Turn Lane Lengths For Various Speed Roads and Evaluation of Determining Criteria

2008-14
Turn Lane Lengths for Various Speed Roads and

Evaluation of Determining Criteria
ve Solutions!
vati
ch . ... Inno
r .. Knowledge
sea
Take the steps... Re
Transportation Research
Technical Report Documentation Page
1. Report No. 2. 3. Recipients Accession No.
MN/RC 2008-14
4. Title and Subtitle 5. Report Date
May 2008
Turn Lane Lengths for Various Speed Roads and Evaluation of 6.
Determining Criteria
7. Author(s) 8. Performing Organization Report No.
Lora Yekhshatyan, Thomas Schnell
9. Performing Organization Name and Address 10. Project/Task/Work Unit No.
Operator Performance Laboratory
Center for Computer Aided Design 11. Contract (C) or Grant (G) No.
University of Iowa (c) 82647
116 Engineering Research Facility
330 S. Madison St.
Iowa City, Iowa 52242
12. Sponsoring Organization Name and Address 13. Type of Report and Period Covered
Minnesota Department of Transportation Final Report
395 John Ireland Boulevard Mail Stop 330 14. Sponsoring Agency Code
St. Paul, Minnesota 55155
15. Supplementary Notes
http://www.lrrb.org/PDF/200814.pdf
16. Abstract (Limit: 200 words)
The purpose of this research project performed for The Minnesota Department of Transportation is to find the
optimal length of right and left turn lanes at intersections from a system design point of view. This research project
will also determine and quantify the influence of the factors that need to be considered when estimating turn
lengths on specific type of intersection. The following parameters that possibly affect right and left turn lane
lengths in Signalized and Unsignalized intersections are investigated in this study: speed, grade, through and
turning traffic volumes, heavy vehicle mixture, and protected/unprotected left and right turn signalization. In this
study, there is also an in depth review of technical literature and a national and international survey of turn lane
design practices. The videotaped observational data was used to calibrate a computer model of the intersection
scenarios that were video taped using traffic analysis software SYNCHRO and SimTraffic. The calibrated
computer models were then used to conduct a sensitivity analysis to determine the factors that could be used to
predict the most optimal turn lane length. Our major challenge was to develop a set of equations that accurately
predicts the queue length of the turning traffic at the standard intersection types.
17. Document Analysis/Descriptors 18. Availability Statement

Turn lane length, No restrictions. Document available from:
queue length, National Technical Information Services,
intersection design, Springfield, Virginia 22161
traffic analysis,
turning traffic
19. Security Class (this report) 20. Security Class (this page) 21. No. of Pages 22. Price
Unclassified Unclassified 121
Turn Lane Lengths for Various Speed Roads and
Evaluation of Determining Criteria
Final Report
Prepared by:
Lora Yekhshatyan
Tom Schnell
Operator Performance Laboratory (OPL)

Center for Computer Aided Design (CCAD)
University of Iowa
May 2008
Published by:
Minnesota Department of Transportation

Research Services Section
395 John Ireland Boulevard, MS 330
St. Paul, Minnesota 55155-1899
This report was prepared by the Operator Performance Laboratory, University of Iowa as part of a
research project sponsored by the Minnesota Department of Transportation. The contents of this report
reflect the views of the authors, who are responsible for the facts and accuracy of the data presented
herein. The contents do not necessarily reflect the official view or policies of the Minnesota Department
of Transportation. This report does not constitute a standard, specification, or regulation, nor is it intended
for construction, bidding, or permit purposes.
The authors and the Minnesota Department of Transportation and Center for Transportation Studies do
not endorse products or manufacturers. Trade or manufacturers’ names appear herein solely because they
are considered essential to this report.
TABLE OF CONTENTS
1. INTRODUCTION..................................................................................................................... 1
1.1 PROBLEM STATEMENT ........................................................................................................... 1
1.2 SOLUTION APPROACH ............................................................................................................ 1
2. BACKGROUND AND TECHNICAL LITERATURE......................................................... 2
2.1 SIMULATION AND PREDICTION TOOLS ................................................................................... 2
2.1.1. Survey of Intersection Design Practices and Standards............................................... 2
2.2 LITERATURE REVIEW ............................................................................................................. 5
2.3 EXISTING GUIDELINES ........................................................................................................... 8
3. DEVELOPMENT OF CALIBRATED SIMULATION MODELS ................................... 15
3.1 SELECTION AND DOCUMENTATION OF INTERSECTIONS ........................................................ 15
3.2 QUEUE LENGTH OBSERVATION IN FIELD ............................................................................. 17
3.3 ANALYSIS OF FIELD DATA ................................................................................................... 18
3.4 SIMULATION MODEL CALIBRATION ..................................................................................... 21
4. GENERATION OF QUEUE LENGTH PREDICTION MODEL..................................... 25
4.1 SIMULATION MODEL PARAMETERS...................................................................................... 25
4.2 SIMULATION MODEL DATA.................................................................................................. 26
5. TURN LANE LENGTH RECOMMENDATIONS ............................................................. 36
5.1 LEFT TURN LANE LENGTH (TAPER + DECELERATION + STORAGE)...................................... 39
5.2 RIGHT TURN LANE LENGTH (TAPER + DECELERATION + STORAGE) ................................... 42
6. DISCUSSION AND CONCLUSIONS .................................................................................. 45
REFERENCES............................................................................................................................ 49
APPENDIX A. QUESTOINNAIRE
APPENDIX B. MAPS
APPENDIX C. DIAGRAMS
APPENDIX D. INTERSECTION DATA SUMMARY
APPENDIX E. SUMMARY DATA FOR THE SIMULATION QUEUE LENGTHS
APPENDIX F. REGRESSION EQUATIONS
APPENDIX G. EXAMPLE
LIST OF TABLES
Table 1. Comparison of length requirements (in number of vehicles): Proposed

