Bicycle Environmental Quality Index (BEQI) Draft Report

Transcription

1 Bicycle Environmental Quality Index (BEQI) Draft Report Program on Health, Equity and Sustainability Environmental Health Section San Francisco Department of Public Health 1390 Market St., Suite 822 San Francisco, CA

2 I. INTRODUCTION Over the past several decades transportation planning has been predominately focused on the motor vehicle, leaving very little funding and thought for alternate forms of transportation, including bicycling. With rising gas prices, climate change, record high obesity rates, decreased air and noise quality, and a struggle for pedestrian safety, there is a growing need for significant investment in planning around bicycle and pedestrian facilities to promote walking and bicycling. Many people experience considerable challenges riding their bicycles in an urban environment. There are several health benefits, however, associated with using a bicycle as a form of transit and recreation verses using a motor vehicle. Bicycle riding as a means of transportation can reduce private automobile trips that result in reduced pollution and injuries. Cycling to work, school, or even baseball games as part of ones regular daily routine can be both a sustainable and time-efficient exercise regimen for maintaining acceptable levels of fitness. Bike lanes and better bicycle infrastructure can help to better organize the flow of traffic and reduce the chance that motorists will stray into cyclists path of travel. With more and more people using bikes for transportation, the need for streets to be safer for bicycling is clearly visible. More recently research has been conducted to determine if increasing the number of bicyclists can decrease the number of bicycle-motor vehicle collisions. A study by Jacobsen (2003) showed collisions between motorists and bicyclists decrease where more people are bicycling. The author concluded motorists adjust their behavior when surrounded by larger numbers of bicyclists and pedestrians. The San Francisco Department of Public Health (SFDPH) has created the Bicycle Environmental Quality Index (BEQI) to assess the bicycle environment on roadways and to evaluate streetscape conditions which promote bicycling in the city. Use of the BEQI can translate environmental variables into a set of provisions for a healthy bicycle environment and a BEQI assessment can inform neighborhood planning and prioritize improvements through the land use plans and environmental assessments. BEQI results reveal the relative quality of the biking environment at a street-level scale in San Francisco neighborhoods. II. RATIONALE AND PREVIOUS RESEARCH In the past 20 years there have been several attempts to create models to assess bicycle compatibility with roadway environments. Existing models rate bicycle routes for suitability, comfort, and safety. A variety of bicycle indices and their indicators were reviewed to determine what indicators would be included in SFDPH s BEQI (see Table 1 for complete bicycle indices matrix). The Bicycle Environmental Quality Index draws on research from multiple sources on the assessment of the environmental determinants of cycling, as well as bicycling work done in other cities. The San Francisco Department of Public Health (SFDPH) has developed empirical measures that either promote or discourage bicycle riding and connectivity to other forms of sustainable travel. The BEQI has a total of 22 indicators. Several of the indicators have been used in other Bicycle Indices from different regions in the country, while others are new concepts found significant through other studies regarding safe bicycle environments. We identified five main bicycle categories which represent important physical environmental factors: intersection design, street design, vehicle traffic, safety, and adjacent land use. Table 2 details each BEQI indicator under its broader environmental domain. These indicators can be

3 aggregated to create the final index (the BEQI), which can be reported as an overall index, and/or deconstructed by bicycle environmental factors shown below. In addition, a detailed field and technical manual was created with instructions on how to conduct the survey and communicate the results using Geographical Information Systems (GIS) (see Appendix 1). There were several indicators in other studies that are not captured in the SFDPH BEQI and need additional research. Those indicators are width of outside lane and presence of right turn lane. Most studies measured the width of the outside lane/curbside lane in addition to bicycle lane width. Many streets do not have marked bicycle lanes and there will need to be a way to determine if a bicyclist has enough room to ride safety next to motor vehicle lanes or adjacent parking if applicable. The presence of a right turn lane was used in three models that were reviewed and was found as a significant factor in bicycle safety at intersections. The summary below briefly describes the important findings that provide evidence for the SFDPH BEQI for each of the five domains.

4 Table 1. Bicycle Environmental Quality Indices Matrix SFDPH BEQI Indicators Index BSIR RCI IHS BLOS BCI Bike ISI Stress Level Dade BFP Bicycle Lane Markings Bicycle Lane Slope Bicycle Parking Bicycle/Pedestrian Scale Lighting Connectivity of Bicycle Lanes Dashed Intersection Bicycle Lane Driveway Cuts X X X X Left Turn Bicycle Lane X Line of Sight X X No Turn on Red Sign(s) Number of Vehicle Lanes X X X X Parallel Parking Adjacent to Bike Lane/Route X X X X X X Pavement Type/Condition X X X X X Percentage of Heavy Vehicles X X X X X Presence of a Marked Area for Bicycle Traffic X X X Presence of Bicycle Lane Sign Presence of Trees Retail Use X Traffic Calming Features Traffic Volume: Average Number of Vehicles Per Day X X X X X X X X Vehicle Speed X X X X X X X X Width of Bike Lane X X X X X X X Not in SFDPH BEQI Angled Parking X X Intersection Crossing Distance X Paved Shoulder X X X X Presence of Buses X Right Turn Lane X X Traffic Signal X Vehicle Lane Width X X Width of Outside Lane X X X X X Bicycle safety index rating (BSIR), Davis 1987; Florida Roadway Condition Index (RCI), Epperson 1994; Interaction Hazard Score (IHS), Landis 1994; Bicycle Level of Service (BLOS), Landis et al. 1997; Bicycle Compatibility Index (BCI), Harkey et al. 1998; Bicycle Intersection Safety Indices (Bike ISI), Carter et al. 2006; Stress Level, Sorton and Walsh 1994; and Dade County Bicycle Facilities Plan (Dade BFP), Dade 2001.

