channeiization

7/25/2019 ChanneIization

http://slidepdf.com/reader/full/channeiization 1/149

Web-Only Document 208:

Design Guidance for Channelized Right-Tur n Lanes

National Cooperative Highway Research Pr ogram

Ingrid B. PottsDouglas W. Harwood

Karin M. Bauer David K. Gilmore

Jessica M. HuttonDarren J. Tor bic

MRIGlobalKansas City, MO

John F. Ringer tAndrew Daleiden

Kittleson & Associates, Inc.Reston, VA

Janet M. Barlow

Accessible Design for the BlindAsheville, NC

Contr actor’s Final Repor t for NCHRP Pr o ject 03-89

Submitted July 2011

NCHRP



ACKNOWLEDGMENT

This work was s ponsored by the Amer ican Association of State

Highway and Trans por tation Off icials (AASHTO), in cooperation with

the Federal Highway Administration, and was conducted in the

National Cooperative Highway R esearch Program (NCHRP), which is

administered by the Trans por tation R esearch Board (TRB) of the

National Academies.

COPYRIGHT INFORMATION

Author s herein are res ponsi ble for the authenticity of their mater ials and

for obtaining wr itten per missions from publisher s or per sons who own

the copyr ight to any previously published or copyr ighted mater ial used

herein.

Cooperative R esearch Programs (CRP) grants per mission to reproduce

mater ial in this publication for classroom and not-for-prof it purposes.Per mission is given with the under standing that none of the mater ial

will be used to im ply TRB, AASHTO, FAA, FHWA, FR A, FTA,

Transit Development Corporation, or AOC endor sement of a par ticular

product, method, or practice. It is expected that those reproducing the

mater ial in this document for educational and not-for-prof it uses will

give appropr iate acknowledgment of the source of any repr inted or

reproduced mater ial. For o ther uses of the mater ial, request per mission

from CRP.

DISCLAIMER

The opinions and conclusions expressed or im plied in this repor t are

those of the researcher s who perfor med the research. They are not

necessar ily those of the Trans por tation R esearch Board, the National

R esearch Council, or the program s ponsor s.

The infor mation contained in this document was taken directly from the

submission of the author(s). This mater ial has not been edited by TRB.



The National Academy of Sciences is a private, nonprofit, self-perpetuating society of distinguished scholars engaged in scientificand engineering research, dedicated to the furtherance of science and technology and to their use for the general welfare. Upon the

authority of the charter granted to it by the Congress in 1863, the Academy has a mandate that requires it to advise the federal

government on scientific and technical matters. Dr. Ralph J. Cicerone is president of the National Academy of Sciences.

The National Academy of Engineering was established in 1964, under the charter of the National Academy of Sciences, as a parallel

organization of o utstanding engineers. It is aut onomous in its administration and in t he selection of its members, sharing with the

Natio nal Academy of Sciences the respo nsibilit y for advising th e federal govern ment . The Nati onal Academy of Engineer ing also

sponsors engineering programs aimed at meeting national needs, encourages education and research, and recognizes the superior

achievements of engineers. Dr. C. D. Mote, Jr., is president of the National Academy of Engineering.

The Institute of Medicine was established in 1970 by the National Academy of Sciences to secure the services of eminent members

of appropriate professions in the examination of policy matters pertaining to the health of the public. The Institute acts under the

responsibility given to the National Academy of Sciences by its congressional charter to be an adviser to the federal government

and, upon its own initiative, to identify issues of medical care, research, and education. Dr. Victor J. Dzau is president of the

Institute of Medicine.

The Nation al Research Council was organized by the National Academy of Sciences in 1916 to associate t he broad community oscience and technology with the Academys purposes of furthering knowledge and advising the federal government. Functioning in

accordance with general policies determined by th e Academy, the Council has become th e pr incipal operating agency of both the

Natio nal Academy o f Sciences an d the Natio nal Academy of Engineerin g in pr ovidin g services to th e governm ent , the pu blic, and

the scientific and engineering communities. The Council is administered jointly by both Academies and the In stitute of Medicine.

Dr. Ralph J. Cicerone and Dr. C. D. Mote, Jr., are chair and vice chair, respectively, of the National Research Council.

The Transpor tation Research Board is one of six major divisions of the National Research Council. The mission of the Transporta-

tion Research Board is to provide leadership in transportation innovation and progress through research and information exchange,

conducted within a setting that is objective, interdisciplinary, and multimodal. The Boards varied activities annually engage about

7,000 engineers, scientists, and other transportation researchers and practitioners from the public and private sectors and academia,

all of whom contribute their expertise in the pu blic interest. The program is supported by state transportation departments, federal

agencies including the component administrations of the U.S. Department of Transportation, and oth er organizations and individu-

als interested in the development of transportation. www.TRB.org

www.national-academies.org



Contents

Figures............................................................................................................................................. iiTables ............................................................................................................................................. iiiAuthor Acknowledgements .............................................................................................................vAbstract .......................................................................................................................................... vi

Executive Summary...................................................................................................................... viiChapter 1. Introduction ...........................................................................................................1

Background............................................................................................................11.1R esearch Ob jectives and Scope.............................................................................31.2Organization of This R epor t ..................................................................................31.3

Chapter 2. State of Knowledge and Practice with Channelized Right-Turn Lanes ................5

Literature R eview ..................................................................................................52.1Highway Agency Exper ience ..............................................................................202.2

Chapter 3. Pedestr ian Behavior at Channelized Right-Turn Lanes ......................................28

Observational Field Studies.................................................................................283.1Interviews with Or ientation and Mobility S pecialists .........................................373.2Summary of Findings from Observational Field Studies and Interviews with3.3Or ientation and Mobility S pecialists ...................................................................39

Chapter 4. Traff ic Operational Analysis of Channelized Right-Turn Lanes ........................41Traff ic Operational Modeling..............................................................................424.1Base Modeling R esults for Each Conf iguration ..................................................464.2

Im pacts of Pedestr ians on Base Conf iguration R esults .......................................524.3 Im pacts of Geometr ic Character istics and Signal Phasing on Channelized4.4Right-Turn Lane Delay........................................................................................56Summary of Traff ic Operational Analysis Findings ...........................................594.5

Chapter 5. Safety Analysis of Channelized Right Turns ......................................................61Database Development ........................................................................................615.1Cross-Sectional Crash Analysis Approach..........................................................625.2

Cross-Sectional Crash Analysis R esults..............................................................645.3Summary of Safety Analysis ...............................................................................915.4

Chapter 6. Interpretation of R esults and Design Guidance...................................................93Application of Channelized Right-Turn Lanes....................................................946.1Design Issues R elated to Channelized Right-Turn Lanes ...................................956.2

Chapter 7. Conclusions and R ecommendations..................................................................100Conclusions........................................................................................................1007.1

R ecommendations..............................................................................................1017.2

Chapter 8. R eferences . ........................................................................................................103Appendix A. Design Guide for Channelized Right-Turn Lanes................................................ A-1Appendix B. R evised Text on Channelized Right-Turn Lanes for the

AASHTO Gr een Book ......................................................................................B-1

i



Figures

Figure 1. Typical Inter section with Channelized Right-Turn Lanes ...............................................1Figure 2. Channelized Right-Turn Lane Study Location for NCHRP Pro ject 3-78A (21) ...........13Figure 3. Blue Pavement Mark ing Treatment at Channelized Right-turn Lane (27 ) ....................15

Figure 4. Signs Used in Oregon Blue Bike Lane Program (27 ) ....................................................17Figure 5. Com bined Bicycle Lane/Right-Turn Lane (27 ) .............................................................18Figure 6. Traditional Bike Lane/Right-Turn Lane (28).................................................................18

Figure 7. Alternative Crosswalk Locations ...................................................................................22Figure 8. Isosceles-Tr iangle Island Shape Used by WisDOT .......................................................23Figure 9. Typical Channelized Right-Turn Lanes with Differ ing Entry Angles to the Cross Street

[Adapted From (14)]...................................................................................................24Figure 10. Channelized Right-Turn Lane with Vehicle Yielding to Pedestr ian............................28Figure 11. Observational Field Study Locations ...........................................................................30Figure 12. Data Collection Field Setup for an Inter section in Baltimore, Maryland.....................30

Figure 13. Data Collection Field Setup for an Inter section in San Francisco, California .............31Figure 14. R aised Crosswalk at Channelized R ight-Turn Lane in Boulder, Colorado..................36Figure 15. Inter section Conf igurations for Traff ic Operational Analysis......................................43Figure 16. R TOR and Pedestr ian Crossing Movement Considered in Analysis ...........................46

Figure 17. Delay Com par ison of Conf iguration 1 - Conventional Right-Turn Lane (With andWithout R TOR ) ..........................................................................................................47

Figure 18. Delay for Conf iguration 2 (Yield-Controlled Channelized Right-Turn Lane) ............48

Figure 19. Delay for Conf iguration 3 (Signalized Channelized Right-Turn Lane).......................48Figure 20. Delay Com par ison of Conf igurations 1, 2, and 3.........................................................49Figure 21. Right-Turn Delay R eduction Due to a Channelized Right-Turn Lane Conf iguration 1

(R TOR ) Vs. Conf iguration 2......................................................................................50Figure 22. Right Turn Delay R eduction Due to a Channelized Right-Turn Lane Conf iguration 1

(Without R TOR ) Ver sus Conf iguration 2 ..................................................................51Figure 23. Delay Due to Pedestr ian Crossings — Conf iguration 2 Channelized Right Turn Lane

(300 veh/h Right-Turn and 800 veh/h Conf licting Through) .....................................53Figure 24. Delay Due to Pedestr ian Crossings — Conf iguration 2 Channelized Right-Turn Lane

(800 veh/h Conf licting Through Volume)..................................................................54

Figure 25. Pedestr ian Delay Waiting for a Gap in Right-Turning Traff ic.....................................55Figure 26. Delay Com par ison With Acceleration Lane (Conf iguration 2) ...................................56Figure 27. Right-Turn Over lap ......................................................................................................58Figure 28. Inter section Turning Movements R elative to an Inter section Approach (Approach 1)

with a S pecif ic Right-Turn Treatment (CR T, R TL, or STR ) .....................................63Figure 29. Summary of Analysis Model 1 Inputs and Cr iter ia......................................................65Figure 30. Summary of Analysis Model 2 Inputs and Cr iter ia......................................................69

Figure 31. Summary of Analysis Model 3 Inputs and Cr iter ia......................................................73Figure 32. Summary of Analysis Model 4 Inputs and Cr iter ia......................................................76Figure 33. Summary of Analysis Model 5 Inputs and Cr iter ia......................................................79Figure 34. Summary of Analysis Model 6 Inputs and Cr iter ia......................................................83

Figure 35. Summary of Analysis Model 7 Inputs and Cr iter ia......................................................88Figure 36. Illustration of Three Right-Turn Treatment Types — STR , R TL, and CR T. ................93Figure 37. Right-Turn Over lap ......................................................................................................99

ii



Tables

Table 1. Com par ison of Crash History for Common Right-Turn Treatments (4)...........................6Table 2. Annual Right-Turn Crashes by Type of Right-Turn Treatment (5) ..................................7Table 3. Locations Where Highway Agencies Place Pedestr ian Crosswalk s at Channelized

Right-Turn R oadways (3)...........................................................................................22Table 4. Innovative Traff ic Control Devices at Channelized Right-Turn R oadways (3)..............25Table 5. General Strategies Used by Highway Agencies to Assist Pedestr ians with Vision

Im pair ment (3) ............................................................................................................26Table 6. Pedestr ian Issues Considered in Deter mining the R adius or Width of Channelized

Right-Turn R oadway (3).............................................................................................27Table 7. Vehicle and Pedestr ian Counts Dur ing Evening Peak Per iod (5:00 p.m. to 6:00 p.m.)..33Table 8. Com par ison of Pedestr ian and Motor ist Behavior Between Boulder, Colorado, Sites

and Other Sites............................................................................................................37Table 9. HCS and Synchro Right-Turn Movement Delay Cali bration R esults.............................45

Table 10. Delay Im pacts of Crosswalk Location (300 vph and 50 ped/h) ....................................54Table 11. Delay Im pacts of Channelized R ight-Turn Lane S peed/R adius for Conf iguration 2....57Table 12. Delay Im pacts of Adding Additional Green Time to Right-Turn Movement ...............58Table 13. Mean and Median Motor-Vehicle Volumes by Inter section Type—Analysis

Model 1.......................................................................................................................66Table 14. Seven (7)-Year Total and FI Motor-Vehicle Crash Counts by Inter section

Approach—Analysis Model 1....................................................................................66

Table 15. R egression R esults for Motor-Vehicle Crash Models —Analysis Model 1...................67Table 16. Year ly Motor-Vehicle Crash Predictions —Analysis Model 1......................................68Table 17. Mean and Median Motor-Vehicle Volumes by Inter section Type—Analysis


Approach—Analysis Model 2....................................................................................70Table 19. R egression R esults for Motor-Vehicle Crash Models —Analysis Model 2...................71

Table 20. Year ly Motor-Vehicle Crash Predictions —Analysis Model 2......................................71Table 21. Mean and Median Motor-Vehicle Volumes by Inter section Type—Analysis

Model 3.......................................................................................................................72

Table 22. Seven (7)-Year Total and FI Motor-Vehicle Crash Counts by Inter sectionApproach—Analysis Model 3....................................................................................72

Table 23. R egression R esults for Motor-Vehicle Crash Models —Analysis Model 3...................74Table 24. Year ly Motor-Vehicle Crash Predictions —Analysis Model 3......................................74

Table 25. Mean and Median Motor-Vehicle Volumes by Inter section Type—AnalysisModel 4.......................................................................................................................75

Table 26. Seven (7)-Year Total and FI Motor-Vehicle Crash Counts by Inter section Approach—

Analysis Model 4........................................................................................................75Table 27. R egression R esults for Motor-Vehicle Crash Models —Analysis Model 4...................77Table 28. Contrast R esults for Motor-Vehicle Crash Models —Analysis Model 4.......................78Table 29. Year ly Motor-Vehicle Crash Predictions —Analysis Model 4......................................78

Table 30. Mean and Median Motor-Vehicle Volumes by Inter section Type—AnalysisModel 5.......................................................................................................................80

iii



Table 31. Seven (7)-Year Total and FI Motor-Vehicle Crash Counts by Inter sectionApproach—Analysis Model 5....................................................................................80

Table 32. R egression R esults for Motor-Vehicle Crash Models —Analysis Model 5...................81Table 33. Year ly Motor-Vehicle Crash Predictions —Analysis Model 5......................................81Table 34. Mean and Median Motor-Vehicle Volumes by Inter section Type—Analysis


Approach—Analysis Model 6....................................................................................84Table 36. R egression R esults for Motor-Vehicle Crash Models —Analysis Model 6...................84

Table 37. Contrast R esults for Motor-Vehicle Crash Models —Analysis Models 6 .....................86Table 38. Year ly Motor-Vehicle Crash Predictions —Analysis Models 6 ....................................87Table 39. Mean and Median Motor-Vehicle and Pedestr ian Volumes by Inter section

Type—Analysis Model 7............................................................................................89Table 40. Seven (7)-Year Total and FI Pedestr ian Counts by Inter section Approach—Analysis

Model 7.......................................................................................................................89Table 41. R egression R esults for Pedestr ian Crash Models —Analysis Model 7..........................90Table 42. Contrast R esults for Pedestr ian Models —Analysis Model 7 ........................................90Table 43. Year ly Pedestr ian Crash Predictions —Analysis Model 7 .............................................91

iv



Author Acknowledgements

The research repor ted herein was perfor med under NCHRP Pro ject 3-89 by MR IGlobal.MR IGlobal was the contractor for this study, with K ittelson and Associates, and Ms. Janet M.Bar low, Accessi ble Design for the Blind, serving as subcontractor s.

Ms. Ingr id B. Potts, P.E., Pr inci pal Traff ic Engineer at MR IGlobal, was the Pr inci palInvestigator for the research. The other author s of this repor t include Mr. Douglas W. Harwood,

Ms. Kar in M. Bauer, Mr. David K. Gilmore, Ms. Jessica M. Hutton, and Dr. Darren J. Torbic,MR IGlobal; Mr. John F. Ringer t and Mr. Andrew Daleiden, K ittelson and Associates; andMs. Janet M. Bar low, Accessi ble Design for the Blind. The work was done under the generalsupervision of Mr. Douglas W. Harwood, MR IGlobals Trans por tation R esearch Center Director.

v



Abstract

This repor t documents and presents the results of research to develop design guidance for channelized r ight-turn lanes. Observational f ield studies were conducted at 35 inter sectionapproaches in cities to assess pedestr ian crossing behavior, motor ist yield behavior, and the

interaction between pedestr ians and motor vehicles at channelized r ight-turn lanes. Simulationmodeling was perfor med to quantify the traff ic operational benef its of channelized r ight-turnlanes with var ious types of traff ic control and to com pare the delay reduction of channelized

r ight-turn lanes and conventional r ight-turn lanes. Crash data for near ly 400 inter sectionapproaches in Toronto, Ontar io, Canada, including inter section approaches with channelizedr ight-turn lanes, conventional r ight-turn lanes, and shared through/r ight-turn lanes, wereanalyzed to com pare the safety perfor mance of the three r ight-turn treatment types. The researchresults indicate that channelized r ight-turn lanes have a def inite role in im proving operations andsafety at inter sections. However, to achieve these benef its they should have consistent design andtraff ic control and should be used at appropr iate locations. The research provides design

guidance for channelized r ight-turn lanes that addresses geometr ic elements such as crosswalk location, s pecial crosswalk signing and mark ing, island type, radius of turning roadway, angle of inter section with cross street, acceleration and deceleration lanes, and traff ic control.

vi



Executive Summary

Many trans por tation agencies use channelized r ight-turn lanes to im prove operations atinter sections, although their im pact on safety for motor ists, pedestr ians, and bicyclists has not

been clear in the past. A key concern with channelized r ight-turn lanes has been the extent of

conf licts between vehicles and pedestr ians that occur at the point where pedestr ians cross ther ight-turn roadway.

For most pedestr ians, crossing the r ight-turn roadway is a relatively easy task because suchroadways are not very wide and because traff ic is approaching from a single direction. However,

pedestr ians with vision im pair ment may have diff iculty detecting approaching traff ic because(a) r ight-turning vehicles are traveling a curved rather than a straight path; (b) there is not asystematic stopping and star ting of traff ic, as there would be at a conventional signal- or stop-controlled inter section; and (c) the traff ic sounds from the ma jor streets may mask the sound of traff ic on the r ight-turn roadway. Des pite their potential challenges for pedestr ians, channelized

r ight-turn lanes also provide advantages for pedestr ians. The provision of a channelized r ight-turn lane, while mak ing it necessary for pedestr ians to cross two roadways, of ten reduces the

pedestr ian crossing distance of the ma jor and cross streets. Fur ther more, the channelizing island, par ticular ly when bounded by raised curbs, serves as a refuge area for pedestr ians, and may

im prove safety by allowing pedestr ians to cross the street in two stages.

The pr imary traff ic operational reasons for providing channelized r ight-turn lanes are to

increase vehicular capacity at an inter section and to reduce delay to dr iver s by allowing them toturn at higher s peeds and reduce unnecessary stops. Channelized r ight-turn lanes appear to

provide a net reduction in motor vehicle delay at inter sections where they are installed, althoughno existing data and no established methodology have been available to directly com pare theoperational perfor mance of urban inter sections with and without channelized r ight-turn lanes.

The safety effects of channelized r ight-turn lanes on motor vehicles, pedestr ians, and

bicyclists have been largely unknown. It is generally accepted that channelized r ight-turn lanesim prove safety for motor vehicles at inter sections where they are used, but there have been onlylimited quantitative data to demonstrate this. No studies have been found concerning pedestr ian

safety at channelized r ight-turn lanes that have used crash data to document the pedestr ian safetyim plications of channelized r ight-turn lanes. There also appear s to be an inherent r isk to

bicyclists at channelized r ight-turn lanes because motor vehicles enter ing the channelized r ight-turn roadway must weave across the path of bicycles traveling straight through the inter section,

but no studies based on crash history are available to suppor t this presum ption. However, asimilar conf lict between through bicyclists and r ight-turn vehicles is present at conventionalinter sections as well.

It is evident that there have been many unanswered questions about channelized r ight-turnlanes. The research conducted for this repor t sought to answer many of these questions and to

provide a basis for decisions, based on sound research results rather than anecdotal evidence,

about where channelized r ight-turn lanes should and should not be used. The ob jective of theresearch was to develop design guidance for channelized r ight-turn lanes, based on balancing theneeds of motor vehicles, pedestr ians, and bicycles. Key studies in the research included f ield

vii



observational studies of vehicle and pedestr ian interactions at channelized r ight-turn lanes,interviews with or ientation and mobility (O&M) s pecialists, traff ic operational analyses of channelized r ight-turn lanes conducted with the VISSIM model, and safety analyses of channelized r ight-turn lanes.

Observational f ield studies were conducted at 35 inter section approaches with channelizedr ight-turn lanes in 11 cities to assess pedestr ian crossing behavior, motor ist yield behavior, andthe interaction between pedestr ians and motor vehicles at channelized r ight-turn lanes. Ama jor ity of the sites (near ly 70 percent) had marked crosswalk s located near the center of the

channelized r ight-turn lane; only about 30 percent of crosswalk s were located at the upstream or downstream end of the channelized r ight-turn lane. Pedestr ians did not appear to have any

par ticular diff iculty crossing channelized r ight-turn lanes. Most pedestr ians crossed in thecrosswalk (75 percent) and, where pedestr ian signals were present, most pedestr ians crosseddur ing the pedestr ian crossing phase (72 percent). The pedestr ians who crossed outside the

crosswalk generally did so when no vehicular traff ic was present and this behavior did not causeany traff ic conf licts. Avoidance maneuver s by a pedestr ian or motor ist appear to be relativelyrare, and were observed in less than 1 percent of the pedestr ian crossings.

Most motor ists yielded to pedestr ians once they were in the crosswalk of a channelizedr ight-turn lane, either by stopping or reducing their s peed. Fewer motor ists in channelized r ight-

turn lanes yielded to pedestr ians waiting at the curb to cross, although such failure to yield istypical of most pedestr ian crossings and is not unique to channelized r ight-turn lanes. The yield

behavior of motor ists was slightly better at sites with s pecial crosswalk treatments (e.g., raised

crosswalk, pavement mark ings, signing).

Interviews with ten or ientation and mobility (O&M) s pecialists were conducted to learn of their exper ience in teaching pedestr ians with vision im pair ment to traver se inter sections with

channelized r ight-turn lanes and to deter mine their recommendations on how channelized r ight-turn lanes might be better designed for all pedestr ians with disabilities. O&M s pecialists did nothave a def inite preference for crosswalk location at channelized r ight-turn lanes, butrecommended that consistency of crosswalk location and traff ic control is im por tant to

pedestr ians with vision im pair ment. They also had a strong preference for raised islands with“cut-through” pedestr ian paths, which provide better guidance and infor mation for pedestr ianswith vision im pair ment than painted islands. O&M s pecialists also indicated that channelized

r ight-turn lanes with acceleration lanes were par ticular ly challenging for pedestr ians with visionim pair ment to cross.

A traff ic operational analysis was conducted to evaluate the traff ic operational perfor mance

of r ight-turning vehicle movements at signalized inter sections for three conf igurations: aconventional r ight-turn lane at a signalized inter section, a yield-controlled channelized r ight-turnlane, and a signalized channelized r ight-turn lane. A ser ies of microscopic simulation runs (using

VISSIM) were conducted to evaluate the traff ic operational perfor mance for both vehicles and pedestr ians. The simulation studies addressed r ight-turn vehicle delay, delay due to pedestr iancrossings, and the im pact of inter section character istics.

viii



Channelized r ight-turn lanes with yield control were shown to reduce r ight-turn delay tovehicles by 25 to 75 percent in com par ison to inter section approaches with conventional r ight-turn lanes. High pedestr ian volumes increased r ight-turn delay by approximately 60 percent on ayield-controlled channelized r ight-turn lane. The addition of an acceleration lane at thedownstream end of a channelized r ight-turn lane can reduce the r ight-turn delay by65 to 85 percent, depending on the conf licting traff ic volume on the cross street. Increasing theradius of a channelized r ight-turn roadway can reduce r ight-turn delay by approximately10 to 20 percent for each 8-k m/h (5-mi/h) increase in turning s peed.

Seven (7) year s of motor-vehicle and pedestr ian crash and volume data were obtained for 103 four-leg signalized inter sections in Toronto, Ontar io, Canada. An overall com par ison was

perfor med of the safety perfor mance of inter section approaches with channelized r ight-turn lanesto inter section approaches with other r ight-turn treatments. S pecif ically, a cross-sectionalanalysis was conducted to com pare the crash exper ience among:

Inter section approaches with channelized r ight-turn lanesInter section approaches with conventional r ight-turn lanesInter section approaches with no r ight-turn treatments (shared through/r ight-turn lanes)

Inter section approaches with channelized r ight-turn lanes were shown to have similar motor-

vehicle safety perfor mance as approaches with conventional r ight-turn lanes or sharedthrough/r ight-turn lanes where the r ight-turning vehicle depar ts from the through traff ic stream(the upstream end of the channelized r ight-turn lane). R esults of the safety analysis suggest that

the three r ight-turn treatments may differ in motor-vehicle safety perfor mance as the r ight-turning vehicle merges with the cross street (the downstream end of the channelized r ight-turnlane), but this was not conclusively established. Inter section approaches with channelized r ight-turn lanes were shown to have similar pedestr ian safety perfor mance as approaches with shared

through/r ight-turn lanes. Inter section approaches with conventional r ight-turn lanes hadsubstantially more pedestr ian crashes (approximately 70 to 80 percent more) than approacheswith channelized r ight-turn lanes or shared/through r ight-turn lanes.

A recommendation of the research is for highway agencies to develop a consistent practicefor crosswalk location. Consistency in crosswalk location is im por tant to pedestr ians with visionim pair ment and would make it easier for O&M s pecialists to teach pedestr ians with vision

im pair ment how to better traver se a channelized r ight-turn lane. Since current practice shows aclear preference for crosswalk locations near the center of a channelized r ight-turn lane, designguidance should recommend placing crosswalk s near the center of the channelized r ight-turnlane for channelized r ight-turn lanes with yield control or no control at the entry to the cross

street. Where STOP sign control or traff ic signal control is provided at the entry to the crossstreet, the crosswalk should be placed immediately downstream of the stop bar, where possi ble.Where the channelized r ight-turn roadway inter sects with the cross street at near ly a r ight angle,

the stop bar and crosswalk can be placed at the downstream end of the channelized r ight-turnroadway.

ix



R aised islands should be considered because they serve as a refuge area so that pedestr iansmay cross the street in two stages. R aised islands with “cut-through” paths also provide better guidance for pedestr ians with vision im pair ment than painted islands.

Channelized r ight-turn lanes with acceleration lanes appear to be diff icult for pedestr ianswith vision im pair ment to cross. Therefore, the use of acceleration lanes at the downstream endof a channelized r ight-turn lane should generally be reserved for locations where no pedestr iansor very few pedestr ians are present. Typically, these would be locations without sidewalk s or

pedestr ian crossings; at such locations, the reduction in vehicle delay resulting from addition of

an acceleration lane becomes very desirable.

Channelized r ight-turn lanes are most appropr iate for im proving traff ic operations atinter sections where pedestr ian volumes crossing the channelized r ight-turn lane are expected to

be low to moderate (e.g., up to approximately 1,000 pedestr ians per day). This recommendation

is based on the 85th percentile pedestr ian volume—1,000 pedestr ians per day— in the Torontointer section database that was used for the safety evaluation. While channelized r ight-turn lanesmay be suitable for higher pedestr ian volumes, the research cannot predict the safety

perfor mance of channelized r ight-turn lanes with pedestr ian volumes beyond the range evaluatedin the research.

The research results indicate that channelized r ight-turn lanes have a def inite role inim proving operations and safety at inter sections. However, to achieve these benef its they shouldhave consistent design and traff ic control and should be used at appropr iate locations. The

research provides design guidance for channelized r ight-turn lanes that addresses geometr icelements such as crosswalk location, s pecial crosswalk signing and mark ing, island type, radiusof turning roadway, angle of inter section with cross street, acceleration and deceleration lanes,and traff ic control.

x



Chapter 1.Intr oduction

This chapter summar izes the background for the research, the research ob jectives and scope,and the organization of this repor t.

Backgr ound1.1

Channelized r ight-turn lanes are turning roadways at inter sections that provide for free-f lowor near ly free-f low r ight-turn movements. Channelization can be provided in a var iety of for msincluding painted pavement areas and curbed islands. Figure 1 illustrates an inter section with

channelized r ight-turn lanes. While the f igure shows channelized r ight-turn lanes in all quadrantsof the inter section, channelized r ight-turn lanes may be appropr iate in some quadrants, but not inother s, depending on inter section geometry and traff ic demands.

Figure 1. Typical Intersection with Channelized R ight-Turn Lanes

The pr imary reasons for providing a channelized r ight-turn lane are (1):

To increase vehicular capacity at inter sections

To reduce delay to dr iver s by allowing them to turn at higher s peeds

To reduce unnecessary stops

Channelized

right-turn roadway

Channelizing

island

Channelized

roadway width

Radius

Channelized

right-turn roadway

Channelizing

island

Channelized

roadway width

Radius

1



To clear ly def ine the appropr iate path for r ight-turn maneuver s at skewed inter sectionsor at inter sections with high r ight-turn volumes

To im prove safety by separating the points at which crossing conf licts and r ight-turnmerge conf licts occur

To per mit the use of large curb return radii to accommodate turning vehicles, includinglarge truck s, without unnecessar ily increasing the inter section pavement area and the

pedestr ian crossing distance

Many trans por tation agencies use channelized r ight-turn lanes to im prove operations atinter sections, although their im pact on safety for motor ists, pedestr ians, and bicyclists has not

been clear. A key concern with channelized r ight-turn lanes has been the extent of conf licts

between vehicles and pedestr ians that occur at the point where pedestr ians cross the r ight-turnroadway. Conf licts with pedestr ians may occur at r ight-turn roadways because the dr iver’sattention may be focused on the cross-street traff ic; the placement of pedestr ian crosswalk s or

pedestr ian signals on channelized turning roadways may violate dr iver expectancy.

For most pedestr ians, crossing the r ight-turn roadway is a relatively easy task because suchroadways are not very wide and because traff ic is approaching from a single direction. However,

pedestr ians with vision im pair ment may have diff iculty detecting approaching traff ic because(a) r ight-turning vehicles are traveling a curved rather than a straight path; (b) there is not asystematic stopping and star ting of traff ic, as there would be at a conventional signal- or stop-controlled inter section; and (c) the traff ic sounds from the ma jor streets may mask the sound of

traff ic on the r ight-turn roadway. The Amer icans with Disabilities Act (ADA) requires that all pedestr ian facilities, including sidewalk s and crosswalk s, be accessi ble to pedestr ians withdisabilities. The U.S. Access Board has published draf t r ights-of-way guidelines requir ing

pedestr ian-activated signals at multi-lane crossings of channelized r ight-turn lanes with pedestr ian signal indications (2). S pecif ically, it states:

“R305.7 Channelized Turn Lanes at Intersections. Where pedestr ian crosswalk s are provided at multi-lane r ight or lef t channelized turn lanes at inter sections with pedestr ian signal indications, a pedestr ian-activated signal com plying with R 306 shall be provided.

Advisory R305.7 Channelized Turn Lanes at Intersections. Accessi ble pedestr iansignal devices installed at s plitter and ‘pork chop islands must be carefully located andseparated so that signal s pillover does not give conf licting infor mation about whichcrossing has the WALK indication dis played.

Additional guidance on signal types is provided in Advisory R 305.6.2.”

Des pite their potential challenges for pedestr ians, channelized r ight-turn lanes also provideadvantages for pedestr ians. The provision of a channelized r ight-turn lane, while mak ing it

necessary for pedestr ians to cross two roadways, of ten reduces the pedestr ian crossing distanceof the ma jor and cross streets. Fur ther more, the channelizing island serves as a refuge area for

pedestr ians, par ticular ly when bounded by raised curbs, and im proves safety by allowing pedestr ians to cross the street in two stages.

2



The pr imary traff ic operational reasons for providing channelized r ight-turn lanes are toincrease vehicular capacity at an inter section and to reduce delay to dr iver s by allowing them toturn at higher s peeds and reduce unnecessary stops. Channelized r ight-turn lanes appear to

provide a net reduction in motor vehicle delay at inter sections where they are installed, althoughno existing data and no established methodology have been available to directly com pare theoperational perfor mance of urban inter sections with and without channelized r ight-turn lanes.Fur ther more, no data are available on the operational effects of installing pedestr ian-activatedsignals along r ight-turn roadways.