model and Simulation model .................................................................................................. 6
Table 2. Comparison of NETSIM with proposed model results .................................................... 7
Table 3. Guide for left-turn lanes on two-lane highways ............................................................... 9
Table 4. Summary table of some existing guidelines ................................................................... 11
Table 5. Intersection descriptions from the field study................................................................. 17
Table 6. Description of intersections for categorization............................................................... 20
Table 7. Evaluation of the accuracy of the simulation model of the intersection
TH 169-CR 4......................................................................................................................... 23
Table 8. Evaluation of the accuracy of the simulation models: difference (in cars)
between average queue length of the simulation model and queue length
obtained from the videotapes ................................................................................................ 24
Table 9. Coefficient of determination for the single intersection in each group .......................... 27
Table 10. Coefficient of determination of the multivariate regression analysis of
the average queue lengths for each group of intersections and their
combinations ......................................................................................................................... 28
Table 11. Deceleration length for right-turning traffic ................................................................. 37
Table 12. Deceleration length for left-turning traffic ................................................................... 37
Table 13. Coefficient of determination of the multivariate regression analysis of
the 95th percentile queue lengths for each group of intersections and their
combinations ......................................................................................................................... 39
Table 14. Three components of the left turn lane length .............................................................. 41
Table 15. Three components of the right turn lane length............................................................ 44
LIST OF FIGURES
Figure 1. Source of the guidelines for the exclusive turn lanes on signalized
intersections ............................................................................................................................ 3
Figure 2. Source of guidelines for signalized and unsignalized turn length................................... 3
Figure 3. Guidelines efficiency....................................................................................................... 4
Figure 4. Scheme of Unsignalized intersection with turning lanes ................................................ 5
Figure 5. Lane overflow and blockage of lane entrance at a signalized intersection ..................... 7
Figure 6. Left turn bay storage requirements as a function of left turn volume ........................... 10
Figure 7. Left-Turn Lane Warrants............................................................................................... 13
Figure 8. Right-Turn Lane Warrants ............................................................................................ 14
Figure 9. Overall map of area ....................................................................................................... 15
Figure 10. Classification chart ...................................................................................................... 21
Figure 11. Channelization design on the right turns in SYNCHRO............................................. 22
Figure 12. Sketch of volumes on the intersection......................................................................... 26
Figure 13. Components of turn lane ............................................................................................. 36
Figure 14. Right-Turn Deceleration Bay ...................................................................................... 38
Figure 15. The flow chart of the turn lane length evaluation procedure....................................... 48
EXECUTIVE SUMMARY
The purpose of this research project performed for Minnesota Department of Transportation
(MN DOT) was to find the optimal length of right and left turn lanes at intersections from the
point of view of a system design. This research project also determined and quantified the
influence of the factors that need to be considered when estimating turn lengths on specific type
of intersection. The primary parameters that significantly affect right and left turn lane lengths at
signalized and unsignalized intersections and which were considered in our study are: speed,
grade, through/turning traffic volumes, and heavy vehicle mixture. The report also includes an
in-depth review of technical literature and a national and international survey of turn lane design
practices.
A turn lane should be designed to allow the turning vehicle to exit the through lane, to
decelerate, and to provide enough storage space for vehicles waiting to complete the turn. Design
and estimation of turn lane lengths are provided in the American Association of State Highway
and Transportation Officials (AASHTO) guideline, the Highway Capacity Manual (HCM),
Traffic Engineering Handbook, and the MNDOT Road Design Manual. A brief summary of
these guidelines is presented below.
The distance of the vehicle deceleration movement consists of two parts, the taper and
deceleration length. The turn lane begins with a taper, the design of which depends on location
and traffic characteristics. The AASHTO, the HCM, and Traffic Engineering Handbook
guidelines specify the taper length as a ratio of 8:1 and 15:1 for design speeds up to 30 mph and
up to 50 mph, respectively (for 12 ft lane width). The Traffic Engineering Handbook specifies
the deceleration distance in the high speed conditions as 100 ft to 130 ft. The sources of the
abovementioned guidelines provide the deceleration lengths according to the design speeds in
desirable conditions and in limiting conditions. The estimation of the storage bay in the
abovementioned guidelines is associated with the average number of vehicles in a 2 min interval.
The recommendations suggest a minimum length of the storage bay to be equal to two car
length.
To predict overflow or blockage of the turn lane, the existing methodologies and the practices of
the storage length design take into account at least a 95% probability of storing all turning
vehicles during the peak hour. For design purposes, we used the 95th percentile queue length
consideration. Our major challenge was to develop a set of equations that accurately predicts the
95th percentile queue length of the turning traffic at the standard intersection types. Prediction
models were generated to derive equations that are easy to use as design guidelines for the use by
practitioners.
Ten intersections were selected in the Twin-Cities area with the help of the Mn/DOT project
panel. These intersections had the following general characteristics:
4-lane and 2-way on arterial road and one turn lane for each turn
2-lane and 2-way on crossing road and one turn lane for each turn
with or without channelization design on right turns
The factors that affect the queue length were obtained in field observations using a remote
wireless monitoring and recording system deployed at the study intersections. Video footage was
analyzed with regard to the independent variables offline, in the laboratory. The observational
data obtained in this fashion was then used to generate and exercise computer models of the
intersections scenarios in traffic analysis software SYNCHRO and SimTraffic. To build a
simulation model of each intersection it was reasonable to group them into categories according
to the following criteria:
Type of signal control: Signalized and Unsignalized
Type of turn: Left and Right
Type of mode on signalized intersections
Protected and Permitted for left turns
Permitted and Free for right turns
Type of sign on unsignalized intersections
Yield for left turns
Yield and Free for right turns
Thus, seven types of computer models based on defined categories were calibrated and
exercised to conduct a sensitivity analysis to determine the factors that could be used to predict
the most optimal turn lane length. The equations for calculating the storage length were derived
by means of multivariate regression analysis. The queue length was considered to be a
dependent variable, and the following list of parameters represented a set of independent
variables:
1. Through Volume (TV) – vehicles per hour per through lanes in through direction
2. Opposing Volume range (OV) – vehicles per hour per through lanes in opposing
direction
3. Crossing Volume (CV) – vehicles per hour per through lanes in crossing direction
4. Left Turn Volume (LTV) / Right Turn Volume (RTV) – vehicles per hour per turn lane
in left / right direction
5. Heavy Vehicle Through Percent (HVT) – percent of heavy vehicles in the through
direction
6. Heavy Vehicle Left (HVL) / Right Turn Percent (HVR) – percent of heavy turning
vehicles
7. Grade (Gr) – the percent of grade on the approach of the through direction
8. Speed (Sp) – miles per hour on the approach
All data from the exercised generic models were tabulated and examined in the MINITAB
statistical software program to generate the regression models and assess the influence of each
input on the queue length. The regression models appeared to result in a good fit: the coefficient
of determination for the most of intersections was greater than 80%, which means that the
regression models relatively accurately explain all of the variation in the response and can
predict the average queue length for each intersection. By applying the regression analysis to
different groups of intersections according to our classification, we got that each group of
intersections shows good correlation, and the coefficient of determination is equal to or greater
than 85%.
Thus, the total length of the turn lane should be calculated as a sum of three components—taper,
deceleration length, and storage length—for each turn type according to our classification and
design speed. The entire procedure of defining a turn lane length is divided into the following
steps:
1. Identification of the type of turn: Left or Right;

2. Identification of the type of intersection: Signalized or Unsignalized;
3. Identification of the type of mode or sign control: Protected, Permitted, Yield, or Free;
4. Identification of the designed speed value;
5. Identification of the total volume of the through traffic - Through Volume;
6. Identification of the Opposing Volume of the traffic (total volume of the opposing
traffic) for left turns and Crossing Volume of the traffic (total volume of the crossing
traffic) for right turns;
7. Identification of the Heavy Vehicles Mixture Percent for the turn lane;
8. Identification of the Heavy Vehicles Mixture Percent for the through lane;
9. Identification of the Grade;
10. Calculation of the storage bay for defined turn type;
11. Determination of the taper and deceleration lengths for defined design speed;
12. Calculation of the sum of the storage length obtained in Step 10 and the deceleration
distance obtained in Step 11.
The influence of each independent variable within the ranges that was investigated from the
regression analysis is as follows:
1. The most important and significant factors are Through Volume, Opposing Volume (for
the left turns), Crossing Volume (for the right turns), Left and Right Turn Volumes,
Heavy Vehicle Turn Percent for left and right turns, and Speed.
2. The influence of Heavy Vehicle Through Percent factor exists but does not appear to be
significant for Permitted left turns on Signalized intersections; Free right turns on both
types of intersections; Yield right and left turns on Unsignalized intersections.
3. The influence of Through Volume on Free right turns on Unsignalized intersections does
not appear practically significant.
4. The Grade factor does not appear to affect the left and right turn queue length.
These recommendations are applicable for the following range of parameters:

1. Volume per Through lane – up to 500 vehicles per hour per lane
2. Turn Volume – up to 250 vehicles per hour
3. Heavy Vehicle Percent – up to 20-25%
4. Speed – from 30 mph to 70 mph
5. Grade – from -4% to +4%.
For the values that are not in these ranges, additional HCM recommendations such as double left
turn lane consideration, changes in the available green time, etc. should be used.
1. INTRODUCTION
1.1 Problem Statement

