Assessing Interactions Between Access Management Treatments and Multimodal Users (2018)

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.

196 A P P E N D I X E : F I N A L S T U D Y D E S I G N S Final Study Designs Introduction This appendix describes the final study design for three of the four high-priority access management (AM) techniques identified in Table 83 of Appendix C. The three techniques that are described are driveway design, right-turn deceleration, and TWLTL vs. non-traversable median. The Phase 2 work plan recommended delaying the start of the study associated with the corner clearance technique. This strategy was used to maximize the probability of success for the three higher priority studies (i.e., those studies identified in the previous paragraphs). As time passed, challenges associated with these three studies were encountered, which required additional project resources to overcome. As a result, the corner clearance study was not undertaken in Phase 2. Each final study design described in this appendix expands the initial study design documented in Appendix D. The initial study design outlined the modal focus, study objectives, analysis scale, and data sources needed to develop quantitative relationships for predicting specific performance measures. The final study design incorporated additional information about site characteristics, sample size, and data collection methods. The final study designs were used to guide the site selection, data collection, and data reduction activities in Phase 2. Driveway Design This section describes the final study design for the driveway design technique. The original title of this technique is install driveways with the appropriate return radii, throat width, and throat length for the type of traffic to be served. The study was intended to produce performance relationships that describe the change in safety and operations associated with a change in key driveway design elements. Objectives and Scope The objective of this study is to develop operations and safety performance relationships describing the effect of key driveway design elements (i.e., return radii and throat width) on the performance of the pedestrian and bicycle travel modes. The panel indicated that research on the transit and truck modes for this technique was low priority. Table 107 identifies the travel models for which new performance relationships are planned for development. Details of the relationship development process are described in subsequent subsections.

197 Table 107. Performance relationships planned for development – driveway design. Performance Measure Category Performance Relationship (PR) Research by Travel Mode Pedestrian Bicycle Transit Truck Operations Develop new PR Develop new PR not addressed not addressed Safety Develop new PR Develop new PR not addressed not addressed The focus of the study was on (1) typical pedestrians crossing the driveway (traveling parallel to the intersecting major street) and (2) typical bicyclists passing through the street-driveway intersection (while traveling along the major street). These two travel paths are shown in Figure 5. Pedestrians and bicycles that travel in other paths through the intersection were not the focus of this study. Figure 5. Intersection boundaries – driveway design. The street-driveway intersection of interest had (1) no control for the two major street approaches and (2) a legal requirement for drivers on the driveway approach to yield the right-of-way to major street vehicles (but no sign or signal control). Summary of Literature Review Findings The literature was reviewed to identify driveway traffic characteristics and design elements that may have an effect on pedestrian or bicycle safety or operation. Design elements that reduce motorized vehicle speed when entering or exiting the driveway can have a major effect on pedestrian safety. The data elements that were identified from the literature review are identified in Table 108. The elements listed in the table are likely to have some correlation with the safety or operation of the pedestrian or bicycle travel modes. A reduction in the curb return radius can reduce vehicle speed a modest amount. However, it can also significantly reduce the pedestrian crossing distance and thus, the pedestrian exposure to vehicle conflict.

198 The presence of on-street parking can have an adverse effect on driver sight distance to the driveway, especially if the parking stalls are located near the driveway and are often occupied by parked vehicles. The presence of a bicycle lane generally improves driver sight distance and can mitigate the negative effect of on-street parking on sight lines. Table 108. Data elements possibly affecting performance – driveway design. Category Data Element Data Need by Travel Mode Ped Bike Transit Truck Design Driveway left-turn prohibition (i.e., prohibit only left turns out of driveway, or prohibit left turns into and out of driveway) Yes Yes Number and width of driveway lanes (entering, exiting) Yes Yes Distance from major street curb line to near edge of sidewalk (at driveway) Yes No Driver sight distance (to/from driveway) Yes Yes Presence and width of outside shoulder Yes Yes Presence of on-street parking within 250 ft. of the driveway Yes Yes Parking setback (distance measured along major street between nearest parking stall and near edge of driveway) Yes Yes Presence and width of marked bicycle lane Yes Yes Curb return radius (use equivalent radius if flare is present) Yes No Presence of a triangular channelizing island Yes No Width of a non-traversable median on the driveway Yes Yes Presence of a right-turn bay on the major street Yes Yes Distance to nearest downstream signalized intersection Yes Yes Distance to nearest upstream driveway (or intersection) Yes Yes Distance to nearest downstream driveway (or intersection) Yes Yes Driveway skew angle Yes Yes Distance from major street curb line to marked (or effective) driveway stop line (i.e., stop line setback) Yes No Presence and radius of horizontal curve in major street Yes Yes Presence of changes in major street vertical alignment through street–driveway intersection (e.g., for drainage) No Yes Pavement condition rating for major street in vicinity of street–driveway intersection (see Chapter 17 of 2010 HCM) No Yes Driveway grade in vicinity of sidewalk Yes No Driveway throat length Yes No Traffic characteristics Bicycle speed No Yes Motorized vehicle volume (left-turn, through, right-turn; all vehicle types; a.k.a. “traffic volume”) Yes Yes Pedestrian volume crossing driveway Yes No Bicycle volume on major street (adjacent to driveway) No Yes Truck volume (left, through, right) entering and exiting driveway Yes Yes Proportion time on-street parking is occupied Yes Yes Traffic control Speed limit of major street Yes Yes Presence of crosswalk markings on driveway Yes No Driveway traffic control (none, yield, stop, signal) Yes Yes

199 Data Collection Plan This section describes the data collection plan for driveway design. The plan is focused on the collection and reduction of data needed to develop relationships describing the effect of driveway design on pedestrian and bicycle performance. The plan consists of two components: (1) development of operations performance relationships (based on an examination of delay and stops measured in the field); and (2) development of safety performance relationships (based on an examination of surrogate safety measures [i.e., conflicts] obtained from field observations). The safety component also includes an investigation of the correlation between conflict frequency and reported crash frequency at the field study sites. Approach A cross-sectional study method was used for the evaluation based on consideration of the study scope and available resources. The database that was assembled consists of many sites that collectively have a range of driveway widths and curb return radii. A study site is defined to be one street-driveway intersection. Operations Relationships Based on Field Data Field data were necessary to evaluate the operational effects of alternative driveway designs. A cross- sectional database was assembled to include data for many driveways. The set of driveways represented in the database collectively include a range of curb return radii and throat widths. These data were evaluated to develop the performance relationships identified in the following list.  An operations-based performance relationship for pedestrian travel; and  An operations-based performance relationship for bicycle travel. These relationships describe the association between performance, curb return radius, and throat width. The collected data describe representative design and traffic conditions at each study site for three one- hour field study periods. In this manner, a total of three hours of traffic operations data were collected for each site. Database Elements Data elements that may influence the effects of driveway design on pedestrian and bicycle operation are listed in Table 108. These data elements were collected for each study site. Performance measures that were measured during the study period include:  Pedestrian delay at driveway crosswalk;  Pedestrian stop rate at driveway crosswalk;  Delay to bicyclists traveling through the street-driveway intersection on the major street; and  Bicycle stop rate when traveling through the street-driveway intersection on the major street. Site Selection Considerations This subsection lists the desirable characteristics for the collective set of study sites. This list was intended to guide the selection of candidate study sites.

