Guide to Pedestrian Analysis (2022)

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.

54 Evaluating pedestrian operations is an important part of designing and operating pedestrian facilities, particularly in high-activity areas such as downtowns and shopping districts, but also in areas that experience intermittent, intense levels of activity, such as around sports venues, audito- riums, transit facilities, and cruise ship terminals. Operational performance measures that different travel modes share in common, such as delay, can be used to compare the effects of proposed projects on individual modes, including pedestrians. Even in areas with relatively low levels of pedestrian activity, operational measures can be useful as proxies for potentially unsafe behavior (e.g., crossing against a red light if the delay experienced is too high, walking in the street if the sidewalk is too narrow) (1). Chapter 4 presents concepts and methods for evaluating pedestrian flow and storage needs along and across roadways and within transit facilities. These methods are particularly applicable to designing pedestrian facilities that function adequately for the number of pedestrians expected to use them. Some methods also support quality-of-service performance measures described in Chapter 5. Chapter 4 presents the pedestrian delay, flow, speed, storage, and circulation area concepts that form the basis for conducting pedestrian operations analysis. Although pedestrian operations performance measures are, in general, similar to operations measures used for the motor vehicle mode, pedestrian flow is different from vehicular flow (2). Compared with motor vehicle traffic, pedestrians move in a less organized manner, at higher densities, and in more complex and constrained spaces (3). Pedestrians’ speed depends on their surroundings, including the pedestrians immediately around them, the pedestrians ahead of them, adjacent land uses, and environmental conditions. Depending on an area’s land use and personal security characteristics, pedestrians may try to access crowded areas or avoid them (4). This chapter focuses on the macroscopic analysis of outdoor pedestrian flow; that is, the analysis of average conditions for groups of pedestrians within a defined area (e.g., a side- walk, a crosswalk, a bus stop) over relatively long periods of time (several minutes to an hour). The analysis of indoor facilities (e.g., transit stations, airport terminals) is typically more complex, in that it involves multiple interacting pedestrian flows, queueing areas (e.g., fare gates, security lines), elevation changes, and shorter analysis periods. Those applications generally require microscopic simulation, in which each person is modeled individually, or mesoscopic analysis, which divides the study area into small cells, assigns groups of individuals to each cell, and then transfers individuals between cells at each small time-step (4). Microscopic pedestrian simulation may also be incorporated into a broader simulation of the street environment in which the interactions of multiple travel modes are modeled. More information about micro- scopic and mesoscopic modeling can be found in Teknomo (5), Liu et al. (6), and Tordeux et al. (7). C H A P T E R 4 Pedestrian Operations Analysis

Pedestrian Operations Analysis 55   Pedestrian Delay Pedestrian delay is defined as the difference between ideal walking time and actual walking time at a location. Delay is most typically evaluated for pedestrian crossings, where it represents the average time a pedestrian waits for a legal opportunity to enter a crosswalk at a signalized intersection or the average time to receive an adequate gap in traffic at an unsignalized crossing (8). However, delay can also be evaluated along a pedestrian facility, for example, by comparing walking time on a crowded sidewalk with walking time on a near-empty sidewalk. Pedestrian delay can also be used as a component of overall intersection person delay, which considers the average delay of all modes using an intersection, weighted by the number of people using each mode (8). Large amounts of pedestrian delay have the potential to discourage pedestrian trips and to increase pedestrian noncompliance with signals when crossing an intersection (8). The HCM 6th ed. (8) uses pedestrian delay directly to determine pedestrian LOS at uncontrolled cross- ings (which can be marked or unmarked crosswalks) and as one of the inputs for determining pedestrian LOS at signalized crosswalks. Field Measurement Pedestrian delay data can be obtained at both signalized or unsignalized intersections. The data can be measured in real time by field personnel or discretely recorded by high-definition, field-mounted video cameras. This type of video data collection is advantageous, as it is rela- tively inconspicuous and less likely to alter pedestrians’ natural behavior. Video data collection is accurate, because recordings can be observed frame by frame during data extraction. Video data should be collected during daylight hours and in relatively clear weather conditions. The cameras may be set up such that the entire crosswalk is visible, along with the face of the pedestrian signals (for signalized intersections) for that crosswalk. Pedestrian delay and crossing time can be measured by a stopwatch to track intervals of the pedestrian’s entire crossing experience. For each pedestrian or groups of pedestrians traveling together, the following times can be recorded: • PT1: Arrival time is defined as the moment the pedestrian pushes the button (if present). If the pedestrian does not push the button or there is no button present, the time the pedestrian arrives at the curb and prepares to cross is recorded. In the case of a group, one person in the group—such as the person who pressed the button or arrived first—should consistently be chosen and that person’s crossing time coded. • PT2: Time when the pedestrian’s first foot steps off the curb. • PT3: Time when the pedestrian arrives at the median (if present). Recorded as when the pedestrian’s first foot crosses the first line of the median. • PT4: Time when the pedestrian’s first foot steps off the median. • PT5: Time when the pedestrian steps onto the curb on the opposite side of the crossing. Pedestrian delay at the start of the crossing is the difference between PT1 and PT2. Pedestrian delay in the median is the difference between PT4 and PT3. Note that this value also includes pedestrian travel time through the median. Pedestrian crossing time is the difference between PT5 and PT2. A pedestrian can be delayed during crossing for multiple reasons. It is useful to record the following types of delays to help inform potential treatment selection if deemed necessary: • Delay due to motorist behavior: Pedestrians are delayed due to motorist behavior if their speed or trajectory is changed as a result of a motorist’s actions, such as waiting for motorists to pass or yield. If a motorist did not have time to stop when the pedestrian arrived, the pedestrian is not considered to be delayed by motorist behavior.