This study was conducted to build a comprehensive set of design guidelines and a tool to
determine right and left turn lane lengths as a function of a fairly exhaustive list of independent
variables. The factors that the road designers from The Minnesota Department of Transportation
(MN DOT) are interested in are through volume, turn volume, heavy vehicle through percent,
heavy vehicle turn percent, grade, and speed. It may be possible to reduce travel delays in urban
roadway networks by optimizing turn lane lengths at intersections according to the
abovementioned factors and additional parameters such as deceleration rates, vehicle classes,
traffic behavior parameters, land availability, level of service, signal timing, etc. We also wanted
to determine the design practices for intersections across the US and worldwide.
The length of the lanes for turning traffic includes three components: taper, deceleration length,
and storage length. The length of the turn lane should be long enough to enable drivers to
decelerate outside the through lane (especially for high-speed highways) and to store vehicles
queued in the auxiliary lane (especially for high-volume roadways). The design and calculation
of taper and deceleration lengths are suggested in the American Association of State Highway
and Transportation Officials (AASHTO) guideline, the Highway Capacity Manual (HCM), and
the Road Design Manual, and we will use those values in the final equation. The estimation of
length of the storage bay is the main purpose of this research.
1.2 Solution Approach

Ten intersections in the state of Minnesota were selected for studying the problem of turn lane
length. The intersections exhibited a wide range of factors that, according to the literature review
and the survey results, were thought to have an effect on the optimum turn lane length.
Documentation of the intersections was collected, such as existing construction drawings, video
recording data and photographs taken during the field survey, and signal and timing plans
provided by MN DOT .
Generation and calibration of simulation models was performed by means of macroscopic and
microscopic simulation packages such as SYNCHRO (to govern vehicle movement) and
SimTraffic (to perform characteristics of a driver). For adjustment of the simulation models, such
factors as cycle length, saturated flow rate, turning speed, max split, headway factor, and vehicle
mixture factor were used. We attempted to calibrate the simulation models to achieve traffic
behavior that was close to that found in real-world intersections and to have comparable queue
lengths.
The next step was to exercise the calibrated models and carry out a sensitivity analysis to
determine the effect of each parameter and the configuration of parameters by varying the wide
range of independent factors. Regression models, generated on the basis of the simulation run
data, provided turning vehicles queue length as a function of the independent variables using
simple equations. The regression model approach is convenient for practitioners, and it appears
to result in a good fit with the data.
1
2. BACKGROUND AND TECHNICAL LITERATURE
2.1 Simulation and Prediction Tools

In this chapter, we delineate the existing practices and methodologies for calculating turning
vehicle queue length. Existing guidelines determine the warrants of turn lanes depending on turn,
through, opposing, and cross traffic; speed; signal timing; cycle length; and g/C ratios for a
signalized intersection. These guidelines are available in Traffic Engineering Handbook, A
Policy of Geometric Design on Highways and Streets (called the Green Book), the Highway
Capacity Manual, and the Road Design Manual. In order to comprehend how the traffic
engineers use the turn lane guidelines, we reviewed the existing turn lane practices by
conducting a survey.
2.1.1. Survey of Intersection Design Practices and Standards
A survey was designed to investigate the current practices of determining the queue length for
right and left turn lanes. The survey questionnaire is in Appendix A. Surveys were mailed out to
approximately 1,000 participants, including the traffic and design engineers of the US state
DOTs, Canadian province, traffic engineers in Europe and Japan, worldwide design offices, and
individually designated specialists. About ten specialists were interviewed over the phone or in
person. We received 37 completed surveys.
The survey was divided into two main categories: determination of left turn lanes and right turn
lanes. The survey included questions about signalized turn lanes and unsignalized turn lanes for
both categories. The questions were grouped as follows:
− Guidelines agencies pursue to warrant an exclusive turn lane at the intersection
− Guidelines agencies pursue when determining the length of the queue
− Agencies’ opinions about guidelines that ensure efficient queue lengths
− Ways in which agencies determine the length of the queue
− Ways in which agencies warrant exclusive left and right turn lanes
To determine the necessity of an exclusive turn lane at signalized intersections, the respondents
referred to the Capacity Analysis, the Highway Capacity Manual (HCM), the Highway Capacity
Software (HCS), Traffic Engineering Handbook, or to a rule of thumb if the number of vehicles
turning during a peak hour is equal to or greater than 300, etc. The sources mentioned represent
the Other Guidelines option. They refer to the American Association of State Highway and
Transportation Officials (AASHTO) and State Guidelines with the same frequency (Figure 1).
On the question about the source of guidelines for determining the length of the turn lane, the
options were the AASHTO Guidelines, the State Guidelines, the Harmelink’s Guidelines, and
the Other guidelines, the respondents mainly mentioned the State Guidelines (Figure 2). The
Other Guidelines option included SYNCHRO analysis, the capacity analysis, the HCS analysis,
Leisch Nomograph, the HCM, etc. The AASHTO guidelines were mentioned often, mostly to
determine the left turn length for a signalized intersection (40.5%).
2
60.0%
Left Turn
Right Turn
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%
AASHTO guidelines State guidelines Harmelink guidelines Other guidelines
Figure 1. Source of the guidelines for the exclusive turn lanes on signalized intersections
60.0% Signalized Left Turn

Unsignalized Left Turn
50.0% Signalized Right Turn
Unsignalized Right Turn
40.0%
30.0%
20.0%
10.0%
0.0%
AASHTO guidelines State guidelines Harmelink guidelines Other guidelines
Figure 2. Source of guidelines for signalized and unsignalized turn length
To the question, “In your opinion, do the guidelines that your agency uses provide efficient turn
lane lengths at all signalized intersections?”, most respondents answered “Yes” or “Other.”
3
Respondents explained that their choice might vary depending on different situations and
conditions (Figure 3).
yes no other
100.0%
90.0%
80.0%
70.0%
60.0%
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%
Signalized Left Turn Unsignalized Left Signalized Right Turn Unsignalized Right
Turn Turn
Figure 3. Guidelines efficiency
The majority of answers to the question about how agencies warrant an exclusive turn lane at
signalized or unsignalized intersections were as follows:
− Volume of turning vehicles during peak hours (up to 62.2%)
− Opposing/Crossing Volume (up to 56.8%)
− Advancing Volume (up to 43.2%)
The Vehicle Mixture parameter was mentioned for left turns on signalized intersections in 100%
of cases. The Driver Complaints option was mentioned in about one-third of answers for all
types of intersections.
To the question, “How does the agency determine the length of the queue for turning lanes?”, the
majority of the respondents answered as follows:
− Volume of turning vehicles during peak hours (up to 78.4%)
− Advancing Volume (54.1%)
− Opposing/Crossing Volume (up to 45.9%)
4
2.2 Literature Review
The research on the evaluation of left-turn lanes at unsignalized grade intersections on four-lane
and two-lane highways that is well known among traffic designers is Harmelink’s work (1967).
He based his derivations on a queuing model with negative exponential distribution for arrival
and service rates. He presented his results in the form of graphs with the following variables:
advancing volume, opposing volume, speed, and left-turn percentage. The range of speed is of
40, 50, and 60 mph, and left-turn volumes are 5, 10, 15, 20, 30, and 40 percent. The AASHTO’s
Green Book contains a table determining the need for a left-turn lane on two-lane highways
based on the values from the Harmelink’s graphs (Table 3).
Kikuchi et al. (1991) indicates that Harmelink’s method overestimates service rates because his
definition of service rates supposes that the sum of residual gaps is also considered a part of the
time available to make a turn. Thus, predicted queues are smaller than the actual ones.
Chakroborty et al. (1995) developed a mathematical model to determine the adequate left turn
lane length at unsignalized intersections (Figure 4). Factors considered in the model are traffic
volumes, vehicle mix, critical gap size, space required per vehicle, and threshold probability.
Recommended lane lengths were created in a tabular form, then were validated by comparing the
recommended lane length with the lengths derived from computer simulation software. The
result showed similar lane lengths (Table 1).
Figure 4. Scheme of Unsignalized intersection with turning lanes

