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 2 Medium-Level Analysis Methods The sections in Part 2 of the Guide describe medium-level analysis methods that work best when evaluating a single freeway, highway, or urban street facility and its component interchanges, segments, and intersections. The sections are organized according to the system elements (e.g., freeways, signalized intersections) used by the HCM, and include sections focusing on the analysis of non-automobile modes: H. Freeway analyses I. Multilane highways J. Two-lane highways K. Urban streets L. Signalized intersections M. Stop-controlled intersections N. Roundabouts O. Pedestrians, bicyclists, and public transit P. Truck level of service Sections H–N have similar structures, to aid the reader in quickly finding information relevant to a particular analysis need. The typical contents of these sections include: • System element definition and overview • Potential applications for the methods presented in the section • Summary of the types of methods presented in the section • Scoping and screening method (e.g., using generalized service volume tables) • Full HCM method with defaults • Simplified version or versions of the full HCM method • Travel time reliability estimation method (if available) • Cross-references to multimodal performance measures provided in Sections O and P • Cross-references to worked examples in the case studies in Part 4 Section O provides medium-level methods for estimating pedestrian, bicycle, and public transit performance measures and is organized by the system elements used in sections H–N. Section O also provides planning methods for off-street pedestrian and bicycle facilities. Section P describes a planning method for estimating truck LOS.

43 H. Freeway Analyses 1. Overview A freeway is a grade-separated highway with full control of access and two or more lanes in each direction dedicated to the exclusive use of motorized vehicles. This section presents medium-level methods suitable for evaluating single freeway facilities or segments. 2. Applications The methodologies presented in this section support the following planning and preliminary engineering applications: • Development of a freeway corridor system management and improvement plan; • Feasibility studies of: – Adding a high-occupancy vehicle (HOV), high-occupancy toll (HOT), or express lane (or converting an existing lane or shoulder lane to HOV, HOT, or toll operation); – Ramp metering; or – Managed lanes, including speed harmonization, temporary shoulder use, and other active transportation and demand management (ATDM) strategies; • Interchange justification or modification studies (the freeway mainline portions of these studies); and • Land development traffic impact studies. The facility-specific procedures described here produce facility-specific performance results that can be aggregated into system performance measures for transportation systems plans. Section R, Areas and Systems, provides more cost-efficient methods for computing system per- formance measures. HCM Chapter 25, Freeway Facilities: Supplemental, presents a model for predicting the per- formance of freeways with extended upgrades, significant volumes of trucks, or both (HCM 2016). This method is not addressed in this Guide and users should be cautious about using this section’s planning methods to predict freeway performance for extended upgrades (i.e., upgrades of greater than 2% persisting for one mile or more). The planning methods described in this section do not explicitly address freeway work zones, ATDM measures, and managed (e.g., toll) lanes. Further information on these topics can be found in HCM Chapter 10, Freeway Facilities Core Methodology.

44 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual 3. Analysis Methods Overview Freeway performance can be directly measured in the field or estimated in great detail using microsimulation. However, the resource requirements of these methods render them generally impractical for most planning and preliminary engineering applications. The HCM provides a less resource-intensive approach to estimating freeway performance; however, it too is generally impractical to use for many planning and preliminary engineering analyses if 100 percent field-measured inputs are to be used. This section presents two medium-level methods for evaluating freeway performance, plus a high-level screening and scoping method that can be used to focus the analysis on only those locations and time periods requiring investigation, as shown by the unshaded boxes in Exhibit 18. The HCM’s segment and facility analysis methods, covered in HCM Chapters 10 to 14, pro- vide a good basis for estimating freeway performance under many conditions. The basic segment analysis method is relatively simple to apply when defaults are used for some difficult-to-obtain inputs. Analysis of on-ramps and off-ramps (merge and diverge segments in HCM parlance) and weaving segments is a bit more challenging with a more complex set of equations, but the computational effort is simplified with software. The freeway facility method is the most chal- lenging, requiring a great deal more data to cover the larger geographic area involved in a full facility analysis. In addition, several computations are iterative. Generally, specialized software is required to implement the HCM facility method. Consequently, this section of the Guide presents a simplified HCM facility analysis method that reduces the overall number of computations and eliminates the dynamic segmentation and iterative computations. The simplified analysis method is designed to be easily programmable in a static spreadsheet without need for macros. Because both the HCM method and the simplified method require a fair amount of data, this section also provides a high-level service volume and volume-to-capacity (v/c) ratio screening method for quickly identifying which portions of the freeway can be evaluated solely using the segment analysis methods and which portions will require a facility-level analysis to properly account for the spillover effects of congestion. The high-level method Exhibit 18. Analysis options for freeways. High Level Medium Level Low Level

H. Freeway Analyses 45 can also be used to quickly compare improvement alternatives according to the capacity they provide. 4. Scoping and Screening Method Generalized Service Volume Table Whether or not a more detailed freeway facility analysis is needed can be determined by com- paring the counted or forecasted peak hour or daily traffic volumes for the sections of the free- way between each on- and off-ramp to the values given in Exhibit 19. If all of the section volumes fall in the LOS E range or better, there will be no congestion spillover requiring a full facility analysis to better quantify the facility’s performance. One can then use the HCM segment analy- sis procedures with defaults for some of the inputs to evaluate the performance of each segment. (Note that “segments” have a special definition in the HCM, while “sections” are defined in this Guide by the freeway on- and off-ramps.) The service volumes in Exhibit 19 can also be used to quickly determine the geographic and temporal extent of the freeway facility that will require analysis. If the counted or forecasted volumes for a section fall below the agency’s target LOS standard, then the section can be excluded from a more detailed analysis. If the volumes fall near or above the vol- ume threshold for the agency’s target LOS, then the section may require more detailed analysis. Any section that exceeds the capacity values in Exhibit 19 will have queuing that may impact upstream sections and reduce downstream demands. In such a situation, a full freeway facility analysis is required to ascertain the freeway’s performance. The facility analysis can be performed either using the HCM method with defaults, or the simplified HCM method, both of which are described later in this section. The analyst may also use the capacities shown in Exhibit 19 to compute the peak hour, peak direction demand-to-capacity ratio for each segment under various improvement options. These options can then be quickly ranked according to their forecasted demand-to-capacity ratios for the critical sections of the freeway. Area Type Terrain Peak Hour Peak Direction (veh/h/ln) AADT (2-way veh/day/ln) LOS A-C LOS D LOS E (capacity) LOS A-C LOS D LOS E (capacity) Urban Level 1,550 1,890 2,150 14,400 17,500 19,900 Urban Rolling 1,480 1,810 2,050 13,700 16,700 19,000 Rural Level 1,460 1,770 2,010 12,100 14,800 16,800 Rural Rolling 1,310 1,600 1,820 11,000 13,400 15,200 Source: Adapted from HCM (2016), Exhibit 12-39 and 12-40. Notes: Entries are maximum vehicle volumes per lane that can be accommodated at stated LOS. AADT = annual average daily traffic. AADT per lane is two-way AADT divided by the sum of lanes in both directions. Urban area assumptions: Free-flow speed = 70 mph, 5% trucks, 0% buses, 0% RVs, peak hour factor = 0.94, 3 ramps/mi, 12-ft lanes, K-factor = 0.09, and D-factor = 0.60. Rural area assumptions: Free-flow speed = 70 mph, 12% trucks, 0% buses, 0% RVs, peak hour factor = 0.94, 0.2 ramps/mi, capacity adjustment factor for driver population = 1.00, 12-ft lanes, 6-ft lateral clearance, K-factor = 0.10, and D-factor = 0.60. Similar tables can be developed by adjusting input values to reflect other assumptions. The K-factor is the ratio of weekday peak hour two-way traffic to AADT. The D-factor is the proportion of peak hour traffic in the peak direction. Exhibit 19. Daily and peak hour service volume and capacity table for freeways.

46 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual Estimating Freeway Service Volumes The approximate maximum AADT (two-way) that can be accommodated by a freeway at a given LOS can be estimated from Exhibit 19. For example, an eight-lane freeway can carry between 120,000 (15,200 × 8 lanes) and 160,000 (19,900 × 8 lanes) AADT at LOS E, depending on its location (urban or rural) and the terrain type. Higher AADTs can be accommodated when the proportion of AADT occurring during the peak hour (i.e., K-factor) is lower, the proportion of traffic in the peak direction during the peak hour (i.e., D-factor) is lower, or both. Single-lane managed lanes (e.g., HOV lanes, HOT lanes) have capacities between 1,500 and 1,800 vehicles per hour per lane (veh/h/ln) depending on the free-flow speed and the type of bar- rier or buffer (if any) separating the single managed lane from the other general purpose lanes. Dual managed lanes have capacities between 1,650 and 2,100 veh/h/ln (HCM 2016). When local traffic data suggest that values different from the assumptions used in Exhibit 19 would be more appropriate, the analyst should modify the daily and hourly service volumes as follows: Equation 130 0 0 ,0 ,0 0 DSV DSV f CAF PHF K D K D f CAF PHF HV p HV p = × × × × × × × × where DSV = daily service volume (veh/day/ln), DSV0 = initial daily service volume in Exhibit 19 (veh/day/ln), fHV, fHV,0 = desired and initial adjustment factors, respectively, for presence of heavy vehicles in the traffic stream, CAFp, CAFp,0 = desired and initial capacity adjustment factors, respectively, for unfamiliar driver populations, PHF, PHF0 = desired and initial peak hour factors, respectively, K, K0 = desired and initial proportions, respectively, of daily traffic occurring during the peak hour, and D, D0 = desired and initial proportions, respectively, of traffic in the peak direction during the peak hour. Equation 13 can also be used to modify the peak hour, peak direction service volumes if the initial peak hour service volumes from Exhibit 19 are used instead of the daily values. The heavy vehicle adjustment factor fHV used in the service volume table is computed using the following adaptation of HCM Equation 12-10 (HCM 2016): ( )= + × − 1 1 1 Equation 14f P E HV HV HV where fHV = heavy vehicle adjustment factor (decimal), PHV = percentage heavy vehicles (decimal), and EHV = heavy vehicle equivalence from Exhibit 20. For convenience, all heavy vehicles are assigned a single passenger car equivalent (PCE) value from Exhibit 20 below. Daily service volumes should be rounded down to the nearest hundred vehicles, given the many default values used in their computation. Peak hour, peak direction service volumes should be rounded down to the nearest ten vehicles. Terrain Type EHV Level 2.0 Rolling 3.0 Mountainous 5.0 Source: Adapted and extrapolated from HCM (2016), Exhibit 12-25. Exhibit 20. Heavy vehicle equivalence values for freeways.

H. Freeway Analyses 47 5. Employing the HCM with Defaults The HCM divides the freeway facility into various uniform segments that may be analyzed to determine capacity and LOS. HCM Chapter 10, Freeway Facilities Core Methodology, provides more details on how each segment type is defined. Exhibit 21 lists the data needed to evaluate the full range of performance measures for freeway facility and segment analysis. Individual performance measures may require only a subset of these inputs. Free-flow speed estimation using the HCM requires the following information about the facility’s geometry: lane widths, right side lateral clearance, and the number of ramps per mile. Capacity (in terms of vehicles per hour) requires the free-flow speed plus additional data on heavy vehicles, terrain type, number of lanes, peak hour factor (the ratio of the average hourly flow to the peak 15-minute flow rate), and the driver population (i.e., familiar or unfamiliar drivers). Once free-flow speed and capacity have been calculated, then speed, LOS, and queue lengths can be estimated if additional information about segment lengths and the directional demand (vehicles per hour) is available. Travel time reliability analysis requires the same data required to estimate speeds plus infor- mation on the variability of demand; the severity, frequencies and durations of incidents; the frequency of severe weather conditions; and the frequencies of work zones by number of lanes closed by duration. 6. Simplified HCM Facility Method The simplified HCM facility method for freeways focuses on facility-level analysis and section-level analysis. A section is defined as extending from ramp gore point to ramp gore point, avoiding the need to subdivide the section into 1,500-foot-long HCM merge and diverge areas. A section may combine several HCM segments. For example, a section extending between an Performance Measure Input Data (units) FFS Cap Spd LOS Que Rel Default Value Lane widths and right side lateral clearance (ft) • • • • • • 12-ft lanes 10-ft lateral clearance Ramp density (per mile) • • • • • • Must be provided Percentage heavy vehicles (%) • • • • • 12% (rural), 5% (urban) Terrain type/specific grade • • • • • Must be provided Number of directional lanes • • • • • Must be provided Peak hour factor (decimal) • • • • • 0.94 Driver population factor (decimal) • • • • • 1.00 (i.e., familiar drivers) Segment length (mi) • • • • Must be provided Directional demand (veh/h) • • • • Must be provided Variability of demand • Must be provided Incident and crash frequencies • Must be provided Severe weather frequencies • Must be provided Work zone frequencies • Must be provided Notes: FFS = free-flow speed (mph), Cap = capacity (veh/h/ln), Spd = speed (mph), LOS = level of service (A–F), Que = queue (veh), and Rel = travel time reliability (several measures). If a service volume table is used to determine LOS, the data requirements consist of AADT; K-factor (proportion of daily traffic occurring in the peak hour); D-factor (proportion of peak hour traffic in the peak direction); and number of lanes. Exhibit 21. Required data for HCM freeway analysis.

48 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual on-ramp and an off-ramp may be composed of three HCM segments: a merge segment, a basic or weave segment, and a diverge segment. Defining Sections for the Simplified Method Input variables are characterized as global or section inputs. Planning analysis sections are defined to occur between points where either demand or capacity changes. For example, if a lane drop exists between an on- and off-ramp, that length will involve two sections (because the reduced number of lanes reduces the capacity of the section). But for a three-segment sequence of merge area, basic segment, and diverge area, the simplified method defines a single section. Significant grade changes (i.e., involving grades steeper than 2%) also should be considered for separate sections. For example, the facility shown in Exhibit 22 with eleven HCM segments would be trans- formed into seven planning sections for use with the simplified method. Data Requirements The data needs for the simplified freeway facility analysis method, shown Exhibit 23, are simi- lar to those of the HCM method listed in Exhibit 21. The differences are that the simplified method uses posted speed limits to estimate the free-flow speed, and the travel time reliability analysis requires only the crash rate for the facility. Global inputs include information about the facility of interest. Those are applied to all sec- tions across all analysis periods. They include free-flow speed, peak hour factor, percentage heavy vehicles (%HV), K-factor, and a traffic growth factor (if used to obtain forecasts). Exhibit 22. Relationship of HCM segments to simplified method sections. Performance Measure Input Data (units) FFS Cap Spd LOS Que Rel Default Value Posted speed limit (mph) • • • • • • Must be provided Percentage heavy vehicles (%) • • • • • 12% (rural), 5% (urban) Terrain type/specific grade • • • • • Must be provided Number of directional lanes • • • • • Must be provided Peak hour factor (decimal) • • • • • 0.94 Driver population factor (decimal) • • • • • 1.00 Segment length (mi) • • • • Must be provided Directional demand (veh/h) • • • • Must be provided Average crash rate • Must be provided Notes: FFS = free-flow speed (mph), Cap = capacity (veh/h/ln), Spd = speed (mph), LOS = level of service (A–F), Que = queue (veh), and Rel = travel time reliability (several measures). If a service volume table is used to determine LOS, the data requirements consist of AADT, K-factor (proportion of daily traffic occurring in the peak hour), D-factor (proportion of peak hour traffic in the peak direction), and number of lanes. Exhibit 23. Required data for simplified freeway facility analysis.

H. Freeway Analyses 49 Estimating Inputs This subsection describes procedures for estimating the free-flow speed, the section type, and the section capacities. Identifying Freeway Section Types The following definitions are used to split the freeway mainline into its component sections: • A basic freeway section is a section of freeway with a constant demand and capacity, without the presence of on-ramps or off-ramps. • A freeway ramp section is a section of freeway with an on-ramp, off-ramp, or both, but without the presence of an auxiliary lane connecting two ramps. • A freeway weaving section occurs wherever an on-ramp is followed by an off-ramp, and the two are connected by an auxiliary lane. Estimating Free-Flow Speed Free-flow speed is the average traffic speed under low-flow conditions. The most-accurate method for estimating segment free-flow speeds is to measure it in the field during low-flow con- ditions (under 800 veh/h/ln, after considering the effects of heavy vehicles and peaking within the peak hour). In urban environments, traffic sensors may be available to allow the estimation of free-flow speeds; however, this is not usually practical for planning applications. HCM Equa- tion 12-2 (HCM 2016) can be used to estimate free-flow speeds based on the facility’s geometric characteristics: = − − − ×75.4 3.22 Equation 150.84FFS f f TRDLW RLC where FFS = free-flow speed (mph), fLW = adjustment for lane width (mph) = 0.0 for 12-ft or wider lanes, 1.9 for 11-ft lanes, or 6.6 for 10-ft lanes (see HCM Exhibit 12-20), fRLC = adjustment for right side lateral clearance (mph), ranges from 0.0 for 6-ft lateral clear- ance to 3.0 for 1-ft clearance with 2 directional lanes (see HCM Exhibit 12-21), and TRD = total ramp density (ramps/mi) = number of on- and off-ramps in one direction for 3 miles upstream and 3 miles downstream, divided by 6 miles. An alternative approach is to assume the free-flow speed is equal to the posted speed limit plus an adjustment reflecting local driving behavior. HCM Exhibit 10-7 (HCM 2016) suggests adding 5 mph to the posted speed limit. All of these approaches for estimating free-flow speed assume all vehicles have the same posted speed limit. Should the posted speed limit for trucks or other vehicle classes be lower than that for other vehicle types, then the analyst will have to apply some judgment based on local experi- ence when employing the above methods to estimate free-flow speed. Estimating Section Capacities Free-flow speed and percent heavy vehicles are used to calculate section capacity using the following equation: ( )( )( ) = + × − + × 2,200 10 min 70, 50 1 % 100 Equation 16c S HV CAFi FFS

50 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual where Ci = capacity of section i (veh/h/ln), SFFS = free-flow speed (mph), %HV = percent of heavy vehicles (decimal), with heavy vehicles consisting of trucks with more than four tires, buses, and recreational vehicles (see Exhibit 23 for suggested default values), and CAF = capacity adjustment factor, described below, that calibrates the basic section capacity to account for the influences of ramps, weaves, unfamiliar driver populations, and other factors. Equation 16 is fully consistent with the HCM speed–flow models. Section inputs include section type (basic, weave, or ramp); section length in miles; number of lanes; and directional AADT. This information, together with the global inputs, is used to calculate free-flow travel rate (the inverse of free-flow speed); capacity adjustment factors (CAFs) for weave and ramp sections; adjusted lane capacity (the product of base capacity and CAF); and section capacity (the product of adjusted lane capacity and number of lanes). Mainline Entry and On-Ramp Capacity Constraint The estimated hourly mainline entry demands should be compared to the estimated capac- ity for the mainline entry. If the mainline entry hourly demands exceed the estimated mainline entry capacity, the hourly demands should be set equal to the mainline entry capacity for the purposes of the freeway analysis. The mainline entry capacity is computed using Equation 16. Similarly, the hourly on-ramp demands should be compared to the estimated on-ramp capac- ities and any demand in excess of the hourly capacity should be reduced to the hourly capacity. HCM Exhibit 14-12 (HCM 2016) provides nominal on-ramp capacities that can be used in determining the capacity constraint. These capacities are in terms of passenger car equivalents and vary by the ramp free-flow speed. For planning purposes, a nominal value of 2,000 vehicles per hour per lane can be used as an on-ramp capacity. Note that this capacity may exceed that of the ramp merge point with the freeway mainline. Off-ramp demands may exceed an off-ramp’s capacity, in which case excess demand would be queued on the off-ramp and potentially the freeway. This effect is not accounted for in the simplified method. The existence of this condition indicates the need for a more detailed analysis. Assigning Section Demands Daily or peak hour demands are required for each freeway segment. These demands are then converted to 15-minute demands for each section, with unserved demand from a prior 15-minute period being carried over to the following 15-minute period. The demand level for each section is determined from entering demand, exiting demand, and carry-over demand from a previous analysis period (in the case of over capacity operations). The demand-to-capacity ratio is then calculated, along with the delay rate, as discussed later. For each section and time period, the method further estimates travel rate, travel time, density per lane, and segment queue length. Section inflow and outflow during each of the four 15-minute time periods during the peak hour (i.e., t = 1 to 4) is computed as follows: q AADT k f t AADT k PHF f i t i gf i g f , , = × × = × ×     × 1 3 1 t AADT k PHF f ti g f = × × −     × =      2 2 1 4 Equation 17

H. Freeway Analyses 51 where qi,t = in- or outflow for section i during analysis period t (veh/h), AADTi = average annual daily traffic for section i (veh/day), k = K-factor (decimal) = proportion of daily traffic during the peak hour, PHF = peak hour factor (decimal), and fgf = growth factor to forecast future demands. Equation 17 assumes (1) the peak 15-minute flow rate will occur during the second 15-minute period within the peak hour, (2) the first and third 15-minute periods will have average flow rates for the peak hour, and (3) the final 15-minute period within the peak hour will have a lower flow rate to ensure that all four 15-minute periods add up to the total peak hour flow, as shown in Exhibit 24. The demand level di,t in section i at time t is computed as the demand level in section i – 1 plus the inflow at section i at time t minus the outflow at the same section at time t, plus any carry-over demand d′i,t–1 in section i during the previous time interval t – 1. The relationship is as follows: ( ) ( )= + − + ′ − − Equation 18, 1, , in , out , 1d d q q di t i t i t i t i t The carry-over demand d′i,t–1 at section i at time t is the difference between the section demand and capacity as follows, where all variables are as defined previously: max , 0 Equation 19, ,d d ci t i t i( )′ = − The carry-over demand is also used as an indication of the presence of a queue on the section. Note that queues are considered to be vertical, and are not carried to an upstream link. Section queue length is estimated by dividing the difference in lane demand and capacity by its density. It essentially provides an estimate for how long the queue would spillback at the given density, assuming a fixed number of lanes upstream of the bottleneck. Estimating Section Volume-to-Capacity Ratios The section volume-to-capacity ratios are computed using the section demands and capacities previously computed. Off-Ramp Volume-to-Capacity Ratio Check There may be cases where capacity constraints on the off-ramp, such as the capacity of the intersection approach at the foot of the off-ramp, may result in lower throughput than predicted. In such a situation, the excess demand may queue up on the off-ramp and eventually back up Exhibit 24. Allocation of peak hour demand to 15-minute periods.

52 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual onto the freeway, affecting mainline operations. The complexity of such a situation goes beyond typical planning analysis and may require microsimulation to adequately assess the severity of the problem and its impacts on freeway mainline operations. Speed Section speeds are estimated based on delay rate curves. The estimated delay is added to the estimated travel time at free-flow speed to obtain the travel time with congestion effects. The congested travel time is divided into the section length to obtain the average speed. Estimating Section Delay Rates for Basic Sections In the following, details for the delay rate estimation are presented for basic sections without the influence of on-ramps or off-ramps. That discussion is followed by recommended adjustments for merge, diverge, and weaving sections. The procedure estimates delay rate per unit distance as a function of the section’s demand-to- capacity ratio. The delay rate is calculated as the difference between actual and free-flow travel time per unit distance. The calculation of the delay rate needs to be performed separately for undersaturated and oversaturated flow conditions. Undersaturated Flow Conditions. For undersaturated flow conditions, the HCM’s speed– flow model for basic freeway segments is used to estimate delay rates. This model, shown in Equation 20 and Exhibit 25, is a polynomial function fitted to the HCM speed–flow curves. The parameter E is related to the breakpoint in the HCM speed–flow curves, that is, the demand at which travel speeds begin to decline from the free-flow speed. ∆ = <   +   +   + ≤ ≤      0 1.00 Equation 20 , , 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 where DRUi,t = delay rate for undersaturated section i at time t (s/mi), di,t = demand for section i at time t (veh/h), ci = section capacity (veh/h), and A, B, C, D, E = equation parameters from Exhibit 25. For each FFS, the sole input to the regression model is the demand-to-capacity ratio. Oversaturated Flow Conditions. For oversaturated flow conditions, the undersaturated model is first applied with a demand-to-capacity ratio of 1.00. An additional oversaturated delay FFS (mph) A B C D E 75 68.99 -77.97 34.04 -5.82 0.44 70 71.24 -85.48 35.58 -5.44 0.52 65 92.45 -127.33 56.34 -8.00 0.62 60 121.35 -184.84 83.21 -9.33 0.72 55 156.43 -248.99 99.20 -0.12 0.82 Exhibit 25. Values for the parameters of Equation 20.

H. Freeway Analyses 53 rate is approximated assuming uniform arrivals and departures at a freeway bottleneck. This oversaturation delay rate is calculated using the following equation: ∆ = − 2 1 Equation 21 , , T L d c RO i i t i i t where DROi,t ____ = additional average delay rate due to oversaturation for section i at time t (s/mi), T = analysis period duration (s), typically 900 s, Li = length of section i (mi), di,t = demand for section i at time t (veh/h), and ci = section capacity (veh/h). Section Travel Time. After determining the delay rate, the section travel rate is determined by adding the delay rate(s) to the travel rate under free-flow conditions. The section travel time is then computed by multiplying the travel rate and the section length, as shown in Equation 22. ( )= + ∆ + ∆3,600 Equation 22, , ,T L FFS Li t i i i RU ROi t i t where Ti,t = travel time for section i at time t (s), Li = length of section i (mi); FFSi = free-flow speed of section i (mph), DRUi,t = delay rate for undersaturated section i at time t (s/mi), and ROi t,∆ = additional average delay rate due to oversaturation for section i at time t (s/mi). Adjustments for Weaving Sections As mentioned above, the basic approach applies the speed–flow model for basic freeway seg- ments to estimate a freeway section’s delay rate and travel speed. When applied to weaving sections, a capacity adjustment factor is required to account for the generally lower capacity in weaving sec- tions compared to basic sections. With this adjusted capacity, the basic section planning method can be applied to weaving sections. The model is as follows: = − + ≤0.884 0.0752 0.0000243 1.00 Equation 23weaveCAF V Lr s where CAFweave = capacity adjustment factor used for a weaving section (decimal), Vr = ratio of weaving demand flow rate to total demand flow rate in the weaving section (decimal), and Ls = weaving section length (ft). For a planning analysis, demand data for specific movements (e.g., freeway-to-ramp, ramp-to- ramp) may not be available. In these cases, it can be conservatively estimated that ramp-to-ramp demand is zero, and that the volume ratio Vr is the total on- and off-ramp demand divided by the sum of the (unconstrained) mainline demand entering the section and the on-ramp demand. Adjustments for Merge and Diverge Sections Merge Sections. Similar to weaving sections, a capacity adjustment factor CAFmerge is used to generate an equivalent merge section capacity that would yield speeds equivalent to a

54 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual basic section speed. In the absence of local data, a value of 0.95 is recommended for CAFmerge regardless of the merge configuration. However, a user-supplied CAF can also be used and is recommended for a merge segment with known capacity constraints and congestion impacts. Diverge Sections. For diverge segments, an average CAFdiverge value of 0.97 is recommended. Again, user-specific calibration of this factor is encouraged. Ramp Section Capacity Calculation. The overall capacity of ramp sections is determined from a length-weighted average of the capacity of the merge, basic, and diverge segments within a given section. Note that the effective length of merge and diverge segments are 1,500 feet each in the HCM. If the section is shorter than 3,000 feet, the length of the basic freeway segment is considered to be zero and the length of the merge and diverge segments is assumed to each be equal to half the section length. Computing Speed The procedure determines the travel rate TRi,t at section i at time t by adding the associated travel rate under free-flow conditions TRFFS and the delay rate DRi,t ___ . It then calculates travel time TTi,t by multiplying the travel rate by the section rate Li: = ∆ + Equation 24, ,TR TRi t R FFSi t = × Equation 25, ,TT TR Li t i t i The average speed Si,t in section i at time t is found as follows: = Equation 26, , S L TT i t i i t Level of Service To calculate level of service (LOS), the facility-wide average density is first computed and then the LOS letter is determined from a look-up table. The density Di,t of section i at time t is found by dividing the section demand di,t by its speed Si,t as follows: = Equation 27, , , D d S i t i t i t This mixed vehicle density is converted to units of passenger cars using Equation 28 and Equation 29: Equation 28D D PHF f PC HV = × ( )= + − 1 1 1 Equation 29f P E HV t HV

H. Freeway Analyses 55 where D = mixed vehicle density (veh/mi/ln), DPC = passenger car density (pc/mi/ln), PHF = peak hour factor (decimal), fHV = heavy vehicle factor (decimal), Pt = percent heavy vehicles (decimal), and EHV = passenger car equivalent for heavy vehicles (pc). Recommended default values for peak hour factor and percent heavy vehicles are provided in Exhibit 23. Default values for EHV are 2.0 for level terrain and 3.0 for rolling terrain. If specific values for grade, grade length, percent heavy vehicles, and the proportion of single-unit trucks to tractor- trailers are known, HCM Exhibits 12-26 through 12-28 provide more precise values for EHV. A weighted average, by lanes and length, of the section densities is used to obtain the average density for the facility. ∑ ∑= × × × Equation 30D D L N L N F i i i i i where DF = average passenger car density for the facility (pc/mi/ln), Di = passenger car density for section i (pc/mi/ln), Li = length of section i (mi), and Ni = number of lanes in segment i (ln). The facility and segment passenger car densities are entered into Exhibit 26 to obtain the level of service. Queues A segment is considered to be in 100% queue if its estimated density is greater than 45 passenger car equivalents per mile per lane (pc/mi/ln). For segments with densities below 45 pc/mi/ln, but demand-to-capacity ratios greater than 1.00, the section queue length is estimated by dividing the difference in lane demand and capac- ity by its density. It essentially provides an estimate for how long the queue would spillback at the given density, assuming a fixed number of lanes upstream of the bottleneck. Level of Service Urban/Suburban Freeway Average Facility or Section Density (pc/mi/ln) Rural Freeway Average Facility or Section Density (pc/mi/ln) A ≤ 11 ≤ 6 B >11–18 >6–14 C >18–26 >14–22 D >26–35 >22–29 E >35–45 >29–39 F >45 or any section has d/c>1.00 >39 or any section has d/c>1.00 Source: Adapted from HCM Exhibit 10-6. Note: d/c = demand-to-capacity ratio. Exhibit 26. Level of service criteria for freeway facilities.

56 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual ( ) = −max , 0 Equation 31, , , QL d c D i t i t i i i where QLi,t = queue length in segment i at time t (veh), di,t = demand on segment i at time t (veh/h), ci = capacity of segment i (veh/h), and Di,t = density on segment i at time t (veh/mi/ln). 7. Reliability The travel time on a facility will vary from hour to hour, day to day, and season to season of the year, depending on fluctuations in demand, weather, incidents, and work zones. Travel time reliability measures are an attempt to characterize this distribution of travel times for a selected period (often the non-holiday, weekday a.m. or p.m. peak period) of a year in some way meaningful to the analyst, the agency’s objectives, and the general public. Exhibit 27 shows two measures (the 95th percentile travel time index and the percent of trips less than 45 mph) out of many possible measures for characterizing the travel time distribution and communicating travel time reliability to decision-makers and the public. The agency and the analyst may choose other measures or other thresholds (such as the 85th percentile travel time index) for character- izing reliability. (The travel time index is the ratio of the actual or average travel time, depending on the context, to the travel time at free-flow speed.) The HCM (2016) provides a relatively data- and computationally intensive method for evalu- ating freeway reliability. The Florida DOT has also developed a reliability analysis procedure (Elefteriadou et al. 2012). Both methods provide defaults for many of the required inputs, but both require custom software to apply. As alternatives, this section describes how the HCM method can be applied with default values and presents a simplified method for estimating the two performance measures shown in Exhibit 27. Exhibit 27. Two measures for characterizing travel time reliability.

