. Section Travel Times and Speeds (Auto) This portion of the example presents the speed and travel time calculations for automobiles on northbound Telegraph Avenue between 45th and 48th Streets. Summary results for both directions of Telegraph Avenue through supersection C follow. Step 1: Calculate Running Time The running time tR for the section is calculated using Equation 58. From Example 1, the posted speed in supersection C is 30 mph. The default user-selected adjustment to the posted speed is 5 mph. The segment length is 655 feet. Then: 3,600 5,280 3,600 655 5,280 30 5 12.8 st L S UserAdj R pl( ) ( )= × × + = × × + = Steps 2–4: Calculate the Downstream Intersection Capacity, Volume-to-Capacity Ratio, and Delay These steps were completed previously as part of the signalized intersection calculations described above. From Exhibit 173, the control delay d in the northbound direction is 18.6 seconds. Exhibit 173. Case study 2: delay and LOS calculations for supersection C. 45th 48th 49th 51st Claremont 55th NB SB NB SB NB SB NB SB NB SB NB SB Cycle length, C (s) 120 120 120 120 120 120 120 120 120 120 120 120 Effective green time, g (s) 89.0 89.0 69.2 85.0 76.8 85.0 45.3 45.3 89.0 89.0 65.3 65.3 Lane group volume, V (tpc/h/ln) 759 574 651 578 992 699 862 641 667 924 1,028 972 Lane group capacity, c (tpc/h/ln) 1,409 1,409 1,096 1,346 1,216 1,346 717 717 1,409 1,409 1,034 1,034 Volume-to-capacity ratio, X 0.54 0.41 0.59 0.43 0.82 0.52 1.20 0.89 0.47 0.66 0.99 0.94 Uniform delay, d1 (s) 6.7 5.8 16.3 7.3 16.4 8.1 37.4 35.0 6.1 7.8 27.0 25.5 Incremental delay, d2 (s) 1.5 0.9 2.3 1 6.3 1.4 103.1 15.5 1.1 2.4 25.7 16.8 Progression quality Avg. Avg. Avg. Avg. Avg. Avg. Avg. Avg. Avg. Avg. Avg. Avg. Progression factor, PF 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 Control delay, d (s) 8.2 6.7 18.6 8.3 22.7 9.5 140.5 50.5 7.2 10.2 52.7 42.3 LOS A A B A C A F D A B D D Note: Avg. = average.

U. Case Study 2: Arterial BRT Analysis 239 Step 5: Compute the Average Travel Speed and Determine Level of Service The average travel time on the segment TT is calculated using Equation 64: 12.8 18.6 31.4 sT t dT R= + = + = The average travel speed on the segment ST,seg is calculated using Equation 65: 3,600 5,280 3,600 655 5,280 31.4 14.2 mph,S L T T seg T = × × = × × = From Exhibit 52, this speed corresponds to LOS D. Summary Speed Results for Supersection C Travel times, speed, and LOS results by section are presented in Exhibit 174 and Exhibit 175 for the northbound and southbound directions, respectively. For the northbound direction of supersection C, the average travel time is 293.7 seconds, the average speed is 6.2 mph, and the LOS is F. In the southbound direction, the average travel time is 137.2 seconds, the average speed is 13.3 mph, and the LOS is E. Section Travel Speeds (BRT) The BRT project includes exclusive bus-only lanes between traffic signals, with right-turning traffic allowed into the bus lanes at traffic signals. There will only be one BRT stop within Exhibit 174. Case study 2: travel time, speed, and LOS calculations for supersection C (northbound). Downstream Intersection 48th 49th 51st Claremont 55th Speed limit, Spl (mph) 30 30 30 30 30 User adjustment, UserAdj (mph) 5 5 5 5 5 Segment length, L (ft) 655 468 478 269 798 Running time, tR (s) 12.8 9.1 9.3 5.2 15.6 Control delay, d (s) 18.6 22.7 140.5 7.2 52.7 Volume-to-capacity ratio, X 0.59 0.82 1.20 0.47 0.99 Segment travel time, Tt (s) 31.4 31.8 149.8 12.4 68.3 Segment speed, ST,seg 14.2 10.0 2.2 14.8 8.0 LOS D F F D F Exhibit 175. Case study 2: travel time, speed, and LOS calculations for supersection C (southbound). Downstream Intersection 45th 48th 49th 51st Claremont Speed limit, Spl (mph) 30 30 30 30 30 User adjustment, UserAdj (mph) 5 5 5 5 5 Segment length, L (ft) 655 468 478 269 798 Running time, tR (s) 12.8 9.1 9.3 5.2 15.6 Control delay, d (s) 6.7 8.3 9.5 50.5 10.2 Volume-to-capacity ratio, X 0.41 0.43 0.52 0.89 0.66 Segment travel time, Tt (s) 19.5 17.4 18.8 55.7 25.8 Segment speed, ST,seg 22.9 18.3 17.3 3.3 21.1 LOS C C D F C

240 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual supersection C, at 49th Street. Section O4 of the Guide provides two options for estimating bus speeds: (1) a generalized method and (2) a modified version of the auto method that was used above. Because Option 2 only applies to buses running in mixed traffic, and not to buses running in exclusive lanes, Option 1 will be followed. Step 1: Unimpeded Bus Travel Time Rate This step calculates the bus’ speed without traffic signal delays, but including deceleration, dwell time, and acceleration delays due to bus stops. The process uses Equation 152 through Equation 157. First, the time for a bus to decelerate from its running speed to a stop tdec is cal- culated from Equation 157, assuming a running speed between stops equal to the posted speed (30 mph) and the default bus deceleration rate of 4.0 feet per second per second from Exhibit 104: 1.47 1.47 30 4.0 11.0 st v d dec run = = × = Second, the time for a bus to accelerate back to its running speed tacc is calculated from Equa- tion 156 in a similar manner, but using the default bus acceleration rate of 3.4 feet per second per second from Exhibit 104. 1.47 1.47 30 3.4 13.0 st v a acc run = = × = Third, the distance traveled during acceleration and deceleration associated with each bus stop Lad is calculated using Equation 155. 0.5 0.5 0.5 3.4 13.0 0.5 4.0 11.0 529 ft2 2 2 2L at dtad acc dec ( ) ( )= + = × ×  + × ×  = Fourth, the portion of each mile of route traveled at running speed Lrs is calculated from Equa- tion 154. This equation uses the average bus stop spacing to determine how often a bus must accelerate or decelerate to serve a stop. In this case, there is one stop within supersection C and the next-closest bus stops are at 40th and 59th Streets, giving an average stop spacing of 2,900 feet (1.82 stops per mile). Then: 5,280 5,280 1.82 529 4,317 ftL N Lrs s ad ( )= − = − × = Fifth, the time spent per mile traveling at running speed trs is determined from Equation 153: 1.47 4,317 1.47 30 97.9 st L v rs rs run = = × = Finally, the unimpeded running time rate tu is determined from Equation 152, applying the default average critical stop dwell time of 30 seconds for urban areas from Exhibit 104. (The critical stop dwell time is used as this is the only BRT stop within supersection C.) 60 97.9 1.82 30 13.0 11.0 60 3.3 min mit t N t t t u rs s dt acc dec( ) ( ) = + + + = + + + = Step 2: Additional Bus Travel Time Delays This step estimates additional bus dwell time delays tl in the exclusive bus lane due to traffic sig- nal and traffic interference, using Exhibit 109. As supersection C is not located within Oakland’s central business district, the “Arterial Roadways Outside the CBD” portion of the exhibit is used, and a value of 0.7 minutes per mile is identified for bus lanes.

