Bike-Car Crash Study

7/30/2019 Bike-Car Crash Study

1/30

Understanding Bicyclist-Motorist Crashes in

Minneapolis, Minnesota

A comprehensive look at crash data from 2000-2010 and

recommendations for improved bicyclist safety

Bicycle and Pedestrian Section

Public Works Department

City of Minneapolis

January 15, 2013


2/30

Understanding Bicyclist-Motorist Crashes in Minneapolis 2

Understanding Bicyclist-Motorist Crashes in Minneapolis, Minnesota:A comprehensive look at crash data rom 2000-2010 and solutions or improved bicyclist saety

January 15, 2013

City o MinneapolisPublic Works DepartmentBicycle and Pedestrian Section350 South Fith Street, Room 203

Minneapolis, Minnesota 55415

www.minneapolismn.gov/bicycles

This report is published by the City o Minneapolis Public Works Department. For questions about this report,please contact Simon Blenski at 612-333-1274 or [email protected].

Some o the images in this report are based on illustrations by Chicago Cartographics or the Minneapolis BicycleMap, published by Hedberg Maps, Inc.


3/30


The 2011 Minneapolis Bicycle Master Plan1 calls or a 10 percent annual reduction in the number o bicyclist-motorist crashes. To better understand what is causing crashes and to meet reduction goals, 2,973 bicyclist-motorist crash records rom 2000-2010 were examined. Specic crash attributes were extracted rom MinnesotaDepartment o Public Saety accident reports, analyzed and mapped. The ndings in this report should be used

to inorm and inuence the design o new bicycle acilities, the redesign o existing roadways, the developmento education programs or bicyclists and motorists, enorcement campaigns, and the creation o bicycle-relatedpolicy in Minneapolis.

Key Findings rom 2000-2010:

When crashes occur:

An average o 270 bicyclist-motorist crashes occur annually in Minneapolis. This is down rom anaverage o 320 rom 1993-1999. (p. 15)

Crashes are most prevalent rom April-October (88.3 percent), on weekdays (79.3 percent) and duringthe aternoon peak period rom 3:00-6:00 p.m. (28.5 percent). (p. 15)

Crashes mostly occur on clear or cloudy days (93.5 percent), when the road surace is dry (89.1 percent)and during daylight hours (72.7 percent).(p. 16)

Who is involved:

Most (93.5 percent) crashes involve bicycles and automobiles. Large trucks, buses, taxis and othervehicles make up the remaining vehicle types. (p. 16)

Bicyclist age is tracked or 2009-2010 data. The cohort aged 18-24 was the most prevalent - involved in21.9 percent o crashes. (p. 16)

Crashes involving known drug use or drinking are limited. Bicyclists are impaired in 5.9 percent ocrashes and motorists in 1.2 percent o crashes. (p. 17)

Approximately one out o ve crashes are hit-and-runs, with the motorist eeing the scene 93.3 percento the time. Motorist condition is unknown in these cases. (p. 17)

Injuries and atalities:

Bicyclists sustained an injury in 87.0 percent o crashes. It is estimated that motorists sustained an injuryin no crashes. (p. 21)

There were 12 bicyclist atalities rom 2000-2010. All cases involved at least one o the three ollowingattributes: rain or wet pavement, aggressive or impaired motorist, or a large motor vehicle. (p. 21)

Causes o crashes:

Assigning ault is a dicult and inexact task. However, it appears that bicyclists and motorists areequally contributing to the causes o crashes. Bicyclists are estimated to have contributing actors in59.0 percent o crashes and motorists in 63.9 percent o crashes. The totals exceed 100 percent as bothparties have contributing actors. (p. 18)

The most common contributing actors or bicyclists are ailure to yield right-o-way (13.3 percent),disregarding a trac control device (12.6 percent) and improper lane use (9.2 percent). (p. 18)

The most common contributing actors or motorists are ailure to yield right-o-way (31.8 percent),driver inattentive or distracted (8.5 percent) and improper lane use (5.2 percent). (p. 18)

The most common pre-crash maneuvers or bicyclists are bicyclist riding across roadway (46.0 percent),bicyclist riding with trac (29.8 percent) and bicyclist riding against trac (15.4 percent). (p. 18)

The most common pre-crash maneuvers or motorists are vehicle ollowing roadway (42.2 percent),vehicle making let turn (18.7 percent) and vehicle making right turn (16.4 percent). (p. 18)

1 City o Minneapolis. Bicycle Master Plan. June 2011. www.minneapolismn.gov/projects/plan

Executive Summary


4/30


Where crashes are occurring:

Forty-one percent o crashes occur at intersections and another 40 percent occur within 50 eet ointersections. (p. 23)

Crashes occur in all areas o Minneapolis, although there is a clear concentration along major arterialswith high volumes o motor vehicles. (p. 23)

The highest crash volume intersections rom 2000-2010 are East Franklin Avenue and South Cedar Avenue(20), Hennepin Avenue and North 7th Street (19), Hennepin Avenue and North 3rd Street (17), Hiawatha

Avenue South and East 26th Street (17), East Franklin Avenue and Nicollet Avenue South (17). (p. 23) Corridors with the highest number o crashes rom 2000-2010 are East-West Lake Street (226 bicyclist-motorist crashes), East-West Franklin Avenue (205), Portland Avenue South (127), Hennepin AvenueSouth in downtown (126) and Lyndale Avenue South (111). (p. 24)

Corridors with the highest crash rates rom 2000-2010 are East-West 28th Street (68.5 crashes per millionbicycle miles traveled), Lowry Avenue North-Northeast (55.4), Marquette Avenue South (39.5), East-West26th Street (39.2) and West Broadway-Broadway Street Northeast (39.1). (p. 24)

Saety in numbers:

There is a clear correlation between the number o bicyclists and the crash rate. Minneapolis has seenthis phenomena occur across both time and space. (pp. 8,9 & 25)

As the number o bicyclists has increased over the past decade, the crash rate has decreased. (pp. 8-9) On streets and corridors with higher volumes o bicycle trac, the crash rate tends to be lower than on

streets with lower volumes o bicycle trac. (p. 25) This saety in numbers efect is documented in other U.S. cities and in academic literature. (pp. 8-9)

Summary

The analysis o the 2,973 bicyclist-motorist crashes ound that crashes are complex events and there is no oneactor that is contributing to crashes. However, three primary conclusions emerge rom the data:

1. Most crashes are occurring at intersections along major arterials.2. Motorists are not seeing or yielding to bicyclists.3. Bicyclists are not riding in a predictable manner.

Recommendations

The recommendations aim to reduce perceived ears o interested but concerned bicyclists and are presentedwithin the ramework o the Six Es o Bicycling: Equity, Engineering, Enorcement, Education, Encouragementand Evaluation.

Equity

Develop a bicycle trac saety work group Ensure that saety talking points equally

address motorists and bicyclists

Engineering

Guide and protect bicyclists at intersectionsand on busy streets

Highlight areas where bicyclists andmotorists cross paths

Provide designated and comortable placesor bicyclists to ride

Enorcement

Expand a relationship with the MPD Ensure bicyclists and motorists are treated

equally under the law Use enorcement as an educational tool

Education

Educate proessional drivers Use media to reach a wide audience Continue rides and classes

Encouragement

Design inrastructure that is perceived to be sae Publish data and document results

Evaluation

Publish a regular saety bicyclist report Increase understanding o crashes


5/30


Chapter 1 - Introduction ......................................................................................................................................................7

1.1 Purpose .................................................................................................................................................................................................................. 7

1.2 Using this Report ................................................................................................................................................................................................ 7

Chapter 2 - Bicyclist Saety in Context ...............................................................................................................................8

2.1 A Changing City or Bicyclists ....................................................................................................................................................................... 8

2.2 Saety in Numbers .............................................................................................................................................................................................. 8

2.3 Previous Work ...................................................................................................................................................................................................... 9

Chapter 3 - Understanding the Data ...............................................................................................................................10

3.1 How a Crash is Reported ................................................................................................................................................................................10

3.2 Unreported Crashes ........................................................................................................................................................................................10

3.4 Interpretation and Assumptions .................................................................................................................................................................12

3.5 What is Not Captured in the Data ...............................................................................................................................................................12

3.6 Notes About Terminology .............................................................................................................................................................................12

Chapter 4 - Approach & Methodology ............................................................................................................................13

4.1 Approach .............................................................................................................................................................................................................13

4.2 Methodology .....................................................................................................................................................................................................13