Source: Chakroborty et al., 1995
5
Table 1. Comparison of length requirements (in number of vehicles): Proposed model and
Simulation model
Source: Chakroborty et al., 1995
Lertworawanich and Elefteriadou (2003) proposed a queuing model to determine lengths of left
turn lanes at unsignalized intersections based on probability of overflow. Their model
incorporated two types of service time: service times incurred by the vehicles that arrive when
the left-turning lane is empty, and service times incurred by the vehicles that arrive when the
left-turning lane is not empty. In their model, determined storage lengths of left-turning lanes are
provided for various combinations of opposing volume, left-turning volumes, and threshold
probabilities of the lane overflow. The lengths recommended by the model were compared with
the lengths recommended in the 2001 Green Book guideline. The result showed similar
recommendation length; however, when the demands of left turn were high, the Green Book
recommendation underestimated the required length.
Kikuchi et al. (1993) have developed a methodology to determine a recommendation for lengths
of left turn lanes at signalized intersections. Two case problems were considered, the probability
of the overflow of vehicles from the turning lane, and the probability of blockage of the entrance
to the turning lane by the queue of vehicles in the adjacent through lane (Figure 5).
6
Source: Kikuchi et al., 1993
Figure 5. Lane overflow and blockage of lane entrance at a signalized intersection
Both cases were modeled using the Markov Chain algorithm that considers previous and current
probabilities of states to predict the next states of the system. The factors used for modeling
were: traffic volumes (left-turn, through, and opposing volumes), vehicle mix, signal timing,
time required to make a left turn, and gap between vehicles. The recommended lengths of left-
turn lanes derived in this research were presented in a tabular form for both lane overflow and
lane blocking problems. The optimum length was determined to be the largest of the two
presented by both case problems. The recommended lengths were compared with the existing
guidelines (AASHTO and HCM) and the simulation results from NETSIM traffic simulation
software. The lengths suggested in AASHTO guidelines are longer and the difference increases
proportionally to the left-turn volume. Recommended and suggested in the HCM, lengths are
close for most values of left-turn volume but in the case of lane blockage, the recommended
lengths differ from the guidelines suggested lengths. The recommended length and simulation
results are highly agreeable (Table 2).
Table 2. Comparison of NETSIM with proposed model results
Source: Kikuchi et al., 1993
7
Oppenlander et al. (1989) applied a rational procedure to determine the design length for left or
right turn lanes with separate signal control. They used arrival and service rate equations, which
are acceptably accurate for predicting traffic operations for both left and right turn lanes.
Another approach to investigate the influence of independent variables on the queue length of the
turning vehicles is a simulation of the traffic operations. McCoy (1994) conducted research to
determine the right turn volume for two-lane and four-lane roadways as a function of directional
volume, roadway speed, number of lanes, and right-of-way cost. Multiple regression analysis of
the outputs of simulation models with and without right turn lanes was conducted. The
guidelines specify the design-hour traffic volumes for which the benefits of right turn lanes, such
as operational and accident cost savings, exceed their cost. The accident cost saving relates to
reduction in accidents expected from the lower speed differentials between right-turn and
through traffic. The operational cost saving relates to reduction in stops, delays, and fuel
consumption. The costs of right turn lanes were estimated from the cost data for the construction
of the right turn lanes in a typical urban area of the state highway system in Nebraska. The
guidelines developed in the research are within the range of those developed previously.
The study of the effectiveness of the “Michigan U-Turn” relative to conventional direct turn lane
operations, conducted by using a computer simulation approach (Dorothy et al.,1997), shows that
− the boulevard designs that used indirect left-turning strategies and signalized crossovers
in most cases were superior to all other designs; and
− with a low percentage of left turns, the operation of an indirect-left-turn boulevard
design had similar results for both signalized and stop-controlled crossovers.
The variables of interest studied in this research were traffic volumes for both arterial and
crossroad approaches, turning percentages, median width, and type of traffic control device at the
crossover.
One more factor that affects completing a turn is speed differential, which is the difference
between the speeds of the through traffic and the turning traffic. Keeping the speed differential
below 20 mph is important to provide safety to the intersection. Misaghi and Hassan (2005)
estimated the speed differential based on the difference of the operating speed of the drivers on
the curved and the tangent sections. The statistic “Δ85V” (differential speed not exceeded by 85%
of the drivers traveling under free flow condition) instead of “ΔV85” (difference of the operating
speeds on two successive elements V85 i and V85 i-1) was used to set up the relationship with the
geometric parameters of the road, traffic volume, and grade. They developed a model to predict
the vehicle’s operating speed at different points of a curve for different classes of vehicles. One
of the findings was that the relationship between speed differential and the geometric features of
the road is considerably strong.
2.3 Existing Guidelines

A summary table that determines the need for left turn lane for two-lane intersections with
speeds between 40 and 60 mph from Harmelink’s graphs (Table 3) is presented in AASHTO
guidelines. The table shows the traffic volumes for (and above) which one should consider
providing left turn lanes. The parameters considered include left turn vehicle volume, opposing
8
vehicle volume, vehicle mix, and average time to make a turn. Some general guidelines for
using protected left turning movements in the Highway Capacity Manual (1997) are as follows:
− Left turn protection is rarely used for left turn volumes of less than 100 vph.
− Left turn protection is almost always used for left turn volumes greater than 250-300
vph.
− Where left turn volumes exceed 300 vph, provision of a double left turn lane should be
considered.
− For arrival rates between these values, the analysts must consider opposing volumes,
number of lanes, accident experience, and other related factors.
− Any approaches having a protected left turn phase must have exclusive left turn lane or
bay with sufficient length to accommodate the expected queues during each signal
cycle.
Table 3. Guide for left-turn lanes on two-lane highways

Opposing Volume Advancing Volume (veh/h)
(veh/h)
5% Left Turns 10% Left Turns 15% Left Turns 20% Left Turns
40-mph operating speed
800 330 240 180 160
600 410 305 225 200
400 510 380 275 245
200 640 470 350 305
100 720 575 390 340
800 280 210 165 135
600 350 260 195 170
400 430 320 240 210
200 550 400 300 270
100 610 445 335 295
800 230 170 125 115
600 290 210 160 140
400 365 270 200 175
200 450 330 250 215
100 505 370 275 240
Source: AASHTO, 2004, exhibit 9-75
The standard practice in many design offices is to make the left turn lane equivalent in length as
1.5 to 2 times the average number of arriving turning vehicles per cycle would be. Figure 6
shows the relationship between left turn volume in passenger car equivalents and the required
length of the storage bay (Highway Capacity Manual, 1994, updated in 1997).
9
Source: Highway Capacity Manual, 1994. Updated December 1997
Figure 6. Left turn bay storage requirements as a function of left turn volume
Table 4 presents a comparison of guidelines from the following sources: the AASHTO Green
Book, the Highway Capacity Manual, and the Road Design Manual. Traffic Engineering
Handbook suggests the design of left and right turn lanes based on operational needs, as follows:
− Lanes should be long enough to enable a driver to decelerate outside the higher speed
lanes at rural intersections.
− Lanes should be of sufficient length to store vehicles queued in the turn-lane at urban
signalized intersections.
10
Table 4. Summary table of some existing guidelines
Source Deceleration Storage Length Taper Length