5 Table 2. BEQI Indicators by Bicycle Environmental Domain INTERSECTION STREET SEGMENT Intersection Safety Vehicle Traffic Street Design Safety Land Use Left Turn Bicycle Lane Number of Vehicle Lanes Presence of a Marked Area for Bicycle Traffic Bicycle/Pedestrian Scale Lighting Line of Sight Dashed Intersection Bicycle Lane No Turn on Red Sign(s) Vehicle Speed Width of Bike Lane Presence of Bicycle Lane Signs Bicycle Parking Traffic Calming Features Bicycle Lane Markings Retail Use Parallel Parking Adjacent to Bicycle Lane/Route Traffic Volume Percentage of Heavy Vehicle Trees Connectivity of Bicycle Lanes Pavement Type/Condition Driveway Cuts Street Slope 2.1 Intersection Design It is well known that a high percentage of bicycle-motor vehicle collisions occur at intersections, which has led to an increase in research to determine how to better design intersection for bicycle safety. Hunter et al. published further results in 1996 with a study of more than 8,000 pedestrian and bicycle crashes from six states. The results provided a summary of the distribution of crash types experienced by pedestrians and bicyclists. Nearly 40% of these crashes occurred at intersections. Some of the most frequently occurring bicycle crash types are listed in Table 2. Table 2. Bicycle Crash Types and Percent Frequency Bicycle Crash Type Percent Frequency A motorist failing to yield 21.7% A bicyclist failing to yield at an intersection 16.8% A motorist turning or merging into the path of the bicyclist 12.1% A bicyclist failing to yield at a mid block location 11.7% A motorist overtaking a bicyclist 8.6% A bicyclist turning or merging into the path of the motorist 7.3% (Hunter et al 1996)

6 A 2006 report published by the Federal Highway Administration discusses models of pedestrian and bicyclist safety at intersections. One of the models is the Bicycle Intersection Safety Indices (Bike ISI), which is intended to give intersections relative rankings according to bicyclist safety so that high-risk intersections can be prioritized for improvements (See Table 1 for specific measures used). The Bike ISI variables were found to be significant determents of level of safety for bicyclists and can be used to prioritize intersections for safety and comfort countermeasures. At intersections where there are high volumes of right turns made by vehicles, a separate right-turn lane should be incorporated, where the bicycle lane would continue straight and to the left of the turn lane (AASHTO, 1999). Right-turn lanes and left-turn lanes are considered a main bicycle-motor vehicle crash type, accounting for approximately 12% of the collisions over a period of five years (North Carolina Department of Transportation, 2005). The Bicycle Safety Index Rating (BSIR) combined a street segment and intersection model as one of the first models to relate safety to physical and operational roadway features (David, 1987). The presence of a right-turn lane, a geometrics factor in the model, was perceived to be the greatest safety hazard at the intersection. A study in Orlando, FL found that 2.6% of crashes occurred when bicyclists attempted to turn left. These crashes contribute to 10.7% of crashes occurring when bicyclists are traveling on the roadway with the flow of traffic (Metropolitan Orlando, 2004). It is only recommended to use dashed bicycle lanes through complex intersections where there is high traffic volume and/or high vehicle speeds (AASHTO, 1999). The Draft San Francisco Bicycle Plan (2006) states bike lanes associated with a busy intersection can have a dashed marking to guide bicyclist and motorist through an intersection. Note, for the average intersection, dashed bicycle lanes are not necessary and therefore the indicator should not be counted against those intersections. Efforts are underway to correctly weight this indicator in the BEQI based on traffic volume and number of lanes Vehicle Traffic Fifty percent of bicycle fatalities occur on roadways with posted speed limits greater than 35 mph, confirming speed limits and travel speed of vehicles to be a reasonable indicator of the design of roadways (Davis, 1987). The number of bicycle-motor vehicle collisions is even greater; three-fourths of such collisions occur where vehicle speed limits are 35mph or higher (Hunter et. al., 1996). Landis (1994) Interaction Hazard Score (IHS) concluded that 79% of the model s value was attributed to vehicle speed. High speed limits, combined with traffic volume, number of through vehicle lanes, and vehicle type (i.e. heavy vehicles) pose the greatest risk to bicyclists on roadways. The Bicycle Level of Service Model (Landis et al., 1997) demonstrated a correlation (R 2 ) of.73, connecting traffic volume, vehicle speed, and number of vehicle lanes with collisions, results which are transferable to the vast majority of U.S. urban areas. Several studies also include percent heavy vehicles as a variable to determine bicycle safety and bicyclist s level of comfort (Epperson, 1994; Harkey et al., 1998; and Dade, 2001). Streets with higher traffic volume are more likely to have a collision between a bicyclist and motor vehicle (SFDPT, 2004). Traffic volume increases the risk of pedestrian, cyclist, and motorist injury and death (Ewing et al, 2006). Bicyclists feel more comfortable and have less stress when there is a lower traffic volume (Harkey et al., 1998). In 1994, Sorton and Walsh used the Bicycle Stress Level concept to relate cyclists perceptions of roadway conditions to index variables. The concept of Stress Level was developed on the assumption that bicyclists want to minimize the physical effort required when choosing a roadway on which to ride, but they also want to minimize mental effort, or stress, that results from conflicts with motor vehicles. The results indicated different types of bicyclists (based on experience) can recognize