The safety effects of channelized r ight-turn lanes on motor vehicles, pedestr ians, and bicyclists have been largely unknown. It is generally accepted that channelized r ight-turn lanesim prove safety for motor vehicles at inter sections where they are used, but there have been onlylimited quantitative data to demonstrate this. No studies have been found concerning pedestr ian

safety at channelized r ight-turn lanes that have used crash data to document the pedestr ian safetyim plications of channelized r ight-turn lanes. There also appear s to be an inherent r isk to

bicyclists at channelized r ight-turn lanes because motor vehicles enter ing the channelized r ight-turn roadway must weave across the path of bicycles traveling straight through the inter section,

but no studies based on crash history are available to suppor t this presum ption. However, thissame type of conf lict between through bicyclists and r ight-turn vehicles is present at

conventional inter sections as well.

It is evident that there have been many unanswered questions about channelized r ight-turn

lanes. The research conducted for this repor t sought to answer many of these questions and to provide a basis for decisions, based on sound research results rather than anecdotal evidence,about where channelized r ight-turn lanes should and should not be used.

Research Ob jectives and Scope1.2

The ob jective of the research was to develop design guidance for channelized r ight-turnlanes, based on balancing the needs of passenger car s, truck s, buses, pedestr ians (including

pedestr ians with disabilities), and bicycles. For the purposes of this pro ject, a channelized r ight-turn lane is character ized by separation from the through and lef t-turn lanes on the approach byan island and separate traff ic control from the pr imary inter section. The channelized r ight-turnlane may have a deceleration lane enter ing it and it may have a merge or an auxiliary lane at the

exiting end. The results of the research include general design guidance on channelized r ight-turn

lanes for highway engineer s and s pecif ic recommended language for consideration in futureupdates to key AASHTO and FHWA documents.

Or ganization of This Repor t1.3

This repor t presents the results of the research on channelized r ight-turn lanes. Theremainder of this repor t is organized as follows. Chapter 2 presents an overview of the state of knowledge and practice with channelized r ight-turn lanes, based on a review of literature and a

3



survey of highway agency exper ience. Chapter 3 discusses pedestr ian behavior at channelizedr ight-turn lanes based on observational f ield studies and interviews with or ientation and mobilitys pecialists. Chapter 4 presents the results of a traff ic operational analysis of channelized r ight-turn lanes. Chapter 5 presents the results of a safety analysis of channelized r ight-turn lanes.Chapter 6 presents the interpretation of the research results. Chapter 7 presents the conclusionsand recommendations of the research. Chapter 8 presents a list of references cited in this repor t.Appendix A includes a design guide for channelized r ight-turn lanes. Appendix B presentssuggested text concerning channelized r ight-turn lanes for consideration by AASHTO for

potential inclusion in future editions of the Gr een Book .

4



Chapter 2.State of Knowledge and Practice with ChannelizedRight-Tur n Lanes

This chapter presents the results of a review of com pleted and ongoing research related tothe design and safety of channelized r ight-turn lanes, and summar izes the results of a survey of highway agency exper ience with channelized r ight-turn lanes.

Literature Review2.1

A literature review was conducted to update the infor mation presented in a recent synthesison channelized r ight turns (3) that was prepared in NCHRP Pro ject 3-72, Lane W id t h s ,

C hanneliz ed Ri ght T ur n s , and Ri ght -T ur n Decel er ation Lane s in U r ban and Subur ban Ar ea s.This section summar izes current knowledge concerning the traff ic safety perfor mance of channelized r ight turns and presents infor mation obtained from the literature review of both

NCHRP Pro ject 3-72 and the current research. Safety for motor vehicles, pedestr ians, and bicycles are addressed separately.

2.1.1 Motor Vehicle Saf ety at Channelized Right-Tur n Lanes

It is generally accepted that channelized r ight turns im prove safety for motor vehicles atinter sections where they are used, but there has been only limited quantitative data to

demonstrate this. The research f indings that are available are summar ized below.

Dixon et al. (4) analyzed the crash history at 17 signalized inter sections with var ious r ight-turn treatments in Cobb County, Georgia, to identify the effects of those r ight-turn treatments onr ight-turn crashes. The inter sections were located on both ma jor and minor ar ter ials. A total of 70 r ight-turn movements were identif ied for evaluation, and 57 of these movements had one of the following f ive common r ight-turn treatments:

Shared r ight-turn lane, no island, merge, and no additional controlExclusive r ight-turn lane, no island, merge, and no additional control

Exclusive r ight-turn lane, raised island, acceleration lane, and no additional control

Exclusive r ight-turn lane, raised island, merge, and yield controlShared r ight-turn lane, raised island, large turning radius, merge, and yield control

Table 1 summar izes the num ber of r ight-turn crashes for each treatment. The analysis was based str ictly upon crash frequencies over a 2-year per iod and did not include exposure datarelated to traff ic volumes.

5



Dixon et al. (4) noted the following general f indings, indicating that they mer it futureresearch:

The use of a traff ic island appear s to reduce the num ber of r ight-angle crashes

The addition of an exclusive r ight-turn lane appear s to corres pond to elevated sideswi pecrashes

The addition of an exclusive lane on the cross street for r ight-turning vehicles (i.e., anacceleration lane) does not appear to reduce the num ber of rear-end crashes when no

additional control is im plemented

Table 1. Comparison of Crash History for Common R ight-Turn Treatments (4 )

Treatment

Shared right-tur nlane,

mer ge,no island,

no additionalcontr ol

Exclusive right-tur n lane,

mer ge,no island,

no additionalcontr ol


acceleration lane,raised island,no additional

contr ol


mer ge,raised island,yield contr ol

Shared right-tur nlane,

lar ge tur ningradius,mer ge,

raised island,yield contr ol

Number of sitesevaluated

29 8 5 7 8

Number of right tur ncr ashes for 2-year period

18 13 14 22 10

Aver age number of right tur n cr ashes per site per year

0.31 0.81 1.40 1.57 0.63

Crash type Percent of Right-Tur n Crashes Observed

Right angle 50 31 22 23 0

Rear-end 28 23 64 59 90

Sideswipe 17 31 7 18 0

Other 5 15 7 0 10

The Texas Depar tment of Trans por tation s ponsored a study (5) similar to that perfor med byDixon (4). In the Texas study, crash data for six inter sections were reviewed. A total of 30approaches with one of the following treatments were evaluated:

Right-turn lane with lane linesRight-turn lane with raised islandShared through-r ight lane

Shared through r ight lane with raised island

Table 2 summar izes the num ber of r ight-turn crashes for each treatment over a 3-year

per iod. The values do not include consideration of r ight-turn volume; however, they can providean appreciation of the var iability in the num ber of r ight-turn crashes among the differenttreatments.

6



Table 2. Annual R ight-Turn Crashes by Type of R ight-Turn Treatment (5 )

Intersection Total

Treatment

RTLw/island

RTLw/striping

Sharedlane

Sharedlane

w/island

Number of appr oaches 30 14 6 8 2

Number of cr ashes 16 9 2 1 4

TxDOT pr o ject— Aver age number of right -tur n cr ashes per site per year 0.18 0.21 0.11 0.04 0.67

Dixon pr o ject—aver age number of right-tur n cr ashes per site per year N/A 1.57 0.81 0.31 0.63

RTL = right-tur n lane; SL = shar ed thr ough-right lane.

The ma jor ity of the crashes (10 out of 16) were rear-end crashes. Of the 10 rear-end crashes,

5 crashes occurred in a r ight-turn lane with a raised island. As shown in Table 2, the shared-laneconf iguration exper ienced the lowest average num ber of r ight-turn crashes per site per year. Ther ight-turn lane separated by a raised island showed the highest num ber of crashes in Dixonsstudy and the second highest num ber of crashes in the TxDOT study.

Staplin et al. (6 ) conducted an accident analysis to examine the problems that older dr iver shave in inter section areas. Approximately 700 accident records were reviewed dur ing the

analysis. In general, older dr iver s had diff iculty yielding the r ight-of-way and mak ing lef t turnsat inter sections, but the accident analysis did not reveal channelized r ight turns as a safety issuefor older dr iver s.

Tarawneh and McCoy (7 ) conducted f ield investigations to study the effects of thegeometr ics of r ight-turn lanes on the turning perfor mance of dr iver s. Right-turn perfor mance of

100 sub jects was evaluated at four signalized inter sections of different r ight-turn lanechannelization and skew. Three of the four inter sections had a channelized r ight-turn lane. Theinvestigation found that dr iver s turn r ight at s peeds 5 to 8 k m/h (4 to 5 m ph) higher oninter section approaches with channelized r ight-turn lanes than they do on approaches with

unchannelized r ight-turn lanes. In addition, it was observed that dr iver s are less likely to come toa com plete stop before turning onto the cross street on approaches with channelized r ight turns.However, no explicit safety f indings were inferred from this result.

McCoy et al. (8) conducted f ield studies on rural two-lane highways and found a higher incidence of merging conf licts from vehicles enter ing the cross street from a channelized r ightturn without an acceleration lane than those with an acceleration lane.

McCoy et al. (8) developed guidelines for channelized r ight-turn lanes at unsignalizedinter sections on rural two-lane highways. In developing the guidelines, McCoy et al. evaluated

the safety effects of channelized r ight-turn lanes based on accident data, f ield studies, andcom puter simulation of truck dynamics. An analysis of the accident history at 89 ruralinter sections with and without channelized r ight-turn lanes over a 5-year per iod found no effectof channelized r ight-turn lanes on the frequency, sever ity, or types of accidents that occur on

approaches to unsignalized inter sections. Thus, it was concluded from the accident analysis thatchannelized r ight-turn lanes do not provide the road user with any safety benef its or dis benef its.Field studies which investigated the tendency of dr iver s to travel faster than the design s peed of

7



the channelized roadway and to over steer at some point through the turn found that some dr iver sexceeded the design s peeds of the channelized r ight-turn roadways, but they seldom exceededthe margin of safety for their vehicles. Vehicle-path data showed that near ly all of the vehiclesobserved were well-positioned in the lane at the center of the curve, indicating that they wereable to follow the curvature of the channelized roadway with ease. R esults from the truck simulation study suppor t the use of the AASHTO cr iter ia for curves on open highways, insteadof the AASHTO cr iter ia for minimum-radii inter section curves for the design of inter sectioncurves on rural highways with substantial truck volumes. R esults of the truck simulation alsoshowed that the margin of safety between the maximum safe s peed for truck s and the design

s peed is narrow.

In a 2006 study by Abdel-Aty and Nawathe (9), ar tif icial neural network s were used toanalyze the safety of signalized inter sections. Geometry, traff ic, and crash data were obtained for 1,562 signalized inter sections from six counties in Flor ida. Neural network trees were used to

deter mine the relationshi p between inter section geometry/conf iguration and the frequency of s pecif ic types of crashes. The study found that:

The presence of channelized r ight-turn lanes on the ma jor road was shown to have nosignif icant effect on total crashes, but was linked to an increase in turning and sideswi pecrashes.

On the minor road, the presence of channelized r ight-turn lanes was associated with adecrease in total crashes and an increase in rear-end crashes.

2.1.2 Pedestrian Saf ety at Channelized Right-Tur n Lanes

No studies have been found that have used crash data to document the pedestr ian safetyim plications of channelized r ight-turn lanes. Pr ior crash studies have focused on the vehicle-

pedestr ian collisions involving turning vehicles, but the geometr ics of the inter section were notavailable to document the type of turning lane present.

A f ive-state analysis of more than 5,000 vehicle-pedestr ian collisions found that 38 percentof all such crashes occurred at inter sections (10). Fur ther examination of the inter sectionaccidents found that 30 percent of these crashes involved a turning vehicle. There was no fur ther

breakdown to deter mine if the vehicle was turning r ight or lef t. From a query of a Nor th Carolinadatabase that includes detailed crash types developed with the Pede strian and Bicycl e Cr a sh

Anal y sis T ool , one can deter mine the breakdown of turning vehicles (11, 12). The Nor th Carolina

system includes 5 year s of data (over 11,000 pedestr ian-motor vehicle collisions). Inter sectioncrashes account for 26 percent of those collisions. Lef t-turning vehicles accounted for 10 percentof the collisions at inter sections, while r ight-turning vehicles accounted for 6 percent. Statistics

gathered by the Oregon DOT (13) show that 19 percent of vehicle-pedestr ian crashes occurr ingat inter sections arose from dr iver s mak ing r ight turns.

The geometry of channelized r ight-turn lanes per mits turns at higher s peeds than in anunchannelized situation. Higher motor-vehicle s peeds represent higher r isk to pedestr ianscrossing the roadway. R esearch by Zegeer et al. (14) has established that, in the event that there

8



is a collision, vehicle s peed directly affects the likelihood that a pedestr ian will be fatally in jured.Should a pedestr ian be hit by a vehicle traveling at 32 k m/h (20 m ph), the chance of being k illedis 5 percent. For a 48-k m/h (30-m ph) vehicle, that likelihood of a fatality r ises to 45 percent,while for vehicles traveling at 64 k m/h (40 m ph), the likelihood of a fatal in jury is 85 percent.Motor ists traveling at higher s peeds have less time to see pedestr ians and require more time toslow, stop, or change direction to avoid str ik ing them.

Geruschat and Hassan (15) evaluated dr iver s behavior in yielding the r ight-of-way tosighted and blind pedestr ians who stood at three different positions relative to the curb at the

crosswalk at entry and exit lanes at two different roundabouts. The researcher s found that adr ivers willingness to yield to a pedestr ian was related to the s peed of the vehicle. S pecif ically,at low s peeds [less than 24 k m/h (15 mi/h)], dr iver s yielded approximately 75 percent of thetime, whereas at higher s peeds [greater than 32 k m/h (20 mi/h)], they typically yielded less than50 percent of the time.

Safety for Pedestrians with Special Needs

Other crash-based analyses have focused on pedestr ians with s pecial needs, par ticular ly theelder ly. One such study that looked s pecif ically at the types of crashes occurr ing at inter sections

showed older pedestr ians to be overrepresented in collisions with both lef t- and r ight-turncollisions (16 ). Collisions involving lef t-turning and r ight-turning vehicles accounted for 17 percent and 13 percent, res pectively, of all inter section accidents involving pedestr ians.

Pedestr ians who were age 75 or older and were involved in a vehicle-pedestr ian collision werestruck by a lef t-turning vehicle in 24 percent of cases and by a r ight-turning vehicle 14 percent of cases. Those aged 65 to 74 were struck by a lef t-turning vehicle in 18 percent of cases and by ar ight-turning vehicle in 19 percent of cases.

Schroeder et al. (17 ) conducted a paired com par ison study of blind and sighted pedestr iansat three channelized r ight-turn locations. At each location, pedestr ians were observed as theyassessed gaps in traff ic and identif ied oppor tunities to cross the channelized r ight-turn roadway.A key issue in the study was whether the geometry of channelized r ight-turn lanes and/or thelack of signal control at channelized r ight-turn roadways negatively affect the delay and safetyfor blind pedestr ians. Blind pedestr ians rely on auditory cues when crossing the roadway because

they cannot use visual infor mation to identify approaching vehicles and assess appropr iate gapsin traff ic. However, at channelized r ight-turn roadways, blind pedestr ians have to judge traff icmoving in a circular path while dealing with a signif icant amount of background noise fromtraff ic at the inter section. The ob jectives of the study were to:

Identify diff iculties exper ienced by blind pedestr ians in crossing a channelized r ight-turnlane safely.

Test the effect of crosswalk location, geometry of the channelized r ight-turn lane, andtraff ic volumes on pedestr ian delay and r isk perfor mance measures.

Study par tici pants consisted of nine blind and nine sighted pedestr ians, who were tested in pair s (one blind and one sighted). The pedestr ians were asked to stand by the roadside as though

9



they were going to cross and indicate when they felt it was safe to cross. No actual crossingswere perfor med in the study. The blind pedestr ians signaled when they would cross by pushing a

button and holding it down as long as they felt it was safe to cross. The par tici pants released the button when they no longer felt it was safe to cross. The sighted pedestr ians gave similar indications by raising and lower ing a white pocket folder they held in their hand.

Key f indings in the research were:

On average, blind pedestr ians require more time to make a crossing decision. The greater

time consists of longer lead and lag times.

Blind pedestr ians make a greater percentage of r isky “go” decisions and a greater percentage of unnecessary “no go” decisions than sighted pedestr ians.

The percentage of r isky “go” decisions tends to increase with higher conf licting vehiclevolumes in the turn lane, while the percentage of unnecessary “no go” decisions tends to

decrease at high volumes.

Background traff ic volumes had similar effects for sighted pedestr ians. This may suggest

that the “noise” of background traff ic, which com plicated the decision process for blind pedestr ians, may be par tially off set by the uncer tainty of sighted pedestr ians in knowingwhether vehicles upstream of the turn lane were enter ing the channelized r ight-turn laneor continuing straight through the inter section. The effect of background volume on

blind pedestr ian perfor mance was not signif icant.

The com par ison of crossing perfor mance at two crosswalk locations (in the center vs.downstream of the channelized r ight-turn lane) showed no signif icant effect for either group of pedestr ians. The center crossing location appeared to result in slightly better

perfor mance in ter ms of reducing r isky and unnecessary decisions, but additional dataare necessary to show signif icance.

Pedestrian Safety at Roundabouts

The geometry of roundabouts is similar to channelized r ight-turn lanes in some res pects, androundabouts and channelized r ight-turn lanes pose similar challenges to pedestr ians with visionim pair ment attem pting to cross the road. In both situations, the pedestr ian with visionim pair ment has to judge traff ic moving on a circular path and the vehicle movement is notcontrolled by signals. In both cases, traff ic may be free-f lowing or may yield to other vehicles(e.g., at the downstream end of the channelized r ight-turn roadway or roundabout approach). The

pedestr ian with vision im pair ment is faced with the task of identifying gaps in traff ic,distinguishing between sounds from traff ic in the turn lane and background traff ic noise at theinter section, and judging yield behavior of dr iver s. R esearcher s from Western Michigan

Univer sity and Vanderbilt Univer sity recently conducted a ser ies of studies of blind pedestr ianscrossing at roundabouts (18, 19).

One study evaluated the crossing behavior of blind and sighted pedestr ians at three

roundabouts in the Baltimore, Maryland, metropolitan area (18). The research evaluated howwell each pedestr ian group judged whether gaps in traff ic were long enough to safely cross to the

10



s plitter island. Using six blind and six sighted par tici pants that each had exper ience maneuver ingurban environments and roundabouts, the study found that the blind pedestr ians were near ly 2.5times less likely to make correct judgments than sighted pedestr ians, took longer to detectcrossable gaps, and were more likely to miss crossable gaps altogether. R esults were found to bemost signif icant at high-volume inter sections. A follow-up study was conducted at three differentroundabouts to deter mine if the presence of visi ble indicator s of a blind pedestr ian (e.g., guidedog or long cane) would produce better yielding patterns from dr iver s. At each location,

pedestr ians holding long canes elicited more yielding maneuver s. Also, dr iver s were more likelyto yield at the entry of a roundabout than at the exit.

Another study evaluated pedestr ian crossings at a double-lane urban roundabout in Nashville, Tennessee (19). Par tici pants included six blind and six sighted pedestr ians. Each pedestr ian par tici pated in two sessions; each session consisted of six tr ials. On three tr ials, pedestr ians crossed the roadway independently, followed by a cer tif ied or ientation and mobility

(O&M) s pecialist. On the other three tr ials, they used a hand signal to indicate when they wouldhave star ted crossing, but did not actually cross the roadway. The research found that:

Blind pedestr ians waited three times longer to cross than sighted pedestr ians.

Near ly 6 percent of the crossing attem pts by blind pedestr ians were judged dangerousenough to require intervention.

In high-volume conditions, blind pedestr ians waited almost twice as long to make acrossing than in low-volume conditions; traff ic volumes had much less of an effect on

the wait time of sighted pedestr ians.Sighted pedestr ians made the crossing 41 percent of the time without waiting for anyapproaching vehicles to pass the location. In contrast, only 15 percent of crossings by

blind pedestr ians were made before a vehicle passed; with most par tici pants waiting for a per iod of time that allowed f ive or six vehicles to pass before crossing.

Dr iver s yielded frequently on the entry lanes but not the exit lanes.

Sighted pedestr ians were much more inclined to accept a dr ivers yield than a blind pedestr ian.

In post-session interviews, blind par tici pants indicated that they would likely take measuresto avoid crossing the roundabout test site on a daily basis due to the signif icant num ber of self-initiated and exper imenter-initiated interventions from potential vehicle-pedestr ian conf licts.While these studies focused on roundabouts, the free-f low conditions of vehicles enter ing a

roundabout are very similar to those of a channelized r ight-turn lane and, therefore, the f indingsmay potentially be applicable to channelized r ight-turn lanes.

In a recent study, Inman et al. (20) tested the ability of an audi ble surface treatment (similar to a rum ble str i p) to im prove pedestr ian safety at crosswalk s upstream of roundabouts. Using acontrolled environment, the research team documented the reactions of seven pedestr ians withvision im pair ment to vehicles approaching an inter section with and without the proposed

treatment. In two lanes of traff ic, test vehicles approached an inter section where pedestr ians werewaiting for audi ble cues to signal their approach and yield. In order to simulate a real-wor ld

11



situation of pedestr ians encounter ing an innovative treatment, study par tici pants were not provided any s pecif ic infor mation on the treatment pr ior to the study. The treatment layoutconsisted of four sound str i ps upstream of the crosswalk that would be activated by all vehiclesand then two more str i ps at the crosswalk that would only be traver sed by non-yielding vehicles.Based on hand gestures from the pedestr ians, the research team documented the num ber of timeseach pedestr ian correctly detected a yielded vehicle, the num ber of false detections, and thenum ber of missed vehicles.

R esults of the study found that, af ter the surface treatment was added, detection accuracy

im proved by an average of approximately 58 percent. There were also a substantial num ber of false detections, which would have resulted in pedestr ians crossing in an environment that they

believed was protected by stopped vehicles but that would have, in fact, been vulnerable toapproaching vehicles. The author s suggest, however, that the treatment may be more effective insingle-lane approach situations where one lane of traff ic would not create the false secur ity for

both lanes.

Pedestrian Research in NCHRP Pr o ject 3-78A

The effects of channelized r ight-turn lanes on pedestr ians with vision im pair ment were

studied as par t of NCHRP Pro ject 3-78A, Cr o ssing Sol ution s at Roundabouts and C hanneliz ed T ur n Lane s f or Pede strian s wit h V ision Disabilitie s (21). The ob jectives of the research were toassess pedestr ians with vision im pair ment crossing a channelized r ight-turn lane and evaluate

treatment installations to assist those pedestr ians in crossing.

The study site for NCHRP Pro ject 3-78A was located in Char lotte, Nor th Carolina, andconsisted of channelized r ight-turn lanes on all four inter section approaches. Data were collected

at two of the four turn lanes, as indicated in Figure 2. Each turn lane has a deceleration lane onthe approach, but does not have an acceleration lane for vehicles to merge onto the cross street.A pedestr ian crosswalk is located at the center of each island, perpendicular to the sidewalk.

Safety im provement treatments were installed at the channelized r ight-turn lanes dur ingSummer 2008:

Sound str i ps —devices intended to provide audi ble cues to pedestr ians with visionim pair ment about approaching vehicles

Sound str i ps in com bination with pedestr ian-actuated beacons

The sound str i ps, s paced approximately 9-m (30-f t) apar t, were installed on each of thedeceleration lanes. The sound str i ps were a modif ication of a treatment used in recent FHWA

research at roundabouts (20), with minor modif ications in treatment placement. The pedestr ian-actuated beacon was installed, in con junction with an audi ble indication that the beacon wasf lashing, at one of the crosswalk s. Lane delineator s were also installed as par t of the treatment to

prevent last-minute entry by vehicles into the deceleration lane.

12



Pedestr ians with vision im pair ment par tici pated in both the pre-treatment and post-treatment phases of the study. Dur ing the study, par tici pants were asked to navigate across the channelizedr ight-turn lane to the best of their ability using the existing cues provided at the crosswalk. In theevent that a par tici pant moved into an unsafe situation, research team mem ber s were available tointervene.

Dur ing the pre-treatment data collection, observations of the research team included:

The channelized r ight-turn lane study site exper ienced high traff ic volumes.

Pedestr ian delays were relatively high.

Gap acceptance and yield utilization were relatively low.

Exper imenter interventions were in the range of 8 to 10 percent, with several other “self-interventions” by the par tici pants.

Figure 2. Channelized R ight-Turn Lane Study Location for NCHRP Project 3-78A (21 )

Post-treatment data collection was conducted to deter mine if sound str i ps, either alone or incom bination with pedestr ian-actuated beacons, have any effect on yield behavior of dr iver s or onr isky crossings (i.e., interventions) by pedestr ians with vision im pair ment. Analysis of post-treatment data suggested the following:

13



Yield rates of dr iver s increased slightly (from 15.2 to 22.0 percent) where sound str i pswere installed in com bination with f lashing beacons. There was no change where soundstr i ps were installed as the only treatment.

Installation of treatments reduced, but did not eliminate, interventions.

Several par tici pants noted that they could hear better to make crossing decisions fromthe curb than from the island. They stated that the sound of traff ic behind them, whenwaiting on the island, made the crossing decision more diff icult.

2.1.3 Bicycle Saf ety at Channelized Right-Tur n Lanes

There is an inherent r isk to bicyclists at channelized r ight turns because motor vehiclesenter ing the channelized r ight-turn roadway must weave across the path of bicycles travelingstraight through the inter section. However, no studies based on crash history are available to

suppor t this presum ption. Fur ther more, a similar conf lict between through bicyclists and r ight-turning vehicles is present at all inter sections, except at inter sections where r ight turns are

prohi bited or three-leg inter sections where there is no leg to the r ight on a given approach. Thereare also no studies that provided data on the r isk of collisions between motor vehicles and

bicycles on the channelized r ight-turn roadway itself or at the point at which the channelizedr ight-turn roadway joins the cross street. The following discussion presents basic statistics on

bicycle safety, followed by available infor mation on bicycle-related safety issues at r ight-turnlanes.

In 2002, 662 bicyclists were k illed and an additional 48,000 were in jured in traff ic accidentsin the U.S. (22). Thus, bicyclists accounted for about 2 percent of all traff ic fatalities. Bicyclists

accounted for approximately 12 percent of all nonmotor ized traff ic fatalities, while pedestr iansaccounted for 86 percent; the remaining 2 percent were skateboard r ider s, roller skater s, etc.

Oregon repor ts that most bicycle crashes (65 to 85 percent) do not involve collisions withmotor vehicles but rather involve falls or collisions with stationary ob jects, other bicyclists, and

pedestr ians. Of the bicycle/motor vehicle crashes, 45 percent occurred at inter sections (13). Inanother evaluation of bicycle/motor vehicle crashes, Tan repor ted that approximately 5 percent

of bicycle/motor vehicle crashes occurred when a motor ist made a r ight turn (23), but noinfor mation was provided on whether the res pective crashes occurred at inter sections withchannelized r ight turns. Tan also repor ted that approximately 4 percent of bicycle/motor vehiclecrashes occurred at an inter section controlled by a signal at which the motor ist struck the

bicyclist while mak ing a r ight-turn-on-red.

Clark and Tracy (24) repor ted that 13 percent of all bicycle/motor vehicle crashes resulted

when motor ists were mak ing a r ight-turn movement, and a ma jor ity of these crashes involved astraight-through bicyclist being struck by a r ight-turning motor vehicle. Clark and Tracyindicated that many bicyclists f ind changing lanes diff icult or choose to ignore signage and

pavement mark ings.

Much of the advice for highway designer s in dealing with inter sections and r ight-turn lanesis applicable only to locations where bicycle lanes already exist (or are planned in the future). As

14



indicated in Chapter 3, the MUTCD (25) and AASHTO bicycle guide (25) recommend break ing bicycle lane mark ings ahead of the inter section and then mark ing the bicycle lane again at theinter section itself, to the lef t of the r ight-turn lane. This positions bicyclists traveling straightthrough the inter section away from any conf lict with r ight-turning vehicles and allows a mergearea for r ight-turning vehicles to get into r ight-turn lane.

Two recently com pleted studies for the FHWA have included observational studies of bicyclists and motor ists as they maneuvered through a var iety of r ight-turn lane conf igurations(27, 28). One of the studies was a before-af ter effor t in which the conf lict zone, def ined as the

place where the paths of bicyclists and motor ists crossed most of ten, was treated with blue pavement mark ings at 10 inter sections in Por tland, Oregon (27 ). Figure 3 illustrates the use of blue pavement mark ings at the entrance to and exit from a channelized r ight-turn lane.

Figure 3. Blue Pavement Marking Treatment at Channelized R ight-turn Lane (27 )

15



Conf igurations addressed in the Oregon blue bike lane program included exit ram ps, r ight-turn lanes, and entrance ram ps. The mark ings were also supplemented with unique signsshowing the blue mark ings and yield signs for motor ists (see Figure 4). Both video observationsand survey feedback were collected as par t of the study, with approximately 850 bicyclists and190 motor ists in the before per iod and 1,020 bicyclists and 300 motor ists in the af ter per iod. Themost im por tant results were as follows:

There was a signif icant increase in motor ists yielding to bicyclists af ter the treatmentwas installed, from 71 percent in the before per iod to 87 percent in the af ter per iod.

Signif icantly more bicyclists followed the path marked for bicyclists af ter the bluemark ings were in place, 85 percent in the before per iod com pared to 93 percent in theaf ter per iod.

There was a decrease in head-turning and scanning on the par t of bicyclists af ter thetreatment was installed, from 43 percent in the before per iod to 26 percent in the af ter

per iod, which was a concern. The author s were not sure of the reason for this result.

While conf licts between the two modes were rare, the conf lict rate decreased from

0.95 conf licts per 100 enter ing bicyclists in the before per iod to 0.59 conf licts per 100enter ing bicyclists in the af ter per iod.

The survey data showed that 70 percent of the motor ists noticed the blue mark ings, and

59 percent noticed the accom panying sign. When asked about safety, 49 percent of themotor ists thought it would increase safety, 20 percent thought it would be the same,

12 percent thought it would be less safe, and the remaining motor ists were not sure.The bicyclists surveyed thought the treatment would increase safety (76 percent). Only1 percent thought it would decrease safety.

Overall, it was found that the treatment resulted in a safer r iding environment and a heightenedawareness on the par t of both bicyclists and motor ists. The City of Por tland continues to use thistreatment at 6 of the 10 locations today.

The second study examined the behavior s of bicyclists and motor ists at a “com bined” bicycle lane/r ight-turn lane used in Eugene, Oregon (28). The results were com pared to

observations made at a more traditional r ight-turn lane. The com bined lane created a 1.5-m (5-f t) bike pocket within a 3.6-m (12-f t) r ight-turn lane, leaving 2.1 m (7 f t) for r ight-turning vehicles(see Figure 5). The traditional lane location used for com par ison was a 3.6-m (12-f t) r ight-turnlane and a 1.5-m (5-f t) bike pocket (see Figure 6). Approximately 600 bicyclists were videotaped

at each location as they approached and continued straight through the inter section. Thedifferences in the two types of r ight-turn lanes can be summar ized as follows:

Bicyclists and motor ists tended to queue up behind one another more of ten in thecom bined lane facility (43 percent of the time) than in the standard lane facility(1 percent of the time).

At both locations, bicyclists were most of ten able to position themselves in the bike pocket (94 percent of the time in the com bined lane and 86 percent of the time in thestandard lane). At the com bined lane inter section, bicyclists tended to use the ad jacent

16



through lane more of ten (2 percent of the time) com pared to vir tually no such positioning at the standard lane. This was pr imar ily due to the occasional bus that neededto turn r ight at the com bined lane inter section, which then forced the approaching

bicyclists to use the through lane.

At both locations, the yielding behavior of each mode was captured. At the com binedlane location, the motor ist yielded to the bicyclist in 93 percent of the cases where thetwo par ties would have collided had someone not slowed or stopped. At the standardlane location, motor ist yield 48 percent of the time. This low percentage of yielding by

motor ists at the standard lane is believed to be an ar tifact of bicyclists having to shif t tothe lef t on the approach to the inter section in order to move from the bicycle lanead jacent to the curb to the bike pocket at the inter section.

No conf licts requir ing either mode to suddenly stop or change direction were observed ateither location.

Figure 4. Signs Used in Oregon Blue Bik e Lane Program (27 )

17



Figure 5. Combined Bicycle Lane/R ight-Turn Lane (27 )

Figure 6. Traditional Bik e Lane/R ight-Turn Lane (28 )

In addition to the observational data, a br ief survey of a sam ple of bicyclists wasadministered at both locations. When asked to com pare the two locations, 18 percent said thecom bined lane was safer, 27 percent said it was less safe, and 55 percent said there was nodifference. Overall, the observational and survey data showed the com bined bicycle lane/r ight-

18



turn lane to be an effective treatment that could be benef icial at locations where r ight-of-wayconstraints exist.

There has also been a perception study conducted for FHWA in which par tici pants wereasked to view a num ber of r ight-turn lane conf igurations and provide a rating of howcomfor table they would be interacting with r ight-turning traff ic in an effor t to continue straightthrough the inter section (29). The conf igurations rated included:

A standard r ight/through lane in which the bicyclist could travel straight on the approach

and continue through the inter section.

An auxiliary r ight-turn only lane that was added at the inter section, which allowed the bicyclist to travel straight on the approach and forced the motor ist to cross the path of

the bicyclist.