Chapter 5 - Results ..............................................................................................................................................................14

5.1 Overview .............................................................................................................................................................................................................14

5.2 When Crashes Occur .......................................................................................................................................................................................15

5.3 Environmental Conditions ............................................................................................................................................................................16

5.4 Motorist Vehicle Type......................................................................................................................................................................................16

5.5 Bicyclist Age .......................................................................................................................................................................................................16

5.6 Rider and Driver Condition ...........................................................................................................................................................................17

5.7 Crash Circumstances (Hit-and-Runs) .........................................................................................................................................................17

5.8 Contributing Factors .......................................................................................................................................................................................17

5.9 What is Causing Crashes? ..............................................................................................................................................................................18

5.10 Specic Crash Types and Other Crash Attributes ...............................................................................................................................19

5.11 Injury Severity .................................................................................................................................................................................................21

5.12 Fatalities.............................................................................................................................................................................................................21

5.13 The Cost o Crashes .......................................................................................................................................................................................22

5.14 Where Crashes are Occurring ....................................................................................................................................................................23

Chapter 6 - Discussion, Approaches and Recommendations .....................................................................................26

6.1 Summary o Findings ......................................................................................................................................................................................26

6.2 Approaching Bicyclist Saety ........................................................................................................................................................................26

6.3 Recommendations or Improved Bicyclist Saety .................................................................................................................................28

Table o Contents


6/30


Appendix

Appendix A: Example MN DPS Accident Reports ........................................................................................................A-1

Appendix B: MN DPS Accident Coding List .............. ............... ................ ............... ............... ................ ............... ........ B-6

Appendix C: Complete Bicyclist-Motorist Crash Data 2000-2010 ............................................................................ C-9

Appendix D: State and Peer City Comparison .......................................................................................................... D-19

Appendix E: Supplemental Context Maps ............. ................ ............... ............... ................ ............... ............... ......... E-21

Minneapolis Bikeways ...........................................................................................................................................................................................22

Bicyclist Estimated Daily Trac ..........................................................................................................................................................................24

Signalized Intersections ........................................................................................................................................................................................27

One-Way Streets ......................................................................................................................................................................................................28

Posted Speed Limit .................................................................................................................................................................................................29

Schools and Parks ....................................................................................................................................................................................................30

Sidewalk Riding ........................................................................................................................................................................................................31

Appendix F: Supplemental Crash Maps ............... ............... ................ ............... ............... ................ ............... ........... F-32

Crash Density by Year .............................................................................................................................................................................................33

Crash Density by Season .......................................................................................................................................................................................45

Crash Density by Day .............................................................................................................................................................................................49

Crash Density by Time ...........................................................................................................................................................................................51

Crash Density by Circumstance ..........................................................................................................................................................................59

Crash Density by Vehicle Type (Non-Bicycle).................................................................................................................................................60

Crash Density by Bicyclist Condition ................................................................................................................................................................63

Crash Density by Motorist Condition ...............................................................................................................................................................64

Crash Density by Bicyclist Age............................................................................................................................................................................65

Crash Density by Bicyclist Contributing Factor ............................................................................................................................................73

Crash Density by Bicyclist Pre-Crash Maneuver ...........................................................................................................................................87

Crash Density by Motorist Pre-Crash Maneuver ..........................................................................................................................................93

Crash Density by Motorist Combined Attributes .........................................................................................................................................99

Crash Density by Injury Severity ......................................................................................................................................................................101

Appendix G: Corridor Crash Rates ............... ............... ............... ................ ............... ............... ................ ............... ... G-105

Appendix H: Supplemental Corridor Analysis .............. ............... ................ ............... ............... ................ ............. H-110

Lowry Avenue North-Northeast.......................................................................................................................................................................111

West Broadway Avenue North | Broadway Street Northeast ................................................................................................................112

Central Avenue Northeast ..................................................................................................................................................................................113

Hennepin Avenue South (Downtown) (First Avenue Northeast) ........................................................................................................114

University Avenue Southeast | Fourth Street Southeast .........................................................................................................................115

Washington Avenue North-South | Riverside Avenue South................................................................................................................116

Marquette Avenue South | Second Avenue South....................................................................................................................................117

Nicollet Mall/Avenue South | Third Avenue South ....................................................................................................................................118

East-West Franklin Avenue .................................................................................................................................................................................119

East-West 24th Street | East-West 26th Street | East-West 28th Street ...............................................................................................120

East-West 35th Street | East-West 36th Street | East-West 38th Street ...............................................................................................121

East-West Lake Street (Lagoon Avenue) | East-West 31st Street ..........................................................................................................122

Hennepin Avenue (South) | Lyndale Avenue South .................................................................................................................................123

Portland Avenue South | Park Avenue South ..............................................................................................................................................124

Cedar Avenue South ............................................................................................................................................................................................125

Hiawatha Avenue South | Minnehaha Avenue South ..............................................................................................................................126


7/30


8/30


Chapter 2 - Bicyclist Saety in Context

2.1 A Changing City or Bicyclists

This report analyzes bicyclist-motorist crash datarom 2000-2010. Riding a bicycle in Minneapolis in2000 was a much diferent experience than in 2010.

At the start o the decade the Midtown Greenwaydid not yet exist, Hennepin Avenue and otherdowntown streets looked much diferent and theconcept o bicycle boulevards had not yet arrivedin Minneapolis. However, over the next ten yearsbikeways and bicycling made substantial gains.

From 1999 to 2011 the miles o bikeways doubledrom 80 to 166, with a signicant increase in thenumber o on-street acilities. Most bikeways addedin the later hal o the decade were made possiblein part by the $25 million ederal Non-Motorized

Transportation Pilot Program grant awarded tothe Minneapolis area in 2007. From 2000-2010 thenumber o regular bicycle commuters increasedrom 1.6 to 3.4 percent1 and rom 2007 to 2011 thenumber o bicycle trips increased 47 percent.

1 U.S. Census Bureau. 2000 Decennial Census and 2011 AmericanCommunity Survey. www.census.gov

2.2 Saety in Numbers

While added inrastructure is correlated to theincreasing number o bicyclists in Minneapolis,the number o bicyclists themselves appear to be

improving saety. Despite the act that the numbero regular bicycle commuters doubled over the pasttwo decades, the number o reported crashes hasheld steady at around 300 crashes per year. Withapproximately 3,000 bicycle commuters in 1993 and7,000 in 2011, the city-wide crash rate decreased byover two old.2

This trend is not unique to Minneapolis and isdocumented in other cities. Analysis o 68 Caliorniacities ound that counter to intuition, a motoristis less likely to collide with a person walking andbicycling when there are more people walking orbicycling.3 Data rom Portland, Oregon reveals a

2 The city-wide crash rate is the ratio o reported crashes to thenumber o regular bicycle commuters as collected by the U.S.Census Bureau.3 Jacobsen, Peter L. Saety in Numbers: More Walkers and Bicyclists,Saer Walking and Bicycling. Injury Prevention. 2003; 9: 205-209.

Bikeways in 1999 Bikeways in 2011

Figure 2.1 - Minneapolis on-street and of-street bikeways in 1999 and 2011.

On-Street Bikeway On-Street Bikeway

Of-Street Bikeway Of-Street Bikeway


9/30


10/30


Chapter 3 - Understanding the Data

3.1 How a Crash is Reported

A trac crash is an unortunate and complex event.There are oten multiple contributing actors,multiple parties involved and several layers o

interpretation and reporting. In Minnesota, anindividual involved in a trac crash that immediatelyresults in property damage or bodily injury has thelegal obligation to remain at the scene o the crashuntil contact inormation is exchanged with allparties involved.1 I a police ocer is not immediatelycalled to the scene, involved parties have up to 72hours to notiy authorities.

Once a police ocer collects the necessaryinormation, he or she completes a MinnesotaDepartment o Public Saety (DPS) accident report.Location, time, personal inormation, weather, roadsurace conditions and other attributes are recordedusing a standardized coding system. To supplementthe codes, a crash narrative and diagram are alsocompleted as part o the report.

The Minneapolis Public Works Department receivescopies o accident reports rom the MinneapolisPolice Department and the Minneapolis Park Police.Select inormation is entered into an internal PublicWorks database and the crash reports are thendestroyed. Attributes entered into the crash databaseare discussed in Section 4.2.