H. Freeway Analyses 57 HCM Method Using Defaults The HCM method for estimating travel time reliability is described in HCM Chapter 11. Exhibit 28 lists the required inputs and identifies which ones have default values available in the HCM. Simplified Method The following equations can be used to estimate freeway facility travel time reliability (Economic Development Research Group et al. 2014, Cambridge Systematics 2014, Elefteriadou et al. 2012). First, the average annual travel time rate (hours per mile), including incident effects, is computed: ( )= + × +1 Equation 32TTI FFS RDR IDRm = − 1 1 Equation 33RDR S FFS [ ]( )= − − × × ≤0.020 2 0.003 1.00 Equation 3412IDR N X X where TTIm = average annual mean travel time index (unitless), FFS = free-flow speed (mph), RDR = recurring delay rate (h/mi), IDR = incident delay rate (h/mi), S = peak hour speed (mph), N = number of lanes in one direction (N = 2 to 4), and X = peak hour volume-to-capacity ratio (decimal). Data Category Description Defaults Time Periods Study period, reliability reporting period Must be selected by the analyst Demand Patterns Day-of-week by month-of-year demand factors Urban: HCM Exhibit 11-18 Rural: HCM Exhibit 11-19 Weather Probabilities of various intensities of rain, snow, cold, and low visibility by month Determined by nearest city; specific values provided in the HCM Volume 4 Technical Reference Library Incidents Crash rate, incident-to-crash ratio, incident type probability, average incident duration by type Crash rate: must be provided Incident-to-crash ratio: 4.9 Others: HCM Exhibit 11-22 Work Zones and Special Events Changes to base conditions (e.g., demand, number of lanes), schedule for occurrence Optional inputs Nearest City Main city in nearest metropolitan area Required to look up weather defaults Traffic Counts Demand multiplier for demand represented in base dataset Must be provided; equals 1.00 when demands represent AADT Source: Adapted from HCM (2016), Exhibit 11-10. Notes: The study period is the portion of the day (e.g., 5 a.m. to 10 a.m.) in which travel time reliability will be evaluated. The reliability reporting period is the specific set of days (e.g., all non-holiday weekdays in a year) for which travel time reliability will be evaluated. AADT = annual average daily traffic. Exhibit 28. Input data needs for HCM travel time reliability analysis of freeways.

58 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual Values of X greater than 1.00 should be capped at 1.00, and values of N greater than 4 should be capped at 4, for use in Equation 34. Also note that Equation 34 does not explicitly account for differences in significant weather events between facilities and regions. Next, the 95th percentile travel time index TTI95 and percent of trips traveling under 45 mph PT45 can be computed from the average annual travel time index TTIm according to the following equations. ( )= + ×1 3.67 ln Equation 3595TTI TTIm [ ]( )= − − × −1 exp 1.5115 1 Equation 3645PT TTIm 8. Adaptations for Advanced Freeway Management Practices Although much remains unsettled as to the precise impacts of advanced freeway management practices on freeway capacities and speeds, there is some research on some practices that can be summarized here. Active Transportation and Demand Management (ATDM) HCM Chapter 37, ATDM: Supplemental, provides a general introduction to ATDM strategies and their likely effects on capacity, speed, and travel time reliability. Ramp Metering Ramp metering can result in more efficient merging at the ramp merge. Zhang and Levinson (2010) suggest that ramp metering can increase freeway mainline bottleneck capacity by 2% to 3% by smoothing out demand surges. Additional information on the capacity and performance analysis of dynamic ramp metering can be found in HCM Chapter 37, Section 4. HOV and HOT Lanes Single-lane high-occupancy vehicle (HOV) and high-occupancy toll (HOT) lanes restrict the ability of vehicles in those lanes to pass each other. Thus, capacities are somewhat lower in these situations than they are for the equivalent mixed flow lanes on the freeway, depending on how the HOV or HOT lane is separated from the rest of the lanes on the freeway. NCHRP Web-only Document 191 (Wang et al. 2012) suggests that capacities on the order of 1,600 to 1,700 vehicles per hour per lane may be appropriate for single HOV and HOT lanes. Section 4 of HCM Chapter 10, Freeway Facilities Core Methodology, provides additional information on the capacity and performance analysis of managed lanes (e.g., HOV and HOT lanes) on freeways. Temporary Shoulder Use Temporary shoulder use opens the shoulder lane to traffic for limited periods each day. Tenta- tive data suggest that the capacity and speed on a temporary shoulder lane are lower than for the adjacent full-time lanes. Work Zones Section 4 of HCM Chapter 10 provides information on the capacity and performance analysis of work zones on freeways.

H. Freeway Analyses 59 Speed Harmonization Variable speed limit and speed harmonization installations are intended to give drivers advance notice of downstream slowing and to provide recommended speeds for upstream driv- ers to reduce the shockwaves on freeways. These installations are intended to improve safety and reduce the effects of primary incidents on freeway operations. The magnitudes of these effects depend on the specifics of the installations. At the time of writing, this topic was the subject of FHWA research and it was not clear what the precise effects would be. Autonomous and Connected Vehicles Autonomous, automated, and connected vehicles have the potential to increase or decrease freeway capacities and speeds, depending on the specifics of their implementation. These vehi- cles may increase reliability by reducing collisions. At the time of writing, this topic was the subject of FHWA research and it was not clear what the precise effects would be. 9. Multimodal Level of Service The HCM does not provide level of service (LOS) measures for trucks, transit, bicycles, and pedestrians on a freeway facility. This section describes alternatives, where applicable. Truck LOS Truck level of service is defined in NCFRP Report 31 (Dowling et al. 2014) as a measure of the quality of service provided by a facility for truck hauling of freight, as perceived by shippers and carriers. It is measured in terms of the percentage of ideal conditions achieved by the facility for truck operations. Section P of the Guide describes how to calculate truck LOS for freeway facilities. Transit LOS The HCM does not provide a transit LOS measure for freeways. In general, buses will expe- rience the same conditions as other vehicles in the general purpose or managed lanes (where applicable) and could be assigned the same LOS as for motorized vehicle traffic generally. Alternatively, where buses stop along the freeway facility to serve passengers, the transit LOS measure for urban streets described in Section 4 of the Guide could be applied to the stops along the freeway facility, with appropriate adjustments to the assumed average passenger trip length and baseline travel time rate, and considering the pedestrian LOS of the access route to the stop. Bicycle and Pedestrian LOS Bicycle and pedestrian LOS are not generally applicable to freeways because access is usually limited to motor vehicles. Where a multilane path is provided within the freeway right-of-way, its LOS can be estimated using the procedure described in Section O8 of this Guide. 10. Example An example application of the simplified freeway facility method is provided in Case Study 1 in Section T of the Guide.

60 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual 11. References Cambridge Systematics, Inc. IDAS User’s Manual, Table B.2.14. http://idas.camsys.com/documentation.htm, accessed August 14, 2014. Dowling, R., G. List, B. Yang, E. Witzke, and A. Flannery. NCFRP Report 31: Incorporating Truck Analysis into the Highway Capacity Manual. Transportation Research Board of the National Academies, Washington, D.C., 2014. Economic Development Research Group, Inc.; Cambridge Systematics, Inc.; ICF International; Texas A&M Transportation Institute; and Weris, Inc. Development of Tools for Assessing Wider Economic Benefits of Transportation. SHRP 2 Report S2-C11-RW-1. Transportation Research Board of the National Academies, Washington, D.C., 2014. Elefteriadou, L., C. Lu, Z. Li, X. Wang, and L. Jin. Multimodal and Corridor Applications of Travel Time Reliability. Final Report, FDOT Contract BDK77 977-10. University of Florida, Gainesville, March 30, 2012. Highway Capacity Manual: A Guide to Multimodal Mobility Analysis. 6th ed. Transportation Research Board, Washington, D.C., 2016. Wang, Y., X. Liu, N. Rouphail, B. Schroeder, Y. Yin, and L. Bloomberg. NCHRP Web-Only Document 191: Analysis of Managed Lanes on Freeway Facilities. Transportation Research Board of the National Academies, Washington, D.C., Aug. 2012. http://onlinepubs.trb.org/onlinepubs/nchrp/nchrp_w191.pdf. Zhang, L., and D. Levinson. Ramp Metering and Freeway Bottleneck Capacity. Transportation Research Part A, Vol. 44. Elsevier Ltd., New York, 2010, pp. 218–235.

61 I. Multilane Highways 1. Overview Multilane highways are roadways with a minimum of two lanes in each direction, with traffic signals, roundabouts, or intersections where highway traffic stops (if any) must be spaced more than 2 miles apart. They have either no access control or partial control of access. This section presents medium-level methods for evaluating single multilane highway sections and facilities. 2. Applications The procedures in this section are designed to support the following planning and preliminary engineering analyses: • Developing a highway corridor improvement plan, • Assessing the impact on facility operations of changing or adding more intersection controls, and • Preparing traffic impact studies for land development. 3. Analysis Methods Overview The HCM provides a method for estimating the performance of multilane highway sections between intersections. It does not provide a method for evaluating multilane highway facilities that combines the operations of uninterrupted-flow sections with the operations of signalized intersections, stop-controlled intersections, or roundabouts located intermittently along the highway. This section presents three analysis methods for planning and preliminary engineering appli- cations, as indicated by the unshaded boxes in Exhibit 29: 1. A high-level screening and scoping method that can be used to focus the analysis on only those locations and time periods requiring investigation; 2. The HCM medium-level method for evaluating multilane highway section performance using defaults; and 3. A medium-level procedure for combining intersection and section performance into an estimate of overall multilane facility performance.

62 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual 4. Scoping and Screening Method Generalized Service Volume Table Whether or not a more detailed multilane highway analysis is needed can be determined by comparing the counted or forecasted peak hour or daily traffic volumes for the sections of the highway between each major intersection to the values given in Exhibit 30. If all of the section volumes fall in the LOS E range or better, there will be no congestion spillover requiring a full facility analysis to better quantify the performance of the facility. One can then use the HCM multilane highway section analysis procedures, with defaults for some of the inputs, to evaluate the performance of each section. The service volumes in Exhibit 30 can also be used to quickly determine the geographic and temporal extent of the multilane highway that will require analysis. If the counted or fore- High Level Medium Level Low Level Exhibit 29. Analysis options for multilane highways. Exhibit 30. Daily and peak hour service volume and capacity table for multilane highway sections. Area Type Terrain Peak Hour Peak Direction (veh/h/ln) AADT (2-way veh/day/ln) LOS A-C LOS D LOS E (capacity) LOS A-C LOS D LOS E (capacity) Urban Level 1,360 1,700 1,940 12,600 15,700 17,900 Urban Rolling 1,270 1,580 1,800 11,800 14,600 16,700 Rural Level 1,220 1,520 1,730 10,200 12,600 14,400 Rural Rolling 1,100 1,370 1,560 9,200 11,400 13,000 Notes: Entries are maximum vehicle volumes per lane that can be accommodated at stated level of service (LOS). AADT = annual average daily traffic. AADT per lane is two-way AADT divided by the sum of lanes in both directions. Urban area assumptions: Free-flow speed = 60 mph, 8% trucks, 0% buses, 0% RVs, peak hour factor = 0.95, capacity adjustment factor for driver population = 1.00, K-factor = 0.09, D-factor = 0.60. Rural area assumptions: Free-flow speed = 60 mph, 12% trucks, 0% buses, 0% RVs, peak hour factor = 0.88, capacity adjustment factor for driver population = 1.00, K-factor = 0.10; D-factor = 0.60. Similar tables can be developed by adjusting input values to reflect other assumptions. The K-factor is the ratio of weekday peak hour two-way traffic to AADT. The D-factor is the proportion of peak hour traffic in the peak direction.

I. Multilane Highways 63 casted volumes for a section fall below the agency’s target LOS standard, then the section can be excluded from a more detailed analysis. Any section that exceeds the capacity values in Exhibit 30 will have queuing that may impact upstream sections and reduce downstream demands. In such a situation, a full facility analysis is required to ascertain the highway’s performance. At present, the HCM does not provide such an analysis procedure, so the analyst would have to resort to microsimulation or some other system analysis approach. The analyst may also use the capacities in Exhibit 30 to compute the peak hour, peak direction demand-to-capacity ratio for each section under various improvement options. The options can then be quickly ranked according to their forecasted demand-to-capacity ratios for the critical sections of the highway. Estimating Multilane Highway Service Volumes The approximate maximum annual average daily traffic (AADT) (two-way) that can be accom- modated by a multilane highway at a given level of service can be estimated from Exhibit 30. For example, a four-lane highway (two lanes in each direction) can carry between 49,600 (12,400 × 4 lanes) and 65,600 (16,400 × 4 lanes) AADT at LOS E, depending on its location (urban or rural) and the terrain type. Higher AADTs can be accommodated at lower K- (peak hour proportion) and D- (directional proportion) factors. Note that the values in this simple example are shown to the nearest hundred but the final result should be considered accurate to the nearest thousand. A multilane highway in an urban setting delivers between 85% and 90% of the capacity per lane as an urban freeway. A rural multilane highway delivers 95% to 98% of the capacity per lane as a rural freeway. When local traffic data suggests that other values for the assumptions than those noted in Exhibit 30 would be more appropriate, the analyst should modify the daily and hourly service volumes using this equation: = × × × × × × × × Equation 370 0 0 ,0 ,0 0 DSV DSV f CAF PHF K D K D f CAF PHF HV p HV p where DSV = daily service volume (veh/day/ln), DSV0 = initial daily service volume in Exhibit 30 (veh/day/ln), fHV, fHV,0 = desired and initial adjustment factors, respectively, for presence of heavy vehicles in the traffic stream, CAFp, CAFp,0 = desired and initial capacity adjustment factors, respectively, for unfamiliar driver populations, PHF, PHF0 = desired and initial peak hour factors, respectively, K, K0 = desired and initial proportions, respectively, of daily traffic occurring during the peak hour, and D, D0 = desired and initial proportions, respectively, of traffic in the peak direction during the peak hour. The same equation can be used to modify the peak hour, peak direction service volumes if the initial peak hour service volumes from Exhibit 30 are used instead of the daily values. The heavy vehicle adjustment factor fHV used in the service volume table is computed using the following adaptation of HCM Equation 12-10 (HCM 2016): ( )= + × − 1 1 1 Equation 38f P E HV HV HV

64 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual where fHV = heavy vehicle adjustment factor (decimal), PHV = percentage heavy vehicles (decimal), and EHV = heavy vehicle equivalence from Exhibit 31. For convenience, all heavy vehicles are assigned a single PCE value from Exhibit 31. Daily service volumes should be rounded to the nearest hundred vehicles, given the many default values used in their computation. Peak hour, peak direction service volumes should be rounded to the nearest ten vehicles. 5. Section Analysis Using HCM with Defaults HCM Chapter 12, Basic Freeway and Multilane Highway Segments, describes the method for evaluating the capacity, speed, density, and LOS for multilane highway sections without major intersections (i.e., intersections that slow down or stop through traffic on the mainline). Data Requirements Exhibit 32 lists the data needed to evaluate the full range of performance measures for HCM multi lane highway section analysis and for the multilane facility analysis method described in this section. To evaluate multilane highway sections at a facility level, all of the HCM section-level data listed in Exhibit 32 are required (including section length), plus the intersection-level data for each of the intersection or interchange types found along the multilane facility. Section Free-Flow Speeds The free-flow speed, representing the speed drivers would choose based only on the highway’s horizontal and vertical alignment, is a critical input for calculating most multilane highway performance measures. Input Data (units) For HCM Section For Facility Method Default Value Hourly directional volume (veh/h) • • Must be provided Number of directional lanes • • Must be provided Terrain type (level, rolling, etc.) • • Must be provided* Lane width (ft) • • 12 Total lateral clearance (ft) • • 12 Access points/mile • • 8 (rural), 16 (low-density suburban), 25 (high-density suburban) Free-flow speed (mph) • • Must be provided Percentage heavy vehicles (%) • • 10 (rural), 5 (suburban)** Peak hour factor (decimal) • • 0.88 (rural), 0.95 (suburban) Section length (mi) • Must be provided Intersection performance data • Must be provided Notes: See HCM Chapter 12 for definitions of the required input data. *Heavy vehicle impacts on traffic flow on long (≥1 mi) and steep (>4%) grades with relatively few (<5%) trucks can be significantly more severe than the default value for mountainous terrain would indicate. Consideration should be given to developing specific passenger car equivalent values for mountainous sections where these conditions are met. **HCM Chapter 26, Section 2, provides state-specific default values. Exhibit 32. Required data for multilane highway section analysis. Terrain Type EHV Level 2.0 Rolling 3.0 Mountainous 5.0 Source: Adapted and extrapolated from HCM (2016), Exhibit 12-25. Exhibit 31. Heavy vehicle equivalence values for multilane highways.

I. Multilane Highways 65 The most-accurate method for estimating free-flow speed is to measure it in the field under low-flow (less than 1,000 vehicles per hour per lane) conditions. The free-flow speed would be the average of the observed spot speeds under those low-flow conditions. The second-best method is to estimate the free-flow speed using the method provided in the HCM. The third-best method to estimate the free-flow speed is to use the posted speed limit plus an adjustment deemed appropriate by the analyst (for example: posted speed limit plus 5 mph). The result should be rounded to the nearest 5 mph. Should the posted speed limit for trucks or other vehicle classes be lower than that for other vehicle types, then the analyst will have to apply some judgment based on local experience to estimate the free-flow speed. Section Capacities The capacity of a multilane highway section depends upon its free-flow speed, the peak hour factor, and the effect of heavy vehicles. The HCM also offers a capacity adjustment factor for driver population that adjusts capacity downward, but planning analyses often assume that drivers are familiar with the highway and, thus, no capacity adjustment is made for the driver population. 6. Multilane Facility Analysis Method The multilane highway facility analysis combines the performance estimates produced by the HCM multilane highway section analysis method with the performance results for any con- trolled intersections on the facility. A controlled intersection is one where the mainline through traffic is required to stop or slow down, such as at a traffic signal, an all-way stop, or a round- about (see Exhibit 33). A stretch of highway between two controlling intersections may be split into multiple highway sections where there are significant changes in the capacity of the highway (usually caused by changes in the grade, alignment, or number of lanes). Estimation of Facility Free-Flow Speed The facility free-flow speed may be estimated three ways. In order of decreasing accuracy, these are: • Field measurement. The free-flow speed may be directly measured in the field at flow rates below 1,000 vehicles per hour per lane, when measured at least one-half mile from a major intersection (i.e., an intersection where a traffic signal, stop sign, or roundabout requires mainline traffic to slow down or stop). The Manual of Transportation Engineering Studies (Schroeder et al. 2010) describes spot speed measurement techniques. • HCM estimation method. The HCM multilane highway section method may be used to esti- mate the free-flow speed. This method is likely to be less accurate than field measurement, but it requires fewer resources. • Estimate from posted speed. The free-flow speed may be estimated as the posted or statutory speed limit plus an adjustment that the analyst judges to be appropriate, often 5 to 7 mph. This method is likely to be the least accurate of the three approaches, but it requires the least Exhibit 33. Controlled intersections and sections on highway facility.

66 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual resources and the accuracy is likely to be sufficient for most planning and preliminary engi- neering applications. Level of Service The HCM does not define LOS at a facility level for multilane highways. However, the multilane highway analysis method described in HCM Chapter 12 can be used to estimate the LOS of the uninterrupted-flow sections between major intersections, while the appropriate HCM method for signalized intersections (Chapter 19), all-way stops (Chapter 21), or roundabouts (Chapter 22) can be used to estimate the LOS of the major intersections. The worst case results can be reported for sections and major intersections. Volume-to-Capacity Ratio The volume-to-capacity ratios are examined for each section and major intersection along the facility. If it is desired to convey a single value to decisionmakers then the highest volume-to- capacity ratio should be reported for the facility. Highway Sections The capacities shown in Exhibit 30 may be used to estimate section capacities between con- trolled intersections. The more detailed HCM section analysis methods with defaults may be used for a more precise estimate. Controlled Intersections The intersection through movement capacities are estimated using the HCM and the proce- dures described later in this Guide in Sections L (signalized intersections), M (stop-controlled intersections), and N (roundabouts). Average Travel Speed and Travel Time The total travel time for the facility is computed by summing the section travel times and the intersection delays to mainline through movements. The average speed for the facility is obtained by dividing the length of the facility by the total travel time. Highway Sections Average travel speed is computed by the HCM method for individual sections. The average travel time for a section (excluding any intersection delays) is calculated as the section length divided by the estimated average section speed: = × 3,600 Equation 39section section section TT L S where TTsection = average section travel time (s), Lsection = section length, including the downstream intersection (mi), Ssection = average section travel speed (mph), and 3,600 = number of seconds in an hour (s/h). The following equation, adapted from HCM Equation 12-1 and Exhibit 12-6, can be used to estimate average section travel speed. The percent base capacity in the equation is used to convert capacities from vehicles per hour per lane into passenger car equivalents.

I. Multilane Highways 67 ( )= + ×1 Equation 40section section S FFS a v c b where Ssection = average section travel speed (mph), FFSsection = section free-flow speed (mph), and a, b = parameters as given in Exhibit 34. Facilities For facility analyses, the effects of intersection delays at intersections need to be accounted for. The average travel time along a multilane highway facility is estimated by adding inter- section delays for through traffic to the estimated section travel times. The average travel speed for through traffic on the facility is then determined by dividing the total travel time into the facility length. ∑ ∑= + Equation 41facility ,TT TT dii i thrui = × 3,600 Equation 42facility facility facility S L TT where TTfacility = average facility travel time (s), TTi = average section travel time for section i (s), di,thru = average through-vehicle intersection control delay at the intersection at the down- stream end of section i (s), Sfacility = average through-vehicle facility travel speed (mph), Lfacility = facility length (mi), and 3,600 = number of seconds in an hour (s/h). Vehicle-Hours of Delay Vehicle-hours of delay are calculated by comparing the travel time at an analyst-defined tar- get travel speed to the average travel time, and multiplying by the number of through vehicles. The HCM defines the target travel speed as the free-flow speed. However, some agencies use the speed limit as the basis for calculating delay, while others choose a threshold or policy speed that the agency considers to be its minimum desirable operating speed. Free-Flow Speed (mph) a b 70 0.37 6.9 65 0.27 7.3 60 0.23 7.5 55 0.18 7.7 50 0.13 8.1 45 0.07 8.9 Note: This equation produces speed estimates for multilane highways within 2 mph of the HCM-estimated speed for v/c ratios ≤1.00. Exhibit 34. Parameters for multilane highway speed estimation.

68 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual = × 3,600 Equation 43target,section section target,section TT L S ( ) = + − × ≥ 3,600 0 Equation 44section section target,section section,thru VHD TT d TT Vthru ∑= Equation 45facilityVHD VHDii where TTtarget, section = target travel time for a section (s), Lsection = section length, including the downstream intersection (mi), Starget,section = target travel speed for the section (mph), 3,600 = number of seconds in an hour (s/h), VHDsection = vehicle-hours of delay to through vehicles in a section (veh-h), TTsection = average section travel time (s), dthru = average through-vehicle intersection control delay at the intersection at the downstream end of the section (s), Vsection,thru = vehicle directional demand volume for the section (veh), VHDfacility = vehicle-hours of delay to through vehicles on the facility (veh-h), and VHDi = vehicle-hours of delay to through vehicles in section i (veh-h). Person-Hours of Delay Person-hours of delay for a section or facility is the corresponding vehicle-hours of delay, multiplied by an assumed average vehicle occupancy. Density Section density is computed according to the following equation, adapted from HCM Equation 12-11: ( ) = Equation 46section section section D V N S where Dsection = section density (pc/mi/ln), Vsection = vehicle directional demand volume for the section (veh), N = number of directional lanes (ln), Ssection = average section travel speed (mph). Queuing A section is considered 100% in queue if its density exceeds 45 pc/mi/ln (the density at capacity given in HCM Exhibit 12-6). Queues are meaningful on multilane highways only at the specific bottlenecks causing the queues. Thus queues are estimated and reported by bottleneck (for example, using the appropriate intersection queuing estimation method). 7. Reliability There is currently no method in the HCM or in the literature for estimating the reliability of rural or urban multilane highways.

I. Multilane Highways 69 8. Multimodal LOS Bicycle LOS The HCM provides a bicycle LOS measure for multilane highways. For details, see Section O3 in this Guide. Pedestrian LOS The HCM does not provide a pedestrian LOS measure for multilane highways. However, the pedestrian LOS measure for urban streets (see Section O4) was developed in part using data from urban multilane highways and can be applied to facilities whose characteristics are within the range of those used to develop the model (in particular, posted speeds of 50 mph or less). Transit LOS The HCM does not provide a transit LOS measure for multilane highways. However, similar to freeways, if bus service exists along the highway and makes stops to serve passengers, the transit LOS measure for urban streets described in Section O4 of the Guide could be applied to the stops along the multilane highway, with appropriate adjustments to the assumed average passenger trip length and baseline travel time rate. Truck LOS The truck LOS estimation procedure described in Section P can be used to estimate truck LOS for multilane highways. 9. Example Preparation of an example problem was deferred to a future edition of the Guide. 10. References Highway Capacity Manual: A Guide to Multimodal Mobility Analysis. 6th ed. Transportation Research Board, Washington, D.C., 2016. Schroeder, B. J., C. M. Cunningham, D. J. Findley, J. E. Hummer, and R. S. Foyle. Manual of Transportation Engineering Studies, 2nd ed. Institute of Transportation Engineers, Washington, D.C., 2010.

70 J. Two-Lane Highways 1. Overview Two-lane highways have one lane for the use of traffic in each direction. The principal characteristic that separates the traffic performance of two-lane highways from other uninterrupted- flow facilities is that passing maneuvers may be allowed to take place in the opposing lane of traffic. Passing maneuvers are limited by the availability of gaps in the opposing traffic stream and by the availability of sufficient sight distance for a driver to discern the approach of an opposing vehicle safely. As demand flows and geometric restrictions increase, opportunities to pass decrease. This creates platoons within the traffic stream, with trailing vehicles subject to additional delay because of the inability to pass the lead vehicles. Because passing capacity decreases as passing demand increases, two-lane highways exhibit a unique characteristic: operating quality often decreases precipitously as demand flow increases, and operations can become “unacceptable” at relatively low volume-to-capacity ratios. For this reason, few two-lane highways ever operate at flow rates approaching capacity; in most cases, poor operating quality has led to improvements or reconstruction long before capacity demand is reached. Two-lane highways have no access control or partial control of access. Traffic signals, round- abouts, or stop signs controlling highway traffic may be found along two-lane highways but must be spaced at least 2 miles apart if the roadway is to be considered a two-lane highway for the purposes of the analysis methods presented in this section. 2. Applications The procedures in this section are designed to support the following planning and preliminary engineering analyses: • Developing a highway corridor improvement plan, • Assessing the impacts on facility performance of changing or adding intersection controls, • Preparing feasibility studies of truck climbing lanes and passing lanes, and • Conducting traffic impact studies for land development. 3. Analysis Methods Overview The HCM provides a method for estimating the performance of two-lane highway sections between intersections. It does not provide a method or LOS measures for evaluating two-lane high-

J. Two-Lane Highways 71 way facilities, combining the operations of sections with signalized intersections, stop-controlled intersections, or roundabouts. This chapter presents three analysis methods for planning and preliminary engineering applications, as indicated by the unshaded boxes in Exhibit 35: 1. A high-level screening and scoping method that can be used to focus the analysis on only those locations and time periods requiring investigation, 2. The HCM medium-level method for evaluating two-lane highway section performance using defaults, and 3. A medium-level procedure for combining intersection and section performance to estimate overall two-lane highway facility performance. 4. Scoping and Screening Generalized Service Volume Table Whether or not a more detailed two-lane highway analysis is needed can be determined by comparing the counted or forecasted peak hour or daily traffic volumes for the sections of the highway between major intersections (i.e., intersections where highway traffic must stop or slow due to a traffic signal or other form of traffic control) to the values given in Exhibit 36. If all of the section volumes fall in the LOS E range or better, there will be no congestion spillover requiring a full facility analysis to better quantify the performance of the facility. One can then use the HCM two-lane highway section analysis procedures with defaults for some of the inputs to evaluate the performance of each section. The service volumes in Exhibit 36 can also be used to quickly determine the geographic and temporal extent of the two-lane highway that will require analysis. If the counted or fore- casted volumes for a section fall below the agency’s target LOS standard, then the section can be excluded from a more detailed analysis. Any section that exceeds the capacity values in Exhibit 36 will have queuing that may impact upstream sections and reduce downstream demands. In such a situation, a full facility analysis High Level Medium Level Low Level Exhibit 35. Analysis options for two-lane highways.

72 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual is required to ascertain the performance of the highway. At present, the HCM does not provide such an analysis procedure, so the analyst would have to resort to microsimulation or some other system analysis approach. The analyst may also use the capacities in Exhibit 36 to compute the peak hour, peak direction demand/capacity ratio for each section under various improvement options. The options can then be quickly ranked according to their forecasted demand/capacity ratios for the critical sections of the highway. Estimating Two-Lane Highway Service Volumes The approximate maximum two-way annual average daily traffic (AADT) that can be accom- modated by a two-lane highway at a given LOS can be estimated from Exhibit 36. For example, a two-lane highway can carry between 24,100 and 24,900 AADT at LOS E, depending on its class and the terrain type. Higher AADTs can be accommodated at lower K- (peak hour proportion) and D- (directional proportion) factors. When local traffic data suggest that other values for the assumptions than those noted in Exhibit 36 would be more appropriate, the analyst should modify the daily and hourly service volumes using the following equation: = × × × × × × Equation 470 0 0 ,0 0 DSV DSV f PHF K D K D f PHF HV HV where DSV = daily service volume (veh/day/ln), DSV0 = initial daily service volume in Exhibit 30 (veh/day/ln), Highway Type Terrain Peak Hour Peak Direction (veh/h) AADT (2-way veh/day) LOS A-C LOS D LOS E (capacity) LOS A-C LOS D LOS E (capacity) Class I Level 440 750 1,490 7,300 12,500 24,900 Class I Rolling 340 690 1,450 5,600 11,500 24,100 Class II Rolling 430 790 1,490 7,100 13,100 24,900 Source: Adapted from HCM (2016), Exhibit 15-5. Notes: AADT = annual average daily traffic, LOS = level of service. Entries are maximum vehicle volumes that can be accommodated at the stated LOS. Class I highways are highways where motorists expect to travel at relatively high speeds. Class II highways are highways where motorists do not necessarily expect to travel at high speed (e.g., access routes to Class I highways, scenic and recreational highways). Assumed values for Class I—level: base free-flow speed = 65 mph and 20% no-passing zones. Assumed values for Class I—rolling: base free-flow speed = 60 mph and 40% no-passing zones. Assumed values for Class II—rolling: base free-flow speed = 50 mph and 60% no-passing zones. The K-factor (ratio of weekday peak hour two-way traffic to AADT) is assumed to be 0.10 for all classes. The D-factor (proportion of peak hour traffic in the peak direction) is assumed to be 0.60 for all classes. The peak hour factor is assumed to be 0.88 for all classes. Values can be adjusted for other assumptions. Exhibit 36. Daily and peak hour service volume and capacity table for two-lane highway sections.