U. Case Study 2: Arterial BRT Analysis 241 Step 3: Base Bus Speed The two running time rates calculated in Steps 1 and 2 are combined into a base bus running time rate tr using Equation 160 and a base bus speed Sb using Equation 161. 3.3 0.7 4.0 min mit t tr u l= + = + = 60 60 4.0 15 mphS t b r = = = This estimated base bus speed applies to both directions, northbound and southbound, as all the inputs used to calculate it are the same in both directions. Step 4: Average Bus Speed The bus rapid transit line is projected to operate initially at 12-minute headways (i.e., 5 buses per hour per direction). In addition, a local version of the route that stops more frequently will operate in the bus lanes at 15-minute headways (i.e., 4 buses per hour per direction). The total number of buses is 9 per hour. From the discussion of Step 4 in Section O4 of the Guide, bus congestion at bus stops typically only affects bus speeds when more than 10–15 buses per hour are scheduled. As the number of scheduled buses is 9 per hour is less than the 10–15 per hour threshold, the bus–bus interference factor fbb is set to 1.00, and the estimated average bus speed equals the base bus speed, namely 15 mph. Interpretation of Speed Results The low auto speed of 6 mph in the northbound direction of Telegraph Avenue is a result of congestion in the sections ending at 49th, 51st, and 55th Streets. The southbound speed of 13 mph is better, but still affected by congestion in the section ending at 51st Street. Average BRT speeds of 15 mph in each direction demonstrate the combined benefit of the exclusive bus lane and the long stop spacing, particularly in the northbound (peak) direction. 6. Example 5: Predicting Queue Hot Spots Approach This short example demonstrates the estimation of queue lengths for the northbound and southbound through movements on Telegraph Avenue in supersection C, following Step 9 of the simplified signalized intersection method described in Section L5 of the Guide. Calculation The deterministic average queue for each lane group Q is determined by dividing the aver- age uniform delay for that lane group by the capacity for that lane group, using Equation 99. This equation uses the uniform delay d1 for the lane group and the per-lane capacity of the lane group c, both of which were computed as part of Example 4 and presented in Exhibit 173. Note that this estimate provides the average queue at the end of red, and is only applicable to lane groups that operate under capacity. For limited storage situations (such for left turn bays or short block lengths), it is desirable to provide storage for random fluctuations in-vehicle arrivals from cycle to cycle. In such a case, a 95th percentile probable queue may be used, estimated as twice the average queue. Results should be rounded to whole vehicles.

242 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual For northbound Telegraph Avenue at 55th Street, the average queue computation is as follows: 3,600 27.0 1,034 3,600 7.8 8 tpc ln 1 Q d c = × = × = → The estimated 95th percentile queue would be 2 × 7.8 = 15.6 tpc/ln, which rounds to 16 tpc/ln. Exhibit 176 summarizes the results for both directions of Telegraph Avenue in supersection C. At an average length of 25 feet per vehicle, the 95th-percentile southbound queue approaching 51st Street would be 350 feet and would exceed the 269-foot spacing between the signalized inter- sections. At 55th Street, the 95th-percentile queue lengths are estimated to be up to 400 feet long, which might require checking for potential blockage issues. The northbound queue at 51st Street cannot be calculated, as this movement operates over capacity; however, it can be determined from Exhibit 173 that the unserved demand at the end of the p.m. peak hour would be 145 tpc/ln (i.e., the difference between the movement demand and the movement capacity). 7. Example 6: Pedestrian, Bicycle, and Transit LOS Approach This example demonstrates the computation of pedestrian, bicycle, and transit LOS for super- section C, following the methods presented in Section O4 of the Guide for urban streets. The BRT project will convert one of the two travel lanes in each direction, plus the parking lane, into an exclusive bus lane and bicycle lane. The remaining width will be added to the sidewalk area and, in most sections, will be used to create an improved landscape buffer area. Link LOS (i.e., between traffic signals) will be calculated for the pedestrian and bicycle modes, to reduce the computational effort while providing a reasonable proxy for a section-level analysis. As pedestrian volumes were determined to be “Medium” in previous examples, an evaluation of pedes- trian density (i.e., sidewalk crowding) will not be performed. Section and facility LOS will be calcu- lated for the transit mode. Detailed calculations will be presented for the northbound direction of Telegraph Avenue (including the adjacent sidewalk) between 45th and 48th Streets. Summary results will then be presented for both directions of Telegraph Avenue for all sections of supersection C. Pedestrian LOS Input Data Exhibit 98 lists the input data requirements for calculating pedestrian LOS. For a link-level analysis, only the number of through travel lanes for motorized vehicles and the directional vehicular volume are required, both of which are available from previous examples in this case Lane group capacity, c (tpc/h/ln) 1,409 1,409 1,096 1,346 1,216 1,346 717 717 1,409 1,409 1,034 1,034 Volume-to-capacity ratio, X 0.54 0.41 0.59 0.43 0.82 0.52 1.20 0.89 0.47 0.66 0.99 0.94 Uniform delay, d1 (s) 6.7 5.8 16.3 7.3 16.4 8.1 37.4 35.0 6.1 7.8 27.0 25.5 Deterministic avg. queue (tpc/ln) 3 2 5 3 6 3 ** 7 2 3 8 7 95th percentile queue (tpc/ln) 5 4 10 5 11 6 ** 14 5 6 16 15 Note: Avg. = average. **Cannot be calculated,v/c ratio > 1. 45th 48th 49th 51st Claremont 55th NB SB NB SB NB SB NB SB NB SB NB SB Exhibit 176. Case study 2: deterministic average queues in supersection C.