200 Sample Size. Consideration of sampling statistics and project resources indicated that data were needed for at least 20 sites. Where applicable, the data collected for these sites were used for both the operations- based field study and the safety-based field study. Geographic Diversity. The collective set of sites represented in the database has sufficient geographic diversity to ensure transferability of the findings. This diversity was achieved by having a portion of the sites located in at least four regions of the U.S. (e.g., West, Southeast, Northeast, Midwest). To maximize the potential to obtain adequate pedestrian and bicycle sample sizes, the sites were located in larger urban areas or communities with universities. Criteria Used to Control Secondary Influences. Selected design, traffic characteristics, and traffic control elements were controlled to minimize their potential influence on pedestrian and bicycle performance. This technique facilitates the examination of key data elements in isolation of possible secondary influences. It is intended to maximize the potential for identifying causal connections and developing reliable performance relationships. In this regard, the following site selection criteria were established:  Include only sites that have no control for the major street and a legal requirement for drivers on the driveway approach to yield to major street vehicles;  Include only driveways with two-way traffic flow;  Exclude sites that have an outside shoulder;  Exclude sites with a speed limit of 45 mph or more;  Exclude sites having pedestrian crossing traffic control devices or traffic calming devices;  Exclude sites having horizontal curvature in the major street alignment;  Exclude sites that are within 350 feet of a signalized intersection;  Exclude sites that are within 350 feet of a bus stop; and  Exclude sites within the right-turn bay associated with a nearby signalized intersection. As discussed in a subsequent section, the selected sites were also used to investigate the relationship between crash frequency and conflict frequency. Thus, the sites were selected such that their crash history (including fatal, injury, and property-damage-only crashes) was confirmed to be available for the most recent five-year period. This crash history was confirmed to not be influenced by secondary factors, as identified in the following list:  Exclude sites that have work zones present during any portion of the most recent five-year period; and  Exclude sites for which major geometric design changes have occurred to the street-driveway intersection during the most recent five-year period. Volume Criteria. Local agency contacts and surrounding land uses were used to determine whether the selected sites would likely have a relatively high driveway volume and a high pedestrian or bicycle volume. Illustrative minimum study-period volumes for these thresholds are: 250 veh/h two-way flow, 20 pedestrians per hour, and 20 bicycles per hour, respectively. To support the investigation of conflict frequency, the major street annual average daily traffic (AADT) volume was needed for each study site. Other Criteria. In addition to the aforementioned criteria, the sites were selected to ensure that they collectively include a range of values for key geometric design elements. These elements include:  About 1/2 of the sites should have adjacent on-street parking (and the other 1/2 of sites should not have on-street parking);

201  About 1/2 of the sites should have a marked bicycle lane on the major street (and the others should have no marked lane);  At least 1/3 of the sites should have driveway left-turn prohibition on the major street (and the others should allow left turns into and out of the driveway);  Range of curb return radius: 0 to 50 feet;  Range of driveway width per lane: 11 to 23 feet; and  Range of major street approach lanes: 1 to 2. Study Design A wide variety of data was planned for collection at each study site. The geometric design and traffic control elements listed in Table 108 were obtained using aerial photographs available from the internet. Traffic Characteristics Data. The traffic characteristic data listed in Table 108 was measured in the field at each site during the study period. The traffic characteristic and performance measure data were collected using two camcorders. The desired data was extracted from the videotape recordings during replay in the office. The two cameras were positioned as shown in Figure 6. One camera was used to observe traffic on the subject driveway leg of the intersection. The second camera was used to observe traffic on the adjacent major street approach. Both cameras were used to observe pedestrians crossing the driveway and bicycles traveling along the major street. Figure 6. Camera location at driveway study site – driveway design. Study Procedures. Camcorders were used to record traffic conditions at each site during each of three one-hour study periods. Each one-hour study period occurred during a time period for which the pedestrian and bicycle delay was at its highest level during the typical day. This time period coincided with the hours having peak pedestrian volume, peak bicycle volume, and moderate-to-high motorized vehicle volume. The study periods occurred during typical weekday daylight time periods. No data were collected during inclement weather, during nighttime hours, or when traffic flow was disrupted by incidents, work zone activity, or special events.

202 Data Reduction. Data reduction procedures were developed and documented. All data reduction staff were trained using these procedures to ensure consistency in the data extracted from the video recordings and from the aerial photographs. Safety Relationships Based on Field-Measured Conflict Data Conflict data were necessary to evaluate the safety effects of alternative driveway designs. A cross- sectional database was assembled to include conflict data for many driveways. The set of driveways represented in the database collectively include a range of curb return radii and throat widths. These data were evaluated to develop the performance relationships identified in the following list.  A conflict-based safety performance relationship for pedestrian travel; and  A conflict-based safety performance relationship for bicycle travel. These relationships describe the influence of curb return radius and throat width on driveway safety and operations. The collected data describe design and traffic conditions at each study site for five one-hour field study periods. In this manner, a total of five hours of conflict data were collected for each site. These study periods overlapped (i.e., include) the three one-hour study periods described previously for the operations-focused field study (refer to the subsection titled operations relationships based on Field Data). The additional two hours of conflict data were believed necessary to achieve the minimum sample size based on typical conflict rates published in the literature. Database Elements Data elements that may influence the effects of driveway design on pedestrian and bicycle safety are listed in Table 108. These data elements were collected for each study site. Performance measures that were measured during the study period include:  Vehicle-pedestrian conflict frequency in the driveway crosswalk; and  Vehicle-bicycle conflict frequency at the street-driveway intersection. Several specific types of conflicts and erratic maneuvers were identified prior to the start of the data extraction process. Conflicts were defined by the direction of travel for each of the two conflicting travel movements and the associated response of the two travelers (e.g., right-turn vehicle-pedestrian conflict). Site Selection Considerations The site selection considerations for this conflict study were the same as described previously for the operations-focused field study (refer to the subsection titled operations relationships based on Field Data). Study Site Locations The study sites for this conflict study are the same as described previously for the operations-focused field study (refer to the subsection titled operations relationships based on Field Data). Study Design The geometric design and traffic control elements listed in Table 108 were obtained using aerial photographs available from the internet. These were shared with the operations-based field study described previously. Traffic Characteristics Data. The traffic characteristic data listed in Table 108 were measured in the field at each site during the study period. The traffic characteristic and performance measure data were

203 collected using two camcorders. The desired data were extracted from the videotape recordings during replay in the office. The two cameras were positioned as shown in Figure 6. The major street AADT volume was obtained from the local transportation agency for each of the most recent five years. Interpolation was used to estimate the AADT volume for intermediate years for which AADT data are unavailable. Representative pedestrian and bicycle hourly volume estimates for the average day were estimated using the pedestrian and bicycle counts measured during the study period. Adjustment factors based on typical time of day, day of week, and month of year variations (obtained from the local agency) were used by the researchers to compute these volume estimates. Crash Data. Electronic copies of the crash reports (including fatal, injury, and property-damage-only crashes) for the major street associated with each study site were requested from the local agency. The crash data requested described crashes that (1) occurred at or within 50 feet of the intersection; (2) are vehicle-pedestrian or vehicle-bicycle crashes; and (3) occurred during the most recent five years for which crash data are available. The data received were screened to retain only those crashes where one or more of the involved vehicles was entering or exiting the subject driveway (i.e., driveway-related). The attributes in the crash data include crash date, crash time of day, crash type, crash severity, crash location (i.e., milepost), work zone presence, contributing factor, parties involved (i.e., pedestrian, bicyclist, motor vehicle), and intersection relationship (i.e., driveway, intersection, other). Data Reduction. Data reduction procedures were developed and documented. All data reduction staff were trained using these procedures to ensure consistency in the data extracted from the aerial photographs. Right-Turn Deceleration This section describes the final study design for the right-turn deceleration technique. The original title of this technique is 4a. install right-turn deceleration lane. The study was intended to produce performance relationships that describe the change in safety and operations associated with the addition of a right-turn deceleration lane at a signalized intersection. Objectives and Scope The objective of this study was to develop operations and safety performance relationships describing the effect of exclusive right-turn deceleration lane presence on select non-passenger-car travel modes at urban signalized intersections. The specific travel modes and associated performance relationships of interest are identified in Table 109. As indicated in the table, performance relationships are available in the literature for three of the eight travel mode/performance categories. As a result, the focus of this research was on the development of new relationships for the remaining five travel mode/performance categories. The findings from the literature review are provided in Appendix A. Table 109. Performance relationships planned for development – right-turn deceleration. Performance Measure Category Performance Relationship (PR) Research by Travel Mode Pedestrian Bicycle Transit Truck Operations Use existing PR Develop new PR Develop new PR Develop new PR Safety Use existing PR Use existing PR Develop new PR Develop new PR