1 Minnesota Statute. 169.09 Accidents. www.revisor.mn.gov/statutes

3.2 Unreported Crashes

This report only examines data rom reportedtraccrashes. Crashes o all types go unreported, but it isestimated that bicycle and pedestrian crashes are

overrepresented among unreported crashes. Reasonsor not reporting a crash may be that no party wasinjured, property damage was marginal, individualsed the scene or were not aware that they arerequired to report a crash.

A possible method or determining the number ounreported crashes is to examine hospital records.A study o bicyclist-motorist crashes in Caliornia,North Carolina and New York ound that only 42.5-66.7 percent o bicyclist emergency room visitsmatched bicyclist-motorist crash records.2 The dataonly examined hospital visits or bicyclists involved inbicyclist-motorist trac crashes - not solo alls. Thesendings imply that 33.3-57.5 percent o bicyclist-motorist crashes go unreported. However, denitiveconclusions cannot be made because all crashes donot automatically result in hospitalization.

Examining Minneapolis hospital records rom theMinnesota Department o Health do not reveala clear conclusion either. According to accidentreports, most, but not all bicyclist-motorist crashes

2 Federal Highway Administration. Injury to Pedestrians andBicyclists: An Analysis based on Hospital Emergency Department

Data. FHWA-RD-99-078. 1999.

1 Crash

occurs

2 Police ocer

collects information

3MN DPS accident

report completed

by police ocer

4 Copy of accident report

submitted to Minneapolis

Public Works

5Minneapolis Public Works

enters select crash

information into database

Figure 3.1 -The crash reporting process rom collision to data entry.


11/30


ocer determines that there was no apparent orobvious actor.

Pre-crash maneuvers describe the actions o eachparty just prior to the collision. Common pre-crashmaneuvers are Vehicle Making Let Turn, BicyclistRiding With Trac, or Bicyclist Riding Against Trac.While other crash attributes are useul in determiningcause, it is primarily contributing actors and pre-crash maneuvers that allow or the determinationo crash causes and crash typing. Example accidentreports and coding lists are included in Appendix Aand B.

A peace oce has the option o including up totwo contributing actors on an accident report. Forexample a bicyclist could have both disregarded atrac signal and been under the inuence. However,most crash reports only include one contributingactor and Minneapolis Public Works only records therst, or primary actor on the report.

While Public Works records the total number ovehicles and injuries, only detailed data is compiledor the rst two vehicles in a crash. For example, ian automobile rear ends another automobile which

then pushes the second vehicle into a bicyclist, onlydetails about the rst two vehicles (automobile oneand automobile two) would be recorded by PublicWorks. These complex events are important, butonly represent less than one percent o the crashesanalyzed in this report. Some records are technicallybicycle-related crashes, although the database mayonly show vehicle one as an automobile and vehicletwo as an automobile.

result in bicyclist injuries. However, even wheninjuries are sustained, the injury may not be severeenough or hospitalization or an individual mayreuse medical attention or other reasons. Also, abicyclist involved in a crash occurring outside the citymay visit a Minneapolis hospital or vice versa.

The data in Figure 3.2 show a very weak correlation toinjuries that may require hospitalization (moderate,severe or unknown injuries), but does not yield anyconclusive results.3 The reason or the low number omoderate and severe injuries rom 2000-2002 is notclear, although coding errors may be a possibility.

Note, this data does not include bicycle injuries romsolo alls.

3.3 Accident Reports and Defnitions

On a crash report, an array o inormation is compiledby the police ocer assigned to the case. Both theMinneapolis Police Department and MinneapolisPark Police use Minnesota DPS accident reports. Aseries o standardized codes are used to ecientlycategorize crash attributes. While all codes helpexplain the cause(s) and circumstance(s) o a crash,the two codes o particular interest are contributing

actor and pre-crash maneuver.

Contributing actors are circumstances that directlylead to the collision. Some o the most commoncontributing actors are Failure to Yield Right-o-Way or Disregarding a Trac Control Device. NoClear Factor is recorded or a party i the police

3 Minnesota Department o Health. Minnesota Injury Data AccessSystem (MIDAS). www.health.state.mn.us/injury/midas

0

100

200

300

400

20102009200820072006200520042003200220012000

Reported crashes

(DPS accident reports)

Reported crashes with moderate, severe

or unknown injury (DPS accident reports)

Reported hospital visit

by bicyclists involved

in a crash (MIDAS)

CrashesbyReportingT

ype

Figure 3.2 - Reported crashes rom MN DPS accident reports, reported hospital visits3 and reportedcrashes rom MN DPS accident reports with moderate, severe or unknown injury, 2000-2010


12/30


3.4 Interpretation and Assumptions

While accident reports are the most reliable sourceo bicyclist-motorist crash inormation, the dataare used with caution. By the time accident reportsare entered in the Public Works crash database therecounting o events has gone through many layerso interpretation and the probability o inaccuracy

may be high.

Ater a crash, victims may be severely injured andunable to recount details to the ocer. They may alsobe trying to avoid ault, insurance claims or ticketing.Witnesses may also provide conicting inormation.

The data passes another layer o interpretation whenthe ocer compiles inormation or the accidentreport. Crashes are complex events and reducing thesituation to codes is a dicult task.

These caveats are not intended to discredit accidentreports as a viable source o inormation. However,

with any data source there are limitations to its use.The primary assumption with this analysis is thataccident reports are accurate. While the precisedetails o every crash may deviate rom the sequenceo actual events, the overall ndings in this reportare consistent with anecdotal evidence, knowledgeo bicyclist trac volumes, turning movements,intersection geometry and the varying nature oMinneapolis streets and neighborhoods.

3.5 What is Not Captured in the Data

A number o attributes are not collected on accident

reports or analyzed as part o this research.

Bicyclist position prior to the crash - While somereports describe the bicyclists riding positionin detail, reporting is not consistent and theinormation is not recorded by Public Works. For thisreport, Public Works examined 800 accident reportsrom 2006-2008 or bicyclist position. The resultso that specic analysis are discussed briey in thisreport.

Driveway, alley and mid-block crashes - Crashesoccurring at driveway entrances, alleys or mid-block

locations are included in this dataset, although thelocation inormation is aggregated to the closestintersection and may not reect the actual locationo the crash. Data is also not available or crashesoccurring on private property such as a store parkinglot. Only crashes occurring in the public right-o-wayare recorded.

Bicyclist and motorist demographics - Gender andhome address are collected on crash reports,although Public Works does not record theinormation. Bicyclist age was collected beginning in2009.

Helmet Use - While some crash reports includebicyclist helmet use, reporting is not consistent and

is not recorded by Public Works.

Specic bicyclist crash types - At this time, PublicWorks does not record specic bicycle crash typessuch as right hooks, let hooks or dooring. However,an efort was made to estimate the prevalence othese crash types and is discussed in Chapter 5.

Other local agencies - This analysis primarily usedaccident reports supplied by the Minneapolis PoliceDepartment and Minneapolis Park Police. Data romother local agencies such as the U o M Police orMetro Transit Police are not consistently reported toPublic Works.

3.6 Notes About Terminology

The terms accident and crash are both used inthis report, although not interchangeably. In thediscussion o trac saety, crash is becoming theaccepted term when describing a collision. Accordingto the National Highway Trac Saety Administration,Continued use o the word accident promotes theconcept that these events are outside o humaninuence or control.1

To reinorce the act that a trac collision is apreventable event, crash is used throughout thisreport. However, accident is the term currently usedby the Minnesota Department o Public Saety ontheir ocial accident reports. In this report, accidentis only used when reerring to Minnesota DPSaccident reports.

1 Amsden, Michael and Thomas Huber. Bicycle Crash Analysisor Wisconsin using a Crash Typing Tool (PBCAT) and Geographic

Inormation System (GIS). Wisconsin Department o Transportation.June 30, 2006.


13/30


4.1 Approach

To yield patterns o statistical and spatial signicance,planning was done to determine an appropriatesample size o bicyclist-motorist crash records.

When evaluating saety projects, a three-year beoreand ater period is oten used or motor vehiclecrashes. This is a logical approach, as a substantialnumber o motor vehicle crashes occur each year ineven small communities. In Minneapolis alone, thereare an average o 6,700 total crashes each year - mosto which involve only motor vehicles.

Comparably, there are only an average o 270 bicyclerelated crashes annually in Minneapolis. Due to therelatively small number o crashes, a longer timeperiod was desired or this analysis, especially withrespect to mapping. In Minneapolis, there are 7,361roadway intersections. I a sample rom one, twoor even three years was used, numeric results mayprove useul. However, using that same sample, itwould be dicult to illustrate many spatial patternsacross the template o the citys street grid.