U. Case Study 2: Arterial BRT Analysis 243 study. (Pedestrian signal delay data shown in Exhibit 98 are only required for a section-level analysis, and the segment length is only required for a facility-level analysis.) All other inputs can be defaulted, but will be substituted with actual values when known. Actual values are known or planned as part of the BRT project for the following inputs: • Sidewalk width by section; • Street tree presence by section; • Landscape buffer width by section; • On-street parking (none); • Outside travel lane width (12 feet); and • Bicycle lane width (5 feet). Calculation The pedestrian LOS score is calculated using Equation 148. It takes the following input values: • The distance from the inner edge of the outside lane to the curb WT is illustrated in Exhibit 99. In this case, it is equal to the bus lane width (12 feet) plus the bicycle lane width (5 feet), for a total of 17 feet in all sections. • The distance from the outer edge of the outside lane to the curb W1 is illustrated in Exhibit 99. In this case, it is equal to the bicycle lane width (5 feet) in all sections. • The on-street parking coefficient fp is always 0.50. • The percentage of the section with occupied on-street parking %OSP is zero in all sections, as on-street parking is prohibited in all sections. • The buffer area coefficient fB is 5.37 for the section between 45th and 48th Streets, as street trees will be provided within this section’s landscape buffer. • The landscape buffer width WB in the section between 45th and 48th Streets will be 5 feet. • The actual sidewalk width in the section between 45th and 48th Streets will be 12 feet, but the sidewalk width value WS is capped at 10 feet. • The sidewalk presence coefficient fSW equals 6 – 0.3WS, which results in a value of 5.7. • Traffic volumes are normally assumed to be evenly distributed between the available travel lanes, and the calculation normally divides the directional traffic volume by the number of travel lanes. However, because the travel lane closest to the curb will be used as an exclusive bus lane, the total bus volume (9 buses per hour) will be used for V and the number of bus lanes (1) will be used for N. • As the volume in the exclusive bus lane is less than or equal to 160 vehicles per hour, the low- volume factor fLV is calculated as 2.00 – 0.005V, which results in a value of 1.96. • The average vehicle speed between intersections is assumed to be the bus running speed deter- mined in Example 4, 30 mph. Pedestrian LOS for the northbound section between 45th and 48th Streets is then calculated as follows: 1.2276 ln 0.5 0.5 % 0.0091 4 0.0004 6.0468 1.2276 ln 1.96 17 0.5 5 0.5 0 5.37 5 3 10 0.0091 9 4 1 0.0004 35 6.0468 0.87 1 2 2 PLOS f W W OSP f W f W V N SPD PLOS PLOS LV T B B SW s[ ] [ ] [ ]( )[ ] [ ] [ ] [ ] [ ] [ ] [ ]( ) ( ) ( ) = − × × + × + × + × + × + + × + = − × × + × + × + × + × + × × + × + = From Exhibit 100, this PLOS score produces pedestrian LOS A.

244 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual Summary Pedestrian LOS Results for Supersection C (Northbound) Exhibit 177 presents the summary results for each northbound section in supersection C, fol- lowing implementation of the BRT project. As can be seen, all sections will operate at pedestrian LOS A, a result of the exclusive bus lane producing low traffic volumes in the travel lane closest to the curb and therefore a wide separation between pedestrians and traffic. For comparison, Exhibit 178 presents the summary results for the “before” condition. The following changes are made to the inputs: • With both travel lanes available to general traffic, the normal process described in Section O4 of the Guide is used to determine the directional volume V and the number of lanes N. The average of the directional volume departing one intersection and the volume arriving at the next downstream intersection, from Exhibit 165, is used to determine V. • Lane widths, landscape buffer widths, and street tree presence take on their existing condi- tions values. • A default value of 50% is assumed for occupied on-street parking, except in the section between 51st and Claremont, where on-street parking is prohibited. • The vehicular free-flow speed of 35 mph used in previous examples is used for the vehicle running speed. It can be seen that in the “before” condition, pedestrian LOS was in the B to C range. There- fore, it is concluded that the proposed project will result in an improvement in pedestrian LOS. Bicycle LOS Input Data Exhibit 101 lists the input data requirements for calculating bicycle LOS. For a link-level analy- sis, only the number of through travel lanes for motorized vehicles and the directional vehicular volume are required, both of which are available from previous examples in this case study. (Inter- Exhibit 177. Case study 2: pedestrian LOS calculations for supersection C (northbound with BRT). Downstream Intersection 48th 49th 51st Claremont 55th Outside lane volume, V (veh/h) 9 9 9 9 9 Low-volume factor, fLV 1.96 1.96 1.96 1.96 1.96 Outside lane width (ft) 12 12 12 12 12 Bicycle lane width (ft) 5 5 5 5 5 Parking lane width (ft) 0 0 0 0 0 Width WT (ft) 17 17 17 17 17 Width W1 (ft) 5 5 5 5 5 Occupied on-street parking (%) 0 0 0 0 0 Street tree presence Yes Yes No Yes Yes Buffer area coefficient, fB 5.37 5.37 1.00 5.37 5.37 Buffer width, WB (ft) 5 5 5 5 5 Sidewalk width (ft) 12 8 8 5 5 Width WS (ft) 10 8 8 5 5 Sidewalk width coefficient, fSW 3.0 3.6 3.6 4.5 4.5 Number of bus lanes, N 1 1 1 1 1 Bus running speed, SPD (mph) 30 30 30 30 30 PLOS score 0.87 0.88 1.22 0.97 0.97 LOS A A A A A

U. Case Study 2: Arterial BRT Analysis 245 section related inputs shown in Exhibit 101 are only required for a section-level analysis, and the segment length is only required for a facility-level analysis.) All other inputs can be defaulted, but will be substituted with actual values when known. Actual values are known or planned as part of the BRT project for the following inputs: • Bicycle lane width (5 feet), • Outside travel lane width (12 feet), • On-street parking (none), • Curb presence (yes), • Median type (undivided), and • Percent heavy vehicles (5%, from Example 2). Default values will be used for the following inputs: • Curb and gutter width (1.5 feet), • Pavement condition rating (3.5), and • Motorized vehicle running speed (the bus running speed of 30 mph from Example 4). Calculation The bicycle LOS score is calculated using Equation 149. It takes the following input values: • Traffic volumes are normally assumed to be evenly distributed between the available travel lanes and the calculation normally divides the directional traffic volume by the number of travel lanes. However, because the travel lane closest to the curb will be used as an exclusive bus lane, the total bus volume (9 buses per hour) will be used for V and the number of bus lanes (1) will be used for N. • The average motorized vehicle running speed S (30 mph for buses in the exclusive lane). • The effective speed factor fs equals (1.1199 × ln[S – 20]) + 0.8103, which results in a value of 3.39. Exhibit 178. Case study 2: pedestrian LOS calculations for supersection C (northbound “before”). Downstream Intersection 48th 49th 51st Claremont 55th Directional volume, V (veh/h) 653 706 809 1,044 978 Low-volume factor, fLV 1.00 1.00 1.00 1.00 1.00 Outside lane width (ft) 12 12 12 20 12 Bicycle lane width (ft) 0 0 0 0 0 Parking lane width (ft) 8 8 8 0 8 Width WT (ft) 20 20 20 20 20 Width W1 (ft) 8 8 8 0 8 Occupied on-street parking (%) 50 50 50 0 50 Street tree presence No No No No No Buffer area coefficient, fB 1.00 1.00 1.00 1.00 1.00 Buffer width, WB (ft) 0 0 0 0 0 Sidewalk width (ft) 14 10 10 7 7 Width WS (ft) 10 10 10 7 7 Sidewalk width coefficient, fSW 3.0 3.0 3.0 3.9 3.9 Number of lanes, N 2 2 2 2 2 Bus running speed, SPD (mph) 35 35 35 35 35 PLOS score 1.92 1.98 2.09 2.99 2.33 LOS B B B C B