204 The research addressed right-turn deceleration lane presence at signalized intersections in urban areas. The addition of a right-turn lane is anticipated to change the safety and operation of the aforementioned five travel mode/performance categories. The bicycle movement of interest is that which travels on the intersection approach on which the turn lane is installed (and in the same direction as the motorized vehicles using the approach). The bicycle crosses the intersection as a through-traffic movement. Bicyclists that turn left or right at the intersection were not the focus of this study. The study of the bicycle mode was focused on the typical bicyclist. The transit movement of interest is that which travels on the intersection approach on which the turn lane is installed. The transit vehicle stops at the intersection (using a near side stop) to pick up passengers and then crosses the intersection as a through movement. Transit vehicles that (1) do not stop, (2) stop on the far side, or (3) stop and then turn left or right are not the focus of this study. The truck movement of interest is that which travels on the intersection approach on which the turn lane is installed. The trucks of interest may turn right at the intersection or continue through the intersection. Summary of Literature Review Findings The literature was reviewed to identify right-turn-lane-related traffic characteristics and design elements that may have an effect on non-passenger-car safety or operation. The data elements that were identified from the literature review are identified in Table 110. The elements listed here are likely to have some correlation with the safety or operation of the bicycle, transit, or truck travel modes. The bicycle travel mode is considered only for operational performance. Table 110. Data elements possibly affecting performance – right-turn deceleration. Category Data Element Data Need by Travel Mode Ped Bike (ops. only) Transit Truck Design Number and width of right-turn lanes on subject approach Yes Yes Yes Number and width of adjacent through lanes Yes Yes Yes Number and width of lanes on the cross-street leg Yes Yes Yes Presence and width of right-turning roadway (if triangular channelizing island is present) No No Yes Right-turn entry angle (if triangular channelizing island is present) a No No Yes Curb return radius No No Yes Length of right-turn bay (full width length, exclude taper) Yes Yes Yes Presence and width of marked bicycle lane on the subject approach Yes Yes Yes Intersection skew angle No Yes Yes Presence of near side bus stop and distance between bus stop and nearest stop line No Yes Yes Presence of on-street parking within 250 ft. of the intersection Yes Yes Yes Parking setback (distance measured along major street between nearest parking stall and nearest stop line) Yes Yes Yes

205 Category Data Element Data Need by Travel Mode Ped Bike (ops. only) Transit Truck Presence and width of marked bus-only lane on the subject approach Yes Yes Yes Presence of right-turn accepting lane on cross street No No Yes Traffic characteristics Motorized vehicle volume (left-turn, through, right-turn; all vehicle types; a.k.a. “traffic volume”) Yes Yes Yes Motorized vehicle queue length in outside approach lane No Yes Yes Truck volume (left, through, right), on subject approach Yes Yes Yes Number of stops per hour by local bus Yes Yes Yes Bicycle volume on the subject approach Yes No No Bicycle speed Yes Yes Yes Bus dwell time No Yes No Traffic control Speed limit on subject approach Yes Yes Yes Average phase duration for subject approach Yes Yes Yes Signal cycle length Yes Yes Yes Prohibition of right turns on red indication Yes No Yes Traffic control for right-turn movement (free, yield, stop, signal) No No Yes Right-turn mode (protected-permitted, permitted-only) No No Yes Note: a – Right-turn entry angle represents the angle between the centerline of the subject right-turning vehicle (when at the stop or yield point) and the right-turning driver’s line of site to approaching traffic in the nearest conflicting through lane. Data Collection Plan This section describes the data collection plan for the investigation of right-turn deceleration lane influence on operation and safety. The plan is focused on the collection and reduction of data needed to develop relationships describing the effect of right-turn deceleration lane presence on bicycle, transit, and truck performance. The plan consists of two components. One component is focused on the development of operations performance relationships. It is based on an examination of delay measures obtained from simulation. A second component of the plan is focused on the development of safety performance relationships. It is based on an examination of surrogate safety measures obtained from simulation. Approach A study site was defined to be one signalized intersection approach. A before–after study method was used for the evaluation based on consideration of the study scope and available resources. The database that was assembled included simulation-based performance measures for sites with the presence of a bike lane and with/without a right-turn deceleration lane. Operations Relationships Based on Simulation Data Simulation data were necessary to evaluate the operational effects of right-turn lane presence. A simulation testbed was established using intersections without right-turn deceleration lanes to represent

206 the “before” condition. A second testbed was based on the “before” condition intersections but they included right-turn deceleration lanes to represent the “after” condition. The set of sites represented in the database collectively include simulation results for a range of traffic characteristics, geometric design elements, and traffic control features. These data were evaluated to develop the performance relationships identified in the following list.  An operations-based performance relationship for bicycle travel;  An operations-based performance relationship for transit travel; and  An operations-based performance relationship for truck travel. These relationships describe the association between operational performance and right-turn lane presence. The VISSIM simulation model was used in this project because it has the capability of simulating cars, pedestrians, bicycles, transit, and trucks. The plan for calibrating this model is discussed in a later section. Database Elements Data elements that may influence the effects of right-turn deceleration lane presence on bicycle, transit, and truck operations are listed in Table 110. Performance measures that were measured during the study period include:  Control delay to bicyclists;  Control delay to transit riders; and  Control delay to trucks. Study Design Simulation Test Bed. The phrase “simulation test bed” is used herein to describe the simulation model input data file set up for the evaluation of a specific technique and containing the values of the calibration parameters. The testbed describes the traffic characteristics, geometric design elements, and traffic control device features of the site of interest. Selected input variables are considered to be independent variables. The values of the independent variables in the testbed were then changed, in conformance with the experimental design, to form supplemental input data sets. The test bed used for this study was based on the signalized intersections used for model calibration, as described in a later section. The use of these intersections provided a real-world character to the test bed. Given that a site is one intersection approach, each four-leg intersection represented in the test bed included up to four sites (i.e., one site per leg). Test Bed Characteristics. The test bed was used to develop a large number of realistic combinations of intersection design (i.e., scenarios). A test bed without the presence of a right-turn deceleration lane was used to establish base performance conditions. This test bed is called the “base” test bed. Its characteristics are identified in Table 111. Table 111. Characteristics of the base test bed – right-turn deceleration. Category Data Element Values Combinations Design Number of right-turn lanes 0 (base condition) 1 Number of adjacent through lanes 2 1 Number of left-turn lanes 1 1 Number of lanes on the cross-street leg 2 1

207 Category Data Element Values Combinations Presence of turning roadway and triangular channelizing island no 1 Curb return radius 25 ft. 1 Presence of marked bicycle lane yes 1 Intersection skew angle 0 degrees (no skew) 1 Presence of near side bus stop no, yes a, b 1 Presence of on-street parking no 1 Presence of bus-only lane no 1 Presence of right-turn accepting lane no 1 Traffic characteristics Motorized vehicle volume (left-turn, through, right-turn; all vehicle types; a.k.a. “traffic volume”) 450, 550, 650 veh/h/ln 3 Percent left-turn motorized vehicles 2% 1 Percent right-turn motorized vehicles 2%, 10%, 20% 3 Truck volume 3%, 6% 2 Number of stops per hour by local bus 0, 1, 2 b 3 Bicycle volume 0, 30, 60 bicycles/h 3 Bicycle running speed 12 mph 1 Bus dwell time 15, 30 s 2 Traffic control Speed limit 35 mph (corresponds to a free-flow speed of 40 mph) 1 G/C ratio (through/left-turn) 0.40/0.10 c 1 Signal cycle length 100, 150 s 2 Right-turn-on-red yes, no 2 Traffic control for right turn yield for 75 ft. radius, signal for other radii 1 Right-turn mode permitted-only (if signalized) 1 Left-turn mode protected 1 Product: 1,296 Notes: a – The bus stop is located such that the front of the stopped bus is 5 ft. in advance of the stop line. b – When a bus stop is present, the number of stops per hour is 1 or 2. When a bus stop is not present, the number of stops per hour is 0. c – G equals the green interval duration. C equals the signal cycle length. The last column in Table 111 indicates that some data elements are shown to be varied and others are shown to be constant. By holding selected elements constant, the effect of the other characteristics can be quantified more efficiently. Those characteristics that are held constant are considered to have either a secondary effect on performance or are not believed to occur with reasonable frequency at most urban signalized intersections. The factorial number of combinations of data elements is computed using the number of value combinations listed in the last column of Table 111. The product of all combinations produces a total of 1,296 unique combinations (i.e., scenarios). The “right-turn lane” test bed data elements that were evaluated are shown in Table 112. A right-turn deceleration lane was added to the base test bed to create the right-turn lane test bed. Specifically, the approximately 1,296 scenarios represented in the base test bed were expanded to include a right-turn lane with one of three lengths (i.e., 100, 200, or 300 ft.), thus creating 3,888 scenarios (= 3 × 1,296) for the right-turn lane test bed. All total, 5,184 scenarios (= 1,296 base + 3,888 right-turn lane) were evaluated using simulation.