J. Two-Lane Highways 73 fHV, fHV,0 = desired and initial adjustment factors, respectively, for presence of heavy vehicles in the traffic stream, PHF, PHF0 = desired and initial peak hour factors, respectively, K, K0 = desired and initial proportions, respectively, of daily traffic occurring during the peak hour, and D, D0 = desired and initial proportions, respectively, of traffic in the peak direction during the peak hour. The same equation can be used to modify the peak hour, peak direction service volumes if the initial peak hour service volumes from Exhibit 36 are used instead of the daily values. The heavy vehicle adjustment factor fHV used in the service volume table is computed using the following adaptation of HCM Equation 15-4 (HCM 2016): 1 1 1 Equation 48f P E HV HV HV( )= + × − where fHV = heavy vehicle adjustment factor (decimal), PHV = percentage heavy vehicles (decimal), and EHV = heavy vehicle equivalence from Exhibit 37. For convenience, all heavy vehicles are assigned a single PCE value from Exhibit 37 above. Daily service volumes should be rounded to the nearest hundred vehicles, given the many default values used in their computation. Peak hour, peak direction service volumes should be rounded to the nearest ten vehicles. 5. Section Analysis Using HCM with Defaults HCM Chapter 15, Two-Lane Highways, describes the method for evaluating the capacity, speed, density, and LOS for two-lane highway sections without major intersections (intersections that slow down or stop through traffic on the mainline). HCM Highway Classes Two-lane highway sections are divided into three classes for the purpose of LOS analysis (HCM 2016): • Class I two-lane highways are highways where motorists expect to travel at relatively high speeds. Two-lane highways that are major intercity routes, primary connectors of major traffic generators, daily commuter routes, or major links in state or national highway networks are generally assigned to Class I. These facilities serve mostly long-distance trips or provide the connections between facilities that serve long-distance trips. • Class II two-lane highways are highways where motorists do not necessarily expect to travel at high speeds. Two-lane highways functioning as access routes to Class I facilities, serving as scenic or recreational routes (and not as primary arterials), or passing through rugged terrain (where high-speed operation would be impossible) are assigned to Class II. Class II facilities most often serve relatively short trips, the beginning or ending portions of longer trips, or trips for which sightseeing plays a significant role. • Class III two-lane highways are highways serving moderately developed areas. They may be portions of a Class I or Class II highway that pass through small towns or developed Terrain Type EHV Level 1.1 Rolling 1.5 Mountainous 3.0 Source: Adapted and extrapolated from HCM (2016), Exhibit 15-11. Exhibit 37. Heavy vehicle equivalence values for two-lane highways.

74 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual recreational areas. On such sections, local traffic often mixes with through traffic, and the density of unsignalized roadside access points is noticeably higher than in a purely rural area. Class III highways may also be longer sections passing through more spread-out recreational areas, also with increased roadside densities. Such sections are often accompanied by reduced speed limits that reflect the higher activity level. Data Requirements Exhibit 38 lists the data needed to evaluate the full range of performance measures for HCM two-lane highway section analyses and for the two-lane facility analysis method described later. To evaluate two-lane highways at a facility level, all of the HCM section-level data listed in Exhibit 38 are required, plus the intersection data for the two-lane facility. Section LOS Section-level LOS is an output of the HCM method; step-by-step calculation details are provided in HCM Chapter 15. The HCM section method starts by estimating the free-flow speed based on the geometry of the section and the characteristics of the traffic demands (percent heavy vehicles). The average travel speed is then estimated, followed by the percent time-spent-following. Finally, the LOS and capacity are estimated. Exhibit 39 presents the automobile LOS criteria for two-lane highway sections for each highway class. The HCM does not define LOS at a facility level for two-lane highways. Input Data (units) For HCM Section For Facility Method Default Value Hourly two-directional volume (veh/h) • • Must be provided Directional split (%) • • 60/40 Locations and lengths of passing lanes • • Must be provided Terrain type (level, rolling, mountainous) • • Must be provided* Highway class (I, II, III) • • Must be provided Lane width (ft) • • 12 Shoulder width (ft) • • 6 Percentage no-passing zones (%) • • Level terrain: 20%, rolling: 40%, more extreme: 80% Access point density, one side (accesses/mi) • • Classes I and II: 8 per mile, Class III: 16 per mile Base free-flow speed (mph) • • Speed limit + 10 mph Percentage heavy vehicles (%) • • 6** Peak hour factor (decimal) • • 0.88 Section length (mi) • • Must be provided Intersection performance data • Must be provided Notes: See HCM Chapter 15 for definitions of the required input data. *Heavy vehicle impacts on traffic flow on long (≥1 mi) and steep (>4%) grades with relatively few (<5%) trucks can be significantly more severe than the default value for mountainous terrain would indicate. Consideration should be given to developing specific passenger car equivalent values for mountainous sections where these conditions are met. **HCM Chapter 26 provides state-specific default values. Exhibit 38. Required data for two-lane highway section analysis.

J. Two-Lane Highways 75 6. Two-Lane Facility Analysis Method The two-lane highway facility analysis combines the performance estimates produced by the HCM two-lane highway section analysis method with the performance results for any controlled intersections along the facility. A controlled intersection is one where mainline through traffic is required to stop or slow down, such as at a traffic signal, an all-way stop, or a roundabout (see Exhibit 40). A stretch of highway between two controlling intersections may be split into multiple highway sections where there are significant changes in the capacity of the highway (usually caused by changes in the grade or alignment). Facility Free-Flow Speed The facility free-flow speed may be estimated three ways: • The most accurate approach is to directly measure speeds under low-flow conditions in the field. The field-measured speeds must still be adjusted following the guidance provided in HCM Chapter 15. (It is difficult to find low enough volumes in the field for direct measurement, so the HCM adjustments are required.) • The next most accurate approach is to use the method in HCM Chapter 15 to estimate the free- flow speed. This method is likely to be less accurate than field measurement, but it requires fewer resources. • Finally, the free-flow speed may be estimated as the posted or statutory speed limit plus an adjustment that the analyst judges to be appropriate. For two-lane highways, the HCM recom- mends an upward adjustment of 10 mph (see HCM Exhibit 15-5). This method is likely to be the least accurate of the three approaches, but it requires the least resources and the accuracy is likely to be sufficient for most planning and preliminary engineering applications. Class I Highways Class II Highways Class III Highways LOS ATS (mph) PTSF (%) PTSF (%) PFFS (%) A >55 ≤35 ≤40 >91.7 B >50–55 >35–50 >40–55 >83.3–91.7 C >45–50 >50–65 >55–70 >75.0–83.3 D >40–45 >65–80 >70–85 >66.7–75.0 E ≤40 >80 >85 ≤66.7 F Demand > capacity Source: Adapted from HCM (2016), Exhibit 15-3. Notes: ATS = average travel speed (excluding intersection delays) (mph), PTSF = percent time-spent-following (%), PFFS = percent of free-flow speed away from signalized or other controlling intersections (e.g., roundabouts and all-way stops) (%). Exhibit 39. Automobile LOS for two-lane highway sections. Exhibit 40. Controlled intersections and sections on highway facility.

76 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual All of these approaches for estimating free-flow speed assume all vehicles have the same posted speed limit. Should the posted speed limit for trucks or other vehicle classes be lower than that for other vehicle types, the analyst will have to apply some judgment based on local experience when employing the above methods to estimate free-flow speed. Level of Service The HCM does not define LOS at a facility level for two-lane highways. However, the two-lane highway section analysis method described in HCM Chapter 15 can be used to estimate the LOS of the sections between controlled intersections, while the appropriate HCM method for signalized intersections (Chapter 19), all-way stops (Chapter 21), or roundabouts (Chapter 22) can be used to estimate the LOS of the controlled intersections. The worst case results can be reported for sections and controlled intersections. Volume-to-Capacity Ratio The volume-to-capacity ratios are examined for each section and controlled intersection along the facility. If it is desired to convey a single value to decision-makers, then the highest volume- to-capacity ratio should be reported for the facility. Highway Sections The capacities shown in Exhibit 36 may be used to estimate section capacities between controlled intersections. The more detailed HCM section analysis methods with defaults may be used for a more precise estimate. Controlled Intersections The intersection through movement capacities are estimated using the HCM and the procedures described in later in this Guide. Average Travel Speed and Average Travel Time The total travel time for the facility is computed by summing the section travel times and the intersection delays to mainline through movements. The average speed for the facility is obtained by dividing the length of the facility by the total travel time. Highway Sections Average travel speed is computed by the HCM method for individual sections. The average travel time for a section (excluding any intersection delays) is calculated as the section length divided by the section speed: 3,600 Equation 49section section section TT L S = × where TTsection = average section travel time (s), Lsection = section length, including the downstream intersection (mi), Ssection = average section travel speed (mph), and 3,600 = number of seconds in an hour (s/h).

J. Two-Lane Highways 77 Equation 50 is used to estimate average speed without the effects of passing lanes. The esti- mated free-flow speed should include the effects of narrow lane widths, restricted right side lateral clearance, and access point density (see HCM Chapter 15 for details). The heavy vehicle factor fHV in the equation is used to convert capacities from vehicles per hour to passenger car equivalents. 0.00776 Equation 50baseS FFS v v PHF f f d o HV NP= − + ×     − where Sbase = average speed in portions of the section not influenced by passing lanes (mph), FFS = free-flow speed (mph), vd = volume in the subject direction (veh/h), vo = volume in the opposite direction (veh/h), PHF = peak hour factor (decimal), fHV = heavy vehicle adjustment factor (decimal) from Equation 48, and fNP = no-passing adjustment factor (mph) from Exhibit 41. If no-passing lanes are provided in the section, the average section speed Ssection equals Sbase. Otherwise, one additional step calculates the average section speed as the length-weighted aver- age of the average speed within passing lanes, the average speed in the passing lanes’ downstream influence areas, and the average speed in the remainder of the section. Average speeds are 8 to 11 percent higher where passing lanes exist, relative to the base speed calculated in Equation 50. In addition, passing lanes provide some speed benefit for up to 1.7 miles beyond the end of the passing lane (HCM 2016). 0 0.5 0 Equation 51section base base section S S N S f L f L L L N pl pl pl pl de npl pl [ ]( ) ( )= = × + × +    > where Ssection = average section speed (mph), Sbase = average speed in portions of the section not influenced by passing lanes (mph), Npl = number of passing lanes in the section in the analysis direction, fpl = speed adjustment factor for passing lanes (decimal) from Exhibit 42, Lpl = total length of passing lanes in the section (mi), Lde = total length of passing lane downstream effect in the section (mi), Free-Flow Speed (mph) 200 veh/h < Opposing Volume < 500 veh/h All Other Opposing Volumes0% No-Passing 50% No-Passing 100% No-Passing 60 2 3 4 1 55 2 3 4 1 50 1 2 4 1 45 1 2 4 1 Source: Adapted from HCM (2016), Exhibit 15-15. Exhibit 41. No-passing adjustment factor (mph) for two-lane highway speed estimation. Directional Volume (veh/h) fpl ≤150 1.08 151–250 1.09 251–550 1.10 >550 1.11 Source: Adapted and extrapolated from HCM (2016), Exhibit 15-28. Exhibit 42. Speed adjustment factor for passing lanes.

78 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual Lnpl = total length of the section not influenced by passing lanes (mi), and Lsection = total section length (mi). The maximum value of Lde is 1.7 miles per passing lane, but this length should be reduced when either a new passing lane begins or the end of the section is reached within 1.7 miles of the end of a passing lane. The total length of the section not influenced by passing lanes Lnpl is then Lsection minus Lpl minus Lde,, with a minimum value of zero. Facilities For facility analyses, the effects of intersection delays at intersections need to be accounted for. The average travel time along a two-lane highway facility is estimated by adding inter section delays for through traffic to the estimated section travel times. The average travel speed for through traffic on the facility is then determined by dividing the total travel time into the facility length. ∑ ∑= + Equation 52facility ,TT TT dii i thrui = × 3,600 Equation 53facility facility facility S L TT where TTfacility = average facility travel time (s), TTi = average section travel time for section i (s), di,thru = average through-vehicle intersection control delay at the intersection at the down- stream end of section i (s), Sfacility = average through-vehicle facility travel speed (mph), Lfacility = facility length (mi), and 3,600 = number of seconds in an hour (s/h). Vehicle-Hours of Delay Vehicle-hours of delay are calculated by comparing the travel time at an analyst-defined target travel speed to the average travel time, and multiplying by the number of through vehicles. The HCM defines the target travel speed as the free-flow speed. However, some agencies use the speed limit as the basis for calculating delay, while others choose a threshold or policy speed that the agency considers to be its minimum desirable operating speed. = × 3,600 Equation 54target,section section target,section TT L S 3,600 0 Equation 55section section target,section section, VHD TT d TT Vthru thru( ) = + − × ≥ ∑= Equation 56facilityVHD VHDii where TTtarget, section = target travel time for a section (s), Lsection = section length, including the downstream intersection (mi), Starget,section = target travel speed for the section (mph), 3,600 = number of seconds in an hour (s/h),

J. Two-Lane Highways 79 VHDsection = vehicle-hours of delay to through vehicles in a section (veh-h), TTsection = average section travel time (s), dthru = average through-vehicle intersection control delay at the intersection at the down- stream end of the section (s), Vsection,thru = vehicle directional demand volume for the through section (veh), VHDfacility = vehicle-hours of delay to through vehicles on the facility (veh-h), and VHDi = vehicle-hours of delay to through vehicles in section i (veh-h). Person-Hours of Delay Person-hours of delay for a section or facility is the corresponding vehicle-hours of delay, multiplied by an assumed average vehicle occupancy. Density Section density is computed according to the following equation, adapted from HCM Equation 12-11: = × + 1 Equation 57section section section section D V S L L pl where Dsection = section density (pc/mi/ln), Vsection = vehicle directional demand volume for the section (veh), Ssection = average section travel speed (mph), Lpl = total length of passing lanes in the section (mi), and Lsection = section length, including the downstream intersection (mi). The 1 + (Lpl /Lsection) term in Equation 57 reduces the density according to the proportion of passing lanes (i.e., two lanes of travel in the analysis direction) in the section. Queuing Queues are meaningful on two-lane highways only at the specific bottlenecks causing the queues. Thus queues are estimated and reported by bottleneck (for example, using the appro- priate inter section queuing estimation method). Note that the HCM does not provide methods for evaluating nonintersection bottlenecks that may occur on two-lane highways where large midsection demand surges or significant changes in geometry (e.g., lane drops, grade changes) might create a bottleneck. Percent Time-Spent-Following Percent time-spent-following is used in determining LOS for Class I and Class II two-lane highways. To estimate this measure, the procedures described in HCM Chapter 15 should be used. 7. Reliability There is no method in the HCM or in the literature for estimating the reliability of rural two-lane highways.

80 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual 8. Multimodal LOS Bicycle LOS The HCM provides a bicycle LOS measure for two-lane highways. For details, see Section O3 in this Guide. Pedestrian LOS The HCM does not provide a pedestrian LOS measure for two-lane highways. Transit LOS The HCM does not provide a transit LOS measure for multilane highways. However, similar to freeways and multilane highways, if bus service exists along the highway and makes stops to serve passengers, the transit LOS measure for urban streets described in Section O4 of the Guide could be applied to the stops along the two-lane highway, with appropriate adjustments to the assumed average passenger trip length and baseline travel time rate. Truck LOS The truck LOS estimation procedure described in Section P can be used to estimate truck LOS for two-lane highways. 9. Example Preparation of an example problem was deferred to a future edition of the Guide. 10. Reference Highway Capacity Manual: A Guide to Multimodal Mobility Analysis. 6th ed. Transportation Research Board, Washington, D.C., 2016.

81 K. Urban Streets 1. Overview Any street or roadway with signalized intersections, stop- controlled intersections, or roundabouts that are spaced no farther than 2 miles apart can be evaluated using the HCM methodology for urban streets and the procedures described in this section. The planning methods for urban streets focus on facility-level analysis, segment-level analysis, and intersection-level analysis. Facility-level performance is estimated by summing the segment (between intersections) and intersection performance results. Interchange ramp terminals are a special case of intersection at the foot of freeway on- and off-ramps. They are addressed in HCM Chapter 23. The uneven nature of lane demands and the tight spacing between signals within a freeway interchange result in conditions that are not typical of an urban street. An urban street segment is a segment of roadway bounded by controlled intersections at either end that require the street’s traffic to slow or stop. An urban street facility is a set of contiguous urban street segments. The control delay at the downstream intersection defining a segment is included in the segment travel time. Exhibit 43 shows the relationship between an urban street facility, an urban street segment, and an intersection, as well as the segment travel time and inter- section control delay. The exhibit shows only one direction of a typical bi-directional urban street analysis. 2. Applications The procedures in this chapter are designed to support the following planning and prelimi- nary engineering analyses: • Development of an urban street corridor improvement plan • Feasibility studies of – Road diets, – Complete streets, – Capacity improvements, – Signal timing improvements, – Transit priority timing, and • Land development traffic impact studies.

82 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual 3. Analysis Methods Overview Urban street performance can be directly measured in the field or it can be estimated in great detail using microsimulation. However, the resource requirements of both of these methods render them generally impractical for most planning and preliminary engineering applications. The HCM provides a less resource-intensive approach to estimating urban street performance; however, it also is generally impractical to use the HCM with 100% field-measured inputs for many planning and preliminary engineering analyses. As shown by the unshaded boxes in Exhibit 44, this section presents two medium-level methods for evaluating urban street performance, as well as a high-level screening and scoping method that can be used to focus the analysis on only those locations and time periods requiring investigation. The HCM facility, segment, and intersection analysis methods (covered in HCM Chapters 16 to 23) provide a good basis for estimating urban street performance under many conditions. However, these methods are complex and specialized software is required to implement them. Consequently, a simplified HCM facility analysis method is presented in this section to reduce the number of computations and to enable programming of the method in a static spreadsheet, without requiring writing macros to implement it. Exhibit 43. Relationships between urban street facility, urban street segments, and intersections. Exhibit 44. Analysis options for urban streets. High Level Medium Level Low Level

K. Urban Streets 83 Because all of these methods still require a fair amount of data and computations, this chapter also provides a high-level service volume and volume-to-capacity ratio screening method for quickly identifying which portions of the street will require more detailed analysis (to properly account for the spillover effects of congestion), and to quickly compare improvement alternatives according to the capacity they provide. 4. Scoping and Screening Generalized Service Volume Tables Whether or not a more detailed urban street facility analysis is needed can be determined by comparing the counted or forecasted daily or peak hour traffic volumes for the urban street segments between each controlled intersection to the values given the service volume tables pre- sented later in this subsection. If all of the segment volumes fall in the LOS E range or better, there will not be congestion spillover requiring a full facility analysis to better quantify the facility’s per- formance. One can then use the HCM intersection and segment analysis procedures with defaults for some of the inputs to evaluate the performance of each segment and intersection. The service volumes can also be used to quickly determine the geographic and temporal extent of the urban street facility that will require analysis. If the counted or forecasted volumes for a segment fall within the agency’s target LOS standard, then the segment and its associated down- stream intersection can be excluded from a more detailed analysis. HCM Daily Service Volume Table HCM Exhibit 16-16 (adapted below as Exhibit 45) provides approximate maximum two-way AADT volumes that can be accommodated by an urban street at a given LOS for two posted speed limits under very specific assumptions of signal timing, signal spacing, access point (unsignal- ized driveway) spacing, and access point volumes. The service volumes are highly sensitive to the selected assumptions. Alternative Daily and Peak Hour Service Volume Table Exhibit 46 provides maximum service volumes (both two-way AADT and peak hour peak direc- tion) that can be accommodated by an urban street under differing assumptions regarding signal timing, signal spacing, and facility length. The values in this table are expressed on a per-lane basis. For example, a six-lane urban street (three lanes each direction) can carry between 52,200 (8,700 × 6 lanes) and 81,600 AADT (13,600 × 6 lanes) at LOS E, depending on the posted speed limit, signal spacing, and traffic signal cycle length. The LOS E service volume is generally also the through capacity at the critical signal on the facility; however, in some situations (as noted in the chart), this volume may be lower than the capacity. Intersection Volume-to-Capacity Ratio Checks The problem with screening at the facility level is that it is possible for the service volume check to show LOS E for the facility when the capacity of one or more intersections along the street has already been exceeded. This condition is especially likely when the signals are widely spaced (i.e., more than one-quarter mile apart). Thus, an intersection volume-to-capacity (v/c) ratio check is recommended to supplement the overall facility service volume screening. The intersection v/c ratios are computed and screened using the methods described in the intersection sections of this Guide (Section L for signalized intersections, Section M for stop- controlled intersections, and Section N for roundabouts). The v/c ratios may be used for study

84 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual scoping purposes to identify those intersections requiring more detailed analysis. They may also be used to quickly screen capacity-related improvement alternatives. Any segment that exceeds the capacity of the downstream intersection will have queuing that may impact upstream segments and reduce downstream demands. In such a situation, a full urban street facility analysis using a method capable of accurately identifying queue spillbacks is required to ascertain the performance of the urban street. The facility analysis can be performed using the HCM method with defaults, described later in this section. In cases of severe conges- tion, a microsimulation analysis may be required to accurately assess queue spillback effects. The analyst may also use the intersection demand-to-capacity (d/c) ratios for each segment to quickly screen various capacity improvement options. Exhibit 47 shows the planning capaci- ties per through lane that may be used to screen for signalized intersection capacity problems. The options can then be quickly ranked according to their forecasted d/c ratios for the critical segments of the urban street. K- D- Two-Lane Streets Four-Lane Streets Six-Lane Streets Factor Factor LOS C LOS D LOS E LOS C LOS D LOS E LOS C LOS D LOS E Posted Speed Limit = 30 mph 0.09 0.55 1,700 11,800 17,800 2,200 24,700 35,800 2,600 38,700 54,000 0.09 0.60 1,600 10,800 16,400 2,000 22,700 32,800 2,400 35,600 49,500 0.10 0.55 1,600 10,700 16,100 2,000 22,300 32,200 2,400 34,900 48,600 0.10 0.60 1,400 9,800 14,700 1,800 20,400 29,500 2,200 32,000 44,500 0.11 0.55 1,400 9,700 14,600 1,800 20,300 29,300 2,100 31,700 44,100 0.11 0.60 1,300 8,900 13,400 1,700 18,600 26,900 2,000 29,100 40,500 Posted Speed Limit = 45 mph 0.09 0.55 7,700 15,900 18,300 16,500 33,600 36,800 25,400 51,700 55,300 0.09 0.60 7,100 14,500 16,800 15,100 30,800 33,700 23,400 47,400 50,700 0.10 0.55 7,000 14,300 16,500 14,900 30,200 33,100 23,000 46,500 49,700 0.10 0.60 6,400 13,100 15,100 13,600 27,700 30,300 21,000 42,700 45,600 0.11 0.55 6,300 13,000 15,000 13,500 27,500 30,100 20,900 42,300 45,200 0.11 0.60 5,800 11,900 13,800 12,400 25,200 27,600 19,100 38,800 41,500 Source: Adapted from HCM (2016), Exhibit 16-16. Notes: Entries are maximum vehicle volumes per lane that can be accommodated at stated LOS. AADT = annual average daily traffic. AADT per lane is two-way AADT divided by the sum of lanes in both directions. This table is built on the following assumptions: No roundabouts or all-way STOP-controlled intersections along the facility. No on-street parking and no restrictive median. Coordinated, semi-actuated traffic signals, with some progression provided in the analysis direction (i.e., arrival type 4). 120-second traffic signal cycle lengths, protected left-turn phases provided for the major street, and the weighted average g/C ratio (i.e., ratio of effective green time for the through movement in the analysis direction to the cycle length) = 0.45. Exclusive left-turn lanes with adequate queue storage are provided at traffic signals and no exclusive right-turn lanes are provided. 2-mile facility length. At each traffic signal, 10% of traffic on the major street turns left and 10% turns right. Peak hour factor = 0.92 and the base saturation flow rate = 1,900 pc/h/ln. Additional assumptions for 30-mph facilities: signal spacing = 1,050 ft and 20 access points/mi. Additional assumptions for 45-mph facilities: signal spacing = 1,500 ft and 10 access points/mi. Exhibit 45. HCM daily service volume and capacity table for urban streets.

Exhibit 46. Daily and peak hour service volume and capacity table for four-lane urban streets. Speed Signal Cycle Peak Hour Peak Direction (veh/h/ln) AADT (2-way veh/day/ln) Limit (mph) Spacing (ft) Length (s) LOS C LOS D LOS E (capacity) LOS C LOS D LOS E (capacity) 25 660 90 630 840 940 5,800 7,800 8,700 25 1,320 120 1,000 1,100 1,100 9,300 10,200 10,200 35 1,320 120 820 1,040 1,100 7,600 9,600 10,200 35 2,640 180 1,300 1,360 1,460 12,000 12,600 13,500 45 1,320 180 630 1,180 1,300* 5,800 10,900 12,000* 45 2,640 180 1,220 1,320 1,400* 11,300 12,200 13,000* 55 2,640 180 1,240 1,320 1,380* 11,500 12,200 12,800* 55 5,280 180 1,340 1,430 1,470 12,400 13,200 13,600 55 10,560 180 1,470 1,470 1,470 13,600 13,600 13,600 Notes: *The LOS F speed threshold is reached before the through movement volume-to-capacity (v/c) ratio reaches 1.00. In all other cases, the v/c ratio limit of 1.00 for LOS F controls. Entries are maximum vehicle volumes per lane that can be accommodated at stated LOS. AADT = annual average daily traffic. AADT per lane is two-way AADT divided by the sum of lanes in both directions. This table is built on the following assumptions: Four-lane facility (two lanes in each direction). No roundabouts or all-way STOP-controlled intersections along the facility. No on-street parking and no restrictive median. Coordinated, semi-actuated traffic signals, with some progression provided in the analysis direction (i.e., arrival type 4). Protected left-turn phases provided for the major street, and the weighted average g/C ratio (i.e., ratio of effective green time for the through movement in the analysis direction to the cycle length) = 0.45. Exclusive left-turn lanes with adequate queue storage are provided at traffic signals and no exclusive right-turn lanes are provided. At each traffic signal, 10% of traffic on the major street turns left and 10% turns right. Peak hour factor = 1.00 and base saturation flow rate = 1,900 pc/h/ln. The facility is exactly two segments long with exactly three signals, so a facility with 1,320 feet (0.25 mile) between signals is 2,640 feet long. Two access points between each traffic signal, regardless of signal spacing. Each access point has two lanes in and two lanes out, with a peak hour volume of 180 veh/h turning into each driveway and 180 veh/h turning out of each driveway. K-factor (ratio of weekday peak hour two-way traffic to AADT) = 0.09 and D-factor (proportion of peak hour traffic in the peak direction) = 0.60. For other K- and D- values, multiply AADTs by the assumed factor values (i.e., 0.09 and 0.60) and divide by the desired values. Saturation Flow Rate (veh/h/ln) Through Movement g/C 0.40 0.45 0.50 1,500 600 675 750 1,600 640 720 800 1,700 680 765 850 1,800 720 810 900 1,900 760 855 950 Notes: Entries are through vehicles per hour per through lane. If exclusive turn lanes are present on the signal approach, then the total approach volumes used to screen for capacity problems should be reduced by the number of turning vehicles. A default value of 20% turns (10% lefts, 10% rights) may be used if both exclusive left- and right-turn lanes are present. Saturation flow rates, in vehicles per hour of green per lane, are effective rates after adjustments for heavy vehicles, turns, peak hour factor, and other factors affecting saturation flow. g/C = ratio of effective green time to traffic signal cycle length. Exhibit 47. Signal approach through movement capacities per lane.

86 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual Sensitivity of Predicted Urban Street Speeds Analysts should be aware of the following sensitivities of the HCM urban street estimation method: • The HCM-predicted average speeds under low-flow conditions may be higher or lower than the posted speed limit, depending on the posted speed limit and the signal spacing. • For through movement v/c ratios below 1.00, average speeds are much more sensitive to changes in v/c ratios than are freeways and highways. For freeways and multilane highways, the speed–flow curve is relatively flat until the v/c ratio at the bottleneck exceeds 1.00. For urban streets, the speed–flow curve drops comparatively rapidly with increasing v/c ratios, even when the v/c ratio is significantly below 1.00. • As demand increases on an urban street (but is still below a v/c of 1.00), there comes a point in the HCM method where the additional through traffic on the urban street at the unsignalized driveways (access points) can be significantly delayed by the driveways, thereby significantly reducing the predicted speed. • The HCM-estimated speed ceases to be sensitive to increases in demand once the v/c ratios on the upstream signal approaches feeding the downstream link reach 1.00. Further increases in demand are stored on the upstream signal approaches. The HCM speed estimation method for urban streets does not currently add in the delay to vehicles stored on the upstream signal approaches. For this reason, the HCM arterial method cannot be currently relied upon for speed prediction when the demands on the upstream signal approaches exceed a v/c of 1.00. 5. Employing the HCM Method with Defaults The HCM facility analysis method is described in HCM Chapter 16 and draws from the seg- ment analysis method in HCM Chapter 18. Urban street reliability analysis is described in HCM Chapter 17. Exhibit 48 lists the data needed to evaluate the full range of performance measures for planning-level urban street analysis. Individual performance measures may require only a subset of these inputs. The estimation of free-flow speeds using the HCM Chapter 17 method requires information on the posted speed limit, median type, presence of a curb, the number of access points per mile, the number of through lanes, and signal spacing. Urban street capacity, which is determined by the through capacities of the controlled inter- sections, requires intersection control data, intersection demands, intersection lane geometry, and the analysis period length. Average speed, motorized vehicle LOS, and multimodal LOS require the intersection capacities and free-flow speed plus additional data on segment lengths, demands, and lanes. Queues are estimated based on the intersection control, demand, and geometric data. Reliability analysis requires all the data required to estimate average speed, plus additional information on demand variability, incident frequencies and duration, weather, and work zones. 6. Simplified HCM Segment Analysis Method This simplified urban street segment analysis method assumes that the segments between intersections have no access points between the intersection boundaries and that there are no turning movements at the intersection. All intersections are assumed to be signalized. The method does not consider the effects of a median. Exhibit 49 provides a flow diagram showing the analysis steps for the method.