A travel lane that became a r ight-turn lane at the inter section, forcing bicyclists to shif t

lef t across the path of motor ists in order to continue straight through the inter section.

A gradual increase in pavement width on the inter section approach that became a r ight-

turn lane at the inter section, also forcing bicyclists to shif t lef t across the path of motor ists in order to continue straight through the inter section.

A regression model was developed using the perception ratings as the dependent var iableand several geometr ic and operational var iables as independent measures. The most signif icant

predictor s of the a bicyclists comfor t level were whether there was a bike lane present on the

approach and whether the bicyclist had to shif t to the lef t across the motor ist path in order tocontinue through the inter section. The presence of a bike lane increased the comfor t level, whilethe requirement to shif t across the motor ists path decreased the comfor t level. This resultconf ir med some of the observational data collected in the com bined bicycle/r ight-turn lane study

previously descr i bed.

An exam ple of a treatment for bicycle lanes at inter sections that is considered inappropr iatesuggests channeling bicyclists onto a sidewalk or bike path and having them behave as

pedestr ians (24). Crash records suggest this approach is ser iously f lawed, es pecially since it canencourage wrong-way r iding.

On streets with bicycle lanes, the current recommended designs ensure straight-though bicyclists are positioned to the lef t of exclusive r ight-turn lanes. On streets without bicycle lanes, bicyclists and motor ists must perfor m the same maneuver s as if separate lanes were marked.

They must do so, however, without the guidance offered by the bicycle lane mark ings andwithout the same amount of s pace available to share the road at the inter section. In bothinstances, there are several im por tant design features to remem ber (24):

As the length of the r ight-turn lane increases, so does the exposure of the bicyclist to

traff ic dr iving on either side of them. In addition, the s peed of vehicles in the r ight-turnlane may be greater. Thus, exclusive r ight-turn lanes should be kept as shor t as possi ble.

19



As both bicyclists and motor ists pass through inter sections, they are concentrating ontheir own position on the road and on traff ic within the inter section. No dr ivewaysshould be positioned near the inter section to cause additional conf licts.

Highway Agency Experience2.2

A for mal highway agency survey on channelized r ight-turn lanes was conducted as par t of

NCHRP Pro ject 3-72, the results of which were ref lected in the Synt he sis on C hanneliz ed Ri ght T ur n s (3) that was also developed as par t of that research. Therefore, no for mal survey wasneeded as par t of the current research. However, because the survey was conducted in 2003,there was a clear need to update the results. To assure that any new research, design practice, or agency exper iences were identif ied, the research team asked the AASHTO Standing Committee

on Highway Traff ic Safety (SCOHTS) to send an email through their list-serve to all 50 statehighway agencies ask ing for new or updated infor mation on channelized r ight-turn lanes. Thefollowing seven states res ponded to the request for infor mation: Idaho, New York, R hode Island,

Tennessee, Texas, Virginia, and West Virginia. This section presents key f indings from bothresearch pro jects —NCHRP Pro ject 3-72 and the current research—related to highway agency

practice with channelized r ight-turn lanes and focuses on geometr ic design issues, traff ic control,and pedestr ian considerations at channelized r ight-turn lanes.

2.2.1 Geometric Design Issues at Channelized Right-Tur n Lanes

The highway agency survey (3) conducted in NCHRP Pro ject 3-72 indicated that 87 percentof state and local highway agencies use channelized r ight turns; all seven highway agencies that

provided additional infor mation in the current research indicated that they utilize channelizedr ight-turn lanes. Even casual observation suggests that channelized r ight-turn lanes are arelatively common geometr ic feature at inter sections on urban and suburban ar ter ials.

Deceler ation Lanes

Right-turn deceleration lanes serve one or more of the following functions (30):

A means for safe deceleration outside the high-s peed through lanes for r ight-turning

traff ic.

A storage area for r ight-turning vehicles to assist in optimization of traff ic signal phasing.

A means for separating r ight-turning vehicles from other traff ic at stop-controlled

inter section approaches.

The addition of a deceleration lane at the approach to a channelized r ight-turn lane provides anoppor tunity for motor ists to safely slow down pr ior to reaching the crosswalk area at the turning

roadway. In res ponse to the survey (3) conducted in NCHRP Pro ject 3-72, 89 percent of the state

20



highway agencies and 70 percent of the local agencies that use channelized r ight-turn lanesindicated that they have used deceleration lanes in advance of those channelized r ight-turn lanesfor at least some locations.

Acceler ation Lanes

Acceleration lanes provide an oppor tunity for vehicles to com plete the r ight-turn maneuver unim peded and then accelerate parallel to the cross-street traff ic pr ior to merging. In res ponse to

the survey (3) conducted in NCHRP Pro ject 3-72, 77 percent of the state highway agencies and43 percent of the local agencies that use channelized r ight turns indicated that they have usedacceleration lanes downstream of those channelized r ight turns for at least some locations. In therecent infor mal survey, one agency res ponded that acceleration lanes are generally used when theangle between turning roadway and inter secting roadway is less than 60 degrees.

Cr osswalks

Crosswalk s that are parallel to and constitute an extension of the sidewalk may provide the best alignment infor mation for pedestr ians with vision im pair ment because they allow the

pedestr ian to continue along a straight travel path. Also, the sounds of traff ic moving parallel tothe pedestr ians line of travel can of ten be used for establishing and maintaining alignment for crossing. Although not s pecif ically illustrated in Figure 7, an additional factor to be considered in

locating the crosswalk is the ability to construct appropr iate curb ram ps to provide access for wheelchair user s. While the alignment of parallel crosswalk s may be preferred by pedestr ianswith vision im pair ment, it is diff icult to build wheelchair ram ps in these locations withoutshif ting grades at the gutter that cause problems in traver sing the ram ps for wheelchair user s. For

that reason, the perpendicular crosswalk alignments would be preferred for wheelchair user s,along with maintaining the crosswalk at a level grade, with a cross slope of less than 2 percent.

Table 3 summar izes highway agency practices concerning the placement of crosswalk s for channelized r ight-turn roadways (3). The most common highway agency practice is to place thecrosswalk near the center of the r ight-turn roadway (i.e., not immediately ad jacent to either of theinter secting streets). The table indicates that 77 percent of state highway agencies and 67 percent

of local highway agencies that use channelized r ight turns have placed pedestr ian crosswalk s inthis center position. The recent infor mal survey produced similar results, where six out of sevenhighway agencies indicated that they place the crosswalk near the middle of the r ight-turnroadway.

21



Figure 7. Alternative Crosswalk Locations

Table 3. Locations Where Highway Agencies Place Pedestrian

Crosswalks at Channelized R ight-Turn Roadways (3 )

Location

Number (percentage) of agencies

State agencies Local agencies Total

At the upstr eam end 8 (22.9) 7 (23.3) 15 (23.1)

In the middle 27 (77.1) 20 (66.7) 47 (72.3)

At the downstr eam end 9 (25.7) 12 (40.0) 21 (32.3)

NOTES: 1. Columns total to mor e than 100 percent because of multiple r esponses.2. Percentages ar e of those h ighway agencies that have used channelized right-tur n

r oadways.

The Wisconsin DOT (WisDOT) has been im plementing a newer design for channelizedr ight-turn lanes that incorporates a different shape for the island—an isosceles tr iangle rather than an equilateral tr iangle conf iguration—as illustrated in Figure 8 (31).

Option 1: Marked crosswalk

Location: Upstream end

Direction: Parallel to sidewalk



Direction: Perpendicular to sidewalk


Location: Center


Option 6: No marked crosswalk

Location: N / A

Direction: N / A


Location: Downstream end












Location: Center


Option 6: No marked crosswalk

Location: N / A

Direction: N / A







22



Figure 8. Isosceles-Triangle Island Shape Used by WisDOT

Angle of Intersection With Cr oss Str eet

The survey results indicate that the alignment of a channelized r ight-turn lane and the angle between the channelized r ight-turn roadway and the cross street can be designed in two different

ways:

A f lat-angle entry to the cross street like the channelized r ight-turn lanes illustrated in

Figures 1 through 3 and Figure 7A near ly-r ight-angle entry to the cross street like that par tially visi ble on the r ight side of the photograph in Figure 8

These two designs are com pared in Figure 9.

The two designs shown in Figure 9 differ in the shape of the island that creates the

channelized r ight-turn lane. The f lat-angle entry design has an island that is typically shaped likean equilateral tr iangle (of ten with one curved side), while the near ly-r ight-angle design istypically shaped like an isosceles tr iangle. The f lat-angle entry design is appropr iate for use inchannelized r ight-turn lanes with either yield control or no control for vehicles at the entry to thecross street. The near ly-r ight-angle entry design can be used with STOP sign control or traff icsignal control for vehicles at the entry to the cross street; yield control can also be used with this

design where the angle of entry and sight distance along the cross street are appropr iate.

2.2.2 Traff ic Contr ol at Channelized Right-Tur n Lanes

The survey results (3) from NCHRP Pro ject 3-72 indicate that only 14 percent of statehighway agencies and 17 percent of local highway agencies have for mal policies concerningtraff ic control devices for channelized r ight-turn roadways. Other agencies rely on the Manual on

23



Figure 9. Typical Channelized R ight-Turn Lanes with Differing Entry

Angles to the Cross Street [Adapted From (14 )]

Unif or m Tr a ffic C ontr ol Device s (MUTCD) (24) for guidance concerning the proper applicationof yield signs, stop signs, and signals; such guidance deals with the general application of thesedevices, but is not s pecif ic to channelized r ight turns. The infor mal survey conducted in thecurrent research had similar results; f ive out of seven agencies did not have a for mal policy

concerning traff ic control at channelized r ight-turn lanes. One agency stated that yield signs arethe most commonly used traff ic control. Another agency has a policy in which if the end of the

S k e t c h e s b y M i c h a e l K i m e l b e r g

24



channelized r ight-turn lane is more than 9 m (30 f t) from the pr imary inter section control, itoperates under its own signal, stop, or yield sign.

Highway agencies were asked to identify innovative traff ic control devices they haveim plemented at channelized r ight-turn roadways. Table 4 summar izes use of innovative traff iccontrol devices by highway agencies. Between 35 and 50 percent of highway agencies have usedhigh-visi bility crosswalk mark ings and f lorescent yellow-green signs, but fewer than 10 percentof agencies have tr ied other innovative devices. In the recent infor mal survey, one agencyrepor ted using the f luorescent yellow-green signs; another has im plemented f lashing warning

signs for motor ists to warn of a pedestr ian crossing. One agency uses a sign that states“TUR NING TR AFFIC MUST YIELD TO PEDESTR IANS;” another agency uses im pr int tohel p create high-visi bility crosswalk s.

Table 4. Innovative Traff ic Control Devices at Channelized R ight-Turn Roadways (3 )

Traff ic contr ol deviceNumber (percentage) of agencies


High-visibility cr osswalk markings 13 (37.1) 13 (43.3) 26 (41.3)

Fluor escent yellow-gr een signs 16 (45.7) 15 (50.0) 31 (49.2)

Real-time war ning devices 2 (5.7) 3 (10.0) 5 (7.9)

Other dynamic message signs 2 (5.7) 2 (6.7) 4 (6.3)

Other 1 (2.9) 0 (0.0) 1 (1.6)

NOTES: 1. Columns total to mor e than 100 percent because of multiple r esponses.2. Percentages ar e of those h ighway agencies that have used channelized right-tur n r oadways.

In the survey (3) conducted in NCHRP Pro ject 3-72, highway agencies were asked whether they install pedestr ian-actuated signals at channelized r ight-turn roadways on urban and

suburban ar ter ials. The res ponses are summar ized below:

Of the highway agencies that use channelized r ight-turn roadways, only 6 percent of state highway agencies and 17 percent of local highway agencies install pedestr ian-actuated signals at all channelized r ight-turn roadways.

The ma jor ity of highway agencies using channelized r ight-turn roadways (83 percent of state highway agencies and 60 percent of local highway agencies) install pedestr ian-actuated signals at sel ect ed locations only.

Of the highway agencies that use channelized r ight-turn roadways, approximately11 percent of state highway agencies and 23 percent of local highway agencies do not

use pedestr ian-actuated signals.

In the recent infor mal survey, four out of the seven agencies indicated that they haveinstalled pedestr ian-actuated signals at select locations.

Highway agencies were also asked whether they have developed or used any strategiess pecif ically intended to assist pedestr ians with vision im pair ment in crossing channelized r ight-

turn roadways without pedestr ian signals. Of the highway agencies that use channelized r ight-turn roadways, 23 percent of state highway agencies and 10 percent of local highway agencies

25



have either developed or used such strategies. The general types of strategies used by theres ponding agencies to assist pedestr ians with vision im pair ment in crossing channelized r ight-turn roadways are summar ized in Table 5. One highway agency that does not currently useaccessi ble pedestr ian signals at channelized r ight-turn lanes is consider ing their use. Onehighway agency repor tedly installed audi ble signals at one inter section, but was requested by anorganization representing pedestr ians with vision im pair ment to deactivate the sound. In therecent infor mal survey, one agency repor ted using different curb conf igurations to facilitatecrossing the r ight-turn lane. One agency uses a shor t crossing distance, perpendicular to theroadway, with detectable warning devices and way-f inding cues within the turning roadway

island. In the recent infor mal survey, four out of seven agencies indicated that they do notim plement any strategies s pecif ically aimed at pedestr ians with vision im pair ment. One agencyuses signs stating “TUR NING TR AFFIC MUST YIELD TO PEDESTR IANS.” One agency usesdifferent curb conf igurations to facilitate crossing the r ight-turn lane. One agency uses a shor tcrossing distance, perpendicular to turning roadway, with detectable warning devices and way-

f inding cues within the turning roadway island.

Table 5. General Strategies Used by Highway Agencies to

Assist Pedestrians with Vision Impairment (3 )

General strategyState

agenciesLocal

agencies Total

Cur b r amps with tr uncated domes 5 1 6

Textur ed cur b r amps 4 1 5

Audible signals 1 0 1

While textured curb ram ps are listed by highway agencies as a strategy used to assist pedestr ians with vision im pair ment, research has indicated that var ious textures are notdetectable or usable by pedestr ians with vision im pair ment. The only texture that is recognized to

provide adequate detectability, both underfoot and under cane, is the truncated dome detectablewarning surface required by the Amer icans with Disabilities Act (ADA). However, neither strategy assists pedestr ians with deter mining the appropr iate time to cross channelized r ight-turn

roadways.

2.2.3 Pedestrian Considerations at Channelized Right-Tur n Lanes

In the survey conducted in NCHRP Pro ject 3-72, highway agencies were asked about pedestr ian considerations in deter mining the radius or width of channelized r ight-turn roadways.

Of the highway agencies that use channelized r ight-turn roadways, 23 percent of state highwayagencies and 40 percent of local highway agencies indicated that they consider pedestr ian issuesin deter mining the radius and/or width of a channelized r ight-turn roadway. On the other hand, inthe infor mal survey of highway agencies, six of the seven agencies stated that they use design

vehicles rather than pedestr ian considerations to deter mine the radius of a channelized r ight-turnlane. One agency stated that pedestr ians have been used in some cases as a factor in deter miningthe radius. Table 6 presents a list of s pecif ic pedestr ian-related issues considered by highway

agencies in deter mining the radius or width of a channelized r ight-turn roadway. One highway

26



agency indicated that they do not use channelized r ight turns, and another is trying to minimizetheir use, because of pedestr ian concerns.

Table 6. Pedestrian Issues Considered in Determining the Radius

or Width of Channelized R ight-Turn Roadway (3 )

Pedestrian issueNumber of agencies


Pedestrian cr ossing distance/time minimized 3 1 4

Vehicle speeds minimized 1 3 4

Pr ovision of pedestrian r efuge location 2 2

Impr oved sight distance of opposing tr aff ic 1 1

Pedestrian volumes 1 1

Gener al consider ation of pedestrians 4 1 5

Highway agencies were also asked if they have encountered any safety problems related to

pedestr ians crossing at channelized r ight-turn roadways on urban and suburban ar ter ials. Of thehighway agencies that use channelized r ight-turn roadways, approximately 23 percent of statehighway agencies and 17 percent of local highway agencies have encountered pedestr ian-related

safety problems at channelized r ight-turn roadways. Highway agencies repor ted the followingsafety concerns related to pedestr ians crossing at channelized r ight turns:

General concern about pedestr ian safety at channelized r ight-turn roadways (5 agencies)

Higher vehicle s peeds put pedestr ians at r isk (3 agencies)Pedestr ians with vision im pair ment may expect approaching traff ic to stop (1 agency)Truck-trailer off track ing onto sidewalk jeopardizes pedestr ian safety (1 agency)Dr iver s may not yield to pedestr ians (1 agency)Larger radii may make pedestr ians less visi ble to dr iver s (1 agency)There is some confusion regarding the most appropr iate crossing location (1 agency)There is greater exposure to pedestr ians (1 agency)

Small islands and snow on islands are not conducive to pedestr ian use (1 agency)

Only one highway agency repor ted a safety problem at a s pecif ic location. One location with

an unsignalized r ight-turn roadway and no pedestr ian signal has a sight distance problem thatwill probably be addressed by providing a signal. In the infor mal survey, f ive out of the sevenagencies repor ted no safety issues. Two agencies repor ted occasional pedestr ian safety concerns.

27



Chapter 3.Pedestrian Behavior at Channelized Right-Tur n Lanes

This chapter presents the results from studies of pedestr ian behavior at channelized r ight-turn lanes conducted as par t of the research, including observational f ield studies of pedestr iancrossing behavior and interviews with or ientation and mobility (O&M) s pecialists who teach

pedestr ians with vision im pair ment to traver se inter sections with channelized r ight-turn lanes.

Observational Field Studies3.1

One of the concerns with channelized r ight-turn lanes has been the interaction between

vehicles and pedestr ians at the point where pedestr ians cross the r ight-turn roadway. Figure 10illustrates a typical channelized r ight-turn lane with vehicles yielding to a pedestr ian. While theyield behavior of dr iver s at channelized r ight-turn lanes has not been well documented,

channelized r ight-turn lanes present a scenar io where a dr iver’s attention could be focused on thecross-street traff ic or the placement of pedestr ian crosswalk s or signals may violate dr iver expectancy.

Figure 10. Channelized R ight-Turn Lane with Vehicle Yielding to Pedestrian

Another concern with channelized r ight-turn lanes has been the potential challenge for pedestr ians with vision im pair ment. For pedestr ians with no vision im pair ment, watching for agap in traff ic and then physically negotiating the crossing of a channelized r ight-turn roadway

may be a relatively easy task because such roadways are not very wide and because traff ic isapproaching from a single direction. However, pedestr ians with vision im pair ment may havediff iculty detecting approaching traff ic because (a) r ight-turning vehicles are traveling a curved

rather than a straight path; (b) there is not a systematic stopping and star ting of traff ic, as therewould be at a conventional signal- or stop-controlled inter section; and (c) the traff ic sounds fromthe ma jor streets may mask the sound of traff ic on the r ight-turn roadway.

To address the concerns of interactions between pedestr ians and vehicles at channelizedr ight-turn lanes, observational f ield studies were conducted at inter section approaches with

28



channelized r ight-turn lanes to document pedestr ian and vehicle behavior s and interactions. Toaddress the concern of pedestr ians with vision im pair ment at channelized r ight-turn lanes,interviews with or ientation and mobility s pecialists were conducted. Each is descr i bed below.

Observational f ield studies were conducted at 35 inter section approaches with channelizedr ight-turn lanes. The pr imary ob jectives of the observational studies were to:

Observe the interaction of motor vehicles, pedestr ians, and bicycles at var iouschannelized r ight-turn lane conf igurations.

Document the frequency of given types of interactions consider ing geometr ic design,traff ic control, and traff ic volume data.

Observe how well pedestr ians “obey” crosswalk s and traff ic control at differentcrosswalk locations (i.e., upstream end, center, and downstream end) and directions

(i.e., perpendicular or parallel to the sidewalk) within a channelized r ight-turn roadway.

Observe how well motor ists yield to pedestr ians at different crosswalk locations

(i.e., upstream end, center, and downstream end) and directions (i.e., perpendicular or parallel to the sidewalk) within a channelized r ight-turn roadway.

3.1.1 Site Selection

Over 100 candidate study sites were identif ied and reviewed for geometr ic, pedestr ian, and

traff ic character istics in Ar izona, California, Colorado, Flor ida, Idaho, Illinois, Maryland,Missour i, Oregon, and Washington. A key pr ior ity was the selection of inter sections where

pedestr ian activity was estimated to be moderate to high. Figure 11 illustrates the geographiclocations (city and state) of the 35 sites that were ultimately selected for conductingobservational f ield studies.

Inter sections were selected to best represent an accurate cross section of channelized r ight-

turn lanes based on differ ing geometr ic design and traff ic control character istics that make upsuch facilities, such as island type and size, presence of deceleration or acceleration lane,crosswalk location, and traff ic control on the channelized r ight-turn roadway.

3.1.2 Data Collection Methodology

At each site, video cameras were used to observe motor vehicle, pedestr ian, and bicycle

interactions, document pedestr ian crossing behavior and motor vehicle yielding behavior, andrecord motor vehicle and pedestr ian volumes at each channelized r ight-turn lane site. Videocameras were mounted unobtrusively on utility poles or traff ic signal poles at the study

inter section approaches. Three video cameras were typically used at each site to cover the entirechannelized r ight-turn lane. The data collection methodology and general camera setup wassimilar at all sites, although the s pecif ic camera locations were deter mined on a site-by-site basis

to optimize the viewing angle for each camera. Figures 12 and 13 illustrate the video camerasetup for two typical sites, one in Baltimore, Maryland, and one in San Francisco, California,res pectively.

29



Figure 11. Observational Field Study Locations

Figure 12. Data Collection Field Setup for an Intersection in Baltimore, Maryland

30



Figure 13. Data Collection Field Setup for an Intersection in San Francisco, California

A minimum of 8 hour s of video was recorded at each site. The following infor mation was

documented for each channelized r ight-turn lane:

Num ber of pedestr ians crossing the channelized r ight-turn lane Num ber of vehicles traver sing the channelized r ight-turn laneGeometr ic design and traff ic control infor mationArea type (residential/commercial/industr ial)Posted s peed limit on ma jor road and minor road

Inter section traff ic control (signalized/unsignalized)Traff ic control on channelized r ight-turn lane (STOP/yield/signal/none)Type (raised or painted) and size (small/medium/large) of channelizing island

Width of channelized r ight-turn roadwayType of pedestr ian signal head (countdown, hand/ per son, other)Presence of crosswalk (yes/no)Location of crosswalk (upstream/middle/downstream)

Presence of bicycle lane on ma jor road (yes/no)Dr iveways present within 76 m (250 f t) of channelized r ight-turn lane

31



3.1.3 Site Characteristics

The site character istics of the observational f ield study sites are summar ized as follows:

Number of sites

Ar ea type Residential 4

Commercial 28

Industrial 3

Intersection tr aff ic contr ol Signalized 35

Unsignalized 0

Island type Painted 0

Raised 35

Island size Small 9

Medium 20

Lar ge 6

Deceler ation lane Yes 21

No 14

Acceler ation lane Yes 7No 28

Cr osswalk location Upstr eam 6

Middle 25

Downstr eam 4

Tr aff ic contr ol for CRT Yield 19

Stop 2

Signal 5

None (f r ee) 9

3.1.4 Vehicle and Pedestrian Counts

Table 7 presents the vehicle and pedestr ian counts that were obtained dur ing the evening peak per iod (5:00 p.m. to 6:00 p.m.) of each observational study. Right-turn volumes rangedfrom 11 veh/h to 719 veh/h, but about half of the study sites had volumes between 100 and300 veh/h. Most sites exper ienced peak-per iod volumes of 200 ped/h or fewer, although two siteshad near ly 350 ped/h dur ing the peak per iod (Boulder and Chicago), and one site in SanFrancisco had over 1,800 ped/h in the peak per iod.

32



Table 7. Vehicle and Pedestrian Counts During Evening Peak Period

(5:00 p.m. to 6:00 p.m.)

Location Intersection

Traff iccontr ol

(direction)

Right-tur nvolume(veh /h)

Conf lictingthr oughvolume(veh /h)

Pedestrianvolume(ped /h)

Boise, ID

Br oadway Str eet andMyr tle Str eet

Unsignalized(EB)

719 690 10

Br oadway Str eet and

Warm Springs Boulevar d

Unsignalized

(EB)

299 875 11

Eagle, IDSH 44 and SH 55

Unsignalized(NB)

549 384 0

Eagle Road andFairview Avenue

Unsignalized(EB)

176 1,070 5

Boulder , CO

Ar apahoe Avenue and28th Str eet

Unsignalized(WB)

299 1,487 104

Color ado Avenue and28th Str eet

Unsignalized(SB)

162 564 73

Unsignalized(EB)

533 2,116 57

Ar apahoe Avenue and30th Str eet

Unsignalized(EB)

222 737 44

Unsignalized(SB)

300 1,301 196

Baseline Avenue andBr oadway Str eet

Unsignalized(WB)

491 694 120

Br oadway Str eet andUniversity Avenue

Unsignalized(NB)

335 152 335

Chicago, IL

S. Cicer o Avenue and55th Str eet

Signalized(SB)

58 771 4

Gr and Avenue andLake Shor e Drive

Signalized(WB)

53 594 429

Fr anklin Boulevar d andSacr amento Boulevar d

Unsignalized(EB)

52 0 0

Unsignalized(SB)

112 62 1

College Park,MD

MD 193 (UniversityBoulevar d) and

New Hampshir e Avenue(MD 650)

Unsignalized(NB)

230 968 163

Unsignalized(EB)

291 678 39

Towson, MD Dulaney Valley Road and

Fairmont Avenue

Unsignalized(WB)

638 1,206 50

Unsignalized

(SB) 219 629 29

Tualatin, OR

Highway 99W and124th Avenue

Signalized(WB)

391 930 0

Bridgepor t Road and72nd Avenue

Unsignalized(WB)

614 95 7

Por tland, OR 23r d Avenue and

Bur nside RoadUnsignalized

(SB) 175 1,111 82

33



Table 7. Vehicle and Pedestrian Counts During Evening Peak Period

(5:00 p.m. to 6:00 p.m.) (Continued)

Location Intersection

Traff iccontr ol

(direction)

Right-tur nvolume(veh /h)

Conf lictingthr oughvolume(veh /h)

Pedestrianvolume(ped /h)

Sacr amento,

C A

J Str eet andCarlson Drive

Unsignalized(EB)

221 617 118

Unsignalized(NB)

631 1,498 7

Fr anklin Boulevar d and

Fr uitridge Boulevar d

Unsignalized

(WB) 174 896 12Fr uitridge Road andFr eepor t Boulevar d

Unsignalized(WB)

556 157 3

Br uceville Road andConsumnes River

Signalized(NB)

759 1,111 27

San Fr ancisco,C A

Market Str eet andDuboce Avenue

Unsignalized(WB)

22 743 166

Ocean Avenue andPhelan Avenue

Unsignalized(SB)

130 992 116

Unsignalized(WB)

604 170 60

Sloat Boulevar d andJuniper o Serr a Boulevar d

Signalized(EB)

463 1,089 48

Stockton Str eet andPost Str eet

Unsignalized(EB)

154 827 1,810

Seattle, W A

Aur or a Avenue andDenny Way

Unsignalized(WB)

189 1,008 74

Lake WashingtonBoulevar d and

E. Madison Str eetUnsignalized

(WB)217 451 22

Yesler Way and3r d Avenue

Unsignalized(SB)

11 307 197

3.1.5 Observational Study Results

Pedestr ian crossing behavior and yield behavior of motor ists were documented from thevideo for each pedestr ian crossing at each of the sites. From the 35 sites, over 2,800 pedestr iancrossing observations were recorded, pr imar ily dur ing the evening peak per iod. In addition togeneral infor mation that was documented (e.g., pedestr ian arr ival and crossing time, num ber of vehicles present dur ing crossing, whether pedestr ian was traveling alone or in a group), the

pedestr ian and motor ist behavior obtained from the video was organized into three categor ies:

Avoidance maneuver sPedestr ian crossing behavior sYield behavior of motor ists

34



Avoidance Maneuvers

The types of avoidance maneuver s that were documented include:

Pedestr ian hesitates, stops, or retreats in crosswalk due to presence of vehicleMotor ist swerves or abruptly stops to avoid pedestr ian

R esults of the observational f ield studies suggest that avoidance maneuver s are relatively rare.Approximately 17 percent of the over 2,800 pedestr ian crossings occurred with a motor vehicle

present in the channelized r ight-turn lane; and out of those, only 12 pedestr ian and 10 vehicleavoidance maneuver s were observed. Thus, avoidance maneuver s were made in less than one

percent of all observations and less than f ive percent of observations with approaching vehicles.

Pedestrian Cr ossing Behaviors

The types of pedestr ian crossing behavior s that were documented include:

Pedestr ian uses the crosswalk for the entire crossingPedestr ian follows (“obeys”) the pedestr ian signal (if present)

Pedestr ian has diff iculty f inding a gapPedestr ian crosses aggressively (e.g., runs) or crosses between stopped vehicles

Overall, the results showed that 72 percent of all pedestr ians crossed the entire channelized r ight-turn roadway within the crosswalk. With res pect to following the pedestr ian signal, the resultsshowed that 72 percent of all pedestr ians crossed dur ing the pedestr ian crossing phase and28 percent crossed against the signal. While 21 percent of pedestr ians crossed the channelized

r ight-turn lane aggressively (i.e., running or dar ting across the roadway, potentially in a smallgap of traff ic), only 4 percent of pedestr ians walked between stopped vehicles to cross thechannelized r ight-turn lane.

Yield Behavior of Motorists

The types of motor ist yield behavior that were documented include:

Motor ist yields to pedestr ian in crosswalk Motor ist or pedestr ian perfor ms an avoidance maneuver

Motor ist yields to pedestr ian waiting at curbMotor ist stops but block s the crosswalk

Over 96 percent of the observations in which a vehicle was present and a pedestr ian was in thecrosswalk, the vehicle yielded to the pedestr ian or was unaffected by the pedestr ian crossing. Insome cases, the vehicle slowed or stopped; in other cases, the pedestr ian crossed withoutsubstantially im pacting the vehicle s peed. This f inding also conf ir med the resulted thatavoidance maneuver s occurred at approximately 1 percent of all observations.

35



When pedestr ians were waiting at the curb, a yield rate of approximately 41 percent wasobserved. Some of the data points included both a pedestr ian in the crosswalk as well a

pedestr ian at the curb at the same time, which could have confounded the observed yield rate.

Stopped or queued vehicles tended to keep the crosswalk open to pedestr ians. Only7 percent of vehicles stopped in a location that blocked the crosswalk. At those locations where astop bar was present, 62 percent of observed vehicles stopped at the proper location relative tothe stop bar.

Effect of Special Cr osswalk Signing and Marking

The observational f ield study sites in Boulder, Colorado, included s pecial crosswalk signingand mark ing. Figure 14 illustrates a raised crosswalk on a channelized r ight-turn roadway withs pecial pavement mark ings. An infor mal com par ison of pedestr ian and motor ist behavior wasmade between the sites in Boulder, which have s pecial crosswalk treatments, and theobservational f ield study sites at all other locations. Table 8 presents the results of the infor malcom par ison. While no statistical analyses can be perfor med due to the small sam ple sizes, thecom par ison suggests that the additional signage and pedestr ian crosswalk treatments mayim prove the motor ist yield behavior and pedestr ian use of the crosswalk.

Figure 14. Raised Crosswalk at Channelized R ight-Turn Lane in Boulder, Colorado

36



Table 8. Comparison of Pedestrian and Motorist Behavior

Between Boulder, Colorado, Sites and Other Sites

Pedestrian or motorist behavior

Percent of total observations

Boulder, CO, sites All other sites

Pedestrians use entir e cr osswalk 86 72Vehicles yield to pedestrians waiting at cur b 47 40Vehicles yield to pedestrians in cr osswalk 99 96Vehicles block cr osswalk when stopped 4 7

Interviews with Orientation and Mobility Specialists3.2

Interviews with ten or ientation and mobility (O&M) s pecialists were conducted to learn of their exper ience in teaching pedestr ians with vision im pair ment to traver se inter sections with

channelized r ight-turn lanes and to deter mine their recommendations on how channelized r ight-turn lanes might be better designed for all pedestr ians with disabilities. Dur ing the interviews,O&M s pecialists were asked to descr i be strategies that pedestr ians with vision im pair ment use atinter sections with channelized r ight-turn lanes to:

Locate the crossing star ting pointAlign to crossDecide when to star t crossing

Aler t dr iver s of their intention to crossOther techniques or comments

These interviews were conducted by Ms. Janet Bar low, the O&M s pecialist on the researchteam. Interviews were conducted via email, phone, or in per son depending on the preference andlocation of the res pondent.