208 Table 112. Characteristics of the right-turn lane test bed – right-turn deceleration. Category Data Element Values Combinations Design Number of right-turn lanes 1 a 1 Length of right-turn bay 100, 200, 300 ft. 3 Note: a – When a right-turn lane is present, the bus stop is located in the right-turn lane. A queue jump signal is used to allow the bus to re-enter the adjacent through lane. The advance green for queue jump is 9 seconds. Data Collection Process. Each intersection was simulated for a one-hour simulation time period. The simulation software tool was programmed to obtain the motorized vehicle, bicycle, transit, and truck control delay on each leg for each one-hour period. The simulation initialization period was five minutes in duration. This duration should ensure that the first vehicles to enter have traversed the simulated road system. The simulation run for a given intersection (i.e., for a given set of four scenarios) was replicated twice to provide some indication of the random variation in the performance measures. Whenever this variation was found to be large, additional replications were made for each scenario. In all cases, the database with simulation results included an equal number of replications for each scenario. That is, if a statistical analysis indicated that three replications were needed, then all 5,184 scenarios were represented in the database by data from three replications. The output for each leg for a one-hour simulation period represents one observation in the output database. The database was structured as a flat file and stored in one Excel worksheet. Each observation in the database included the value associated with each data element identified in Table 111 and Table 112. In addition, each observation included the following attributes:  Simulation run number (1, 2, 3…);  Replication number (1, 2);  Right-turn motorized vehicle average control delay;  Bicycle average control delay;  Transit average control delay;  Transit dwell time;  Number of stops by bus before clearing the intersection; and  Truck average control delay. Safety Relationships Based on Simulated Conflict Data Simulated conflict data were necessary to evaluate the safety effects of right-turn lane presence at a signalized intersection. The simulation testbed established for the operations-focused study was used for this purpose (refer to the section titled Operations Relationships Based on Simulation Data). The set of sites represented in the database collectively include simulation results for a range of traffic volume, geometric design, and traffic control conditions. These data were evaluated to develop the performance relationships identified in the following list.  A safety-based performance relationship for transit travel; and  A safety-based performance relationship for truck travel. These relationships describe the association between safety performance and right-turn lane presence.

209 The VISSIM simulation model was used in this project because it has the capability of simulating cars, pedestrians, bicycles, transit, and trucks. The plan for calibrating this model is discussed in a later section. Database Elements The vehicle trajectory files generated by VISSIM were post-processed in the SSAM tool to extract surrogate conflict measures for the safety analysis. Conflicts were defined based on user-calibrated thresholds of time-to-collision (TEC) and Post-Encroachment Time (PET). Conflicts were filtered by attributes like vehicle speed (e.g. removing conflicts that happen at low speeds, which are unlikely to result in crashes). Conflicts were further extracted separately for different vehicle classes, to isolate conflicts for truck and transit vehicles. For the SSAM data analysis, the team used best practices available in the literature, supported by additional calibration performed as part of this project. The simulation models were configured in a way that transit vehicles and trucks are assigned their own vehicle classes. This approach allowed the team to extract only conflicts pertaining to those modes. Specific performance measures include:  Frequency and rate of simulated conflicts involving transit vehicles; and  Frequency and rate of simulated conflicts involving trucks. Study Design Simulation Test Bed. The simulation test bed for this study was the same as described previously for the operations-focused study (refer to the subsection titled Operations Relationships Based on Simulation Data). Test Bed Characteristics. The test bed characteristics for this study are the same as described previously for the operations-focused study (refer to the subsection titled Operations Relationships Based on Simulation Data). Data Collection Process. The output from the simulation software is described in the previous subsection titled Operations Relationships Based on Simulation Data. The extraction of safety surrogates from this simulation output is described in this subsection. The surrogate safety data were extracted using the Surrogate Safety Assessment Methodology (SSAM), which was developed by FHWA as a post-processing tool for simulation trajectory data. To use SSAM, the team generated a trajectory file output from VISSIM, which was then post-processed in the SSAM tool. The surrogate safety method is based on Time-to-Contact (TTC) principles, where a short time to contact is indicative of a potential conflict. In SSAM, TTC is calculated based on the evaluation of the position and velocity vectors of each object (vehicle, pedestrian, bicycle, or transit vehicle), searching for instances where two or more objects are likely to occupy the same space at the same time (i.e., a simulated crash). The SSAM tool can be calibrated using user-specified thresholds for TTC, speed, and other parameters. The SSAM analysis process results in large file sizes (trajectory files store the position of each object ten times per second), which can be time-consuming to process. As such, the team limited the SSAM post-processing to a single, representative, run of the operational modeling results. TWLTL vs. Non-Traversable Median This section describes the final study design for the comparison of the TWLTL and non-traversable median (NTM) as access management techniques. The study was intended to produce one set of performance relationships that describe the safety and operation of the TWLTL, and a second set of relationships that describe the safety and operation of the NTM. When both sets of relationships are used

210 together, the results can be compared to obtain information about the relative performance of the two techniques. Objectives and Scope The objective of this study was to develop operations and safety performance relationships describing the effect of TWLTLs and NTMs on select non-passenger-car travel modes on urban street segments bounded by signalized intersections. The specific travel modes and associated performance relationships of interest are identified in Table 113. As indicated in the table, performance relationships are available in the literature for four of the eight travel mode/performance categories. As a result, the focus of this research was on the development of new relationships for the remaining four travel mode/performance categories. The findings from the literature review are provided in Appendix A. Table 113. Performance relationships planned for development – TWLTL vs. non-traversable median. Performance Measure Category Performance Relationship (PR) Research by Travel Mode Pedestrian Bicycle Transit Truck Operations Use existing PR Develop new PR Use existing PR Develop new PR Safety Use existing PR Use existing PR Develop new PR Develop new PR The research addressed TWLTL and NTM presence on arterial streets in urban areas. The safety and operation of a street with a TWLTL was expected to have definable relationships with traffic volume, driveway density, speed, geometric design elements, traffic control features, and traffic characteristics. Similarly, the safety and operation of a street with a NTM was expected to have definable relationships with several geometric design elements, traffic control features, and traffic characteristics. The bicycle movement of interest is that which travels along the street segment in a marked bicycle lane or in the traffic lanes. The study of the bicycle mode is focused on the typical bicyclist. The transit movement of interest is that which travels along the street and stops to pick up passengers at mid-segment locations. Transit vehicles include public transit bus and school bus. The truck movement of interest is that which enters the segment as a through vehicle or turns onto the segment at a driveway. It then exits the segment as a through vehicle or turns off the segment at a driveway. Trucks include all heavy vehicles (e.g., motor home, tractor-trailer, local delivery truck) with more than two axles, and vehicles pulling trailers. A pickup truck is not considered a truck for the purposes of this study. Summary of Literature Review Findings The literature was reviewed to identify median-type-related traffic characteristics and design elements that may have an effect on non-passenger-car safety or operation. The data elements that were identified from the literature review (conducted in Phase I) are identified in Table 114. The elements listed here are likely to have some correlation with the safety or operation of the bicycle, transit, or truck travel modes. The bicycle travel mode is considered only for operational performance, and the transit mode is considered only for safety performance.