246 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual • The percentage of heavy vehicles HV in the direction of travel is 5%. Because the bicycle LOS score is highly sensitive to HV, the percentage of heavy vehicles in the exclusive bus lane (100%, capped at 50% by the methodology) is not used, as it produces unrealistic results. • The pavement condition rating PC will use the default value of 3.5. • There is no on-street parking allowed. • The width of the outside lane, bicycle lane, and parking lane or shoulder Wt is illustrated in Exhibit 102. In this case, it is equal to the bus lane width (12 feet) plus the bicycle lane width (5 feet), for a total of 17 feet in all sections. • The width of the bicycle lane and parking lane or shoulder Wl is illustrated in Exhibit 102. In this case, it is equal to the bicycle lane width (5 feet) in all sections. • The effective width of the outside through lane as a function of traffic volume Wv equals Wt × (2 – 0.005V) = 33.2 feet as the street is undivided and the bus lane volume is under 160 vehicles per hour. • The average effective width of the outside through lane We is Wv + Wl – (0.2 × %OSP) = 38.2 feet as Wl ≥ 4. Bicycle LOS for the northbound section between 45th and 48th Streets is then calculated as follows: 0.507 ln 4 0.199 1 0.1038 7.066 1 0.005 0.760 0.507 ln 9 4 1 0.199 3.39 1 0.1038 5 7.066 1 3.5 0.005 38.2 0.760 4.01 2 2 2 2 2 2 BLOS V N f HV PC W BLOS BLOS s e ( ) ( ) [ ] [ ] ( ) ( ) = ×     + × × + + ×       − × + = × ×     + × × + × + ×       − × + = − From Exhibit 103, this BLOS score produces bicycle LOS A. Summary Bicycle LOS Results for Supersection C (Northbound) Exhibit 179 presents the summary results for each northbound section in supersection C, following implementation of the BRT project. As can be seen, all sections will operate at bicycle LOS A, a result of a combination of the provision of the bicycle lane and the exclusive bus lane acting as an additional buffer between bicyclists and the majority of the traffic. For comparison, Exhibit 180 presents the summary results for the “before” condition. The following changes are made to the inputs: • With both travel lanes available to general traffic, the normal process described in Section O4 of the Guide is used to determine the directional volume V and the number of lanes N. The average of the directional volume departing one intersection and the volume arriving at the next downstream intersection, from Exhibit 165, is used to determine V. • Lane widths take on their existing conditions values. • A default value of 50% is assumed for occupied on-street parking, except in the section between 51st and Claremont, where on-street parking is prohibited. • The vehicular free-flow speed of 35 mph used in previous examples is used for the vehicle running speed. It can be seen that in the “before” condition, bicycle LOS was mostly LOS E, with one section with LOS D. The poor LOS is a result of bicycles having to operate in mixed traffic adjacent

U. Case Study 2: Arterial BRT Analysis 247 to parked cars. Therefore, it is concluded that the proposed project will result in substantial improvement to bicycle LOS. Transit LOS Input Data Exhibit 104 lists the input data requirements for calculating transit LOS. The key required inputs are bus frequency, average bus speeds, and average passenger load factor; other inputs can be defaulted if not known. Bus frequency was identified in Example 4 (5 BRT buses per hour and Exhibit 179. Case study 2: bicycle LOS calculations for supersection C (northbound with BRT). Downstream Intersection 48th 49th 51st Claremont 55th Outside lane volume, V (veh/h) 9 9 9 9 9 Number of bus lanes, N 1 1 1 1 1 Bus running speed, S (mph) 30 30 30 30 30 Speed factor, fs 3.39 3.39 3.39 3.39 3.39 Percent heavy vehicles (%) 5 5 5 5 5 Pavement condition rating 3.5 3.5 3.5 3.5 3.5 Occupied on-street parking (%) 0 0 0 0 0 Outside lane width (ft) 12 12 12 12 12 Bicycle lane width (ft) 5 5 5 5 5 Parking lane width (ft) 0 0 0 0 0 Width Wt (ft) 17 17 17 17 17 Width Wl (ft) 5 5 5 5 5 Width Wv (ft) 33.2 33.2 33.2 33.2 33.2 Effective width We (ft) 38.2 38.2 38.2 38.2 38.2 BLOS score -4.01 -4.01 -4.01 -4.01 -4.01 LOS A A A A A Exhibit 180. Case study 2: bicycle LOS calculations for supersection C (northbound “before”). Downstream Intersection 48th 49th 51st Claremont 55th Directional volume, V (veh/h) 653 706 809 1,044 978 Number of lanes, N 2 2 2 2 2 Vehicle running speed, S (mph) 35 35 35 35 35 Speed factor, fs 3.84 3.84 3.84 3.84 3.84 Percent heavy vehicles (%) 5 5 5 5 5 Pavement condition rating 3.5 3.5 3.5 3.5 3.5 Occupied on-street parking (%) 50 50 50 0 50 Outside lane width (ft) 12 12 12 20 12 Bicycle lane width (ft) 0 0 0 0 0 Parking lane width (ft) 8 8 8 0 8 Width Wt (ft) 12 12 12 20 12 Width Wl (ft) 0 0 0 0 0 Width Wv (ft) 12.0 12.0 12.0 20.0 12.0 Effective width We (ft) 7.0 7.0 7.0 20.0 7.0 BLOS score 5.09 5.13 5.20 3.57 5.29 LOS E E E D E

248 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual 4 local buses per hour). Average BRT bus speed was determined in Example 4, but will need to be determined for local buses. Average passenger load factor will need to be estimated for future conditions. Ridership modeling conducted for the BRT project indicates that the average BRT load factor will be 110% in supersection C, while the average local bus load factor will be 70%. Other input data values are as follows: • Average excess wait time (default value of 3 minutes), • Stops with shelter and bench (only in the 48th to 51st section for both BRT and local), • Average passenger trip length (default value of 3.7 miles), and • Pedestrian LOS score by segment (calculated previously in Example 6). Local Bus Speed Local buses will use the exclusive bus lanes, but will stop more frequently than BRT buses (i.e., at every signalized intersection), with an average stop spacing of 535 feet (10 stops per mile). The local bus speed calculation then proceeds similarly to that demonstrated in Example 4 for BRT buses. Step 1: Unimpeded Bus Travel Time Rate. The time for a bus to decelerate from its running speed to a stop tdec (11.0 s) and accelerate to its running speed from a stop time tacc (13.0 s), and the distance traveled while accelerating and decelerating (529 ft) are the same as for BRT buses. At 10 stops per mile, the distance traveled at running speed per mile is: 5,280 5,280 10 529 0 ftL N Lrs s ad ( )= − = − × ≈ indicating that buses will rarely be able to accelerate to running speed before starting to decelerate to their next stop. The time spent at running speed per mile trs is also 0. Finally, the unimpeded running time rate tu is determined from Exhibit 152, applying the default average dwell time of 20 seconds for urban areas from Exhibit 104. (In contrast to the BRT route, which used the critical stop dwell time, the average dwell time is used here as the local route makes multiple stops within the supersection.) 60 0 10 20 13.0 11.0 60 7.3 min mit t N t t t u rs s dt acc dec( ) ( ) = + + + = + + + = Step 2: Additional Bus Travel Time Delays. Local buses using the exclusive bus lane will experience the same additional bus travel time delay tl as BRT buses. From Example 4, this value is 0.7 minutes per mile. Step 3: Base Bus Speed. The base bus running time rate tr is the sum of the unimpeded travel time rate and the additional travel time delay rate, or 8.0 minutes per mile. The corresponding speed is 7.5 mph. Step 4: Average Bus Speed. As discussed in Example 4, the scheduled number of buses is low enough that bus–bus interference will be rare and the average local bus speed is therefore the same as the base bus speed, or 7.5 mph. Transit LOS Calculation The transit LOS score for a section is a function of the section’s transit wait–ride score and its pedestrian LOS score. The transit wait–ride score, in turn, is a function of a headway factor and a perceived travel time factor.