K. Urban Streets 87 Exhibit 48. Required data for urban street analysis with the HCM. Performance Measures Input Data (units) FFS Cap Spd LOS MMLOS Que Rel Default Values Posted speed limit (mph) • • • • • Must be provided Median type • • • • • Must be provided Curb presence • • • • • Must be provided Access points per mile • • • • • HCM Exhibit 18-7 Number of through lanes • • • • • • Must be provided Segment length (mi) • • • • • Must be provided Directional demand (veh/h) • • • • • Must be provided Percentage trucks (%) • • • • • 3% Intersection control data • • • • • • See Section L, M, or N Intersection demands • • • • • • See Section L, M, or N Intersection geometry • • • • • • See Section L, M, or N Analysis period length (h) • • • • • • 0.25 h Seasonal demand variation • HCM Exhibits 17-5 through 17-7 Crash rate (crashes/yr) • Must be provided Incident frequency, duration • HCM Exhibits 17-9through 17-12 Local weather history • HCM Volume 4 Work zone probability • Optional Notes: See appropriate sections in text for definitions of the required input data. Data required for intersection analysis is not shown here. See Section L (signalized intersections), M (stop-controlled intersections), or N (roundabouts) as appropriate. FFS = free-flow speed (default = speed limit plus 5 mph), Cap = capacity (veh/h/ln), Spd = average speed (mph), LOS = auto level of service, MMLOS = multimodal LOS (pedestrian, bicycle, transit), Que = queue (vehicles), and Rel = travel time reliability (multiple measures). Exhibit 49. Simplified urban street segment analysis method steps.

88 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual Input Requirements The method requires data for four input parameters: 1. The through movement volume along the segment vm (veh/h), 2. The number of through lanes on the segment NTH, 3. The segment length L (ft), and 4. The posted speed limit Spl (mph). Default values are assumed for five other input parameters: • Through movement saturation flow rate s = 1,900 veh/h/ln, • Effective green ratio g/C = 0.45, • Traffic signal cycle length C = 120 s, • Progression quality along the segment = average, and • Analysis period duration T = 0.25 h. As a default, the cycle length is assumed to be 120 seconds and the g/C ratio is assumed to be 0.45. The latter value assumes that the green time is evenly divided between the north–south and east–west intersection approaches and that lost time accounts for ten percent of the cycle length. The analyst can and should override these defaults based on local knowledge (such as coordi- nation plans). The quality of progression is assumed to be average (random arrivals), but the analyst can also select good (if there is some degree of coordination between the two signalized intersections) or poor (if there is poor coordination between the intersections). Step 1: Calculate Running Time The running time tR is calculated as follows: 3,600 5,280 Equation 58t L S UserAdj R pl( )= × × + where tR = running time excluding intersection delays (s), Spl = posted speed limit (mph), UserAdj = user-selected adjustment (mph) to reflect the difference between the facility’s posted speed limit and the free-flow speed (default = 5 mph), and L = segment length (ft). The default value for UserAdj assumes that the facility’s free-flow speed between controlled intersections is 5 mph greater than the posted speed limit. The analyst may wish to choose an alternative assumption to better reflect local conditions. Step 2: Calculate the Capacity of the Downstream Intersection The capacity of the downstream intersection is calculated as follows: Equation 59c g C N sTH= × × where c = capacity of the downstream intersection (veh/h), g/C = effective green ratio for the through movement (default = 0.45) (unitless),

K. Urban Streets 89 NTH = number of through lanes, and s = saturation flow rate for the through movement (veh/h/ln). Step 3: Calculate the Volume-to-Capacity Ratio The volume-to-capacity ratio for the through movement X is calculated as follows: Equation 60X v c m = where X = volume-to-capacity ratio for the through movement (unitless), vm = through movement volume along the segment (veh/h), and c = capacity of the downstream intersection (veh/h). Step 4: Calculate the Control Delay The control delay d in seconds per vehicle is determined either from the signalized inter- section planning method (see Sections L5) or calculated as described herein. The uniform delay d1 is calculated using Equation 61. 0.5 1 1 min 1, Equation 611 2 d C g C X g C[ ] ( ) ( )( )= − − where d1 = uniform delay for through vehicles (s/veh), C = traffic signal cycle length (s), g/C = effective green ratio for the through movement (unitless), and X = volume-to-capacity ratio for the through movement (unitless). The incremental delay d2 is calculated as follows: 225 1 1 16 Equation 622 2d X X X cNTH ( ) ( )= − + − +   where d2 = incremental delay for through vehicles (s/veh), X = volume-to-capacity ratio for the through movement (unitless), c = capacity of the downstream intersection (veh/h), and NTH = number of through lanes. The average control delay d for through vehicles is calculated using Equation 63. Equation 631 2d d PF d= + where d = average control delay for through vehicles (s/veh), d1 = uniform delay for through vehicles (s/veh), PF = progression factor reflecting the quality of signal progression (unitless) from Exhibit 50, and d2 = incremental delay for through vehicles (s/veh).

90 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual Step 5: Calculate the Average Travel Speed and Determine Level of Service The average travel time on the segment TT is calculated using Equation 64. Equation 64T t dT R= + where TT = average though movement travel time (s), tR = running time (s), and d = average control delay for through vehicles (s/veh). The average travel speed on the segment ST,seg is calculated using Equation 65. 3,600 5,280 Equation 65,S L T T seg T = × × where ST,Seg = average travel speed for the through movement (mph), L = segment length (ft), and TT = average though movement travel time (s). A spreadsheet-based computational engine has been developed for use in computing each of the data elements. Worksheets for completing the calculations are provided in Exhibit 51. Once the average speed is estimated, the level of service is looked up in Exhibit 52. Extension to Oversaturated Conditions Cases in which demand exceeds capacity are common in urban street networks, particularly when considering future planning scenarios. This condition is considered to be sustained when demand exceeds capacity over an entire analysis period, not just for one or two signal cycles. The condition is illustrated in Exhibit 53, where the arrival volume v1 during the analysis period t1 exceeds the capacity c for the downstream intersection approach. During the second analysis period t2 the arrival volume v2 is sufficiently low such that the queue that formed during t1 clears before the end of t2. The area between the demand line and the capacity line represents the over- flow delay experienced by all vehicles arriving during these two analysis periods. Each of the two analysis periods shown in Exhibit 53 represents a number of signal cycles. In contrast, the delay resulting from the failure of an individual cycle (“the occasional over- flow queue at the end of the green interval”) is accounted for by the d2 term of the delay equation for signalized intersections and urban street segments. This condition is illustrated in Exhibit 54 where a queue exists for two cycles, but clears in the third cycle. The non-zero slope of the departure Progression Quality Progression Factor (PF) Good (some degree of coordination between the two signalized intersections) 0.70 Average (random arrivals) 1.00 Poor (poor coordination between the intersections) 1.25 Exhibit 50. Progression factor.

K. Urban Streets 91 Exhibit 51. Simplified urban street method worksheets. Simplified Urban Street Method, Input Data Worksheet Input Data Direcon 1 (EB/NB) Direcon 2 (WB/SB) Through movement volume vm (veh/h) Number of through lanes NTH Segment length L ( ) Posted speed limit Spl (mph) Through move saturaon flow rate s (veh/h/ln) (default = 1,900) Effecve green rao g/C (default = 0.45) Cycle length C (s) (default = 120) Progression quality (good, average, poor) (default = average) Analysis period T (h) (default = 0.25) Simplified Urban Street Method, Calcula on Worksheet Step 1. Running Time Direc on 1 (EB/NB) Direc on 2 (WB/SB) Running me (s): = , × , × ( ) Step 2. Capacity Direcon 1 (EB/NB) Direcon 2 (WB/SB) Capacity (veh/h): = / × × Step 3. Volume-to-Capacity Rao Direcon 1 (EB/NB) Direcon 2 (WB/SB) Volume-to-capacity ra o: = Step 4. Control Delay Direcon 1 (EB/NB) Direcon 2 (WB/SB) Uniform delay (s): = . ( / )[ ( , )( / )] Incremental delay (s): = 225 ( − 1) + ( − 1) + Progression factor PF: 0.70 (good), 1.00 (average), 1.25 (poor) Control delay (s): = + Step 5. Average Travel Speed Direc on 1 (EB/NB) Direc on 2 (WB/SB) Travel me (s): = + Travel speed (mph): , = , × , × Note: EB = eastbound, NB = northbound, WB = westbound, SB = southbound. Exhibit 52. Urban street LOS average speed thresholds. Base Free-Flow Speed (mph) LOS 55 50 45 40 35 30 25 A >44 >40 >36 >32 >28 >24 >20 B >37 >34 >30 >27 >23 >20 >17 C >28 >25 >23 >20 >18 >15 >13 D >22 >20 >18 >16 >14 >12 >10 E >17 >15 >14 >12 >11 >9 >8 F ≤17 ≤15 ≤14 ≤12 ≤11 ≤9 ≤8 or any v/c > 1.0 Source: HCM (2016), Exhibit 16-3. Notes: Entries are minimum average travel speeds (mph) for a given LOS. The base free-flow speed is estimated as described in HCM Chapter 18, page 18-28, or can be approximated by adding 5 mph (or other appropriate adjustment) to the posted speed limit. v/c = volume-to-capacity ratio for the through movement in the analysis direction at the boundary intersection.

92 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual line during the green interval is equal to the saturation flow rate. The slope of the capacity line is the product of the saturation flow rate and the green ratio. The condition shown in Exhibit 54 is not considered to be sustained oversaturation and is therefore not addressed by the method described in this section. Overview of the Method The urban street segment planning method for oversaturated conditions predicts the overflow delay that results when the demand volume on an urban street segment exceeds its capacity. The method also predicts the v/c ratio for the first analysis period. The method considers only the Exhibit 53. Overflow delay when demand exceeds capacity over the analysis period. Exhibit 54. Delay resulting when demand is less than capacity over the analysis period.

K. Urban Streets 93 through traffic on the segment. The method considers a queue that may exist at the beginning of the analysis period, the queue that exists at the end of the analysis period, and the time that it takes for this queue to clear during a second analysis period. The framework for determining the effect of oversaturation in the urban street segment is shown in Exhibit 55. Limitations of the Method The method does not consider mid-section movements or turning movements at the down- stream intersection. The method does not consider the operational impacts of the queue spillback that result from the oversaturated conditions. The method can be used to analyze oversaturated conditions that result from demand exceeding capacity during several analysis periods. However, during the final analysis period, the demand must be such that the queue clears during this period. Input Data Requirements The input data requirements for the method include the following nine parameters: • Arrival volumes v1 and v2 (veh/h) for the through movement at the downstream intersection during analysis period 1 (the period of oversaturation) and analysis period 2 (the period when the queue clears); • Analysis period duration T (h); • Segment length L (ft); • Initial queue Q0 (veh) existing at the beginning of analysis period 1 for the through movement at the downstream intersection; • Number of through lanes in the segment NTH; • Saturation flow rate s for the downstream signalized intersection (veh/h/ln); and • Cycle length C (s) and effective green ratio g/C at the downstream signalized intersection. Default values are assumed for four of these parameters: • T = 0.25 h, • s = 1,900 veh/h/ln, • C = 120 s, and • g/C = 0.45. Computational Steps The planning method for urban street segments during periods of oversaturation is a sim- plified version of the operational analysis method for urban street segments for oversaturated Exhibit 55. Oversaturated urban street segment planning method analysis framework.

94 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual conditions described in HCM Chapter 30. The method includes nine steps, shown in Exhibit 56 and described below. Step 1: Calculate Queue Storage Capacity The queue storage capacity Qcap is the number of vehicles that can be stored in the segment, assuming an average vehicle length of 25 ft. The queue storage capacity is calculated as follows: 25 Equation 66Q N L cap TH = where Qcap = queue storage capacity (veh), NTH = number of though lanes in the subject direction, and L = segment length (ft). Step 2: Calculate Available Queue Storage This step calculates the available queue storage Qa in the segment during analysis period 1 after accounting for any initial queue Q0 that is present at the beginning of the analysis period. The available queue storage is calculated using Equation 67. Equation 670Q Q Qa cap= − where Qa = available queue storage capacity (veh) during analysis period 1, Qcap = queue storage capacity (veh), and Q0 = initial queue (veh) at the beginning of analysis period 1. Exhibit 56. Urban street segment planning method, oversaturated conditions.

K. Urban Streets 95 The available queue storage Qa is compared to the estimated maximum queue (computed later) to identify queue overflow problems. Step 3: Calculate Through Movement Capacity Equation 68 is used to calculate the capacity of the through movement cTH at the downstream signalized intersection. Equation 68c N s g C TH TH=     where cTH = through movement capacity at the downstream signal (veh/h), s = saturation flow rate for the through movement (veh/h), g = effective green time for the through movement (s), and C = traffic signal cycle length (s). Step 4: Calculate Volume-to-Capacity Ratio The volume-to-capacity ratio X for the segment during analysis period 1 is calculated as follows: Equation 69 1 X v cTH = where X = volume-to-capacity ratio for the through movement (unitless), v1 = arrival volume (veh/h) during analysis period 1, and cTH = through movement capacity at the downstream signal (veh/h). Step 5: Calculate Rate of Queue Growth This step calculates the rate of queue growth rqg during analysis period 1. If the through move- ment arrival volume v1 is less than the capacity, no queue forms and this method is not needed. Equation 70 is used to calculate the rate of queue growth. 0.0 Equation 701r v cqg TH= − ≥ where rqg = rate of queue growth (veh/h) during analysis period 1, v1 = arrival volume (veh/h) during analysis period 1, and cTH = through movement capacity at the downstream signal (veh/h). Step 6: Calculate Queue Length The length of the queue Qmax at the end of analysis period 1 is determined as follows: Equation 711Q r tmax qg= where Qmax = queue length (veh) at the end of analysis period 1, rqg = rate of queue growth (veh/h) during analysis period 1, and t1 = duration of analysis period 1 (h).

96 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual Step 7: Calculate Queue Clearance Rate The rate of queue clearance rqc during analysis period 2 is calculated as follows: Equation 722r c vqc TH= − where rqc = rate of queue clearance (veh/h) during analysis period 2, cTH = through movement capacity at the downstream signal (veh/h), and v2 = arrival volume (veh/h) during analysis period 2. Step 8: Calculate Queue Clearance Time The time for the queue to clear depends on the length of the queue at the end of analysis period 1, the arrival volume during analysis period 2, and the capacity of the through movement for the downstream intersection. If the queue does not clear before the end of analysis period 2, the volumes during subsequent analysis periods must be considered and the queue clearance time calculation must be modified to account for this result. The queue clearance time tc is calculated using Equation 73. Equation 73 1 2 t r t r Q c v c qg qc max TH = = − where tc = queue clearance time (h), rqg = rate of queue growth (veh/h) during analysis period 1, t1 = duration of analysis period 1 (h), rqc = rate of queue clearance (veh/h) during analysis period 2, Qmax = queue length (veh) at the end of analysis period 1, cTH = through movement capacity at the downstream signal (veh/h), and v2 = arrival volume (veh/h) during analysis period 2. Step 9: Calculate Oversaturated Delay The final step calculates the delay resulting from oversaturation dsat. Exhibit 57 shows the queue accumulation polygon for oversaturated conditions in which a queue grows during analysis period 1 and clears during analysis period 2. The area of the polygon that is formed by these conditions is the delay resulting from the oversaturated conditions. The average delay per vehicle is calculated as follows: 0.5 0.5 Equation 74 0 1 1 1 2 d Q Q t t Q v t v t sat max c max c ( ) = − + + where dsat = delay resulting from oversaturation (s/veh), Qmax = queue length at the end of analysis period 1 (veh), Q0 = initial queue (veh) at the beginning of analysis period 1, t1 = duration of analysis period 1 (h), tc = queue clearance time (h), v1 = arrival volume (veh/h) during analysis period 1, and v2 = arrival volume (veh/h) during analysis period 2.

K. Urban Streets 97 Computational Tools A spreadsheet has been developed for use in calculating each of the data elements. A work- sheet for completing the calculations is provided as Exhibit 58. 7. Reliability Analysis HCM Chapter 17 describes a method for estimating urban street reliability that is sensitive to demand variations, weather, incidents, and work zones. The Florida DOT has also developed a method for estimating reliability for urban streets (Elefteriadou et al. 2013). Both methods are data- and computationally intensive, requiring custom software to implement. As such, neither method is readily adaptable to a planning and preliminary application that could be programmed in a simple, static spreadsheet. Analysts wishing to perform a reliability analysis of urban streets should consult these sources. 8. Multimodal LOS Bicycle, Pedestrian, and Transit LOS The HCM provides methods for evaluating bicycle, pedestrian, and transit LOS on urban streets, which are described in Section O4. Truck LOS The HCM does not provide a truck LOS method. However, the truck LOS estimation proce- dure described in Section P can be used to estimate truck LOS for urban streets. 9. Example Case Study 2 (Section U) provides an example application of the screening and simplified analysis methods described in this section. Exhibit 57. Queue accumulation polygon for oversaturated conditions.

98 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual 10. References Elefteriadou, L., Z. Li, and L. Jin. Modeling, Implementation, and Validation of Arterial Travel Time Reliability. Final Report, FDOT Contract BDK77 977-20, University of Florida, Gainesville, Nov. 30, 2013. Highway Capacity Manual: A Guide to Multimodal Mobility Analysis. 6th ed. Transportation Research Board, Washington, D.C., 2016. Exhibit 58. Oversaturated urban street segment planning method worksheet. Oversaturated Urban Street Segment Planning Method, Input Data Worksheet Input Data Arrival volume, me period 1 v1 (veh/h) Arrival volume, me period 2 v2 (veh/h) Analysis period duraon T (h) Segment length L () Inial queue Q0 (veh) Number of through lanes NTH Through movement saturaon flow rate s (veh/h/ln) Effecve green rao g/C Cycle length C (s) Oversaturated Urban Street Segment Planning Method, Calculaon Worksheet Step 1: Queue Storage Capacity (veh) = 25 Step 2: Available Queue Storage (veh) = − Step 3: Capacity of Through Movement (veh/h) = Step 4: Volume-to-Capacity Rao = Step 5: Rate of Queue Growth (veh/h) = − ≥ 0.0 Step 6: Length of Queue (veh) = Step 7: Rate of Queue Clearance (veh/h) = − Step 8: Time of Queue Clearance (h) = = − Step 9: Oversaturaon Delay (s) = 0.5( − ) + 0.5 +

99 L. Signalized Intersections 1. Overview A signalized intersection is an intersection or midblock crosswalk where some or all conflicting movements are controlled by a traffic signal. The pro- cedures presented here can also be adapted to the analysis of freeway ramp meters and traffic signals used to meter traffic flow into a roundabout. Signalized interchange ramp terminals are a special case of signalized intersections at the foot of freeway on- and off-ramps. They are addressed in HCM Chapter 23. The uneven nature of lane demands and the tight spacing between signals within a freeway interchange result in conditions than are not typical of an urban street. HCM Chapter 23 also presents methods for analyzing signalized alternative intersections, where one or more movements are rerouted to secondary intersections. To the extent that movements of interest to a planning analysis are not diverted (e.g., through movements on an arterial street), the planning-level procedures in this section can be used. The analysis of movements that are diverted requires a more detailed analysis, such as the methods described in HCM Chapter 23. 2. Applications The procedures in this section are designed to support the following planning and preliminary engineering analyses: • Feasibility studies of – Intersection improvements, and – Signal timing improvements, and • Land development traffic impact studies. 3. Analysis Methods Overview Intersection performance can be directly measured in the field or it can be estimated in great detail using microsimulation. The resource requirements of both of these methods render them generally impractical for most planning and preliminary engineering applications. HCM Chapter 19 provides a much less resource-intensive approach to estimating intersection performance; however, it is generally impractical to use the HCM methods with 100 percent field-measured inputs for many planning and preliminary engineering analyses.

100 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual Employing the HCM method with defaults identified in HCM Chapter 19 reduces the data requirements, but still requires specialized software to implement the complex computations. As indicated by the unshaded boxes in Exhibit 59, this section presents a medium-level method for evaluating signalized intersections, portions of which can be used to perform a high-level screening and scoping analysis to focus the planning and preliminary engineering analysis on only those intersections and time periods requiring investigation. This simplified volume-to- capacity (v/c) ratio and level of service (LOS) method can be easily programmed in a static spreadsheet without requiring knowledge of macros. Exhibit 60 lists the input data required for conducting a planning analysis for signalized inter- sections. The analyst is required to specify values for two of the parameters, the volume for each movement and the number of lanes (and the turn designation for each) on each approach. If only approach volumes are known, one of the methods described in Section D8 can be used to generate turning-movement volumes. Default values can be assumed for the other seven input parameters, or the analyst can specify the parameter values if they are known. High Level Medium Level Low Level Exhibit 59. Analysis options for signalized intersections. Performance Measure Input Data (units) Cap Del LOS MMLOS Que Default Value Number of turn lanes • • • • • Must be provided Other geometry • • • • • HCM Exhibit 19-11 Signal timing • • • • • HCM Exhibits 19-11 and 19-17 Peak hour factor (decimal) • • • • 0.90 (total entering volume<1,000 veh/h), 0.92 (otherwise) Percentage heavy vehicles (%) • • • • • 3% Parking activity • • • • • None Pedestrian activity • • • • • None Volumes by movement (veh/h) • • • • Must be provided Analysis period length (h) • • • 0.25 h Notes: See the text for definitions of the required input data. Cap = capacity (veh/h/ln), Del = delay (s), LOS = auto level of service, MMLOS = multimodal LOS (bicycle, pedestrian, transit), Que = queue (veh). “Other geometry” data include lane widths, bus stops, and pedestrian crossings. “Signal timing” data include cycle length, effective green time, lost time, progression, and phasing. Exhibit 60. Required data for signalized intersection analysis.

L. Signalized Intersections 101 Parking activity at the intersection is characterized as either allowed or prohibited (default). Pedestrian activity is characterized as follows: • None (default) • Low – 50 pedestrians per hour • Medium – 200 pedestrians per hour • High – 400 pedestrians per hour • Very High – 800 pedestrians per hour 4. Simplified Method, Part 1: Volume-to-Capacity Ratio Calculation Whether an intersection requires more detailed analysis can be determined quickly by estimating its volume-to-capacity (v/c) ratio. These ratios can also be used to quickly compare different capacity improvement alternatives and select the more cost-effective alternatives for further analysis. A critical movement analysis is used to predict the critical v/c ratio of the intersection and make an assessment of the sufficiency of the intersection to accommodate the forecasted peak hour traffic volumes. Exhibit 61 shows that five steps are used to assess the sufficiency of intersection capacity based on the v/c ratio. Step 1: Determine the Left-Turn Phasing The left-turn phasing can be permitted, protected, protected plus permitted, or split. • Permitted phasing allows left-turn movements to proceed when gaps in traffic permit. • Protected phasing provides a left-turn arrow that allows left turns to proceed without conflicts. • Protected plus permitted phasing provides both a protected and a permitted phase. Exhibit 61. Intersection capacity sufficiency analysis steps.

102 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual • Split phasing means that all movements on an approach, including the left turns, proceed at the same time with no opposing movements. The analyst can select one of these four phasing types if the phasing is known. If it is not known, the computational procedure will determine the left-turn phasing. The method will select protected left-turn phasing if any of the following three conditions are met; otherwise, permitted left-turn phasing will be selected: • Left-turn volume exceeds 240 veh/h; • The product of the left-turn volume and the opposing through volume exceeds a given threshold (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 both opposing approaches have exclusive left-turn lanes, and one of those meets the above thresholds for left-turn protection, then both approaches will have left-turn protection. Step 2: Identify Lane Groups A lane group is a lane or set of lanes designated for separate analysis. All traffic movements for a given approach (i.e., left, through, and right) must be assigned to a lane group. A lane group can consist of one or more lanes. There are two guidelines for assigning traffic movements to lane groups: 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. When a right-turn movement is shared with a through movement, it is considered to be a part of the through movement lane group. When a right-turn movement is shared with a left-turn move- ment (such as at a T-intersection), it is considered to be a part of the left-turn movement lane group. Lane groups should first be checked to determine if a de facto turn lane exists. A de facto turn lane occurs on approaches with multilane lane groups where either a left- or right-turn movement is shared with a through movement, but that lane is only used by turning vehicles. This occurs in situations where the turning movements are high, there are significant impedances for the turning movements, or both. In these situations, de facto turn lanes should be analyzed as exclusive turn lanes and all through movements should be assumed to occur from the through-only lane(s). In cases where there are multiple turn lanes and one lane is shared with a through move- ment, that combination of lanes should be treated as a single-lane group and all the lanes should be associated with the through lane group. For approaches at a T-intersection where there are only left- and right-turn movements and multiple lanes, and one of the lanes is shared, the analyst has the option of coding all lanes as either the right-turn lane group or the left-turn lane group. Step 3: Convert Turning Movements to Through 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.

L. Signalized Intersections 103 Step 3a: Heavy Vehicle Adjustment The adjustment for heavy vehicles EHVadj is calculated using Equation 75. E P EHVadj HV HV( )= + −1 1 Equation 75 where EHVadj = heavy vehicle adjustment factor (unitless), PHV = proportion of heavy vehicles in the movement (decimal), and EHV = passenger car equivalent for heavy vehicles in the movement (default = 2.0). 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 Equation 76 where EPHF = flow variation adjustment factor (unitless), and PHF = peak hour factor (unitless, ranges from 0.25 to 1.00, default = 0.92). Step 3c: Turn Impedance Adjustment The turn impedance adjustment factors ELT and ERT adjust for impedances experienced by left- and right-turning vehicles, respectively. Left-turning vehicles served by permitted left-turn phasing must find acceptable gaps in the opposing through traffic stream to complete their turns. Left-turning vehicles served by protected left-turn phasing also flow more slowly than through vehicles. The methods used to determine ELT and ERT depend on the signal phasing used for the turns. Through vehicles do not experience the impedances that turning vehicles do, so the flows for these movements are not adjusted. Permitted Left-Turn Phasing. The values for ELT for permitted left turns are given in Exhibit 62. Protected and Split Left-Turn Phasing. If the left turn is protected, or uses split phasing, then ELT = 1.05 regardless of volume. Protected-Permitted Left-Turn Phasing. Equation 77 is used to calculate ELT when protected- permitted phasing is used. The signal timing must be known or estimated by the analyst. Note that the effective green time for the first portion of the protected-permitted phase includes the yellow interval between the two portions. Opposing Through and Right-Turn Volumes (veh/h) ELT <200 1.10 200–599 2.00 600–799 3.00 800–999 4.00 ≥1,000 5.00 Exhibit 62. Left-turn impedance adjustment factor ELT values for permitted left turns.

104 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual E E g E g g g LT LT prot LT prot LT perm LT perm LT prot LT perm ( )( ) = + Equation 77 , , , , , , where ELT = left-turn impedance adjustment factor (unitless), ELT,prot = left-turn impedance adjustment factor for the protected portion of the left-turn phase (unitless) = 1.05, ELT,perm = left-turn impedance adjustment factor for the permitted portion of the left-turn phase (unitless), gLT,prot = effective green time for the protected portion of the left-turn phase (s), and gLT,perm = effective green time for the permitted portion of the left-turn phase (s). Permitted Right-Turn Phasing. Right-turning vehicles are sometimes impeded by pedes- trians. The values for ERT for permitted right turns are given in Exhibit 63. Protected and Split Right-Turn Phasing. When protected right turns are provided, the ERT value for “none or low” pedestrian activity in Exhibit 63 (1.2) should be used. Step 3d: Parking Adjustment Factor The parking adjustment factor Ep is a function of the presence of on-street parking and applies to through and right-turn volumes. Values for Ep are given in Exhibit 64. Step 3e: Lane Utilization Factor The lane utilization factor ELU recognizes the volume imbalance between lanes when there are two or more lanes on an approach. Values for this factor are given in Exhibit 65. Step 3f: Adjustment Factor for Other Effects The analyst may wish to incorporate the saturation flow rate effects of work zones (if any), mid-segment lane blockage, and sustained spillback from downstream segment in a comprehen- sive volume adjustment factor for other effects Eother. The analyst should consult the HCM for Pedestrian Activity ERT None or low 1.20 Medium 1.30 High 1.50 Very high 2.10 Exhibit 63. Right-turn impedance adjustment factor ERT values for permitted right turns. Parking Activity Number of Lanes in Lane Group Ep No parking lane All 1.00 Adjacent parking 1 1.20 2 1.10 3 1.05 Exhibit 64. Parking adjustment factor Ep.

L. Signalized Intersections 105 guidance on the magnitude of these other effects on saturation flow rates. The default value for Eother is 1.00 (i.e., no other effects). Step 3g: Through Passenger Car Equivalent Flow Rate The through passenger car equivalent flow rate vadj is calculated using Equation 78, applying the adjustment factors determined in Steps 3a through 3f. v VE E E E E E Eadj HVadj PHF LT RT p LU= Equation 78other where vadj = through passenger car equivalent flow rate (through passenger cars per hour, tpc/h), V = turning-movement volume (veh/h), EHVadj = heavy vehicle adjustment factor (unitless), EPHF = flow variation adjustment factor (unitless), ELT = left-turn impedance adjustment factor (unitless), ERT = right-turn impedance adjustment factor (unitless), Ep = parking adjustment factor (unitless), ELU = lane utilization adjustment factor (unitless), and Eother = adjustment factor to account for other conditions determined by the analyst (unitless). Step 3h: Equivalent Per-Lane Flow Rate Finally, the equivalent per-lane flow rate vi for a given lane group i is calculated using Equation 79. v v N i adj i i = Equation 79 , where vi = equivalent per-lane flow rate for lane group i (tpc/h/ln), vadj,i = through passenger car equivalent flow rate for lane group i (tpc/h), and Ni = number of lanes within lane group i, accounting for de facto lanes. Step 4: Calculate Critical Lane Group Volumes Critical lane groups represent the combination of conflicting lane groups from opposing approaches that have the highest total demand. These critical lanes groups thus dictate the amount of green time required during each phase as well as the total cycle length required for the intersection. The movements and phasing for the north–south and east–west approaches are Lane Group Movements No. of Lanes in Lane Group ELU Through or shared 1 2 ≥3 1.00 1.05 1.10 Exclusive LT 1 ≥2 1.00 1.03 Exclusive RT 1 ≥2 1.00 1.13 Exhibit 65. Lane utilization factor ELU.