211 Table 114. Data elements possibly affecting performance – TWLTL vs. non-traversable median. Category Data Element Data Need by Travel Mode Ped Bike (ops. only) Transit (safety only) Truck Design Number of through lanes Yes Yes Yes Segment length Yes Yes Yes Width of through lanes No Yes Yes Median or TWLTL width No Yes Yes Number of unsignalized access points (per mile) with right- in-right-out only (no right-turn bay on major street) Yes Yes Yes Number of unsignalized access points (per mile) with all movements allowed (no right-turn bay on major street) Yes Yes Yes Number of unsignalized access points (per mile) with right- turn bay on major street Yes Yes Yes Presence and width of outside shoulder Yes Yes Yes Presence of on-street parking Yes Yes Yes Presence and width of marked bicycle lane Yes Yes Yes Presence and width of marked bus-only lane Yes Yes Yes Presence of mid-segment transit stop No Yes No Land use adjacent to each unsignalized access point No Yes Yes Traffic characteristics Through passenger car volume Yes Yes Yes Turn movement percentage for boundary signals Yes No No Through bicycle volume Yes Yes Yes Through truck volume Yes No Yes Passenger car turn movement volume at unsignalized access points (including U-turn volume) Yes Yes Yes Truck turn movement volume at unsignalized access points (including U-turn volume) No No Yes Number of stops per hour by local bus No Yes No Proportion of time on-street parking is occupied Yes Yes Yes Traffic control Speed limit Yes Yes Yes Presence of mid-segment signalized pedestrian crossing Yes Yes Yes Average phase duration at boundary signals Yes Yes Yes Signal cycle length at boundary signals Yes Yes Yes Motorized vehicle volume at the unsignalized access points is not expected to be available at the sites used for the safety-focused study. As a result, volume surrogates were included in the database to provide some information about unsignalized access point volume level. These surrogates are identified in the following list:  Number of unsignalized access points per mile; and  Land use adjacent to unsignalized access point (e.g., residential, office, commercial). Data Collection Plan This section describes the data collection plan for the investigation of TWLTL and NTM influence on operation and safety. The plan is focused on the collection and reduction of data needed to develop

212 relationships describing the effect of TWLTL presence and NTM presence on bicycle, transit, and truck performance. The plan consists of two components. One component is focused on the development of operations performance relationships. It is based on an examination of travel speed and stop rate measures obtained from simulation. A second component of the plan is focused on the development of safety performance relationships. It is based on an examination of the reported crash history of several street segments including marked bicycle lanes and having either a TWLTL or a NTM. Approach For the operations-focused study, a site was defined to include one direction of travel along a street segment bound by signalized intersections. For the safety-focused study, a site was defined to include both directions of travel along a street segment. A cross-sectional study method was used for the evaluation based on consideration of the study scope and available resources. The database that was assembled consists of many sites that include marked bicycle lanes and collectively have either a TWLTL or NTM and a range of volumes, speeds, and unsignalized access point densities. Operations Relationships Based on Simulation Data Simulation data were necessary to evaluate the operational effects of the TWLTL and the NTM. A simulation testbed was established using several different prototype street segments. One-half of the segments included a TWLTL, the other one-half of the segments included a NTM. The set of sites represented in the database collectively include simulation results for a range of traffic characteristics, geometric design elements, and traffic control features. These data were evaluated to develop the performance relationships identified in the following list.  An operations-based performance relationship for bicycle travel on a street with a TWLTL;  An operations-based performance relationship for bicycle travel on a street with a NTM;  An operations-based performance relationship for truck travel on a street with a TWLTL; and  An operations-based performance relationship for truck travel on a street with a NTM. These relationships describe the association between performance and various factors for both the TWLTL and the NTM. The VISSIM simulation model was used in this project because it has the capability of simulating cars, pedestrians, bicycles, transit, and trucks. The plan for calibrating this model is discussed in a later section. Database Elements Data elements that may influence the effects of median type on bicycle and truck operations are listed in Table 114. Performance measures that were measured during the study period include:  Bicyclist travel speed; and  Through truck travel speed. Study Design Simulation Test Bed. The phrase “simulation test bed” is used herein to describe the simulation model input data file set up for the evaluation of a specific technique and containing the values of the calibration parameters. The testbed describes the traffic characteristics, geometric design elements, and traffic control device features of the site of interest. Selected input variables are considered to be independent variables.

213 The values of the independent variables in the testbed were then changed, in conformance with the experimental design, to form supplemental input data sets. The test bed used for this study was based on the street segments used for model calibration, as described in a later section. The use of these intersections provided a real-world character to the test bed. Given that a site is one direction of travel along a street segment, each street segment represented in the test bed may include up to two sites (i.e., one site per direction of travel). Test Bed Characteristics. The test bed was used to develop a large number of realistic combinations of street segments (including bicycle lanes) with a TWLTL or with a NTM (i.e., scenarios). A test bed with the presence of a TWLTL was used to describe performance conditions for the TWLTL. This test bed is called the “TWLTL” test bed. Its characteristics are identified in Table 115. The last column in Table 115 indicates that some data elements are shown to be varied and others are shown to be constant. By holding select elements constant, the effect of the other characteristics can be quantified more efficiently. Those characteristics that are held constant are considered to have either a secondary effect on performance or are not believed to occur with reasonable frequency on most urban street segments. The factorial number of combinations of data elements is computed using the number of value combinations listed in the last column of Table 115. The product of all combinations produces a total of 432 unique combinations (i.e., scenarios).

214 Table 115. Characteristics of the TWLTL test bed – TWLTL vs. non-traversable median. Category Data Element Values Comb- inations Design Number of through lanes 2 1 Segment length 2 sites at 0.22 mi, 2 sites at 0.48 mi 4 Number of left-turn lanes at signal 1 1 Number of unsignalized access points with right- in-right-out only (no right-turn bay) 0 1 Number of unsignalized access points with all movements allowed (no right-turn bay) 2, 3, 5 for 0.22 mi site 1 2, 3, 4 for 0.22 mi site 2 4, 6, 8 for 0.48 mi site 1 4, 6, 10 for 0.48 mi site 2 3 Number of unsignalized access points with right- turn bay on major street 0 1 Presence of outside shoulder width no 1 Presence of on-street parking no 1 Presence of marked bicycle lane yes 1 Presence of bus-only lane no 1 Presence of mid-segment transit stop no 1 Traffic characteristics Through motorized vehicle volume (all vehicle types; a.k.a. “traffic volume”) 450, 550, 650 veh/h/ln 3 Percent left-turn motorized vehicles at signal 10% 1 Percent right-turn motorized vehicles at signal 10% 1 Bicycle free-flow speed 11 mph 1 Through bicycle volume 0, 30, 60 bicycles/h 3 Through truck percentage (as a percentage of through motorized vehicle volume) 3%, 6% 2 Passenger car turn percentage at unsignalized access points (as a percentage of through passenger car volume) 1% right turns off major, 1% on 4.5% left turns off major, 5% on 0.5% U-turn via major street 1 Truck turn percentage at unsignalized access points (as a percent of though truck volume) 1% right turns off major, 1% on 4.5% left turns off major, 5% on 0.5% U-turn via major street 1 Number of stops per hour by local bus 0 1 Portion of time on-street parking is occupied not applicable - Traffic Control Speed limit 35 mph (corresponds to a free-flow speed of 40 mph) 1 G/C ratio (through/left-turn) 0.40/0.10 a 1 Signal cycle length 100, 150 s 2 Right-turn-on-red yes 1 Right-turn mode permitted-only 1 Left-turn mode protected 1 Product: 432 Note: a – G equals the green interval duration. C equals the signal cycle length. A test bed with the presence of a NTM was used to describe performance conditions for the NTM. This test bed is called the “NTM” test bed. Its characteristics are identified in Table 116. There are 288 unique combinations identified for this test bed.