Length
(ft)
Design Deceleration Signalized Unsignalized
Speed Length
(mph)
30 170* Length = 2 or Length = average 1:15 for design speed 50
AASHTO 40 275* 1.5 x average number of vehicles mph,
2004 45 340* number of 2 min interval. 1:8 for design speed up to
50 410* vehicles that Minimum length 30 mph
55 485* would store = 2 cars, with max
* for grades per cycle. 10% heavy OR
less than 3 vehicles.
percent Minimum Minimum length L = S2xW/60, where
A 10 mph speed Length = 2 = car and truck, S - design speed less than
differential* is passenger with min 10% 40mph,
commonly acceptable cars heavy vehicles. W - offset
on arterial roads.
Higher speed L = S x W, where
differentials may be
S - design speed greater
acceptable on collector
highways and streets. than 45 mph,
W – offset
(AASHTO, 1990)**
Road Taper is a part of See AASHTO recommendations A divergence angle
Design deceleration lane, but between 2 and 5 degrees
Manual, not always. Additional length should be is acceptable.
June See Table 5-3.01A in provided on downgrades. See 1:15 - with 3.8 degree
2000. Road Design Table 2-5.09B in Road Design angle,
Manual. Manual. 1:5 – when a turn lane on
curve or when greater
storage length is needed.
Typical length for a turn lane is 90 m (300′) of full width (3.6 m + 0.6 m (14′) where
curb or gutter are presented) + additional 54 m (180′) taper section (1:15 from 3.6 m
(12′)) + additional length for downgrades.
HCM Exclusive left-turn lane should be provided for fully protected left-turn phasing
Single exclusive left-turn lane – min Turn Volume = 100 veh/hour
Double exclusive left-turn Lane – min Turn Volume = 300 veh/hour
Exclusive right-turn lane – min Turn Volume = 300 veh/hour
* Speed Differential is the difference between the speeds of the through traffic and turning traffic
** This method of taper length estimation is mentioned in other manuals and handbooks as well and is still in
use
11
The desirable length of storage should be enough to accommodate twice the average arrival rate
of turning traffic during the red cycle. The storage length requirement for left turn lanes is based
on peak 15-min flow rates. The following formulas are based on the probability of event
occurrence and are offered in Traffic Engineering Handbook (1999) to estimate the length of
storage bay for a left-turn lane (the length of a typical passenger car is considered to be 25 ft):
L = V*K*25*(1+p) / Nc
where L – storage length, in feet,
V – peak 15-min flow rate, in vph,
K – constant to reflect random arrival of vehicles, usually 2,
Nc – number of cycles per hour, and
P – percentage of trucks and buses.
For evaluation of right-turn lanes, the formula is
L = (1-G/C)*Volume*(1+p)*K*25 / Nc*n
where
L – storage length, in feet,
G – green time,
C – cycle length,
K – random arrival for vehicles equals 2 when right-turn-on-red is not permitted and 1.5 when
right-turn-on-red is allowed,
n – number of traffic lanes,
V – peak 15-min flow rate, in vph,
K – constant to reflect random arrival of vehicles, usually 2,
Nc – number of cycles per hour, and
P – percentage of trucks and buses.
Figure 7 and Figure 8 present a summary of warrants from recent research represented in the
Traffic Engineering Handbook.
12
Source: Traffic Engineering Handbook, 1999
Figure 7. Left-Turn Lane Warrants
13
Source: Traffic Engineering Handbook, 1999
Figure 8. Right-Turn Lane Warrants
14
3. DEVELOPMENT OF CALIBRATED SIMULATION MODELS
3.1 Selection and Documentation of Intersections

A total of ten intersections from the several evaluated intersections were selected and designated
with the help of the MN DOT project panel (see Figure 9). The intersections were initially
grouped by location to expedite data collection and minimize travel time.
Figure 9. Overall map of area
15
Two types of intersections were chosen for the study: Signalized (7 intersections) and
Unsignalized (3 intersections). The main characteristics of the intersections are as follows:
− The majority have channelization (island design).
− A few have grades that were observed on the field study but were not documented in the
construction drawings provided by MN DOT.
− Signalized intersections are actuated-uncoordinated with non-constant cycle length.
− Signalized approaches have permitted and/or protected phases for left turns and
permitted and free phases for right turns.
The main characteristics of the intersections are presented in
Table 5. The intersections were documented by using existing construction drawings and signal
and timing plans provided by MN DOT, and video recording data and photographs taken during
the field survey. Geometry data of the intersections such as lane width, number of lanes, existing
storage length, and detector location were obtained from construction plans and videotapes.
Geometrical models of the intersection were built (Appendix C). Signal and timing plans
provided information about cycle length, turn type, maximum and minimum splits, etc. The
following list represents the geometric design and traffic control variables:
− Number of through lanes
− Number of turn lanes
− With or without channelization for right turn
− Approach grades
− Through and turn lane widths
− Total turn lane length (taper, deceleration, and storage length)
− Type of traffic control
− Type of turn phasing (if signalized)
− Warning signs
− Posted speed
− Turning speed (a speed inside the intersection; default is 15 mph for left turns and 9
mph for right turns)
− Detector location
16
Table 5. Intersection descriptions from the field study
Intersection Name Lat/Lon Signal Type Description

North of St. Paul
N452637.2
SR-169 & CR-4 W933505.3 Signalized 4 free rights with channelization
N452934.0
SR-65 & SR-5 (Isanti) W931414.8 Signalized 4 free rights with channelization
TH 10 at Big Lake Industrial N455559.5
W934307.6 Unsignalized 4 free rights with channelization
Park
N452440.6 4 lane with right turn lanes into side
TH 169 & CR25 W933359.5 Unsignalized roads
Near Brainard/Baxter
N462308.1 4 lane with right turn lanes into side
SR-371 & CR-125 W941710.4 Unsignalized roads, short turn lanes
N463000.7
SR-371 & CR77-CR49 W941752.1 Signalized Rural intersection of two highways
St. Paul Metro Area
N445559.5 2 islands with some grade on a turn
SR-7 & CR-73 W932517.1 Signalized lane
N445617.2 4 free rights with channelization
SR-7 & Louisiana Ave W932212.3 Signalized and some grade
N445328.0
TH 3 (Robert ST) & Marie Ave W930447.2 Signalized Short left turn lane
N450114.6 2 islands. Roads intersect at 45
SR36 & SR120 (Century Ave) W925903.3 Signalized degrees not perpendicular
Traffic characteristics of each intersection were identified from the videotapes and include:
− Volume of approach
− Volume of turning traffic
− Heavy vehicle mixture
− Queue length in two-minute intervals
All geometric design, traffic characteristics, and control data were collected for each study
intersection and were tabulated for each individual intersection approach (Appendix D).
3.2 Queue Length Observation in Field

The major data sources for this research included traffic data obtained by means of an Automatic
Traffic Recording (ATR) station, road geometry data from construction drawings, and
photographs and video taken during the field survey during the visit to each intersection site. The
major apparatus components of field data collection included:
− a trailer with a video mast and video recording equipment;
17
− measuring wheels and equipment to collect dimensional information not apparent in the
drawings or not available;
− a LIDAR gun to collect speed information: the traffic turn speeds as well as average
traffic speeds; and
− basic photographic equipment to have an overview of the intersection layout and
placement of equipment.
To capture traffic data such as volumes, queue lengths, heavy traffic mixture, and headways, the
station with video recording equipment was used. The video recording trailer was situated at an
unobtrusive location at the selected intersections. The location was optimized to give views of
the intersection so it would be possible to evaluate turn lanes, queue lengths, overall volumes,
etc. The trailer was equipped with four cameras on remote pan tilt mounts that allowed
orientating cameras and recording video of the different traveling directions of the intersection.
Video recordings of all intersections initially were broken into three field missions. The first
recording of the intersection in the north of the St. Paul metro area (Appendix B.1, Figure B.1.1)
was done in the first week of November 2005. At the first four intersections (Appendix B.1,
Figures B.1.2-B.1.5), 8 hours of video at each were collected. Analyzing the obtained data, we
concluded that we got more information than was needed. For our research, we needed only one
hour of peak time traffic information. To save time and travel costs on the follow-up trips, we
decided to collect a minimum of one hour during peak daytime conditions. Additional data
would be collected as needed. Thus, the traffic was recorded at the rest of the intersections
during the daytime peak hours for the next batches of intersections in the Brainard/Baxter area
(Appendix B.2) and St. Paul metro area (Appendix B.3). The video recordings of intersections in
the abovementioned areas were done in the second week of April 2006 (Appendix B, Figures
B.2.1-B.2.3 and B.3.1- B.3.5).
Extraction of traffic data during a one-hour run such as through volume and turning volumes on
each approach, number of heavy vehicles on each type of lane (through, left turn, and right turn),
volume, driver behavior data such as headways, and queue lengths on turn lanes was done from
the video records. Queue lengths of average, 95th percentile, and maximum in two-minute
intervals during a one-hour run were calculated for each bound on left and right turns (Appendix
E). Field data of ten selected intersections were collected to build, calibrate, and exercise models
in simulation programs.
3.3 Analysis of Field Data