7 variation in traffic volume, motor vehicle speed, and lane width and that these differences are consistently reflected in their comfort or stress level. Decreased traffic volume will result in an effective safety measure for bicyclists, where most bicyclists will choose a street to ride on with less traffic volume (NCDOT, 2005). There is an ongoing effort to lower vehicle speeds to provide an adequate sense of safety for bicyclists. Traffic Calming Features (TCFs) are one avenue to accomplish this goal. Motor vehicle driver s behavior at non-signalized intersections found that speed-reducing measures, such as speed bumps, elevated bicycle crossings, and stop signs, help drivers to recognize bicyclists early and properly (Summala et al., 1996). A recent study (Jacobsen, 2003) explains that traffic calming is intended to lower motor vehicle speed, reducing bicycle-motor vehicle collisions; head injuries more than any other injury are reduced by reducing these collisions. Decreasing the actual or apparent size of a street with TCFs (e.g., streets continuously lined with trees) will also help reduce vehicle speed and therefore, the driver s vision and attention to the street border improve where pedestrians and bicyclists are located. Multiple factors can influence and slow down vehicle speed including round-abouts or traffic circles, speed bumps, crosswalk treatments, narrow streets, and extended curbs at intersections (bulb-outs). Reducing vehicle speeds will increase pedestrian/bicycle safety and slightly increase a pedestrian/bicyclists response time if encountered by a vehicle. Most pedestrian/bicyclist fatalities involving vehicles happen at speeds over 25 mph (City of Vancouver, 2002). According to the San Francisco Bicycle Plan (SFBP) (2006), there are certain TCFs not acceptable for certain street types. Arterial streets should not have rumble strips in commercial or residential areas. Arterial streets should also not have speed humps, chicanes, traffic circles on most streets, diverters/forced turns, and street closers. Commercial streets should not have rumble strips, diverters/forced turns, or street closers. In addition, local area tracks and school tracks should not have rumble strips. Pavement and crosswalk treatments, speed tables, raised crosswalks, commercial/local/school speed humps, and angled parking all need specific field testing for each area to determine acceptability (SFBP: Policy Framework). Bicycle boulevards should use traffic calming features such as traffic circles, chokers, speed tables, and medians to slow down traffic, as well as speed limits at 20 mph (SFBP: SDG). Vehicle parking also has a significant effect on bicyclists safety. Bicyclist have a high risk of being involved in a bicycle-motor vehicle collision where motor vehicles are exiting or entering on-street parking, either hitting the bicyclist or causing the bicyclist to swerve into the adjacent lane or into the sidewalk curb (Hunter et al, 1999). A bicyclist's level of comfort riding on a street with on-street parking will be lower due to vehicles frequently moving in and out of the spaces. Presence of on-street parking ranks as one of the highest factors in streets being compatible with bicycle lanes and use (Harkey et al., 1998). Given that parking was significant for the behavioral data for the Bike ISI and is known by bicycle researchers to cause potential safety hazards, street parking was included as a variable in the final Bike ISI (Carter et al., 2006). The San Francisco Bicycle Plan (2006) recommends avoiding diagonal parking configurations adjacent to bike routes. Also, parallel parking should only be on one side of the street, where the bicycle lane could be located on the opposite side. This is only feasible if the street is not a high traffic volume/bus route area. As long as there is adequate space (five feet or more ) between parallel parking markings and bicycle lane marking, the two can remain on the same side of the street. Parking lanes are between seven and eight feet depending on parking turnover, traffic volumes, and street widths (SFBP: SDG).

8 2.3 Street Design The use of bicycles is a healthy alternative as a primary mode of transportation, therefore the addition of bike lanes to roadways helps define road space for both vehicle traffic and bicyclists. Bike lanes provide safety for bicyclists by organizing the flow of traffic, which encourages bicyclists to ride in the correct direction and reduces the chance for vehicles to stray into a bicyclist. A study (Moritz, 1998) revealed that serious bicycle injuries occur on streets without bike lanes. Additional bicycle lanes were added in Davis, California which reduced accidents by 31% (Federal Highway Administration, 1995). In addition to bike lanes, shared lane pavement markings can also have a positive impact on motorist and bicyclist behavior, positions, and safety (Alta Planning and Design, 2004). The Bicycle Compatibility Index (BCI) attempted to introduce the bicyclists perspective of safety and comfort into an index of compatibility, which reflects comfort levels of bicyclists on the basis of observed geometric and operational conditions (Harkey et al., 1998). In the model, bike lane presence, wider bike lane, and wider curb lane all had a positive effect on comfort level. Developed in 1991, the Florida Roadway Condition Index (RCI) was based on the BSIRs model, but used only the roadway segment portion of the model and changed relative values of pavement factors and location factors. The RCI presented bicycle lane or paved shoulder with the highest positive value in the model for bicycle safety. The model was used to predict crashes and linked to crash data over a 20-month period (Epperson, 1994). Harkey et al (1998) utilized the video method of Sorton and Walsh (1994), which they validated in a pilot study. The model separated bicycle lanes/paved shoulders from standard wide curb lanes, as well as types of bicyclists as casual users and experienced users. Mean comfort level for casual bikers was higher than others; experienced users were therefore less comfortable. The variable with greatest effect was absence of a bike lane. Bicycle lane width is also a vital issue because it increases the level of service for bicyclists and provides a defined space. The proposed minimum width varies by report, but typically is no less than four feet for a bicycle lane with no curb and gutter. In most cases, it is recommended a bicycle lane should be five feet, leaving room for a one to two-foot gutter which would allow at least three feet for the bicycle lane itself. If there is parallel parking on the street, the optimal bicycle width is five feet. The width of the bicycle lane should be at least 12 feet in situation where there is parallel parking with no marked parking stripes, and feet if there is high parking volume and/or turnover (AASHTO, 1990). The majority of other bicycle indices researched include bicycle lane width to determine bicycle safety and/or level of comfort (See Table 1). Different types of bicyclists (based on experience) can recognize variation in lane width and these differences are consistently reflected in their comfort or stress level (Sorton and Walsh, 1994). In order for bicycle lanes to be effective, the lane should be continuous from street to street, displaying connectivity throughout the bicycle route (NCDOT, 2005). It is important for bicyclists to have a high level of comfort on the road, therefore a bicycle network should be implemented and continuous for bicyclists to be able to travel between destinations and residential neighborhoods (Dill and Carr, 2003). Bicycle routes should be interconnected and not end suddenly with barriers (AASHTO, 1999). Bicycle lane markings can improve a bicyclist s experience. There are different varieties of bicycle lane markings to choose from. The presence of a double striped bike lane allows enough room for parked motor vehiclists to exit the vehicle without contacting a bicyclist (Hunter et al, 1999). A study (Landis et al., 1997) explains that lane stripping increases the quality of