215 Table 116. Characteristics of the NTM test bed – TWLTL vs. non-traversable median. Category Data Element Values Comb- inations Design Number of through lanes 2 1 Segment length 2 sites at 0.22 mi, 2 sites at 0.48 mi 4 Number of left-turn lanes at signal 1 1 Number of unsignalized access points with right- in-right-out only (no right-turn bay) 1, 2 for 0.22 mi site 1 1, 1 for 0.22 mi site 2 2, 3 for 0.48 mi sites 1 and 2 2 Number of unsignalized access points with all movements allowed (no right-turn bay) 1, 2 for 0.22 mi site 1 1, 1 for 0.22 mi site 2 2, 3 for 0.48 mi sites 1 and 2 Number of unsignalized access points with right- turn bay on major street 0 1 Presence of outside shoulder width no 1 Presence of on-street parking no 1 Presence of marked bicycle lane yes 1 Presence of bus-only lane no 1 Presence of mid-segment transit stop no 1 Traffic characteristics Through motorized vehicle volume (all vehicle types; a.k.a. “traffic volume”) 450, 550, 650 veh/h/ln 3 Percent left-turn motorized vehicles at signal 10% 1 Percent right-turn motorized vehicles at signal 10% 1 Bicycle free-flow speed 11 mph 1 Through bicycle volume 0, 30, 60 bicycles/h 3 Through truck percentage (as a percentage of through motorized vehicle volume) 3%, 6% 2 Passenger car turn percentage at unsignalized access points (as a percentage of through passenger car volume) 3% right turns off major, 3% on 13.5% left turns off major, 15% on 1.5% U-turn via major 1 Truck turn percentage at unsignalized access points (as a percent of though truck volume) 3% right turns off major, 3% on 13.5% left turns off major, 15% on 1.5% U-turn via major 1 Number of stops per hour by local bus 0 1 Portion of time on-street parking is occupied not applicable - Traffic control Speed limit 35 mph (corresponds to a free-flow speed of 40 mph) 1 G/C ratio (through/left-turn) 0.40/0.10 a 1 Signal cycle length 100, 150 s 2 Right-turn-on-red yes 1 Right-turn mode permitted-only 1 Left-turn mode protected 1 Product: 288 Note: a – G equals the green interval duration. C equals the signal cycle length. With one exception, the data elements, values and combinations are the same in Table 116 and Table 115. The one exception is the use of some right-in-right-out access points along the NTM segment. Specifically, Table 116 indicates that, for a given segment length, the number of access points with right- in-right-out is the same as the number of full-movement access points. That is, for the 0.48-mile segment

216 length case, one scenario will include two right-in-right-out access points and two full-movement access points. A second scenario will include three right-in-right-out access points and three full-movement access points. The pattern is similar for the 0.22-mile segment length, but the number of access points provided is reduced. Data Collection Process. Each segment was simulated for a one-hour simulation time period. The simulation software tool was programmed to obtain the motorized vehicle, bicycle, and truck travel speed and stop rate for each direction of travel for each one-hour period. The simulation initialization period was sufficiently long to ensure that the first vehicles to enter have traversed the simulated road system. The simulation run for a given segment (i.e., for a given set of two scenarios) was replicated twice to provide some indication of the random variation in the performance measures. Whenever this variation was found to be large, additional replications were made for each scenario. In all cases, the database with simulation results included an equal number of replications for each scenario. That is, if a statistical analysis indicated that three replications were needed, then all scenarios would be represented in the database by data from three replications. The output for each travel direction for a one-hour simulation period represents one observation in the output database. The database was structured as a flat file and stored in one Excel worksheet. Each observation in the database included the value associated with each data element identified in Table 115 and Table 116. In addition, each observation included the following attributes:  Simulation run number (1, 2, 3…);  Replication number (1, 2);  Motorized vehicle average travel speed;  Motorized vehicle stop rate;  Bicycle average travel speed;  Bicycle stop rate;  Truck average travel speed; and  Truck stop rate. Safety Relationships Based on Crash Data Crash and field data were necessary to evaluate the safety effects of the TWLTL and NTM. A cross- sectional database was assembled to include crash and field data for many sites. The set of sites represented in the database collectively include a range of traffic characteristics, geometric design elements, and traffic control features. These data were evaluated to develop the performance relationships identified in the following list.  A crash-based safety performance relationship for transit travel on a street with a TWLTL;  A crash-based safety performance relationship for transit travel on a street with a NTM;  A crash-based safety performance relationship for truck travel on a street with a TWLTL; and  A crash-based safety performance relationship for truck travel on a street with a NTM. These relationships describe the association between performance and various factors for both the TWLTL and the NTM. The assembled data describe representative design and traffic conditions at each study site during a five-year study period. The study period represents the most recent five-year period for which crash data are available.

217 Database Elements Data elements that may influence the effects of median type on transit and truck safety are listed in Table 114. Performance measures that should also be measured during the study period are identified in the following list.  Transit vehicle-related crash frequency (and severity); and  Truck-related crash frequency (and severity). Site Selection Considerations This subsection lists the desirable characteristics for the collective set of study sites. This list was intended to guide the selection of candidate study sites. Sample Size. Consideration of sampling statistics and project resources indicated that data were needed for at least 260 sites (i.e., about 130 miles of urban street segments). Geographic Diversity. The collective set of sites represented in the database has sufficient geographic diversity to ensure transferability of the findings. This diversity was achieved by having a portion of the sites located in at least four regions of the U.S. (e.g., West, Southeast, Northeast, Midwest). The sites were located in larger urban areas to ensure adequate transit volumes. Criteria Used to Control Secondary Influences. Selected design, traffic characteristics, and traffic control elements were controlled to minimize their potential influence on transit and truck performance. This technique facilitates the examination of key data elements in isolation of possible secondary influences. It is intended to maximize the potential for identifying causal connections and developing reliable performance relationships. In this regard, the following criteria were established:  Include only sites that have two or three through lanes;  Include only sites that have either a TWLTL or NTM for the full length of the segment;  Include only sites with a marked bicycle lane;  Exclude sites that have an outside shoulder;  Exclude sites with a speed limit of 50 mph or more;  Exclude sites that have a marked bus-only lane;  Exclude sites with a mid-segment signalized pedestrian crossing;  Exclude sites having pedestrian crossing traffic control devices or traffic calming devices;  Exclude sites having sharp horizontal curvature in the major street alignment;  Exclude sites that have work zones present during any portion of the study period; and  Exclude sites for which major changes have occurred to the site during the study period. Volume Criteria. Transit stop frequency and truck volume data should be available for the selected sites. These volume data should be suitable to describe typical daily demand levels for the subject site during the five-year study period. Major street annual average daily traffic (AADT) volume will also need to be available for each study site. Unsignalized access point volume is not required for site selection. Its effect was indirectly accounted for in the data analysis using surrogate variables (e.g., adjacent land use, access point spacing). Other Criteria. In addition to the aforementioned criteria, the sites were selected to ensure that they collectively include a range of values for key geometric design elements. These elements include:  About 1/2 of the sites should have adjacent on-street parking (and the other 1/2 of sites should not have on-street parking),  About 1/2 of the sites should have a mid-segment transit stop,

218  One-half of the sites should have a TWLTL and the other 1/2 should have a NTM. Study Design A wide variety of data was planned for assembly. The geometric design and traffic control elements listed in Table 114 were obtained using aerial photographs available from the internet. Traffic Characteristics Data. The traffic characteristic data were obtained from local transportation agencies. The major street AADT volume was obtained for each year of the study period. Interpolation was used to estimate the AADT volume for intermediate years for which AADT data were unavailable. Representative transit stop frequency and truck volume estimates for the average day were obtained for each year of the study year. Some processing of the data provided by the local agency was required by the researchers to obtain these estimates. Crash Data. Electronic copies of the crash reports for the street associated with each study site (including boundary intersections) were requested from the local agency. Copies were requested for all transit-related, truck-related, vehicle-pedestrian, and bicycle-pedestrian crashes that occurred during the study period. The crash reports included fatal, injury, and property-damage-only crashes. The pedestrian- and bicycle-related crash data was used to validate the safety performance relationships developed from information found in the literature. The attributes in the crash data included crash date, crash time of day, crash type, crash severity, crash location (i.e., milepost), work zone presence, contributing factor, parties involved (i.e., pedestrian, bicyclist, truck, bus, automobile), and intersection relationship (i.e., driveway, intersection, other). Data Reduction. Data reduction procedures were developed and documented. All data reduction staff were trained using these procedures to ensure consistency in the data extracted from the aerial photographs. Crashes associated with a given site (i.e., street segment) were extracted from the electronic crash data provided by the local agency. Site-based crashes were identified as those that are not related to the boundary intersections. Simulation Model Calibration The development of performance relationships for several access management techniques was based on data from two traffic simulation software products. This section describes the activities undertaken to calibrate these simulation products. The VISSIM simulation model was used to generate the data needed to develop the desired operations- based performance relationships. VISSIM has the capability of simulating cars, pedestrians, bicycles, transit, and trucks. The Surrogate Safety Assessment Model (SSAM) was used to generate the surrogate safety measures needed to develop the desired safety-based performance relationships. These measures were extracted from the vehicle trajectory data output from VISSIM. Site Selection for Simulation Calibration A small number of sites were identified to provide the data needed for model calibration. These sites include a combination of signalized intersections and street segments (bounded on each end by a signalized intersection). Arterial street segment sites were selected for the TWLTL vs. non-traversable median technique. Signalized intersection sites were selected for the right-turn deceleration technique. All sites are in urban areas.