To build a simulation model of each intersection and exercise them to make a database for the
regression analysis, it was reasonable to group intersections into categories (Table 6) according
to the
− type of signal control;
− type of mode on signalized intersections;
− type of sign on unsignalized intersections;
18
− number of link directions;
− number of through lanes by approach;
− number of auxiliary lanes by approach.
Thus, the major groups are Signalized and Unsignalized. All signalized and unsignalized
intersections have four lanes on an arterial road, are two-way, and have one auxiliary lane for
each turn. The Signalized group was divided according to the type of mode or sign on the arterial
road into the following subgroups (Figure 10):
for Left Turns
− Protected Left Turn
− Protected + Permitted Left Turn
for Right Turns
− Free Right Turn
− Permitted Right Turn
The Unsignalized group was divided according to the arterial road’s type of sign control into the
following groups:
for Left Turns
– Yield Left Turn
for Right Turns

− Free Right Turn
− Yield Right Turn
19
Table 6. Description of intersections for categorization
# Intersection Arterial Road Crossroad Number of

Turn signal type Number Turn signal type Number approaches
# (Left/Right) of lanes (Left/Right) of lanes
UNSIGNALIZED
TH 10 &
1
Industrial Yield/Free Yield/Free 4
Park
TH 169 &
2 CR25
Yield/Free Stop 3
SR-371 &
3 Yield Stop 4
CR-125
SIGNALIZED
SR-169 &
1 Prot/Free Prot/Free 4
CR-4
SR-65 &
2 Prot/Free Prot + Perm/Free 4
SR-5 (Isanti)
SR-371 &
3 Prot/Perm Perm/Perm 4
CR77-CR49
Robert ST &
4 Prot + Perm/Perm Prot/Perm 4
Marie Ave
SR36 &
5 Prot + Perm/Free Prot/Perm 4
SR120
SR-7 &
6 Prot/Perm Prot/Free 4
CR-73
SR-7 &
7 Prot/Free Prot/Free 4
Louisiana Ave
20
INTERSECTION
Signalized Unsignalized
Left Turn Right Turn Left Turn Right Turn
Protected Permitted Permitted Free Yield Free Yield
Figure 10. Classification chart
3.4 Simulation Model Calibration

All the data about each intersection was collected and tabulated (Appendix D). The macroscopic
simulation model SYNCHRO (to govern vehicle movement) and the microscopic simulation
model SimTraffic (to perform characteristics of the driver, vehicle performance, and vehicles
interactions with network geometry) were chosen to build intersections and simulate traffic on
them. To calibrate the simulation models for the purpose of correctly predicting the real-world
traffic conditions, we adjusted the following parameters:
− heavy vehicle mixture that could affect the acceleration rate and consequently queue
length;
− driver behavior parameters through headways and turning speeds;
− roadway parameters through increasing storage lane length by adding a portion of the
taper, adjusting saturated flow rates, and designing acceleration lanes;
− signal timing data such as cycle length, max split, and minimum gap.
To design the channelization on the right turns, we tried two types intersection design: with and
without acceleration lanes (Figure 11). The first model without acceleration lanes was a better fit
for the real-world scenario of signalized intersections. By activating the “Right Turn
Channelized” option in the SYNCHRO signing window for unsignalized intersections, the
models performed better as well.
21
Figure 11. Channelization design on the right turns in SYNCHRO
The following features of the simulation packages should be taken into consideration when
evaluating the results of the calibration outputs:
− The vehicle is considered queued when its speed is less than 7 mph. We calculated the
stopped vehicles and queued behind them because we could not differentiate the speed
value from the videotapes.
− SimTraffic reports the queue length based on the distance that the model run observes.
We calculated the number of cars in the queue and then multiplied that number by the
average car length.
− Vehicle length is considered 25 ft for passenger cars. Since average vehicle length has
changed with the tendency to increase (SUVs and pickup trucks replacing some of
smaller cars that were popular before), the distance of 25 ft per vehicle (average vehicle
length for passenger cars, SUVs, and vans is around 15 ft) including 12 ft of inter-
vehicle gap is short (Long, 2002). On the other hand, some preceding studies verify that
the mentioned length is reasonable. At any rate, since SYNCRO considers this value in
its calculations, we follow this rule.
− Three types of queue length are reported: average queue—average of maximum queue
length in the two-minute intervals; 95th percentile queue—average queue + 1.65
standard deviations for two-minute intervals (this queue length might not be observed
during an actual simulation); and maximum queue—the longest queue length observed
in two-minute intervals. For calibration purposes, we compared the average queue
length of the simulation model with the real-world scenario.
To evaluate the accuracy of the simulation models, we compared the simulation models’ average
queue lengths on the turn lane with the real-world queue lengths obtained from the video (in feet)
and calculated their difference expressed in percents. This evaluation was not accurate in the
situations of short queue length. In those cases, even a very short difference in queue lengths
between the real-world and simulation models caused significant differences, which does not
22
express real estimation. To elaborate the estimation, the difference of the turn lane queue lengths
was calculated in cars as well. Table 7 shows the accuracy of our simulation model of the
intersection TH 169-CR 4. The calibrated simulation model has less than one car difference with
the real-world model.
All simulation models were evaluated in the same manner (Table 8). For exercising purposes,
models where the difference between the average queue length of the simulation model and the
real-world queue length does not exceed one car or is very close to one car (in the case of SR-65
& SR-5 (Isanti) intersection) were preferred. Thus, for exercising purposes the following
intersections were chosen:
− TH 10 & Industrial Park
− TH 169 & CR25
− SR-371 & CR-125
− SR-169 & CR-4,
− SR-65 & SR-5 (Isanti),
− SR-371 & CR77-CR49
− Robert St. & Marie Ave
Table 7. Evaluation of the accuracy of the simulation model of the intersection TH 169-CR 4
EB WB NB SB
Evaluation value
L R L R L R L R
Difference (in %) of average queue
length between simulated and real- 1% 26% 23% -55% -12% 49% 16% -53%
world models
Difference (in cars) of average
queue length between simulated 0.0 0.2 0.6 -0.3 -0.9 0.3 0.4 -0.7
and real-world models
Standard deviation of queue length
(in cars) obtained from video in 2 2.0 0.9 1.6 0.8 3.4 0.7 1.9 1.8
min intervals
Standard deviation of average
queue length (in cars) for 0.4 0.3 0.4 0.2 0.6 0.3 0.3 0.1
simulation model
Standard deviation of 95th
percentile queue length (in cars) 0.7 1.2 1.1 1.2 1.2 1.0 0.6 0.3
for simulated model
23
Table 8. Evaluation of the accuracy of the simulation models: difference (in cars) between
average queue length of the simulation model and queue length obtained from the
videotapes
EB WB NB SB
## Intersection Name
L R L R L R L R
UNSIGNALIZED
1 TH 10 & Industrial Park 0.0 0.2 0.1 0.0 0.0 0.3 -0.1
2 TH 169 & CR25 -0.1 -0.7 -0.9
3 SR-371 & CR-125 -0.5 -0.3 0.1 0.0 -0.7 -0.7
SIGNALIZED
1 SR-169 & CR-4 0.0 0.2 0.6 -0.3 -0.9 0.3 0.4 -0.7
2 SR-65 & SR-5 (Isanti) -0.7 1.1 0.4 0.3 -1.2 0.0 0.6 1.1
3 SR-371 & CR77-CR49 0.4 -0.9 0.4 -1.1 -1.2 0.7 -0.6 0.4
4 Robert ST & Marie Ave -0.1 0.1 0.2 -0.1 0.2 -0.2
5 SR36 & SR120 -0.9 0.3 -0.5 -1.5 -1.2 -0.5 -1.4 -0.3
6 SR-7 & CR-73 -1.0 2.6 0.0 1.9 -1.4 -0.2 -1.0 -1.5
7 SR-7 & Louisiana Ave -1.0 -3.3 -2.8 -2.0 -0.9/-0.7* -1.4 -1.5/-2.3* -1.0
Note: * There are two left turn lanes on this approach.
24
4. GENERATION OF QUEUE LENGTH PREDICTION MODEL
4.1 Simulation Model Parameters