9 service for bicyclists by 31%. Dashing the bike lane stripe at busy driveways is also recommended, not only to alert a motorist that a bicyclist may be approaching because of the presence of the bike lane but also to alert a bicyclist that a motorist may be emerging from the driveway adjacent to the dashed stripe (Hunter et al., 1999). Most bike indices did not include the type of bicycle lane marking, only focusing on whether there was an area designated for bicyclists to ride. Another variable in promoting safe bicycle facilities is the condition of the pavement bicyclists ride on; a smooth and uniform pavement surface is desired. A bicyclist can lose control if wide cracks, joints, or drop-offs along the edge of a bicycle lane are present. In addition, potholes or bumps in the road surface could cause a bicyclist to swerve and/or merge into traffic (AASHTO, 1999). Street paving is prioritized in San Francisco by three surface features: cracking, raveling (erosion), and motor vehicle ride quality (SFBP, 2006). Pavement conditions, especially major obstruction such as potholes, should be addressed if the roadway is a bicycle network. Bicyclists are negatively affected when pavement conditions change suddenly, especially with no forewarning (AASHTO, 1999). The Interaction Hazard Score (IHS) concluded that 13% of the model s value was attributed to pavement conditions. These values were in accordance with cyclist perceptions (Landis, 1994). The model did not specifically define exposure or explain validation, so subjectivity remained an issue. However, Landis and colleagues later found a statistically significant correlation between pavement condition and a bicyclist s level of service (Landis et al., 1997). A high number of bicycle-motor vehicle conflicts occur when motor vehicles are exiting and entering driveways (Hunter et al., 1999). A City of Vancouver study (2002) recommended that the number of driveways should be limited on each street due to pedestrian/vehicle collisions where the driveway and sidewalk intersect. The BCI (Harkey et al., 1998) included vehicles turning right into driveways as a variable and identified it as being important to a bicyclist s comfort level. Driveway cuts were included as a traverse roadway environment variable in the HIS and combined with on-street parking presence, representing 8% of the model s value (Landis, 1994). Another challenge bicyclists can face is the slope or grade of the street, especially if a bicycle lane is present. Bicyclists prefer street grade at 5% or less because most riders find it difficult going up hill and are uncomfortable with increased speeds associated with downhill grades. It is recommended that grades larger than 5%, a maximum distance of the grade be associated with it. For example, the grade should be no more than 50 feet with an 11% grade or 200 feet with a 9% grade (AASHTO, 1999). If a street is steeper than 15%, alternate bicycle routes in San Francisco are chosen with a lower grade. Although the standard is less than 15%, the lowest possible grade is always a priority (SFBP, 2006). As mentioned in the Vehicle Traffic section, traffic calming features are beneficial to bicyclists, potentially reducing motor vehicle speeds. Tree-lined streets are a well known and used traffic calming feature, aiming to reduce the speed of motor vehicles by giving a sense the road is narrower than it appears (Massachusetts Highway Department, 2006). Dumbaugh s (2005) study shows decreased vehicle speeds when street trees are present. In San Francisco, street trees are a recommended and acceptable traffic calming feature for all newly constructed and reconstructed streets including arterial streets, commercial streets, local streets tracks, and school area tracks (SF Bike Plan, 2006).