Motivation or a larger time period was also selectedto avoid the possibility o alse readings. A SanFrancisco study on corridor level analysis o bicyclistand pedestrian crashes ound that, Basing decisionson individual intersections and single years is o

limited ecacy and will yield substantial numberso...alse positives and alse negatives.1 In AppendixF, the maps o crashes by year highlight this dilemma(see p. F-34-44). The same study recommends a threeyear period as it, provides a good balance betweenreducing statistical variation and accounting orchanges in the intersections over time. Although, it isnoted that ve years is better or intersections with arelatively low number o crashes.

Public Works selected a sample period o 10 yearsbecause little was understood about local bicyclistsaety and there was a desire to gain a broad

understanding o crashes in Minneapolis. However,because many corridors and intersections havechanged over the 10-year period, changes that mayinuence the results are noted.

1 Ragland, David, et. al. Strategies or Reducing Pedestrian andBicyclist Injury at the Corridor Level. UC Berkeley Sae TransportationResearch & Education Center. July 2011.

4.2 Methodology

As outlined in Chapter 3, Public Works records selectcrash attributes rom DPS accident reports in a crashdatabase. The primary attributes available or each

crash are:

Context, Environment and Injury Severity

Date Time Intersection Distance rom Intersection Weather Road Surace Circumstance (Hit-and-Run) Injury Severity

Bicyclist Inormation

Bicyclist Contributing Factor Bicyclist Pre-Crash Maneuver Bicyclist Condition Bicyclist Date o Birth (2009-2010 only)

Motorist Inormation

Motorist Contributing Factor Motorist Pre-Crash Maneuver Motorist Condition Motorist Vehicle Type

These attributes are available or all years between2000 and 2010, except or bicyclist date o birth.Public Works began recording bicyclist andpedestrian date o birth in 2009. A ull list o DPScodes can be ound in Appendix B.

Summary data is presented or each o theseattributes. In some cases multiple attributes arecombined to bring increased understanding tocrashes. However, caution was used when biurcatingthe data. Even with a sample size o nearly 3,000crashes, layering more than two attributes yieldedresults o little signicance.

Chapter 4 - Approach & Methodology


14/30


Chapter 5 - Results

5.1 Overview

The analysis o 2,973 bicyclist-motorist crashes oundthat crashes are complex events and there is no oneactor that is contributing to crashes. That said, three

primary conclusions emerge rom the data:

1. Most crashes are occurring at intersectionsalong major arterials.

2. Motorists are not seeing or yielding tobicyclists.

3. Bicyclists are not riding in a predictablemanner.

This chapter provides support or these conclusionsand highlights other prevalent crash attributes.

Topics presented in this chapter are:

Background

When crashes occur (Section 5.2)

Environmental conditions (5.3) Motorist vehicle type (5.4) Bicyclist age (5.5) Rider and driver condition (5.6) Crash circumstances (5.7)

Causes o Crashes

Contributing Factors (5.8) What is causing crashes? (5.9) Specic crash types and other crash

attributes (5.10)

Injuries and Costs

Injuries (5.11) Fatalities (5.12) The cost o crashes (5.13)

Where Crashes are Occurring

Top crash corridors (5.14) Top crash intersections (5.14) Saety in numbers (5.14)

Lake

ofthe

Isles

Cedar

Lake

Lake

Nokomis

Lake

Hiawatha

LakeCalhoun

LakeHarriet

M

i

s

s

i

s

s

i

p

p

i R

i

v

e

r

Figure 5.1 - Citywide bicyclist-motorist crash density, 2000-2010

Crashes PerIntersection

1

2-3

4-5

6-9

10-20


15/30


16/30


5.3 Environmental Conditions

Weather

Weather conditions at the time o crashes weregenerally avorable. Conditions were clear 72 percento the time and cloudy 22 percent o the time. It wasraining or ve percent o crashes and snowing orless than one percent o crashes.

Road Surace

The road surace at the time o crashes was generallyavorable. Conditions were dry 89 percent o the timeand wet eight percent o time. Snow, slush or icewere present or 1.3 percent o crashes.

Light

Conditions tended to be light during the time ocrashes. About 73 percent o crashes occurred whenthere was natural daylight and 27 percent occurredwhen it was dark. Crash time was analyzed withsunrise and sunset times between 2000 -2010. Lightconditions were not adjusted or cloud cover, rainor other actors. Street light inormation is listed onaccident reports, but is not recorded in the PublicWorks crash database.

5.4 Motorist Vehicle Type

Most motor vehicles were automobiles - ninety-threepercent were either a car, small van, pick up truck orSUV. All other vehicle types each accounted or lessthan two percent o crashes. There were 43 trucks,41 taxis, 40 buses, 18 emergency vehicles, ninemotorcycles and three limousines. Other vehicles,non-bicyclists pairings and pedestrians round out thetotal. Note that Public Works only records details orthe primary two vehicles, so the pedestrian crashesmay be related to crashes involving more than twovehicles.

5.5 Bicyclist Age

Public Works started recording bicyclist age in 2009.Age is not currently recorded or motorists. Bicyclistsage 18-24 is the most prevalent cohort and is theonly cohort that is overrepresented when compared

to the overall population o Minneapolis. Elementary-aged youth1 and adults aged 45 or older areunderrepresented when compared to the populationas a whole.

1 Analysis o bicyclist crashes near schools was originally desiredas part o this report. However, the limited sample size rom 2009-2010 did not yield noticeable patterns related to school locations.In act, more youth crashes occurred on weekends or during thesummer months than during the during the school week.

5%

10%

15%

20%

25%

Fog/Smoke/

Smo

Sleet/

Hail

SnowUnknown/Other

RainCloudy

Conditions were clear for 72% of crashes

Percentage

ofCrashes

Figure 5.3.1 - Percentage o crashes by weather conditions

2%

4%

6%

8%

10%

Not-Applicable

DebrisIce/PackedSnow

Snow/Slush

Unknown/Other

Wet

The road surface was dry for 89% of crashes

Percentage

ofCrashes

Figure 5.3.2 - Percentage o crashes by road surace

0.3%

0.6%

0.9%

1.2%

1.5%

Unknow

n/

Other

Not-A

pplicabl

e

Limo

Pedestri

an

Motorcy

cle

Emergen

cy

Bu

s

Taxi

Truck

93% of crashes involved automobiles

Percentage

ofCrashes

Figure 5.4 - Percentage o crashes by motorist vehicle type

0

5%

10%

15%

20%

25%

Unknown65+55-6445-5435-4425-3418-2413-174-12

PercentageofCrashes

Bicyclist Age

Minneapolis

Population

(2010)

Population lowerlimit extends

to birth.

Figure 5.5 - Percentage o crashes by bicyclist age,2009-2010 only and Minneapolis population by age.Source: 2010 American Community Survey.


17/30


5.6 Rider and Driver Condition

Bicyclist Condition

Bicyclists had a normal condition in 83 percent ocrashes. Condition was unknown or 10 percento crashes. Bicyclists were impaired (under theinuence, had been drinking or drug use) six percento the time and ve bicyclists, or 0.2 percent o

bicyclists were aggressive.

Motorist Condition

Motorists had a normal condition in 77 percent ocrashes. Condition was unknown or 21 percent ocrashes and coincides with most hit-and-runs (seeSection 5.7 below). Motorists were impaired (underthe inuence, had been drinking or drug use) 1.2percent o the time and ve motorists, or 0.2 percento motorists were aggressive.

5.7 Crash Circumstances (Hit-and-Runs)

Crash circumstances document hit-and-runs andpolice chases. Twenty-one percent o bicyclist-motorist crashes are hit-and-runs. Three o the 2,973crashes involved a police chase. While high, bicycle-related hit-and-run crashes in Minneapolis areactually underrepresented compared to all crashes.O all trac crashes in Minneapolis rom 2000-2010,41.3 percent o crashes were hit-and-run crashes.Removing crashes involving parked vehicles or xedobjects, hit-and-runs still account or 29.7 percent oall crashes.

While bicycle-related hit-and-run crashes areunderrepresented, bicyclists are disproportionatelythe victims o hit-and-run crashes. Detailed analysiso 800 accident reports rom 2006-2008 ound that obicycle-related hit-and-run crashes, motorists ed 92.8percent o the time and bicyclists ed 7.2 percent.