3.2.1 Design and Cr osswalk Location

It was apparent from the interviews that there is considerable var iation in the design of channelized r ight-turn lanes. Most of the O&M s pecialists stated that the channelized r ight-turnlanes in their area were yield-controlled, although two of the O&M s pecialists were aware of

locations in their area where the channelized r ight-turn roadway was signalized. One per soncommented that many channelized r ight-turn lanes in her area had STOP signs, rather than yieldsigns. Another noted that the designs in his area were mostly low s peed without deceleration or acceleration lanes.

Var iation in locations for crosswalk s was noted by all, with frustration expressed about not being able to easily provide clients with a clear under standing of channelized r ight-turn lane

layout because of the lack of consistency in geometry, crosswalk location, and control. Oneinstructor noted that “the more rounded the corner, the more diff icult it seems to be to get a goodread of the traff ic. These lanes appear to make the inter section wider and add to confusion tosound cues.” Most commented that channelized r ight-turn lanes with acceleration lanes were

37



vir tually uncrossable for their clients/students, because of the s peed of vehicles and lack of yielding by dr iver s.

The interview par tici pants opinions about crosswalk location (upstream, center, or downstream) var ied. Four of the instructor s indicated a preference for the crosswalk at thedownstream end because the vehicles on the downstream street serve as “blocker car s,”

preventing vehicles from exiting the channelized r ight-turn lane and providing a cue about a timeto begin crossing. Additional arguments for the downstream location were:

“ I like down str eam bett er becau se t hat’s wher e t he two str eets act uall y mer ge (and t hu swher e t he mer g ing car mu st yiel d if tr a ffic on t he ent ering str eet is moving) and t hat’s

wher e I t each my clients t o cr o ss in mo st ca se s. T he one s t hat ar e up str eam I t hink ar edone t hat way becau se fiel d o f view t o see oncoming cars is bett er . T hat o ft en con f u se s a

l ot o f my clients who have some vision but poor dept h per ception.”

“ I pr e f er t he down str eam end becau se t his is wher e d rivers t end t o sl ow down and l ook ar ound be f or e continuing. O f course, t his make s mor e o f a d iff er ence if t he pede strian is

going fr om t he d river’s l e ft t o his ri ght ( isl and t o cur b), becau se t he d river is alr eady

l ook ing l e ft f or oncoming tr a ffic.”

Four other instructor s suppor ted the center location because they felt their clients were morevisi ble at that location, as well as for reasons associated with alignment for crossing:

Cr o sswal k s at t he cent er o f t he t ur n “seem be st , a s t his may put t he pede strian and

d river in bett er l ine o f si ght , and also it’s ea sier f or a blind pede strian t o hear , a s t hecars about t o t ur n ar e l e ss ma sked by tr a ffic in t he int ersection. Also, it may be ea sier f or

some t o get t heir ali gnment a s t he cur b is str ai ght er her e.”

Two instructor s descr i bed pros and cons of each location, with one stating that the upstream position, es pecially with a deceleration lane, usually provides the best visi bility position for the pedestr ian to see/hear and be seen and allowed another crossing timing (of crossing in front of a

stopped car without dr iver contact). The other instructor stated that the center crosswalk locationmade it easiest to f ind the island, but more diff icult to use traff ic cues. Both said that the center

position allows the pedestr ian with vision im pair ment to use the curb, ram p, and detectablewarning edges for alignment to cross, while the alignment to cross is more challenging atupstream and downstream crosswalk locations.

O&M s pecialists were asked about the techniques and strategies they teach at channelized

r ight-turn lanes. Some commented that painted islands were of little use to their clients/students because they do not provide the tactile cues provided by a raised island, par ticular ly a raisedisland with a “cut-through” pedestr ian path. In fact, the im por tance of cut-through areas with

detectable warnings was em phasized. One instructor said that it was better to have some slope onthe cut-through area to provide an additional cue to the pedestr ian with vision im pair ment thatthey had reached the island, and to prevent accumulation of debr is in the cut-through area. Someinstructor s stated that landscaping on islands (if diff icult to step into, such as holly bushes), or low signs on islands, could be problematic if a pedestr ian with vision im pair ment is not alignedcorrectly when crossing to an island. Some instructor s commented that channelized r ight-turn

38



lanes with acceleration lanes are very diff icult for pedestr ians with vision im pair ment to crossdue to higher vehicle s peeds and lower yield rates by motor ists.

3.2.2 Relevant Findings f r om NCHRP Pr o ject 3-78A

As par t of NCHRP Pro ject 3-78A, Cr o ssing Sol ution s at Roundabouts and C hanneliz ed

Ri ght T ur n Lane s f or Pede strian s wit h V ision Disabilitie s (20), data were collected on crossings by pedestr ians with vision im pair ment at a channelized r ight-turn lane in Char lotte, NC. The data

were collected by Ms. Janet Bar low, the same O&M s pecialist who conducted the interviews for the current research. The data from NCHRP Pro ject 3-78A showed that many of the pedestr ianswith vision im pair ment had very little exper ience with channelized r ight-turn lanes. Althoughthey may have crossed a channelized r ight-turn roadway, they had minimal under standing of thelayout and var iations in design of channelized r ight-turn lanes. Some pedestr ians crossed the

channelized r ight-turn lane when it seemed quieter, or when they thought there was a break intraff ic, but it was diff icult for most par tici pants to clear ly descr i be their strategy for crossing achannelized r ight-turn lane. Interesting anecdotal infor mation from that research is that several

par tici pants said they could hear better to make crossing decision from the curb as opposed tofrom the island. They stated that the sound of traff ic behind them made the crossing decisionmore diff icult when waiting on the island. This is an interesting point because it is expected that

pedestr ians are more visi ble to dr iver s when on the island, since dr iver s are predominantlyfocusing to their lef t, on the traff ic they will be merging into.

Summary of Findings f r om Observational Field Studies and3.3Interviews with Orientation and Mobility Specialists

Channelized r ight-turn lanes generally operate well for motor ists and most pedestr ians.

There are very few instances of motor ist or pedestr ian avoidance maneuver s, and motor istsgenerally yield to pedestr ians. Pedestr ians with vision im pair ment, however, f ind channelizedr ight-turn lanes to be more of a challenge to traver se. Key f indings from the observational f ieldstudies and interviews with O&M s pecialists are as follows:

Field observational studies of vehicle and pedestr ian inter sections were conducted at35 channelized r ight-turn lanes with pedestr ian crossings. A ma jor ity of the sites (near ly70 percent) had marked crosswalk s located near the center of the channelized r ight-turnroadway; only about 30 percent of crosswalk s were located at the upstream or

downstream end of a channelized r ight-turn lane. The highway agency survey conductedin NCHRP Pro ject 3-72 (3) conf ir ms that highway agencies prefer a crosswalk locationnear the center of a channelized r ight-turn lane; over 70 percent of highway agenciesrepor ted that their practice was to place crosswalk s near the center of channelized r ight-turn lanes.

Over 2,800 pedestr ian crossings of channelized r ight-turn lanes were observed in theobservational f ield studies. Avoidance maneuver s (e.g., pedestr ian hesitates, stops, or

retreats in crosswalk due to presence of vehicle; motor ist swerves of abruptly stops to

39



avoid pedestr ian) appear to be relatively rare and were made in less than one percent of all observations.

Approximately 25 percent of pedestr ians did not cross within the crosswalk. This mayindicate the need to better plan pedestr ian routing ad jacent to the inter section such thatthe crosswalk provides the most direct route. However, the pedestr ians who crossedoutside the crosswalk generally did so when no vehicular traff ic was presented and this

behavior did not cause any traff ic conf licts.

At sites with pedestr ian signals, approximately 72 percent of pedestr ians crossed dur ing

the pedestr ian crossing phase; the remainder crossed against the signal. However, the pedestr ians who crossed against the pedestr ian signal indication generally did so whenno vehicular traff ic was present and this behavior did not cause any traff ic conf licts.

While 21 percent of pedestr ians crossed aggressively, only 4 percent walked betweenstopped vehicles to cross the channelized r ight-turn lane.

Less than half (40 percent) of observed vehicles yielded to a pedestr ian waiting at thecurb, but near ly all motor ists yielded to pedestr ians when they were in the crosswalk

(rather than waiting at the curb). The failure of vehicles in yielding to pedestr ianswaiting to cross at a marked crosswalk is a general problem at pedestr ian crossing that isnot unique to channelized r ight-turn lanes. The yield behavior of motor ists was slightly

better (47 percent vs. 40 percent) at sites with s pecial crosswalk treatments (e.g., raised

crosswalk, pavement mark ings, signing). This may indicate that additional em phasis onsigning or other treatments may be needed to increase yielding for pedestr ians waiting at

the curb.Only 7 percent of vehicles stopped in a location that blocked the crosswalk.

O&M s pecialists do not have a unif ied preference for crosswalk location at channelizedr ight-turn lanes, but would like to see more of a consistency in crosswalk locations. Thiswould make it easier to teach pedestr ians with vision im pair ment how to better traver se a

channelized r ight-turn lane.

O&M s pecialists have a strong preference for raised islands with “cut-through”

pedestr ian paths, which provide better guidance for pedestr ians with vision im pair mentthan painted islands.

Use of a consistent design with res pect to traff ic control and crosswalk location is

recommended.

Channelized r ight-turn lanes with acceleration lanes are very diff icult for pedestr ians

with vision im pair ment to cross due to higher vehicle s peeds and lower yield rates bymotor ists.

40



Chapter 4.Traff ic Operational Analysis of ChannelizedRight-Tur n Lanes

This chapter presents the results of a traff ic operational analysis of channelized r ight-turnlanes conducted with the VISSIM simulation model.

The pr imary traff ic operational reasons for providing channelized r ight-turn lanes are toincrease vehicular capacity at an inter section and to reduce delay to dr iver s by allowing them toturn at higher s peeds and reduce unnecessary stops. Channelized r ight-turn lanes appear to

provide a net reduction in motor vehicle delay at inter sections where they are installed, althoughno existing data and no established methodology have been available to directly com pare theoperational perfor mance of urban inter sections with and without channelized r ight-turn lanes. Atraff ic operational evaluation of channelized r ight-turn lanes, with and without pedestr ian signals

on the r ight-turn roadway, was conducted to quantify the differences between alternative designs.

Four key questions related to traff ic operations at channelized r ight-turn lanes were centralto the traff ic operational analysis. They are:

What is the traff ic operational perfor mance of channelized r ight-turns lanes?

What traff ic operational benef its would be lost if channelized r ight-turn lanes were not

used?

What are the effects of different geometr ic designs of channelized r ight-turn lanes(location of crosswalk, turning radius, etc.) on traff ic operational perfor mance?

What are the effects of traff ic control strategies on the operation of channelized r ight-turn lanes?

The answer s to these questions play a key role in the decision of whether a channelizedr ight-turn lane should be installed at an inter section. In order to answer these questions, theoperational research focused on s pecif ic issues that could be addressed through the use of traditional traff ic analysis modeling. These included:

Im pact of providing a channelized r ight-turn lane on traff ic delay at var ious traff icvolume levels

Im pact of pedestr ians on signalized and unsignalized r ight-turn movements

Im pact of key design features on the operational perfor mance of a channelized r ight-turnlane including:

- Location of pedestr ian crosswalk - R adius of channelized r ight-turn roadway- S peed of the cross-street onto which the r ight-turn vehicle is turning- Provision of acceleration and deceleration lanes- Effects of signal timing strategies

41



Traff ic Operational Modeling4.1

The ma jor ity of the traff ic operational studies were perfor med using simulation modeling,which can test many design, traff ic volume, and pedestr ian volume com binations. Simulationmodeling allows for the evaluation of vehicle-to-vehicle and vehicle-to-pedestr ian interactions ina controlled environment.

The traff ic operational analysis was conducted to evaluate the traff ic operational

perfor mance of r ight-turning vehicle movements at signalized inter sections for threeconf igurations (illustrated in Figure 15): a conventional r ight-turn lane at a signalizedinter section, a yield-controlled channelized r ight-turn lane, and a signalized channelized r ight-turn lane. These conf igurations were chosen because they represent the most typical urbansituations in which a channelize r ight-turn lane is either used or being considered for use. A

ser ies of microscopic simulation runs (using VISSIM) were conducted to evaluate the traff icoperational perfor mance for both vehicles and pedestr ians. Three key simulation studies were

perfor med for this evaluation:

Ri ght -t ur n vehicl e del ay: The im pact of the r ight-turn volume and conf licting throughvolume on delay to the r ight-turning vehicle.

Del ay due t o pede strian cr o ssing s: The im pact of pedestr ian volume crossing theconf licting crosswalk on the delay to the r ight-turning vehicle.

Impact o f int ersection char act eristic s: The changes in r ight-turn vehicle delay due to

design of the channelized r ight-turn lane and different signal strategies, including:

- S peed of vehicles on the cross street- S peed/radius of the r ight-turning movement- Effects of the a r ight-turn over lap phase for the signalized movements- Im pacts of an acceleration lane on the delay to r ight turns

4.1.1 Modeling Conf igurations

The three modeling conf igurations used for the simulation analysis are descr i bed andillustrated below. In order to reduce the num ber of var iables that might affect the results of theanalysis, many assum ptions with res pect to the inter section design were held constant. Theseincluded roadway approach s peed, signal cycle length, and lane widths. In addition, thefollowing assum ptions were made for the r ight-turn lane:

Approach s peeds of 48 k m/h (30 mi/h)

Inf inite storage in the sub ject r ight-turn lane

Standard 3.6-m (12-f t) travel lanes

A two-phase signal with a 90-second cycle and 50 percent of the time allocated to ther ight-turn phase

42



Each conf iguration is descr i bed below.

Conf iguration 1 is a typicalsignalized inter section with aconventional (i.e., non-channelized) r ight-turn lane.The evaluation for this

conf iguration focused on thedelay for r ight-turning vehicles

based on a range of r ight-turnvolumes and conf lictingthrough volumes. In addition,

delay to r ight-turning traff icdue to pedestr ian crossings wasevaluated.

Conf iguration 2 is a typicalsignalized inter section with ayield-controlled channelized

r ight-turn lane. The evaluationfor this conf iguration focusedon the delay to r ight-turningvehicles due to conf lictingtraff ic on the cross street and

pedestr ians crossing the

channelized r ight-turn lane. In

addition, key geometr iccharacter istics of thechannelized r ight-turn lane

(turning radius, location of crosswalk) were evaluated.

Figure 15. Intersection Conf igurations for Traff ic Operational Analysis

43



Conf iguration 3 assumesa signal-controlledchannelized r ight-turn lane.

The evaluation for thisconf iguration focused onthe differences in r ight-turndelay due to signalizationof the r ight turn assumingsimilar signal timing to thatassumed inConf iguration 1.

Figure 15. Intersection Conf igurations for Traff ic Operational Analysis (Continued)

4.1.2 Modeling Appr oach

A total of 349 scenar ios were modeled in order to cover a range of vehicle and pedestr ianvolumes. For each scenar io, 30 simulations were run using VISSIM. Thir ty runs were com pletedfor each scenar io to ensure a large enough sam ple size to provide a 95 percent conf idence levelin the results. A total of 10,470 simulation runs were conducted.

4.1.3 Model Calibration and Validation

To ensure the results generated in the VISSIM model were accurate and reasonable, two

model validation analyses were conducted. The f ir st evaluated the delay parameter s for the threeconf igurations descr i bed above using the H i ghway C apacit y Manual (32) procedures asim plemented in the Highway Capacity Sof tware R elease 5 (HCS), and by using the Synchro 7sof tware program. Volume and delay data were collected at two of the observational f ield studylocations and used in the second validation analysis.

In VISSIM, saturation f low rates values cannot be modif ied directly because they are afunction of the car following model, and therefore deter mined through safety distance

parameter s. Conver sations with PTV Amer ica, vendor of VISSIM in Nor th Amer ica, indicated

44



VISSIM simulation can result in var ious saturation f low rates depending on a num ber of factor s.A ser ies of tests were conducted to deter mine the correct set of safety distance parameter s to

produce results similar to those produced by HCS and Synchro. Saturation f low rates, r ight-turndelay, and queue lengths produced in the VISSIM model closely correlated to the resultsmeasured in the f ield.

HCS and Synchr o Comparison

For Conf iguration 1, HCS and Synchro were used to analyze the r ight-turn delay for var iousvolume alternatives, and results such as delay, queue length, and saturation f low rates wereobtained and used for validation. The HCS analysis and Synchro analysis produced similar saturation f low rates and delay results. Table 9 com pares the results of the HCS, Synchro, andVISSIM analyses.

Table 9. HCS and Synchro R ight-Turn Movement Delay Calibration Results

Traff icvolume(veh /h)

Right-tur n volume (veh /h)

50 250 500

HCS Synchr o VISSIM HCS Synchr o VISSIM HCS Synchr o VISSIM

200 14.5 14.2 13.7 25.1 17.2 27.8

800 17.7 16.7 17.4

1600 14.5 14.5 13.7 25.1 25.4 27.8

As shown in Table 9, the VISSIM results were similar to the HCS and Synchro for all caseswith the exception of the case when the r ight-turn volume is 500 veh/h per hour and theconf licting through traff ic volume is 200 veh/h. Fur ther review identif ied that the VISSIM model

delay for the r ight-turning movement increased rapidly around 500 veh/h es pecially at lowconf licting through traff ic volumes. However, it was deter mined that the other measurements atlower r ight-turning volumes were well cali brated and changing the VISSIM parameter s fur ther might reduce the overall accuracy for the range of modeling scenar ios.

Field Observation Comparison

For Conf iguration 2, two sites were chosen to collect volume and delay data for thecali bration. The two sites used for the cali bration are located in Por tland, Oregon, and Boise,Idaho. The results of the f ield data and VISSIM model com par ison are shown below:

Por tland, Oregon: 23rd Street and Burnside Street

- Field measurement of r ight-turn delay—8.3 sec- VISSIM model estimate of r ight-turn delay—8.3 sec

Boise, Idaho: Broadway Avenue and War m S pr ings Avenue (skewed inter section)

- Field measurement of r ight-turn delay—8.3 sec- VISSIM model estimate of r ight-turn delay—7.2 sec

45



As shown above, the VISSIM results for r ight-turn delay matched the f ield measurements at theinter section in Por tland, and were only 0.9 sec (13.3 percent) lower than the f ield measurementsfor the site in Boise.

Based on the com par ison of the VISSIM results to the delays from HCS, Synchro, and f ield

data, it was deter mined that the VISSIM model was adequately cali brated for use.

Base Modeling Results for Each Conf iguration4.2The base modeling results for Conf igurations 1, 2, and 3 are presented below.

4.2.1 Conf iguration 1

For Conf iguration 1, two signal operational scenar ios were evaluated. The f ir st scenar ioassumes the typical r ight-turn-on-red (R TOR ) operation in which a r ight turn can be made af ter

stopping if there is a suff icient gap in the conf licting through traff ic stream and if no pedestr iansare present near the crosswalk conf licting with the turning vehicle. The second scenar io assumesthat R TOR is not allowed and the r ight-turning vehicle cannot proceed until it receives a greensignal indication and the conf licting through traff ic is stopped. Figure 16 illustrates the R TOR

movement and the pedestr ian crossing movement.

Figure 16. R TOR and Pedestrian Crossing Movement Considered in Analysis

Figure 17 shows the delay for r ight-turning vehicles for three r ight-turn volumes for eachscenar io. As shown in Figure 17, the conf licting through-traff ic volume has a substantial im pact

on the amount of delay for a r ight-turning vehicle when R TOR is per mitted. When conf lictingthrough-traff ic volumes approach 1,600 veh/h, and few gaps exist for r ight-turning vehicles, thedelay exper ienced by vehicles at R TOR inter sections approach that of vehicles at inter sections

where R TOR is not per mitted. When R TOR is not per mitted, the delay is greater for higher r ight-turn volumes, but is not im pacted by the volume of conf licting through traff ic, since the traff icsignal green phase is f ixed. The greatest benef it of R TOR appear s to be achieved whenconf licting through traff ic volumes are low.

46



Figure 17. Delay Comparison of Conf iguration 1 - Conventional R ight-Turn Lane

(With and Without R TOR)


Figure 18 shows the delay for the r ight-turn movement for Conf iguration 2 (channelizedr ight-turn lane with Yield control). As shown in Figure 17, delay increases from approximately

1 to 2 sec/veh for the lower conf licting through volume of 200 veh/h to approximately 10 to15 sec/veh at a conf licting through volume of 1,600 veh/h. As conf licting through volumeincreases, the im pact of r ight-turn volume on delay also increases.


Figure 19 shows the delay for the r ight-turn movement for Conf iguration 3 (channelizedr ight-turn lane with a signal-controlled r ight turn). For this conf iguration, it was assumed thatR TOR was not per mitted, although it is recognized that many inter sections with signalizedchannelized r ight-turn lanes allow R TOR for single-lane r ight-turn conf igurations. The reasons

for assuming no R TOR is because, in most cases, signalization of the r ight-turn is typicallyim plemented at locations with high r ight-turn volumes, high pedestr ian crossing volumes, or

both. Under these circumstances, R TOR is typically considered problematic due to pedestr ian

conf licts and not as benef icial to vehicles since a ma jor ity of the r ight-turning vehicles will beserved dur ing the green phase of signal. In addition, a pr imary purpose of evaluatingConf iguration 3 is to com pare the results to Conf iguration 1 under similar signal operationalassum ptions to deter mine the extent of delay change by channelizing the signalized r ight turn.

Right-Tur n Hour ly Volume

47



Figure 18. Delay for Conf iguration 2 (Yield-Controlled Channelized R ight-Turn Lane)

Figure 19. Delay for Conf iguration 3 (Signalized Channelized R ight-Turn Lane)

Right-Tur n Hourly Volume


48



As shown in Figure 19, the r ight-turn delays ranged from approximately 14 sec for a r ight-turn volume of 100 veh/h to approximately 20 sec for a r ight-turn volume of 500 veh/h. This iscom pared to approximately 15 and 29 sec, res pectively, for Conf iguration 1.

When com par ing the results shown in Figure 19 with those shown in Figure 17 for the “NoR TOR ” scenar io, there appear s to be a much larger difference in delay between the three volumelevels for Conf iguration 1 (conventional r ight-turn lane) than for Conf iguration 3 (signalizedchannelized r ight-turn lane). Since traff ic volumes and traff ic control assum ptions were identicalin both scenar ios, this difference does not seem reasonable. Fur ther review found that the delay

for r ight-turn volume increased substantially between 400 and 500 veh/h for Conf iguration 1with No R TOR . This delay increase did not occur for Conf iguration 3. Other programs such asHCS and Synchro model both conf igurations identically and therefore do not suppor t such asubstantial difference. Therefore, caution is recommended with regard to the s pecif ic valuesshown repor ted for the 500 veh/h r ight-turn scenar ios.

4.2.4 Right-Tur n Delay Reduction Due to Channelized Right-Tur n Lane for BaseConf igurations

Figure 20 shows the r ight-turn vehicle delay for each conf iguration for r ight-turning

volumes of 100 veh/h, 300 veh/h, and 500 veh/h. The No R TOR option for Conf iguration 1 isnot included in Figure 20 because it is typically only used at com plex inter sections where signalstrategies, such as over lap phasing for the r ight-turns, are utilized.

Figure 20. Delay Comparison of Conf igurations 1, 2, and 3

0

5

10

15

20

25

30

200 400 600 800 1000 1200 1400 1600

D e l a y ( s e c / v e h )

Config#1 (Signal with RTOR)

Config#2 (CRTL Yield Controlled)

Config#3 (Signalized CRTLNo RTOR)

Through Volume (veh/h)C1-100 C1-300 C1-500 C2-100 C2-300C2-500 C3-100 C3-300 C3-500

e

49



The yield-controlled channelized r ight-turn lane has the lowest delay of the threeconf igurations. At very low r ight-turn volumes, the im pact of a yield-controlled channelizedr ight-turn lane com pared with a conventional traff ic signal with R TOR is relatively small. This isdue to the fact that delays are low in general when r ight-turn volumes are low. However, theim pact of a yield-controlled channelized r ight-turn lane increases as the conf licting throughtraff ic volume increases.

When the conf licting through traff ic volume is between 1,200 veh/h and 1,600 veh/h,Conf iguration 3 (signalized channelized r ight-turn lane) exper iences similar delays to

Conf iguration 1 (signalized conventional r ight-turn lane with R TOR ), which indicates that thereare fewer gaps for R TOR vehicles at these conf licting through-traff ic volumes. At a conf lictingthrough volume of approximately 1,400 veh/h, the r ight-turn vehicle delays for Conf iguration 3are similar to those in Conf iguration 2.

Figure 21 shows the resulting delay reductions due to the channelized r ight-turn lane inConf iguration 2 ver sus the conventional r ight-turn lane with R TOR in Conf iguration 1.

Figure 21. R ight-Turn Delay Reduction Due to a Channelized R ight-Turn Lane

Conf iguration 1 (R TOR) Vs. Conf iguration 2

As shown in Figure 21, the average vehicle delay reduction for r ight-turning vehicles at the

yield-controlled channelized r ight-turn lane increases slightly as the r ight-turn volume increases.This indicates that the channelized r ight-turn lane reduces delay com pared to a conventionalr ight-turn lane with R TOR , even at high conf licting through traff ic volumes. The delay shown inFigure 21 equates to a 25 to 75 percent reduction for r ight-turning vehicles.


50



The greater reduction in delay that results from the increase in conf licting through traff icvolumes is likely because the R TOR vehicles for Conf iguration 1 require a larger gap in traff icthan the merging vehicles for Conf iguration 2 dur ing the red signal phase. Figure 22 shows asimilar com par ison with the Conf iguration 1 without R TOR option. As shown in Figure 22, thedelay reduction due to the channelized r ight-turn lane decreases as the conf licting through traff icvolume increases, even without the R TOR for Conf iguration 1.

Figure 22. R ight Turn Delay Reduction Due to a Channelized R ight-Turn Lane

Conf iguration 1 (Without R TOR) Versus Conf iguration 2

4.2.5 Summary of Results for Base Conf igurations

The analysis results of r ight-turn delay for each base conf iguration are summar ized below:

For Conf iguration 1 (conventional r ight-turn lane), conf licting through-traff ic volumehas a substantial im pact on the amount of delay for a r ight-turning vehicle when R TOR is per mitted. When conf licting through-traff ic volumes approach 1,600 veh/h, few gaps

exist for r ight-turning vehicles, and the delay exper ienced by vehicles at R TOR

inter sections approaches that of vehicles at inter sections where R TOR is not per mitted.Thus, the greatest benef it of R TOR appear s to be achieved when conf licting through

traff ic volumes are low.

For Conf iguration 2 (yield-controlled channelized r ight-turn lane), delay increases from

approximately 1 to 2 sec/veh at a conf licting through volume of 200 veh/h toapproximately 10 to 15 sec/veh at a conf licting through volume of 1,600 veh/h. Asconf licting through volume increases, the im pact of r ight-turn volume on delay alsoincreases.


51



For Conf iguration 3 (signalized channelized r ight-turn lane), delay is generally notim pacted by the volume of conf licting traff ic on the cross street, unless the signal timingis changed to accommodate conf licting traff ic volumes. Therefore, signalization of thechannelized r ight-turn lane provides traff ic operational benef its only at high conf lictingvolumes on the cross street. Delays ranged from approximately 14 to 20 seconds for r ight-turn volumes of 100 to 500 veh/h, res pectively.

Based on these f indings, the following can be concluded:

The use of a yield-controlled channelized r ight-turn lane can substantially reduce thedelay exper ienced by r ight-turning vehicles. Com par ing Conf iguration 2 toConf iguration 1 with R TOR , the channelized r ight-turn lane provides a delay reductionof between 25 and 75 percent for r ight-turning vehicles, and provides a delay reductioneven at high conf licting through traff ic volumes.

At lower r ight-turn volumes, the traff ic operational benef it of a channelized r ight-turnlane is relatively small as com pared with the conventional r ight-turn lane in

Conf iguration 1.

At high conf licting traff ic volumes on the cross street, delay exper ienced by r ight-

turning vehicles becomes similar for yield- and signal-controlled conditions atchannelized r ight-turn lanes.

Impacts of Pedestrians on Base Conf iguration Results4.3

A key focus of this research was on how pedestr ians use the channelized r ight-turn lane andthe effects of pedestr ians on the operation of the channelized r ight-turn lane. In order to evaluatethe operational effects of pedestr ians, the following issues were evaluated:

The im pact of pedestr ian crossings on r ight-turn delay for Conf iguration 2 (yieldcondition)

The effect of crosswalk location on the delay to r ight-turning vehicles for Conf iguration 2

Delay to pedestr ians waiting to f ind a gap in r ight-turning traff ic

Each of these issues was evaluated using the VISSIM models developed for the baseconf igurations.

4.3.1 Impact of Pedestrian Cr ossings on Delay at Channelized Right-Tur n Lanes

Figure 23 shows the im pact to r ight-turning delay of var ious pedestr ian volumes for the

three study conf igurations. The model assumes that all vehicles yield to pedestr ians approachingthe crosswalk ; however, the observational studies discussed in Chapter 3 of this repor t indicatethat dr iver s of ten do not yield to pedestr ians waiting to cross, but instead only to those who have

begun crossing. Therefore, the delay data shown in Figure 23 may be slightly greater than what

52



would be exper ienced in a location in which a por tion of the vehicles do not yield to pedestr iansapproaching the crosswalk.

Figure 23. Delay Due to Pedestrian Crossings —Conf iguration 2 Channelized R ight Turn

Lane (300 veh/h R ight-Turn and 800 veh/h Conf licting Through)

As shown in Figure 23, pedestr ian volume has a clear effect on the r ight-turn delay inConf igurations 1 and 2. The increase in delay from zero pedestr ians to a pedestr ian crossingvolume of 200 ped/h is about 50 percent for Conf iguration 1 and 60 percent for Conf iguration 2.

For all pedestr ian volumes, the yield-controlled channelized r ight-turn lane conf iguration resultsin lower delay for r ight-turning vehicles than a conventional r ight-turn lane, but at the higher

pedestr ian volumes the delay for r ight-turning vehicles near ly doubles. The delay for Conf iguration 1 with R TOR is close to the delay for Conf iguration 3 (which does not haveR TOR ) at the highest pedestr ian volumes, indicating that pedestr ian crossings substantiallyreduce the ability of vehicles to turn r ight dur ing the red signal phase (when the pedestr ian havethe “walk” indication at the traff ic signal). This suppor ts the notion that if pedestr ian crossing

volumes are very high, R TOR has marginal benef it.

Figure 24 shows the analysis results for Conf iguration 2 with r ight-turn volumes of 100,300, and 500 veh/h. As shown in the f igure, all three r ight-turn volume scenar ios are affected

similar ly by the pedestr ian crossings. At r ight-turn volumes of 500 veh/h, there is an increase indelay of 4.5 sec (70 percent)—from 0 to 200 ped/h—and at r ight-turn volumes of 100 veh/h, theincrease in delay from 0 to 200 ped/h is about 2.5 sec (70 percent).

0

2

4

6

8

10

12

14

16

18

20

0 50 100 150 200

R i g h t T u r n D e l a

y ( s e c / v e h )

Pedestrian Volume (ped/h)

Con fig 1 Con fi g 2 Con fig 3

53



Figure 24. Delay Due to Pedestrian Crossings —Conf iguration 2 Channelized R ight-Turn

Lane (800 veh/h Conf licting Through Volume)

4.3.2 Impact of Cr osswalk Location

Three separate crosswalk locations (upstream, center, and downstream) were modeled for Conf iguration 2. Table 10 shows the results of the crosswalk location on r ight-turn delay for alocation with a r ight-turn volume of 300 veh/h and a pedestr ian volume of 50 ped/h. The resultsshow a marginal difference in average delay.

Table 10. Delay Impacts of Crosswalk Location (300 vph and 50 ped/h)

Cr osswalk location Upstream Middle Downstream

Delay (sec) 5.7 6.2 5.9

4.3.3 Potential Delay to Pedestrians

Figure 25 shows the potential delay a pedestr ian might exper ience for pedestr ian volumesranging from 50 to 200 ped/h and r ight-turn volumes of 100 to 500 veh/h and assuming that

vehicles do not yield to pedestr ians. This scenar io represents a “wor st case scenar io” for pedestr ians, since observational studies revealed that approximately 40 percent of the vehicles doyield to pedestr ians waiting to cross and near ly all yield once a pedestr ian enter s the crosswalk.

As shown in Figure 25, the potential delay for pedestr ians is relatively large at r ight-turnvolumes of 300 and 500 veh/h. At such high r ight-turn volumes, pedestr ians may becomefrustrated while trying to cross a channelized r ight-turn lane and may run across or accept verysmall gaps in traff ic.

0

2

4

6

8

10

12

14

0 50 100 150 200

R i g h t T u r n

D e l a y ( s e c / v e h )


100 300 500Right Tur n Hourly Volume

54



Figure 25. Pedestrian Delay Waiting for a Gap in R ight-Turning Traff ic

4.3.4 Summary of Results for Pedestrian Impacts on Traff ic Operations

The results of the pedestr ian analysis are summar ized below:

Pedestr ian volume increases r ight-turn vehicle delay by 50 to 70 percent for

Conf igurations 1 and 2. This is due to a substantially reduced ability of motor ists to turnr ight dur ing a red signal phase.