10 2.4 Safety Many individuals only mode of transportation is a bicycle and therefore these bicyclists are forced to ride at night (e.g., commuting to and from a job). Thirty-one percent of bicycle fatalities occur between 5p.m. and 9p.m. (National Highway Traffic Safety Administration, 2003), where there is limited or no light available for bicyclists. In addition, a study in Orlando, FL found approximately 22% of crashes occurred during the night, demonstrating the importance of proper bicycle/pedestrian lighting (Metropolitan Orlando, 2004). Unlit streets can be dangerous because there is potential a motorists will not see the bicyclist, which could result in a collision and/or injury. Even with bike headlights and flashers, the motorist cannot always see these features until they are very close to the bicyclist (AASHTO, 1999). Presence of pedestrian and bicycle lighting provides necessary visibility and safety for bicyclists by allowing them to see obstacles and faults in the pavement which could cause a collision. Lighting standards should be set at a pedestrian level compared to street lighting for vehicles, and have horizontal illumination levels of 5 lux to 22 lux. Levels should be raised if there is a safety issue at certain areas or intersections, as well as underpasses or tunnels (AASHTO, 1999). Bicycle/pedestrian lighting is an important component in a bicycle network. The San Francisco Bicycle Plan (2006) states ground-level bicycle lighting must be present at all bicycle parking areas. Motor vehicle drivers are not hesitant to use or cross over a bike lane when exiting a building or turning at an intersection. A Yield to Bicyclists sign will inform the driver to slow down and be cautious of bicyclists (Hunter et al., 1999). In addition to having a bicycle lane sign present, it is recommended to have an "Ahead" sign mounted right below the bicycle lane sign before the bicycle lane begins. The same is true for an "End" sign before the bicycle lane ends. To remind motorist to lookout for bicyclists, a "Share the Road" sign can be mounted below the bicycle lane sign (USDOT, 2004). There are a few standard bicycle signs used in San Francisco that are recommended but not approved by the Manual of Uniform Traffic Control Devices. These include Local Bicycle Route Sign, Cross Town Bicycle Route Sign, and Bicycle Route Detour Sign. There are also bicycle-specific signals which inform bicyclists which streets are appropriate to ride on (red, yellow, or green bicyclists), and a track crossing warning sign may be used to inform bicyclists of train/trolley tracks coming up ahead. It is important to note the overuse of these signs can decrease the effectiveness, as well as the street aesthetics. In highconflict areas, colored bicycle lanes can be used, paired with Yield to Bicycle signs warning the bicyclist and motorist (SFBP: SDG). 2.5 Adjacent Land Use Mixed-use land zoning is said to promote more walking and bicycling (Metropolitan Area Planning Council, 2006). The majority of bicycle trips are 2 miles or less, therefore if there is not a mix of uses (e.g., parks, offices, neighborhood retail) near by where people live and work, there will be fewer bicycle trips made (USDOT, 2002). According to the draft Prop K 5-Year Prioritization Program (2006), bicycle facilities are favored if they are intra-connected or interconnected to neighborhoods close to retail shopping, access to transit, and open space. Retail and commercial areas are more likely to attract bicyclists and should be considered throughout bicycle demand analysis (AASHTO, 1999). Land use is a variable included in the HIS (see Table 1), where commercial zoning (i.e. when 30% or more of the land use is used for other purposes than residential or agricultural) was given a high positive value in the model (Landis, 1994). A main reason bicycling is chosen as a main mode of transportation is the presence of bicycle parking. It is important to locate bicycle parking on sidewalks wide enough, not to get in the way

11 of pedestrians walking. It is recommended to put bicycle parking on sidewalks that are at least 10 feet (City of Portland, OR, 2009). The San Francisco Bicycle Plan (2006) suggests placing bicycle racks frequently throughout commercial/retail areas. It is know that people will not leave their bicycles too far from where they are shopping due to high theft rates in the city. Over 1.5 million bicycles are reported stolen in the United States each year and this figure is associated with lack of safe and convenient bicycle parking (Pedestrian and Bicycle Information Center, 2009). It is important to note the San Francisco Planning Code requires bicycle parking in buildings with 10 or more automobile parking spaces; and requires parking for employees and visitors for new and renovated commercial buildings based on square footage. More specifically, if there are up to 120 vehicle parking garage spaces there must be six bicycle parking spaces. If there are more than 500 vehicle spaces in a parking garage there must be 25 bicycle spaces plus one bicycle space per every additional 40 auto spaces, up to a maximum of 50 bicycle spaces. There is currently no specific bicycle parking requirement for residential developments in San Francisco, other than the requirement that applies to all developments of one bicycle parking space for every 20 auto spaces provided. Line of sight is thought to be a variable which improves bicycle safety on the roadway. Bicycle safety is improved if the bicyclist has a long view of oncoming traffic (City of Auburn, 1998). Motor vehicle stopping sight distances are usually measured and can be used and are appropriate for bicycle stopping distances as well. The stopping sight distance depends on the grade of the street and the speed limit. For example, if the speed limit is 25 mph with a zero percent grade, the stopping distance would be 155 feet. As the speed increases, the stopping sight distance will increase as well. If traveling uphill, the distance will decrease, and it will increase for downgrades (Massachusetts Highway Department, 2006). III. METHODLOGY The BEQI was created using the methodology of the Pedestrian Environmental Quality Index created by the San Francisco Department of Public Health, Program on Health, Equity, and Sustainability (SFDPH, 2008). 3.1 Survey Instrument and Data Collection BEQI data is primarily collected with an observational survey (Appendix B) based upon the visual assessment of street segments and intersections by a trained observer (training provided by SFDPH staff). A survey form is completed for each individual intersection and street segment (i.e., the segment of a street between two intersections). We designed the two-page survey as a checklist of closed-ended questions that is fairly simple to use in the field. We also created a Data Collection and Analysis Manual with pictures and detailed instructions for each indicator (Appendix A). The manual includes technical instructions regarding data entry in the customdesigned BEQI Microsoft Access database, geocoding street and intersection data, and preparing maps and presenting results using ArcGIS. Certain indicator data cannot be collected during the surveying, including traffic volume, street slope, and percentage of heavy vehicles. For traffic volume and percentage of heavy vehicles SFDPH uses existing street segment traffic counts from a citywide dataset, created by researchers at the San Francisco Department of Public Health and the University of California- Berkeley and based on San Francisco traffic count data produced and maintained by San Francisco County Transportation Authority (Seto et al., 2007; SFCTA, 2002). Street slope was calculated using GIS 3-D Analysis tools...