U. Case Study 2: Arterial BRT Analysis 249 Headway Factor. Equation 169 is used to calculate the headway factor. For sections with only local bus service, buses arrive every 15 minutes (i.e., 60 minutes divided by 4 buses per hour), and the headway factor for these sections is: 4 exp 0.0239 15 2.79fh ( )= × − × = Similarly, for the sections with both local and BRT service, buses arrive every 6.7 minutes on average and the headway factor is 3.41. Perceived Travel Time Factor. The perceived travel time factor is calculated using Equa- tion 170. First, an in-vehicle travel time rate (IVTTR) is calculated. Equation 172 could be used with average bus speed as an input, but these speeds have already been calculated above (for local bus service, 8.0 minutes per mile) and in Example 4 (for BRT service, 4.0 minutes per mile). Next, a perceived travel time rate PTTR is calculated, which combines the perceived travel time rates of waiting for the bus to come (incorporating the effects of both stop amenities and late buses) and the perceived travel time rate while traveling on the bus. For local bus service, the following inputs are provided to Equation 171 in calculating the perceived travel time rate: • The travel time perception coefficient for passenger load a1 is 1.00, as the average local bus load factor was given as 70%. • The in-vehicle travel time rate IVTTR was determined previously to be 8.0 minutes per mile. • The travel time perception coefficient for excess wait time a2 uses the default value of 2.0. • The excess wait time rate EWTR is the excess wait time (3 minutes) divided by the average passenger trip length (3.7 miles), which equals 0.8 minutes per mile. • The amenity time rate ATR is 0.4 minutes per mile for stops with a shelter and bench and 0.0 otherwise. For local bus service in the section between 48th and 51st Street (e.g., the section providing a bus stop with a shelter and bench), PTTR is then: 1.00 8.0 2.0 0.8 0.4 9.2 min mi1 2PTTR a IVTTR a EWTR ATR( ) ( ) ( ) ( )= × + × − = × + × − = In the other sections without shelters or benches, local bus PTTR is 9.6 minutes per mile. BRT PTTR is calculated similarly, except that coefficient a1 = 1.28 by interpolation, using a 110% load factor and IVTTR = 4.0 minutes per mile. The resulting PTTR value is 6.3 minutes per mile. Finally, Equation 170 can now be used to calculate the perceived travel time factor fPTT. In addition to the PTTR value just calculated, this equation uses a default ridership elasticity e value of -0.4 and a baseline travel time rate of 4.0 minutes per mile (as supersection C is not located within the central business district of a metropolitan area of 5 million population or greater). For BRT service, fPTT is: 1 1 1 1 0.4 1 4 0.4 1 6.3 0.4 1 6.3 0.4 1 4 0.84f e BTTR e PTTR e PTTR e BTTR ptt [ ] [ ] [ ] [ ] ( ) ( ) ( ) ( ) ( )( ) ( )( ) ( )( ) ( )( )= − − + − − + = − − − − + − − − − + = Similarly, fPTT for local bus service is 0.73 for the section with a shelter and bench at the bus stop and 0.72 elsewhere. In the section between 48th and 51st Street, which is served by both local and BRT buses, an average fPTT needs to be calculated, weighted by the number of buses on each route: 0.84 5 buses 0.73 4 buses 9 buses 0.79f ptt ( )( ) ( )( ) = + =

250 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual In the other sections, served only by local buses at stops without shelters or benches, the fPTT is 0.72. Transit Wait-Ride Score. The transit wait–ride score sw-r is calculated by Equation 168. For the section between 48th and 51st Streets, it is: 3.41 0.79 2.69s f fw r h ptt= × = × =− For the other sections, it is: 2.79 0.72 2.01s f fw r h ptt= × = × =− Transit LOS Score. The transit LOS score TLOS is given by Equation 167, using a section’s wait–ride score and pedestrian LOS score as inputs. For the section between 48th and 51st Street, it is: 6.0 1.50 0.15 6.0 1.50 2.69 0.15 1.22 2.15TLOS s PLOSw r( ) ( ) ( ) ( )= − × + × = − × + × =− From Exhibit 111, this corresponds to transit LOS B. Exhibit 181 summarizes the results for all sections in the northbound direction. As can be seen, the section served by a BRT stop provides transit LOS B and is close to the LOS A threshold. The other sections served only by local buses provide transit LOS C. For comparison, Exhibit 182 presents the summary results for the “before” condition. The following changes are made to the inputs: Downstream Intersection 48th 49th 51st Claremont 55th Bus frequency (bus/h) 6 6 6 6 6 Shelter at stop? No No Yes No No Bench at stop? No No Yes No No Headway factor, fh 3.15 3.15 3.15 3.15 3.15 Perceived travel time factor, fptt 0.62 0.62 0.62 0.62 0.62 Transit wait–ride score, fs 1.95 1.95 1.95 1.95 1.95 Pedestrian LOS score, PLOS 1.92 1.98 2.09 2.99 2.33 TLOS score 3.36 3.37 3.39 3.52 3.42 LOS C C C D C Exhibit 182. Case study 2: transit LOS calculations for supersection C (northbound “before”). Downstream Intersection 48th 49th 51st Claremont 55th Bus frequency (bus/h) 4 4 9 4 4 Shelter at stop? No No Yes No No Bench at stop? No No Yes No No Headway factor, fh 2.79 2.79 3.41 2.79 2.79 Perceived travel time factor, fptt 0.72 0.72 0.79 0.72 0.72 Transit wait–ride score, fs 2.01 2.01 2.69 2.01 2.01 Pedestrian LOS score, PLOS 0.87 0.88 1.22 0.97 0.97 TLOS score 3.12 3.12 2.15 3.13 3.13 LOS C C B C C Exhibit 181. Case study 2: transit LOS calculations for supersection C (northbound with BRT).