The location of the crosswalk is not a key factor with res pect to delay for r ight-turn

vehicles.

Three separate crosswalk locations (upstream, center, and downstream) were modeled

for Conf iguration 2. R esults suggest that crosswalk location has a marginal effect (nomore than 0.5 sec) on delay.

At moderate r ight-turn volumes (300 veh/h), the potential delay to pedestr ians waitingfor a gap under Conf iguration 2 is between 15 and 30 seconds, which is likely similar to

the level of delay that might be exper ienced for a signalized crossing.

0

5

10

15

20

25

30

50 100 150 200

R i g h t T u

r n D e l a y ( s e c / v e h )


100 300 500Ri

55



Impacts of Geometric Characteristics and Signal Phasing on4.4Channelized Right-Tur n Lane Delay

The geometry of the channelized r ight-turn lane for Conf iguration 2 as well as the type of

signal phasing for Conf iguration 3 are thought to affect the delay exper ienced by r ight-turningvehicles. The VISSIM simulation models created for the delay studies were used to quantify therelative im pact of these factor s:

Addition of an acceleration lane (Conf iguration 2)Channelized r ight-turn lane radius (Conf iguration 2)

Im pact of adding additional green time by im plementing an over lap phase for asignalized channelized r ight-turn lane (Conf iguration 3)

4.4.1 Acceleration Lane

Figure 26 shows the simulation results for Conf iguration 2 with and without a 61-m (200-f t),full-width acceleration lane. The addition of an acceleration lane reduces the delay for the fullrange of conf licting through volumes and r ight-turning volumes. The acceleration lane reduces

the r ight-turn delay by 65 percent at low conf licting through volumes and by 85 percent at higher conf licting through volumes.

Figure 26. Delay Comparison With Acceleration Lane (Conf iguration 2)


56



4.4.2 Channelized Right-Tur n Lane Radius and Speed Impacts

The effect of channelized r ight-turn lane radius on r ight-turn delay was evaluated bychanging the s peed of the channelized r ight-turn lane for var ious conf licting through volumesand for two roadway s peeds. Vehicle s peed along the channelized r ight-turn lane was used as asurrogate for channelized r ight-turn lane radius, since vehicle s peed is limited on narrower curves (smaller radius channelized r ight-turn lanes) and higher for channelized r ight-turn laneswith larger radii. Table 11 shows the results of the analysis for Conf iguration 2.

Table 11. Delay Impacts of Channelized R ight-Turn Lane Speed/Radius

for Conf iguration 2

Two thr oughlanes traff ic volume

(veh /h)

Delay for right-tur ning vehicles

10 mi/h(15- to 20-ft radius)



Conf licting thr ough volume speed = 35 mi/h

600 4.3 3.4 2.91,000 8.4 7.1 6.51,400 13.3 11.5 10.8

Conf licting thr ough volume speed = 45 mi/h

600 4.5 3.6 3.11,000 8.0 6.8 6.11,000 12.6 11.0 10.3

As shown in Table 11, increasing the radius of the r ight turn (which increases the travel

s peed along the channelized r ight-turn lane) reduces the delay by approximately 10 to 20 percentfor each 8-k m/h (5-mi/h) increase in turning s peed. Larger delay reductions are seen at lower through-lane volumes. Delay decreases slightly when the s peed of the conf licting through

volume is increased from 56 to 72 k m/h (35 to 45 mi/h) for through volumes of 1,000 veh/h or greater, but increases slightly for the lower volumes. Based on the observational studies, most of the medium island sizes had a radius of 18 to 31 m (60 to 100 f t).

4.4.3 Traff ic Signal Phasing for Signalized Channelized Right-Tur n Lanes

At a signalized channelized r ight-turn lane (Conf iguration 3), it is common for traff icengineer s to provide extra green time to the r ight-turn movement without increasing cycle length

by over lapping the r ight turn with the cross street lef t turn. Figure 27 illustrates the concept of ar ight-turn over lap. A potential drawback of the r ight-turn over lap phasing is that U-turns cannot

be per mitted from the cross-street lef t-turn lane, as they may conf lict with r ight-turning vehicles.The elimination of the U-turn can be a signif icant issue in states where U-turns are allowed at allinter sections and in areas that have median access control and U-turns are the pr imary means of

providing lef t-turn access to businesses.

57



Figure 27. R ight-Turn Overlap

Table 12 com pares the delay for the base condition to providing additional green timethrough use of an over lap signal phase. An analysis was conducted for provision of 10, 15, and

20 seconds of additional green time to the 45-sec signal phase under the base condition for ther ight-turn movement.

Table 12. Delay Impacts of Adding Additional Green Time to R ight-Turn Movement

Overlap /added greentime for right tur n

(sec)

Average right-tur n delay (sec)(for pedestrian volume of 100 ped /h)

Right-tur n volume (veh /h)

100 300 500

0 6.2 10.7 19.910 5.9 9.5 16.015 5.8 9.0 14.820 5.5 8.5 13.5

As shown in Table 12, the additional green time is more effective at reducing delay as the

r ight-turn volume increases from 100 to 500 veh/h. Each increment of addition green time for the

r ight turn movement provides approximately a 5 to 10 percent decrease in delay for the r ight-turnvehicles.

58



4.4.4 Summary of Results for Geometric and Signal Element on ChannelizedRight-Tur n Lane Traff ic Operations

The results of the geometr ic and signal operations analysis of r ight-turn delay suggest thefollowing:

Acceleration lanes substantially reduce r ight-turn delay at all volume levels.

Increasing the radius of the channelized r ight-turn roadway reduced the delay by

approximately 10 to 20 percent for each 8 k m/h (5-mi/h) increase in turning s peed.Larger delay reductions were observed at lower through-lane volumes.

Use of an over lap phase or other methods of providing additional green time to r ight-

turning vehicles can substantially reduce the delay for a signalized channelized r ight-turnlane but may result in other im pacts to inter section operations such as restr icting U-turnsmaneuver s.

Summary of Traff ic Operational Analysis Findings4.5

Based on the results of the simulation modeling, channelized r ight-turn lanes cansubstantially reduce delay for r ight-turning vehicles in near ly every traff ic volume scenar io. Site-

s pecif ic factor s, such as pedestr ian volumes and the geometry of the channelized r ight-turn lane,have an effect on the level of im provement. Therefore, these factor s are im por tant in deter mining

the delay benef its that may result from installation of the channelized r ight-turn lane. Followingare the key f indings:

A yield-controlled channelized r ight-turn lane can decrease r ight-turn delay by25 to 75 percent com pared with a conventional r ight-turn lane at a signalizedinter section. At most volume levels, a conventional r ight-turn lane at a signalizedinter section provides lower delay than a signalized channelized r ight-turn lane due to theuse of R TOR . At high pedestr ian volumes or high conf licting cross-street traff ic

volumes, a conventional r ight-turn lane with R TOR results in similar delays to signalcontrol without R TOR . Thus, the greatest traff ic operational benef it of a conventionalr ight-turn lane with R TOR appear s to be achieved when conf licting traff ic volumes onthe cross street are low to moderate.

A pedestr ian volume of approximately 200 ped/h increases r ight-turn delay byapproximately 60 percent on a yield-controlled channelized r ight-turn lane com pared to

a base condition of no pedestr ians, assuming vehicles yield to pedestr ians. However, for all pedestr ian volumes, the channelized r ight-turn lane results in lower delay for r ight-turning vehicles than a conventional r ight-turn lane.

When r ight-turn volumes are greater than 300 veh/h, the potential delay for pedestr ianswaiting for a gap to cross a channelized r ight-turn lane could be between 15 and

30 seconds.

The addition of an acceleration lane reduces the r ight-turn delay by 65 to 85 percent

depending on the conf licting through traff ic volume.

59



Increasing the radius of the yield-controlled channelized r ight-turn lane fromapproximately 5 to 6 m (15 to 20 f t) up to 24 to 31 m (80 to 100 f t) results in a decreasein r ight-turn delay of approximately 20 to 50 percent.

Three separate crosswalk locations (upstream, center, and downstream) were modeledfor Conf iguration 2. R esults suggest that crosswalk location has a marginal effect (nomore than 0.5 sec) on delay.

Use of an over lap phase, or other methods of providing additional green time to r ight-turning vehicles, can substantially reduce the delay for a signalized channelized r ight-

turn lane, but may result in other im pacts to inter section operations, such as restr ictingU-turn maneuver s.

60



Chapter 5.Saf ety Analysis of Channelized Right Tur ns

This chapter presents the results of a safety analysis of channelized r ight-turn lanes.

A safety analysis of channelized r ight-turn lanes, in com par ison to other r ight-turntreatments, was under taken because of concerns expressed at the outset of the research that

channelized r ight-turn lanes might exper ience more crashes (and, in par ticular, more pedestr iancrashes) than other r ight-turn treatments. To investigate this concern, two key questions relatedto safety at channelized r ight-turn lanes were addressed in the safety analysis:

What is the safety perfor mance of channelized r ight-turn lanes? S pecif ically, how doesthe safety perfor mance of inter section approaches with channelized r ight-turn lanes

com pare to that of inter section approaches with conventional r ight-turn lanes or sharedthrough/r ight-turn lanes?

What safety benef its would be lost if channelized r ight-turn lanes were not used?

An overall com par ison was perfor med of the safety perfor mance of inter section approacheswith channelized r ight-turn lanes to inter section approaches with other r ight-turn treatments.S pecif ically, a cross-sectional analysis was conducted to com pare the crash exper ience among:

Inter section approaches with channelized r ight-turn lanesInter section approaches with conventional r ight-turn lanesInter section approaches with no r ight-turn treatments (shared through/r ight-turn lanes)

The cross-sectional analysis involved com par ing mean and median crash frequencies andrates for each of the three inter section approach types and developing negative binomialregression relationshi ps for crash frequencies as a function of traff ic volume. The crash dataanalyses looked separately at motor-vehicle crashes and pedestr ian crashes.

Database Development5.1

Seven (7) year s (1999 to 2005) of motor-vehicle and pedestr ian crash and volume data wereobtained for a total of 103 four-leg signalized inter sections in Toronto, Ontar io, Canada. TheToronto data represent a unique resource because they include both vehicles turning movementvolumes and pedestr ian crossing volumes by inter section approach, as well as crash data that can

be classif ied by inter section approach and turning movement. These are the same data that were

used to develop the pedestr ian safety prediction model for Chapter 12 of the H i ghway Sa f et y

Manual (HSM) (33, 34).

Initially, a com par ison of inter sections with and without channelized r ight-turn lanes was planned, but the inter sections with channelized r ight-turn lanes did not have a consistent patternof approaches with and without channelized r ight-turn lanes. That is, some inter sections had a

61



channelized r ight-turn lane on one approach; other s had a channelized r ight-turn lane on two or more approaches. Therefore, the analysis was conducted at the inter section approach rather thanat the inter section level. The num ber of inter section approaches of each r ight-turn treatment typeincluded in the analysis was:

Shared through/r ight-turn lane (designated as STR ): 217 inter section approachesConventional r ight-turn lane (designated as R TL): 95 inter section approachesChannelized r ight-turn lane (designated as CR T): 83 inter section approaches

Cr oss-Sectional Crash Analysis Appr oach5.2

The safety analysis focused on r ight-turn crashes involving motor vehicles and/or pedestr ians; separate analyses were conducted for motor-vehicle crashes and pedestr ian crashes.The effect of r ight-turn channelization on motor-vehicle and pedestr ian crashes of interest wasestimated by means of a cross-sectional analysis in which a single statistical model, including anindicator var iable for approach type (STR , R TL, or CR T) and r ight-turn motor-vehicle and

pedestr ian volumes, was developed using all three approach types. A negative binomial (NB)regression model was used, with the general for m as follows, for motor-vehicle and pedestr ian

crashes, res pectively:

(1)

or

(2)

where: NMV = predicted num ber of motor-vehicle crashes per year per approach NPed = predicted num ber of pedestr ian crashes per year per approachIType(i) = indicator (0,1) var iable for approach type i, i = 1, 2, or 3; for STR

approaches, the value of the coeff icient is b1; for R TL approaches,the value of the coeff icient is b2; for CR T approaches, the value of the coeff icient is 0

Vol1 = r ight-turning motor-vehicle (MV) volume (24-hour count); sameturning movement for all analysis models

Vol2 = MV volume (24-hour count); turning movements represented byVol2 vary with analysis model

Vol3 = MV volume (24-hour count); turning movements represented byVol3 vary with analysis model

VolPed = pedestr ian volume crossing inter section approaches of interest(24-hour count)

a, bi, c, d, e = regression coeff icientsln() represents the natural logar ithm function

Figure 28 illustrates all of the possi ble vehicle turning movements at a typical inter section,where the r ight-turn movement, designated in red (Movement 3), represents the r ight-turnmovement at a par ticular inter section approach. The other turning movements at the inter sectionare num bered as shown in the f igure in relation to the inter section approach being analyzed. A

)]Volln(e+)Volln(d+)Volln(c+I b+aexp[= N 321)i(TypeiMV

)]Volln(d+)Volln(c+I b+aexp[= N ped1)i(TypeiPed

62



multi ple-vehicle r ight-turn crash involving the inter section approach being analyzed bydef inition includes a r ight-turning vehicle (Movement 3). The other involved vehicle could

potentially be mak ing any other turning movement at the inter section, but the r ight-turningvehicle is mor e likel y to conf lict with cer tain turning movements (Movements 1, 2, 3, 7, and 11)than with other s.

The research team did not want to presuppose which movements would most likely conf lictwith the r ight-turning vehicle without evaluating all possi ble movements f ir st. Therefore, amulti-tiered analysis approach was conducted in which all movements were initially included in

the analysis, and then each subsequent analysis became more focused on the movements mostlikely to conf lict with the r ight-turn movement in question (Movement 3). Each individualanalysis is referred to in the following discussion as an “analysis model.” In all, nine differentanalysis models were investigated— labeled as Analysis Models 1, 2, 3, 4, 5, 6a, 6b, 6c, and 7— three of which are only slight var iations of one another. A detailed descr i ption of each analysis

model is provided later in this section.

Figure 28. Intersection Turning Movements Relative to an Intersection Approach

(Approach 1) with a Specif ic R ight-Turn Treatment (CR T, R TL, or STR)

63



The following sections present the statistical analysis results, separately for each analysismodel. Each section presents the following infor mation:

1. Summary of analysis inputs and cr iter ia: this includes a discussion of the choice of vehicle maneuver s, initial im pact types, directional vehicle movements, and aggregatedmotor-vehicle volumes considered; exceptions are also discussed

2. Mean and median motor-vehicle and pedestr ian volumes (based on 24-hour counts),separately for each inter section approach type

3. Motor-vehicle and pedestr ian crash statistics — minimum, maximum, and sum of crashesin the seven-year per iod, separately for total and fatal-and-in jury (FI) crashes and for each inter section approach type

4. R egression results, separately for total and FI crashes and for each inter section approachtype

5. Statistical com par ison of crash rates between CR T and STR or R TL approaches (theseresults are only provided when the r ight-turn treatment had an overall statistically

signif icant effect on safety), separately for total and FI crashes

6. Predicted year ly crash counts, separately for total and FI crashes and for each

inter section approach type

Cr oss-Sectional Crash Analysis Results5.3

5.3.1 Analysis Model 1 Results

Analysis Model 1 includes consideration of all possi ble vehicle maneuver s at eachinter section and all possi ble initial im pact types involving Movement 3 and any other vehicle,including single-vehicle crashes involving only Movement 3. While cer tain conf licts appear to

be highly unlikely (e.g., a conf lict between Movements 3 and 9), this f ir st analysis wasconducted to represent a com prehensive com par ison of the safety perfor mance of the three

inter section approach types (CR T, R TL, STR ) consider ing all possi ble maneuver s and im pacttypes. Pedestr ian crashes are excluded.

The layout for this analysis model is presented in Figure 29. The graphic in Figure 29 shows

all vehicle turning movements, labeled 1 through 12, at a typical four-way inter section. Eachsuch inter section provided four inter section approaches as separate observations for the statistical

analysis, with each approach categor ized depending on its r ight-turn treatment. Approach 1always refer s to the analysis approach considered as the pr imary approach for a par ticular observation— it can therefore be the NB, WB, SB, or EB inter section approach. The infor mationin Figure 29 (and similar subsequent f igures) is explained from the per s pective of Approach 1.

The three motor-vehicle volumes used for Analysis Model 1 in Equation 1 are explained inFigure 29, based on the 12 vehicle movements as follows:

64



Figure 29. Summary of Analysis Model 1 Inputs and Criteria

65



Vol1 always represents the r ight-turning MV volume (i.e., Vol1 corres ponds to vehicleMovement 3, indicated in red)

Vol2 always represents the sum of the remaining MV volumes of the same directionincluded in the model (i.e., Vol2 corres ponds to the sum of vehicle Movements 1 and 2)

Vol3 always represents the sum of the MV volumes of the other directions included inthe model (i.e., Vol3 corres ponds to the sum of vehicle Movements 4 through 12)

Initial im pact types and possi ble vehicle maneuver s are indicated in the subtables in

Figure 29. The possi ble vehicle maneuver s are those used in standard Ontar io crash data for mat.The subtable in Figure 29 (and in subsequent f igures with the same for mat), entitled “vehiclemaneuver s for crashes included in the analysis,” def ines which crashes were considered in the

analysis. Vehicle 1 always represents a r ight-turning vehicle. Vehicle 2 identif ies the vehiclemaneuver s considered for the second crash-involved vehicle in multi ple-vehicle crashes. Theaccom panying text in the f igure identif ies whether single-vehicle crashes involving a r ight-

turning vehicle were included or excluded.

Analysis Model 1 uses the largest dataset (with res pect to total MV crash count) of all themodels considered in the cross-sectional crash analysis. Volume statistics (mean and median) are

shown in Table 13, where Vol1, Vol2, and Vol3 are as def ined in Figure 29. Total and FI crashstatistics are shown in Table 14.

Table 13. Mean and Median Motor-Vehicle Volumes by

Intersection Type—Analysis Model 1Intersection

appr oachtype

Number of appr oaches

Mean MV volumes(24-hr counts)

aMedian MV volumes

(24-hr counts)a

Vol1 Vol2 Vol3 Vol1 Vol2 Vol3

STR 217 1,435 10,208 36,209 1,091 9,118 30,891

RTL 95 2,274 14,507 47,841 2,169 13,845 50,256

CRT 83 1,799 12,070 42,770 1,581 12,189 42,211a

Vol1, Vo l2, and Vol3 ar e as def ined in Figur e 29.

Table 14. Seven (7)-Year Total and FI Motor-Vehicle Crash Counts by

Intersection Approach —Analysis Model 1

Intersectionappr oach

typeNumber of

appr oaches

7-year total crash counts 7-year FI crash counts

Minimum Maximum Sum Minimum Maximum Sum

STR 217 0 10 365 0 3 52

RTL 95 0 14 276 0 3 36

CRT 83 0 13 185 0 5 38

The regression results for the total and FI models are summar ized in Table 15. These includethe regression coeff icients [see Equation (1)] and their 90 percent conf idence limits. The Type 3

p-value provides the signif icance level of each parameter in the model; the last column indicates

whether the parameter is statistically signif icant at the 90 percent conf idence level.

66



Table 15. Regression Results for Motor-Vehicle Crash Models —Analysis Model 1

Parameter a

Regression coeff icients

Type 3p-value

Parameter signif icant

at 90%conf idence

level?Estimate

90% lower conf idence

limit

90% upper conf idence

limit

Total motor -vehicle crashes—Analysis Model 1

Intercept –13.49 –15.99 –11.01

Intersection

appr oachtype

STR –0.04 –0.25 0.17

0.67 NoRTL 0.07 –

0.16 0.29CRT 0

Vol1 0.38 0.25 0.51 < .0001 YES

Vol2 0.15 0.01 0.29 0.09 YES

Vol3 0.76 0.50 1.03 < .0001 YES

Dispersion 0.35 0.25 0.47

Fatal-and-in jury motor -vehicle crashes—Analysis Model 1

Intercept –14.72 –20.21 –9.45

Intersectionappr oachtype

STR –0.38 –0.79 0.04

0.26 NoRTL –0.39 –0.84 0.06

CRT 0

Vol1 0.58 0.29 0.87 0.0007 YES

Vol2 0.12 –0.18 0.44 0.54 No

Vol3 0.62 0.05 1.20 0.07 YES

Dispersion 0.71 0.25 1.36a

Vol1, Vo l2, and Vol3 ar e as def ined in Figur e 29.b

Using the model form in Equation (1).

Based on this analysis model, the r ight-turn treatment has no statistically signif icant effecton total crashes (p = 0.67) or FI crashes (p = 0.26). The coeff icients (third column) for the threetypes of inter section approaches are to be interpreted as follows:

The CR T treatment is the base (com par ison) treatment in all models and its coeff icient istherefore always zero on the log-scale or one on the or iginal scale.

If the coeff icient for either STR or R TL approaches is positive, then approaches withthat type of r ight-turn treatment exper ience, on average, higher crash counts than CR Tapproaches, all volumes held constant between inter section approach types.

If the coeff icient for either STR or R TL approaches is negative, then approaches withthat type of r ight-turn treatment exper ience, on average, lower crash counts than CR T

approaches, all volumes held constant between inter section approach types.

The coeff icient estimates corres ponding to the inter section approach types shown in

Table 15 therefore provide a means for rank ing the r ight-turn treatments with res pect to their predicted crash counts (all volumes held constant between inter section approach types): thesmallest coeff icient of the three corres ponds to the lowest predicted crash count. This, however,

does not im ply statistical signif icance; the last column in Table 15 provides that infor mation.Because the r ight-turn treatment for Analysis Model 1 was not statistically signif icant at the90 percent conf idence level, the NB regression analysis was not followed-up with a directcom par ison of CR T approaches to either STR or R TL approaches.

67



Using Equation (1) and the regression coeff icients shown in Table 15, year ly MV crashcounts were predicted for all three approach types. To account for the differences in MVvolumes among the three inter section approach types (as shown in Table 13), the three modelswere applied to all three inter section approaches with MV volumes (Vol1, Vol2, and Vol3) set atthe observed mean and median volumes for CR T approaches, as shown in Table 13. The

predicted average crash frequencies are shown in Table 16 for each approach type using meanand median volumes, res pectively.

Table 16. Yearly Motor-Vehicle Crash

Predictions —Analysis Model 1


type

Yearly crash predictionsper appr oach

At mean CRTvolumes

At median CRTvolumes


STR 0.316 0.298

RTL 0.351 0.331

CRT 0.329 0.310


STR 0.046 0.042

RTL 0.045 0.041

CRT 0.066 0.061

For Analysis Model 1, the r ight-turn treatment has no statistically signif icant effect on totalcrashes or FI crashes. This may be largely due to the fact that all possi ble maneuver s and im pacttypes, several of which are unlikely to affect r ight-turn crashes involving Movement 3, have beenincluded in Analysis Model 1.


Analysis Model 2 is similar to Analysis Model 1 with the exception that two turning

movements (Movement 9 and 10), which appear to be highly unlikely to conf lict withMovement 3, have been omitted from the analysis. Analysis Model 2 includes multi ple-vehiclecrashes that involve one vehicle mak ing Movement 3 and another vehicle mak ing one of thefollowing movements: 1, 2, 3, 4, 5, 6, 7, 8, 11, or 12. Also, single-vehicle crashes and crashes

where the maneuver of the second vehicle was coded as “unknown” or “other” or was lef t blank have been excluded from the analysis. The layout for this analysis model is presented inFigure 30.

Volume statistics (mean and median) corres ponding to Analysis Model 2 are shown inTable 17, where Vol1, Vol2, and Vol3 are def ined in Figure 30. Total and FI crash statistics areshown in Table 18. The selection of possi ble vehicle maneuver s for Analysis Model 2 resulted in

a considerably smaller num ber of crashes than that for Analysis Model 1 (com pare to Table 14).

68




69




Intersection Type—Analysis Model 2


typeNumber of

appr oaches


aMedian MV volumes

(24-hr counts)a


STR 217 1,435 10,208 32,647 1,091 9,118 29,215

RTL 95 2,274 14,507 43,124 2,169 13,845 45,493

CRT 83 1,799 12,070 39,288 1,581 12,189 39,023

a

Vol1, Vol2, and Vol3 ar e as def ined in Figur e 30.

Table 18. Seven (7)-Year Total and FI Motor-Vehicle Crash Counts

by Intersection Approach —Analysis Model 2


typeNumber of

appr oaches

7-yr total crash counts 7-yr FI crash counts


STR 217 0 8 221 0 3 32

RTL 95 0 10 182 0 2 18

CRT 83 0 9 115 0 2 20

The regression results for the total and FI models are summar ized in Table 19. Based on thisanalysis model, the r ight-turn treatment has no statistically signif icant effect on total crashes(p = 0.48) or FI crashes (p = 0.39). Because the r ight-turn treatment for Analysis Model 2 was

not statistically signif icant at the 90 percent conf idence level, the NB regression analysis was notfollowed-up with a direct com par ison of CR T approaches to either STR or R TL approaches.

Using Equation 1 and the regression coeff icients shown in Table 19, year ly MV crash countswere predicted for all three approach types. To account for the differences in MV volumesamong the three inter section approach types (as shown in Table 17), the three models were

applied to all three inter section approaches with MV volumes (Vol1, Vol2, and Vol3) set at theobserved mean and median volumes for CR T approaches. The predicted average crashfrequencies are shown in Table 20 for each approach type using mean and median volumes,res pectively.

For Analysis Model 2, the r ight-turn treatment has no statistically signif icant effect on totalcrashes or FI crashes. Again, this may be largely because there are still several maneuver s andim pact types included in the analysis that are not necessar ily related to the r ight-turn movementin question.

70



Table 19. Regression Results for Motor-Vehicle Crash

Models —Analysis Model 2

Parameter a

Regression coeff icientsb

Type 3p-value


at 90%conf idence

level?Estimate

90% Lower conf idence

limit

90% Upper conf idence

limit


Intercept –16.05 –18.97 –13.19


STR –0.03 –0.27 0.200.48 NoRTL 0.12 –0.12 0.37

CRT 0

Vol1 0.42 0.26 0.57 < .0001 YES

Vol2 0.13 –0.03 0.30 0.18 No

Vol3 0.96 0.65 1.27 < .0001 YES



Intercept –19.42 –26.64 –12.61


STR –0.18 –0.69 0.34

0.39 NoRTL –0.48 –1.06 0.10

CRT 0

Vol1 0.56 0.18 0.94 0.01 YES

Vol2 0.12 –0.25 0.54 0.61 No

Vol3 1.02 0.27 1.79 0.02 YES

Dispersion 0.49 –0.11 1.48a

Vol1, Vol2, and Vol3 ar e as def ined in Figur e 30.b

Using the model form in Equation 1.



Intersectionappr oach type


At mean CRTvolumes



STR 0.195 0.184

RTL 0.227 0.214CRT 0.201 0.190


STR 0.029 0.027

RTL 0.021 0.020

CRT 0.034 0.032

71




Analysis Model 3 focuses on those vehicle maneuver s more likely to conf lict with the r ight-turning vehicle. This analysis includes the r ight-turn movement (Movement 3) and any other movement that could potentially conf lict with the r ight-turning vehicle, either at the depar tureend of the r ight turn (Movements 1 and 2) or as the r ight-turning vehicle merges into traff ic onthe cross street (Movements 7 and 11). The layout for this analysis model is presented inFigure 31.

Volume statistics (mean and median) corres ponding to Analysis Model 3 are shown inTable 21, where Vol1, Vol2, and Vol3 are def ined in Figure 31. Total and FI crash statistics areshown in Table 22. The selection of possi ble vehicle maneuver s for Analysis Model 3 resulted ina yet smaller num ber of crashes than that for Analysis Model 2 (com pare to Table 18).

Table 21. Mean and Median Motor-Vehicle Volumes by Intersection

Type—Analysis Model 3


typeNumber of

appr oaches


aMedian MV volumes

(24-hr counts)a


STR 217 1,435 10,208 10,323 1,091 9,118 8,937

RTL 95 2,274 14,507 13,220 2,169 13,845 13,697

CRT 83 1,799 12,070 12,786 1,581 12,189 13,743

aVol1, Vol2, and Vol3 ar e as def ined in Figur e 31.




typeNumber of

appr oaches



STR 217 0 8 172 0 2 20

RTL 95 0 9 154 0 2 17

CRT 83 0 9 86 0 2 14

The regression results for the total and FI models are summar ized in Table 23. Based on thisanalysis model, the r ight-turn treatment has no statistically signif icant effect on total crashes(p = 0.28) or FI crashes (p = 1). Because the r ight-turn treatment for Analysis Model 3 was not

statistically signif icant at the 90 percent conf idence level, the NB regression analysis was notfollowed-up with a direct com par ison of CR T approaches to either STR or R TL approaches.

Using Equation (1) and the regression coeff icients shown in Table 23, year ly MV crash

counts were predicted for all three approach types. To account for the differences in MVvolumes among the three inter section approach types (as shown in Table 21), the three modelswere applied to all three inter section approaches with MV volumes (Vol1, Vol2, and Vol3) set atthe observed mean and median volumes for CR T approaches. The predicted average crashfrequencies are shown in Table 24 for each approach type using mean and median volumes,res pectively.

72




73



Table 23. Regression Results for Motor-Vehicle Crash Models —Analysis Model 3

Parameter a


Type 3p-value


at 90%conf idence

level?Estimate


limit


limit


Intercept –15.01 –17.51 –12.59

Intersection

appr oachtype

STR 0.00 –0.26 0.27

0.28 NoRTL 0.21 –

0.06 0.48

CRT 0

Vol1 0.58 0.41 0.76 < .0001 YES

Vol2 0.38 0.20 0.57 0.0005 YES

Vol3 0.56 0.37 0.75 < .0001 YES



Intercept –5.16 –5.16 –5.16


STR –0.09 –0.09 –0.09

1 NoRTL –0.08 –0.08 –0.08

CRT 0

Vol1 0.19 0.19 0.19 1 No

Vol2 0.02 0.02 0.02 0.02 YES

Vol3 0.08 0.08 0.08 1 No

Dispersion –0.50 –0.50 –0.48a

Vol1, Vol2, and Vol3 ar e as def ined in Figur e 31.b






At mean CRTvolumes



STR 0.162 0.157

RTL 0.200 0.194

CRT 0.162 0.157Fatal-and-in jury motor -vehicle crashes—Analysis Model 3

STR 0.054 0.053

RTL 0.055 0.054

CRT 0.059 0.058

For Analysis Model 3, the r ight-turn treatment has no statistically signif icant effect onyear ly predictions of total or FI crashes. However, while not statistically signif icant, the year ly

74



total crash predictions tend to show that CR T and STR approaches have similar safety perfor mance (0.162 crashes per year per approach based on mean CR T volumes or 0.157 crashes per year per approach based on median CR T volumes).


Analysis Model 4 focuses on those vehicle maneuver s more likely to conf lict with the r ight-turning vehicle (Movement 3) as it merges into traff ic on the cross street (Movements 7 and 11).

Thus, Analysis Model 4 includes multi ple-vehicle crashes that involve one vehicle mak ingMovement 3 and another vehicle mak ing Movement 7 or 11. The ob jective of this analysis wasto assess whether CR T approaches exper ience more crashes than R TL or STR approaches as ther ight-turning vehicle merges with the cross street, par ticular ly since a CR T approach positionsthe dr iver at more of a skew angle at the point of the merge. The layout for this analysis model is

presented in Figure 32.

Volume statistics (mean and median) corres ponding to Analysis Model 4 are shown inTables 25, where Vol1 and Vol3 are def ined in Figure 32. Total and FI crash statistics are shownin Table 26. The selection of possi ble vehicle maneuver s for Analysis Model 4 resulted in a yetsmaller num ber of crashes than that for Analysis Model 3 (com pare to Table 22). Note that the

7-year FI crash counts are very low for this scenar io; indeed, at most one FI crash occurred atany single approach over the 7-year study per iod.

Table 25. Mean and Median Motor-Vehicle Volumesby Intersection Type—Analysis Model 4


typeNumber of

appr oaches


aMedian MV volumes

(24-hr counts)a


STR 217 1,435 10,323 1,091 8,937

RTL 95 2,274 13,220 2,169 13,697

CRT 83 1,799 12,786 1,581 13,743

aVol1 and Vol3 ar e as def ined in Figur e 32.

Table 26. Seven (7)-Year Total and FI Motor-Vehicle Crash

Counts by Intersection Approach —Analysis Model 4


type

Number of

appr oaches


Minimum Maximum Sum Minimum Maximum SumSTR 217 0 3 56 0 1 7

RTL 95 0 5 75 0 1 7

CRT 83 0 5 41 0 1 6

75




76



The regression results for the total and FI models are summar ized in Table 27. Thealgor ithm to estimate the regression coeff icients for the FI crash model did not converge,therefore no model for FI crashes for Analysis Model 4 is available. Based on this analysismodel, the r ight-turn treatment has a statistically signif icant effect on total crashes (p = 0.01).