Factors associated with eeing the scene o a crashare not listed on accident reports and the MPDlimits investigations o hit-and-runs to crashesinvolving severe or atal injuries. However, researchopedestrian-motoristhit-and-run crashes atthe national level ound that alcohol use and aninvalid license are among the leading driver actorsassociated with hit-and-runs.1 Identied hit-and-run

drivers are also more likely to have had a previousarrest or driving while intoxicated.2 Whilepedestrian-motoristcrashes difer rom bicyclist-motoristcrashes,the research ofers possible explanations or the highpercentage o eeing motorists.

5.8 Contributing Factors

As outlined in Chapter 3, a crash is a complex eventand determining ault is dicult. Fault is not explicitlyassigned on accident reports, although analysiso contributing actors can help provide useulinormation.

O the 2,973 crashes, there were 52 cases in whichboth vehicles were assigned No Clear Factor orUnknown or Other. Removing these crashes romthe sample, we nd that bicyclist and motoristcontributing actors are about equally represented.In 59.0 percent o crashes, bicyclists had contributingactors. Motorists had contributing actors in 63.9percent o crashes. These totals exceed 100 percentas both the bicyclist and motorist had contributingactors in some instances.

1 MacLeod, Kara E. et. al., Factors Associated with Hit-and-RunPedestrian Fatalities and Driver Identication.Accident Analysisand Prevention. 45 (2012)2 Solnick, Sara J. & David Hemenway. Hit the Bottle and Run:

The Role o Alcohol in Hit-an-Run Pedestrian Fatalities.Journal oStudies o Alcohol. November 1994.

20%

40%

60%

80%

100%

MotoristBicyclist

PercentofCrashes

59.0%63.9%

Figure 5.8 - Contributing actors by bicyclists andmotorists. These totals exceed 100 percent as both thebicyclist and motorist had contributing actors in some

instances.

Motorist

Fled

(92.8%)

All Bicyclist-Motorist

Crashes

Bicyclist-Motorist

Hit & Run Crashes

Bicyclist

Fled (7.2%)

Hit-and-Runs

(21.7%)

Figure 5.7 - Percentage obicyclist-motorist crashes by hit-and-run (let) and percentage ohit-and-run crashes by eeing unit(right).


18/30


5.9 What is Causing Crashes?

Using this data set, contributing actors and pre-crash maneuvers are the best indicators o the causeo a crash. This section outlines the top contributingactors and pre-crash maneuvers or both bicyclistsand motorists.

Bicyclist FactorsBicyclists had no clear actor in 43 percent o crashes.

The next ve most prevalent bicyclist contributingactors were ailure to yield right-o-way (13 percent),disregarding a trac control device (13 percent),improper lane use (nine percent), driver (bicyclist)inattentive or distracted (ve percent) and non-motorist error (ve percent). Non-motorist erroris a catch all term or non-motorist (bicyclist andpedestrian) errors. All other actors occurred in lessthan two percent o crashes.

Motorist Factors

Motorists had no clear actor in 38 percent o crashes.The next ve most prevalent motorist contributingactors were ailure to yield right-o-way (32 percent),driver inattentive or distracted (nine percent),improper lane use (ve percent), disregarding atrac control device (ve percent), vision obstructedby other actors (two percent). All other actorsoccurred in less than two percent o crashes.

Bicyclist Pre-Crash Maneuvers

Bicyclist pre-crash maneuver document the actions

o a bicyclist just prior to the collision. The top vepre-crash maneuvers are bicyclist riding acrossroadway (46 percent), bicyclist riding with trac (30percent), bicyclist riding against trac (15 percent),bicyclist making let turn (three percent), bicyclistslowing, stopping or starting (two percent).

Motorist Pre-Crash Maneuvers

Motorist pre-crash maneuver document the actionso a motorist just prior to the collision. The top vepre-crash maneuvers are vehicle ollowing roadway(42 percent), vehicle making let turn (19 percent),vehicle making right turn (16 percent), vehiclestarting in trac (seven percent) and vehicle makingright turn on red (ve percent).

0 10% 20% 30% 40% 50%Failure to Use Headlights

Non-Motorist Error

Driver Inattentive or Distracted

Improper Lane Use

Disregarding a Trac Control Device

Failure to Yield Right-of-Way

No Clear Factor

ContributingF

actor

Percentage of Crashes

Figure 5.9.1 -Top bicyclist contributing actors

0 10% 20% 30% 40% 50%

Other Human Factors

Vision Obstructed by Other Factors

Disregarding a Trac Control Device

Improper Lane Use

Driver Inattentive or Distracted

Failure to Yield Right-of-Way

No Clear Factor

ContributingF

actor


Figure 5.9.2 -Top motorist contributing actors

0 10% 20% 30% 40% 50%

Bicyclist Making Right Turn

Bicyclist Slowing, Stopping or Starting in Trac

Bicyclist Making Left Turn

Bicyclist Riding Against Trac

Bicyclist Riding With Trac

Bicyclist Riding Across Roadway

Pr

e-CrashManeuver


Figure 5.9.3 -Top bicyclist pre-crash maneuvers

0 10% 20% 30% 40% 50%

Vehicle Parked Legally

Vehicle Making Right Turn on Red

Vehicle Starting in Trac

Vehicle Making Right Turn

Vehicle Making Left Turn

Vehicle Following Roadway

Pre-CrashManeuver

Percenta e of Crashes

Figure 5.9.4 -Top motorist pre-crash maneuvers


19/30


5.10 Specifc Crash Types and Other Crash

Attributes

Motorist Failure to Yield Right-o-Way While Turning

While contributing actors and pre-crash maneuversprovide understanding o what is causing crashes,combining the two attributes can bring urtherspecicity to crash events. A greater level o detail is

especially useul or unspecic contributing actorssuch as ailure to yield right-o-way.

Failure to yield right-o-way is the most prevalentcontributing actor among motorists, accountingor 32 percent o crashes. In 38 percent o motoristailure to yield crashes, or 12 percent o all bicyclist-motorist crashes, the motorist ailed to yield right-o-way while making a let turn. In 22 percent omotorist ailure to yield cases, or seven percent o allbicyclist-motorist crashes, the motorist ailed to yieldwhile making a right turn.

case b

Motoristfailure to yield

right-of-way

while making a

left turn

Combining contributing factors withpre-crash maneuvers can provide a

greater understanding of crash events.

However, even with increased detail,

many attributes remain uncertain. In this

case, the large number of possible

bicyclist pre-crash maneuvers

highlight the issue.

12% of crashes 7% of crashes

Motoristfailure to yield

right-of-way

while making a

right turn

Typical

Right Hook

Motorist

turning

right

Motorist

turning

left

Motorist

turning

right

case c

Motorist

turning

left

Bicyclist riding

across roadway

Bicyclist riding

with trac

Bicyclist riding

with trac

Bicyclist riding

across roadway

Typical

Left Hook

case a

STOP

STOP STOP

STOP

Figure 5.10.1 - Example cases o motorists ailing to yield right-o-way while turning.

While the task o combining motorist attributes canyield greater understanding, it can stil l leave manydetails unknown. Below, Figure 5.11.1 illustrates ourexamples o this challenge. In each case, motoristsare ailing to yield while making turns. This is asimple event to depict, as a turning motorist is astraightorward maneuver.

However, what were the actions o the bicyclist?Using the case in the upper right, a typical righthook crash would involve a bicyclist riding withtrac in the roadway ( case a). However, a bicyclistcould have also been riding with trac, but on thesidewalk ( case b or c ). With little knowledge aboutsidewalk riding, we cannot say or sure. This issueis repeated in other examples below, and likelyincludes many other scenarios not shown here.


20/30


Disregarding a Trafc Control Device

Disregarding a trac control device was a prevalentcontributing actor or both bicyclists and motorists.Bicyclists disregarded a signal in 13 percent ocrashes and motorists in ve percent o crashes.When a bicyclist disregards a trac control device,it is a trac signal 74.5 percent o the time. In the

remainder o cases it is a stop sign or other device.When a motorist disregards a trac control deviceit is a trac signal 67.8 percent o the time. In theremainder o cases it is a stop sign or other device.

Bicyclist Failure to Use Headlight

Bicyclist ailure to use a headlight was cited as aprimary contributing actor in 1.6 percent o allcrashes. Isolating unique crashes occurring when itwas dark, raining or snowing, bicyclist ailure to usea headlight was a primary contributing actor in 5.8percent o crashes. Motorist ailed to use headlightsin less than 0.1 percent o all crashes.