56 Guide to Pedestrian Analysis • Delay due to pedestrian not pushing the button (if present): This applies to signalized cross- walks, where a “Walk” phase is not provided because the pedestrian did not push the button. Failure to press a button can also be recorded at unsignalized crosswalks with pedestrian- activated safety countermeasures (e.g., RRFBs, in-pavement flashing lights); however, delay should still be recorded, because pedestrians have the right-of-way, even if they do not activate the crosswalk visibility treatment. • Delay when pedestrian motions driver through. • No delay. • Unable to tell. Delay Estimation Uncontrolled Crossings Chapter 20 of the sixth edition of the Highway Capacity Manual (HCM 6th ed.) (8) provides a method for estimating pedestrian delay at uncontrolled crossings of the major street at a two-way stop-controlled intersection or at a midblock location. However, the portion of the methodology that accounts for the delay-reducing effects of motorist yielding produced illogical results in certain circumstances and has been revised to correct these issues (9). Details of the revised method are presented in Appendix A of this guide, and a computational engine for implementing the method is described in Appendix B. Estimation of delay involves the following steps: • Step 1. Identify two-stage crossings. Crossings that include a median refuge island allow pedes- trians to perform two short crossings (crossing one direction of traffic at a time) instead of performing one long crossing, thus reducing pedestrian delay. • Step 2. Determine critical headway. The critical headway is the minimum amount of time an average pedestrian needs to cross the street with no conflicting traffic (when motorists do not yield). • Step 3. Estimate probability of a delayed crossing. The greater the traffic volume, the greater the chance that one or more vehicles will arrive at the crosswalk within the time defined by the critical headway, resulting in delay to an arriving pedestrian unless motorists yield. • Step 4. Calculate average delay to wait for an adequate gap. This step calculates the average pedestrian delay in the event motorists do not yield, which forces the pedestrian to wait for a gap in traffic exceeding the critical headway. • Step 5. Estimate average pedestrian delay for the crossing stage. This step calculates the reduced delay to motorist yielding on the basis of an analyst-provided average yielding rate for the crossing. Average yielding rates associated with a number of crossing treatments are given in Table 3-4. The average delay for the crossing stage is the weighted average of the delay when (a) motorists yield and (b) pedestrians must wait for an adequate gap. If the crossing involves more than one crossing stage, the analysis returns to Step 2 to calculate the delay for the next crossing stage. • Step 6. Calculate average pedestrian delay. This step totals the average delay for each cross- ing stage to produce an average delay for the crossing as a whole. Input data required to estimate delay consist of the following (for items marked with an asterisk, local or national default values can be substituted): • Crosswalk length (ft)—measured curb to curb; • Number of through lanes crossed; • Presence of two-stage crossings (yes/no); • Conflicting vehicular flow rate (vehicles per second, veh/s); • Pedestrian speed (ft/s);* • Pedestrian start-up and end clearance time (s)* (Start-up time is the average time between when the vehicle forming the start of the critical headway clears the crosswalk and the first

Pedestrian Operations Analysis 57   waiting pedestrian enters the crosswalk; end clearance time is extra time buffer that pedes- trians provide themselves prior to the arrival of the vehicle forming the end of the critical headway.); and • Average motorist yielding rate (%).* If pedestrian demand at the crossing is high enough that pedestrians arrange themselves into rows when crossing, the following additional input data are required: • Crosswalk width (ft) and • Pedestrian demand (pedestrians per second, p/s). The choice of pedestrian walking speed depends on the purpose of the analysis: • For estimating average pedestrian delay or pedestrian LOS under existing conditions, a locally measured average walking speed for uncontrolled crossings is recommended. In the absence of local data, research has found an average pedestrian speed of 4.7 ft/s at uncontrolled crossings (9–11). • For planning and design purposes—for example, to assess the adequacy of the crossing to accommodate pedestrians of all abilities—a walking speed of 3.5 ft/s, representative of a 15th-percentile pedestrian, would be appropriate. The choice of pedestrian start-up and end clearance time also depends on the purpose of the analysis. The research that developed the revised HCM method (9) found that field-measured values of average delay best matched the estimated delay when the pedestrian start-up time and end clearance time were 0 s. This value implies that pedestrians anticipated the arrival of an adequate gap (i.e., they did not require any start-up time) and started immediately upon its arrival. It also implies that pedestrians did not require any end clearance time. However, it is more likely that the vehicles defining the start and end of the adequate gap are often not traveling in the first and last lanes, respectively, crossed by the pedestrian (hence, crossing safety is assured spatially by lane separation rather than temporally by a second or two of clearance time). The use of a local value for start-up and end clearance time is encouraged when one is available. For design purposes, a value of 3.0 s provides a more conservative estimate of start-up and end clearance time. Signalized Crossings Chapter 19 of the HCM 6th ed. (8) provides a method for estimating pedestrian delay at signalized pedestrian crossings. The method accurately predicts delay for pedestrians arriving randomly for a one-stage crossing but does not account for the fact that pedestrians generally arrive at the second stage of a two-stage crossing (or at the corner) to cross a second leg of the intersection as a group. In the latter two cases, delay for the subsequent crossing depends on the traffic signal timing (9). Appendix A provides updated methods for calculating delay in any of these three crossing situations, while Appendix B describes a computational engine that imple- ments these methods. One-Stage Crossing Delay. Average pedestrian delay for a one-stage crossing is given by the following equation (8): ( )= − 2 Walk 2 d C g C p where dp = average pedestrian delay (seconds per pedestrian, s/p), C = length of traffic signal cycle (s), and gWalk = effective walk time (s)—the amount of time pedestrians have to begin crossing.