The factors that the road designers are interested in were discussed at the meeting with the panel
from MN DOT. The list of those parameters is as follows: Through Volume (TV), Opposing
Volume (OV), Crossing Volume (CV), Left Turn Volume (LTV), Right Turn Volume (RTV),
Heavy Vehicle Through Percent (HVT), Heavy Vehicle Left Turn Percent (HVL), Heavy
Vehicle Right Turn Percent (HVR), Grade (Gr), and Speed (Sp) (Figure 12). The effect of each
parameter as well as their combination was considered in the sensitivity analysis. Simulations
were conducted in the following ranges:
− Through Volume range – we started with the range from 200 to 800 vehicles per hour
(vph) with increments of 100 vph, and then made the decision to change the range from
300 to 1100 vph with increments of 300 vph, taking into account that the arterial roads
are four-lane and that the regression model reveals a good fit
− Opposing Volume range – the same as the Through Volume
− Crossing Volume range – from 50 to 350 vph with increments of 100 vph, taking into
account that the opposing approaches are two-lane
− Left Turn Volume range – from 0 to 300 vph with increments of 50 vph
− Heavy Vehicle Through Percent – from 0 to 25% with increments of 5%
− Heavy Vehicle Left / Right Turn Percent – from 0 to 25% with increments of 5% at the
outset
− Grade – from -4% to +4% with increments of 1%; increments of 0.5% were selected at
the meeting, but SYNCHRO accepts only integers
− Speed – from 30 to 70 mph with increments of 5 mph
25
Figure 12. Sketch of volumes on the intersection
Exercising of generic models based on the calibrated real-world models for sensitivity analysis to
determine the most optimum turn lane length was done. Thus, the queue length is considered to
be a dependent variable, and the list of parameters upon which the prediction is based is a set of
independent variables.
The simulation model was run with five different random numbers of seeds to determine the
average values of the average, maximum, and 95th percentile queue lengths for each set of
inputs. The results were compiled into a spreadsheet and averaged for each alternative. A total of
14,170 simulation runs was performed. All data were tabulated to be examined in the MINITAB
statistical software program.
4.2 Simulation Model Data

To assess the influence of each input on the queue length, we used multivariate regression
analysis to define the queue length as a function of the independent variables. The regression
models were generated on the basis of the simulation run data for each turn lane on the
intersections. The regression model approach is convenient for practitioners and at this time
appears to result in a good fit (Table 9). One of the quantitative characteristics of the degree of fit
of a regression equation is a coefficient of determination. The percent of the variations that can
be explained by the regression equation could be assessed by the value of the coefficient of
determination (R2).
26
Table 9. Coefficient of determination for the single intersection in each group
Left Turn
Intersection Approach COEFFICIENT OF
DETERMINATION (R2 )
Protected Signalized
SR-169 & CR-4 North 97.9%
SR-169 & CR-4 (2-lane) West 98.3%
SR-7 & CR-73 West 88.7%
Robert ST & Marie Ave South 97.3%
SR-371 & CR77-CR49 South-East 89.9%
Permitted Signalized
SR-169 & CR-4 South 95.6%
SR-65 & SR-5 (Isanti) North 66.9%
Yield Unsignalized
TH 10 & Industrial Park North 81.3%
TH 169 & CR25 West 90.5%
SR-371 & CR-125 North 84.7%
Right Turn
Permitted Signalized
SR-7 & CR-73 East 98.8%
SR-371 & CR77-CR49 North-East 89.5%
Free Signalized
SR-169 & CR-4 North 86.0%
SR-65 & SR-5 (Isanti) North 88.9%
Free Unsignalized
TH 10 & Industrial Park East 86.7%
SR-371 & CR-125 South 90.7%
Yield Unsignalized
TH 10 & Industrial Park West, East 73.1%, 79.8%
TH 169 & CR25 South 59.7%
SR-371 & CR-125 South 94.0%
In most cases, the coefficient of determination is greater than 85%, which means that the
regression models relatively accurately explain all of the variation in the response and can
predict the average queue length for each intersection. By applying the regression analysis to the
different intersections, groups of intersections, and their combinations, we get the following
results:
27
− Each intersection shows good correlation, and the coefficient of determination is equal
to or greater than 80% in 18 cases out of 21.
− Each group of intersections according to our classification shows good correlation, and
the coefficient of determination is greater than or equal to 85% (Table 10).
− The regression model of the combination of Permitted left turns for Signalized (S)
intersections and Yield left turns for Unsignalized (U) intersections shows the
acceptable fit to the function with the coefficient of determination equal to 78.4%.
− The regression model of the combination of Permitted left turns and Protected left turns
for Signalized (S) intersections shows the acceptable fit to the function with the
coefficient of determination equal to 76.2%.
− The regression model of the combination of Free right turns and Yield right turns for
Unsignalized (U) intersections shows good fit to the data with the coefficient of
determination equal to 86.2%.
Table 10. Coefficient of determination of the multivariate regression analysis of the average
queue lengths for each group of intersections and their combinations
LEFT RIGHT
Prot(S) Perm(S) Yield(U) PERM(S) FREE(S) FREE(U) Yield(U)
89.7% 91.3% 85.0% 85.7% 85.1% 85.7% 87.1%
76.2% 47.2%
78.4% 66.6%
72.1% 86.2%
46.4% *
Note: * Coefficient of determination of the combination of Permissive (Signalized) and Yield (Unsignalized)
intersections.
The equations for calculating the queue length derived from multivariate regression analysis are
presented below. The absolute value of T-statistics indicates which variables add the most
specifics to the regression model. The low value of P-statistics designates the significance of the
variable. The lower the P value and the higher the T-statistics, the more important the variable is.
28
Left Turns
For Protected left turns on Signalized intersections
Aver = -8.64 + 0.00618*TV + 0.881*LTV - 0.119*Sp - 2.24*HVT + 0.863*HVL + 0.055*Gr
(Parameters are defined in Section 4.1 )

Predictor T P
Const. -1.99 0.048
TV 1.63 0.104
LTV 70.09 0.000
Sp -1.58 0.116
HVT -5.77 0.000
HVL 5.39 0.000
Gr 0.09 0.929
S = 21.51; R2 = 89.8%; R2 (adjusted) = 89.7%
For Permitted left turns on Signalized intersections
Aver = 20.7 - 0.00159*TV + 0.0233*OV + 0.483*LTV - 0.580*Sp - 0.192*HVT + 0.382*HVL -