106 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual assessed independently. The combination of movements that make up the critical movements are different for protected and permitted left-turn phasing, and for split phasing. Step 4a: Identify Critical Movements Protected Left-Turn Phasing. When opposing approaches use protected left-turn phasing, the critical lane group volumes will be the maximum of the two sums of the left-turn lane vol- ume and the opposing through (or shared through) lane volume, or right-turn lane volume if that is greater. For the east–west approaches, the critical lane group volume Vc,EW is calculated using Equation 80. v v v v v v v c EW EBLT WBTH WBRT WBLT EBTH EBRT ( ) ( )= + +  max max , max , Equation 80, where Vc,EW = critical east–west lane group volume (tpc/h/ln), vEBLT = equivalent flow rate for the eastbound left-turn lane group (tpc/h/ln), vEBTH = equivalent flow rate for the eastbound through lane group (tpc/h/ln), vEBRT = equivalent flow rate for the eastbound right-turn lane group (tpc/h/ln), vWBLT = equivalent flow rate for the westbound left-turn lane group (tpc/h/ln), vWBTH = equivalent flow rate for the westbound through lane group (tpc/h/ln), and vWBRT = equivalent flow rate for the westbound right-turn lane group (tpc/h/ln). Similarly for the north–south approaches, the critical volume Vc,NS is calculated using Equation 81. v v v v v v v c NS NBLT SBTH SBRH SBLT NBTH NBRH max max , max , Equation 81, ( ) ( )= + +  where Vc,NS = critical north–south lane group volume (tpc/h/ln), vNBLT = equivalent flow rate for the northbound left-turn lane group (tpc/h/ln), vNBTH = equivalent flow rate for the northbound through lane group (tpc/h/ln), vNBRT = equivalent flow rate for the northbound right-turn lane group (tpc/h/ln), vSBLT = equivalent flow rate for the southbound left-turn lane group (tpc/h/ln), vSBTH = equivalent flow rate for the southbound through lane group (tpc/h/ln), and vSBRT = equivalent flow rate for the southbound right-turn lane group (tpc/h/ln). Permitted Left-Turn Phasing. When opposing approaches use permitted phasing, the critical lane group volume will be the highest lane group volume of all lane groups for a pair of approaches. For the 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. v v v v v v vc EW EBLT EBTH EBRT WBLT WBTH WBRTmax , , , , , Equation 82, ( )= v v v v v v vc NS NBLT NBTH NBRT SBLT SBTH SBRTmax , , , , , Equation 83, ( )= where all variables are as defined previously. Split Phasing. When opposing approaches use split phasing (where only one approach is served during the phase), the critical lane group volume will be the highest lane group volume

L. Signalized Intersections 107 of all lane groups for that approach. For the east–west approaches, the critical volume Vc,EW will be the sum of: v v v v v v vc EW EBLT EBTH EBRT WBLT WBTH WBRTmax , , max , , Equation 84, ( ) ( )= + where all variables are as defined previously. Similarly, for the north–south approaches with split phasing, the critical volume Vc,NS will be the sum of: v v v v v v vc NS NBLT NBTH NBRT SBLT SBTH SBRTmax , , max , , Equation 85, ( ) ( )= + where all variables are as defined previously. Protected-Permitted Left-Turn Phasing. The signal timing must be known or estimated by the analyst, which would have been done as part of Step 3c. To find the critical lane group volumes, the equivalent through-car volume in the left lane during the protected portion of the phase is found using Equation 86 by splitting the total demand in proportion to the length of the protected portion of the phase to the overall protected-permitted phase. V V g g g LT prot LT LT prot LT prot LT perm = +     Equation 86, , , , where VLT,prot = left-turn demand during the protected portion of the phase (tpc/h/ln), VLT = overall left-turn demand during the left-turn phase (tpc/h/ln), gLT,prot = effective green time for the protected portion of the left-turn phase (s), and gLT,perm = effective green time for the permitted portion of the left-turn phase (s). The critical lane volumes are then found using only the protected portion of the compound phase. The critical lane group volume is the highest total of a through lane volume and its oppos- ing protected left-turn volume. The remainder of the methodology does not change. In the delay module (optional Step 7), the overall left-turn demand VLT is used to find delay. Step 4b: Calculate the Sum of the Critical Lane Volumes The sum of the critical lane volumes Vc is calculated using Equation 87. V V Vc c EW c NS= + Equation 87, , where Vc = critical intersection volume (tpc/h/ln), Vc,EW = critical east–west volume (tpc/h/ln), and Vc,NS = critical north–south volume (tpc/h/ln). Step 5: Determine Intersection Sufficiency Step 5a: Calculate the Critical Volume-to-Capacity Ratio The critical volume-to-capacity ratio Xc is calculated using Equation 88. X V c c c i = Equation 88

108 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual where Xc = critical volume-to-capacity ratio (unitless), Vc = critical intersection volume (tpc/h/ln), and ci = intersection capacity (tpc/h/ln). Intersection capacity is the maximum per-lane through movement flow rate that can be accom- modated by the intersection, accounting for lost time. A value of 1,650 tpc/h/ln can be used as a default if local data are not known. This value reflects a saturation flow rate of 1,900 tpc/h/ln, a lost time of 4 seconds per critical phase, and a cycle length of 30 seconds per critical phase. A value of 1,500 tpc/h/ln may be used for signalized intersection capacity in smaller urban areas (under 250,000 population). Higher values may be appropriate for suburban or rural signals with high- speed approaches (≥45 mph). Step 5b: Assess the Intersection Sufficiency The final step of the v/c analysis is to assess the sufficiency of the intersection to accommodate a given demand level. Exhibit 66 provides the assessment of intersection sufficiency (under, near, or over) based on the critical volume-to-capacity ratio. 5. Simplified Method, Part 2: Delay, LOS, and Queue Calculation Part 2 of the method includes four steps and produces estimates of delay, LOS, and queue. It applies the results from Part 1. The steps are shown in Exhibit 67 and are described herein. Step 6: Calculate Capacity Step 6a: Calculate Cycle Length The traffic signal cycle length C is assumed to be 30 seconds per critical phase. The analyst can use another value based on local practice or conditions. C n= 30 Equation 89 where C = traffic signal cycle length (s), and n = number of critical phases. Xc Description Capacity Assessment <0.85 All demand is able to be accommodated; delays are low to moderate. Under 0.85–0.98 Demand for critical lane groups near capacity and some movements require more than one cycle to clear the intersection; all demand is able to be processed at the end of the analysis period; delays are moderate to high. Near >0.98 Demand for critical movements is just able to be accommodated within a cycle but more oftentimes requires multiple cycles to clear the intersection; delays are high and queues are long. Over Exhibit 66. Intersection sufficiency.

L. Signalized Intersections 109 Step 6b: Calculate the Total Effective Green Time The total effective green time gTOT available during the cycle is calculated using Equation 90. Equation 90g C LTOT = − where gtot = total effective green time (s), C = traffic signal cycle length (s), and L = lost time per cycle (s) (default = 4 seconds per critical phase). The total effective green time is then allocated to each critical phase in proportion to the criti- cal lane group volume for that movement using Equation 91: Equation 91g g V V i TOT ci c =   where gi = effective green time for phase i (s), gtot = total effective green time (s), Vci = critical lane group volume for phase i (tpc/h/ln), and Vc = critical intersection volume (tpc/h/ln). For the non-critical phase (and the movements served by these phases), the effective green time is set equal to the green time for the phase on the opposing approach that serves the same directional movement. The green time for each phase should be reviewed against policy Exhibit 67. Signalized intersection planning method, part 2.

110 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual requirements and other considerations such as the minimum green time and the time required for pedestrians to cross the approach. All green time and cycle length calculations should be adjusted to meet minimum requirements for all users. Step 6c: Calculate Capacity and Volume-to-Capacity Ratio The capacity ci and volume-to-capacity ratio Xi for each lane group i are calculated using Equation 92 and Equation 93. c BaseSat g C i i =     Equation 92 x v c i i i = Equation 93 where ci = capacity of lane group i (tpc/h/ln), vi = volume for lane group i (tpc/h/ln), BaseSat = 1,900 for large urban areas (over 250,000 population) and 1,750 otherwise (pc/h/ln), gi = effective green time for lane group i (s), and C = traffic signal cycle length (s). For the intersection as a whole, the critical degree of saturation Xc is calculated using Equation 94 and Equation 95. Equation 941X v c c cii n SUM ∑ = = 1,900 Equation 95SUM 1c g C cii n∑ =    = where Xc = critical degree of saturation (unitless), vci = volume for critical phase i (tpc/h/ln), cSUM = intersection capacity (tpc/h/ln), gci = effective green time for critical phase i (s), and C = traffic signal cycle length (s). Step 7: Estimate Delay The control delay for each lane group di is calculated using Equation 96. d d PF d di unsig= + + Equation 961 2 where di = control delay for lane group i (s/veh), d1 = uniform delay (s/veh), PF = progression adjustment factor (unitless), d2 = incremental delay (s/veh), and dunsig = analyst-provided estimate of unsignalized movement delay, if any (s/veh).

L. Signalized Intersections 111 The unsignalized movement delay dunsig is the average delay (if any) for turns at the intersection that are not controlled by a signal head. This delay is usually zero but may be non-zero for situations such as a stop-controlled channelized right-turn lane. It may also be non-zero for alternative inter- section concepts, such as Michigan U-turns and others. The uniform delay d1 is calculated using Equation 97. d C g C X g C[ ] ( ) ( )( )= − − 0.5 1 1 min 1, Equation 971 2 where d1 = lane group uniform delay (s/veh), C = traffic signal cycle length (s), g/C = lane group effective green ratio (unitless), and X = lane group volume-to-capacity ratio (unitless). The progression factor PF is given in Exhibit 68 and is selected based on the quality of pro- gression from an upstream signalized intersection. Possible values for the progression factor are 0.70 if the quality of progression is good and 1.25 if the quality is poor. The default value is 1.00 if progression is average, indicating that vehicles arrive in a random manner. The incremental delay d2 is calculated as follows: d X X X c ( ) ( )= − + − +  225 1 1 16 Equation 982 2 where d2 = lane group incremental delay (s/veh), X = lane group volume-to-capacity ratio (unitless), and c = lane group capacity (tpc/h/ln). Step 8: Determine LOS The LOS for each lane group or for the intersection as a whole is given in Exhibit 69 on the basis of average control delay. Note that if the volume-to-capacity ratio exceeds 1.0, then the LOS will be F regardless of the control delay. Step 9: Estimate Queues The deterministic average queue for each lane group (i.e., the average queue at the end of red) is determined by dividing the average uniform delay for that lane group by the capacity for that lane group. Progression Quality Progression Factor PF Good (some degree of coordination between the two signalized intersections) 0.70 Average (random arrivals) 1.00 Poor (poor coordination between the intersections) 1.25 Exhibit 68. Progression adjustment factor. Control Delay (s/veh) LOS ≤10 A >10–20 B >20–35 C >35–55 D >55–80 E >80 or X > 1.00 F Source: Adapted from HCM (2016), Exhibit 19-8. Note: X = volume-to-capacity ratio. Exhibit 69. Level of service for signalized intersections.

112 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual Q d c = × 3,600 Equation 99 1 where Q = deterministic average queue for the lane group (tpc/ln), d1 = uniform delay for the lane group (s), and c = per-lane capacity of the lane group (tpc/h/ln). The deterministic average queue for the lane group does not take into account random bunching of traffic arrivals within the analysis period. The deterministic average queue may be multiplied by 2.0 (approximately the ratio of the 95th percentile to the mean for a Poisson process) to obtain an approximation of the 95th percentile longest queue likely to be observed during a traffic signal cycle. Equation 99 only applies when the lane group operates under capacity, and, on average, the queue is able to fully dissipate each cycle. When a lane group operates over capacity, the differ- ence between the lane group demand and the lane group capacity, divided by the number of lanes in the lane group, provides the number of vehicles per lane not served (i.e., in queue) at the end of the analysis hour. 6. Worksheets The worksheets provided as Exhibit 70 through Exhibit 73 illustrate how the computations might be laid out in a spreadsheet. 7. Reliability Analysis The HCM 2016 does not provide a method for estimating the variability of delay at a signal- controlled intersection. The analyst might perform a sensitivity analysis by repeating the plan- ning computations using the 25th percentile and 75th percentile highest demands of the year and the 25th percentile and 75th percentile highest capacities of the year (taking into account incidents) and report the results in a table such as shown in Exhibit 74. Signalized Intersection Planning Method (Part 1), Inputs NB SB EB WB LT TH RT LT TH RT LT TH RT LT TH RT Volume Lanes PHF % HV Parking activity Ped activity LT phasing Notes: NB = northbound, SB = southbound, EB = eastbound, WB = westbound, LT = left turn, TH = through, RT = right turn, PHF = peak hour factor, and HV = heavy vehicles. Exhibit 70. Signalized intersection planning method (Part 1), input worksheet.

L. Signalized Intersections 113 Signalized Intersection Planning Method (Part 1), Calculations NB SB EB WB LT TH RT LT TH RT LT TH RT LT TH RT Step 1: Determine the Left-Turn Phasing Check #1 Check #2 Check #3 LT phasing Step 2: Assign Volumes to Lane Groups Step 3: Convert Turning Movements to Passenger Car Equivalents EHVadj EPHF ELT ERT EP ELU vadj vi Step 4: Calculate Critical Lane Groups vcEW vcNS vc Step 5: Determine Intersection Sufficiency vc/ci Intersection sufficiency Notes: NB = northbound, SB = southbound, EB = eastbound, WB = westbound, LT = left turn, TH = through, RT = right turn. Exhibit 71. Signalized intersection planning method (Part 1), calculations worksheet. Signalized Intersection Planning Method (Part 2), Calculations NB SB EB WB LT TH RT LT TH RT LT TH RT LT TH RT Step 6: Calculate Capacity C L gTOT vci vc gi ci Xij cSUM Xc Steps 7 and 8: Estimate Delay and Level of Service d1 d2 PF d LOS d (int.) LOS (int.) Notes: NB = northbound, SB = southbound, EB = eastbound, WB = westbound, LT = left turn, TH = through, RT = right turn, int. = intersection. Exhibit 72. Signalized intersection planning method (Part 2), calculations worksheet.

114 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual Signalized Intersection Planning Method, Protected-Permitted Left-Turn Worksheet NB SB EB WB gLTPT gLTPM ELTPT ELTPM ELTC VLTTOT VLTPT Notes: NB = northbound, SB = southbound, EB = eastbound, WB = westbound. Exhibit 73. Signalized intersection planning method, protected-permitted left-turn worksheet. 8. Multimodal LOS Bicycle and Pedestrian LOS Procedures for evaluating bicycle and pedestrian LOS at signalized intersections are provided in Section O5. Transit LOS (No Method Available) The HCM does not provide procedures for assessing transit LOS at signalized intersections. Instead, transit LOS is measured at the urban segment and facility levels, with the measure incor- porating the effects of traffic signal delay on overall transit speed. Truck LOS (No Method Available) The HCM does not provide a truck LOS measure for signalized intersections. 9. Example Case Study 2 (Section U in the Guide) provides an example application of the screening and simplified analysis methods described in this section. 10. Reference Highway Capacity Manual: A Guide to Multimodal Mobility Analysis. 6th ed. Transportation Research Board, Washington, D.C., 2016. Exhibit 74. Example sensitivity analysis table for signalized intersection reliability. Percentile Demand (veh/h) Capacity (veh/h) 25th Median (50th) 75th 25th percentile 50th percentile (median) 75th percentile Note: Table is intentionally blank. Entries would be average delays in seconds per vehicle.

115 M. Stop-controlled Intersections 1. Overview Stop-controlled intersections may be all-way Stop-controlled or partially Stop-controlled. A two-way Stop intersection is an example of a partially Stop- controlled intersection. Neither the HCM nor this Guide provides a method for intersections that falls between two-way and all-way Stop control (e.g., four-legged intersections where three legs are Stop-controlled). A two-way Stop-controlled (TWSC) intersection is an intersection in which the movements on one street (the minor street) are controlled by Stop signs, while the movements on the other street (the major street) are not Stop-controlled. An all-way Stop- controlled intersection (AWSC) intersection is one where all movements are Stop-controlled. 2. Applications The procedures in this section are designed to support the following planning and preliminary engineering analyses: • Feasibility studies of intersection improvements, and • Land development traffic impact studies. 3. Analysis Methods Overview Intersection performance can be directly measured in the field or it can be estimated in great detail using microsimulation. However the resource requirements of both of these methods render them generally impractical for most planning and preliminary engineering applications. HCM Chapters 20 and 21 provide a much less resource-intensive approach to estimating intersection performance; however, it is generally impractical to use the HCM methods with 100 percent field-measured inputs for many planning and preliminary engineering analy- ses. Employing the HCM methods with the defaults identified in HCM Chapters 20 and 21 reduces the data requirements, but still requires specialized software to implement the complex computations. This section presents a simplified HCM medium-level method for evaluating all-way and two-way Stop-controlled intersections, as indicated by the unshaded boxes in Exhibit 75. The data needs, assumptions, and limitations of each analysis approach are described as part of the procedures for each approach.

116 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual 4. Simplified HCM Method for All-Way Stop-controlled Intersections An AWSC intersection is an intersection in which all movements are Stop-controlled. The operational analysis method for AWSC intersections, described in HCM Chapter 21, uses an iterative approach to calculate the delay on one approach of the intersection, based on the flow rate on that approach and the flow rates on the other approaches. The method is complex enough to require a computational engine to produce the predictions of delay for even the most basic conditions. The planning method for AWSC intersections is based on the HCM operational analysis method. The method predicts the delay for each intersection approach and for the intersection. Because of the computational complexity of the operational analysis method, the planning method is presented in a series of figures from which the analyst can determine the approach delay and the intersection delay based on the volumes of the two intersecting streets and the number of lanes on each approach. Assumptions, Limitations, and Input Requirements The following assumptions are made in applying the planning method for AWSC intersections: • There are no pedestrians at the intersection, • There are one or two lanes on each approach, • Opposite approaches (e.g., north and south) have the same number of lanes, and • Turning movements account for 20 percent of the traffic on each approach. The AWSC intersection planning method requires two inputs. The analyst is required to specify values for the volume for each movement (in vehicles per hour) and the number of lanes on each approach. Volume-to-Capacity Ratio Estimation For the purposes of computing approximate volume-to-capacity ratios for the intersection, Exhibit 76 can be used. The capacity available to any single approach depends on how much capacity is consumed by the other approaches. High Level Medium Level Low Level Exhibit 75. Analysis options for stop-controlled intersections.

M. Stop-controlled Intersections 117 Delay Estimation The delay on each approach of an AWSC intersection is estimated by entering the Street 1 (subject approach) approach volume and the higher of the Street 2 (cross street) approach volumes in Exhibit 77 for a single-lane approach, or by using Exhibit 78 for an approach with two or more lanes. The delay for the Street 1 volume is then read on the graph’s y-axis. The average intersection delay is then computed by taking a weighted average of the approach delays. Equation 100d v d v i i i ∑ ∑= × where d = average intersection delay (s/veh), vi = volume on approach i (veh/h), and di = delay on approach i (s/veh). Number of Lanes Total Entering Capacity, Street 1 Approach Street 2 Approach All Approaches (veh/h) 1 1 1,200 1 2 1,500 2 2 1,800 Source: Adapted from HCM (2016), Equation 21-14. Note: Assumes average adjusted headways of 3 seconds for single-lane approaches, and two-lane approaches increase capacity by 50%. Exhibit 76. Total entering capacity for AWSC intersections. Exhibit 77. AWSC intersection planning method, street 1 delay, 20% turns, one-lane approaches.

118 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual 5. Simplified HCM Method for Two-Way Stop-controlled Intersections A TWSC intersection is an intersection where the movements on one street (the minor street) are controlled by Stop signs, while the movements on the other street (the major street) are not Stop-controlled. The planning method for TWSC intersections is based on the operational analysis method described in HCM Chapter 20. The TWSC intersection planning method predicts the capacity and delay for all minor-stream movements at a TWSC intersection. The method estimates the capacity of a minor-stream movement based on the conflicting flows of higher-priority traffic streams, and the critical headway and follow-up headway of the minor traffic stream. Assumptions, Limitations, and Data Requirements The planning method for TWSC intersections has the following assumptions and limitations: • No pedestrians at the intersection; • No median on the major street; • Left turns and through movements must be made in one step; • Random vehicle arrivals on the major street, with no platooning from upstream traffic signals; • Exclusive left-turn lane(s) provided on the major street; • No short right-turn lanes are provided; and • No U-turns occur. The TWSC intersection planning method requires four inputs: • Demand volumes Vi (veh/h) for each movement; • Proportion of heavy vehicles PHV for each movement; • Number of lanes (and the turn designation for each) on each approach; and Exhibit 78. AWSC intersection planning method, street 1 delay, 20% turns, two-lane approaches.

M. Stop-controlled Intersections 119 • Intersection peak hour factor PHF for the intersection, either supplied by the analyst or assuming a default value of 0.92. If only approach volumes are known, one of the methods described in Section D8 can be used to generate demand volumes by movement. Capacity Estimation The method uses eight steps to estimate capacity, as shown in Exhibit 79 and described herein. Step 1: Determine and Label Movements and Priorities Movements and priorities are determined and labeled using the numbering scheme from Exhibit 80. The movements are ranked according to the following priorities: • Rank 1 movements are the major street through movements (movements 2 and 5) and major street right turns (3 and 6), • Rank 2 movements are the major street left turns (1 and 4) and the minor street right turns (9 and 12), • Rank 3 movements are the minor street through movements (8 and 11), and • Rank 4 movements are the minor street left turns (7 and 10). Step 2: Convert Movement Demand to Flow Rates Movement demand volumes are converted to flow rates using Equation 101. Equation 101v V PHF i i = where vi = demand flow rate for movement i (veh/h) Vi = demand volume for movement i (veh/h), and PHF = peak hour factor (decimal, default = 0.92). Exhibit 79. TWSC intersection planning method, computational steps.

120 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual Step 3: Determine Conflicting Flows Each non–rank 1 movement faces a unique set of conflicting flows through which the movement must maneuver. For example, a minor street through movement conflicts with one higher ranked movement (its opposing major street left-turn movement) while the minor street left turn conflicts with up to three higher ranked movements (the major street left turns, the opposing minor street through movement, and the opposing minor street right turns). The conflicting flows vc,x for each movement are calculated using the equations herein. The demand flow rates vi , where i ranges from 1 to 12 as shown in Exhibit 80, are the independent variables in these equations. Conflicting flows for the major street left turns (movements 1 and 4) are calculated using Equation 102 and Equation 103: v v vc = + Equation 102,1 5 6 v v vc = + Equation 103,4 2 3 where vc,1, vc,4 = conflicting flow rates for movements 1 and 4, respectively (veh/h), and v2, v3, v5, v6 = demand flow rates for movements 2, 3, 5, and 6, respectively (veh/h). Conflicting flows for the minor street right turns (movements 9 and 12) are calculated using Equation 104 through Equation 107, depending on the number of lanes on the major street: Two-lane major streets: 0.5 Equation 104,9 2 3v v vc = + 0.5 Equation 105,12 5 6v v vc = + Four- and six-lane major streets: 0.5 0.5 Equation 106,9 2 3v v vc = + 0.5 0.5 Equation 107,12 5 6v v vc = + 1 2 3 7 8 9 101112 4 5 6 Exhibit 80. Turning-movement numbering for TWSC intersections.

M. Stop-controlled Intersections 121 where vc,9, vc,12 = conflicting flow rates for movements 9 and 12, respectively (veh/h), and v2, v3, v5, v6 = demand flow rates for movements 2, 3, 5, and 6, respectively (veh/h). Conflicting flows for the minor street through movements (8 and 11) are calculated using Equation 108 and Equation 109: v v v v v v vc = + + + + +2 0.5 2 Equation 108,8 1 2 3 4 5 6 v v v v v v vc = + + + + +2 0.5 2 Equation 109,11 4 5 6 1 2 3 where vc,8, vc,11 = conflicting flow rates for movements 8 and 11, respectively (veh/h); and v1, v2, v3, v4, v5, v6 = demand flow rates for movements 1, 2, 3, 4, 5, and 6, respectively (veh/h). Conflicting flows for the minor street left turns (movements 7 and 10) are calculated using Equation 110 through Equation 115, depending on the number of lanes on the major street: Two-lane major streets: 2 0.5 2 0.5 0.5 0.5 Equation 110,7 1 2 3 4 5 6 12 11v v v v v v v v vc = + + + + + + + 2 0.5 2 0.5 0.5 0.5 Equation 111,10 4 5 6 1 2 3 9 8v v v v v v v v vc = + + + + + + + Four-lane major streets: 2 0.5 2 0.5 0.5 Equation 112,7 1 2 3 4 5 11v v v v v v vc = + + + + + 2 0.5 2 0.5 0.5 Equation 113,10 4 5 6 1 2 8v v v v v v vc = + + + + + Six-lane major streets: 2 0.5 2 0.4 0.5 Equation 114,7 1 2 3 4 5 11v v v v v v vc = + + + + + 2 0.5 2 0.4 0.5 Equation 115,10 4 5 6 1 2 8v v v v v v vc = + + + + + where vc,7, vc,10 = conflicting flow rates for movements 7 and 10, respectively (veh/h); and v1, v2, v3, v4, v5, v6, v8, v9, v11, v12 = demand flow rates for movements 1, 2, 3, 4, 5, 6, 8, 9, 11, and 12, respectively (veh/h). Step 4: Determine Critical and Follow-up Headways The critical headway tc,x is calculated for each movement x as follows: t t t Pc x c c HV HV= + Equation 116, ,base , where tc,x = critical headway for movement x (s), tc,base = base critical headway from Exhibit 81 (s), tc,HV = heavy vehicle adjustment factor (s) = 1.0 for major streets with one lane in each direction and 2.0 for major streets with two or three lanes in each direction, and PHV = proportion of heavy vehicles for movement (decimal).

122 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual The follow-up headway tf,x for each movement x is calculated using Equation 117. Equation 117, ,base ,t t t Pf x f f HV HV= + where tf,x = follow-up headway for movement x (s), tf,base = base follow-up headway from Exhibit 82 (s), tf,HV = heavy vehicle adjustment factor = 0.9 for major streets with one lane in each direction and 1.0 for major streets with two or three lanes in each direction, and PHV = proportion of heavy vehicles for movement (decimal). Step 5: Calculate Potential Capacities The potential capacity cp,x for movement x is calculated using Equation 118. 1 Equation 118, , 3,600 3,600 , , , , c v e e p x c x v t v t c x c x c x f x = − − − where cp,x = potential capacity for movement x (veh/h), vc,x = conflicting flow rate for movement x (veh/h), tc,x = critical headway for movement x (s), and tf,x = follow-up headway for movement x (s). Step 6: Calculate Movement Capacities The movement capacities cm,j for the Rank 2 movements j (major street left turns, movements 1 and 4, and minor street right turns, movements 9 and 12) are calculated using Equation 119. c cm j p j= Equation 119, , Number of Lanes on Major Street Vehicle Movement 2 4 6 Major street left turn (1, 4) 4.1 4.1 5.3 Minor street right turn (9, 12) 6.2 6.9 7.1 Minor street through movement (8, 11) 6.5 6.5 6.5 Minor street left turn (7, 10) 7.1 7.5 6.4 Exhibit 81. Base critical headways. Number of Lanes on Major Street Vehicle Movement 2 4 6 Major street left turn (1, 4) 2.2 2.2 3.1 Minor street right turn (9, 12) 3.3 3.3 3.9 Minor street through movement (8, 11) 4.0 4.0 4.0 Minor street left turn (7, 10) 3.5 3.5 3.8 Exhibit 82. Base follow-up headways.

M. Stop-controlled Intersections 123 where cm,j = movement capacity for Rank 2 movements j ( j = 1, 4, 9, or 12), and cp,j = potential capacity for Rank 2 movements j. The movement capacities cm,k for the Rank 3 movements k (minor street though movements 8 and 11) are calculated using Equation 120 and Equation 121. 1 1 Equation 120,8 ,8 1 ,1 4 ,4 c c v c v c m p m m = −   −  1 1 Equation 121,11 ,11 1 ,1 4 ,4 c c v c v c m p m m = −   −  where cm,1, cm,4, cm,8, cm,11 = movement capacity for movements 1, 4, 8, and 11, respectively (veh/h), cp,8, cp,11 = potential capacity for movements 8 and 11, respectively (veh/h), and v1, v4 = demand flow rates for movements 1 and 4, respectively (veh/h). The movement capacities cm,l for the Rank 4 movements 7 and 10 (minor street left turns) are calculated using Equation 122 through Equation 126. c c p pm p( )( )( )= ′ Equation 122,7 ,7 0,12 c c p pm p( )( )( )= ′ Equation 123,10 ,10 0,9 1 Equation 1240, , p v c i i m i = − p p p p p′ = ′′ − ′′ ′′ + + ′′0.65 3 0.6 Equation 125 movement 7 movement 10 Equation 126 0,1 0,4 0,11 0,1 0,4 0,8 p p p p p p p ( ) ( )′′ =  where cm,i = movement capacity for movement i (veh/h), cp,7, cp,10 = potential capacity for movements 7 and 10, respectively (veh/h), p0,i = probability of a queue-free state for movement i (decimal), vi = demand flow rate for movement i (veh/h), and p′, p″ = adjustments to the impedance created by higher-ranked movements (decimal). Step 7: Calculate Shared Lane Capacities The shared lane capacities cSH (veh/h) of the two minor street approaches are calculated as follows: Equation 127 , c v v c SH yy y m y y ∑ ∑ =

124 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual where cSH = shared lane capacity of a minor street approach (veh/h), vy = demand flow rate of movement y in the subject shared lane (veh/h), and cm,y = movement capacity of movement y in the subject shared lane (veh/h). Step 8: Calculate Delay Estimation The average control delay d for a movement is calculated using Equation 128. 3,600 900 1 1 3,600 450 5 Equation 128 , , , 2 , ,d c T v c v c c v c Tm x x m x x m x m x x m x = + − + −   +           + where d = average control delay (s/veh), vx = demand flow rate for movement x (veh/h), cm,x = movement capacity of movement x (veh/h), and T = analysis time period (h), default = 0.25. The average control delay for all vehicles on an approach dA is calculated using Equation 129. d d v d v d v v v v A r r t t l l r t l = + + + + Equation 129 where dA = average control delay for the approach (s/veh), dr, dt, dl = control delay for the right-turn, through, and left-turn movements on the approach, respectively (s/veh), and vr, vt, vl = demand flow rate of the right-turn, through, and left-turn movements on the approach, respectively (veh/h). The average intersection control delay dI is calculated using Equation 130. d d v d v d v d v v v v v I A A A A A A A A A A A A = + + + + + + Equation 130 ,1 ,1 ,2 ,2 ,3 ,3 ,4 ,4 ,1 ,2 ,3 ,4 where dI = average control delay for the intersection (s/veh), dA,x = average control delay for approach x (s/veh), and vA,x = demand flow rate for approach x (s/veh). 6. Level of Service Analysis (AWSC and TWSC) The LOS ranges for Stop-controlled intersections are given in Exhibit 83 on the basis of control delay. Note that if the volume-to-capacity ratio exceeds 1.00, the LOS will be F regardless of the control delay.