58 Guide to Pedestrian Analysis Two-Stage Crossing Delay. The two-stage crossing delay procedure begins by calculating the average delay for the first stage, as described above for a one-stage crossing. The delay in waiting to start the second stage depends on how long an average pedestrian requires to com- plete the first-stage crossing, along with the start time of the “Walk” phase for the second stage, relative to the start time of the “Walk” phase for the first stage. Depending on signal timing, any given pedestrian may be able to complete the crossing without waiting in the median or may need to wait for the second-stage “Walk” phase to begin. The delay associated with the second- stage crossing will generally vary by crossing direction. In addition to the data required to esti- mate delay for a one-stage crossing, the following additional data are required (local or national default values can be substituted for items marked with an asterisk): • First-stage crossing length (ft), • Pedestrian speed (ft/s)*, • Effective walk time for the second stage (s), and • Start time of the “Walk” phase for the first and second stages. If the signal is not fixed timed (i.e., is actuated or semiactuated, meaning that the start times of certain signal phases vary in response to the presence of vehicles or pedestrians), additional details about the signal timing will be required, as described in Appendix A. As with uncon- trolled crossings, the choice of pedestrian speed will depend on the purpose of the analysis. When average pedestrian delay is being estimated, an average pedestrian walking speed should be used. When operations for persons with slower walking speeds (e.g., seniors, small children, persons with disabilities) are being evaluated, a speed representative of a 15th-percentile pedes- trian (3.5 ft/s) would be appropriate. Depending on the start time of the second-stage crossing relative to the first-stage crossing, the calculated delay for a slower pedestrian could be less than that for an average pedestrian (a slower pedestrian is still making the first-stage crossing while an average pedestrian has arrived in the median and is waiting to start the second-stage crossing). The slower pedestrian’s delay could also be greater than the average person’s delay (an average pedestrian can arrive in time to start the second-stage crossing during the same cycle, while a slower pedestrian has to wait one cycle length to begin the second-stage crossing). Diagonal Crossing Delay. The procedure for estimating delay for pedestrians to cross two legs of an intersection (i.e., to cross to a diagonally opposite corner at a four-leg intersection using two crosswalks) is similar to the procedure for a two-stage crossing, in that pedestrians are assumed to arrive randomly when crossing the first leg but arrive at the corner to cross the second leg at a predictable time. Other Situations. NCHRP Web-Only Document 312 describes extensions of the above methods to address the following situations (9): • Crosswalk closures, • Exclusive pedestrian phases (e.g., Barnes Dance), and • Signal timing at crosswalks with traffic signals or pedestrian hybrid beacons (HAWK signals) that attempts to serve pedestrians as soon as possible after the pedestrian pushes the button. These extensions are based in theory and at the time of writing had not yet been validated through field data collection or simulation. Pedestrian Flow The dimensions of pedestrian facilities greatly influence pedestrian flow, or the number of pedes- trians a facility can serve in a given period of time. Pedestrian speed, density (the number of pedestrians per unit area), and space (the average area available per pedestrian and the inverse

Pedestrian Operations Analysis 59   of density) are all related to pedestrian demand. Understanding these relationships is impor- tant for planning and designing pedestrian circulation facilities such as sidewalks, crosswalks, underpasses, overpasses, ramps, and stairways. ey are also important in designing pedestrian waiting and queuing areas such as street corners, bus stops, transit station platforms, transit station fare gates, and security checkpoints. Relationships Between Speed, Flow, and Density Figure 4-1 illustrates fundamental relationships between speed, ow, and density observed in various studies (12). Although individual studies vary slightly in their depiction of these relationships, the overall pattern is apparent. At low pedestrian volumes, pedestrians can move at their desired speed without hindrance from other pedestrians. As pedestrian demand increases, it becomes more dicult for pedes- trians to maintain their desired course and speed; thus, average speed decreases. At a certain density—typically around 0.2 to 0.25 pedestrians per square foot (p/2) (4 to 5 2/p)—ow is maximized (13−18). is density represents the facility’s capacity; however, pedestrian circu- lation facilities are usually not designed for capacity because high levels of pedestrian crowding occur, resulting in low speeds, lack of accommodation for persons with disabilities, and poten- tially unsafe conditions (19). Sidewalks reach their functional capacity at levels below their theoretical capacity, as faster-moving pedestrians spill out beyond the designated walking area (e.g., onto lawns, into the street) to get around slower-moving pedestrians (8). When demand exceeds capacity in a conned space (e.g., a hallway, an overpass), pedestrian speeds drop to a slow shue and queuing develops. Source: Daamen et al. (12 ). Figure 4-1. Diagrams of fundamental relationships in pedestrian ow characteristics.