U. Case Study 2: Arterial BRT Analysis 251 • The existing local bus service operates every 10 minutes, has an average peak hour load factor of 125%, and has a scheduled travel speed through supersection C of 6 mph. • The “before” PLOS values previously calculated are used. As can be seen, service in supersection C generally operates at LOS C, with one section operat- ing at LOS D due to a worse pedestrian environment. The section served by BRT will improve from a low LOS C to a high LOS B. The other sections will remain at LOS C, with the reduction in local bus headways compensated by reduced crowding, faster speeds, and an improved pedes- trian environment along the street. 8. References AC Transit. East Bay Rapid Transit Project in Alameda County: Traffic Analysis Report. Oakland, Calif., January 2012. Manual on Uniform Traffic Control Devices. Federal Highway Administration, Washington, D.C., 2009.

252 V. Case Study 3: Long-Range Transportation Plan Analysis 1. Overview This case study illustrates the use of highway capacity and operations analysis in support of the development and update of a long-range trans- portation plan (LRTP) for a large region. Planning Objective The objective is to perform the necessary transportation performance and investment alternatives analyses required to update the 2040 LRTP for the region. Auto, truck, bus, bicycle, and pedestrian analyses are to be performed. Background The hypothetical metropolitan planning organization (MPO) for this example consists of a dozen cities and the unincorporated areas within a county covering approximately 6,000 square miles with a combined population of slightly fewer than 1 million persons (see Exhibit 183). The largest city within the MPO accounts for about half of the total county population. The county’s population is forecasted to increase by 35% over the next 25 years. The highway network includes 2,100 directional miles of urban and rural arterial roads, 1,900 miles of collector roads, and 250 directional miles of urban and rural freeways. Example Problems Worked in this Case Study The planning problems that will be illustrated by worked examples within this case study are: • Example Problems that Develop Demand Model Inputs – Example 1: Estimating Free-Flow Speeds and Capacities for Model Input – Example 2: HCM-based Volume–Delay Functions for Model Input • Example Problems that Post-Process Demand Model Outputs – Example 3: Predicting Density, Queues, and Delay – Example 4: Predicting Reliability These planning problems illustrate the development, selection, and application of defaults for use in system-level planning analyses of large areas.

V. Case Study 3: Long-Range Transportation Plan Analysis 253 2. Example 1: Estimating Free-Flow Speeds and Capacities for Model Input Approach This example illustrates how the HCM can be used to develop a look-up table of free-flow speeds and capacities for use in quickly coding large highway networks for a regional travel demand model. It illustrates the application of the methods described in Section R of the Guide. The example will first identify an appropriate set of categories for representing the diversity of facility types, free-flow speeds, and per-lane capacities of the region’s roadway facilities. Values for free-flow speeds and capacities will then be selected from the appropriate tables in Section R. Procedure Step 1: Identify Facility Categorization Scheme for Look-up Table Based on local knowledge of the diversity of facility types, area types, terrains, capacities, and free-flow speeds in the region, the analyst tentatively identifies five facility types and four area types for stratifying the regional roadway network into 20 possible different categories for the speed and capacity look-up table. The HCM identifies four basic facility types: freeways, multilane highways, two-lane high- ways, and urban streets. The analyst, wishing to distinguish the differing capacity and operating characteristics between major and minor functional class facilities, has subdivided each of the HCM non-freeway facility types into major and minor facility subtypes. The first two columns of Exhibit 184 show the resulting categories. Additional categories and subdivisions are possible depending on the analyst’s needs and resources. The facilities may be further stratified by the general terrain in which they are located (level, rolling, and mountain- ous). Additional facility types may be created for on-ramps, off-ramps, collector-distributor roads within an interchange, and freeway-to-freeway ramps. A local road category may be added as well, if local roads will be included in the demand model. Exhibit 183. Case study 3: MPO planning area.

254 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual Step 2: Identify Free-Flow Speeds The analyst then identifies the appropriate default free-flow speed to be assumed for each facility and area type. These speeds are generally rounded to the nearest 5 miles per hour for the purposes of the look-up table and the initial coding of the highway network. Later, during model development and calibration, the analyst may fine tune the default free-flow speeds for specific links to obtain improved demand estimates from the model. The procedures described in Section R4 of the Guide are followed to identify free-flow speeds by facility type, area type, and subtype (major and minor). The values shown in the third column of Exhibit 184 are taken from the illustrative look-up table of free-flow speed defaults in Section R4 (Exhibit 124). The analyst may follow the procedures described in Section R4 to develop his or her own values based on local conditions. Step 3: Identify Capacities The analyst identifies the appropriate default per-lane capacities to be assumed for each facil- ity and area type. These capacities are generally rounded to the nearest 50 vehicles per hour per lane for the purposes of the look-up table and the initial coding of the highway network. Later, during model development and calibration, the analyst may fine tune the default capacities for specific links to obtain improved demand estimates from the model. The procedures described in Section R4 of the Guide are followed to identify capacities by facility type, area type, and subtype (major and minor). The values shown in Exhibit 184 are taken from the 80% HCM Capacity column of Exhibit 128, Illustrative Per-lane Capac- ity Look-up Table. The 80% HCM Capacity is selected (rather than the 90% HCM Capacity) because the facilities in this region tend to experience higher truck percentages and lower peak hour factors. The analyst might choose to use a mix of 80% and 90% capacities to reflect dif- ferent facility characteristics in the central business district of the region’s primary city relative to elsewhere in the region. The analyst may follow the procedures described in Section R4 to develop their own values based on local conditions and to develop capacities for additional facility subtypes. Facility Type Area Type Free-Flow Speed (mph) Capacity (veh/ln) Freeway Downtown 55 1,800 Urban 60 1,800 Suburban 65 1,900 Rural 70 1,900 Principal Highway Rural Multilane 55 1,700 Rural Two-lane 55 1,300 Minor Highway Rural Multilane 45 1,500 Rural Two-lane 45 1,300 Arterial Downtown 25 700 Urban 35 700 Suburban 45 600 Collector Downtown 25 600 Urban 30 600 Suburban 35 600 Exhibit 184. Case study 3: capacity and free-flow speed look-up table for highway system coding.