M. Stop-controlled Intersections 125 7. Queuing Analysis (AWSC and TWSC) The deterministic average queue for each Stop-controlled approach at an intersection is determined by dividing the approaches’ average delay by its capacity: Q d c A A SH = 3,600 Equation 131 where QA = deterministic average queue on approach (veh), dA = average approach delay (s/veh), and cSH = shared lane capacity of a minor street approach (veh/h). The deterministic average queue does not take into account the bunching of vehicle arrivals within the analysis period. An approximate estimate of the stochastic 95th percentile queue can be obtained by multiplying the deterministic average queue by 2.0 (the approximate ratio of the 95th percentile to the mean for a Poisson process). For approaches with multiple lanes, the queue per lane can be estimated by dividing by the number of lanes and applying an uneven lane usage adjustment factor to the result. QPL Q LU N = × Equation 132 where QPL = queue per lane (veh/ln), Q = queue (veh), LU = adjustment factor for uneven lane utilization (unitless), default = 1.10, and N = number of lanes on the approach (ln). 8. Worksheets The worksheets shown in Exhibit 84 through Exhibit 86 illustrate how the computations might be laid out in a spreadsheet and can be used to organize manual calculations, as desired. Volume-to-Capacity Ratio, X Control Delay (s/veh) X ≤ 1.0 X > 1.0 ≤10 A F >10–15 B F >15–25 C F >25–35 D F >35–50 E F >50 F F Source: Adapted from HCM (2016), Exhibit 20-2. Exhibit 83. Level of service: stop-controlled intersections.

126 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual All-Way STOP Control (AWSC) Intersec on Planning Method Worksheet Approach NB SB EB WB Turning movement LT TH RT LT TH RT LT TH RT LT TH RT Volume Lanes Delay Delay Notes: NB = northbound, SB = southbound, EB = eastbound, WB = westbound, LT = le turn, TH = through, RT = right turn. Exhibit 84. AWSC intersection delay computation worksheet. Two-Way STOP Control (TWSC) Intersec on Planning Method, Input Data Worksheet Movement 1 2 3 4 5 6 7 8 9 10 11 12 Demand volume, Vi Lanes Peak hour factor PHF Flow rate, vi Proporon of heavy vehicles, PHV Exhibit 85. TWSC input data worksheet. Two-Way STOP Control (TWSC) Intersec on Planning Method, Capacity and Delay Worksheet Movements 1 3 7 8 9 10 11 12 Flow rate, vi Conflicng flows, vc Crical headway, tc Follow-up headway, tf Potenal capacity, cp,x Movement capacity, cm,l Control delay, d Approach control delay, dA Intersec on control delay, di Exhibit 86. TWSC capacity and delay computation worksheet.

M. Stop-controlled Intersections 127 9. Reliability Analysis The HCM does not provide a method for estimating the variability of delay at an intersection. The analyst might perform a sensitivity analysis by repeating the planning computations using the 25th percentile and 75th percentile demands of the year and the 25th percentile and 75th percen- tile capacities of the year (taking into account incidents) and report the results in a table such as shown in Exhibit 87. 10. Multimodal LOS The HCM does not provide bicycle, pedestrian, transit, or truck LOS measures for Stop- controlled intersections. 11. Example Preparation of an example problem was deferred to a future edition of the Guide. 12. Reference Highway Capacity Manual: A Guide to Multimodal Mobility Analysis. 6th ed. Transportation Research Board, Washington, D.C., 2016. Percentile Demand (veh/h) Capacity (veh/h) 25th Median (50th) 75th 25th percentile 50th percentile (median) 75th percentile Note: Table is intentionally blank. Entries would be average delays in seconds per vehicle. Exhibit 87. Example sensitivity analysis table for intersection reliability.

128 N. Roundabouts 1. Overview A roundabout is a circular intersection in which vehicles within the circulatory roadway have the right-of-way. Movements entering the roundabout must yield to traffic already circulating. The planning method for roundabouts is based on the operational analysis method described in HCM Chapter 22. 2. Applications The procedures in this chapter are designed to support the following planning and preliminary engineering analyses: • Feasibility studies of intersection improvements, and • Land development traffic impact studies. 3. Analysis Methods Overview Intersection performance can be directly measured in the field or it can be estimated in great detail using microsimulation. However, the resource requirements of both of these methods render them generally impractical for most planning and preliminary engineering applications. HCM Chapter 22 provides a much less resource-intensive approach to estimating intersection performance; however, it is generally impractical for many planning and pre- liminary engineering analyses to use the HCM methods with 100 percent field-measured inputs. Employing the HCM methods with the defaults identified in HCM Chapter 22 reduces the data requirements but still requires specialized software to implement the complex computations. As indicated by the unshaded boxes in Exhibit 88, this chapter presents a simplified HCM medium-level method for evaluating roundabouts. 4. Simplified HCM Method The roundabout planning analysis approach predicts the capacity and delay for each round- about approach, as well as the delay for the intersection. The planning method is a simplification of the HCM operational analysis method.

N. Roundabouts 129 Data Needs, Assumptions, and Limitations The following assumptions and limitations apply to the simplified HCM planning method for roundabouts: • No pedestrians, • No bypass lanes, and • No more than two lanes within the roundabout and on any entry. The planning method for roundabouts requires four inputs: • The volume for each movement, • The number of lanes on each approach, • The peak hour factor (default = 0.92), and • The proportion of heavy vehicles for each movement (default = 3%). Volume-to-Capacity Ratio Estimation The roundabout planning method includes eight steps to ultimately estimate delay, as shown in Exhibit 89 and described herein. The first seven steps are used to obtain the volume-to-capacity ratios. Step 1: Estimate Flow Rates from Demands Movement demand volumes are converted to flow rates using Equation 133. v V PHF i i = Equation 133 where vi = demand flow rate for movement i (veh/h), Vi = demand volume for movement i (veh/h), and PHF = peak hour factor (decimal), default = 0.92. If only approach volumes are known, one of the methods described in Section D8 can be used to generate turning-movement volumes. High Level Medium Level Low Level Exhibit 88. Analysis options for roundabouts.

130 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual Step 2: Heavy Vehicle Adjustment Demand flow rates in vehicles per hour are adjusted for the presence of heavy vehicles using Equation 134 and Equation 135, producing adjusted flow rates in passenger cars per hour (pc/h). v v f i pce i HV = Equation 134, f P HV T = + 1 1 Equation 135 where vi,pce = adjusted flow rate for movement i (pc/h), vi = demand flow rate for movement i (veh/h), fHV = heavy vehicle adjustment factor (decimal), and PT = proportion of heavy vehicles for movement i (decimal). Step 3: Determine Circulating Flow Rates The circulating flow rates vc,xx,pce are calculated for each approach direction xx of the round- about using Equation 136 through Equation 139. v v v v v v vc NB pce WBU pce SBL pce SBU pce EBT pce EBL pce EBU pce= + + + + + Equation 136, , , , , , , , v v v v v v vc SB pce EBU pce NBL pce NBU pce WBT pce WBL pce WBU pce= + + + + + Equation 137, , , , , , , , v v v v v v vc EB pce NBU pce WBL pce WBU pce SBT pce SBL pce SBU pce= + + + + + Equation 138, , , , , , , , v v v v v v vc WB pce SBU pce EBL pce EBU pce NBT pce NBL pce NBU pce= + + + + + Equation 139, , , , , , , , Exhibit 89. Planning method for roundabouts.

N. Roundabouts 131 where vc,xx,pce = circulating flow rate opposing approach direction xx (pc/h), where xx = NB (north- bound), SB (southbound), EB (eastbound), or WB (westbound), and vxxy,pce = adjusted flow rate for turning-movement y from approach direction xx (pc/h), where y = U (U-turn), L (left turn), or T (through movement). Step 4: Determine Entry Flow Rates by Lane For single-lane entries, the entry flow rate is the sum of all movement flow rates using that entry. For two-lane entries, the following procedure may be used to assign flows to each lane: 1. If only one lane is available for a given movement, the flow for that movement is assigned only to that lane. 2. The remaining flows are assumed to be distributed across the two lanes, subject to the constraints imposed by any designated or de facto lane assignments and any observed or estimated lane utilization imbalances. Five generalized multilane cases may be analyzed with this procedure. For cases in which a movement may use more than one lane, a check should first be made to determine what the assumed lane configuration may be. This may differ from the designated lane assignment based on the specific turning-movement patterns being analyzed. These assumed lane assignments are given in Exhibit 90. For intersections with a different number of legs on each approach, the analyst should exercise reasonable judgment in assigning volumes to each lane. On the basis of the assumed lane assignment for the entry and the lane utilization effect described above, flow rates can be assigned to each lane by using the formulas given in Exhibit 91. Step 5: Determine Capacity of Entry Lane The entry lane capacity ce,pce is determined on the basis of the number of entry and conflicting lanes, using the appropriate equation given in Exhibit 92. Step 6: Convert Lane Flow Rates and Capacities to Vehicles per Hour The flow rates and capacities by lane, in passenger cars per hour, are converted back into vehicles per hour using Equation 140 and Equation 141. v v fj j pce HV= Equation 140, Designated Lane Assignment Assumed Lane Assignment LT, TR If vU + vL > vT + vR,e: L, TR (de facto left-turn lane) If vR,e > vU + vL + vT: LT, R (de facto right-turn lane) Else LT, TR L, LTR If vT + vR,e > vU + vL: L, TR (de facto through–right lane) Else L, LTR LTR, R If vU + vL + vT > vR,e: LT, R (de facto left–through lane) Else LTR, R Notes: vU, vL, vT, vR,e are, respectively, the U-turn, left-turn, through, and nonbypass right-turn flow rates (pc/h) using a given entry. L = left, LT = left–through, TR = through–right, LTR = left–through–right, and R = right. Exhibit 90. Assumed (de facto) lane assignments.

132 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual c c fj j pce HV= Equation 141, where vj = demand flow rate for lane j (veh/h), vj,pce = adjusted flow rate for lane j (pc/h), fHV = heavy vehicle adjustment factor (decimal) from Equation 135, cj = capacity of lane j (veh/h), and cj,pce = capacity of lane j (pc/h). Step 7: Calculate Volume-to-Capacity Ratios The volume-to-capacity ratio xj for each lane j is calculating using Equation 142. x v c j j j = Equation 142 where vj = demand flow rate of the subject lane j (veh/h), and cj = capacity of the subject lane j (veh/h). Step 8: Calculate Delay Estimation If average control delay is desired to be computed, the volume-to-capacity ratio results from Step 7 are carried forward into Step 8. Flow Rate Assignment (pc/h) Assumed Lane Assignment Left Lane Right Lane L, TR vU + vL vT + vR,e LT, R vU + vL + vT vR,e LT, TR (%LL)ve (%RL)ve L, LTR (%LL)ve (%RL)ve LTR, R (%LL)ve (%RL)ve Notes: flow rates (pc/h) using a given entry, and ve is the total entry flow (pc/h). vU, vL, vT, vR,e are, respectively, the U-turn, left-turn, through, and nonbypass right-turn L = left, LT = left–through, TR = through–right, LTR = left–through–right, and R = right. %RL = percentage of entry traffic using the right lane and %LL = percentage of entry traffic using the left lane, with %RL + %LL = 1. Exhibit 91. Flow rate assignments for two-lane entries. Ce,pce = 1,130e–.003vc,pce Ce,pce = 1,130e–.007vc,pce Both lanes: Ce,pce = 1,130e–.003vc,pce Right lane: Ce,pce = 1,130e–.007vc,pce Left lane: Ce,pce = 1,130e–.0075vc,pce Entry Lanes Conflicting Lanes Capacity Equation 1 1 2 1 1 2 2 2 Note: ce,pce = entry lane capacity (pc/h) and vc,pce = conflicting flow rate for the entry (pc/h). Exhibit 92. Capacity equations for roundabouts.

N. Roundabouts 133 Step 8a: Calculate Average Control Delay per Entry Lane. The average control delay d for each entry lane is calculated using Equation 143. 3,600 900 1 1 3,600 450 5 min , 1 Equation 1432 ,d c T x x c x T xm x [ ]( )( )= + − + − +         + where d = average control delay of the subject lane (s/veh), c = capacity of the subject lane (veh/h), T = analysis period duration (h) (default = 0.25 h), x = volume-to-capacity ratio of the subject lane, and cm,x = movement capacity of movement x in the subject lane. Step 8b: Calculate Average Control Delay per Approach. For a single-lane entry, the aver- age control delay for the approach dapproach is the same as the average control delay for the approach’s entry lane. For two-lane entries, the average control delay for the approach is calculated using Equation 144. d d v d v v v LL LL RL RL LL RL = + + Equation 144approach where dapproach = average control delay for the approach (s/veh), dLL = average control delay for the left lane (s/veh), vLL = demand flow rate in the left lane (veh/h), dRL = average control delay for the right lane (s/veh), and vRL = demand flow rate in the right lane (veh/h). Step 8c: Calculate Intersection Average Control Delay. The average control delay for the intersection dintersection is calculated using Equation 145. d d v v i i i ∑ ∑= Equation 145intersection where dintersection = average control delay for the intersection (s/veh), di = average control delay for approach i (s/veh), and vi = demand flow rate for approach i (veh/h). Level of Service Analysis The LOS ranges for motor vehicles are given in Exhibit 93, in the basis of control delay. Note that if the volume-to-capacity ratio exceeds one, the LOS will be F regardless of the control delay. Queuing Analysis The deterministic average queue Q for each approach at an intersection is determined by dividing the average delay for that approach by the capacity for that approach.

134 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual = 3,600 Equation 146Q d c A where QA = deterministic average queue on approach (veh), d = average control delay on approach (s/veh), and c = capacity of approach (veh/h). The deterministic average queue does not take into account the bunching of vehicle arrivals within the analysis period. An approximate estimate of the stochastic 95th percentile queue can be obtained by multiplying the deterministic average queue by 2.0 (the approximate ratio of the 95th percentile to the mean for a Poisson process). For approaches with multiple lanes, the queue per lane can be estimated by dividing by the number of lanes and applying an uneven lane usage adjustment factor to the result. QPL Q LU N = × Equation 147 where QPL = queue per lane (veh/ln), Q = queue (veh), LU = adjustment factor for uneven lane utilization (unitless), default = 1.10, and N = number of lanes on the approach (ln). 5. Worksheets The worksheets provided in Exhibit 94 and Exhibit 95 illustrate how the computations might be laid out in a spreadsheet and can be used to organize manual calculations, as desired. 6. Reliability Analysis The HCM does not provide a method for estimating the variability of delay at an intersection. The analyst might perform a sensitivity analysis by repeating the planning computations using the 25th percentile and 75th percentile demands during the year and the 25th percentile and 75th percentile capacities during the year (taking into account incidents) and report the results in a table such as shown in Exhibit 96. Volume-to-Capacity Ratio, X Control Delay (s/veh) X ≤ 1.0 X > 1.0 ≤10 A F >10–15 B F >15–25 C F >25–35 D F >35–50 E F >50 F F Source: Adapted from HCM (2016), Exhibit 22-8. Exhibit 93. Level of service, roundabouts.

N. Roundabouts 135 Percentile Demand (veh/h) Capacity (veh/h) 25th Median (50th) 75th 25th percentile 50th percentile (median) 75th percentile Note: Table is intentionally blank. Entries would be average delays in seconds per vehicle. Exhibit 96. Example sensitivity analysis table for intersection reliability. Vi rate, Roundabouts Planning Method, Input Worksheet Approach NB SB EB WB Turning- movement LT TH RT LT TH RT LT TH RT LT TH RT Demand volume, Lanes Peak hour factor Demand flow vi Notes: NB = northbound, SB = southbound, EB = eastbound, WB = westbound, LT = left turn, TH = through, RT = right turn. Exhibit 94. Roundabout input worksheet. Roundabouts Planning Method, Volume Adjustments Approach NB SB EB WB Turning- movement LT TH RT LT TH RT LT TH RT LT TH RT Demand flow rate, vi Heavy vehicle adjustment factor, fHV Adjusted flow rate, vi,pce Circulating flow rates, vxx,pce NB Lane 1 NB Lane 2 SB Lane 1 SB Lane 2 EB Lane 1 EB Lane 2 WB Lane 1 WB Lane 2 Entry flow rates by lane, vj Capacity of entry lane, cj Volume-to- capacity ratio, xj Lane control delay, d Approach control delay, dapproach Interstion control delay, dintersection Notes: NB = northbound, SB = southbound, EB = eastbound, WB = westbound, LT = left turn, TH = through, RT = right turn. Exhibit 95. Roundabout volume-to-capacity ratio and delay computation worksheet.

136 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual 7. Multimodal LOS The HCM does not provide bicycle, pedestrian, transit, or truck LOS measures for round- abouts. 8. Example Preparation of an example problem was deferred to a future edition of the Guide. 9. Reference Highway Capacity Manual: A Guide to Multimodal Mobility Analysis. 6th ed. Transportation Research Board, Washington, D.C., 2016.

137 O. Pedestrians, Bicyclists, and Public Transit 1. Overview In addition to providing performance measures and compu- tational methods for the motorized vehicle mode, the HCM also provides a variety of measures for pedestrians and bicycles on vari- ous types of on- and off-street facilities. The HCM also provides a transit LOS measure for evaluating on-street public transit service in a multimodal context. A sister publication, the Transit Capacity and Quality of Service Manual (TCQSM) (Kittelson & Associates et al. 2013), provides a variety of performance measures, com- putational methods, and spreadsheet tools to evaluate the capac- ity, speed, reliability, and quality of service of on- and off-street transit service. The HCM’s pedestrian and bicycle performance measures focus on (1) the impacts of other facility users on pedestrians and bicyclists and (2) facility design and operation features under the control of a transportation agency. However, some analyses may also be interested in the effects of urban design on pedestrians’ and bicyclists’ potential comfort and enjoyment while using a facility. In those cases, additional measures, such as the Walkability Index (Hall 2010) or the Bicycle Environment Quality Index (San Francisco Department of Public Health 2009), could be appropriate. This section is organized by HCM system element, providing guidance on applying the HCM and TCQSM’s pedestrian, bicycle, and transit methods to a planning and preliminary engineering study. As research has not yet been conducted to quantify the pedestrian and bicycle experience for all types of HCM system elements, not every mode is addressed in every subsection below. 2. Freeways Pedestrians and Bicycles In most cases, pedestrians and bicycles are prohibited on freeways; therefore, the operations and quality of service of these modes on freeways is not assessed. In some cases, a multiple-use path is provided within the freeway facility, with a barrier separating non-motorized and motorized traffic. In these situations, the pedestrian and bicycle facility should be analyzed as an off-street pathway (see Section O8). In situations where bicycles are allowed on freeway shoulders, the HCM provides no guidance on evaluating performance. It is not recommended to use the HCM’s multi- lane highway method for bicycles to evaluate bicycle quality of service on freeway shoulders, as the method was developed from urban street and suburban multilane highway data and has not been calibrated to freeway environments.

138 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual Transit Buses operating on freeways in level terrain will generally operate at the same speed as other vehicular traffic, although buses designed to primarily operate on urban streets may not have the power to travel at higher freeway speeds (e.g., over 55 mph). In addition, buses designed to primarily operate on urban streets may have poor performance on steep grades—particularly when fully loaded with passengers—and are recommended to be evaluated as a truck in these cases. Buses designed for freeway travel (i.e., motor coaches designed for long-distance trips) generally do not experience these issues. When bus routes stop along a freeway facility (e.g., at a stop or station in the freeway median or within a freeway interchange), the TCQSM can be consulted for guidance on estimating the delay associated with each stop. The TCQSM can also be consulted for performance measures for rail transit operating within a freeway right-of-way. In general, buses operating on freeway facilities will experience the same conditions as other vehicles in the general purpose or managed lanes (where applicable) and could be assigned the same LOS as for motorized vehicle traffic generally. Alternatively, where buses stop along the freeway facility to serve passengers, the transit LOS measure for urban streets described in Sec- tion O4 could be applied to the stops along the freeway facility, with appropriate adjustments to the assumed average passenger trip length and baseline travel time rate, and considering the pedestrian LOS of the access route to the stop. 3. Multilane and Two-Lane Highways Pedestrians When pedestrian facilities exist along a multilane highway (e.g., a sidewalk along a multilane highway in a suburban area), the facility can be analyzed as an urban street pedestrian facility (see Section O4). However, if the pedestrian facility is separated from a multilane or two-lane highway by a barrier, or is generally located more than 35 feet away from the travel lanes, it should be analyzed as an off-street facility (see Section O8). Lower-speed two-lane highways (posted speeds of 45 mph or less) can be evaluated using the urban street pedestrian method (Section O4), whether or not a sidewalk exists. However, the HCM’s urban street pedestrian method is not calibrated for, and not recommended for use with, higher speed two-lane high- ways or multilane highways lacking sidewalks or sidepaths. Bicycles HCM Chapter 15 provides a method for evaluating bicyclist perceptions of quality of service along multilane and two-lane highways. The method generates a bicycle LOS score, which can be translated into a bicycle LOS letter or used on its own. Exhibit 97 lists the required data for this method and provides suggested default values. Of the inputs listed in Exhibit 97, the LOS result is highly sensitive to shoulder width and heavy vehicle percentage and is somewhat sensitive to lane width and pavement condition (par- ticularly very poor pavement). The calculation of the bicycle LOS score is readily performed by hand, following the steps given in HCM Chapter 15, or can be easily set up in a spreadsheet. Transit The guidance presented above for transit operating on freeways (Section O2) is also applicable to multilane and two-lane highways.

O. Pedestrians, Bicyclists, and Public Transit 139 4. Urban Streets Pedestrians The HCM provides three pedestrian performance measures for urban street segments and facilities: space (reflecting the density of pedestrians on a sidewalk); speed (reflecting intersection delays); and a pedestrian LOS score (reflecting pedestrian comfort with the walking environment). Exhibit 98 lists the data required for these measures and provides suggested default values. Calculating the pedestrian LOS score requires a number of inputs. Most of these can be defaulted, and the ones that cannot be defaulted are used by the urban street motorized vehicle LOS method. Given that different pedestrian design standards are typically used for different combinations of roadway functional classifications and area types, it is recommended that ana- lysts develop sets of default values covering the most common combinations for their study area, based on typical local conditions or design standards. Pedestrian space and speed are sensitive to effective sidewalk width, representing the portion of the sidewalk that is actually used by pedestrians. Common effective width reductions are 1.5 feet adjacent to the curb and 2.0 feet adjacent to a building face; Exhibits 24-8 and 24-9 in HCM Chapter 24, Off-Street Pedestrian and Bicycle Facilities, provide effective width reductions for many other types of objects (e.g., street trees, street light poles, bus stop shelters, café tables). The effective width used for analysis purposes should be based on the narrowest point of the sidewalk from an effective width standpoint. As the types of objects that create effective width reductions will vary depending on the sidewalk design (e.g., use of landscape buffers, street tree presence) and the adjacent land uses, it is recommended that analysts develop a set of local effec- tive width default values that cover the most common situations. The HCM provides a pedestrian LOS score (and associated LOS letter) for urban street links (between signalized intersections), segments (a link plus the downstream intersection), and facil- ities (multiple contiguous segments) that relates to pedestrian perceptions of quality of service for each element. The pedestrian LOS score uses the same scale as related bicycle and transit LOS scores for urban streets, and a related urban street automobile traveler perception score, which allows for multimodal analyses in which the relative quality of service of each travel mode can be evaluated and compared. At present, at a facility level, the HCM methodology only evaluates signalized urban streets, and not streets with all-way stops, roundabouts, or interchanges. How- ever, the link methodology can be used to evaluate pedestrian facilities along any urban street section between intersections. Input Data (units) Default Value Speed limit (mph) Must be provided Directional automobile demand (veh/h)* Must be provided Number of directional lanes 1 (two-lane highway), 2 (multilane highway) Lane width (ft)* 12 Shoulder width (ft)* 6 Pavement condition rating (FHWA 5-point scale) 3.5 (good) Percentage heavy vehicles (decimal)* 0.06** Peak hour factor (decimal)* 0.88 Percent of segment with occupied on-highway parking 0.00 Notes: See HCM Chapter 15 for definitions of the required input data. *Also used by the multilane or two-lane highway LOS methods for motorized vehicles. **HCM Chapter 26 provides state-specific default values. Exhibit 97. Required data for multilane and two-lane highway bicycle analysis.

140 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual As noted above, the pedestrian LOS methodology requires a number of input values, but most of these can be defaulted, particularly when local default values have been established for differ- ent combinations of roadway functional class and area type. The calculations can be performed by hand or (preferably when large numbers of segments will be evaluated) incorporated into a spreadsheet. Equations in HCM Chapter 18, Urban Street Segments, are used to calculate a link LOS score. This score can be converted to a LOS letter and reported by itself, if the purpose of the analysis is to evaluate the pedestrian environment between intersections. Otherwise, the analyst can proceed to calculate a segment LOS score. The segment LOS score combines the link LOS score and the signalized intersection LOS score (see Section O5), weighting the two scores by the relative amounts of time that pedestrians expe- rience each element. It is calculated using HCM Equation 18-39. A roadway crossing difficulty factor also enters into this equation. This factor incorporates the lesser of the delays pedestrians experience when (1) trying to cross the street at an unsignalized midblock location (if legal), or (2) walking to the nearest traffic signal, crossing the street, and walking back on the other side of the street. The segment LOS score can be converted to a LOS letter and reported by itself (using HCM Exhibit 18-2), if the purpose of the analysis is to evaluate the pedestrian environment Input Data (units) For SPC For SPD For PLOS Default Value Sidewalk width (ft) • • • 12 (CBD), 5 (other) Effective sidewalk width (ft) • • 8.5 (CBD), 3.5 (other) Bi-directional pedestrian volume (ped/h) • • Must be provided Free-flow pedestrian speed (ft/s) • • • 4.4 Segment length (ft)* • • Must be provided Signalized intersection delay walking along street (s)* • • See Section O5 or use 12 (CBD), 30 (suburban) Signalized intersection delay crossing street (s)* • See Section O5 or use 12 (CBD), 50 (suburban) Outside lane width (ft)* • 12 Bicycle lane width (ft) • 0 Shoulder/parking lane width (ft) • 1.5 (curb and gutter only) 8 (parking lane provided) Percentage of segment with occupied on- street parking (decimal) • 0.00 (no parking lane) 0.50 (parking lane provided) Street trees or other barriers (yes/no)** • No Landscape buffer width (ft) • 0 (CBD), 6 (other) Curb presence (yes/no) • Yes Median type (divided/undivided) • Undivided Number of travel lanes* • Must be provided Directional vehicle volume (veh/h)* • Must be provided Vehicle running speed (mph)* • See Section K6 or use the posted speed Intersection pedestrian LOS score (unitless) • Calculated, see Section O5 Average distance to nearest signal (ft) • One-third the segment length Notes: See HCM Chapter 18 for definitions of the required input data. SPC = space, SPD = speed, PLOS = pedestrian level of service, CBD = central business district. *Input data used by or calculation output from the HCM urban street motorized vehicle LOS method. **Street trees, bollards, or other similar vertical barriers 3 feet or more tall, or a continuous barrier at least 3 feet tall. Exhibit 98. Required data for urban street pedestrian analysis.

O. Pedestrians, Bicyclists, and Public Transit 141 along a street segment, including intersection and street crossing effects. Otherwise, the analyst can proceed to calculate a facility LOS score. The facility LOS score is calculated similarly to the segment LOS score, weighting the LOS scores of the individual links and signalized intersections that form the facility by the relative amounts of time that pedestrians experience each element. It is calculated using Equation 16-7 in HCM Chapter 16, Urban Street Facilities. Planning Procedure for Estimating Pedestrian LOS When pedestrian crowding and delays at signals are not a concern, then this procedure (adapted from the HCM segment method) can be used to quickly evaluate the pedestrian LOS for stretches of urban streets between signalized intersections. Signalized intersection effects, pedestrian density, and midblock roadway crossing difficulty are not considered in this proce- dure. For high pedestrian volume locations (over 1,000 pedestrians per hour), the HCM proce- dure for evaluating pedestrian space should be used. The pedestrian segment LOS is determined by the perceived separation between pedestrians and vehicle traffic: • Higher traffic speeds and higher traffic volumes reduce the perceived separation, • Physical barriers and parked cars between motorized vehicle traffic and the pedestrians increase the perceived separation, and • Sidewalks wider than 10 feet do not further increase the perceived separation. The segment pedestrian LOS is calculated as follows: PLOS f W W OSP f W f W V N SPD LV T B B SW S[ ] [ ] [ ][ ]( )[ ] ( ) = − × × + × + × + × + × + + × + 1.2276 ln 0.5 0.5 % 0.0091 4 0.0004 6.0468 Equation 148 1 2 where PLOS = pedestrian level of service score for a segment (unitless), ln = natural logarithm, fLV = low volume factor (unitless) = 1.00 if V > 160 veh/h and (2.00 – 0.005V) otherwise, WT = distance from the inner edge of the outside lane to the curb (ft) (see Exhibit 99), W1 = distance from the outer edge of the outside lane to the curb (ft) (see Exhibit 99), Exhibit 99. Measurement of widths for pedestrian LOS analysis.

142 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual %OSP = percent of segment with occupied on-street parking (percent), fB = buffer area coefficient (unitless) = 5.37 if a barrier is provided and 1.00 otherwise, WB = buffer width (ft), the distance between the curb and sidewalk (see Exhibit 99), fSW = sidewalk presence coefficient (unitless) = 6 – 0.3WS, WS = sidewalk width (ft) (see Exhibit 99), with a maximum allowed value of 10 ft, V = directional volume of vehicles in the direction closest to pedestrians (veh/h), N = number of through lanes of traffic in the direction closest to pedestrians, and SPD = average vehicle speed between intersections (excluding stops) (mph). Vertical objects at least 3 feet tall, such as street trees, bollards, or concrete barriers, that are sufficiently dense to be perceived as a barrier are treated as barriers for the purposes of determin- ing the buffer area coefficient fb. The furnishings zone portion of a sidewalk (e.g., the area with street furniture, planters, and tree wells), such as often found in central business districts with wide sidewalks, is treated as part of the buffer strip width WB. In these cases, the portion of the sidewalk allocated to pedestrian circulation would be used to determine the sidewalk width WSW. The pedestrian LOS method has not been designed or tested for application to rural highways and other roads where a sidewalk is not present and the traffic volumes are low but the speeds are high. The PLOS score value is converted into a LOS letter using Exhibit 100. Special Cases This section gives guidance on the analysis of special cases. Treatment of Sections with Significant Grades. The pedestrian LOS equations are designed for essentially flat grades (grades of under 2% of any length). For steeper grades, the analyst should consider applying an adjustment to the LOS estimation procedure to account for the negative impact of both upgrades and downgrades on pedestrian quality of service. This adjust- ment probably should be sensitive both to the steepness of the grade and its length. However, research available at the time this Guide was produced did not provide a basis for computing such an adjustment. The precise adjustment is left to the discretion of the analyst. Pedestrian LOS and ADA Compliance. The Americans with Disabilities Act (ADA) sets various accessibility requirements for public facilities, including sidewalks on public streets. The United States Access Board (www.access-board.gov) has developed specific accessibility guide- lines that apply to sidewalks and pedestrian paths. Because pedestrian LOS is defined to reflect the average perceptions of the public, it is not designed to specifically reflect the perspectives of any particular subgroup of the public. Thus, the analyst PLOS Score LOS ≤1.50 A >1.50–2.50 B >2.50–3.50 C >3.50–4.50 D >4.50–5.50 E >5.50 F Source: Adapted from HCM (2016), Exhibit 18-2. Exhibit 100. Level of service, pedestrians on urban streets.