60 Guide to Pedestrian Analysis Design Applications e design of pedestrian circulation facilities (e.g., sidewalks) typically involves nding the minimum facility width to accommodate a certain number of pedestrians per minute at a desired QOS (described in terms of speed, density, space, or an LOS based on one of these characteristics). Later sections in this chapter describe design levels suggested in the literature. An important concept when a pedestrian facility is being designed or evaluated is eective width—the portion of the facility’s physical width usable for pedestrian circulation. In a sidewalk context, the eective width may be reduced by landscaping features (e.g., street trees), street furniture (e.g., benches, art), roadway infrastructure (e.g., light poles), transit stops (e.g., shelters, people queued waiting for a bus), pedestrians’ shy distance from curbs and walls, commer- cial uses (e.g., café tables), and people window-shopping. Each feature reduces the sidewalk’s eective width, with the narrowest eective portion of the sidewalk forming the bottleneck constraining the pedestrian ow. Figure 4-2 illustrates the concept of eective width. Figure 4-3 shows an extreme example of a wide sidewalk whose eective width is reduced to single-le operation due to the presence of café tables, street trees, and bicycle parking, with pedestrians choosing to walk in the bicycle facility instead. Platoon Flow Platoon ow is dened as the grouping or bunching of pedestrians caused by internal or external trac impedances (20). For example, a group of pedestrians crossing a signalized crosswalk together will typically continue down the street as a platoon. Platoons are particularly evident in corridors because limited walkway width and the presence of walls inhibits passing oppor- tunities. Platoons are characterized by increasing behavioral consistencies manifested in the adoption of a prevalent group speed and positioning arrangements (20). In general, a greater amount of space per pedestrian (i.e., lower density) is required under platoon ow to maintain the same level of freedom of movement possible under random pedestrian ow (8). Cross-Flow Cross-ow is dened as pedestrian ow that is roughly perpendicular to and crosses another pedestrian stream. Pedestrian cross-ows are common in major activity centers and in special Source: HCM 6th ed. (8 ). Figure 4-2. Sidewalk effective width.

Pedestrian Operations Analysis 61   event transportation systems such as universities, transit stations, art galleries, museums, and places of entertainment (21). Pedestrian cross-ows are also common in corridors, passageways, hallways, and on sidewalks near building exits. Cross-ows disrupt pedestrian streams by causing pedestrians to change course, break their normal walking gait, or even stop. Cross-ow begins to aect pedestrian streams when average pedestrian space drops below 35 2/p; below 15 2/p, nearly every crossing movement is aected (13, 21). e facility’s capacity is reached earlier (i.e., at a lower density or greater average pedes- trian space) when cross-ows exist (8). Pedestrian Speed Pedestrian walking speed is an important parameter for designing pedestrian facilities, and the choice of walking speed used in an analysis will depend on the specic application being analyzed. For signal timing applications, the Manual on Uniform Trac Control Devices (MUTCD) suggests timing pedestrian phases on the basis of a pedestrian walking speed of 3.5 /s, except when provisions are made for automatically or manually extending the pedes- trian phase to accommodate slower pedestrians (22). A walking speed of 3.5 /s is equivalent to the 15th-percentile speed of pedestrians as a whole (23), but elderly pedestrians have lower 15th-percentile speeds in the range of 3.0 to 3.1 /s (23−25). For design purposes, or when the ability of pedestrian facilities to accommodate pedestrians with a variety of abilities is being evalu- ated, the use of a walking speed toward the lower end of the spectrum would be appropriate. When the purpose of an analysis is to estimate actual eld conditions or to compare average pedestrian delay with the average delay experienced by other travel modes (e.g., as part of a person delay analysis), it is appropriate to use average values of pedestrian speed representative of pedestrians using a given facility. For example, the HCM suggests using an average speed of 4.0 /s for evaluating sidewalks and signalized crossings, but 3.3 /s if more than 20% of pedes- trians are elderly (8). e average speed of pedestrians using uncontrolled crossings is higher than that of pedestrians walking on sidewalks. e HCM suggests an average speed of 4.4 /s (8), while average speeds of 4.7 /s were observed in the eld when the update to the HCM’s method for uncontrolled crossing delay was being validated (see Appendix A) (9). Whenever possible, local values for average pedestrian speed and other analysis inputs are preferred to national default values. Source: Kittelson & Associates, Inc./Paul Ryus. Figure 4-3. Comparison of physical and effective sidewalk widths.