V. Case Study 3: Long-Range Transportation Plan Analysis 255 3. Example 2: HCM-Based Volume–Delay Functions for Model Input Approach This example illustrates how the HCM can be used to develop volume–delay functions for a regional travel demand model. This example can also be followed when post-processing a travel demand model’s forecasted demands for air quality analysis purposes to more-accurately estimate average vehicle speeds under congested conditions. The procedures described in Section R5 of this Guide will be demonstrated. The example will develop a speed–flow equation that accurately reflects the queuing delays for use within the travel demand model’s traffic assignment process. It is important to note there are many possible volume–delay functions that can be and are used in travel demand modeling. This example illustrates how the methods described in this Guide can be used with one of the more traditional volume–delay functions, the Bureau of Pub- lic Roads (BPR) function. This use is not meant to imply the superiority of the BPR function for use with the HCM or for any other purpose. In fact, other functions may be superior for demand modeling or predicting speeds, depending on the context of the specific application. Procedure Step 1: Identify Speed–Flow Curve Calibration Parameters While not required to estimate free-flow speeds and capacities, the speed–flow curve calibra- tion parameters are also identified in this step for each facility type, area type, and subtype. These parameters will be applied in Example 3 where link speeds are to be estimated, either during the traffic assignment stage of the demand model process, or as a post-processing step to be per- formed after the demand model run is complete. (The latter case might occur if more accurate congested speeds are needed for air quality analysis purposes.) The appropriate speed at capacity and the consequent BPR speed–flow curve calibration parameters are obtained from Exhibit 129, reproduced below as Exhibit 185). The analyst finds the nearest equivalent to the facility, area, and subtypes selected for Exhibit 184, and the HCM Facility Type Area Type Free-Flow Speed (mph) Capacity (veh/ln) HCM Base Speed at Capacity (mph) BPR A Parameter BPR B Parameter Freeway Downtown 55 1,800 50.0 0.10 7 Urban 60 1,800 51.1 0.17 7 Suburban 65 1,900 52.2 0.24 7 Rural 70 1,900 53.3 0.31 7 Principal Highway Rural Multilane 55 1,700 46.7 0.18 8 Rural Two-lane 55 1,300 42.5 0.29 8 Minor Highway Rural Multilane 45 1,500 42.2 0.07 9 Rural Two-lane 45 1,300 32.5 0.38 9 Arterial Downtown 25 700 6.7 2.71 3 Urban 35 700 11.0 2.19 2 Suburban 45 600 11.4 2.95 2 Collector Downtown 25 600 6.7 2.71 3 Urban 30 600 10.4 1.89 3 Suburban 35 600 11.0 2.19 3 Exhibit 185. Case study 3: speed–flow equation parameters.

256 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual facility types and subtypes listed in Exhibit 185. In this case, the calibration parameters for the freeway basic sections in Exhibit 185 are selected to represent all freeway links in Exhibit 184. The multilane and two-lane rural highway values in Exhibit 185 are selected to represent the Principal and Minor Highway types in Exhibit 184. The urban and suburban values for segments with signals are selected in Exhibit 185 to represent the arterial and collector subtypes in Exhibit 184. Finally, the speed at capacity for a 30-mph free-flow speed facility is interpolated between the values shown for 25 and 35 mph free-flow speed facilities. Step 2: Apply the Speed–Flow Curve Exhibit 186 shows the speed estimates for six example freeway and arterial links. The compu- tations proceed as follows: 1. The length, facility type, area type, and number of lanes are input by the modeler for each link. 2. The capacity and free-flow speed are obtained from the look-up table (Exhibit 185) according to the link’s facility type and area type. 3. The demand for the link is predicted by the demand model. 4. The speed at capacity is obtained from the look-up table (Exhibit 185) according to the link’s facility type and area type. 5. The ratio of the free-flow speed to the speed at capacity is used to estimate the BPR curve’s A parameter, and is given in the look-up table (Exhibit 185). 6. The B parameter of the BPR curve is obtained from the look-up table (Exhibit 185). 7. The demand-to-capacity ratio is computed for each link. 8. As free-flow speeds are available, the speed-based version of the BPR curve (Equation 203) is used to estimate travel speeds for each link. For example, Link A001 in the model is an eight-lane urban freeway (i.e., four lanes per direc- tion). From the look-up table (Exhibit 185), the per-lane capacity of an urban freeway is 1,800 veh/h; therefore, the capacity of a four-lane link would be 4 × 1,800 = 7,200 veh/h. The look-up table also provides values for free-flow speed (60 mph) and the BPR curve’s A and B parameters (0.17 and 7, respectively). Finally, the travel demand model estimates a demand of 8,220 veh/h for the link, which results in a demand-to-capacity ratio of 1.14. With this information in hand, the link’s speed can then be estimated using Equation 203: 1 60 1 0.17 8,220 7,200 42.0 mph0 7S S Ax B( ) ( ) ( )= + = + ×  = Exhibit 186 demonstrates the estimation of travel speeds for six example links in the model. Link ID Facility Type Lanes Demand (veh/h) Capacity (veh/h) Free- Flow Speed (mph) BPR A BPR B d/c Ratio Speed (mph) A001 Urban freeway 4 8,220 7,200 60 0.17 7 1.14 42.0 A002 Urban arterial 3 1,740 2,100 35 2.19 2 0.83 14.0 A003 Urban collector 2 1,170 1,200 30 1.89 3 0.98 10.9 A004 Rural freeway 2 2,790 3,800 70 0.31 7 0.73 67.6 A005 Rural principal highway 2 1,490 3,400 55 0.18 8 0.44 55.0 A006 Rural minor highway 1 250 1,300 45 0.38 9 0.19 45.0 Notes: d/c = demand-to-capacity ratio. For reasonably precise travel time estimates, it is necessary to carry more significant digits than shown here throughout the speed computations. Exhibit 186. Case study 3: example application of speed–flow curve.

V. Case Study 3: Long-Range Transportation Plan Analysis 257 4. Example 3: Predicting Density, Queues, and Delay Approach This example demonstrates post-processing travel demand model output to generate addi- tional useful performance measures. In this case, measures of density, queues (vehicle-hours in queue, VHQ), and delay for each model link will be identified, tabulated, and reported, fol- lowing the guidance in Section R5 of the Guide. The required model outputs are directional demand, capacity, lanes, and travel speed by link. Procedure Step 1: Compute Density The average peak hour density D of vehicle traffic (in vehicles per mile per lane) for each direc- tion of the link is computed by dividing the predicted demand rate V (in vehicles per hour) by the number of lanes N and the predicted average speed of traffic S (in miles per hour). If it is desired to convert the density into a LOS for use with freeways and multilane highways, the calculated density must be converted into units of passenger cars, as described in Sections H6 (freeways) and I6 (multilane highways) of the Guide. In lieu of making the facility-specific com- putations, a default passenger car equivalent PCE value of 1.2 may be used to account for most typical conditions. The density calculation then becomes: Equation 206D V PCE N S = × × where all variables are as described above. Exhibit 187 summarizes the HCM’s LOS criteria by facility type and (where appropriate) location. Step 2: Compute Vehicle-Hours In Queue The vehicle-hours traveled (VHT) for each direction of each link with a predicted d/c greater than 1.00 is accumulated to obtain total VHQ for the highway system. Person-hours in queue can be obtained by multiplying the VHQ by an assumed average vehicle occupancy rate. Step 3: Compute Vehicle-Hours Delay Vehicle-hours and person-hours of delay are typically output by the travel demand model using thresholds specified by the analyst. Vehicle-hours and person-hours of travel time are LOS Urban Freeway Facilities and Basic Segments Rural Freeway Facilities and Basic Segments Freeway Weaving Segments Freeway Merge and Diverge Segments Multilane Highways A ≤11 ≤6 ≤10 ≤10 ≤11 B >11–18 >6–14 >10–20 >10–20 >11–18 C >18–26 >14–22 >20–28 >20–28 >18–26 D >26–35 >22–29 >28–35 >28–35 >26–35 E >35–45 >29–39 >35–43 >35 >35–45 F >45 or any section has d/c>1.00 >39 or any section has d/c>1.00 >43 or d/c>1.00 d/c>1.00 >45 or d/c>1.00 Source: Adapted from HCM (2016), Exhibit 10-6, Exhibit 12-15, Exhibit 13-6, and Exhibit 14-3. Exhibit 187. Case study 3: LOS criteria for freeway and multilane highway facilities.