219 Site Sample Size Consideration of sampling statistics and project resources indicate that calibration data were needed for at least eight sites. Four of the sites would be urban arterial street segments (one site is one direction of travel). Four of the sites would be urban signalized intersections (one site is one intersection approach). The signalized intersections that bound the street segment were sometimes used to develop the calibration parameters for the right-turn deceleration technique. The collected data describe representative design and traffic conditions at each study site for two one- hour field study periods. Site Selection Considerations Criteria to Ensure Applicability to Techniques Being Evaluated The segment sites for the TWLTL vs. non-traversable median technique were selected to have either a non-traversable (i.e., restrictive) median or a two-way left-turn lane median. Also, the set of segment sites include at least one site with a transit stop at a mid-segment location. The segment sites include bicycle, transit, and truck traffic. The sites were not specifically selected to include pedestrian traffic because existing performance relationships are available for pedestrian operations and safety for this technique. All sites were known to have moderate to high levels of bicycle, transit, and truck traffic volume. All of the intersection sites had signal control. Only one-half of the sites had a right-turn deceleration lane on a major street approach. Similarly, the set of sites had some locations with a driveway adjacent to the major street and in the functional area of the intersection; other sites had no driveways in the functional area. At least one intersection had a transit stop located on the approach to the intersection (i.e., a near side stop). All intersection sites had moderate to high levels of pedestrian, bicycle, transit, and truck volume. Geographic Diversity The collective sites represented in the database were selected to have geographic diversity to ensure transferability of the findings. This diversity was achieved by having a portion of the sites located in Northwestern U.S. and the others in the Southeastern U.S. The sites were located in larger urban areas to ensure adequate pedestrian and bicycle sample sizes. Crash Data Availability The selected sites were used to investigate the relationship between crash frequency and surrogate measure frequency for the Right-Turn Deceleration technique. Thus, the intersection sites were selected such that their crash history (including fatal, injury, and property-damage-only crashes) is confirmed to be available. The need for crash data is discussed in the section titled SSAM Simulation Model Calibration. In addition, each site’s crash history was checked to ensure that it was not influenced by secondary factors, as identified in the following list:  Exclude sites that have work zones present during any portion of the most recent five-year period; and  Exclude sites for which major changes have occurred to the site during the most recent five-year period.

220 Study Design VISSIM Calibration Field Data Description Field data were collected at the selected sites for the purpose of calibrating key parameters in the simulation model. One global set of calibration parameters was obtained from the collective set of sites. The focus of this data collection was on the data needed to calibrate the pedestrian, bicycle, transit, and truck travel modes. In VISSIM, a “vehicle” can be defined as an automobile, pedestrian, bicycle, bus, truck, motorcycle, etc. Hence, any reference to “vehicle” as related to VISSIM can mean any of these travel modes. The data collected at each of the calibration sites is listed in Table 117. The top half of the table identifies “site description” data that describe the traffic characteristics, geometric design, signal operation, and traffic control devices present at each site. The traffic characteristics, cycle length and green interval duration, transit operation, and calibration data were collected during the field study. The geometric design, traffic control device, and some signal settings were collected just before or after the field study. Table 117. Field data collected at each calibration site – VISSIM calibration. Application Category Data Element Data Need by Travel Mode Data Collection Approach Auto Ped Bicycle Transit Truck Segment and intersection description Traffic characteristics Midblock speed distribution Yes Yes Yes Yes Yes Video Midblock free-flow speed Yes No No Yes Yes Sensors Turn movement volume Yes Yes Yes Yes Yes Video Mid-segment 48-hour volume distribution Yes No No Yes Yes Sensors Parking maneuver frequency Yes Video Geometric design Number of lanes by movement Yes Yes Aerial/CAD file or field measurement Width of each lane Yes Yes Aerial/CAD file or field measurement Distance to all driveways near intersection, or along street (if segment) Needed in general (independent of mode) Aerial/CAD file or field measurement Width of driveways Needed in general (independent of mode) Aerial/CAD file or field measurement Presence of on-street parking Needed in general (independent of mode) Aerial/CAD file or field measurement Median type Needed in general (independent of mode) Aerial/CAD file or field measurement Right-turn radius and island presence Needed in general (independent of mode) Aerial/CAD file or field measurement Turn bay length Needed in general (independent of mode) Aerial/CAD file or field measurement Signal Average green duration by phase Needed in general (independent of mode) Video a Average cycle length Needed in general (independent of mode) Video b Change interval by phase Needed in general (independent of mode) Video or timing sheets Left-turn mode by approach Needed in general (independent of mode) Video or timing sheets

221 Application Category Data Element Data Need by Travel Mode Data Collection Approach Auto Ped Bicycle Transit Truck Right-turn-on-red allowance Needed in general (independent of mode) Video or timing sheets Control device Major street speed limit Needed in general (independent of mode) Field observation Additional segment description Traffic characteristics Through volume (by direction) Yes Yes Yes Yes Yes Video Turn movement volume at each driveway Yes Yes Yes Yes Yes Video Mid-segment crossing volume No Yes No No No Video Geometric design Segment length Needed in general (independent of mode) Aerial/CAD file or field measurement Number of though lanes along segment Needed in general (independent of mode) Aerial/CAD file or field measurement Proportion of street with parking Needed in general (independent of mode) Aerial/CAD file or field measurement Transit operation description Transit stop Dwell time distribution Yes Video Route operation Transit headway Yes Video Calibration Performance measures Queue length (for intersections) Yes Yes Yes Yes Yes Video Average travel speed (for segments) Yes Yes Yes Yes Yes Video Driver behavior parameters Look-ahead distance No Yes Yes Yes Yes VISSIM default Observed vehicles No Yes Yes Yes Yes VISSIM default Look-back distance No Yes Yes Yes Yes VISSIM default Minimum headway No Yes Yes Yes Yes VISSIM default plus calibration Safety distance reduction factor No Yes Yes Yes Yes VISSIM default plus calibration Vehicle parameters Vehicle length No Yes Yes Yes Yes VISSIM default Vehicle acceleration No Yes Yes Yes Yes VISSIM default Vehicle deceleration No Yes Yes Yes Yes VISSIM default Gap acceptance No Yes Yes No No Video Compliance with signal or sign No Yes Yes No No Video Turn speed (for intersections) No Yes No Yes Video Travel path choice No Yes Yes No No Video Validation Safety performance measures Field-observed conflicts No No No Yes Yes Video Crash reports (5 years) Yes Yes Yes Yes Yes Crash reports from agency Notes: a – If the signal operates as pretimed, then the green duration can be obtained from controller timing sheets. b – If the signal operates with a fixed cycle length, then this cycle length can be obtained from timing sheets. During the field study of one site, the desired data were recorded using a combination of video cameras and pavement sensors. The study took place during one or more peak hours and during one or more off- peak hours at each site. Where possible, the data categorized as “calibration” data was extracted from the video tape (during replay) or reduced from the electronic sensor data files. These data were used to calibrate the vehicle and driving behavior parameters for the pedestrian, bicycle, transit, and truck travel modes. The phrases used to identify the calibration data elements are taken from the VISSIM User Manual (PTV, 2011). The definition of each element is provided in this manual.