12 3.2 Data Entry After the BEQI data is collected for all desired streets and intersections, the data is entered into a database so that indicator responses can be converted into numeric values and then scored. SFDPH created a customized Microsoft Access database for this purpose. Data entry consists of selecting the indicator responses using drop-down boxes, which reduces data entry error. Once data entry is complete, the database is designed to execute a series of programs that apply weights to each indicator response, and calculate the BEQI summary scores for street segments, intersections, and the other domains (see Table 3) when the user selects a specified button. Details regarding the Microsoft Access Database and data entry instructions are included in the BEQI Data Collection and Analysis Manual (Appendix A). 3.3 Scoring Indicator and Indicator Category Scores The values of the indicators were obtained by sending a survey in July 2007 to bicycle experts and members of the bicycle community to determine the level of importance of each bicycle environmental quality indicator. The survey was promoted through the SFBC newsletter and ed to individuals in the bicycle community. Final BEQI scores were determined by n=88 surveys. For each BEQI indicator, we asked respondents two questions: 1) Indicators: Overall importance for bicycle quality. Response options included not important, somewhat important, important, very important, and essential, on a scale from 1-5. We re-scaled the responses to a scale from 1-3 for the final indicator scoring, and weighted each indicator by the median value of its survey response score. 2) Indicator response categories: Relative importance of indicator response categories for bicycle quality. Within each indicator, indicator response categories were assessed on a scale of 1 to 11 (from extremely detrimental to ideal for bicyclists). We weighted indicator response categories by the median value of their survey response scores. In a few cases, indicator response categories required further refinement after the survey responses were received; those scores were informed by survey responses as much as possible. Some indicators were also added to the survey after the scoring survey was completed. In those cases, we assigned the indicator the median weight of all the indicators in its BEQI domain category. Table 3 details the scores for each Indicator and Indicator Response Category, as well as the Indicator Response Category Weighted Scores which are used to calculate the Domain and Overall Street Segment and Intersection Scores, described next.

13 Table 3. The Bicycle Environmental Quality Index (BEQI): Indicator, Domain, and Overall Street Segment Score Values Based on Expert Survey Findings (n= 88) Intersection or Street Segment Assessment Domain (Domain Score Weight a ) Intersection Intersection Safety (0.42) Indicator Indicator Score Indicator Response Category Indicator Response Category Score Indicator Response Category Score, Weighted b Left Turn Bicycle Lane: * * Street Segment Street Design (2.05) Street Segment Vehicle Traffic (1.39) Street Segment Safety (0.42) Dashed Intersection Bicycle Lane: * * No Turn on Red Sign(s) Presence of a Marked Area for Bicycle Traffic: 4.00 Bike Lane w/ Parking Adjacent to Right Bike Lane w/ Sidewalk Adjacent to Right (Without Parking) Bike Lane w/ HOV or Public Transit Adjacent to Right Bike Lane w/ Traffic Lane Adjacent to Right Shared Traffic Lane w/ Sharrow (or Painted Bicycle Marking on Pavement) Combined Bike Lane/Parking Bike Path None Width of Bike Lane: 4.00 < 5 ft ft > 6 ft None Bicycle Lane Markings: 4 One Stripe Left Side of Bike Lane Stripes on Both Sides of Bike Lane None Connectivity of Bicycle Lanes: 4.00 Yes No Pavement Type/Condition: 4.00 Smooth Surface Mild Obstructions (e.g., cracks) Medium Obstruction (e.g raised cracks or pavement) Large Obstructions (e.g., Potholes or Bumps) Street Slope <5% % - 10% % - 15% > 15% Driveway Cuts: 4.00 More Than Five Few (Less than Five) None Presence of Trees: 4.00 Continuously Lined Sporatically Lined None Posted Speed Limit: * * > Traffic Volume - Average Number of Vehicles Per Day: Less than 1, ,000-5, ,000-10, , Percentage of Heavy Vehicle: 3.00 Less than 5% % % Greater than 20% Parallel Parking Adjacent to Bicycle Lane/Route: 3.00 Parallel Parking (PP) < 7ft Parallel Parking (PP) 7ft - 9ft Parallel Parking (PP) > 9ft Time-restricted Parallel Parking (TPP) < 7ft Time-restricted Parallel Parking (TPP) 7ft - 9ft Time-restricted Parallel Parking (TPP) > 9ft None Traffic Calming Features Streets: TCF TCFs TCFs or More TCFs Number of Lanes No Lanes Presence of Bicycle Lane Signs: 4.00 Yes No Bicycle/Pedestrian Scale Lighting: 4.00 Yes - Public Yes - Private Yes - Public and Private No Street Segment Land Use (0.66) Bicycle Parking 4.00 Yes No Retail Use: or More Line of Site 4.00 Line of Sight Obstructed or Compromised Adequate Sight Distance Clear Line of Sight Street Segment Overall (4.25) c a The Domain Score Weight is used to obtain Domain Scores for each of the 5 BEQI Domains. The Domain Score is calculated by adding together all the Weighted Indicator Category Scores in the Domain, and then multiplying by the Domain Score Weight for a maximum Domain Score of 100 in each Domain. b Indicator Category scores are weighted by the Indicator Scores by multiplying the two values. c Combines the four Street Segment Domains to create an overall Street Segment BEQI Score.