Dooring

Dooring is a colloquial term describing a crashwhere a motorist opens a door o a parked vehicleinto the path o the bicyclist. There is currently noDPS code that explicitly labels a crash as dooring.However, pre-crash maneuvers and contributingactors can provide an estimate or the number odooring cases. In 2.7 percent o crashes the motoristhad their vehicle legally parked. In two out o three othese cases, the driver was inattentive or distractedor cited or other human actors. Most crashes with

these attributes occur along primary arterials withparking on both sides o the streets. Based on thiscombination o actors it can be presumed thatmany o these crashes are cases o dooring. Detailedanalysis o 800 crash reports rom 2006-2008 oundthat dooring occurred in 2.3 percent o crashes, sothis assumption may be accurate.

Sidewalk Riding and Bicyclist Position

Bicyclist pre-crash maneuver position is notconsistently collected on accident reports, althoughthe riding position is oten noted in the crash

narrative or diagram. Detailed analysis o 800 crashreports rom 2006-2008 ound that 33 percent ocrashes involved a bicyclist entering trac romthe sidewalk or path. While sidewalk riding itselis not a contributing actor, decreased visibility orboth bicyclists and motorists makes this attributenoteworthy and a consideration or urther research.

P

?

Disregarding a

trac control device

Bicyclist

failure to use

headlights

1.6% of

all crashes

5.8% of

crashes at

night,

raining,

snowing

Dooring

Analysis based on parked motor

vehicle + inatttentive or distracted

motorist or motorist other human

factors.

Bicyclist riding

on sidewalk

Examples based

on common cases

listed on DPS

accident reports -

specic results arenot yet tabulated.

When disregarding a trac

control device, both bicyclists

and motorists tend to run trac

signals, rather than stop signs

or other devices.

Bicycles are

vehicles and

are legally

required to

use headlights

at night.

STOP

Motoristdisregarding

a trac control device

(5% of crashes)

Motorist opening car door

into path of bicyclist

(estimated 2.7% of crashes)

Bicyclistdisregarding

a trac control device

(12% of crashes)

Motorist entering

trac from alley

Motorist turning

into driveway

Bicyclist

entering

trac from

sidewalk

Figure 5.10.2 - Other specic bicyclist-motorist crash typesand attributes


21/30


cases involved hit-and-runs, so no rst-hand partywas present to provide inormation.

Due to the small sample, it is dicult to drawdenitive conclusions about atal crashes. However,all crashes had at least one o the ollowing three

characteristics:

Large motor vehicle, or Rain or wet pavement, or Motorist had been drinking or was

aggressive

The two bicyclist atalities that occurred in 2011ollowed these trends (truck and aggressivemotorist). While inclement weather and aggressivemotorists are circumstances o little control, there isa large potential surrounding the education driverso large vehicles, as most are proessional drivers o

trucks or buses. These opportunities are discussed inChapter 6.

5.11 Injury Severity

Public Works records the most severe injury listed onaccident reports. Injuries are categorized as Type A, Bor C and are listed in decreasing order o severity:

Type A: incapacitating injury Type B: non-incapacitating injury

Type C: possible injury (victim complains opain or discomort)

Eighty-seven percent o bicyclist-motorist crashesresult in some level o injury. Sixty percent o crashesresulted in Type C injuries, 23 percent Type B and ourpercent Type A. For 13 percent o crashes, the injurywas unknown or other. Detailed analysis o 800 crashreports rom 2006-2008 ound that when an injurywas sustained, it was always the bicyclist. Motoristssustained injuries in no crashes.

5.12 Fatalities

There were 12 bicyclist atalities rom 2000-2010, oran average o 1.1 bicyclist atalities per year. Therewere no atalities in 2001, 2004 and 2005, whilethree occurred in 2008. Fatal crashes are prevalentthroughout the year, although they are concentratedin warm weather months. All crashes occurred on aweekday and are most prevalent in the morning peakperiod and during the midday. Two crashes occurredwhen it was dark.

O the 12 atal crashes, contributing actors and pre-crash maneuvers are representative o all crashes.

However, because the bicyclist was killed, thebicyclist could not recount any inormation. Only themotorist and available witnesses were present. Two

Figure 5.12.2 -Characteristics o atalbicyclist-motoristcrashes, 2000-2011

*Hit & Run** Involved a motorized

bicycle

Figure 5.12.1 - A ghost bike memorial at the s ite oa atal crash on 15th Ave SE.

Year LocationLargeMotor

Vehicle

Rain or WetPavement

MotoristImpaired orAggressive

2000 12th St N & W Linden Ave X X

2002 3rd St NE & Lowry Ave NE X

2003 52nd Ave N & James Ave N X

2006 University Av SE & Washington Av SE X

2007W Lake St & Dean Pkwy S X

Broadway St NE & Quincy St NE* X X

2008

Washington Av N & Dowling Av N X X

W Excelsior Blvd & W 32nd St* X

5th St S & Nicollet Mall X X

2009 14th St E & Park Av S X

2010Monroe St NE & 15th Ave NE** X

1st Ave N & 5th St N X

201115th Ave SE & 4th St SE X

W River Pkwy & E Franklin Ave X


22/30


5.13 The Cost o Crashes

The cost o crashes is dicult to calculate andapplying a monetary value to a severe or atal eventcan be a sensitive assignment. Nonetheless, boththe State o Minnesota1 and the Federal HighwayAdministration2 have developed estimates todetermine the economic impact o crashes.

The dollar gures support cost-benet estimateswhich can be used to analyze and prioritize the costo saety countermeasures. For example, an agencymay be able to justiy transportation improvementsor actions i improvements would be expected tocorrect specic crash types.

Using these costs estimates, a Type C injury rangesrom $12,400 to $44,900 and a atal injury rangesrom $1,290,000 to $4,008,900. From 2000-2010bicyclist-motorist crashes in Minneapolis had anestimated cost between $63.7 and $320.0 million, or

$5.8 to $29.1 million per year.

1 Minnesota Department o Public Saety. Minnesota Motor VehicleCrash Facts 2010. 2011. www.dps.mn.gov2 Federal Highway Administration. Pedestrian and Bicyclist TracControl Device Evaluation Methods. Publication No. FHWA-HRT-11-035. U.S. Department o Transportation. May 2011.

Figure 5.12.4 - From 2000-2010 the cost obicyclist-motorists crashes is estimated to bebetween $63.7 to $320.0 million.

Figure 5.12.3 Estimate Cost o Bicyclist-Motorist Crashes, 2000-2010

In Minneapolis

InjurySeverity

MN DPS TotalCost

FHWA HumanCapital Cost

FHWA SocialCost

Number oCrashes

Cost Estimate(millions)

Fatal $1,290,000 $1,245,600 $4,008,900 12 $15.5 - 63.1

Type-A $67,800 $111,400 $216,000 122 $8.3 - 39.9

Type-B $21,900 $41,900 $79,000 671 $14.7 - 81.1

Type-C $12,400 $28,400 $44,900 1,781 $22.1 - 130.5

None $8,200 $6,400 $7,400 387 $3.2 - 5.3

Total 2,973 $63.7 - 320.0


23/30


bicyclist-motorist crashes are occurring at or nearintersections. Forty-one percent o crashes occur atintersections, 40 percent occur within 50 eet andless than 19 percent occur at distances greater than50 eet. For reerence a standard city block is 250 eetwide by 600 eet long.

Top Crash Intersections

The most prevalent locations or crashes are alongthe citys busiest streets. Primary arterials haveseen the highest numbers o crashes comparedto residential streets. Four o the top ten crashintersections are located on Franklin Avenue.

5.14 Where Crashes are Occurring

Bicyclist-motorist crashes are occurring in all partso the city, although crashes are most prevalent indowntown, South and Southeast Minneapolis. Thereis also a clear concentration along major arterials.

To simpliy the discussion and illustration o the

results, crash locations are aggregated to the closestintersection. This is a good assumption as most

Figure 5.14.2 -Top Bicyclist-MotoristsCrash Intersections, 2000-2010

Intersection Crashes

1 E Franklin Ave - Cedar Ave S 20

2 Hennepin Ave S - 7th St N 19

3 Hennepin Ave S - 3rd St N 17

4 Hiawatha Ave S - E 26th St 17

5 W Franklin Ave - Nicollet Ave S 17

6 W Franklin Ave - Lyndale Ave S 16

7 University Ave SE - I-35W NB Ramp 14

8 Portland Ave S - E 28th St 14

9 Lyndale Ave S - W Vineland Place 14

10 E Franklin Ave - Chicago Ave S 13

Lake

ofthe

Isles

Cedar

Lake

Lake

Nokomis

Lake

Hiawatha

LakeCalhoun

LakeHarriet

M

i

s

s

i

s

s

i

p

p

i R

i

v

e

r

Figure 5.14.1 - Citywide bicyclist-motorist crash density, 2000-2010

Crashes PerIntersection

1

2-3

4-5

6-9

10-20


24/30


Corridor Crashes and Crash Rates

Examining the map on the previous page, it is clearto see that crashes are concentrated along majorarterials, although this pattern is not exclusive tobicyclist crashes. For this research, 28 corridors wereselected or urther analysis. Corridors were basedon the number o crashes and the signicance o the

street or bicyclists.