222 Subject Driveway 1 Major Street Camera 1 Bike Lane Camera 2 Subject Driveway 2 Bike Lane Bike Lane Intersection 1 Camera 4 Bike Lane Bike Lane Intersection 2 Camera 3 Sensor 1 Sensor 2 Sensor 3 Sensor 4 For a given travel mode, one “vehicle” class was established for each of the parameter sets established. A speed distribution was obtained from the data for each of these classes. Data collected in the field (or obtained from previous research) were used to define the length, acceleration, and deceleration for each vehicle class. For the transit mode, a dwell time distribution and a passenger location distribution were measured, if appropriate, using field data (or similar from previous research). Additionally, the transit headway, occupancy, boarding time, and alighting time as measured for each transit stop. Data Collection Data collection for the simulation calibration was focused on urban street segments with one or more mid-segment driveways, and bounded by two intersections. Data collection relied on video-based observations, with cameras positioned in a way that they capture the various traffic operational and signal timing variables listed in Table 117. In addition, pavement sensors were needed to gather a 48-hour volume count. This count was used to extrapolate peak hour turning movement counts obtained from video to other times of day. Figure 7 shows a schematic of the setup envisioned for this data collection. Figure 8 and Figure 9 show enlarged views of the intersection and driveway camera configuration, respectively. In addition, aerial photographs available from the Internet were used to obtain geometric variables. Figure 7. Camera locations along urban street for simulation calibration – VISSIM calibration.

223 Camera 4 Bike Lane Bike Lane Intersection 2 Sensor 4 Subject Driveway 2 Bike Lane Bike Lane Camera 3 Sensor 3 Figure 8. Camera location for intersection data collection – VISSIM calibration. Figure 9. Camera location for driveway data collection – VISSIM calibration. Traffic Characteristics Data. The traffic characteristic data listed in Table 117 were measured in the field at each site during the study period. Prior research on surrogate safety assessment from microsimulation has demonstrated that a focus on peak hour operations is sufficient to obtain conflicts for peak travel periods, which can then be correlated to historical daily crash patterns. Accordingly, the focus of the field data collection was on one two-hour peak period for each day. The selection of which peak period to study was based on expected volumes of pedestrians and bicycles on each specific study corridor. The traffic characteristic and performance measure data were collected using several camcorders. One camcorder was positioned at each of the boundary intersections, as well as at each mid-segment driveway along the corridor, for each direction being studied.

224 Figure 7 illustrates this setup with two mid-segment driveways for one direction. Additional cameras were needed for longer sites. The intersection and driveway cameras were positioned at an overhead location to provide a vantage point that captures all vehicle and multimodal movements. For large intersections, supplemental cameras were sometimes used to ensure that all movements were easily captured. The desired data were extracted from the videotape recordings during replay in the office. Before this post-processing, the various video recordings were synchronized and combined into a single, split-screen video for evaluation. The synchronized video angles were needed to extract travel time information as vehicles, buses, pedestrians, and cyclists traverse the corridor. A video multiplexer or multi-channel DVR was used to achieve this synchronization. For the in-field recording, it was critical that a common synchronization point was identified on the videos (such as a person waving a flag). Study Procedures. The camcorders were used to record traffic conditions at each site during a two-hour study period. Each study period occurred during a time period for which the pedestrian and bicycle delay is at its highest level during the typical day. This time period typically coincided with the hours of peak pedestrian volume, peak bicycle volume, and moderate-to-high motorized vehicle volume. The study periods occurred during typical weekday daylight time periods. No data were collected during inclement weather, during nighttime hours, or when traffic flow was disrupted by incidents, work zone activity, or special events. Data Reduction. Data reduction procedures were developed and documented. All data reduction staff were trained using these procedures to ensure consistency in the data extracted from the video recordings and from the aerial photographs. Model Calibration Once the data were collected for the identified sites, they were used to determine the appropriate values for selected calibration parameter in the VISSIM model. The objective of the calibration activity was to identify parameter values that are logical and provide an acceptably good fit to the observed performance measures. The guidelines developed by Dowling et al. (2004) were used to assess the quality of fit between the observed and predicted measures. In some instances, the video and sensor data were not adequate to extract some of the identified calibration data elements. In this situation, the performance measure data that were collected were used as an indirect means of calibrating these “unmeasurable” parameters. Specifically, the calibration parameter value was adjusted such that the predicted performance matched the observed performance. In this manner, it was assumed that the adjusted calibration parameter was in reasonable agreement with its true value (as would have been determined if it could have been measured during the field study). SSAM Calibration The sites used to provide VISSIM calibration data for operations relationships were also used to obtain the SSAM calibration data. The criteria used to select these sites and the data collected are described in the previous subsection titled Site Selection for Simulation Calibration. The process for calibrating the VISSIM model was described in the previous subsection. The focus of discussion in this subsection is on the calibration of the SSAM software tool. Safety simulation is planned only for the evaluation of the right-turn deceleration technique. For this technique, performance relationships exist in the literature for the pedestrian and bicycle modes (see the discussion associated with Table 109). As a result, the focus of the SSAM calibration process was on transit and truck safety. Pedestrian and bicycle safety was not evaluated using simulation.

225 Simulation Data Description The field data needed to calibrate the SSAM software tool are identified in Table 118. These data include four different types of conflict for each of the truck and transit travel modes of interest. The conflict data were extracted from the video recordings made during the field studies conducted to obtain the VISSIM calibration data. Table 118. Supplemental conflict data collected in the field at each calibration site. Application Category Data Elementa Data Need by Travel Mode Auto Ped Bike Transit Truck Calibration Performance measures Right turn, same-direction conflict with... No No No Yes Yes Slow vehicle, same-direction conflict with... No No No Yes Yes Lane-change conflict with... No No No Yes Yes Right-turn, cross-traffic from right conflict with... No No No Yes Yes Right-turn, cross-traffic from left conflict with... No No No Yes Yes Opposing right-turn-on-red conflict with... No No No Yes Yes Right-turn into driveway conflict with... No No No Yes Yes Note: a – Conflict definitions developed by Parker and Zegeer (1989). Crash Data The reported crash history was obtained for each of the intersection sites from the local transportation agency. These data describe all crashes (including fatal, injury, and property-damage-only crashes) in the most recent five consecutive years at each site. Only those crashes that involve a bus or truck were requested. A copy of the crash diagram and narrative was also requested for each crash of interest. These data were reviewed to confirm the association of each crash with the subject site. Annual traffic volume variation characteristics representative of each intersection site were also obtained from the local planning agency. Specifically, the annual average daily traffic (AADT) volume, annual truck volume, and annual local bus volume were requested. The hourly volume distribution for the typical day, day of week, and month of year adjustment factors was requested. Model Calibration SSAM was not developed to specifically identify pedestrian, bicycle, transit, or truck conflict frequency. However, the VISSIM output files were screened to include only those trajectories that (1) potentially conflict and (2) involve a “vehicle” type (e.g., pedestrian, bicycle, bus, or truck) of interest. The screened output file was submitted to SSAM to obtain the time-to-collision (TTC) and post- encroachment time (PET) parameters. These parameters were then used to identify the desired conflicts. The TTC is defined as the time distance to a collision of two road users, assuming that their travel direction and velocity is unchanged. PET is defined as the period of time from first road user’s departure from the conflict area to the second road user’s arrival at the conflict area. The TTC and PET parameter thresholds were adjusted to calibrate SSAM. To calibrate the TTC and PET parameters, the trajectory files were submitted to SSAM using specified values of TTC and PET. The predicted conflict frequency was then compared with the frequency obtained from field data on a

226 site-by-site basis. The process was repeated for a range of values for TCC and PET. The pair of values providing the best agreement between the predicted and observed conflicts was used to calibrate SSAM. Assessing Relationship of Surrogates to Safety A surrogate is considered valid for safety-based decisions if it is shown to be correlated to crash frequency or severity. The conflicts extracted from the video recordings were compared with the reported crash frequency on a site-by-site basis. A statistical analysis was used to quantify the correlation between the selected surrogates and crash frequency or severity. When a plausible correlation was found, then there it served as support for the use of surrogate measures to assess the relative safety of various techniques. Moreover, it also served as support for using the field-measured conflict data to calibrate the SSAM software tool. References Dowling, R., A. Skabardonis, and V. Alexiadis. (2004). Traffic Analysis Toolbox – Volume III: Guidelines for Applying Traffic Microsimulation Software. Report No. FHWA-HRT-04-040. Federal Highway Administration, Washington, D.C. Parker, M., and C. Zegeer. (1989) Traffic Conflict Techniques for Safety and Operations – Engineer’s Guide. Report No. FHWA-IP-88-026. Federal Highway Administration, Washington, D.C. PTV. (2011). VISSIM User Manual. Version 5.30. Planung Transport Verkehr AG, Karlsruhe, Germany.