62 Guide to Pedestrian Analysis Pedestrian Space Pedestrian space is the inverse of pedestrian density. Although pedestrian flow theory typi- cally uses units of density (e.g., persons per square foot), in practice, it is usually more conve- nient to work with units of space (e.g., square feet per pedestrian). Both the HCM (8) and the Transit Capacity and Quality of Service Manual (TCQSM) (19) methods for evaluating pedes- trian circulation and storage facilities work with units of pedestrian space. Pedestrian Circulation Area Analysis This section describes methods for evaluating the operation of pedestrian circulation areas such as sidewalks, stairways, and crosswalks to ensure sufficient width is provided to serve pedestrian demand. Methods for evaluating pedestrian satisfaction (i.e., QOS) with pedestrian circulation facilities are presented in Chapter 5. The methods described here assume a linear relationship between pedestrian flow and speed, where flow (in terms of pedestrians per minute per foot of facility width) is the product of speed and density. Because pedestrian space is the inverse of density, this relationship can also be expressed as (19) ( ) ( ) ( ) =pedestrian flow p ft min pedestrian speed ft min pedestrian space ft p2 Tables in the subsections below provide maximum pedestrian flow rates for a given LOS for various types of pedestrian facilities. Given a known or forecasted pedestrian demand and the use of a maximum pedestrian flow rate for design, the required effective width for the facility is ( ) ( ) ( ) =effective width ft pedestrian demand p min design pedestrian flow p ft min For example, assume pedestrian demand is forecasted to be 50 p/min, and a sidewalk with platoon flow is being designed to accommodate LOS C. The maximum pedestrian flow at LOS C is 6 p/ft/min (see Table 4-1 in the subsection on sidewalks below). The minimum effective width to accommodate this demand at LOS C is then (50/6) = 8.3 ft. Additional width (e.g., shy distance to the curb, space required for street trees) would then be added as appropriate to arrive at the required physical sidewalk width. Average Flow Platoon Flow LOS Space (ft2/p) Flow Rate (p/ft/min) Space (ft2/p) Flow Rate (p/ft/min) Description A >60 ≤5 >530 ≤0.5 Ability to move in desired path; no need to alter movements B >40−60 >5−7 >90−530 >0.5−3 Occasional need to adjust path to avoid conflicts C >24−40 >7−10 >40−90 >3−6 Frequent need to adjust path to avoid conflicts D >15−24 >10−15 >23−40 >6−11 Speed and ability to pass slower pedestrians restricted E >8*−15 >15−23 >11*−23 >11−18 Speed restricted; very limited ability to pass slower pedestrians F ≤8* Variable ≤11* Variable Speed severely restricted; frequent contact with other users *13 ft2/min is the minimum value for locations with significant cross-flows. Source: HCM 2000 (26). Table 4-1. Level-of-service criteria for sidewalks and walkways.

Pedestrian Operations Analysis 63   On the other hand, if the analysis seeks to evaluate the operation of an existing sidewalk, the analyst measures actual pedestrian demand on the minimum effective width of each sidewalk segment being analyzed. For example, if the minimum effective width is 4 ft and pedestrian demand is 25 p/min, the flow rate is (25/4) = 6.25 p/ft/min. Table 4-1 shows that this flow rate corresponds to LOS B for average flow and LOS D for platoon flow. Sidewalks and Walkways Table 4-1 provides pedestrian space and flow rates for different levels of service in both average-flow and platoon-flow conditions. The values in Table 4-1 apply to bidirectional flow on sidewalks. The two directions of flow tend to use sidewalk width in proportion to their respective volumes. The exception is with highly directional flow (e.g., 90/10), where low-volume opposing flow uses a greater proportion of the sidewalk width (i.e., the width of a person); reductions in sidewalk capacity of up to 15% have been observed in these situations (8). In situations where pedestrians are not constrained by barriers to remain on the walkway surface, faster-moving pedestrians will often choose to walk outside the walkway (e.g., in the sidewalk furniture zone, in the street, on the lawn) at flow rates below the maximum flow rate shown in Table 4-1. Stairways Table 4-2 provides pedestrian space and flow rates for stairways in both outdoor and transit station conditions. In contrast to sidewalks, stairways are more severely affected by low-volume, reverse-direction flows. Pedestrians tend to form lanes when using stairways, and stairway capacity tends to increase stepwise according to the number of lanes available rather than in proportion to the stairway width. A typical lane width for design is 30 inches, but other values are possible, as indicated in Table 4-3 (19). When a design stairway width is being identified, it is recommended that the calculated effec- tive width be rounded up to the next integer increment of lane width and that one additional Outdoors Transit Stations LOS Space (ft2/p) Flow Rate (p/ft/min) Space (ft2/p) Flow Rate (p/ft/min) Description A >20 ≤5 ≥20 ≤5 Sufficient area to freely select speed and to pass slower- moving pedestrians. Reverse flows cause limited conflicts. B >17−20 >5−6 15−20 5−7 Sufficient area to freely select speed with some difficulty in passing slower-moving pedestrians. Reverse flows cause minor conflicts. C >12−17 >6−8 10−15 7−10 Speeds slightly restricted due to inability to pass slower- moving pedestrians. Reverse flows cause some conflicts. D >8−12 >8−11 7−10 10−13 Speeds restricted due to inability to pass slower-moving pedestrians. Reverse flows cause significant conflicts. E >5−8 >11−15 4−7 13−17 Speeds of all pedestrians reduced. Intermittent stoppages likely to occur. Reverse flows cause serious conflicts. F ≤5 Variable ≤4 Variable Complete breakdown in pedestrian flow with many stoppages. Forward progress dependent on slowest-moving pedestrians. Sources: Fruin (13), HCM 2000 (26), TCQSM (19). Table 4-2. Level-of-service criteria for stairways.