O the 28 corridors analyzed, those with thehighest number o crashes rom 2000-2010 wereEast-West Lake Street (226 crashes), East-WestFranklin Avenue (205), Portland Avenue South(127), Hennepin Avenue South in Downtown (126)and Lyndale Avenue South (111). Note that somecrashes are duplicated across corridors because ointersecting streets.

While the number o crashes is a good indicator orsaety, crash rate provides an alternative perspective.

Developing crash rates provides context and allows

or better comparison across corridors o varyingmagnitude. For example, a corridor with 1,000bicyclists per day has a higher exposure index thana corridor with only 100 bicyclists per day. Crashrates were developed using bicycle trac countsconducted by Public Works. A map o Minneapolisbicycle trac can be ound in Appendix E and acomplete explanation o the crash rate model used is

provided in Appendix F. Crash rates are expressed ascrashes per million bicycle miles traveled (BMT).

The corridors with the highest crash rates rom2000-2010 are East-West 28th Street (68.5 crashes permillion BMT), Lowry Avenue North-Northeast (55.4),Marquette Avenue South (39.5), East-West 26thStreet (39.2) and West Broadway-Broadway StreetNortheast (39.1). Detailed analysis o all 28 corridorsis supplied in Appendix H.

Lake

ofthe

Isles

Cedar

Lake

Lake

Nokomis

Lake

Hiawatha

LakeCalhoun

LakeHarriet

M

i

s

s

i

s

s

i

p

p

i R

i

v

e

r

Figure 5.14.3 -Top 28 bicyclist-motorist crash corridors by number o crashes (let) and crash rate (right), 2000-2010

Lake

ofthe

Isles

Cedar

Lake

Lake

Nokomis

Lake

Hiawatha

LakeCalhoun

LakeHarriet

M

i

s

s

i

s

s

i

p

p

i R

i

v

e

r

Top Bicyclist-Motorist CrashCorridors by Number o Crashes

Top Bicyclist-Motorist CrashCorridors by Crash Rate

100 - 226 30.0 - 68.5

Number o Crashes Crash Rate*

*Crash rate is expressed as crashes per one million bicycle miles traveled (BMT)

50 - 99 15 - 29.9

20-49 7.7-14.9


25/30


Concentrations o Crash Attributes

Just as crashes can be mapped, crash attributescan, too. Knowing where prevalent contributingactors and pre-crash maneuvers are occurring canpinpoint needs and inorm appropriate enorcement,education and engineering countermeasures.While many attributes do not have any specic

geographical concentrations, several do. The maps tothe right show two examples: the top intersectionsor bicyclists disregarding a trac control device (let)and the top locations where motorists ailed to yieldthe right-o-way while making a right turn (right).Maps o nearly all crash attributes mentioned in thischapter can be ound in Appendix F.

Saety in Numbers

As discussed in Chapter 2, recent increases inbicycle trac have exhibited a saety in numbersphenomenon. Counter to intuition, a motorist is lesslikely to collide with a bicyclist when there are morepeople bicycling. Saety in numbers has been shownover time at a city-wide level, but it also holds trueacross localized corridors in Minneapolis.

In developing the corridor crash rates on theprevious page, a slight, but noticeable trendemerged. While speculative, it appears that corridorswith more bicycle trac tend to have lower crashrates. Streets like Hennepin Avenue in downtownand University Avenue Southeast have lower crashrates than streets like Lowry Avenue North-Northeastor East-West 28th Street. The presence o a bicycle

acility along a corridor also appears to be a actor osaety. Below, Figure 5.14.5 illustrates this promisingtrend.

The results are especially encouraging because itmay be a sign that motorists are coming to expectbicyclists on certain streets. Passing a bicyclist on thestreet is more and more a normal event rather thana rare occurrence. While engineering, enorcementand education should support a sae environment,bicycle trac itsel is playing a large role in makingstreets sae.

Figure 5.14.4 - Examples o mapped crash attributes. Topintersections or bicyclist disregarding a trac controldevice (let) and top locations or motorists ailure to yieldright-o-way while making a right turn (right).

0 100 200 300 400 500 600 700 800 900 1,000 1,100 1,200 1,300 1,400 1,500 1,6000

10

20

30

40

50

60

70

80

Daily Bicyclist Trac

CrashRate(crashespermillionBMT)

From 2000-2010:

Hennepin Ave S

(downtown)University

Ave SE

E-W 28th St

Lowry Ave N-NE

Most of the corridor did nothave a bicycle facility.Most of the corridor did have a bicycle facility.

Logorithmic Trend Line (R2 = .56)

Figure 5.14.5 - Bicyclist-motorist crash rate vs. daily bicyclist trac


26/30


6.1 Summary o Findings

To summarize the data presented in the previouschapter,

Crashes are most prevalent in the summermonths, on weekdays and in the aternoonpeak period.

Crash weather conditions are generally clearand dry and occur in daylight.

One out o ve crashes involves a hit-and-run - bicyclists are disproportionately thevictims.

Bicyclists and motorists are generally notimpaired at the time o crashes.

Bicyclists and motorists appear to be equallycontributing to crashes.

Motorists are oten inattentive or aredistracted at the time o crashes, or areailing to yield the right-o-way to bicyclists.

Bicyclists are oten ailing to the yield right-o-way, disregarding trac control devicesand riding against trac.

Bicyclists sustain injuries in most crashes. Nomotorists appear to have sustained injuries.

The circumstances o bicyclist atalitiesadhere to a clear pattern involving largemotor vehicles, inclement weather oraggressive or impaired motorists.

Most crashes are occurring at or near

roadway intersections. There is a clear concentration o crashesalong major arterials.

There is an apparent saety in numbers crash rates tend to be lower on streets withmore bicycle trac.

Reducing these ndings urther, three primaryconclusions emerge:

1. Most crashes are occurring at intersectionsalong major arterials,

2. Motorists are not seeing or yielding to

bicyclists and3. Bicyclists are not riding in a predictable

manner.

6.2 Approaching Bicyclist Saety

These three conclusions help simpliy the complexnature o crashes. However, translating the ndingsinto efective countermeasures is the next task.

While posed with good intentions, the discussiono countermeasures can quickly become detailedand itemized: Which intersections should be improvedrst? How should bicyclists be educated? How should

motorists be educated? How can the police be involved?

Beore moving orward, a ramework orimplementation should be established and a clearunderstanding o who the countermeasures areintend or is needed. Saety is an evolving goaland it may be better to ront load the discussionwith high-level considerations, rather than speciccountermeasures. The approaches to bicyclist saetydiscussed in this section revolve around two ideas:1) The Six Es o Bicycling and 2) The Four Types o

Transportation Cyclists.

Six Es o Bicycling

In order to support a great bicycling community,the League o American Bicyclists recommends abalanced approach o the ollowing six categories:

Equity, Engineering, Enorcement,

Education, Encouragement and Evaluation

Known as the Six Es o Bicycling1, this straightorwardapproach is becoming the norm in cities acrossthe U.S. and was used as a ramework or the 2011Minneapolis Bicycle Master Plan. While originallyintended as a checklist or increasing bicycling, it caneasily be applied to decreasing crashes. Developinga set o countermeasures to increase bicyclist saetyshould use the Six Es approach.

Four Types o Transportation Cyclists

The Portland Bureau o Transportation developeda demographic spectrum known as the Four Types

1 League o American Bicyclists. Cyclists Equity Statement.www.bikeleague.org

Chapter 6 - Discussion, Approaches and Recommendations


27/30


may have sparked the next segment o the spectrum.As the demographic o bicyclists turns a corner, soshould approaches to bicyclist saety.