Table 27. Regression Results for Motor-Vehicle

Crash Models —Analysis Model 4

Parameter a


Type 3p-value


at 90%conf idence

level?Estimate


limit


limit


Intercept -14.47 -17.69 -11.49


STR -0.34 -0.71 0.03

0.01 YESRTL 0.25 -0.10 0.62

CRT 0 0.00 0.00

Vol1 0.83 0.58 1.08 < .0001 YES

Vol3 0.60 0.31 0.91 0.0004 YES



No r egr ession model availablea

Vol1 and Vol3 ar e as def ined in Figur e 32.b Using the model form in Equation 1.

For Analysis Model 4, the r ight-turn treatment was statistically signif icant at the 90 percentconf idence level. In this analysis, CR T approaches have a lower estimate of total crashes thanR TL approaches, but a higher estimate than STR approaches.

Since the r ight-turn treatment for Analysis Model 4 was statistically signif icant at the90 percent conf idence level, the NB regression analysis was followed-up with a directcom par ison of CR T approaches to either STR or R TL approaches; these com par isons aresummar ized in Table 28 for total crashes. Although the overall effect of the r ight-turn treatmentwas statistically signif icant, neither one-to-one com par ison was statistically signif icant at the90 percent conf idence level (p = 0.13 for CR T vs. STR; p = 0.24 for CR T vs. R TL). The secondcolumn in Table 28 indicates whether average total crash rate for CR T approaches is lower thanthat for either of the other two approach types. Each answer is sim ply based on the com par ison

of the corres ponding parameter estimates shown in Table 27—a lower estimate corres ponds to alower crash rate (this is just a mathematical assessment, not a statistical one).

77



Table 28. Contrast Results for Motor-Vehicle


ComparisonCrash rate lower

at CRT? Chi2

p-value

Contrastsignif icant

at 90%conf idence level?


CRT vs. STR No 0.13 No

CRT vs. RTL Yes 0.24 No


No r egr ession model available

Using Equation (1) and the regression coeff icients shown in Table 27, year ly MV crashcounts were predicted for all three approach types. To account for the differences in MVvolumes among the three inter section approach types (as shown in Table 27), the three modelswere applied to all three inter section approaches with MV volumes (Vol1 and Vol3) set at theobserved mean and median volumes for CR T approaches. The predicted average crashfrequencies are shown in Table 29 for each approach type using mean and median volumes,res pectively.





At meanCRT volumes

At medianCRT volumes


STR 0.051 0.048

RTL 0.093 0.087

CRT 0.072 0.068


No r egr ession model available

For Analysis Model 4, the year ly total crash predictions for CR T approaches are lower than

the year ly total crash predictions for R TL approaches but higher than for STR approaches.


Analysis Model 5 focuses on those vehicle maneuver s that could potentially conf lict withthe r ight-turning vehicle (Movement 3) as it turns fr om the inter section approach (i.e., the

depar ture end). Thus, Analysis Model 5 includes sideswi pe and rear-end crashes that involve onevehicle mak ing Movement 3 and another vehicle mak ing Movement 1, 2, or 3. The ob jective of this analysis was to assess whether CR T approaches exper ience more crashes at the depar tureend than R TL or STR approaches. The layout for this analysis model is presented in Figure 33.

78




79



Volume statistics (mean and median) corres ponding to Analysis Model 5 are shown inTables 30, where Vol1 and Vol2 are def ined in Figure 33. Total and FI crash statistics are shownin Table 31. The selection of possi ble vehicle maneuver s for Analysis Model 5 resulted in asmall num ber of crashes. NOTE: The seven-year FI crash counts are again very low for thisscenar io; indeed, at most two FI crashes occurred at any single STR or R TL approach over theseven-year study per iod; at most one FI crashes occurred at any single CR T approach over thesame per iod.


Intersection Type—Analysis Model 5


typeNumber of

appr oaches


aMedian MV volumes

(24-hr counts)a


STR 217 1,435 10,208 1,091 9,118

RTL 95 2,274 14,507 2,169 13,845

CRT 83 1,799 12,070 1,581 12,189





typeNumber of

appr oaches



STR 217 0 7 74 0 2 12

RTL 95 0 9 50 0 2 12

CRT 83 0 6 49 0 1 8

The regression results for the total and FI models are summar ized in Table 32. Based on this

analysis model, the r ight-turn treatment has no statistically signif icant effect on total crashes(p = 0.37) or FI crashes (p = 1). Because the r ight-turn treatment for Analysis Model 5 was notstatistically signif icant at the 90 percent conf idence level, the NB regression analysis was notfollowed-up with a direct com par ison of CR T approaches to either STR or R TL approaches.

Using Equation (1) and the regression coeff icients shown in Table 32, year ly MV crashcounts were predicted for all three approach types. To account for the differences in MVvolumes among the three inter section approach types (as shown in Table 30), the three modelswere applied to all three inter section approaches with MV volumes (Vol1 and Vol2) set at the

observed mean and median volumes for CR T approaches. The predicted average crashfrequencies are shown in Table 33 for each approach type using mean and median volumes,

res pectively.

For Analysis Model 5, the r ight-turn treatment has no statistically signif icant effect on

year ly total or FI crash predictions.

80



Table 32. Regression Results for Motor-Vehicle


Parameter a


Type 3p-value


at 90%conf idence

level?Estimate


limit


limit


Intercept -13.66 -16.86 -10.65


STR -

0.32 -0.73 0.080.37 NoRTL -0.32 -0.76 0.13

CRT 0

Vol1 1.04 0.76 1.33 < .0001 YES

Vol2 0.36 0.07 0.67 0.04 YES



Intercept -4.36 -5.26 -4.35


STR -0.15 -0.15 0.27

1 NoRTL -0.06 -0.06 -0.05

CRT 0

Vol1 0.17 0.17 0.17 < .0001 YES

Vol2 -0.02 -0.02 -0.02 1 No

Dispersion -

0.50 -

0.50 -

0.46a

Vol1 and Vol2 ar e as def ined in Figur e 33.b






At mean CRTvolumes



STR 0.060 0.053

RTL 0.061 0.053

CRT 0.084 0.073

Fatal-and-in jury motor -vehicle crashes—Analysis Model 5STR 0.035 0.034

RTL 0.039 0.038

CRT 0.041 0.040

81



5.3.6 Analysis Models 6 Results

Analysis Models 6a, 6b, and 6c are similar to Analysis Model 5, except that the lef t-turnmovement (Movement 1) is excluded from the analysis. The analysis is broken into three sub-analyses in order to hone in on cer tain potential problems (rear-end crashes vs. sideswi pecrashes, crashes between a r ight-turning vehicle and a through vehicle vs. crashes between twor ight-turning vehicles, etc). The three sub-analyses are:

Analysis Model 6a focuses on rear-end and sideswi pe crashes that involve one vehicle

mak ing Movement 3 and another vehicle mak ing either Movement 2 or 3.

Analysis Model 6b focuses exclusively on rear-end crashes that involve one vehiclemak ing Movement 3 and another vehicle mak ing either Movement 2 or 3.

Analysis Model 6c focuses exclusively on rear-end crashes that involve two r ight-turning vehicles (Movement 3).

The layout for Analysis Models 6a, 6b, and 6c is presented in Figure 34.

Volume statistics (mean and median) corres ponding to Analysis Model 6 are shown in

Tables 34, where Vol1 and Vol2 are def ined in Figure 34. Total and FI crash statistics are shownin Table 35. The selection of possi ble vehicle maneuver s for Analysis Model 6 resulted in asmall num ber of crashes. NOTE: The 7 year FI crash counts are again very low for this scenar io;indeed, at most two FI crashes occurred at any single STR or R TL approach over the 7 year

study per iod; at most one FI crashes occurred at any single CR T approach over the same per iod.

Table 34. Mean and Median Motor-Vehicle Volumes

by Intersection Type—Analysis Model 6

Analysismodel


typeNumber of

appr oaches


aMedian MV volumes

(24-hr counts)a


6a or 6b

STR 217 1,435 8,841 1,091 7,965

RTL 95 2,274 12,215 2,169 11,564

CRT 83 1,799 9,407 1,581 8,691

6c

STR 217 1,435 1,091

RTL 95 2,274 2,169

CRT 83 1,799 1,581


82




83





Analysismodel


typeNumber of

appr oaches



6a(rear -endandsideswipecrashes)

STR 217 0 7 71 0 2 12

RTL 95 0 9 50 0 2 12

CRT 83 0 6 48 0 1 8

6b or 6c(rear -endcrashesonly)

STR 217 0 7 51 0 2 10

RTL 95 0 7 37 0 2 12

CRT 83 0 6 42 0 1 8

The regression results for the total and FI models are summar ized in Table 36. Based on thisanalysis model, the r ight-turn treatment has no statistically signif icant effect on t ot al crashes(p = 0.34 for Model 6a; p = 0.15 for Model 6b; p = 0.30 for Model 6c). However, the r ight-turntreatment has a statistically signif icant effect on FI crashes for all analysis models (p < = 0.001).


Models —Analysis Model 6

Parameter a


Type 3

p-value


at 90%conf idence

level?Estimate


limit


limit

Total motor -vehicle crashes—Analysis Model 6a

Intercept -13.42 -16.36 -10.69


STR -0.36 -0.77 0.06

0.34 NoRTL -0.32 -0.77 0.14

CRT 0

Vol1 1.13 0.84 1.42 < .0001 YES

Vol2 0.27 0.05 0.51 0.05 YES


Fatal-and-in jury motor -vehicle crashes—Analysis Model 6a

Intercept -17.26 -23.21 -12.11


type

STR -0.25 -1.00 0.53

< .0001 YESRTL -0.09 -0.84 0.70

CRT 0

Vol1 1.47 0.94 2.04 < .0001 YES

Vol2 0.20 -0.17 0.68 0.40 No

Dispersion -0.13 -0.50 1.46

84




Models —Analysis Model 6 (Continued)

Parameter a


Type 3p-value


at 90%conf idence

level?Estimate


limit


limit

Total motor -vehicle crashes—Analysis Model 6b

Intercept -16.28 -19.98 -12.89

Intersection

appr oachtype

STR -0.53 -1.00 -0.06

0.15 NoRTL -

0.49 -

1.00 0.03CRT 0

Vol1 1.35 1.02 1.71 < .0001 YES

Vol2 0.38 0.11 0.67 0.02 YES


Fatal-and-in jury motor -vehicle crashes—Analysis Model 6b

Intercept -17.60 -23.93 -12.16


STR -0.42 -1.21 0.40

< .0001 YESRTL -0.09 -0.84 0.71

CRT 0

Vol1 1.53 0.98 2.14 < .0001 YES

Vol2 0.18 -0.20 0.68 0.46 No

Dispersion -0.05 -0.50 1.68

Total motor -vehicle crashes—Analysis Model 6c

Intercept -13.40 -16.20 -10.81


STR –0.44 -0.90 0.03

0.30 NoRTL -0.35 -0.86 0.17

CRT 0

Vol1 1.41 1.08 1.77 < .0001 YES


Fatal-and-in jury motor -vehicle crashes—Level 6c

Intercept -16.12 -20.94 -11.79


STR -0.38 –1.17 0.43

< .0001 YESRTL -0.03 -0.76 0.76

CRT 0

Vol1 1.55 1.01 2.16 < .0001 YESDispersion -0.08 -0.50 1.59a

Vol1 and Vol2 ar e as def ined in Figur e 34.b

Using the model form in Equation 1.

Because the r ight-turn treatment for Analysis Models 6a, 6b, and 6c for FI crashes wasstatistically signif icant at the 90 percent conf idence level, the NB regression analysis wasfollowed-up with a direct com par ison of CR T approaches to either STR or R TL approaches;these com par isons are summar ized in Table 37 for FI crashes and show that:

85



Although the overall effect of the r ight-turn treatment was statistically signif icant (for FIcrashes) in Analysis Model 6a, neither one-to-one com par ison was statisticallysignif icant at the 90 percent conf idence level (p = 0.59 for CR T vs. STR; p = 0.84 for CR T vs. R TL).

Similar ly for Analysis Model 6b, neither one-to-one com par ison was statisticallysignif icant at the 90 percent conf idence level (p = 0.39 for CR T vs. STR; p = 0.85 for CR T vs. R TL).

Similar ly for Analysis Model 6c, neither one-to-one com par ison was statistically

signif icant at the 90 percent conf idence level (p = 0.43 for CR T vs. STR; p = 0.95 for CR T vs. R TL).

Table 37. Contrast Results for Motor-Vehicle

Crash Models —Analysis Models 6

ComparisonCrash rate lower

CRT? Chi2

p-value




Over all effect of intersection appr oach type is not statistically signif icant


CRT vs. S TR No 0.59 No

CRT vs. RTL No 0.84 No








Fatal-and-in jury motor -vehicle crashes—Analysis Model 6c



Using Equation (1) and the regression coeff icients shown in Table 36, year ly MV crashcounts were predicted for all three approach types for each analysis approach. To account for the

differences in MV volumes among the three inter section approach types (as shown in Table 34),the three models were applied to all three inter section approaches with MV volumes (Vol1 andVol2 or Vol1 only) set at the observed mean and median volumes for CR T approaches. The

predicted average crash frequencies are shown in Table 38 for each approach type using meanand median volumes, res pectively.

86



Table 38. Yearly Motor-Vehicle

Crash Predictions —Analysis Models 6



At mean CRTvolumes



STR 0.057 0.048

RTL 0.059 0.050

CRT 0.081 0.069


STR 0.010 0.008

RTL 0.011 0.009

CRT 0.012 0.010


STR 0.039 0.032

RTL 0.041 0.033

CRT 0.066 0.054


STR 0.008 0.006

RTL 0.011 0.009

CRT 0.012 0.010


STR 0.039 0.033

RTL 0.043 0.036

CRT 0.061 0.051

Fatal-and-in jury motor -vehicle crashes—Analysis Model 6c

STR 0.008 0.006

RTL 0.011 0.009

CRT 0.011 0.009

While the overall analysis of treatment type suggests that CR T approaches may exper iencemore rear-end and sideswi pe crashes between the r ight-turning vehicle and either (1) a throughvehicle or (2) another r ight-turning vehicle than R TL or STR approaches (Analysis Model 6a),neither of the one-to-one com par isons was statistically signif icant. This is true for AnalysisModels 6b and 6c as well.

5.3.7 Analysis Models 7 Results

Analysis Model 7 focuses on crashes between r ight-turning vehicles (Movement 3) and pedestr ians, and includes the two pedestr ian maneuver s/crossings that would potentially conf lictwith a r ight-turning vehicle. Figure 35 illustrates the r ight-turn movement and the two conf licting

87




88



pedestr ian movements of interest. Movement 13 consists of the pedestr ian crossing the approachthat the r ight-turning vehicle turns fr om (in this case, Approach 1); Movement 14 consists of the

pedestr ian crossing the approach that the r ight-turning vehicle turns int o (in this case,Approach 2).

Volume statistics (mean and median) corres ponding to Analysis Model 7 are shown inTables 39, where Vol1 and Vol3 are def ined in Figure 35. Total and FI pedestr ian crash statisticsare shown in Table 40. NOTE: The difference between total and FI pedestr ian crashes is minor,as expected, since pedestr ian crashes are almost all FI crashes.

Table 39. Mean and Median Motor-Vehicle and Pedestrian Volumes

by Intersection Type—Analysis Model 7


typeNumber of

appr oaches

Mean volumes(24-hr counts)

aMedian volumes(24-hr counts)

a

Vol1

(motor vehicle) Vol2

Vol3

(pedestrian)

Vol1

(motor vehicle) Vol2

Vol3

(pedestrian)

STR 217 1,435 2,077 1,091 1,133

RTL 95 2,274 1,120 2,169 775

CRT 83 1,799 510 1,581 277


Table 40. Seven (7)-Year Total and FI Pedestrian Counts

by Intersection Approach —Analysis Model 7Intersection

appr oachtype

Number of appr oaches



STR 217 0 4 51 0 4 49

RTL 95 0 3 45 0 3 44

CRT 83 0 2 12 0 2 11

The regression results for the total and FI models are summar ized in Table 41. Based on thisanalysis model, the r ight-turn treatment has a statistically signif icant effect on total crashes(p = 0.05) and on FI crashes (p = 0.04).

The r ight-turn treatment for Analysis Model 7 was statistically signif icant at the 90 percentconf idence level for both total and FI pedestr ian crashes (p = 0.05 and 0.04, res pectively).

Therefore, the NB regression analysis was followed-up with a direct com par ison of CR Tapproaches to either STR or R TL approaches; these com par isons are summar ized in Table 42 for total and FI pedestr ian crashes and show that although the overall effect of the r ight-turntreatment was statistically signif icant:

The com par ison between CR T and STR approaches was not statistically signif icant atthe 90 percent conf idence level (p = 0.95 for total crashes; p = 0.86 for FI crashes).

The com par ison between CR T and R TL approaches was statistically signif icant at the90 percent conf idence level (p = 0.10 for total crashes; p = 0.07 for FI crashes).

89



Table 41. Regression Results for Pedestrian Crash Models —Analysis Model 7

Parameter a


Type 3p-value


at 90%conf idence

level?Estimate


limit


limit

Total pedestrian crashes—Analysis Model 7

Intercept –12.13 –14.96 –9.49

Intersection

appr oachtype

STR 0.02 –0.55 0.64

0.05 YESRTL 0.57 0.02 1.18

CRT 0

Vol1 0.71 0.40 1.02 0.0001 YES

Vol3 0.50 0.33 0.67 < .0001 YES

Dispersion 0.31 –0.06 0.86

Fatal-and-in jury pedestrian crashes—Analysis Model 7

Intercept –11.94 –14.83 –9.25


STR 0.07 –0.53 0.71

0.04 YESRTL 0.65 0.07 1.28

CRT 0

Vol1 0.68 0.36 0.99 0.0003 YES

Vol3 0.50 0.33 0.68 < .0001 YES

Dispersion 0.38 –0.03 0.98a Vol1 and Vol3 ar e as def ined in Figur e 35.b


Table 42. Contrast Results for Pedestrian Models —Analysis Model 7

ComparisonCrash rate lower atchannelized RTL? Chi

2p-value




CRT vs. STR Yes 0.95 No

CRT vs. RTL Yes 0.10 YES


CRT vs. STR Yes 0.86 No

CRT vs. RTL Yes 0.07 YES

Using Equation (2) and the regression coeff icients shown in Table 41, year ly pedestr iancrash counts were predicted for all three approach types for each analysis approach. To accountfor the substantial differences in MV and pedestr ian volumes, res pectively, among the three

inter section approach types (as shown in Table 39), the three models were applied to all threeinter section approaches with MV and pedestr ian volumes (Vol1 and Vol3 ) set at the observedmean and median volumes for CR T approaches. The predicted average crash frequencies areshown in Table 43 for each approach type using mean and median volumes, res pectively.

90



Table 43. Yearly Pedestrian Crash




At mean CRTvolumes



STR 0.025 0.017

RTL 0.043 0.029

CRT 0.024 0.016


STR 0.024 0.016

RTL 0.043 0.029

CRT 0.023 0.015

The year ly pedestr ian crash predictions for CR T and STR approaches are similar (0.024 total crashes per CR T approach vs. 0.025 total crashes per STR approach; 0.023 FIcrashes per CR T approach vs. 0.024 FI crashes per STR approach). These com par isons also holdwhen the predictions are made based on median CR T volumes. The year ly pedestr ian crash

prediction for R TL approaches is approximately 70 to 80 percent higher than for the other two

types of r ight-turn treatments. This suggests that CR T approaches do not exper ience a par ticular safety problem with res pect to pedestr ians and, in fact, may have fewer pedestr ian safety

problems than R TL approaches.

Summary of Saf ety Analysis5.4

A cross-sectional safety analysis was conducted to deter mine the safety perfor mance of channelized r ight-turn lanes. An overall com par ison was perfor med of the safety perfor mance of the following types of inter section approaches:

Inter section approaches with channelized r ight-turn lanes (CR T approaches)Inter section approaches with conventional r ight-turn lanes (R TL approaches)Inter section approaches with shared through/r ight-turn lanes (STR approaches)

A multi-tiered analysis approach was conducted in which all movements were initially

included in the analysis, and then each subsequent analysis became more focused on thosemovements more likely to conf lict with the r ight-turn vehicle in question. In all, nine differentanalysis models were investigated; four of them are of interest and are highlighted next.

Analysis Model 3 focused on those vehicle maneuver s more likely to conf lict with ther ight-turning vehicle, either at the depar ture end of the r ight turn or as the r ight-turningvehicle merges into traff ic on the cross street. While the r ight-turn treatment had nostatistically signif icant effect on total or FI MV crashes, the year ly t ot al crash

predictions showed that CR T and STR approaches have similar safety perfor mance

91



(0.162 total crashes per year, on average, for both approach types), while the crashfrequencies for R TL approaches were slightly higher (0.2 total crashes per year, onaverage, per approach). However, this observation is not statistically signif icant.

Analysis Model 4 assessed whether CR T approaches exper ience more crashes than R TLor STR approaches as the r ight-turning vehicle merges with the cross street, par ticular lysince the CR T approach positions the dr iver at a higher skew angle at the point of themerge. Based on this analysis model, the r ight-turn treatment was statistically signif icantat the 90 percent conf idence level; CR T approaches had a lower estimate of total crash

frequency (0.072 MVcrashes per year per approach) than R TL approaches (0.093 MVcrashes per year per approach) but a higher estimate than STR approaches (0.051 MVcrashes per year per approach). While the overall effect of r ight-turn treatments wasstatistically signif icant, the com par isons between the individual r ight-turn treatment

types were not. This suggests that the three r ight-turn treatments may differ in safety perfor mance as the r ight-turning vehicle merges with the cross street, but this is notconclusively established.

Analysis Model 6 focused exclusively on rear-end and sideswi pe crashes between ther ight-turning vehicle (Movement 3) and either another r ight-turning vehicle or a throughvehicle on the same inter section approach. Based on this analysis model, the r ight-turn

treatment had no statistically signif icant effect on t ot al crashes. For FI crashes, CR Tapproaches had a slightly higher estimate than R TL or STR approaches, but thedifferences were very small —for exam ple, 0.012 vs. 0.011 vs. 0.010 FI MV crashes per year per CR T, R TL, and STR approach, res pectively. Fur ther more, when the analysis

was followed-up with direct one-to-one com par ison of CR T approaches to either STR or R TL approaches, neither of the com par isons was statistically signif icant at the90 percent conf idence level.

Analysis Model 7 focused on crashes between r ight-turning vehicles and pedestr ians,and included the two pedestr ian maneuver s/crossings that would potentially conf lict with

a r ight-turning vehicle. Based on this analysis model, the r ight-turn treatment had astatistically signif icant effect on total and FI crashes. The year ly pedestr ian crash

predictions for CR T and STR approaches were similar (0.024 total crashes per CR T

approach vs. 0.025 total crashes per STR approach; 0.023 FI crashes per CR T approachvs. 0.024 FI crashes per STR approach). The year ly pedestr ian crash prediction for R TLapproaches was approximately 70 to 80 percent higher than for the other two types of r ight-turn treatments. This suggests that CR T and STR have similar pedestr ian safety

perfor mance, while R TL approaches have more pedestr ian crashes, on average, thaneither CR T or STR approaches.

92



Chapter 6.Inter pretation of Results and Design Guidance

This chapter presents the interpretation of the research results and provides a basis for thedesign guidance presented in the guidelines in Appendix A. This guidance com pares approacheswith channelized r ight-turn lanes (CR T), shared through/r ight-turn lanes (SR T), andconventional r ight-turn lanes (R TL). These conf igurations are illustrated in Figure 36.

Figure 36. Illustration of Three R ight-Turn Treatment Types — STR, R TL, and CR T.

93



Application of Channelized Right-Tur n Lanes6.1

The research results indicate that channelized r ight-turn lanes have a def inite role inim proving operations and safety at inter sections. However, to achieve these benef its they shouldhave consistent design and traff ic control and should be used at appropr iate locations.

R esults of the observational f ield studies in Chapter 3 of this repor t suggest that, overall, pedestr ians do not have diff iculty crossing channelized r ight-turn lanes. Most pedestr ians cross

in the crosswalk and obey pedestr ian signals. Most motor ists yield to pedestr ians once they are inthe crosswalk of a channelized r ight-turn lane, either by stopping or reducing their s peed. Fewer motor ists in channelized r ight-turn lanes yield to pedestr ians waiting at the curb to cross,although such failure to yield is typical of most pedestr ian crossings and is not unique tochannelized r ight-turn lanes. When stopped in a queue, motor ists keep the crosswalk open to

pedestr ians. Avoidance maneuver s, either by a pedestr ian or a motor ist, were observed in lessthan 1 percent of the pedestr ian crossings.

The traff ic operational analysis in Chapter 4 of this repor t found that channelized r ight-turnlanes with yield control reduce r ight-turn delay to vehicles by 25 to 75 percent in com par ison tointer section approaches with conventional r ight-turn lanes. This advantage in delay reductionwas observed over the full range of vehicle and pedestr ian volumes considered. Vehicle delay

increased propor tionally faster with increasing pedestr ian crossing volume for channelized r ight-turn lanes than for conventional r ight-turn lanes, but the vehicle delay for channelized r ight-turnlanes was always lower than for conventional r ight-turn lanes.

Where a signal is provided for pedestr ians to cross a channelized r ight-turn lane on a cyclecoordinated with the pr imary signal at the inter section, vehicle delay with the channelized r ight-turn lane is, generally, greater than the vehicle delay for both conventional r ight-turn lanes andyield-controlled channelized r ight-turn lanes. For the conf iguration with a coordinated signal, thevehicle delay is independent of the pedestr ian volume because the signal phase for crossing the

channelized r ight-turn lane is called every cycle when the pr imary signal changes. This effectcan be reduced in instances where pedestr ian crossing times are dictating signal timing by

providing a pedestr ian-actuated signal for crossing the channelized r ight-turn lane; with anactuated signal, vehicle delay would be expected to increase propor tionally with pedestr iancrossing volume.

Chapter 5 of this repor t presents an analysis of total crashes, rear-end crashes, sideswi pe

crashes, and merging crashes, for three r ight-turn treatment types —channelized r ight-turn lanes,

conventional r ight-turn lanes, and shared through/r ight-turn lanes. No difference in the frequencyof motor-vehicle crashes was found between channelized r ight-turn lanes and the other r ight-turntreatments.

Pedestr ian crash frequencies for CR T and STR approaches were very similar ; pedestr iancrash frequencies for R TL approaches were approximately 70 to 80 percent higher than CR T and

R TL approaches, for average levels of pedestr ian crossing volume. Thus, the results indicate thatCR T and STR approaches have similar pedestr ian safety perfor mance, while R TL approacheshave substantially more pedestr ian crashes than either CR T or STR approaches. This is likely

94



because R TL approaches have longer pedestr ian crossing distances on the through roadway thanCR T or STR approaches with similar overall cross sections, because pedestr ians on an R TLapproach must cross not only the through lanes but also the r ight-turn lane. A fur ther advantageof CR T approaches is that they have a refuge island for pedestr ians, providing the oppor tunityfor crossing the inter section in a two-step process. In addition, those pedestr ians that cross twolegs of an inter section with a channelized r ight-turn lane can com plete their crossing without theneed to cross the channelized r ight-turn lane. This eliminates a conf lict with r ight-turning traff icthat would occur for the other inter section types.

The f inding that CR T approaches have similar pedestr ian crash frequencies to STR approaches and substantially lower pedestr ian crash frequencies than R TL approaches must beconsidered in light of current highway agency practices for using channelized r ight-turn lanes.Most highway agencies choose to provide channelized r ight-turn lanes for locations with lower

pedestr ian volumes than inter sections in general. For exam ple, at the Toronto inter sections

evaluated in Chapter 5, the average pedestr ian crossing volume was 510 pedestr ians per day for CR T approaches; 1,120 pedestr ians per day for R TL approaches; and 2,077 pedestr ians per dayfor STR approaches. While the f indings repor ted above were nor malized for pedestr ian volume,the general highway agency practice of using channelized r ight-turn lanes at inter sections withlower pedestr ian volumes suggests that this is a reasonable approach, and that caution should beexercised in using channelized r ight-turn lanes where pedestr ian crossing volumes are high. For

exam ple, most channelized r ight-turn lanes in the Toronto database had pedestr ian crossingvolumes under 1,000 pedestr ians per day.

Design Issues Related to Channelized Right-Tur n Lanes6.2

Cr osswalk Location

A ma jor ity of the sites (near ly 70 percent) evaluated in the observational studies, presented

in Chapter 3 of this repor t, had marked crosswalk s located near the center of the channelizedr ight-turn lane; only about 30 percent of crosswalk s were located at the upstream or downstreamend of the channelized r ight-turn lane. The highway agency survey conducted in NCHRPPro ject 3-72 (3) found that highway agencies prefer a crosswalk location near the center of achannelized r ight-turn lane; over 70 percent of highway agencies repor ted in the survey that their

practice was to place crosswalk s near the center of channelized r ight-turn lanes.

There has been little research that evaluates how the crosswalk location affects crossings by

pedestr ians with vision im pair ment in ter ms of their ability to identify the appropr iate time tocross or eff iciently locate the crosswalk. While research would be desirable to provide moreconcrete recommendations, or ientation and mobility (O&M) s pecialists, who teach pedestr ians

with vision im pair ment how to better traver se inter sections, recommend that consistency of crosswalk location is im por tant to pedestr ians with vision im pair ment. Such consistency wouldmake it easier for O&M s pecialists to descr i be a typical channelized r ight-turn lane to

pedestr ians with vision im pair ment and teach procedures for crossing it.

95



There is no strong technical basis for recommending one crosswalk location over another,other than the need for consistency. However, consistency in locating crosswalk s is im por tant,es pecially to pedestr ians with vision im pair ment, and current practice shows a clear preferencefor crosswalk locations near the center of a channelized r ight-turn lane. And, a crosswalk location at the center of the channelized r ight-turn lane moves vehicle-pedestr ian conf licts awayfrom both the diverge maneuver at the upstream end of the channelized r ight-turn lane and,es pecially, from the merge maneuver at the downstream end of the channelized r ight-turn lane.The only potential exception to a center crosswalk location for channelized r ight-turn lanes isthat, where STOP sign or traff ic signal control is provided on the channelized r ight-turn lane, the

crosswalk should be located beyond the stop line. In addition, at locations where the channelizedr ight-turn lane inter sects the cross street at near ly a r ight angle, the stop line and crosswalk may

be better placed at the downstream end of the channelized r ight-turn lane, depending on theisland size and location of sidewalk approaches. To summar ize the recommended guidance for the placement of crosswalk s at channelized r ight-turn lanes:

Where the entry to the cross street at the downstream end of the channelized r ight-turn

lane has yield control or no control, place the crosswalk near the center of thechannelized r ight-turn lane.

Where the channelized r ight-turn lane has STOP sign control or traff ic signal control, place the crosswalk immediately downstream of the stop bar, where possi ble. Where thechannelized r ight-turn roadway inter sects with the cross street at near ly a r ight angle, the

stop bar and crosswalk can be placed at the downstream end of the channelized r ight-turn roadway.

Special Cr osswalk Signing and Marking

Seven of the observational f ield study sites (see Chapter 3 of this repor t) included s pecialcrosswalk signing and mark ing treatments. An infor mal com par ison of pedestr ian and motor ist

behavior was made between the sites with s pecial crosswalk signing and mark ing and the other observational f ield study sites. The com par ison suggested that the additional signage and

pedestr ian crosswalk treatments may im prove the motor ist yield behavior and pedestr ian use of

the crosswalk. For exam ple, the yield behavior of motor ists yielding to pedestr ians waiting at thecurb was slightly better (47 percent vs. 40 percent) at sites with s pecial crosswalk treatments.This suggests that additional em phasis on signing or other treatments may be needed to increaseyielding for pedestr ians waiting at the curb.

Enhanced features of crosswalk signing and mark ing that are desirable include:

Use of a raised crosswalk to im prove visi bility of crosswalk for motor ists and to better

def ine crosswalk boundar ies for pedestr ians (raised crosswalk s are par ticular ly hel pful to pedestr ians with vision im pair ment).

Addition of f luorescent yellow-green signs both at the crosswalk and in advance of thecrossing location (to supplement the high-visi bility mark ings).

96



Use of a real-time warning device to indicate to the motor ist when a pedestr ian is presentin the area (may be activated via passive detection technologies such as microwave or infrared or via traditional methods such as push buttons).

Use of dynamic message signs (for real-time or static warning messages to motor ists).

Island Type

The channelizing island that def ines a channelized r ight-turn lane, par ticular ly when

bounded by raised curbs, serves as a refuge area for pedestr ians, and im proves safety andaccessi bility by allowing pedestr ians to cross the street in two stages. O&M s pecialists have astrong preference for raised islands with “cut-through” pedestr ian paths, which provide better guidance and infor mation about the location of the island for pedestr ians with vision im pair mentthan painted islands.