64 Guide to Pedestrian Analysis lane of width be provided in situations with highly unbalanced directional flows to accommodate reverse-direction flows. The physical width of the stairway should account for handrail needs. In transit station applications, emergency evacuation needs may dictate stairway size, and the current version of the NFPA 130 Standard for Fixed Guideway Transit and Passenger Rail Systems should be consulted (27). The TCQSM (19) also provides analysis methods for other vertical circulation elements found within transit stations, such as escalators and elevators. Ramps and Grades Ramps that are compliant with the requirements of the Americans with Disabilities Act do not affect pedestrian speeds. Grades of 10% or more reduce average pedestrian speeds by 0.3 ft/s (8). Crosswalks The crosswalk circulation area measures the average space available within a crosswalk while pedestrians are using it. This measure can be used to evaluate existing conditions by comparing the calculated circulation area to the LOS values given in Table 4-1. It can also be used in a design application to identify the crosswalk width required to serve a particular pedestrian demand at a desired LOS. Chapter 19 of the HCM (8) provides a method for determining the crosswalk circulation area. The data required to apply this method are as follows: • Crosswalk length; • Crosswalk effective width—typically the physical width of the crosswalk—although the analyst may reduce this if “vehicles are observed to encroach regularly into the crosswalk area or when an obstruction in the median (e.g., a signal pole or reduced-width cut in the median curb) narrows the walking space” (8, p. 19-75); • Average directional pedestrian volumes per traffic signal cycle; • Length of the traffic signal cycle; • Timing information for the pedestrian phase; • Average pedestrian speed; and • Traffic volumes conflicting with pedestrian movements: permitted left turn, right turn, and right turn on red. Research conducted by the New York City DOT found that pedestrian speeds in signalized crosswalks can be reduced by as much as 1.0 ft/s when significant opposing pedestrian volume is present, but the current HCM procedure does not account for this effect (28). The New York City DOT has also found that pedestrian speeds in crosswalks vary by pedestrian gender, pedes- trian signal display (“Walk,” flashing “Don’t Walk,” or “Don’t Walk”), time of day (a.m. or p.m.), and whether the pedestrian is alone or in a group (29). Lane Width (in.) Approximate Capacity (p/min/lane) Comments 21–27 30 Notable friction, not recommended for daily use 28–30 38 Recommended for general use 31–33 42 Provides extra space and slightly greater capacity ≥34 Little or no additional capacity May be beneficial where pedestrians carry items Source: TCQSM (19). Table 4-3. Stairway capacity by lane width.

Pedestrian Operations Analysis 65   Transit Station Corridors Table 4-4 provides LOS criteria for corridors within transit stations. Pedestrian Storage Area Analysis This section describes methods for evaluating pedestrian storage areas, such as transit plat- forms and signalized intersection corners, to ensure adequate space is available to store waiting pedestrians. Methods for evaluating pedestrian satisfaction (i.e., QOS) with pedestrian storage facilities are presented in Chapter 5. Transit Station Platforms The process for identifying the amount of space needed for transit platform storage is described in the TCQSM (19). The basic process is similar to that for pedestrian circulation areas: identify the passenger demand to be accommodated, identify the minimum space per pedestrian at a desired LOS by using Table 4-5 (LOS C or D is typical), and multiply the two values together to identify the space required for storing passengers waiting for their transit vehicle. Additional platform area is required to accommodate the circulation needs of passengers arriving on transit vehicles (including potential queueing at exit points from the platform, such as escalators), to provide a shy distance to the platform edge and to account for area taken up by displays, fare collection infrastructure, and so forth. This method can also be applied to identify sidewalk space used by passengers waiting for buses. Passenger demand estimates typically account for some degree of service irregularity, which results in more passengers waiting than normal (19). LOS Space (ft2/p) Flow Rate (p/ft/min) A ≥35 0–7 B 25–35 7–10 C 15–25 10–15 D 10–15 15–20 E 5–10 20–25 F <5 Variable Source: TCQSM (19). Table 4-4. Level-of-service criteria for transit station corridors. LOS Space (ft2/p) Interperson Spacing (ft) A ≥13 ≥4.0 B 10–13 3.5–4.0 C 7–10 3.0–3.5 D 3–7 2.0–3.0 E 2–3 <2.0 F < 2 Variable Source: TCQSM (19). Table 4-5. Level-of-service criteria for transit station platforms.