14 3.3.2 Overall Street Segment, Intersection, and Domain Scores To create the overall Street Segment, Intersection, and other Domain BEQI scores, we aggregated the Indicator Response Category Weighted Scores and standardized those BEQI summary scores so that the maximum score is 100 by multiplying that summed total by the corresponding Domain or Street Segment (Overall) weight Score Interpretation The survey responses were used to devise numerical scores and weights for the BEQI. The total score for each street segment and intersection will reflect the bicycle quality for the area the BEQI is applied to. The BEQI street segments and intersections both receive a score on a scale from We are currently using the following categories for scoring, a priori, with equal intervals of 20 points for all categories: = highest quality, many important bicycle conditions present = high quality, some important bicycle conditions present = average quality, bicycle conditions present but room for improvement = low quality, minimal bicycle conditions 20 and below = poor quality, bicycle conditions absent 3.4 Mapping and Results Presentation We use ArcGIS to create maps to visually display street and intersection findings for selected areas by spatially joining the BEQI score to its corresponding street segment or intersection using unique street and intersection identifiers. Detailed instructions regarding this process are included in the BEQI Data Collection and Analysis Manual (Appendix A). Maps can be made to represent the overall BEQI Street Segment or Intersection Score, Domain Score, or score for each indicator in the index. The area selected for the study and how results are scored and presented are informed by the research questions as determined by the BEQI users. 3.5 Pilot Testing and Revisions The San Francisco Department of Public Health has applied the BEQI as a pilot application in Lakeshore and Treasure Island. The BEQI pilot test in San Francisco s Lakeshore neighborhood was specifically along Ocean Avenue (bike route 84), Geneva Avenue (bike route 90), and Holloway Avenue (bike route 90) between Highway 1 (19 th Avenue) and Alemany Boulevard. The pilot location and streets were chosen because of the need for bicycle facility improvements on these three main bicycle routes. The San Francisco Bicycle Coalition (SFBC) also has a specific interest in the Lakeshore area and has been working with the Lakeshore community to improve the bicycle facilities. A group of SFBC members volunteered to survey both areas and participated in a BEQI training. See Figure 1 survey results. The pilot resulted in survey revisions, including adding an intersection indicator - No Turn on Red Sign(s) - and an additional indicator measure for pavement conditions to separate out the different levels of severity in pavement cracks/obstructions. Currently, there has been one application of the BEQI conducted: assessment of the physical bicycle environmental condition on Treasure Island, in collaboration with the San Francisco Bicycle Coalition, as a part of a community-based planning effort funded by CalTrans to create a walkable, bikeable Treasure Island. The following section describes the application in more detail.

15 Figure 1 Lakeshore Pilot: Ocean Ave, Geneva Ave, and Holloway Ave South/West side of the street. IV. APPLICATION The Treasure Island Community Transportation Plan Treasure Island is a former naval base situated midway between San Francisco and West Oakland. The island is under going a unique planning process, with conversion from a former Naval base to a new San Francisco neighborhood. Currently, Treasure Island supports a residential population of just over 3,000 people in 905 residential units, with a high proportion of low-income families, and a daily employee population of nearly 2,000. Future redevelopment plans for Treasure Island call for an additional 6,000 residential units and 2,500 new permanent jobs. This plan offers a rare opportunity to create a new, walkable, livable neighborhood in a dense, otherwise built-out urban environment.

16 The California Department of Transportation awarded a Community-based Transportation Planning Grant to SFDPH and the San Francisco Bicycle Coalition (SFBC) to create a Community Transportation Plan to ensure a walkable/bikeable Treasure Island. This participatory based effort is aimed at developing sound transportation polices and street and bicycle design guidelines for a healthy neighborhood. As part of the project the SFDPH created an Existing Conditions Report and Health Impact Assessment of Treasure Island. The BEQI was used to assess the existing conditions of the bicycle environment (see Figure 2). Yerba Buena Island was not included in the study area due to challenging topography (i.e. extremely high street grade) and unsafe walking conditions for the surveyor. Data was collected on 142 street segments and 91 intersections by SFBC volunteers and SFDPH staff in Fall The goal of the BEQI application is to show the need for bicycle improvements and document the current bicycle conditions before the area is redeveloped. The Treasure Island Transportation Development Plan intends to integrate new bicycle facilities. Figure 2 Treasure Island Pilot: All streets North side of the street

17 V. RELEVANCE FOR LAND USE AND TRANSPORTATION PLANNING PROCESSES Results from the BEQI reveal the relative quality of the biking environment at a street-level scale in select San Francisco neighborhoods. Use of the BEQI can translate environmental variables into a set of provisions for a healthy bicycle environment and a BEQI assessment can inform neighborhood planning and prioritize improvements through the land use plans and environmental assessments. An application of the BEQI asks the following questions: 1) Does a place have adequate and safe bicycle facilities throughout the neighborhood? BEQI indicators are used to assess baseline conditions 2) Does a plan or project advance bicycle facilities in the area? Plans/projects are assessed to evaluate the extent to which BEQI indicators are present 3) What recommendations for planning policies, implementing actions, or project design would advance the bicycle environment? Concrete, specific recommendations are provided to the plan/project based on the evaluation There are several different avenues through which the BEQI can be applied. Through the neighborhood planning process, the BEQI could inform residents of where priority bicycle improvements need to be made. The BEQI would also be beneficial in the planning process for development and redevelopment projects, where a portion of development fees and community benefits funds could be allocated for bicycle facilities to improve access to the project site and other neighborhood services. In addition to planning applications, BEQI results may be used in health research, for example, to see whether a statistical correlation exists between a visual assessment of bicycle environmental quality and other neighborhood vitality indicators and/or health outcomes VI. STRENGTHS, LIMITATIONS, AND NEXT STEPS While the BEQI aims to assess the experience a bicyclist encounters on the streets, there are limitations to the index. First, there are several measures that were not included in the index that possibly contribute to people bicycling. For example, traffic signals were not included but could contribute to why a bicyclist would travel on one route instead another. Many experienced bicyclists might prefer to be on a route with traffic signals so they are not stopping as frequently due to timed signals. Another important caveat is that the route chosen may vary depending on the level of experience of the bicyclist and the bicyclist s comfort around motor vehicles. The BEQI scores are weighted due to importance of each index measure to the average bicyclist but additional weighting might need to be explored to better capture why bicyclists feel safer on certain streets. Forming focus groups around re-weighting indicator values would determine more accurate scores for bicycle conditions. The focus groups could also determine if all indicators are present in the tool.