O. Pedestrians, Bicyclists, and Public Transit 143 should use caution if applying the pedestrian LOS methodology to facilities that are not ADA com- pliant. Pedestrian LOS is not designed to reflect ADA compliance or non-compliance, and therefore should not be considered a substitute for an ADA compliance assessment of a pedestrian facility. Treatment of Street Sections with a Parallel Multiuse Path. Pedestrian LOS for urban streets applies to sidewalks and sidepaths located within 35 feet of the street (i.e., the distance within which research has demonstrated that vehicular traffic influences pedestrians’ percep- tions of quality of service). When a pedestrian pathway is located parallel to the street, but more than 35 feet from the street, it should be evaluated as an off-street pathway (see Section 08). Treatment of Streets with Sidewalk on Only One Side. The pedestrian LOS analysis for both sides of the street proceeds normally. On one side, the sidewalk is evaluated. On the other side, the pedestrian LOS is evaluated using a sidewalk width of 0 feet. Treatment of Discontinuous Sidewalks. Segments with relatively long gaps (over 100 feet) in the sidewalk should be split into sub-segments and the LOS for each evaluated separately. The pedestrian LOS methodology is not designed to take into account the impact of short gaps in sidewalk (under 100 feet). Until such a methodology becomes available, short gaps may be neglected in the pedestrian LOS calculation. However, the analyst should report the fact that there are gaps in the sidewalk in addition to reporting the LOS grade. Treatment of One-Way Traffic Streets. The pedestrian LOS analysis proceeds normally for both sides of the street, even when it is one-way. Note, however, that the lane and shoulder width for the left-hand lane are used for the sidewalk on the left-hand side of the street. Treatment of Streets with Pedestrian Prohibitions or Sidewalk Closures. If pedestrians are prohibited from walking along the street by local ordinance or a permanent sidewalk closure, then the pedestrian LOS is F. No pedestrian LOS computations are performed. Treatment of Streets with Frontage Roads. In some cases a jurisdiction will provide front- age roads to an urban street. There will usually be no sidewalks along the urban street, but there will be sidewalks along the outside edge of each frontage road. If the analyst has information indicating that pedestrians walk along the urban street without the sidewalks, then the pedestrian LOS analysis should be performed for the urban street. If the analyst has information indicating that pedestrians walk exclusively along the frontage roads, then the pedestrian LOS analysis should be performed for the frontage roads. Treatment of Pedestrian Overcrossings. The pedestrian LOS methodology is not designed to account for pedestrian bridges, either across the urban street or along the urban street. Treatment of Railroad Crossings. The pedestrian LOS methodology is not designed to account for the impacts on pedestrian LOS of railroad crossings with frequent train traffic. Treatment of Unpaved Paths/Sidewalks. The pedestrian LOS methodology is not designed to account for unpaved paths in the urban street right-of-way. The analyst should use local knowledge about the climate and the seasonal walkability of unpaved surfaces to determine whether an unpaved surface can be considered as almost as good as a paved sidewalk for the purpose of the pedestrian LOS computation. Otherwise the unpaved path should be considered the same as no sidewalk for the purpose of pedestrian LOS computation. Treatment of Major Driveways. The HCM pedestrian LOS method and the planning pro- cedure presented here are not designed to address the impacts of high-volume driveways on the pedestrian experience.

144 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual Bicycles The HCM provides two bicycle performance measures for urban street segments and facilities: average travel speed (reflecting intersection delays) and a bicycle LOS score (reflecting bicyclist comfort with the bicycling environment). Exhibit 101 lists the data required for these measures and provides suggested default values. As can be seen in Exhibit 101, calculating the bicycle LOS score requires a number of inputs. Most of these can be defaulted, and the ones that cannot be defaulted are used by the urban street motorized vehicle or pedestrian LOS methods. Given that different bicycle design standards are typically used for different combinations of roadway functional classifications and area types, it is recommended that analysts develop sets of default values covering the most common combi- nations for their study area, based on typical local conditions or design standards. Bicycle LOS Score The HCM provides a bicycle LOS score (and associated LOS letter) for urban street links (between signalized intersections), segments (a link plus the downstream intersection), and facilities (multiple contiguous segments) that relates to bicyclist perceptions of quality of service for each element. The bicycle LOS score uses the same scale as related pedestrian and transit LOS scores, and a related urban street automobile traveler perception score, which allows for multimodal analyses in which the relative quality of service of each travel mode can be evaluated and compared. At present, at a facility level, the HCM methodology only evaluates signalized urban streets and not streets with all- way stops, roundabouts, or interchanges. However, the link methodology can be used to evaluate bicycle facilities along any urban street section between intersections. Input Data (units) For SPD For BLOS Default Value Bicycle running speed (mph) • 12 Signalized intersection delay (s) • • See Section O5 or use 10 (CBD), 22 (suburban) Segment length (ft)* • • Must be provided Bicycle lane width (ft)** • 5 (if provided) Outside lane width (ft)** • 12 Shoulder/parking lane width (ft)** • 0 (curb and gutter only) 8 (parking lane provided) Percentage of segment with occupied on-street parking (percent)** • 0 (no parking lane) 50 (parking lane provided) Pavement condition rating (1–5) • 3.5 (good) Curb presence (yes/no)** • Yes Median type (divided/undivided)** • Undivided Number of travel lanes* • Must be provided Directional vehicle volume (veh/h)* • Must be provided Vehicle running speed (mph)* • See Section K6 or use the posted speed Percentage heavy vehicles (%)* • 3% Access points on the right side (points/mi) • 17 (urban arterial), 10.5 (suburban arterial), 30.5 (urban collector), 24 (suburban collector) Intersection bicycle LOS score (unitless) • Calculated, see Section O5 Notes: See HCM Chapter 18 for definitions of the required input data. SPD = speed, BLOS = bicycle level of service, CBD = central business district. *Input data used by or calculation output from the HCM urban street motorized vehicle LOS method. **Input data used by the HCM urban street pedestrian LOS method. Exhibit 101. Required data for urban street bicycle analysis.

O. Pedestrians, Bicyclists, and Public Transit 145 As noted, the bicycle LOS methodology requires a number of input values, but most of these can be defaulted, particularly when local default values have been established for different combi- nations of roadway functional class and area type. The calculations can be performed by hand or (preferably when large numbers of segments will be evaluated) incorporated into a spreadsheet. Equations 18-41 through 18-44 in HCM Chapter 18, Urban Street Segments, are used to calculate a link LOS score. This score can be converted to a LOS letter and reported by itself, if the purpose of the analysis is to evaluate the bicycling environment between intersections. Otherwise, the analyst can proceed to calculate a segment LOS score. The segment LOS score combines the link LOS score and the signalized intersection LOS score (see Section O5), weighting the two scores by the relative amounts of time that bicyclists experi- ence each element. It is calculated using HCM Equation 18-46. The number of access points per mile on the right side of the road (e.g., driveways, unsignalized cross-streets) also enters into this equation as a factor that causes discomfort to bicyclists. The segment LOS score can be converted to a LOS letter and reported by itself (using HCM Exhibit 18-3), if the purpose of the analysis is to evaluate the bicycling environment along a street segment, including intersection and access point effects. Otherwise, the analyst can proceed to calculate a facility LOS score. The facility LOS score is calculated similarly to the segment LOS score, weighting the LOS scores of the individual links and signalized intersections that form the facility by the relative amounts of time that bicyclists experience each element. It is calculated using Equation 16-10 in HCM Chapter 16, Urban Street Facilities. Planning Procedure for Evaluating Bicycle LOS If bicyclist perceptions of signalized intersections are not a significant concern, the fol- lowing planning method (adapting the HCM segment LOS method) can be used to quickly assess bicycle LOS for a street. The segment bicycle LOS is calculated according to the following equation: 0.507 ln 4 0.199 1 0.1038 7.066 1 0.005 0.760 Equation 149 2 2 2 BLOS V N f HV PC W s e ( )[ ] ( ) = ×     + × × + + ×        − × + where BLOS = bicycle level of service score for a segment (unitless), ln = natural logarithm, V = directional volume of vehicles in the direction closest to bicyclists (veh/h), N = number of through lanes of traffic in the direction closest to bicyclists, fs = effective speed factor (unitless) = (1.1199 × ln[S – 20] + 0.8103, HV = proportion of heavy vehicles in the motorized vehicle volume (%), PC = pavement condition rating, using FHWA’s five-point scale (1 = poor, 5 = excellent), We = average effective width of the outside through lane (ft) = Wv – (0.1 × %OSP) if Wl < 4 and Wv + Wl – (0.2 × %OSP) otherwise, with a minimum value of 0, Wv = effective width of the outside through lane as a function of traffic volume (ft) = WT if V > 160 veh/h or the street is divided, and WT × (2 – 0.005V) otherwise, %OSP = percent of segment with occupied on-street parking (percent), Wl = width of the bicycle lane and paved shoulder (ft); a parking lane can only be counted as shoulder if 0% occupied (see Exhibit 102) and the gutter width is not included, and

146 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual WT = width of the outside through lane, bicycle lane if present, and paved shoul- der if present (ft); a parking lane can only be counted as shoulder if 0% occupied (see Exhibit 102) and the gutter width is not included. If the traffic volume V is less than 200 veh/h, the value of HV must be less than or equal to 50% to avoid unrealistically poor LOS results for the combination of low volume and high percentage of heavy vehicles. Note that this method does not account for bicycle-to-bicycle interference and should not be used where bicycle flows are expected to be high enough that significant bicycle-to-bicycle interference occurs. The bicycle LOS score is converted into a letter using Exhibit 103. Simplifications from the HCM The HCM method for estimating bicycle level of service for urban streets is documented in HCM Chapters 16 (Urban Street Facilities), 18 (Urban Street Segments), and 19 (Signalized Intersections). This Guide makes the following simplifications to the HCM method to improve its utility for planning applications: • Intersection analysis and facility analysis are excluded, • Estimation of bicycle speeds and delays is excluded, Exhibit 102. Widths used in bicycle LOS analysis. BLOS Score LOS ≤1.50 A >1.50–2.50 B >2.50–3.50 C >3.50–4.50 D >4.50–5.50 E >5.50 F Source: Adapted from HCM (2016), Exhibit 18-3. Exhibit 103. Level of service, bicycles on urban streets.

O. Pedestrians, Bicyclists, and Public Transit 147 • Bicycle link LOS is used to characterize the segment (intersection plus link), and • No provision is made for characterizing overall facility bicycle LOS. For these features, the analyst must apply the HCM method as described in the HCM, applying default values as needed. Special Cases This section explains the evaluation of bicycle LOS for special cases. Treatment of Sections with Significant Grades. The bicycle LOS equations are designed for essentially flat grades (grades of under 2% of any length). For steeper grades, the analyst should consider applying an adjustment to the LOS estimation procedure to account for the negative impact of both upgrades and downgrades on bicycle LOS. This adjustment probably should be sensitive both to the steepness of the grade and its length. However, research available at the time of production of this Guide did not provide a basis for computing such an adjustment. It is left to the discretion of the analyst. Treatment of Sections with Parallel Multiuse Path. The bicycle LOS is computed separately for bicycles using the street and for bicycles using the parallel path. The bicycle LOS for the path is computed using the off-street path procedures described in Section O8. Treatment of Bus Lanes, Bus Streets, and High Bus Volumes. The bicycle LOS methodol- ogy is not designed to adequately represent bicyclist perceptions of quality of service when they are operating on streets with frequent bus service with bus stops requiring bicyclists to move left to pass stopped buses. The analyst may choose to impose a weighting factor on the bus volume to better reflect the greater impact of the stopping buses on bicyclist LOS. The weighting factor would be at the analyst’s discretion. Treatment of Railroad Crossings and In-Street Tracks. The LOS methodology is not designed to account for the impacts of railroad crossings and the presence of tracks in the street (which may constitute a crash risk for bicyclists traveling parallel to the tracks) on bicycle LOS. The analyst may choose to adjust the pavement condition factor to a lower value to reflect the impacts of parallel in-pavement tracks and railroad crossings on bicycle LOS. Transit The HCM provides a transit LOS score for urban streets that reflects passenger comfort as they walk to a bus stop, wait for a bus, and ride on the bus. In addition, the TCQSM (Kittelson & Associates et al. 2013) provides the most up-to-date methods for calculating bus capacities and estimating average bus speeds on urban streets. Exhibit 104 lists the data required for these measures and suggests default values. The HCM’s transit LOS measure can be used to evaluate fixed-route transit service (e.g., bus, streetcar) that operates on the street and makes periodic stops to serve passengers. The TCQSM (Kittelson & Associates et al. 2013) can be used to evaluate the quality of service provided by other transit modes that travel within, above, or below the street right-of-way. Bus Capacity Bus capacity on an urban street is usually controlled by the capacity of the bus stops to accept and discharge buses. Bus capacity reflects the number of buses per hour that can serve the critical bus stop along a facility, at a desired level of reliability. The critical bus stop is typi- cally the bus stop with the highest dwell time (i.e., serves the greatest number of passengers),

148 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual but a lower-passenger-volume stop with short green times for buses or that experiences high right-turning traffic volumes can also be the critical stop. Bus capacity is calculated using Equation 150 and Equation 151, adapted from the TCQSM: 3,600 Equation 150B N f g C t t g C Zc t el tb c d v d ( ) ( )= + + f f v c tb l cl cl = −  1 Equation 151 where B = bus capacity (bus/h), Nel = number of effective loading areas at a bus stop, from Exhibit 105, ftb = traffic blockage adjustment factor (decimal), 3,600 = number of seconds in 1 hour, Input Data (units) For CAP For SPD For TLOS Default Value Dwell time at critical stop (s) • • ○ 60 (CBD, major transfer point), 30 (urban), 15 (suburban) Average dwell time along facility (s) • ○ 45 (CBD), 20 (urban), 15 (suburban) Coefficient of variation of dwell times (decimal) • • ○ 0.60 Through traffic g/C ratio at critical stop (decimal)* • • ○ 0.45 (CBD), 0.35 (other) Curb lane v/c ratio at critical stop’s intersection* (decimal) • • ○ Must be provided Busiest stop location (online/offline) • • ○ Offline Clearance time at critical stop (s) • • ○ 10 (online stop, queue jump), 14 (far-side/midblock offline stop), 25 (near-side offline stop) Number of loading areas at critical stop • • ○ 1 Design failure rate (%) • • ○ 10% (CBD), 2.5% (other), 25% (when calculating speed) Bus frequency (bus/h) • • Must be provided Average bus stop spacing (stops/mi) • ○ 8 (CBD), 6 (urban), 4 (suburban) Posted speed limit (mph)* • ○ Must be provided Average bus acceleration rate (ft/s2) • ○ 3.4 Average bus deceleration rate (ft/s2) • ○ 4.0 Bus lane type (4 categories) • ○ Mixed traffic Traffic signal progression (3 categories) • ○ Typical Average passenger load factor (p/seat) • Must be provided Average excess wait time (min) • 3 Percentage of stops with shelter (%) • 25% Percentage of stops with bench (%) • 25% Average passenger trip length (mi) • 3.7 Pedestrian LOS score (decimal)** • Must be provided Notes: See the TCQSM for definitions of the required input data. CAP = capacity, SPD = speed, TLOS = transit level of service, CBD = central business district. ○ = required input if bus speeds are not already known (e.g., when evaluating future conditions). *Input data used by or calculation output from the HCM urban street automobile LOS method. **Calculation output from the HCM pedestrian LOS method. Exhibit 104. Required data for urban street transit analysis.

O. Pedestrians, Bicyclists, and Public Transit 149 g/C = ratio of effective green time to total traffic signal cycle length (decimal), tc = clearance time (s), td = average (mean) dwell time (s), Z = standard normal variable corresponding to a desired failure rate, from Exhibit 106, cv = coefficient of variation of dwell times (decimal), fl = bus stop location factor (decimal), from Exhibit 107, vcl = curb lane traffic volume at intersection (veh/h), and ccl = curb lane capacity at intersection (veh/h). When more than one bus can use the critical bus stop at a time (i.e., more than one loading area is provided), the bus stop’s capacity will be greater than if only one loading area was pro- vided. Exhibit 105 gives the number of effective loading areas for a given number of physical loading areas, for both online stops (buses stop in the travel lane) and offline stops (buses stop out of the travel lane). Exhibit 106 provides values for Z, the standard normal variable, for different design failure rates—the percentage of time that a bus should arrive at a bus stop only to have to wait for other buses to finish serving their passengers before space opens up for the arriving bus to enter the stop. Capacity is maximized when a queue of buses exists to move into a bus stop as soon as other buses leave, but this situation causes significant bus delays and schedule reliability problems. Therefore, a lower design rate is normally used as an input for determining a design capacity, balancing capacity with operational reliability. However, the TCQSM’s method for estimating Number of Physical Loading Areas Bus Stop Type Online Offline 1 1.00 1.00 2 1.75 1.85 3 2.45 2.60 4 2.65 3.25 5 2.75 3.75 Source: Adapted from TCQSM (Kittelson & Associates et al., 2013), Exhibit 6-63. Note: Values are numbers of effective loading areas for a given number of physical loading areas. Exhibit 105. Efficiency of multiple loading areas at bus stops. Design Failure Rate Z 1.0% 2.330 2.5% 1.960 5.0% 1.645 7.5% 1.440 10.0% 1.280 15.0% 1.040 20.0% 0.840 25.0% 0.675 Source: Adapted from TCQSM (Kittelson & Associates et al., 2013), Exhibit 6-56. Exhibit 106. Values of Z associated with given failure rates.

150 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual bus speed is calibrated to maximum capacity and therefore uses a 25% (maximum practical) failure rate in its calculation. The location of the critical bus stop relative to the nearest intersection and the ability of buses to avoid right-turning traffic also influence capacity. Exhibit 107 gives values for the bus stop location factor fl used in Equation 151. The curb lane capacity can be estimated using the procedure given in Section L4 or estimated from Exhibit 108, for a given combination of g/C ratio (effective green time divided by the traffic signal cycle length) and conflicting pedestrian volume for right turns. Bus Speed Two options are provided for planning-level estimates of bus speeds along urban streets: 1. If only a planning estimate of bus speeds is desired, then Option 1 can be followed. This option requires less data and is faster to calculate. It accounts for traffic and traffic signal delays in a generalized way. 2. If it is desired to estimate both automobile and bus speeds, then Option 2 can be followed. This option applies the same basic method used for automobiles, but makes adjustments to reflect (a) overlapping signal delay time and bus dwell time to serve passengers, (b) bus delays waiting to re-enter the traffic stream, and (c) bus congestion at bus stops when more than half of the facility’s bus capacity is being used. Option 1: Generalized Bus Speed Method. This option is based on the TCQSM’s bus speed estimation method. In this option, bus speeds are calculated in four steps. First, an unimpeded Bus Freedom to Maneuver Bus Stop Location Buses Restricted to Right Lane Buses Can Use Adjacent Lane Right Turns Prohibited or Dual Bus Lanes Near-side of intersection 1.0 0.9 0.0 Middle of the block 0.9 0.7 0.0 Far-side of intersection 0.8 0.5 0.0 Source: Adapted from TCQSM (Kittelson & Associates et al., 2013), Exhibit 6-66. Exhibit 107. Bus stop location factor fl values. Conflicting Pedestrian g/C Ratio for Curb Lane Volume (ped/h) 0.35 0.40 0.45 0.50 0.55 0.60 0 510 580 650 730 800 870 100 440 510 580 650 730 800 200 360 440 510 580 650 730 400 220 290 360 440 510 580 600 70 150 220 290 360 440 800 * * 70 150 220 290 1,000 * * * * 70 150 Source: HCM (2016), based on 1,450 × (g/C) × [1 – (pedestrian volume × (g/C) / 2,000)] Note: *Vehicles can only turn at the end of green, assume one or two per traffic signal cycle. Values shown are for CBD locations, multiply by 1.1 for other locations. with PHF = 1. Exhibit 108. Approximate curb lane capacities.

O. Pedestrians, Bicyclists, and Public Transit 151 bus travel time rate, in minutes per mile, is calculated for the condition in which a bus moves along a street without traffic or traffic signal delays, with the only source of delay being stops to serve passengers. Second, additional delays due to traffic and traffic signals are estimated. Third, the bus travel time rate is converted to an equivalent speed. Finally, the speed is reduced to reflect the effects of bus congestion. Step 1: Unimpeded Bus Travel Time Rate. The unimpeded bus travel time rate is based on the posted speed, the number of stops per mile, the average dwell time per stop, and typical bus acceleration and deceleration rates. It is based on the delay experienced with each bus stop (deceleration, dwell time, and acceleration) and the time spent traveling at the bus’s running speed (typically the posted speed) between stops. It is calculated using Equation 152 through Equation 157: t t N t t t u rs s dt acc dec( ) = + + + 60 Equation 152 t L v rs rs run = 1.47 Equation 153 L N Lrs s ad= − ≥5,280 0 Equation 154 L at dtad acc dec= +0.5 0.5 Equation 1552 2 t v a acc run = 1.47 Equation 156 t v d dec run = 1.47 Equation 157 where tu = unimpeded running time rate (min/mi), trs = time spent at running speed (s/mi), Ns = average stop spacing (stops/mi), tdt = average dwell time of all stops within the section (s/stop), tacc = acceleration time per stop (s/stop), tdec = deceleration time per stop (s/stop), 60 = number of seconds per minute, Lrs = distance traveled at running speed per mile (ft/mile), 1.47 = conversion factor (5,280 ft/mi/3,600 s/h), vrun = bus running speed on the facility (typically the posted speed) (mph), Lad = distance traveled at less than running speed per stop (ft/stop), a = average bus acceleration rate to running speed (ft/s2), and d = average bus deceleration rate from running speed (ft/s2). If the calculated length traveled at running speed in Equation 155 is less than zero, the bus cannot accelerate to the input running speed before it must begin decelerating to the next stop. In this case, the calculation sequence must be performed again with a lower run- ning speed selected. The maximum speed that can be reached before the bus has to begin decelerating again can be computed using Equation 158 and Equation 159; however, the analyst may wish to choose a lower speed to reflect that bus drivers will typically cruise at a constant speed for some distance between stops, rather than decelerating immediately after accelerating.

152 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual t N a a d acc dc s = + 5,280 0.5 Equation 158, 2 v a t max acc dc = × 1.47 Equation 159 , where tacc,dc = distance-constrained acceleration time (s), Ns = average stop spacing (stops/mi), a = bus acceleration rate (ft/s2), d = bus deceleration rate (ft/s2), and vmax = maximum speed achievable between stops (mph). Step 2: Additional Bus Travel Time Delays. Next, additional bus travel time delays tl (in minutes per mile) are estimated directly from Exhibit 109, using the bus facility type, traffic signal progression quality, and area type as inputs. Step 3: Base Bus Speed. The unimpeded bus travel time rate from Step 1 and the additional bus travel time delays from Step 2 are added together to obtain a base bus travel time rate tr , which is then converted into a base bus speed Sb: t t tr u l= + Equation 160 S t b r = 60 Equation 161 where tr = base bus running time rate (min/mi), tu = unimpeded running time rate (min/mi), tl = additional running time losses (min/mi), 60 = number of minutes in an hour, and Sb = base bus speed (mph). Step 4: Average Bus Speed. When at least half of a facility’s maximum bus capacity is sched- uled, bus congestion at bus stops reduces bus speeds below the base speed calculated in Step 3. The amount of this speed reduction is given by the bus–bus interference factor fbb, which can be Condition Bus Lane Bus Lane, No Right Turns Bus Lane With Right-Turn Delays Bus Lanes Blocked by Traffic Mixed Traffic Flow CENTRAL BUSINESS DISTRICT Typical 1.2 2.0 2.5–3.0 3.0 Signals set for buses 0.6 1.4 Signals more frequent than bus stops 1.75 2.75 3.25 3.75 ARTERIAL ROADWAYS OUTSIDE THE CBD Typical 0.7 1.0 Source: Adapted from TCQSM (Kittelson & Associates et al., 2013), Exhibit 6-73. Exhibit 109. Estimated bus running time losses on urban streets tl (min/mi).

O. Pedestrians, Bicyclists, and Public Transit 153 estimated from Exhibit 110. The input to this exhibit is the bus volume–to–maximum capacity ratio, where maximum bus capacity is estimated by using a 25% failure rate in Exhibit 106 when determining the value of the standard normal variable Z used in the bus capacity equation (Equa- tion 150). Under typical conditions and if bus stops can only serve one bus at a time (i.e., one loading area per stop), at least 10–15 buses per hour need to be scheduled before bus speeds are affected. Equation 162 is used to estimate the average bus speed on the urban street facility. S S fbus b bb= Equation 162 where Sbus = average bus speed along facility (mph), Sb = base bus speed (mph), and fbb = bus–bus interference factor (decimal). Option 2: Modified Auto Speed Method. This option modifies the auto speed estimation method for urban street segments with signalized intersections (see Section K6) to reflect addi- tional delays experienced by buses and to account for potentially overlapping traffic signal delay and dwell time delay. The auto equation for estimating segment travel time is modified as follows for buses: T FFS L d d di bus i mb bs= + + + 5,280 3,600 Equation 163, where Ti,bus = base bus travel time for segment i (s), FFS = midblock free-flow speed (mph), 5,280 = number of feet per mile, 3,600 = number of seconds per hour, Li = distance from upstream intersection stop bar to downstream intersection stop bar for segment i (ft), d = average control delay (s), dmb = midblock bottleneck delay (if any) (s), and dbs = total bus stop delay in the segment (s). Bus Volume–to– Maximum Capacity Ratio Bus–Bus Interference Factor <0.5 1.00 0.5 0.97 0.6 0.94 0.7 0.89 0.8 0.81 0.9 0.69 1.0 0.52 1.1 0.35 Source: TCQSM (Kittelson & Associates et al., 2013), Exhibit 6-75. Exhibit 110. Bus–bus interference factor values.

154 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual Total bus stop delay in the segment is calculated as follows: d N t t t tbs s dt acc dec re( )= + + + Equation 164 where dbs = total bus stop delay in the segment (s), Ns = number of bus stops in the segment (stops), tdt = average dwell time per stop (s/stop), tacc = bus acceleration time per stop (s/stop), tdec = bus deceleration time per stop (s/stop), tre = average re-entry delay per stop (s/stop) = tcl – 10, and tcl = average clearance time per stop (s/stop). When applying Equation 164, the number of bus stops in the segment includes all mid-block stops and any bus stop associated with the downstream intersection (even if far-side and technically located in the next segment). Similarly, any bus stop associated with the upstream intersection is excluded from the count of bus stops. Average bus speed in the segment is calculated as follows: S L T fi bus i i bus bb= 3,600 5,280 Equation 165, , where Si,bus = average bus speed for segment i including all delays (mph), Li = distance from upstream intersection stop bar to downstream intersection stop bar for segment i (ft), Ti,bus = base bus travel time for segment i (s), and fbb = bus–bus interference factor (decimal) from Exhibit 110. Average facility bus speed is calculated as follows: 3,600 5,280 Equation 166 , S L T bus i i bus ∑ ∑= where Sbus = average bus speed along facility (mph), Li = distance from upstream intersection stop bar to downstream intersection stop bar for segment i (ft), 5,280 = number of feet per mile, 3,600 = number of seconds per hour, Ti,bus = base bus travel time for segment i (s). Transit LOS Score The HCM provides a transit LOS score (and associated LOS letter) for urban street segments (a link plus the downstream intersection) and facilities (multiple contiguous segments). The segment score relates to transit passengers’ experiences walking to or from bus stops in the seg- ment, waiting for buses at bus stops in the segment, and riding on buses within the segment. The transit LOS score uses the same scale as related pedestrian and bicycle LOS scores, and a related auto traveler perception score, allowing for multimodal analyses in which the relative quality of service of each travel mode can be evaluated and compared to each other. The calculations

O. Pedestrians, Bicyclists, and Public Transit 155 can be performed by hand or (preferably when large numbers of segments will be evaluated) incorporated into a spreadsheet. HCM Equations 18-56 through 18-63 are used to calculate a link LOS score. This score can be converted to a LOS letter and reported by itself (using HCM Exhibit 18-3), if the purpose of the analysis is to evaluate transit conditions within a segment. Otherwise, a facility score is calculated by weighting the LOS scores of the individual segments that form the facility by the relative length of each segment. It is calculated using HCM Equation 16-13. The transit LOS score is particularly sensitive to the bus frequency provided as an input, and is somewhat sensitive to the average bus speed and passenger load factor provided as inputs. The HCM transit LOS score computations can be applied without change using defaults as needed. Alternatively, the transit LOS score computation steps shown below provide a few sim- plifications on the HCM procedure for planning applications. TLOS s PLOSw r( ) ( )= − × + ×−6.0 1.50 0.15 Equation 167 where TLOS = transit LOS score (unitless), sw-r = transit wait and ride score (unitless), and PLOS = pedestrian LOS score (unitless). The computed transit LOS score is converted to an LOS letter using the equivalencies given in Exhibit 111. Pedestrian LOS Estimation. The pedestrian LOS score for the urban street is estimated using the pedestrian LOS model described earlier in this section. Better PLOS values (i.e., LOS A-C) improve the TLOS score relative to what it would be if only transit factors were considered, while worse PLOS values (i.e., LOS D-F) reduce the TLOS score. Transit Wait-Ride Score Estimation. The transit wait–ride score is a function of a bus head- way factor fh that reflects the multiplicative change in ridership along a route at a given headway, relative to the ridership at 60-minute headways, and a perceived travel time factor fptt that reflects the multiplicative change in ridership along a route at a given perceived travel time rate (PTTR), relative to the ridership at a baseline travel time rate (BTTR). The suggested baseline travel time rate is 4 min/mi (15 mph), except in the central business districts of metropolitan areas with over 5 million population, in which case it is 6 min/mi (10 mph). (These values can be adjusted by the analyst to reflect local passenger expectations of travel speeds.) Equation 168 shows the calculation of the transit wait-ride score. TLOS Score LOS ≤2.00 A >2.00–2.75 B >2.75–3.50 C >3.50–4.25 D >4.25–5.00 E >5.00 F Source: Adapted from HCM (2016), Exhibit 18-3. Exhibit 111. Level of service, transit on urban streets.