66 Guide to Pedestrian Analysis Corner Circulation Area The corner circulation area measures the average space available for pedestrians at a street corner, some of whom are waiting to cross a street, some of whom may be actively crossing the other street, and some of whom may be turning the corner as they walk from one street to the other. Although this measure could be used to evaluate existing conditions, it works well for design applications by identifying the space required to serve a particular pedestrian demand at a desired LOS, based on Table 4-1. Chapter 19 of the HCM (8) provides a method for determining the corner circulation area. The data required to apply this method are as follows: • Widths of both sidewalks approaching the corner, • Corner radius, • Directional pedestrian volumes on both crosswalks, • Length of the traffic signal cycle, and • Timing information for the pedestrian phases for both crosswalks. Note that the LOS thresholds are based on moving pedestrians, although corners serve both standing and moving pedestrians, and therefore may overestimate the area required to provide a given QOS. An alternative approach would be to divide the corner into waiting and circulation areas, apply appropriate LOS thresholds to each area, and evaluate conditions at different points during the traffic signal cycle. Traffic Signal Warrants Pedestrian signals are a potential countermeasure for improving the safety and pedestrian opera- tion of midblock pedestrian crossings. The MUTCD (22) defines a pedestrian volume traffic signal warrant to determine when a signal may be justified. Meeting the warrant does not by itself justify installing a signal; other factors should also be considered (22). The warrant is intended for application at locations where the traffic volume on a major street is so heavy that pedestrians experience excessive delay in crossing the major street. The warrant is based on pedestrian and motor vehicle volumes and motor vehicle speed. If the 15th-percentile pedestrian speed at the location is less than 3.5 ft/s, the pedestrian volumes needed to satisfy the warrant may be reduced. The warrant can be satisfied in one of two ways: (a) when pedestrian volumes in 1 hour (four consecutive 15-minute periods) exceed the values given in the warrant’s peak-hour graph or (b) when the pedestrian volumes in any 4 hours of an average delay exceed the values given in the warrant’s 4-hour graph (22). Some jurisdictions apply pedestrian warrants based on forecasted pedestrian demand rather than existing demand, similar to the way fore- casted traffic demand from a future development is often used to justify installing a traffic signal. The MUTCD also provides a warrant for signalizing a school crossing. This warrant requires analyzing the number of gaps in the traffic stream. The warrant is met when (a) at least 20 school children use the crossing during the highest-volume hour and (b) the number of adequate gaps in the traffic stream during the period when school children use the crossing is less than the number of minutes in that period (22). Summary This chapter presents concepts and methods for evaluating pedestrian flow and storage needs along and across roadways and within transit facilities. Compared to motor vehicle traffic, pedes- trians move in a less-organized manner, at higher densities, and in more complex, constrained spaces. The methods described in this chapter are particularly applicable to designing pedestrian