Radius of Tur ning Roadway

The traff ic operational analysis in Chapter 4 of this repor t found that increasing the radius of a channelized r ight-turn roadway reduces r ight-turn delay by approximately 10 to 20 percent for each 8-k m/h (5-mi/h) increase in turning s peed. Larger delay reductions generally occur whenconf licting traff ic volumes on the cross street are lower. However, vehicle delay should not bethe only consideration when selecting an appropr iate radius. Larger radii generally encouragehigher turning s peeds, and previous research (15) has shown that higher s peeds can result in a

decrease in yielding to pedestr ians by motor ists. Thus, smaller radii may be more appropr iate atlocations where pedestr ians are antici pated.

Angle of Intersection with Cr oss Street

The alignment of a channelized r ight-turn lane and the angle between the channelized r ight-turn roadway and the cross street can be designed in two different ways (as illustrated in Figure 9in Chapter 2):

A f lat-angle entry to the cross streetA near ly-r ight-angle entry to the cross street

The two designs shown in Figure 9 differ in the shape of the island that creates the

channelized r ight-turn lane. The f lat-angle entry design has an island that is typically shaped likean equilateral tr iangle (of ten with one curved side), while the near ly-r ight-angle design istypically shaped like an isosceles tr iangle. The f lat-angle entry design is appropr iate for use inchannelized r ight-turn lanes with either yield control or no control for vehicles at the entry to thecross street. The near ly-r ight-angle entry design can be used with STOP sign control or traff icsignal control for vehicles at the entry to the cross street; yield control can also be used with thisdesign where the angle of entry and sight distance along the cross street are appropr iate.

97



Acceleration Lanes

Acceleration lanes at the downstream end of a channelized r ight-turn lane provide anoppor tunity for vehicles to com plete the r ight-turn maneuver unim peded and then accelerate

parallel to the cross-street traff ic pr ior to merging. The traff ic operational analysis in Chapter 4of this repor t found that the addition of an acceleration lane can reduce the r ight-turn delay by65 to 85 percent, depending on the conf licting through traff ic volume, and may be consideredwhere r ight-turn delay is a par ticular problem. Acceleration lanes are only desirable where thereare no dr iveways or other access points on the cross street within or close to the acceleration

lane. In addition, channelized r ight-turn lanes with acceleration lanes at their downstream endappear to be very diff icult for pedestr ians with vision im pair ment to cross. Therefore, the use of acceleration lanes at the downstream end of a channelized r ight-turn lane should generally be

reserved for locations where no pedestr ians or very few pedestr ians are present. Typically, thesewould be locations without sidewalk s or pedestr ian crossings; at such locations, the reduction invehicle delay resulting from addition of an acceleration lane becomes very desirable.

Traff ic Contr ol

Channelized r ight-turn lanes with signals tied to the signal cycle of the pr imary inter sectionconsistently exper ience more delay to r ight-turning vehicles than yield-controlled channelizedr ight-turn lanes. If signalization of the r ight-turn movement is used, a conventional r ight-turnlane conf iguration (including r ight-turn on red) generally perfor ms better than signalization of the channelized r ight-turn lane. However, at signalized channelized r ight-turn lanes, it is

common for traff ic engineer s to provide extra green time to the r ight-turn movement withoutincreasing cycle length by over lapping the r ight turn with the cross street lef t turn. Figure 37illustrates the concept of a r ight-turn over lap. A potential drawback of the r ight-turn over lap

phasing is that U-turns cannot be per mitted from the cross-street lef t-turn lane, as they may

conf lict with r ight-turning vehicles. Thus, use of an over lap phase, or other method of providingadditional green time to r ight-turning vehicles, can substantially reduce the delay for a signalizedchannelized r ight-turn lane, but may result in other im pacts to inter section operations, such as the

need to restr ict U-turn maneuver s.

Where a signal is provided for pedestr ians to cross a channelized r ight-turn lane, a pedestr ian-actuated signal should be considered. This can reduce vehicle delay because the phasefor crossing the channelized r ight-turn lane is called only when pedestr ians are present.

98




99



Chapter 7.Conclusions and Recommendations

This chapter presents the conclusions and recommendations based on the results of theresearch.

Conclusions7.1

The following conclusions are based on f ield observational studies of vehicle and pedestr ianinteractions at channelized r ight-turn lanes, interviews with or ientation and mobility (O&M)s pecialists, traff ic operational analyses of channelized r ight-turn lanes conducted with VISSIM,

and safety analyses of channelized r ight-turn lanes:

A ma jor ity of the sites (near ly 70 percent) had marked crosswalk s located near the center

of the channelized r ight-turn lane; only about 30 percent of crosswalk s were located atthe upstream or downstream end of the channelized r ight-turn lane. The highway agencysurvey conducted in NCHRP Pro ject 3-72 (3) found that highway agencies prefer acrosswalk location near the center of a channelized r ight-turn lane; over 70 percent of highway agencies repor ted in the survey that their practice was to place crosswalk s near the center of channelized r ight-turn lanes.

Pedestr ians did not appear to have any par ticular diff iculty crossing channelized r ight-turn lanes. Most pedestr ians crossed in the crosswalk (75 percent) and, where pedestr iansignals were present, most pedestr ians crossed dur ing the pedestr ian crossing phase

(72 percent). The pedestr ians who crossed outside the crosswalk generally did so whenno vehicular traff ic was present and this behavior did not cause any traff ic conf licts.Avoidance maneuver s by a pedestr ian or motor ist appear to be relatively rare, and wereobserved in less than 1 percent of the pedestr ian crossings.

Near ly all motor ists (over 96 percent) yielded to pedestr ians in the crosswalk, or wereunaffected by the pedestr ian crossing (i.e., the pedestr ian crossed without substantially

im pacting the vehicles s peed). The observed yield rate was somewhat lower (approximately 40 percent) for pedestr ians waiting at the curb. The failure of vehicles toyield to pedestr ians waiting to cross at a marked crosswalk is not unique to channelizedr ight-turn lanes, but is a general problem at pedestr ian crosswalk s. The yield behavior of motor ists was slightly better (47 percent vs. 40 percent) at sites with s pecial crosswalk

treatments (e.g., raised crosswalk, pavement mark ings, signing). This may indicate thatadditional em phasis on signing or other treatments may be needed to increase yieldingfor pedestr ians waiting at the curb.

There has been little research that evaluates how the crosswalk location affects crossings

by pedestr ians with vision im pair ment in ter ms of their ability to identify the appropr iatetime to cross of eff iciently locate the crosswalk. O&M s pecialists recommend thatconsistency of crosswalk location and traff ic control is im por tant to pedestr ians with

vision im pair ment. They also have a strong preference for raised islands with “cut-

100



through” pedestr ian paths, which provide better guidance and infor mation for pedestr ians with vision im pair ment than do painted islands.

Channelized r ight-turn lanes with yield control were shown to reduce r ight-turn delay tovehicles by 25 to 75 percent in com par ison to inter section approaches with conventionalr ight-turn lanes. High pedestr ian volumes increase r ight-turn delay by approximately60 percent on a yield-controlled channelized r ight-turn lane.

The addition of an acceleration lane at the downstream end of a channelized r ight-turnlane can substantially reduce the r ight-turn delay (by 65 to 85 percent) depending on the

conf licting traff ic volume on the cross street. However, channelized r ight-turn lanes withacceleration lanes appear to be very diff icult for pedestr ians with vision im pair ment tocross due to vehicle s peeds and lack of yielding by motor ists.

Increasing the radius of a channelized r ight-turn roadway can reduce r ight-turn delay byapproximately 10 to 20 percent for each 8-k m/h (5-mi/h) increase in turning s peed.

However, in previous research (15), higher s peeds have been shown to result in adecrease in yielding to pedestr ians by motor ists.

For channelized r ight-turn lanes with signal control, use of an over lap phase, or other method of providing additional green time to r ight-turning vehicles, can substantiallyreduce the delay for a signalized channelized r ight-turn lane. However, this may result inother im pacts to inter section operations, such as restr icting U-turn maneuver s.

Inter section approaches with channelized r ight-turn lanes appear to have similar motor-vehicle safety perfor mance as approaches with conventional r ight-turn lanes or shared

through/r ight-turn lanes. This was found to be the case both at the downstream end of the channelized r ight-turn lane (where the r ight-turning vehicle merges with the crossstreet) as well as at the upstream end of the channelized r ight-turn lane (where the r ight-

turning vehicle begins the r ight-turn maneuver).

Inter section approaches with channelized r ight-turn lanes appear to have similar

pedestr ian safety perfor mance as approaches with shared through/r ight-turn lanes.Inter section approaches with conventional r ight-turn lanes have substantially more

pedestr ian crashes (approximately 70 to 80 percent more) than approaches withchannelized r ight-turn lanes or shared/through r ight-turn lanes.

Recommendations7.2

The following recommendations were developed in the research:

Channelized r ight-turn lanes have a def inite role in im proving operations and safety atinter sections. However, to achieve these benef its, they should have consistent design and

traff ic control and should be used at appropr iate locations.

There has been little research that evaluates how the crosswalk location affects crossings by pedestr ians with vision im pair ment, and more research would be desirable to provide

more concrete recommendations.

101



Since consistency in locating crosswalk s is im por tant and since current practice shows aclear preference for crosswalk locations near the center of a channelized r ight-turn lane,design guidance should recommend placing crosswalk s near the center of thechannelized r ight-turn lane for channelized r ight-turn lanes with yield control or nocontrol at the entry to the cross street.

Where the channelized r ight-turn lane has STOP sign control or traff ic signal control, thecrosswalk should be placed immediately downstream of the stop bar. If the channelizedr ight-turn roadway inter sects with the cross street at near ly a r ight angle, the stop bar and

crosswalk can be placed at the downstream end of the channelized r ight-turn roadway.R aised islands should be considered because they serve as a refuge area so that

pedestr ians may cross the street in two stages. R aised islands with “cut-through” also

provide better guidance for pedestr ians with vision im pair ment than painted islands.

Channelized r ight-turn lanes with acceleration lanes appear to be very diff icult for

pedestr ians with vision im pair ment to cross. Therefore, the use of acceleration lanes atthe downstream end of a channelized r ight-turn lane should generally be reserved for locations where no pedestr ians or very few pedestr ians are present. Typically, these

would be locations without sidewalk s or pedestr ian crossings; at such locations, thereduction in vehicle delay resulting from addition of an acceleration lane becomes verydesirable.

Where a signal is provided for pedestr ians to cross a channelized r ight-turn lane, a pedestr ian-actuated signal should be considered. This can reduce vehicle delay because

the phase for crossing the channelized r ight-turn lane is called only when pedestr ians are present.

Use of an over lap phase (or other method of providing additional green time to r ight-

turning vehicles) should be considered as it can substantially reduce delay at channelizedr ight-turn lanes with signals tied to the signal cycle of the pr imary inter section.However, consideration should be given to other possi ble im pacts to inter sectionoperations, such as restr icting U-turn maneuver s.

The general highway agency practice of using channelized r ight-turn lanes at inter sectionswith lower pedestr ian volumes suggests that this is a reasonable approach, and that caution

should be exercised in using channelized r ight-turn lanes where pedestr ian crossing volumes arehigh (e.g., over 1,000 pedestr ians per day). This guidance is based on the 85th percentile

pedestr ian volume—1,000 pedestr ians per day— in the Toronto inter section database that wasused for the safety evaluation. While channelized r ight-turn lanes may be suitable for higher

pedestr ian volumes, the research cannot predict the safety perfor mance of channelized r ight-turnlanes with pedestr ian volumes beyond the range evaluated in the research.

102



Chapter 8.Ref erences

1. Amer ican Association of State Highway and Trans por tation Off icials, A Policy on Geometric

De si gn o f H i ghway s and S tr eets, 2004.

2. U.S. Access Board, R evised Draf t Guidelines for Accessi ble Public Rights of Way, Novem ber 23, 2005, www.access-board.gov/ prowac/draf t.htm.

3. Potts, I. B., D. W. Harwood, D. J. Torbic, D. K. S. A. Hennum, C. B. Tiesler, J. D. Zegeer, J.F. Ringer t, D. L. Harkey, and J. M. Bar low, Synt he sis on C hanneliz ed Ri ght T ur n s on U r ban

and Subur ban Art erials, Final R epor t, NCHRP Pro ject 3-72, Midwest R esearch Institute,August 2006.

4. Dixon, K. K., J. L. Hi bbard, and H. Nyman, “Right-Turn Treatment for SignalizedInter sections,” Urban Street Sym posium, Tr an s port ation Re sear ch Cir cul ar E-C 019, 1999.

5. Fitzpatr ick, K., W. H. Schneider, and E. S. Park, “Operations and Safety of Right-Turn-LaneDesigns,” Tr an s port ation Re sear ch Recor d: Jour nal o f t he Tr an s port ation Re sear ch Boar d,

N o. 1961, Trans por tation R esearch Board of the National Academies, Washington, D.C.,2006.

6. Staplin, L., D. L. Harkey, K. H. Lococo, and M. S. Tarawneh, Int ersection Geometric De si gn

and Oper ational Guideline s f or Ol der Drivers and Pede strian s , Vol ume I: F inal Report ,

R epor t No. FHWA-R D-96-132, Federal Highway Administration, June 1997.

7. Tarawneh, M. S. and P. T. McCoy, “Effect of Inter section Channelization and Skew onDr iver Perfor mance,” Tr an s port ation Re sear ch Recor d 1523, TRB, National R esearchCouncil, Washington, D.C., 1996.

8. McCoy, P. T., R . R Bishu, J. A. Bonneson, J. W. Fitts, M. D. Fowler, S. L. Gaber, M. E.Lutjehar ms, B. A. Moen, and D. L. Sick ing, Guideline s f or F r ee Ri ght -T ur n Lane s at Un si gnaliz ed Int ersection s on Rur al Two- Lane H i ghway s— F inal Report , R epor t No. TRP-02-32-95, Univer sity of Nebraska-Lincoln, Nebraska Depar tment of R oads, March 1995.

9. Abdel-Aty, M. and P. Nawathe, “A Novel Approach for Signalized Inter section CrashClassif ication and Prediction,” Advances in Trans por tation Studies, Vol. 9, 2006.

10. Hunter, W. W., J. C. Stutts, W. E. Pein, and C. L. Cox, Pede strian and Bicycl e Cr a sh T ype s

o f t he Earl y 1990’ s, R epor t No. FHWA-R D-95-163, Federal Highway Administration,June 1996.

11. N ort h C ar olina Bicycl e and Pede strian Cr a she s (www.pedbikeinfo.org/ pbcat), Division of Pedestr ian and Bicycle Nor th Carolina Depar tment of Trans por tation, R aleigh, Nor th

Carolina, Novem ber 2003.

12. Harkey, D. L., J. Mekemson, M. C. Chen, and K. Krull, Pede strian and Bicycl e Cr a sh

Anal y sis T ool , Product No. FHWA-R D-99-192, Federal Highway Administration,Decem ber 1999.

13. Oregon Depar tment of Trans por tation, Or egon Bicycl e and Pede strian P l an, 1995.

103



14. Zegeer, C. V., C. Seider man, P. Lagerwey, M. Cyneck i, M. R onk in, and B. Schneider, Pede strian Facilitie s U sers Guide: P r ovid ing Sa f et y and Mobilit y, R epor t No. FHWA-R D-01-102, Federal Highway Administration, March 2002.

15. Geruschat, D. R . and S. E. Hassan, “Dr iver Behavior in Yielding to Sighted and BlindPedestr ians at R oundabouts,” Jour nal o f V isual Impair ment & Blindne ss, Vol. 99, No. 5,2005.

16. Schroeder, B. J., N. M. R ouphail, and R . S. W. Emer son, “Exploratory Analysis of CrossingDiff iculties for Blind and Sighted Pedestr ians at Channelized Turn Lanes,” Tr an s port ation

Re sear ch Recor d: Jour nal o f t he Tr an s port ation Re sear ch Boar d, N o. 1956 , Trans por tationR esearch Board of the National Academies, Washington, D.C., 2006.

17. Guth, D., D. Ashmead, R . Long, R . Wall, and P. Ponchillia. “Blind and Sighted Pedestr iansJudgments of Gaps in Traff ic at R oundabouts,” Human Fact ors, Volume 47, 2005.

18. Ashmead, D., D. Guth, R . Wall, R . Long, and P. Ponchillia. “Street Crossing by Sighted andBlind Pedestr ians at a Modern R oundabout,” AS C E Jour nal o f Tr an s port ation Eng ineering ,Vol. 131, No. 11, Novem ber 2005.

19. Inman, V. W., G. W. Davis, and D. Sauerburger, Pede strian Acce ss t o Roundabouts: A sse ssment o f Mot orists’ Yiel d ing t o V isuall y Impair ed Pede strian s and Pot ential Tr eat ments

t o Impr ove Acce ss, Federal Highway Administration, FHWA-HR T-05-080, McLean,Virginia, 2006.

20. Schroeder, B., R . Hughes, N. R ouphail, C. Cunningham, K. Salamati, R . Long, D. Guth,R .W. Emer son, D. K im, J. Bar low, B.L. (Beezy) Bentzen, L. R odegerdts, and E. Myer s, NC HRP Report 674: Cr o ssing Sol ution s at Roundabouts and C hanneliz ed T ur n Lane s f or

Pede strian s wit h V ision Disabilitie s, Trans por tation R esearch Board of the NationalAcademies, Washington, D.C., January 2011.

21. U.S. Depar tment of Trans por tation, National Highway Traff ic Safety Administration, Tr a ffic

Sa f et y Facts 2002, Pedal cyclists , R epor t No. DOT HS 809 613, National Highway Traff icSafety Administration, 2003.

22. Tan, C., Cr a sh-T ype Manual f or Bicyclists, R epor t No. FHWA-R D-96-104, Federal HighwayAdministration, U.S. Depar tment of Trans por tation, 1996.

23. Clarke, A. and L. Tracy, Bicycl e Sa f et y-Rel at ed Re sear ch Synt he sis, R epor t No. FHWA-94-062, Federal Highway Administration, 1995.

24. U.S. Depar tment of Trans por tation, Federal Highway Administration, Manual on Unif or m

Tr a ffic C ontr ol Device s f or H i ghway s and S tr eets, Washington, D.C., 2009.

25. Amer ican Association of State Highway and Trans por tation Off icials (AASHTO), Guide f or t he Devel opment o f Bicycl e Facilitie s , Washington, D.C., 1999.

26. Hunter, W. W., D. L. Harkey, J. R . Stewar t, and M. L. Birk, Eval uation o f t he Bl ue Bike

Lane Tr eat ment U sed in Bicycl e / Mot or Vehicl e C on flict Ar ea s in Portl and, Or egon, R epor t No. FHWA-R D-00-150, Federal Highway Administration, August 2000.

27. Hunter, W. W., Eval uation o f a C ombined Bicycl e Lane / Ri ght -T ur n Lane in Eugene, Or egon,Publication No. FHWA-R D-00-151, Federal Highway Administration, August 2000.

104



28. Harkey, D. L., D. W. R einfur t, M. Knuiman, J. R . Stewar t, and A. Sor ton, Devel opment o f

t he Bicycl e C ompatibilit y Index: A Level o f Ser vice C oncept , R epor t No. FHWA-R D-98-072,Federal Highway Administration, Decem ber 1998.

29. Neuman, T. R ., NC HRP Report 279: Int ersection C hanneliz ation De si gn Guide, TRB, National R esearch Council, Washington, D.C., 1985.

30. Conver sation with Mr. Jeffrey Bagdade, Opus International Consultants, at a meeting of the National Committee on Unifor m Traff ic Control Devices, Mobile, Alabama, June 2008.

31. H i ghway C apacit y Manual , TRB, National R esearch Council, Washington, D.C., 2000.32. Harwood, D. W., D. J. Torbic, D. K. Gilmore, C. D. Bokenkroger, J. M. Dunn, C. V. Zegeer,

R . Sr inivasan, D. Car ter, C. R aborn, C. Lyon, and B. Per saud, NC HRP Web-Onl y Document

129: Pha se III Pede strian Sa f et y P r ed iction Met hodol ogy. Trans por tation R esearch Board of the National Academies, Washington, D.C., 2008.

33. Amer ican Association of State Highway and Trans por tation Off icials, H i ghway Sa f et y

Manual , 2010.

105



Appendix A

Design Guide for Channelized Right-Tur n Lanes



Chapter 1Intr oduction

Channelized r ight-turn lanes are turning roadways at inter sections that provide for free-f lowor near ly free-f low r ight-turn movements. Channelization can be provided in a var iety of for msincluding painted pavement areas and curbed islands. Figure 1 illustrates an inter section withchannelized r ight-turn lanes. While the f igure shows channelized r ight-turn lanes in all quadrantsof the inter section, channelized r ight-turn lanes may be appropr iate in some quadrants, but not in

other s, depending on inter section geometry and traff ic demands.

Figure 1. Typical Intersection With Channelized R ight-Turn Lanes

The pr imary reasons for providing a channelized r ight-turn lane are (1):

To increase vehicular capacity at inter sections

To reduce delay to dr iver s by allowing them to turn at higher s peeds

To reduce unnecessary stops

To clear ly def ine the appropr iate path for r ight-turn maneuver s at skewed inter sectionsor at inter sections with high r ight-turn volumes

To im prove safety by separating the points at which crossing conf licts and r ight-turnmerge conf licts occur

To per mit the use of large curb return radii to accommodate turning vehicles, including

large truck s, without unnecessar ily increasing the inter section pavement area and the pedestr ian crossing distance

Channelized

right-turn roadway

Channelizing

island

Channelized

roadway width

Radius

Channelized

right-turn roadway

Channelizing

island

Channelized

roadway width

Radius

A-1



Chapter 2Consideration of User Types

Many trans por tation agencies use channelized r ight-turn lanes to im prove operations atinter sections, although there may be advantages and disadvantages of channelized r ight-turnlanes for different roadway user types including motor vehicles, pedestr ians, and bicyclists.

Motor Vehicles

The pr imary traff ic operational reasons for providing channelized r ight-turn lanes are toincrease vehicular capacity at an inter section, reduce delay to dr iver s by allowing them to turn athigher s peeds, and reduce unnecessary stops. Channelized r ight-turn lanes appear to provide a

net reduction in motor vehicle delay at inter sections where they are installed. A yield-controlledchannelized r ight-turn lane can reduce r ight-turn delay by 25 to 75 percent in com par ison tointer section approaches with conventional r ight-turn lanes (2). This advantage in delay reductionwould like be observed at channelized r ight-turn lanes with no traff ic control on the r ight-turnroadway.

It is generally accepted that channelized r ight-turn lanes im prove safety for motor vehicles atinter sections where they are used, but there have been concerns that channelized r ight-turn lanesmay exper ience a higher propor tion of rear-end crashes than other r ight-turn treatments. R ecentresearch (2) assessed whether channelized r ight-turn lanes exper ience more crashes than

inter section approaches with conventional r ight-turn lanes or shared through/r ight-turn lanes.R esearch results found that any difference in motor-vehicle safety perfor mance between the threer ight-turn treatments was not conclusively established.

At inter sections with substantial truck volumes, channelized r ight-turn lanes per mit the useof large curb return radii to accommodate large truck s, without unnecessar ily increasing theinter section pavement area and the pedestr ian crossing distance.

Pedestrians

A key concern with channelized r ight-turn lanes has been the extent of conf licts between

vehicles and pedestr ians that occur at the point where pedestr ians cross the r ight-turn roadway.

Conf licts with pedestr ians may be likely at r ight-turn roadways because the dr iver’s attentionmay be focused on the cross-street traff ic or the placement of pedestr ian crosswalk s or pedestr iansignals on channelized r ight-turn lanes may violate dr iver expectancy. For most pedestr ians,

crossing the channelized r ight-turn roadway may be a relatively easy task because such roadwaysare not very wide and because traff ic is approaching from a single direction. However,

pedestr ians with vision im pair ment may have diff iculty detecting approaching traff ic because(a) r ight-turning vehicles are traveling a curved rather than a straight path; (b) there is not asystematic stopping and star ting of traff ic, as there would be at a conventional signal- or stop-

A-2



controlled inter section; and (c) the traff ic sounds from the ma jor streets may mask the sound of traff ic on the r ight-turn roadway.

Des pite these potential problems, channelized r ight-turn lanes also provide advantages for pedestr ians. The provision of a channelized r ight-turn lane, while mak ing it necessary for pedestr ians to cross two roadways, of ten reduces the pedestr ian crossing distance. Fur ther morethe channelizing island, par ticular ly when bounded by raised curbs, serves as a refuge area for

pedestr ians, and im proves safety by allowing pedestr ians to cross the street in two stages.

Bicyclists

There may be an inherent r isk to bicyclists at channelized r ight-turn lanes because motor

vehicles enter ing the channelized r ight-turn roadway must weave across the path of bicyclestraveling straight through the inter section, but no studies based on crash history are available tosuppor t this presum ption. However, similar conf licts between through bicyclists and r ight-turn

vehicles are present at conventional inter sections as well.

A-3



Chapter 3Traff ic Operational and Saf ety Considerations for Channelized Right-Tur n Lanes

Channelized r ight-turn lanes have a def inite role in im proving traff ic operations and safety at

inter sections. S pecif ic traff ic operational and safety considerations related to channelized r ight-turn lanes are discussed below.

Traff ic Operational Considerations

Channelized r ight-turn lanes can play a key role in reducing r ight-turn delay to vehicles.Channelized r ight-turn roadways may have no control or may be controlled by stop signs, yield

signs, or signals. In com par ison to inter section approaches with conventional r ight-turn lanes,channelized r ight-turn lanes with yiel d control can reduce r ight-turn delay to vehicles by 25 to75 percent (2). This advantage in delay reduction has been observed for pedestr ian crossingvolumes up to 200 ped/h. Similar delay reductions would likely be observed for channelized

r ight-turn lanes with no traff ic control.

Right-turn delay at si gnaliz ed channelized r ight-turn lanes is generally higher than at yield-controlled channelized r ight-turn lanes, and is t ypically higher than or similar to the r ight-turn

delay at conventional r ight-turn lanes at signalized inter sections where r ight-turn-on-red (R TOR )operation is per mitted. At signalized channelized r ight-turn lanes, it is common for traff icengineer s to provide extra green time to the r ight-turn movement without increasing cycle length

by over lapping the r ight turn with the cross street lef t turn. Figure 2 illustrates the concept of ar ight-turn over lap. A potential drawback of the r ight-turn over lap phasing is that U-turns cannot

be per mitted from the cross-street lef t-turn lane, as they may conf lict with r ight-turning vehicles.

The elimination of the U-turn can be a signif icant issue in states where U-turns are allowed at allinter sections and in areas that have median access control and U-turns are the pr imary means of

providing lef t-turn access to businesses. Thus, use of an over lap phase or other methods of providing additional green time to r ight-turning vehicles can substantially reduce the delay for a

signalized channelized r ight-turn lane, but may result in other im pacts to inter section operations.

Where a signal is provided for pedestr ians to cross a channelized r ight-turn lane on a cyclecoordinated with the pr imary signal at the inter section, vehicle delay with the channelized r ight-turn lane is generally greater than vehicle delay for conventional r ight-turn lanes and for yield-controlled channelized r ight-turn lanes. The vehicle delay is independent of the pedestr ianvolume because the phase for crossing the channelized r ight-turn lane can be called every cycle

in con junction with the cross-street signal phase. A disadvantage of providing a pedestr ian signalis that a pedestr ian must wait for the pedestr ian crossing phase to cross the channelized r ight-turnlane and then wait again for the pedestr ian crossing phase of the pr imary inter section to cross the

ma jor or cross street. However, pedestr ian signals are very im por tant to pedestr ians with visionim pair ment who cross at a channelized r ight-turn lane. Provision of a pedestr ian-actuated signalshould be considered.

A-4




Traff ic Saf ety Issues

It is generally accepted that channelized r ight-turn lanes im prove safety for motor vehicles, but there have been concerns about the safety of channelized r ight-turn lanes for pedestr ians.R ecent research (2) evaluated seven year s of motor-vehicle and pedestr ian crash and volume datafrom near ly 400 four-leg signalized inter section approaches in Toronto, Ontar io, Canada. TheToronto data represented a unique resource because they included both vehicle turning-

movement volumes and pedestr ian crossing volumes by inter section approach, as well as crashdata that could be classif ied by inter section approach and turning movement. The results of thisresearch provide insights into the safety of motor vehicles and pedestr ians at channelized r ight-turn lanes.

Potts et al. (2) assessed whether channelized r ight-turn lanes exper ience more motor-vehiclecrashes than other r ight-turn treatments, such as conventional r ight-turn lanes or shared

through/r ight-turn lanes. The research results found that any difference between the three r ight-

turn treatment types was not conclusively established. Concerns have been raised thatchannelized r ight-turn lanes may exper ience a higher propor tion of rear-end crashes, par ticular lyat the downstream end of a channelized r ight-turn lane as vehicles merge into cross-street traff ic.

The same research (2) com pared rear-end crash exper ience at channelized r ight-turn lanes to thatof conventional r ight-turn lanes and shared through/r ight-turn lanes and found no difference

between the r ight-turn treatments that could be conclusively established.

A key concern with channelized r ight-turn lanes has been pedestr ian safety as pedestr ianscross the r ight-turn roadway. S pecif ically, there have been concerns that the dr iver’s attention

A-5



may be focused on cross-street traff ic and that the placement of pedestr ian crosswalk s or pedestr ian signals on channelized r ight-turn lanes may violate dr iver expectancy. R esearchresults (2) showed that channelized r ight-turn lane approaches and approaches with sharedthrough/r ight-turn lanes have similar pedestr ian crash frequencies, while conventional r ight-turnlane approaches have substantially more (70 to 80 percent) pedestr ian crashes than either of theother two r ight-turn treatments, for average levels of pedestr ian crossing volumes. This is likely

because conventional r ight-turn lane approaches have longer pedestr ian crossing distances. Afur ther advantage of channelized r ight-turn lanes is that they have a refuge island for pedestr ians,

providing the oppor tunity for crossing the inter section in two stages. In addition, those

pedestr ians that cross two legs of an inter section with a channelized r ight-turn lane can com pletetheir crossing without the need to cross the channelized r ight-turn lane. This eliminates a conf lictwith r ight-turning traff ic that would occur with the other inter section types.

Most highway agencies choose to provide channelized r ight-turn lanes for locations with

lower pedestr ian volumes than inter sections in general. For exam ple, at the inter sectionconsidered in a safety evaluation by Potts et al. (2), the average pedestr ian crossing volume was510 pedestr ians per day for channelized r ight-turn lane approaches; 1,120 pedestr ians per day for conventional r ight-turn lane approaches; and 2,077 pedestr ians per day for shared through/r ight-turn lane approaches. While the f indings repor ted by Potts et al. were nor malized for pedestr ianvolume, the general highway agency practice of using channelized r ight-turn lanes at

inter sections with lower pedestr ian volumes suggests that this is a reasonable approach. Thef indings (2) are based on a database with an 85th percentile pedestr ian volume of 1,000

pedestr ians per day. Therefore, caution should be exercised in using channelized r ight-turn lanes

where pedestr ian crossing volumes are high (i.e., greater than 1,000 ped/day).

For most pedestr ians, crossing the channelized r ight-turn roadway can be a relatively easytask because such roadways are not very wide and because traff ic is approaching from a single

direction. However, pedestr ians with vision im pair ment may have diff iculty detectingapproaching traff ic. Cer tain design considerations at channelized r ight-turn lanes may better facilitate the safety crossing of channelized r ight-turn lanes by pedestr ians with visionim pair ment. These design considerations are discussed in Chapter 4.

A-6



Chapter 4Design Issues Related to Channelized Right-Tur nLanes

The AASHTO Policy on Geometric De si gn o f H i ghway s and S tr eets (1), commonly known

as the Gr een Book , provides guidance on the design of channelized r ight turns under the topic of turning roadways. The policy descr i bes the design controls and cr iter ia for turning roadways and

recommends values for the design elements of the hor izontal and ver tical alignment and the crosssection of turning roadways. Although the Gr een Book is the pr imary reference for roadwaydesign, channelized r ight turns are also addressed in the AASHTO Guide f or P l anning, De si gn,

and Oper ation o f Pede strian Facilitie s (2). R ecent research (3) provides additional guidance on

var ious design issues for channelized r ight turns.

This section presents a discussion of the following geometr ic design issues as they relate to

channelized r ight-turn lanes:

Crosswalk locationS pecial crosswalk signing and mark ingIsland typeR adius of turning roadwayAngle of inter section with cross street

Deceleration and acceleration lanesTraff ic controlBicycle treatments

Each of these issues is addressed below.

Cr osswalk Location

There is not univer sal agreement on where the pedestr ian crosswalk should be placed on a

channelized r ight-turn roadway. A crosswalk could potentially be placed at any location along achannelized r ight-turn roadway. It is obviously desirable to place the crosswalk at whatever location would maximize safety, presumably the location where pedestr ians who are crossing or about to cross the r ight-turn roadway are most visi ble to motor ists and where motor ists are mostlikely to yield to pedestr ians, but there are no research f indings to ver ify which of the potentialcrosswalk locations shown in the f igure is most desirable.

A ma jor ity of the sites (near ly 70 percent) evaluated by Potts et al. (3) had markedcrosswalk s located near the center of the channelized r ight-turn lane; only about 30 percent of crosswalk s were located at the upstream or downstream end of the channelized r ight-turn lane.