156 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual = × − Equation 168s f fw r h ptt where sw-r = transit wait–ride score (unitless), fh = headway factor (unitless), and fptt = perceived travel time factor (unitless). The headway factor calculation incorporates assumed ridership elasticities that relate the per- centage change in ridership to the percentage change in bus headways. Only the buses and bus routes that actually stop to pick up or drop off passengers within the study section of the street should be included in determining the average bus headway on the street. Express bus service without at least one bus stop on the street would be excluded. Equation 169 is used to calculate the headway factor. f hh ( )= × −4 exp 0.0239 Equation 169 where fh = headway factor (unitless), and h = average number of minutes between buses. Perceived Travel Time Factor. The perceived travel time factor calculation incorporates assumed ridership elasticities that relate the percentage change in ridership to the percentage change in the perceived travel time rate. The perceived travel time rate, in turn, is a function of actual bus speeds (travel time rates) and factors that have been found to make the time spent waiting for or riding on the bus seem longer than the actual time. These factors include late bus arrivals; provision of shelters, benches, or both at bus stops; and crowding on board the bus. The perceived travel time factor is calculated using Equation 170 through Equation 172. f e BTTR e PTTR e PTTR e BTTR ptt [ ] [ ] ( ) ( ) ( ) ( )= − − + − − + 1 1 1 1 Equation 170 PTTR a IVTTR a EWTR ATR( ) ( )= × + × − Equation 1711 2 IVTTR Sbus = 60 Equation 172 where fptt = perceived travel time factor (unitless), e = ridership elasticity with respect to changes in the travel time rate (unitless), default = -0.40, BTTR = baseline travel time rate (min/mi), default = 6 for the central business district of metropolitan areas with populations of 5 million or greater, and 4 otherwise, PTTR = perceived travel time rate (min/mi), a1 = travel time perception coefficient for passenger load (unitless) = 1.00 when 80% or fewer of seats are occupied, 1.19 when all seats are occupied, and 1.42 with a standing load equal to 25% of the seating capacity; HCM Equation 18-59 can also be used, IVTTR = in-vehicle travel time rate (min/mi), a2 = travel time perception coefficient for excess wait time (unitless), default = 2.0, EWTR = excess wait time rate (min/mi) = (average wait for buses beyond the scheduled arrival time)/(average passenger trip length), default = 0.8, and

O. Pedestrians, Bicyclists, and Public Transit 157 ATR = amenity time rate (min/mi) = (perceived wait time reduction due to bus stop amenities)/(average passenger trip length); default = 0.1 (bench provided), 0.3 (shelter only), and 0.4 (shelter and bench). When field measurement of average bus speeds along the street is not feasible, the in-vehicle travel time rate can be estimated from the bus schedule as the travel time between timepoints on either side of the study section, divided by the on-street distance between the timepoints. The bus speed estimation procedure presented earlier can also be used. The excess wait time is the average difference between the scheduled and actual arrival times for buses at the timepoint prior to the study section. For example, if buses arrive 3 minutes behind schedule on average at the timepoint, the excess wait time is 3 minutes. An early arrival at the timepoint without a corresponding early departure is treated as 0 minutes of excess wait time, but an early arrival combined with an early departure is counted as being one headway late. Special Cases. This section gives guidance on the analysis of special cases. Gaps in Transit Service. The portions of street where there is no transit service should be split into their own segments for the purpose of transit LOS analysis (if not already split for other reasons). The transit LOS should be set at F for these segments. The rest of the transit LOS analysis proceeds normally, with the overall transit LOS being a length-weighted average includ- ing the segments with no transit service. No Through Transit Service for the Full Length of the Study Facility. The TLOS score is measured on a segment-by-segment basis and reflects in part actions that a roadway agency can take to improve bus speeds. It also reflects the amount of bus service provided within a given segment. It can be compared on a segment-by-segment basis to the LOS scores available for other travel modes, reflecting the quality of service provided within that segment. In this respect, it does not measure origin–destination service quality for transit passengers. Therefore, by default, no adjustment is made to the score if passengers would need to transfer from one route to another to make a complete trip through the study facility. However, if the analyst is interested in measuring origin–destination service quality along a facility, one option would be to calculate the TLOS score as described above, but (1) double the assumed average trip length to reflect the linked (i.e., involving a transfer) trip, and (2) add a perceived transfer time rate equal to the average transfer time multiplied by a perceived waiting time factor (suggested default = 2) and divided by the average trip length. Single-Direction Transit Service on a Two-Way Street. The direction of travel for which there is no transit service is assigned transit LOS F. The other direction of travel is evaluated normally. Bus Lanes and Bus Streets. The methodologies are not specifically designed to handle bus streets and bus lanes, but with some judicious adjustments, they can be adapted to these special situations. In the case of bus streets, the auto LOS is, by definition, LOS F (since autos cannot access this street). The transit, bicycle, and pedestrian LOS are computed normally, with transit vehicles being the only motorized vehicles on the street. In the case of bus lanes, the auto, transit, bicycle, and pedestrian LOS analyses proceed nor- mally. The only difference is that only transit vehicles (and carpools, if allowed) are assigned to the bus lane.

158 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual Simplifications from the HCM The HCM method for estimating transit level of service for urban streets is documented in HCM Chapters 16 (Urban Street Facilities), 18 (Urban Street Segments), and 19 (Signalized Intersections). The transit LOS method presented above makes the following simplifications to the HCM method to improve its utility for planning applications: • Bus running speeds are based solely on bus acceleration and deceleration characteristics rather than on motor vehicle running speeds (which are discounted in the HCM for midblock inter- ference along the street segment). • Bus stop delay is not adjusted for the location of the bus stop (e.g., near-side or far-side). • Bus stop re-entry delay is not computed. • Default values are provided for the a1 passenger load travel time perception factor in lieu of the HCM equation that uses the exact passenger load as an input. • A default value of 3 minutes excess wait time was used in lieu of computing it from on-time arrival statistics. To take full advantage of these features the analyst must apply the HCM method as described in HCM Chapter 18, applying defaults as needed. 5. Signalized Intersections Pedestrians The HCM provides two pedestrian performance measures suitable for planning analyses of signalized intersections: average pedestrian delay and a pedestrian LOS score that reflects pedes- trian comfort while crossing an intersection. Exhibit 112 lists the data required for these mea- Used By Input Data (units) DEL PLOS Default Value Traffic signal cycle length (s)* • • 60 (CBD), 120 (suburban) Major street walk time (s) • • See Section L or use 19 (CBD), 31 (suburban), 7 (minimum) Minor street walk time (s) • • See Section L or use 19 (CBD), 7 (suburban), 7 (minimum) Number of lanes crossed on minor street crosswalk* • Must be provided Number of channelizing islands crossed on minor street crosswalk • 0 15-minute volume on major street (veh)* • Must be provided Number of major street through lanes in the direction of travel* • Must be provided Mid-block 85th percentile speed on major street (mph) • Posted speed limit Right-turn on red flow rate over the minor street crosswalk (veh/h) • 0 (right turns on red prohibited) Must be provided (otherwise) Permitted left-turn volume over the minor street crosswalk (veh/h) • 0 (protected left-turn phasing) 10% of through 15-minute volume (permitted left-turn phasing) 5% of through 15-minute volume (protected-permitted left-turn phasing) Notes: See HCM Chapter 19 for definitions of the required input data. DEL = delay, PLOS = pedestrian level of service, CBD = central business district. *Input data used by or calculation output from the HCM urban street automobile LOS method. Exhibit 112. Required data for signalized intersection pedestrian analysis.

O. Pedestrians, Bicyclists, and Public Transit 159 sures and provides suggested default values. The HCM also provides calculation methods for assessing intersection corner circulation area and crosswalk circulation area, but these typically require more detailed data than would be available for a planning analysis. Pedestrian Delay Average pedestrian delay for a given signalized crosswalk is calculated as follows: d C g C p ( ) = − 2 Equation 173 Walk 2 where dp = average pedestrian delay (s), C = cycle length (s), and gWalk = effective walk time for the crosswalk (s). Pedestrian LOS Score The HCM provides a method (Equations 19-71 through 19-76 in Chapter 19, Signalized Intersections) for calculating a pedestrian LOS score (and associated LOS letter using HCM Exhibit 19-9) for signalized intersections. This score can be used on its own or integrated into the urban street pedestrian LOS procedures. Most of the method’s inputs are required by the auto LOS method for signalized intersections or can be defaulted. An exception is the right-turn- on-red flow rate over the crosswalk being analyzed. The LOS score is sensitive to this input and a wide range of values are possible. The HCM recommends developing local default values for this variable for use in planning analyses. Bicycles The HCM provides two bicycle performance measures for signalized intersections: average bicycle delay and a bicycle LOS score that reflects bicyclist comfort while crossing an intersection. Exhibit 113 lists the data required for these measures and provides suggested default values. Used By Input Data (units) DEL BLOS Default Value Traffic signal cycle length (s)* • 60 (CBD), 120 (suburban) Effective green time for bicycles (s) • Effective green time for parallel through automobile traffic* 15-minute bicycle flow rate (bicycles/h) • Must be provided 15-minute automobile flow rate (veh/h)* • Must be provided Cross street width (ft) • Must be provided Bicycle lane width (ft) • 5 (if provided) Outside lane width (ft)* • 12 Shoulder/parking lane width (ft) • 1.5 (curb and gutter only) 8 (parking lane provided) Percentage of intersection approach and departure with occupied on-street parking (decimal) • 0.00 (no parking lane) 0.50 (parking lane provided) Number of parallel through lanes (shared or exclusive)* • Must be provided Notes: See HCM Chapter 19 for definitions of the required input data. DEL = delay, BLOS = bicycle level of service, CBD = central business district. *Input data used by or calculation output from the HCM urban street automobile LOS method. Exhibit 113. Required data for signalized intersection bicycle analysis.

160 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual Bicycle Delay When bicyclists share the lane with automobile traffic, bicyclist delay is the same as automo- bile delay and can be calculated using Equation 97 (see Section L5). When bicyclists have their own lane, bicycle delay is calculated as follows: 0.5 1 1 min , 1.0 Equation 174 2 d C g C v c g C b b bic b b ( ) = − −     c s g C b b b = Equation 175 where db = average bicycle delay (s), gb = effective green time for the bicycle lane (s), C = cycle length (s), vbic = bicycle flow rate (bicycles/h), cb = bicycle lane capacity (bicycles/h), and sb = bicycle lane saturation flow rate (bicycles/h) = 2,000. Bicycle LOS Score The HCM provides a method (Equations 19-79 through 19-82) for calculating a bicycle LOS score (and associated LOS letter using HCM Exhibit 19-9) for signalized intersections. This score can be used on its own or integrated into the urban street bicycle LOS procedures. Most of the method’s inputs are required by the auto LOS method for signalized intersections or can be defaulted. Transit The HCM does not provide a transit LOS score for signalized intersections; the impacts of signalized intersections on bus speeds are incorporated into the segment and facility LOS scores (see Section O4). 6. Stop-controlled Intersections Pedestrians Two-Way Stops and Midblock Crossings The HCM 2016 provides a method for estimating pedestrian delay crossing the major street at two-way stop-controlled intersections and at midblock crosswalks. Exhibit 114 lists the required data. Input Data (units) Default Value Crosswalk length (ft) Must be provided Average pedestrian walking speed (ft/s) 3.5 Pedestrian start-up time and end clearance time (s) 3 Number of through lanes crossed Must be provided Vehicle flow rate during the peak 15 min (veh/s) Must be provided; note the units of veh/s Note: See HCM Chapter 20 for definitions of the required input data. Exhibit 114. Required data for two-way stop-controlled intersection pedestrian delay calculation.

O. Pedestrians, Bicyclists, and Public Transit 161 When a pedestrian refuge area is available in the street median, pedestrians can cross the street in two stages. In this case, delay should be calculated separately for each stage of the crossing and totaled to determine the overall delay. First, pedestrian delay is calculated for the scenario in which motorists do not yield to pedes- trians (i.e., pedestrians must wait for a suitable gap in traffic). This calculation neglects the additional delay that occurs when pedestrian crossing volumes are high enough that pedestrian platoons form (i.e., some pedestrians have to wait for the pedestrians ahead of them to step off the curb before they can enter the crosswalk). The following equations are used: t L S tc p s= + Equation 176 1 Equation 177P eb t v N c L= − − 1 1 Equation 178P Pd b NL( )= − − d v e vtg vt cc( )= − −1 1 Equation 179 d d P gd g d = Equation 180 where tc = critical headway for a single pedestrian (s), Sp = average pedestrian walking speed (ft/s), L = crosswalk length (ft), ts = pedestrian start-up time and end clearance time (s), Pb = probability of a blocked lane (i.e., an approaching vehicle at the time the pedestrian arrives at the crosswalk that prevents an immediate crossing), Pd = probability of a delayed crossing, NL = number of through lanes crossed, v = vehicular flow rate (veh/s), dg = average pedestrian gap delay (s), and dgd = average gap delay for pedestrians who incur nonzero delay. When motorists yield to pedestrians, pedestrian delay is reduced. The average pedestrian delay in this scenario is calculated as follows: d h i P Y P P Y dp i n i d i i n gd∑ ∑( ) ( ) ( )= − + −  = = 0.5 Equation 181 1 1 where dp = average pedestrian delay (s), i = sequence of vehicle arrivals after the pedestrian arrives at the crosswalk, n = average number of vehicle arrivals before an adequate gap is available = Int(dgd/h), h = average vehicle headway for each through lane (s), Pd = probability of a delayed crossing, P(Yi) = probability that motorist i yields to the pedestrian, from Exhibit 115, and dgd = average gap delay for pedestrians who incur nonzero delay. The motorist yielding rate My is an input to the equations in Exhibit 115, and all other vari- ables in the exhibit are as defined previously. Yielding rates for a selection of pedestrian crossing treatments are given in Exhibit 20-24 in HCM Chapter 20, Two-Way stop-controlled Inter- sections. Alternatively, local values can be developed from field observations.

162 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual All-Way Stops The HCM 2016 provides a qualitative discussion of contributors to pedestrian delay at all-way stop-controlled intersections. However, the research base does not exist to provide a calculation method. Bicycles The HCM 2016 provides qualitative discussions of bicycle delay at two-way and all-way stop- controlled intersections. However, the research base does not exist to provide calculation methods. Transit Buses will experience the same amount of control delay as other motor vehicles at these intersections. 7. Roundabouts Pedestrian delay at roundabouts can be estimated using the methods for two-way stop- controlled intersections (see Section O6). The HCM provides no quantitative method for esti- mating bicycle delay, although it can be expected to be similar to vehicular delay, if bicyclists circulate as vehicles, or to pedestrian delay, if bicyclists dismount and use the crosswalks. Buses will experience the same amount of control delay as other motor vehicles. 8. Off-Street Pathways The HCM 2016 provides LOS measures for three combinations of modes and facility types: • Pedestrians on an exclusive off-street pedestrian facility, • Pedestrians on a shared-use path, and • Bicyclists on an exclusive or shared off-street facility. Exhibit 116 lists the required data for analyzing each of these situations. Lanes Crossed Probability of Vehicle i Yielding 1 = − Equation 182 2 = − − + Equation 183 3 = − + − + − Equation 184 4 = − × + − + − + − Equation 185 Exhibit 115. Equations for calculating probability of vehicles yielding to a crossing pedestrian.

O. Pedestrians, Bicyclists, and Public Transit 163 Pedestrians on an Exclusive Off-Street Facility Pedestrian LOS on an exclusive facility is based on the average space available to pedestrians. It is calculated using the following three equations: v v PHF h = ×4 Equation 18615 v v W p E = ×15 Equation 187 15 A S v p p p = Equation 188 where v15 = pedestrian flow rate during peak 15 min (p/h), vh = pedestrian demand during analysis hour (p/h), PHF = peak hour factor, vp = pedestrian flow per unit width (p/ft/min), WE = effective facility width (ft), Used By Input Data (units) PEX PSH BIKE Default Value Facility width (ft) • • Must be provided Effective facility width (ft) • Same as facility width Pedestrian volume (ped/h) • Must be provided Bicycle volume (bicycles/h) • Must be provided Total path volume (p/h) • Must be provided Bicycle mode split (%) • 55% of path volume Pedestrian mode split (%) • 20% of path volume Runner mode split (%) • 10% of path volume Inline skater mode split (%) • 10% of path volume Child bicyclist mode split (%) • 5% of path volume Peak hour factor (decimal) • • • 0.85 Directional volume split (decimal) • • 0.50 Average pedestrian speed (ft/min) • 300 Average pedestrian speed (mph) • • 3.4 Average bicycle speed (mph) • • 12.8 Average runner speed (mph) • 6.5 Average inline skater speed (mph) • 10.1 Average child bicyclist speed (mph) • 7.9 SD of pedestrian speed (mph) • 0.6 SD of bicycle speed (mph) • 3.4 SD of runner speed (mph) • 1.2 SD of inline skater speed (mph) • 2.7 SD of child bicyclist speed (mph) • 1.9 Segment length (mi) • Must be provided Walkway grade ≤ 5% (yes/no) • Yes Pedestrian flow type (random/platooned) • Random Centerline stripe presence (yes/no) • No Source: Default values from Hummer et al. (2006), except for effective facility width. Notes: See HCM Chapter 24 for definitions of the required input data. PEX = pedestrian LOS on an exclusive path, PSH = pedestrian LOS on a shared path, BIKE = bicycle LOS on all types of off-street pathways, SD = standard deviation. Exhibit 116. Required data for off-street pathway analysis.

164 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual Ap = average pedestrian space (ft 2/p), and Sp = average pedestrian speed (ft/min). Average pedestrian space is converted into an LOS letter using Exhibit 24-1 (for random pedestrian flow) or Exhibit 24-2 (when pedestrian platoons form) in HCM Chapter 24, Off- Street Pedestrian and Bicycle Facilities. HCM Exhibit 24-18 can be used to estimate the reduc- tion in average pedestrian speed that occurs when walkway grades exceed 5%. The LOS result is highly sensitive to the average pedestrian speed provided as an input. Pedestrians on a Shared Off-Street Facility Pedestrian LOS on a shared off-street facility is based on the number of times per hour an average pedestrian meets or is passed by bicyclists using the path. The weighted number of meet- ing and passing events is calculated as follows: F Q PHF S S p sb p b = −  1 Equation 189 F Q PHF S S m ob p b = + 1 Equation 190 F F Fp m( )= + 0.5 Equation 191 where Fp = number of passing events (events/h), Fm = number of meeting events (events/h), Qsb = bicycle demand in same direction (bicycles/h), Qob = bicycle demand in opposing direction (bicycles/h), PHF = peak hour factor, Sp = mean pedestrian speed on path (mph), Sb = mean bicycle speed on path (mph), and F = weighted total events on path (events/h). The weighted total events F is converted into an LOS letter using HCM Exhibit 24-4. The LOS result is sensitive to the peak hour factor provided as an input. Bicyclists on an Off-Street Facility Bicycle LOS on all types of off-street facilities is based on a bicycle LOS score that considers: • The average number of times per minute a bicyclist meets or is overtaken by other path users, • The path width, • The presence or absence of a centerline stripe, and • The average number of times per minute a bicyclist is delayed in passing another path user (for example, because an oncoming path user is in the way). At a minimum, total path width and the total number of hourly path users must be pro- vided, although results will be more accurate if the actual mode split of path users (bicyclists, pedestrians, runners, inline skaters, and child bicyclists) is known or can be defaulted using local values. The bicycle LOS score is particularly sensitive to the bicycle mode split, the peak hour factor, and the directional distribution provided as inputs, and somewhat sensitive to whether or not a centerline stripe is present. HCM Exhibit 24-5 is used to convert the bicycle LOS score into an LOS letter.

O. Pedestrians, Bicyclists, and Public Transit 165 The calculation process requires a large number of computations, and the use of a computational engine is recommended. The FHWA project (Hummer et al. 2006) that developed the method devel- oped an engine, which can be downloaded from http://www.fhwa.dot.gov/publications/research/ safety/pedbike/05138/SharedUsePathsTLOSCalculator.xls. The FHWA computational engine applies the peak hour factor in a different order in the computational sequence than the HCM implementation of the method does. However, any difference between the two methods is negligible for planning purposes. 9. References Hall, R. A. HPE’s Walkability Index—Quantifying the Pedestrian Experience. Compendium of Technical Papers, ITE 2010 Technical Conference and Exhibit, Savannah, Ga., March 2010. Highway Capacity Manual: A Guide to Multimodal Mobility Analysis. 6th ed. Transportation Research Board, Washington, D.C., 2016. Hummer, J. E., N. M. Rouphail, J. L. Toole, R. S. Patten, R. J. Schneider, J. S. Green, R. G. Hughes, and S. J. Fain. Evaluation of Safety, Design, and Operation of Shared-Use Paths—Final Report. Report FHWA-HRT-05-137. Federal Highway Administration, Washington, D.C., July 2006. Kittelson & Associates, Inc., Parsons Brinckerhoff, KFH Group, Inc., Texas A&M Transportation Institute, and Arup. TCRP Report 165: Transit Capacity and Quality of Service Manual, 3rd Edition. Transportation Research Board of the National Academies, Washington, D.C., 2013. San Francisco Department of Public Health. Bicycle Environmental Quality Index (BEQI) Draft Report. San Francisco, Calif., June 2009.

166 P. Truck Level of Service 1. Overview The HCM does not provide a truck LOS measure. However, NCFRP Report 31 (Dowling et al. 2014) does provide a truck LOS measure, which is presented in this section. 2. Truck Level of Service Index Truck LOS is defined as a measure of the quality of service provided by a facility for truck hauling of freight as perceived by shippers and carriers. It is measured in terms of the percent- age of ideal conditions achieved by the facility for truck operations. A logistic function is used to compute the percentage of ideal conditions achieved by the facility for truck operations. TLOS e U x( )= + ( )−% 1 1 0.10 Equation 192 200 where %TLOS = truck LOS index as a percentage of ideal conditions (decimal), U(x) = truck utility function, and e = exponential function. Ideal conditions are defined as a facility usable by trucks with legal size and weight loads, with no at-grade railroad crossings, that provides reliable truck travel at truck free-flow speeds, at low cost (i.e., no tolls). Reliable performance is defined as 100% probability of on-time arrival for the truck. A facility is considered to deliver 100% probability of on-time arrival as long as its travel time index for trucks falls below 1.33 for uninterrupted-flow facilities (i.e., freeways and highways) and 3.33 for interrupted-flow facilities (i.e., streets and highways with signals, roundabouts, or stop control no more than 2 miles apart). (These values are approximately the automobile LOS E/F thresholds for these facility types.) The truck travel time index is the ratio of the truck free-flow speed to the actual truck speed. Truck free-flow speed is defined as the maximum sustainable speed that an average truck can achieve under low traffic flow conditions given the prevailing grades, exclusive of intersection delays. Truck Utility Function A truck utility function is used for computing the truck LOS index. U x A POTA B TTI C Toll mi D TFI( ) ( ) ( ) ( ) ( )= × − + × − + × + × −1 1 1 Equation 193

P. Truck Level of Service 167 where U(x) = utility of facility for truck shipments, A = weighting parameter for reliability, sensitive to shipping distance = 5 / ASL, ASL = average shipment length (mi) = 200 mi (lower 48 states), 280 mi (Alaska), and 30 mi (Hawaii), B = weighting parameter for shipment time, sensitive to free-flow speed = -0.32 / FFS, FFS = free-flow speed, C = weighting parameter for shipment cost = -0.01, D = weighting parameter for the facility’s truck friendliness = 0.03, POTA = probability of on-time arrival = 1 if the mixed flow (autos and trucks) travel time index is ≤1.33 (freeways and highways) or ≤3.33 (urban streets), TTI = truck travel time index for the study period, the ratio of truck free-flow speed to actual truck speed, Toll/mi = truck toll rate (dollars per mile), a truck volume–weighted average for all truck types, and TFI = truck friendliness index, where 1.00 = no constraints or obstacles to legal truck load and vehicle usage of facility and 0.00 = no trucks can use the facility. The truck friendliness index for a facility can be reduced below 1.00 at an agency’s discretion to reflect the effects of restrictions on truck load, length, width, height, turning radius, or a com- bination of these (Dowling et al. 2014). The utility function is weighted so that truck friendliness indices of 0.60 or less will always result in LOS F, regardless of a facility’s speed or reliability. Note that the utility function is designed to work with data on the truck’s experience: prob- ability of on-time arrival for the truck shipment, the travel time index for trucks, tolls paid by trucks, and the truck friendliness index. For many of these data, the HCM and this Guide pro- vide methods only for estimating mixed flow (auto and truck) speeds and reliability. Until truck specific performance estimation procedures become available, the analyst must decide if the mixed flow results produced by the HCM and this Guide are applicable to trucks, and whether or not to apply an adjustment to the mixed flow performance to obtain truck specific performance for the purposes of estimating truck LOS. Truck LOS Thresholds The truck LOS index is the ratio of the utility for actual conditions over the utility for ideal conditions. The truck LOS index is converted into an equivalent letter grade based on its freight facility class, according to the thresholds given in Exhibit 117. The thresholds for a given letter grade are higher for the higher class facilities. MAP-21, the Moving Ahead for Progress in the 21st Century Act, requires the Department of Transportation to establish a national freight network to assist states in strategically directing resources toward improved movement of freight on highways (Federal Register 2013). At the LOS Class I (Primary Freight Facility) Class II (Secondary Facility) Class III (Tertiary Facility) A ≥90% ≥85% ≥80% B ≥80% ≥75% ≥70% C ≥70% ≥65% ≥60% D ≥60% ≥55% ≥50% E ≥50% ≥45% ≥40% F <50% <45% <40% Exhibit 117. Truck LOS thresholds by truck LOS index and freight facility class.

168 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual time of writing, guidelines on establishing freight facility classes had not yet been developed. Until these guidelines are set, Exhibit 118 provides a tentative three-class system that employs some of the general criteria outlined in MAP-21 for classifying highway facilities by their relative importance to the regional and national economy. 3. Estimating Probability of On-Time Arrival from TTI If the cumulative distribution of travel time indices for the facility is available, it is a simple matter for the analyst to read the probability of on-time arrival for any selected on-time arrival threshold (for example, the threshold might be defined as 1.33 times the free-flow travel time). If only the median (50th percentile) and 95th percentile travel time index (TTI) are avail- able to the analyst, then the probability of on-time arrival (POTA) can be estimated through extrapolation. Straight-line extrapolation is used if the 95th percentile TTI is ≤1.33 (freeways) or ≤3.33 (urban streets); otherwise, the value is determined by interpolating between the 50th and 95th percentile TTIs. For example, if one is evaluating the probability of on-time arrival for a freeway, the selected target TTI is 1.33 (for mixed auto and truck traffic). That is the threshold above which the free- way is congested (i.e., speeds are below the speed at capacity). If the 50% TTI is 1.10 and the 95% TTI is 5.00, then the probability of on-time arrival is 53%, computed as follows: POTA ( ) ( )( )= + − × − − 50% 95% 50% 1.33 1.10 5.00 1.10 Equation 194 4. A Service-TTI Lookup Table for Truck LOS The estimation of truck level of service can be expedited by estimating the average peak hour mixed auto and truck speed. After making an adjustment to the mixed traffic speed to get the truck speed, one can apply Equation 192 to estimate the 95th percentile peak hour speed for trucks. From this information, plus other assumed defaults, one can then construct a “Service-TTI” look-up table. TTI TTIm( )= + ×1 3.67 ln Equation 19595 Exhibit 118. Facility freight classification system. Facility Class Description Suggested Criteria Examples Class I Highway facility critical to the inter- regional or within region movement of goods. Facility carries a high volume of goods by truck (by tonnage or by value). Trucks may account for a high volume or percentage of AADT compared to other facilities in the region. Interstate freeway, inter-regional rural principal arterial. Class II Highway facility of secondary importance to goods movement within or between regions. Facility carries lesser volumes of goods (by tonnage or value). Trucks account for a lesser volume or percentage of AADT. Urban principal arterial, connector to major intermodal facilities (maritime port, intermodal rail terminal, airports). Class III Highway facility of tertiary importance to goods movement within or between regions. Connectors to significant single origins/destinations of goods, such as major manufacturing facilities, sources of raw materials (mines, oil, etc.). Connectors to truck service facilities and terminals. Access roads to mines, energy production facilities, factories, truck stops, truck terminals. Note: AADT = annual average daily traffic.

P. Truck Level of Service 169 The average peak hour truck TTI is estimated from the peak hour auto speed by applying a local adjustment factor fLA to reflect local driving characteristics. This factor might apply, for example, if the truck speed limit is set lower than the auto speed limit and trucks comply with the lower limit, or where extended upgrades reduce truck speeds significantly below auto speeds. Otherwise, a default value of 1.00 can be used for fLA. TTI TTI fLA( ) ( )= ×truck mixed Equation 196 where TTI(truck) = truck travel time index, TTI(mixed) = ratio of the free-flow speed to the actual speed for mixed auto and truck traffic, and fLA = the local adjustment factor to account for local truck driving behavior (decimal). The analyst enters Exhibit 119 for the appropriate facility type and (for urban streets only) free-flow speed using the computed truck TTI for average peak hour conditions. Interpolation in the table is allowed. The table shows the estimated 95th percentile TTI, the Facility Freight Facility Class Type Truck TTI 95% TTI POTA Utility %TLOS Class I Class II Class III Fr ee w ay s an d Ru ra l H ig hw ay s 1.05 1.18 99.83% 0.000 90.39% A A A 1.10 1.35 93.77% -0.002 86.81% B A A 1.15 1.51 81.16% -0.006 76.86% C B B 1.20 1.67 69.34% -0.009 63.56% D D C 1.25 1.82 60.20% -0.011 51.15% E E D 1.30 1.96 53.33% -0.013 41.31% F F E 1.35 2.10 48.04% -0.015 33.88% F F F 1.40 2.23 43.86% -0.016 28.27% F F F Si gn al iz ed U rb an S tr ee ts FF S = 55 m ph 1.20 1.67 100.00% -0.001 88.79% B A A 1.40 2.23 99.89% -0.002 86.20% B A A 1.60 2.72 98.93% -0.004 82.51% B B A 1.80 3.16 96.51% -0.006 76.80% C B B 2.00 3.54 92.67% -0.008 68.40% D C C 2.20 3.89 87.70% -0.010 57.23% E D D 2.40 4.21 81.91% -0.013 44.25% F F E FF S = 45 m ph 1.20 1.67 100.00% -0.001 88.27% B A A 1.40 2.23 99.89% -0.003 84.92% B A A 1.60 2.72 98.93% -0.005 80.15% B B A 1.80 3.16 96.51% -0.007 72.91% C C B 2.00 3.54 92.67% -0.009 62.57% D D C 2.20 3.89 87.70% -0.012 49.53% F E E 2.40 4.21 81.91% -0.014 35.60% F F F FF S = 35 m ph 1.20 1.67 100.00% -0.002 87.40% B A A 1.40 2.23 99.89% -0.004 82.72% B B A 1.60 2.72 98.93% -0.006 75.99% C B B 1.80 3.16 96.51% -0.008 66.04% D C C 2.00 3.54 92.67% -0.011 52.68% E E D 2.20 3.89 87.70% -0.014 37.60% F F F 2.40 4.21 81.91% -0.017 23.83% F F F Notes: TTI = travel time index, the ratio of the free-flow speed to the actual speed; POTA = probability on-time arrival; %TLOS = truck LOS index as a percentage of ideal conditions; and FFS = free-flow speed. Exhibit 119. Truck TTI level of service look-up table.

170 Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual estimated probability of on-time arrival (POTA), the estimated utility for trucks, and the %TLOS index. The LOS letter is then read directly from the table for the appropriate freight facility class. 5. References Dowling, R., G. List, B. Yang, E. Witzke, and A. Flannery. NCFRP Report 31: Incorporating Truck Analysis into the Highway Capacity Manual. Transportation Research Board of the National Academies, Washington, D.C., 2014. Federal Register. Establishment of the National Freight Network. 78 FR 8686, Feb. 6, 2013.