Pulling it All Together

The photo below depicts a typical bicyclist at a majorintersection in Minneapolis. The bicyclist appears

condent and motorists are giving the rider space.However, the scene still brings validity to the threemain conclusions o this report. The city s busy streetsand intersections were not originally designed orbicycle trac. Bicyclists are relatively new to modernroadways and not all motorists have come to expectthem. And lastly, the lack o designated space andthe large number o novice riders likely contributeto the large number o unpredictable maneuvers bybicyclists.

A complete retrot o Minneapolis or bicycle trac

is decades in the making. Nonetheless, there arenumerous small scale eforts that can reduce crashesand encourage the interested but concernedto ride a bicycle or more trips. Where large-scaleengineering is not easible, many o the other Escan ll in the gaps.

o Transportation Cyclists.2 Based on surveys andacademic research, the spectrum is a revealingestimate o who bikes or transportation and whodoes not. Most importantly, the research nds that alarge part o the population mayride or regular trips,although they currently have reservations aboutdoing so. The our categories include the strong and

earless, enthused and condent, interested butconcerned, and no way no how.

The strong and earless tend to be experiencedbicyclists and will ride on any street regardless othe acility. Enthused and condent riders maybe newer to bicycling or transportation. They arecondent riding in trac, but preer a bike lane orother acility. Interested but concerned riders arethe largest group, comprising nearly two-thirds othe population. They may ride on trails or recreation,but are currently earul o riding on streets andwith trac. For this group, saety is strongly tied to

comort. Lastly, the no way no how group havenever biked beore and are not interested in bicyclingor transportation, or otherwise.

Throughout the 1990s, most bicyclists ell into thestrong and earless group. As bikeways were addedthroughout the 2000s, the enthused and condentgroup likely burgeoned. A bicycle commute modeshare data o 3.4 percent supports this estimate.3In 2011 and 2012, ater signicant increases inon-street bikeways such as bufered bike lanes orbicycle boulevards, there was a small, but noticeableincrease in women, youth, seniors and amiliesbicycling in Minneapolis. These groups are seenas good indicators o interested but concernedbicyclists. While too early to measure, Minneapolis

2 Geller, Roger. The Four Types o Transportation Cyclists. PortlandBureau o Transportation. 2007.3 U.S. Census Bureau. 2011 American Community Survey.www.census.gov

Strong and Fearless (


28/30


6.3 Recommendations or Improved Bicyclist

Saety

Over the past decade, Minneapolis has made greatstrides in the area o bicyclist saety. This analysisconrms that many o the improvements made areefective and should continue. The ndings alsohighlight the need or new ocus areas, including an

expanded relationship with the Minneapolis PoliceDepartment and continued use o best practices inengineering. The recommendations or improvedbicyclist saety are the ollowing:

Equity

Assign Bicycle and Pedestrian Section staf

to the Crash Reduction Committee. PublicWorks currently has a Crash ReductionCommittee that develops countermeasuresor all types o crashes. Representation romthe Bicycle and Pedestrian Section can help

incorporate the ndings o this report intoroutine decisions and evaluation.

Ensure that saety talking points equallyaddress motorists and bicyclists. It iseasy to assign blame to the other party;however, data has revealed that motoristsand bicyclists are equally contributing to thecauses o crashes. Any orthcoming saetycampaigns should remain balanced.

Engineering

Guide and protect bicyclists atintersections and on busy streets. Mostcrashes are occurring at intersections andalong major arterials. Protected signalphases or separated approaches may givebicyclists condence when riding throughcomplex spaces.

Highlight areas where bicyclists andmotorists cross paths. There are manyknown locations in the city where turningmotorists are ailing to yield the right-o-wayto bicyclists. Potential conict areas maybe highlighted with enhanced pavementmarkings.

Provide designated and comortable

places or bicyclists to ride. Bicyclists arenot always riding in a predictable manner.While much o this is simply improperriding and illegal behavior, existing roadwaydesign may contribute to risky maneuvers.Providing designated space or bicycletrac can oster more predictable riding andincrease bicyclist comort.

Most crashes are occurring at busy intersections.Bicycle signals can provide riders wi th an exclusiveor protected signal phase, allowing them to saelynavigate the intersection.

Saety campaigns like this one rom Fort Collins,Colorado address motorist andbicyclist behavior. Anyorthcoming saety messages should remain balanced.

Providing designated space or bicyclists can increasethe predictability o where bicyclists ride and create amore comortable riding experience.

Colored pavement markings can alert motorists thatthey are crossing a bike lane and should yield tobicyclists beore turning or merging.


29/30


Enorcement

Expand a relationship with the MPD.Theprimary actions o Public Works and thePolice Department are separate. However,both departments share the goal o creatingsaer streets. This common goal should beexplicitly recognized between department

management and expand collaboration atall ranks.

Ensure bicyclists and motorists are

treated equally under the law. Bicycles arelegally trac and should be enorced andprosecuted no diferently than other modes.Current practices indicate this may not bethe norm or either bicyclist inractions orbicycle-related motorists inractions.

Use enorcement as an educational

tool. While enorcement should be anauthoritative action, it can also educate.

Many motorists and bicyclists simply donot know the rules o the road. Diversionprograms can allow rst time ofenders achance to learn the rules while avoiding aull penalty.

Education

Educate proessional drivers. Getting theword out to all drivers is dicult. Luckily,there are a large number o proessionaldrivers who are held to high saetystandards and receive regular training.

Reaching out to proessional drivers hasalready proven successul. A curriculumdeveloped by the City, Minneapolis PublicSchools, the Minnesota Truckers Associationand the Bicycle Alliance o Minnesota hasbeen presented to over 400 truck andtransit drivers in 2012 alone. Other eventslike Behind the Big Wheels have allowedbicyclists to sit in the drivers seat o trucksand buses, educating bicyclists about saetyaround large vehicles.

Use media to reach a wide audience. Fornon-proessional drivers and bicyclists,media can eciently disseminate saetymessages quickly and broadly. Public servicemessages produced by the City are currentlyavailable online and are played regularly onlocal television. Trac saety is also a topico great public interest. Using news mediato circulate messages has proven to be avaluable option.

Rea

rWh

eelTr

ack

MidtownGreenway

M i d t o w n G r e e n w a y

Th

ru

B i

cy

c l i

s

tTrail

t t tt i t t it t i

i i i tt i i t it i

t i t t it t i i t it

MidtownGreenway

M i d t o w n G r e e n w a y

Educating proessional drivers should remain apriority. A curriculum developed by the City and otherlocal partners (above) has been presented to over 400truck and transit drivers in its rst year.

Above, the Bicycle Advisory Committee meets withMPD staf to discuss enorcement eforts. Expandinga relationship with the MPD is essential to improvingsaety or all road users.

Saety videos produced by the City have been widelyviewed by the public. Media should continue to beused to eciently disseminate saety messages.

Behind the Big Wheels events allow bicyclists toexperience the challenges o driving a large vehicle- bringing greater awareness to blind spots andturning movements.


30/30

Continue rides and classes. Local certiedbicycling instructors lead dozens o ridesand classes each year, reaching hundredso bicyclists. While curriculum oten ocuseson commuting, maintenance and routeplanning, saety is always an underlyingtheme. Small group instruction shouldcontinue to occur and the ndings o this

research should be incorporated into uturecurriculum.

Encouragement

Design inrastructure that is perceived

to be sae. Encouragement can utilizesot orms like media, events, literatureand online materials. However, the mostvisible bicycling element in the city isthe inrastructure itsel. Interested butconcerned bicyclists will not ride unlessthey see a comortable place to ride.

Publish data and document results.Actual saety can help inorm perceivedsaety. Publishing data and reporting oncountermeasures will hold Public Worksaccountable and can let road users know isaety is improving.

Evaluation

Publish a regular saety report. This reportis the rst step in understanding bicyclistsaety in Minneapolis. To monitor changesand evaluate uture countermeasures,

continuous and regular reporting is needed.

Increase understanding o crashes. Whilea there is now a greater understanding owhat is causing crashes, many circumstancesremain unclear. When and where easible,additional attributes rom DPS accidentreports should be recorded by Public Works.

Facilities like bufered bike lanes (above) and cycletracks have a high degree o perceived saety and canattract a new demographic o bicyclists.

Public Works and other local organizations teachdozens o bicycle classes every year. Curriculumshould incorporate the new ndings o this report.

Public Works currently maintains an online c rashdatabase that is available to the public. This datashould remain accessible to all stakeholders.

When and where easible, additional attributes romDPS accident reports should be analyzed to gain agreater understanding o crash events.

Bike-Car Crash Study

Documents

Transcript of Bike-Car Crash Study