The highway agency survey conducted in NCHRP Pro ject 3-72 (4) found that highway agencies prefer a crosswalk location near the center of a channelized r ight-turn lane; over 70 percent of

A-7



highway agencies repor ted in the survey that their practice was to place crosswalk s near thecenter of channelized r ight-turn lanes.

There has been little research that evaluates how the crosswalk location affects crossings by pedestr ians with vision im pair ment in ter ms of their ability to identify the appropr iate time tocross or eff iciently locate the crosswalk. While fur ther research would be desirable to providemore com plete recommendations based on f ield observations of pedestr ians with visionim pair ment, or ientation and mobility (O&M) s pecialists, who teach pedestr ians with visionim pair ment how to better traver se inter sections recommend that consistency of crosswalk

location is im por tant. Such a consistency would make it easier for O&M s pecialists to descr i be atypical channelized r ight-turn lane to pedestr ians with vision im pair ment and teach proceduresfor crossing such lanes.

There is no strong technical basis for recommending one crosswalk location over another,

other than the need for consistency. However, consistency in locating crosswalk s is im por tant,es pecially to pedestr ians with vision im pair ment, and current practice shows a clear preferencefor crosswalk locations near the center of a channelized r ight-turn lane. And, a crosswalk location at the center of the channelized r ight-turn lane moves vehicle-pedestr ian conf licts awayfrom both the diverge maneuver at the upstream end of the channelized r ight-turn lane and,es pecially, from the merge maneuver at the downstream end of the channelized r ight-turn lane.

The only potential exception to a center crosswalk location for channelized r ight-turn lanes isthat, where STOP sign or traff ic signal control is provided on the channelized r ight-turnroadway, the crosswalk should be located beyond the stop line. In addition, at locations where

the channelized r ight-turn lane inter sects the cross street at near ly a r ight angle, the stop line andcrosswalk may be better placed at the downstream end of the channelized r ight-turn lane,depending on the island size and location of sidewalk approaches. To summar ize therecommended guidance for the placement of crosswalk s at channelized r ight-turn lanes:

Where the entry to the cross street at the downstream end of the channelized r ight-turnlane has yield control or no control, place the crosswalk near the center of thechannelized r ight-turn lane.

Where the channelized r ight-turn lane has STOP sign control or traff ic signal control, place the crosswalk immediately downstream of the stop bar, where possi ble. Where thechannelized r ight-turn roadway inter sects with the cross street at near ly a r ight angle, thestop bar and crosswalk can be placed at the downstream end of the channelized r ight-turn roadway.

A-8




Marked crosswalk s are the pr imary means of indicating the presence of a pedestr iancrossing. However, dr iver s do not always yield the r ight-of-way to pedestr ians sim ply becausethey are in a crosswalk. Other s pecial crosswalk signing and mark ing treatments have beenconsidered for use at pedestr ian crossings on channelized r ight-turn roadways to enhancecrossing safety for pedestr ians, in general, and for pedestr ians with vision im pair ment. Theseinclude:

Use of a raised crosswalk to im prove visi bility of crosswalk for motor ists and to better def ine crosswalk boundar ies for pedestr ians (R aised crosswalk s are par ticular ly hel pfulto pedestr ians with vision im pair ment)

Addition of ref lective yellow-green signs both at the crosswalk and in advance of thecrossing location (to supplement the high-visi bility mark ings)

Use of a real-time warning device to indicate to the motor ist when a pedestr ian is presentin the area (may be activated via passive detection technologies such as microwave or infrared or via traditional methods such as push buttons)

Use of dynamic message signs (for real-time or static warning messages to motor ists)

None of these traff ic control approaches has been evaluated to prove its effectiveness for pedestr ian crossings on channelized r ight-turn roadways. However, an infor mal com par ison (3)of pedestr ian and motor ist behavior between observational f ield study sites with and without

s pecial crosswalk signing and mark ing (i.e., raised crosswalk s with contrast pavement mark ingsand additional signing) suggested that the additional signing and pedestr ian crosswalk treatmentsmay im prove the motor ist yield behavior and pedestr ian use of the crosswalk. For exam ple, the

percentage of motor ists yielding to pedestr ians waiting at the curb was slightly better (47 percentvs. 40 percent) at sites with s pecial crosswalk treatments. This suggests that additional em phasison signing or other treatments may be needed to increase yielding for pedestr ians waiting at thecurb.

Island Type

A channelized r ight-turn lane consists of a r ight-turning roadway at an inter section,separated from the through travel lanes of both ad joining legs of the inter section by achannelizing island. At r ight-angle inter sections, such channelizing islands are roughly tr iangular in shape, although the sides of the island may be curved, where appropr iate, to match the

alignment of the ad jacent roadways. Islands serve three pr imary functions: (a) channelization— to control and direct traff ic movement, usually turning; (b) division— to divide opposing or samedirection traff ic streams; and (c) refuge— to provide refuge for pedestr ians. Most islandscom bine two or all of these functions. Islands for channelized r ight-turn lanes typically serve allthree functions.

The edges of channelizing islands may be def ined by raised curbs or may consist of painted pavement or turf that is f lush with the pavement. Most channelizing islands in urban areas aredef ined by raised curbs. Curbed islands are considered most favorable for pedestr ians because

A-9



curbs most clear ly def ine the boundary between the traveled way, intended for vehicle use, andthe island, intended for pedestr ian refuge. Curbed islands can im prove the safety for pedestr ians

by allowing them to cross the street in two stages. Or ientation and mobility (O&M) s pecialistshave a strong preference for raised islands with “cut-through” pedestr ian paths because they

provide better guidance and infor mation about the location of the island for pedestr ians withvision im pair ment than painted islands. Where curb ram ps are provided, truncated domedetectable warnings are required at the base of the ram p, where it joins the street, to indicate thelocation of the edge of the street to pedestr ians with vision im pair ment.


Design cr iter ia for the radii of channelized r ight-turn roadways are a function of turnings peeds, truck considerations, pedestr ian crossing distances, and resulting island sizes. Suchcr iter ia are established in current design policy, but the needs of pedestr ians and truck s are inconf lict in setting such cr iter ia. For exam ple, large turning radii better accommodate large truck snegotiating through r ight-turn maneuver s, but may result in higher turning s peeds on the r ight-turn roadway as pedestr ians are crossing. On the other hand, channelized r ight turns provide onemethod for accommodating larger turning radii without widening the ma jor-street pedestr iancrossings and without increasing the inter section pavement area.

R eduction of delay to turning vehicles is a key reason for providing a channelized r ight-turnlane, and recent research (3) found that increasing the radius of a channelized r ight-turn roadwayreduces r ight-turn delay by approximately 10 to 20 percent for each 8-k m/h (5-mi/h) increase in

turning s peed. Where r ight-turn volumes are high and pedestr ian and bicycle volumes arerelatively low, capacity considerations may dictate the use of larger radii, which enable higher-s peed, higher-volume turns. However, small corner radii, which promote low-s peed r ight turns,are appropr iate where such turns regular ly conf lict with pedestr ians, as higher s peeds have beenshown to result in a decrease in yielding to pedestr ians by motor ists.

Angle of Intersection with Cr oss Street

The alignment of a channelized r ight-turn lane and the angle between the channelized r ight-turn roadway and the cross street can be designed in two different ways (as illustrated inFigure 3):


The two designs shown in Figure 3 differ in the shape of the island that creates the channelizedr ight-turn lane. The f lat-angle entry design has an island that is typically shaped like anequilateral tr iangle (of ten with one curved side), while the near ly-r ight-angle design is typicallyshaped like an isosceles tr iangle. The f lat-angle entry design is appropr iate for use in channelizedr ight-turn lanes with either yield control or no control, such locations with an acceleration lane,for vehicles at the entry to the cross street. The near ly-r ight-angle entry design can be used withSTOP sign control or traff ic signal control for vehicles at the entry to the cross street; yield

A-10



control can also be used with this design where the angle of entry and sight distance along thecross street are appropr iate.

Deceleration and Acceleration Lanes

Dr iver s mak ing a r ight-turn maneuver at an inter section are usually required to reduce s peed before turning. Similar ly, dr iver s enter ing a roadway from a turning roadway accelerate until thedesired s peed is reached. Substantial deceleration or acceleration that takes place directly on the

through traveled way may disrupt the f low of through traff ic and increase the potential for

Figure 3. Typical Channelized R ight-Turn Lanes With Differing Entry

Angles to the Cross Street [Adapted From (5 )]

S k e t c h e s b y M i c h a e l K i m e l b e r g

A-11



conf licts with through vehicles. To minimize deceleration and acceleration in the through travellanes, s peed-change lanes, both for deceleration and for acceleration, may be provided byhighway agencies. Figure 4 shows the typical use of deceleration and acceleration lanes incon junction with channelized r ight-turns lanes.

Figure 4. Channelized R ight-Turn Lanes With Deceleration And Acceleration Lanes.

There are no generally established cr iter ia concerning where deceleration and accelerationlanes should be provided in con junction with channelized r ight-turn lanes. The AASHTOGr een Book (1) does not give def initive warrants for the use of s peed-change lanes, but identif iesseveral factor s that should be considered when deciding whether to im plement s peed-changelanes: vehicle s peeds, traff ic volumes, percentage of truck s, capacity, type of highway, service

provided, and the arrangement and frequency of inter sections.

Deceleration Lanes

Right-turn deceleration lanes provide one or more of the following functions (6 ):

A means of safe deceleration outside the high-s peed through lanes for r ight-turningtraff ic.

Deceleration

lane

Acceleration

lane

Deceleration

lane

Acceleration

lane

A-12




A means of separating r ight-turning vehicles from other traff ic at stop-controlledinter section approaches.

The addition of a deceleration lane at the approach to a channelized r ight-turn lane providesan oppor tunity for motor ists to safely slow down pr ior to reaching the crosswalk area at theturning roadway.

Acceleration Lanes

Acceleration lanes provide an oppor tunity for vehicles to com plete the r ight-turn maneuver unim peded and then accelerate parallel to the cross-street traff ic pr ior to merging. The addition

of an acceleration lane at the downstream end of a channelized r ight-turn lane can reduce ther ight-turn delay by 65 to 85 percent, depending on the conf licting through traff ic volume, andmay be considered where r ight-turn delay is a par ticular problem. Channelized r ight-turn laneswith acceleration lanes appear to be very diff icult for pedestr ians with vision im pair ment to

cross. Therefore, the use of acceleration lanes at the downstream end of a channelized r ight-turnlane should generally be reserved for locations where no pedestr ians or very few pedestr ians are

present. Typically, these would be locations without sidewalk s or pedestr ian crossings; at suchlocations, the reduction in vehicle delay resulting from addition of an acceleration lane becomes

very desirable.

Traff ic Contr ol

Channelized r ight-turn lanes are used at signalized inter sections (i.e., where traff ic at themain junction between the inter secting streets is controlled by traff ic signals) and at unsignalizedinter sections, typically two-way stop-controlled inter sections (i.e., locations at which traff ic on

the minor road is controlled by stop signs, but there is no traff ic control on the ma jor road).Traff ic on a channelized r ight-turn roadway generally proceeds independently of the signals or stop signs for through traff ic on the inter secting streets. Laws concerning motor ist obedience totraff ic control devices generally apply to traff ic control devices which motor ists are facing, andmotor ists on a channelized r ight-turn roadway are not generally considered to be facing thesignals or stop signs at the main inter section. Thus, any traff ic control for motor ists on a r ight-

turn roadway must be provided by traff ic control devices intended s pecif ically for motor ists onthat roadway.

The pr imary traff ic control decision for a channelized r ight-turn roadway concerns the type

of traff ic control device to be provided at the downstream end of the r ight-turn roadway, where itenter s the cross street. Traff ic control alternatives include no control, yield control, stop control,and signal control. R ecent research (3) found that signalized channelized r ight-turn lanesconsistently exper ience more delay to r ight-turning vehicles than yield-controlled channelized

r ight-turn lanes. Use of an over lap phase, as discussed in Chapter 3 and illustrated in Figure 2 of

A-13



this Guide, can substantially reduce the delay for a signalized channelized r ight-turn lane, butmay result in other im pacts to inter section operations, such as the need to restr ict U-turnmaneuver s.

Pedestr ian signals can be used at pedestr ian crossings on channelized r ight-turn roadways toenhance crossing safety for pedestr ians, par ticular ly for pedestr ians with vision im pair ment.Where a signal is provided for pedestr ians to cross a channelized r ight-turn lane, a pedestr ian-actuated signal should be considered. Actuation reduces vehicle delay for other movements

because the phase associated to crossing the channelized r ight-turn lane is called only when

pedestr ians or adequate vehicular demand is present.

Bicycle Treatments

There is a potential r isk to bicyclists at channelized r ight-turn lanes because motor vehiclesenter ing the channelized r ight-turn roadway must weave across the path of bicycles traveling

straight through the inter section. However, the same type of conf lict between through bicyclistsand r ight-turning vehicles is present at all inter sections, except at inter sections where r ight turnsare prohi bited or at three-leg inter sections where there is no leg to the r ight on a given approach.

Highway agencies may want to consider using s pecial pavement mark ings and signing, suchas those illustrated in Figures 5 and 6, to highlight a preferred bicycle path through channelizedr ight-turn lanes and to increase motor ist awareness of bicyclists at channelized r ight-turn lanes.

Hunter et al. conducted a before-af ter evaluation of the blue pavement mark ings and signing inFigures 5 and 6 and found that they resulted in an increase in motor ists yielding to bicyclists andan increase in bicyclists following the marked path. Overall, the treatment appeared to result in aheightened awareness on the par t of both bicyclists and motor ists.

A-14



Figure 5. Blue Pavement Marking Treatment at Channelized R ight-Turn Lane (7 )

A-15



Figure 6. Signs Used in Oregon Blue Bik e Lane Program (7 )

A-16



Chapter 5Typical Channelized Right-Tur n Lane Designs

This chapter presents diagrams of the following four typical channelized r ight-turn lanedesigns:

Channelized r ight-turn lane with center crosswalk and yield control

Channelized r ight-turn lane with center crosswalk, acceleration lane, and no controlChannelized r ight-turn lane with center crosswalk and stop controlChannelized r ight-turn lane with center crosswalk, signal control, and pedestr ian signal

Advantages and disadvantages of each design are presented below each diagram.

A-17



Channelized Right-Tur n Lane With Center Cr osswalk and Yield Contr ol

Advantages

Channelized r ight-turn lanes with yield control have been shown to reduce r ight-turn delay

to vehicles by 25 to 75 percent in com par ison to inter section approaches with conventionalr ight-turn lanes.

Channelized r ight-turn lanes generally result in reduced pedestr ian crossing distances and

exposure.

Shor ter signal phase times (of pr imary signal) are possi ble due to shor ter pedestr ian crossing

distance.

A consistent practice of placing the crosswalk near the center of the channelized r ight-turnlane would be benef icial to pedestr ians with vision im pair ment, par ticular ly if accom panied

by a raised island with a cut-through path.

The addition of a deceleration lane at the approach to a channelized r ight-turn lane provides

an oppor tunity for motor ists to safely slow down pr ior to reaching the crosswalk area at the

turning roadway.

Disadvantages

The lack of a pedestr ian signal can be es pecially challenging for pedestr ians with vision

im pair ment.

Potential for queued vehicles to stack across the crosswalk.

Channelized r ight-turn lanes generally have a greater inter section inf luence area in which

dr iveways would need to be restr icted.

A-18



Channelized Right-Tur n Lane With Center Cr osswalk, Acceleration Lane, and NoContr ol

Advantages

The addition of an acceleration lane can reduce the r ight-turn delay by 65 to 85 percent,

depending on the conf licting traff ic volume on the cross street.

The addition of an acceleration lane may allow motor ists to focus on the crosswalk area pr ior to having to address the task of merging with the cross-street traff ic.

Channelized r ight-turn lanes generally result in reduced pedestr ian crossing distance and

exposure.


distances.

A consistent practice of placing the crosswalk near the center of the channelized r ight-turn

lane would be benef icial to pedestr ians with vision im pair ment, par ticular ly if accom panied by a raised island with a cut-through path.

The addition of a deceleration lane at the approach to a channelized r ight-turn lane providesan oppor tunity for motor ists to safely slow down pr ior to reaching the crosswalk area at the

turning roadway.

Disadvantages

The addition of an acceleration lane to a channelized r ight-turn lane makes crossing very

challenging for pedestr ians with vision im pair ment.

High-s peed merging and weaving maneuver s may be challenging for bicyclists traveling on

the cross street.

Higher turning s peeds may be a disadvantage for pedestr ian crossings.

The lack of a pedestr ian signal can be es pecially challenging for pedestr ians with vision

im pair ment.



A-19



Channelized Right-Tur n Lane with Center Cr osswalk and STOP Contr ol

Advantages

Providing a STOP sign on the channelized r ight-turn lane may provide more crossingoppor tunities for pedestr ians than yield control. The crosswalk should be placed

immediately downstream of the stop line, where possi ble.

Consideration should be given to designing the channelized r ight-turn roadway such that it

inter sects with the cross street at near ly a r ight angle. This provides the dr iver with less of askew angle when searching for gaps in cross-street traff ic.


lane would be benef icial to pedestr ians with vision im pair ment, par ticular ly if accom panied by a raised island with a cut-through path. However, where the channelized r ight-turn

roadway inter sects with the cross street at near ly a r ight angle, the stop line and crosswalk

can be placed at the downstream end of the channelized r ight-turn roadway.


DisadvantagesSTOP control results in increased delay for r ight-turning vehicles than yield control.

There is a greater potential for longer r ight-turn queues with STOP control than with yieldcontrol.

The lack of a pedestr ian signal can be es pecially challenging for pedestr ians with visionim pair ment.



A-20



Channelized Right-Tur n Lane With Center Cr osswalk and Signal Contr ol andPedestrian Signal

Advantages

Providing a signal on the channelized r ight-turn lane may provide more crossing

oppor tunities for pedestr ians than STOP or yield control, par ticular ly at locations with high

r ight-turn volumes and pedestr ian volumes.

Provision of a pedestr ian signal is es pecially benef icial pedestr ians with vision im pair ment

but may make crossing easier for all pedestr ians.

Ability to install dual r ight-turn lanes to increase capacity without im pacting the pedestr ian

crossing distance for the inter section legs and the associated pedestr ian crossing times in the

traff ic signal timing.

Channelized r ight-turn lanes generally result in reduced pedestr ian crossing distances and

exposure.


distance.


lane would be benef icial to pedestr ians with vision im pair ment, par ticular ly if accom panied by a raised island with a cut-through path.

The addition of a deceleration lane at the approach to a channelized r ight-turn lane provides

an oppor tunity for motor ists to safely slow down pr ior to reaching the crosswalk area at the

turning roadway.

A-21



Disadvantages

Signal control results in increased delay for r ight-turning vehicles than STOP or yield

control, par ticular ly when R TOR is not per mitted.

Some pedestr ians will likely cross against the pedestr ian signal.

There is an additional cost of installing the traff ic signal equi pment for the channelized

r ight-turn movement.



A-22



Ref erences



2. Amer ican Association of State Highway and Trans por tation Off icials, Guide f or P l anning,

De si gn, and Oper ation o f Pede strian Facilitie s, 2004.

3. Potts, I. B., D. W. Harwood, K. M. Bauer, D. K. Gilmore, J. M. Hutton, D. J. Torbic,J. F. Ringer t, A. Daleiden, and J. M. Bar low, “Design Guidance for Channelized Right-Turn

Lanes,” Final R epor t, NCHRP Pro ject 3-89, Midwest R esearch Institute, July 2011.

4. Potts, I. B., D. W. Harwood, D. J. Torbic, D. K. S. A. Hennum, C. B. Tiesler, J. D. Zegeer,J. F. Ringer t, D. L. Harkey, and J. M. Bar low, Synt he sis on C hanneliz ed Ri ght T ur n s on

U r ban and Subur ban Art erials, Final R epor t, NCHRP Pro ject 3-72, Midwest R esearchInstitute, August 2006.

5. Zegeer, C. V., C. Seider man, P. Lagerwey, M. Cyneck i, M. R onk in, and B. Schneider, Pede strian Facilitie s U sers Guide: P r ovid ing Sa f et y and Mobilit y, R epor t No. FHWA-R D-01-102, Federal Highway Administration, March 2002.

6. Neuman, T. R ., NC HRP Report 279: Int ersection C hanneliz ation De si gn Guide, TRB, National R esearch Council, Washington, D.C., 1985.

7. Hunter, W. W., D. L. Harkey, J. R . Stewar t, and M. L. Birk, Eval uation o f t he Bl ue Bike

Lane Tr eat ment U sed in Bicycl e / Mot or Vehicl e C on flict Ar ea s in Portl and, Or egon, R epor t

No. FHWA-R D-00-150, Federal Highway Administration, August 2000.

A-23



Appendix B

Revised Text on Channelized Right-Tur n Lanes for theAASHTO Gree n B o o k

This appendix presents revised text on channelized r ight-turn lanes for consideration byAASHTO for potential inclusion in a future edition of the AASHTO Policy on Geometric De si gno f H i ghway s and S t r eet s (1), commonly known as the Gr een Book. The f ir st page of text should

replace the discussion in Chapter 9 in the section entitled “Types and Exam ples of Inter sections,”subsection “General Considerations,” that appear s on Page 558 of the 2004 Gr een Book . On thisf ir st page of text, modif ied text is shown in a bold font. It is not intended that this bold font

should be used in the Gr een Book .

The subsequent pages of text should be inser ted in Chapter 9 as a new subsection under

“Types of Turning R oadways,” which begins on Page 583 of the 2004 Gr een Book . Therecommended title of this new subsection is “Channelized Right-Turn Lanes.” The newsubsection could potentially be inser ted before the subsection on “Minimum Edge-of-Traveled-Way Designs.” This text is entirely new, so no bold font is used to identify changes.

The research team recommends that the ter m “free-f low r ight turns” be replaced with theter m “channelized r ight-turn lanes” throughout Chapter 9 of the Gr een Book.

The Gr een Book is currently in the publication process for a ma jor update. If this update iscom plete, this page in the revised f inal repor t for Pro ject 3-89 will be updated accordingly.



Types and Examples of Intersections

Gener al Consider ations

The basic types of inter sections are the three-leg or T, the four-leg, and the multileg. At each

par ticular location, the inter section type is deter mined pr imar ily by the num ber of inter sectinglegs, the topography, the character of the inter secting highways, the traff ic volumes, patterns, ands peeds, and the desired type of operation.

Any of the basic inter section types can vary greatly in scope, shape, and degree of channelization. Once the inter section type is established, the design controls and cr iter iadiscussed in Chapter 2 and the elements of inter section design presented in Chapter 3, as well asin this chapter, should be applied to arr ive at a suitable geometr ic plan. In this section each typeof inter section is discussed separately, and likely var iations of each are shown. It is not practicalto show all possi ble var iations, but those presented are suff icient to illustrate the general

application of inter section design. Many other var iations of types and treatment may be found inthe NCHRP R epor t 279, Int ersection C hanneliz ation De si gn Guide (2), which shows detailedexam ples that are not included in this policy.

Although many of the inter section design exam ples are located in urban areas, the pr inci plesinvolved apply equally to design in rural areas. Some minor design var iations occur withdifferent k inds of traff ic control, but all of the inter section types shown lend themselves to

cautionary or non-stop control, stop control for minor approaches, four-way stop control, and both f ixed-time and traff ic-actuated signal control. Right turns without stop or yield control aresometimes provided at channelized inter sections. Such channelized right-turn lanes should be

used only where an adequate merge is provided. Where motor vehicle conf licts with pedestr iansor bicyclists are antici pated, provisions for pedestr ians and bicycle movements must beconsidered in the design. Channelized right-turn lanes have a def inite role in improving

operations and safety at intersections. However, at locations with high pedestrian volumes,

the use of channelized right-turn lanes should be considered only where signif icant traff iccapacity or safety problems may occur without them and adequate pedestr ian crossings can be

provided.

Sim ple inter sections are presented f ir st, followed by more com plex types, some of which ares pecial adaptations. In addition, conditions for which each inter section type may be suited arediscussed below.

B-1



Channelized Right-Tur n Lanes

Channelized r ight-turn lanes have a def inite role in im proving operations and safety atinter sections. However, to achieve these benef its they should have consistent design and traff iccontrol and should be used at appropr iate locations.

Cr osswalk Location

A pedestr ian crosswalk could potentially be placed at any location along a channelized r ight-turn roadway (e.g., upstream, center, or downstream). It is obviously desirable to place thecrosswalk at whatever location would maximize safety, presumably the location where

pedestr ians who are crossing or about to cross the r ight-turn roadway are most visi ble tomotor ists and where motor ists are most likely to yield to pedestr ians.

A ma jor ity of the sites (near ly 70 percent) evaluated by Potts et al. (2) had marked

crosswalk s located near the center of the channelized r ight-turn lane; only about 30 percent of crosswalk s were located at the upstream or downstream end of the channelized r ight-turn lane.The highway agency survey conducted in NCHRP Pro ject 3-72 (3) found that highway agencies

prefer a crosswalk location near the center of a channelized r ight-turn lane; over 70 percent of highway agencies repor ted in the survey that their practice was to place crosswalk s near thecenter of channelized r ight-turn lanes.

Consistency of crosswalk location at channelized r ight-turn lanes is im por tant to pedestr ianswith vision im pair ment, and current highway agency practice indicates a preference for crosswalk locations near the center of a channelized r ight-turn lane. And, a crosswalk location at

the center of the channelized r ight-turn lane moves vehicle-pedestr ian conf licts away from boththe diverge maneuver at the upstream end of the channelized r ight-turn lane and, es pecially, fromthe merge maneuver at the downstream end of the channelized r ight-turn lane. The only potentialexception to a center crosswalk location for channelized r ight-turn lanes is that where STOP sign

or traff ic signal control is provided at the entry to the cross street, the crosswalk should belocated beyond the stop line at that point. To summar ize the recommended guidance for the

placement of crosswalk s at channelized r ight-turn lanes:

Where the entry to the cross street at the downstream end of the channelized r ight-turnlane has yield control or no control, place the crosswalk near the center of the

channelized r ight-turn lane.

Where the entry to the cross street at the downstream end of the channelized r ight-turnlane has STOP sign control or traff ic signal control, place the crosswalk immediately

downstream of the stop bar, where possi ble. Where the channelized r ight-turn roadwayinter sects with the cross street at near ly a r ight angle, the stop bar and crosswalk can be

placed at the downstream end of the channelized r ight-turn roadway.

B-2




Marked crosswalk s are the pr imary means of indicating the presence of a pedestr iancrossing. However, dr iver s do not always yield the r ight-of-way to pedestr ians sim ply becausethey are in a crosswalk. Other s pecial crosswalk signing and mark ing treatments have beenconsidered for use at pedestr ian crossings on channelized r ight-turn roadways to enhancecrossing safety for pedestr ians, in general, and for pedestr ians with vision im pair ment. Theseinclude:

Use of a crosswalk to im prove visi bility of crosswalk for motor ists and to better def inecrosswalk boundar ies for pedestr ians (R aised crosswalk s are par ticular ly hel pful to

pedestr ians with vision im pair ment)

Addition of f luorescent yellow-green signs both at the crosswalk and in advance of thecrossing location (to supplement the high-visi bility mark ings)

Use of a real-time warning device to indicate to the motor ist when a pedestr ian is presentin the area (may be activated via passive detection technologies such as microwave or infrared or via traditional methods such as push buttons)

Use of dynamic message signs (for real-time or static warning messages to motor ists)

Additional signing and pedestr ian crosswalk treatments may im prove the motor ist yield behavior and pedestr ian use of the crosswalk.

Island Type

A channelized r ight-turn lane consists of a r ight-turning roadway at an inter section,separated from the through travel lanes of both ad joining legs of the inter section by a

channelizing island. At r ight-angle inter sections, such channelizing islands are roughly tr iangular in shape, although the sides of the island may be curved, where appropr iate, to match thealignment of the ad jacent roadways. Islands serve three pr imary functions: (a) channelization—

to control and direct traff ic movement, usually turning; (b) division— to divide opposing or samedirection traff ic streams; and (c) refuge— to provide refuge for pedestr ians. Most islandscom bine two or all of these functions. Islands for channelized r ight-turn lanes typically serve allthree functions.

The edges of channelizing islands may be def ined by raised curbs or may consist of painted

pavement or turf that is f lush with the pavement. Most channelizing islands in urban areas aredef ined by raised curbs. Curbed islands are considered most favorable for pedestr ians becausecurbs most clear ly def ine the boundary between the traveled way, intended for vehicle use, andthe island, intended for pedestr ian refuge. Curbed islands can im prove the safety for pedestr ians

by allowing them to cross the street in two stages. R aised islands with “cut-through” pedestr ian

paths are im por tant to pedestr ians with vision im pair ment because they provide better guidanceand infor mation about the location of the island than painted islands. Where curb ram ps are

provided, truncated dome detectable warnings are required at the base of the ram p, where it joins

the street, to indicate the location of the edge of the street to pedestr ians with vision im pair ment.

B-3




Design cr iter ia for the radii of channelized r ight-turn roadways are a function of turnings peeds, truck considerations, pedestr ian crossing distances, and resulting island sizes.Channelized r ight-turn lanes provide one method for accommodating larger turning radii withoutwidening the ma jor-street pedestr ian crossings and without increasing the inter section pavementarea. Where r ight-turn volumes are high and pedestr ian and bicycle volumes are relatively low,

capacity considerations may dictate the use of larger radii, which enable higher-s peed, higher-volume turns. However, small turning radii, which promote low-s peed r ight turns, areappropr iate where such turns regular ly conf lict with pedestr ians, as higher s peeds have beenshown to result in a decrease in yielding to pedestr ians by motor ists.

Angle of Intersection With Cr oss Street

The alignment of a channelized r ight-turn lane and the angle between the channelized r ight-turn roadway and the cross street can be designed in two different ways:


The two designs differ in the shape of the island that creates the channelized r ight-turn lane.The f lat-angle entry design has an island that is typically shaped like an equilateral tr iangle(of ten with one curved side), while the near ly-r ight-angle design is typically shaped like anisosceles tr iangle. The f lat-angle entry design is appropr iate for use in channelized r ight-turn

lanes with either yield control or no control for vehicles at the entry to the cross street. Thenear ly-r ight-angle entry design can be used with STOP sign control or traff ic signal control for vehicles at the entry to the cross street; yield control can also be used with this design where theangle of entry and sight distance along the cross street are appropr iate.

Deceleration Lanes

Dr iver s mak ing a r ight-turn maneuver at an inter section are usually required to reduce s peed before turning. Signif icant deceleration that takes place directly on the through traveled way maydisrupt the f low of through traff ic and increase the potential for conf licts with through vehicles.

To minimize deceleration in the through travel lanes, deceleration lanes should be considered.

Right-turn deceleration lanes provide one or more of the following functions (4):

A means of safe deceleration outside the high-s peed through lanes for r ight-turningtraff ic.


B-4



A means of separating r ight-turning vehicles from other traff ic at stop-controlledinter section approaches.


Acceleration Lanes

Acceleration lanes provide an oppor tunity for vehicles to com plete the r ight-turn maneuver unim peded and then accelerate parallel to the cross-street traff ic pr ior to merging. Channelizedr ight-turn lanes with acceleration lanes appear to be very diff icult for pedestr ians with visionim pair ment to cross. Therefore, the use of acceleration lanes at the downstream end of a

channelized r ight-turn lane should generally be reserved for locations where no pedestr ians or very few pedestr ians are present. Typically, these would be locations without sidewalk s or

pedestr ian crossings; at such locations, the reduction in vehicle delay resulting from addition of an acceleration lane becomes very desirable.

Pedestrian Signals

Pedestr ian signals can be used at pedestr ian crossings on channelized r ight-turn roadways to

enhance crossing safety for pedestr ians, par ticular ly for pedestr ians with vision im pair ment.Where a signal is provided for pedestr ians to cross a channelized r ight-turn lane, a pedestr ian-actuated signal should be considered.

B-5



Ref erences



2. Potts, I. B., D. W. Harwood, K. M. Bauer, D. K. Gilmore, J. M. Hutton, D. J. Torbic, J. F.

Ringer t, A. Daleiden, and J. M. Bar low, NC HRP Web-Onl y Document 208: De si gn Guidance

f or C hanneliz ed Ri ght -T ur n Lane s, Trans por tation R esearch Board of the NationalAcademies, Washington, D.C., 2011.

3. Potts, I. B., D. W. Harwood, D. J. Torbic, D. K. S. A. Hennum, C. B. Tiesler, J. D. Zegeer, J.F. Ringer t, D. L. Harkey, and J. M. Bar low, Synt he sis on C hanneliz ed Ri ght T ur n s on U r banand Subur ban Ar t er ial s, Final R epor t, NCHRP Pro ject 3-72, Midwest R esearch Institute,

August 2006.

4. Neuman, T. R ., NC HRP Report 279: Int ersection C hanneliz ation De si gn Guide, NCHRPR epor t 279, TRB, National R esearch Council, Washington, D.C., 1985.

channeiization

Documents