258 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual compared to an agency specified minimum speed goal for each link. The speed goal may be the link free-flow speed, or some other value reflecting agency policy. The vehicle-hours of delay VHD for a link is computed by taking the difference between the actual VHT and the estimated VHT if all vehicles could have traversed the link at the agency’s policy minimum acceptable speed Sp (in miles per hour) for the facility represented by the link: Equation 207VHD V L S V L Sp = × − × where all variables are as defined previously. Step 4: Interpreting Results Density can be compared to HCM thresholds to obtain a sense of the degree of congestion represented by the density. System VHD and VHQ can be compared across alternatives to obtain a sense of the degree to which one alternative is better than the other. VHD can also be used in economic cost–benefit analyses. 5. Example 4: Predicting Reliability Approach This example demonstrates the prediction of automobile travel time reliability for a freeway link, following the simplified HCM method described in Section H7 of the Guide. The travel demand model outputs required are the link’s peak hour speed, free-flow speed, number of lanes, and demand-to-capacity ratio. Procedure Step 1: Compute Average Annual Travel Time Index for Each Link The average annual travel time index for a freeway link is computed from its free-flow speed, peak hour speed, and its peak hour volume/capacity ratio. First, the recurring delay rate (RDR) for the link (in hours per mile) is calculated from the peak hour and free-flow speeds using Equa- tion 33. Using the data for Link A001 in Example 2 (Exhibit 186), the free-flow speed is 60 mph, the peak hour speed is 42.0 mph, and the recurring delay rate is then: 1 1 1 42.0 1 60 0.0071 h miRDR S FFS = − = − = Next, the incident delay rate (IDR) (Equation 34) for the link (in hours per mile) is calcu- lated from the number of directional lanes on the link and the link’s demand-to-capacity ratio (capped at 1.00). Again using the data for Link A001, the number of directional lanes is 4 and the d/c ratio is 1.14, which is reduced to 1.00 for use in calculating IDR. Then: 0.020 2 0.003 0.020 4 2 0.003 1.00 0.014 h mi12 12IDR N X[ ] [ ]( ) ( )= − − × × = − − × × = Finally, the average annual travel time index TTIm is calculated using Equation 32. The link free-flow speed is an input to the equation. For Link A001, the free-flow speed is 60 mph and TTIm is then: 1 1 60 0.0071 0.014 2.27TTI FFS RDR IDRm ( ) ( )= + × + = + × + =

V. Case Study 3: Long-Range Transportation Plan Analysis 259 Similarly, applying the data for Link A004 from Example 2, RDR is 0.0005, IDR is 0.0005, and TTIm is 1.07. Step 2: Compute Average Annual Travel Time Index for the System The average annual travel time index for the system is the system VHT (sum of link VHTs) divided by the theoretical VHT if all VMT were completed at free-flow speeds. If, for the sake of example, one defines the system to consist solely of the two freeway links from Example 2, A001 and A004, the calculations proceed as follows. For Link A001, the link length is 0.85 miles, the free-flow speed is 60 mph, the peak hour speed is 42.0 mph, and the link demand is 8,220 veh/h. VHT at free-flow speed would be (0.85/60) × 8,220 = 116, while peak hour VHT is (0.85/42.0) × 8,220 = 166. Similarly, for Link A004, the link length is 2.50 miles, the free-flow speed is 70 mph, the peak hour speed is 67.6 mph, and the link demand is 2,790 veh/h. VHT at free-flow speed would be (2.50/70) × 2,790 = 100, while peak hour VHT is (2.50/67.6) × 2,790 = 103. The system average annual travel time index TTIm,sys is then: 166 103 116 100 1.25,TTIm sys = + + = Step 3: Compute Reliability Statistics for System The 95th percentile travel time index TTI95 and percent of trips traveling under 45 mph PT45 for the system can be estimated from the system average annual travel time index TTIm,sys using Equation 35 and Equation 36, respectively. 1 3.67 ln 1 3.67 ln 1.25 1.8295, ,TTI TTIsys m sys( ) ( )= + × = + × = 1 exp 1.5115 1 1 exp 1.5115 1.25 1 0.3145, ,PT TTIsys m sys[ ] [ ]( ) ( )= − − × − = − − × − = 6. Reference Highway Capacity Manual: A Guide to Multimodal Mobility Analysis. 6th ed. Transportation Research Board, Washington, D.C., 2016.

Abbreviations and acronyms used without definitions in TRB publications: A4A Airlines for America AAAE American Association of Airport Executives AASHO American Association of State Highway Officials AASHTO American Association of State Highway and Transportation Officials ACI–NA Airports Council International–North America ACRP Airport Cooperative Research Program ADA Americans with Disabilities Act APTA American Public Transportation Association ASCE American Society of Civil Engineers ASME American Society of Mechanical Engineers ASTM American Society for Testing and Materials ATA American Trucking Associations CTAA Community Transportation Association of America CTBSSP Commercial Truck and Bus Safety Synthesis Program DHS Department of Homeland Security DOE Department of Energy EPA Environmental Protection Agency FAA Federal Aviation Administration FAST Fixing America’s Surface Transportation Act (2015) FHWA Federal Highway Administration FMCSA Federal Motor Carrier Safety Administration FRA Federal Railroad Administration FTA Federal Transit Administration HMCRP Hazardous Materials Cooperative Research Program IEEE Institute of Electrical and Electronics Engineers ISTEA Intermodal Surface Transportation Efficiency Act of 1991 ITE Institute of Transportation Engineers MAP-21 Moving Ahead for Progress in the 21st Century Act (2012) NASA National Aeronautics and Space Administration NASAO National Association of State Aviation Officials NCFRP National Cooperative Freight Research Program NCHRP National Cooperative Highway Research Program NHTSA National Highway Traffic Safety Administration NTSB National Transportation Safety Board PHMSA Pipeline and Hazardous Materials Safety Administration RITA Research and Innovative Technology Administration SAE Society of Automotive Engineers SAFETEA-LU Safe, Accountable, Flexible, Efficient Transportation Equity Act: A Legacy for Users (2005) TCRP Transit Cooperative Research Program TDC Transit Development Corporation TEA-21 Transportation Equity Act for the 21st Century (1998) TRB Transportation Research Board TSA Transportation Security Administration U.S.DOT United States Department of Transportation

TRA N SPO RTATIO N RESEA RCH BO A RD 500 Fifth Street, N W W ashington, D C 20001 A D D RESS SERV ICE REQ U ESTED ISBN 978-0-309-37565-8 9 780309 375658 9 0 0 0 0 N O N -PR O FIT O R G . U .S. PO STA G E PA ID C O LU M B IA , M D PER M IT N O . 88 Planning and Prelim inary Engineering A pplications G uide to the H ighw ay Capacity M anual N CH RP Report 825 TRB