Pedestrian Operations Analysis 67   facilities that function adequately for the number of pedestrians expected to use them. Some of these methods also support QOS performance measures described in Chapter 5. This chapter presents pedestrian delay, flow, speed, storage, and circulation area concepts that form the basis for conducting pedestrian operations analysis. Appendix A presents methods for estimating pedestrian delay at uncontrolled and signalized pedestrian crossings that update the methods in the HCM 6th ed. References 1. Kennedy, J. F., M. Crabtree, J. Castle, J. Martin, and M. Elliott. 2009. The Effect of Traffic Signal Strategies on the Safety of Pedestrians. Report PPR414. TRL Limited, Wokingham, Berkshire, UK. 2. Tipakornkiat, C., H. Kim, and T. Limanond. 2011. Method to Estimate the Speed and Density of the Pedes- trian Sidewalk. Journal of the Eastern Asia Society for Transportation Studies, Vol. 9, pp. 1100–1114. 3. Ismail, K., T. Sayed, and N. Saunier. 2009. Automated Pedestrian Safety Analysis Using Video Data in the Context of Scramble Phase Intersections. Presented at Annual Conference of the Transportation Association of Canada, Vancouver, British Columbia. 4. Bauer, D., S. Seer, and N. Brändle. 2007. Macroscopic Pedestrian Flow Simulation for Designing Crowd Control Measures in Public Transport after Special Events. Proceedings of the 2007 Summer Computer Simu- lation Conferences, San Diego, CA, July 16–19, 2007, pp. 1035–1042. 5. Teknomo, K. 2006. Application of Microscopic Pedestrian Simulation Model. Transportation Research Part F: Traffic Psychology and Behaviour, Vol. 9, pp. 15–27. 6. Liu, M., W. Zeng, P. Chen, and X. Wu. 2017. A Microscopic Simulation Model for Pedestrian–Pedestrian and Pedestrian–Vehicle Interactions at Crosswalks. PLoS One, Vol. 12. https://doi.org/10.1371/journal. pone.0180992 (as of January 6, 2020). 7. Tordeux, A., G. Lammel, F. S. Hanseler, and B. Steffen. 2018. A Mesoscopic Model for Large-Scale Simula- tion of Pedestrian Dynamics. Transportation Research Part C: Emerging Technologies, Vol. 93, pp. 128–147. 8. Transportation Research Board. 2016. Highway Capacity Manual: A Guide for Multimodal Mobility Analysis, 6th ed. Washington, DC. 9. Ryus, P., A. Musunuru, K. Lausten, J. Bonneson, S. Kothuri, C. Monsere, N. McNeil, K. Nordback, S. LaJeunesse, W. Kumfer, L. Thomas, and S. I. Guler. 2022. NCHRP Web-Only Document 312: Enhancing Pedestrian Volume Estimation and Developing HCM Pedestrian Methodologies for Safe and Sustainable Communities. Transportation Research Board, Washington, DC. 10. Fitzpatrick, K., S. M. Turner, M. Brewer, P. J. Carlson, B. Ullman, N. D. Trout, E. S. Park, J. Whitacre, N. Lalani, and D. Lord. 2006. TCRP Report 112/NCHRP Report 562: Improving Pedestrian Safety at Unsignalized Crossings. Transportation Research Board, Washington, DC. 11. Rodegerdts, L., M. Blogg, E. Wemple, E. Myers, M. Kyte, M. P. Dixon, G. F. List, A. Flannery, R. Troutbeck, W. Brilon, N. Wu, B. N. Persaud, C. Lyon, D. L. Harkey, and D. Carter. 2007. NCHRP Report 572: Round- abouts in the United States. Transportation Research Board, Washington, DC. 12. Daamen, W., S. Hoogendoorn, and P. H. L. Bovy. 2005. First-Order Pedestrian Traffic Flow Theory. Transportation Research Record: Journal of the Transportation Research Board, No. 1934, pp. 43–52. 13. Fruin, J. J. Pedestrian Planning and Design. 1971. Metropolitan Association of Urban Designers and Environ- mental Planners, New York. 14. Pushkarev, B., and J. M. Zupan. 1975. Urban Space for Pedestrians: A Report of the Regional Plan Association. Massachusetts Institute of Technology Press, Cambridge. 15. Transportation Research Board. 1985. Special Report 209: Highway Capacity Manual. Washington, DC. 16. Tanaboriboon, Y., S. S. Hwa, and C. H. Chor. 1986. Pedestrian Characteristics Study in Singapore. Journal of Transportation Engineering, Vol. 112, pp. 229–235. 17. Tanaboriboon, Y., and J. Guyano. 1989. Level-of-Service Standards for Pedestrian Facilities in Bangkok: A Case Study. ITE Journal, pp. 39–41. 18. May, A. D. 1990. Traffic Flow Fundamentals. Prentice Hall, Englewood Cliffs, NJ. 19. Kittelson & Associates, Inc.; Parsons Brinckerhoff; KFH Group, Inc.; Texas A&M Transportation Institute; and Arup. 2013. TCRP Report 165: Transit Capacity and Quality of Service Manual, 3rd ed. Transportation Research Board, Washington, DC. 20. Davis, D. G., and J. P. Braaksma. 1987. Level-of-Service Standards for Platooning Pedestrians in Transportation Terminals. ITE Journal, Vol. 57, No. 4, pp. 31–35. 21. Khisty, C. J. 1985. Pedestrian Cross Flow Characteristics and Performance. Environment and Behavior, Vol. 17, No. 6, pp. 679–695.

68 Guide to Pedestrian Analysis 22. Federal Highway Administration. 2009. Manual on Uniform Traffic Control Devices. U.S. Department of Transportation, Washington, DC. http://mutcd.fhwa.dot.gov/. 23. LaPlante, J. N., and T. P. Kaeser. 2004. The Continuing Evolution of Pedestrian Walking Speed Assumptions. ITE Journal, Vol. 74, No. 9, pp. 32–40. 24. Fitzpatrick, K., S. M. Turner, M. Brewer, P. J. Carlson, B. Ullman, N. D. Trout, E. S. Park, J. Whitacre, N. Lalani, and D. Lord. 2006. TCRP Report 112/NCHRP Report 562: Improving Pedestrian Safety at Unsignal- ized Crossings. Transportation Research Board, Washington, DC. 25. Knoblauch, R. L., M. T. Pietrucha, and M. Nitzburg. 1996. Field Studies of Pedestrian Walking Speed and Start-Up Time. Transportation Research Record 1538, pp. 27–38. 26. Transportation Research Board. 2000. Highway Capacity Manual. Washington, DC. 27. National Fire Protection Association. 2020. NFPA 130 Standard for Fixed Guideway Transit and Passenger Rail Systems. Washington, DC. 28. Park, H. J., W. Yang, W. Yu, I. Wagner, and S. Ahmed. 2014. Investigation of Pedestrian Crossing Speeds at Signalized Intersections with Heavy Pedestrian Volumes. Transportation Research Record: Journal of the Transportation Research Board, No. 2463, pp. 62–69. 29. Peters, D., L. Kim, R. Zaman, G. Haas, J. Cheng, and S. Ahmed. 2015. Pedestrian Crossing Behavior at Signalized Intersections in New York City. Transportation Research Record: Journal of the Transportation Research Board, No. 2519, pp. 179–188.