18 In finalizing the BEQI, it is important to meet with other transportation planners, including the San Francisco Municipal Transportation Agency, to review the BEQI and provide feedback on the index and how helpful the BEQI would be to planners in the planning process.

19 Reference List Alta Planning and Design San Francisco's Shared Lane Pavement Markings: Improving Bicycle Safety. American Association of State Highway and Transportation Officials Guide for the Development of Bicycle Facilities. Washington, DC. Carter, Daniel J; University of North Carolina (System); Highway Safety Research Center; Pedestrian and Bicycle Information Center, and Turner-Fairbank Highway Research Center Pedestrian and bicyclist intersection safety indices. McLean, VA, Springfield, VA: Turner- Highway Research Center, Federal Highway Administration. National Technical Fairbank Information Service; vii, 88 p microform. Notes: Includes bibliographical references (p ). City of Auburn Auburn Bicycle Plan. City of Portland, Oregon Title 17 Public Improvements, Chapter Sidewalks, Curbs and Driveways: Bicycle Parking. City of Vancouver Southeast Vancouver Neighborhood Traffic Management Plan. Prepared by DKS Associates in association with Otak, Inc., JD White Company, Alta Transportation Consulting, Janice Kelley, and Dan Burden. Davis, J Bicycle Safety Evaluation. Auburn University, City of Chattanooga, and Chattanooga- Hamilton County Regional Planning Commission, Chattanooga, TN. Dill, J. and T. Carr Bicycle commuting and facilities in major U.S. cities: if you build them, commuters will use them, Transportation Research Record 1828: Dumbaugh, Eric Safe Streets, Livable Streets. Journal of the American Planning Association 71(3): Epperson, Bruce Evaluating suitability of roadways for bicycle use toward a cycling levelof-service standard. Transportation Research Record. 1438; p Notes: Includes bibliographical references (p. 16). Ewing R, Frank L, Kreutzer R Understanding the Relationship between Public Health and the Built Environment: A Report to the LEED-ND Core Committee. Federal Highway Administration, Bicycle Safety-Related Research Synthesis, Harkey, David L and Knuiman, Matthew Assessing the bicycle compatibility of roadways. Traffic Congestion and Traffic Safety in the 21st Century: Challenges, Innovations, and Opportunities. p Notes: Includes bibliographical references (p. 131). Harkey, David L; Reinfurt, Donald W, and Knuiman, Matthew Development of the bicycle compatibility index. Transportation Research Record. 1636; p Notes: Includes bibliographical references (p. 20). Harkey, David L; Reinfurt, Donald W; Sorton, Alex; United States; Federal Highway

20 Administration; Office of Safety and Traffic Operations Research and Development; University of North Carolina at Chapel Hill, and Highway Safety Research Center. The bicycle compatibility index a level of service concept, implementation manual. McLean, VA, Springfield, Va: Federal Highway Administration, Research and Development, Turner-Fairbank Highway Research Center. Available through the National Technical Information Service; v, [47] p. Notes: Includes bibliographical references (p. [47]). Hunter, William W; Pein, Wayne E, and Stutts, Jane C Bicycle-motor vehicle crash types the early 1990s. Transportation Research Record. 1502; p Notes: Includes bibliographical references (p. 74). Hunter, William W; Stewart, J. Richard, and Stutts, Jane C Study of bicycle lanes versus wide curb lanes. Transportation Research Record. 1674; p Notes: Includes bibliographical references (p. 77). Jacobsen, P. L. Safety in numbers: more walkers and bicyclists, safer walking and bicycling. Inj Prev Sep; 9(3): Landis, Bruce W Bicycle interaction hazard score a theoretical model. Transportation Research Record. 1438; p Notes: Includes bibliographical references (p. 8). Landis, Bruce W; Vattikuti, Venkat R, and Brannick, Michael T Real-time human perceptions toward a bicycle level of service. Transportation Research Record. 1578; p Notes: Includes bibliographical references (p. 126). Massachusetts Highway Department Project Development and Design Guidelines: Chapter 16 Traffic Calming and Traffic Management. Metropolitan Area Planning Council Mixed Use Zoning: A Planners Guide. Metropolitan Orlando Orlando Area Bicyclist Crash Study: A Role-Based Approach to Crash Countermeasures. Orlando, FL. Miami Dade Metropolitan Planning Organization. Bicycle Facilities Plan Moritz, William E. Ph.D Adult Bicyclists in the United States: Characteristics and Riding Experience in Transportation Research Board. National Highway Traffic Safety Administration Traffic Safety Facts U.S. Department of Transportation, National Highway Traffic Safety Administration: Washington, DC, n.d. Accessed at: [DOT HS ]. North Carolina Department of Transportation, Division of Bicycle and Pedestrian Transportation North Carolina Bicycle Crash Data: Bicycle Crash Types Summary Report. Pedestrian and Bicycle Information Center. Bicycle Parking. Accessed on May 17,