0.32*Gr
(Parameters defined in Section 4.1 )

Predictor T P
Const. 2.14 0.033
TV -0.28 0.781
OV 8.14 0.000
LTV 52.44 0.000
Sp -4.73 0.000
HVT -0.68 0.497
HVL 2.09 0.037
Gr -0.30 0.765
S = 12.70; R2 = 91.5%; R2 (adjusted) = 91.3%
29
For Yield left turns on Unsignalized intersections
Aver = -9.75 + 0.00179*TV + 0.0216*OV + 0.226*LTV - 0.0626*Sp + 0.249*HVT -

0.007*HVL + 0.204*Gr

Predictor T P
Const. -4.16 0.000
TV 1.20 0.233
OV 14.32 0.000
LTV 34.51 0.000
Sp -1.58 0.115
HVT 1.40 0.161
HVL -0.06 0.953
Gr 0.37 0.709
S = 5.987; R2= 85.5%; R2 (adjusted) = 85.0%
For the combination of Permitted left turns on Signalized and Yield turns on Unsignalized
intersections
Aver = -29.4 + 0.00248*TV + 0.0203*OV + 0.416*LTV + 0.133*Sp - 0.315*HVT + 0.922*HVL

+ 0.22*Gr

Predictor T P
Const. -5.25 0.000
TV 0.70 0.482
OV 6.85 0.000
LTV 38.82 0.000
Sp 1.50 0.134
HVT -1.03 0.303
HVL 5.27 0.000
Gr 0.20 0.843
S = 17.82; R2 = 78.7% ; R2 (adjusted) = 78.4%
30
For the combination of Permitted and Protected left turns on Signalized intersections
Aver = -7.69 + 0.0379*TV - 0.0409*OV + 0.859*LTV - 0.298*Sp - 2.92*HVT+ 0.106*HVL +

0.081*Gr

Predictor T P
Const. -1.25 0.210
TV 7.84 0.000
OV -8.97 0.000
LTV 55.73 0.000
Sp -3.04 0.002
HVT -5.80 0.000
HVL 0.53 0.596
Gr 0.09 0.932
S = 38.19; R2 = 76.4%; R2 (adjusted) = 76.2%
For the combination of Permitted and Protected left turns on Signalized and Yield left turns on
Unsignalized intersections
Aver = -6.24 + 0.0282*TV - 0.0565*OV + 0.799*LTV - 0.101*Sp - 2.78*HVT+ 0.269*HVL +

0.124*Gr

Predictor T P
Const. -1.07 0.284
TV 6.16 0.000
OV -17.23 0.000
LTV 50.63 0.000
Sp -1.60 0.288
HVT -5.22 0.000
HVL 1.37 0.172
Gr 0.13 0.900
S = 40.71; R2 = 72.3%; R2 (adjusted) = 72.1%
31
Right Turns
For Free right turns on Signalized intersections
Aver = - 24.1 + 0.0102 TV + 0.109 CV + 0.142 RTV - 0.153 Sp - 0.150 HVT+ 0.157 HVR +
0.176 Gr

Predictor T P
Const. -9.98 0.000
TV 4.37 0.000
CV 23.50 0.000
LTV 31.21 0.000
Sp -5.21 0.000
HVT -0.77 0.443
HVL 1.59 0.113
Gr 0.36 0.721
S = 6.571 R2 = 85.5% R2 (adjusted) = 85.1%
For Permitted right turns on Signalized intersections
Aver = 25.2 + 0.00929 TV + 0.0265 CV + 0.152 RTV - 0.367 Sp - 0.475 HVT + 0.375 HVR -
0.129 Gr

Predictor T P
Const. 9.28 0.000
TV 4.22 0.000
CV 5.81 0.000
LTV 29.92 0.000
Sp -9.97 0.000
HVT -4.48 0.000
HVL 3.23 0.001
Gr -0.36 0.719
S = 5.543 R2 = 86.2% R2 (adjusted) = 85.7%
32
For Yield right turns on Unsignalized intersections
Aver = 1.71 - 0.00184 TV + 0.0149 CV + 0.0456 RTV - 0.0077 Sp - 0.0497 HVT - 0.0360 HVR +
0.075 Gr

Predictor T P
Const. 1.49 0.139
TV -1.54 0.125
CV 10.66 0.000
LTV 24.97 0.000
Sp -0.52 0.603
HVT -1.22 0.223
HVL -1.96 0.051
Gr 0.52 0.606
S = 1.735 R2 = 87.6% R2 (adjusted) = 87.1%
For Free right turns on Unsignalized intersections
Aver = 2.65 - 0.00193 TV + 0.00981 CV + 0.0441 RTV - 0.0080 Sp - 0.0554 HVT- 0.0493 HVR
+ 0.045 Gr

Predictor T P
Const. 2.52 0.013
TV -1.76 0.080
CV 7.93 0.000
LTV 25.83 0.000
Sp -0.58 0.560
HVT -1.33 0.185
HVL -2.84 0.005
Gr 0.33 0.741
S = 1.621 R2 = 86.2% R2 (adjusted) = 85.7%
33
For the combination of Free and Permitted right turns on Signalized intersections
Aver = 7.75 + 0.00144 TV + 0.0470 CV + 0.129 RTV - 0.172 Sp + 0.243 HVT - 0.147 HVR +
0.142 Gr

Predictor T P
Const. 2.19 0.029
TV 0.45 0.655
CV 7.23 0.000
LTV 18.92 0.000
Sp -3.76 0.000
HVT 1.28 0.201
HVL -0.99 0.324
Gr 0.23 0.816
S = 12.50 R2 = 47.9% R2 (adjusted) = 47.2%
For the combination of Free right turns on Signalized and Unsignalized intersections
Aver = - 20.8 + 0.0192 TV + 0.0700 CV + 0.113 RTV - 0.0782 Sp - 0.512 HVT - 0.0111 HVR -
0.094 Gr

Predictor T P
Const. -7.35 0.000
TV 6.94 0.000
CV 15.76 0.000
LTV 21.86 0.000
Sp -2.18 0.030
HVT -2.98 0.003
HVL -0.14 0.888
Gr -0.19 0.853
S = 9.124 R2 = 67.1% R2 (adjusted) = 66.6%
34
For the combination of Yield right turns on Unsignalized and Permitted right turns on Signalized
intersections
Aver = 5.13 + 0.00707 TV + 0.0443 CV + 0.123 RTV - 0.215 Sp - 0.082 HVT + 1.44 HVR +
0.142 Gr

Predictor T P
Const. 0.92 0.357
TV 1.50 0.135
CV 5.30 0.000
LTV 12.34 0.000
Sp -2.91 0.004
HVT -0.39 0.699
HVL 7.94 0.000
Gr 0.19 0.847
S = 13.21 R2 = 47.7% R2 (adjusted) = 46.4%
For the combination of Yield and Free right turns on Unsignalized Intersections
Aver = 2.34 - 0.00203 TV + 0.0121 CV + 0.0449 RTV - 0.0083 Sp - 0.0550 HVT- 0.0442 HVR +
0.0530 Gr

Predictor T P
Const. 2.99 0.003
TV -2.51 0.013
CV 13.10 0.000
LTV 35.70 0.000
Sp -0.82 0.413
HVT -1.90 0.058
HVL -3.49 0.001
Gr 0.53 0.595
S = 1.691 R2 = 86.5% R2 (adjusted) = 86.2%
35
5. TURN LANE LENGTH RECOMMENDATIONS
The length of the turn lanes includes three components: taper, deceleration length, and storage
length (Figure 13). The turn lane begins with a taper, the design of which depends on location
and traffic characteristics. The AASHTO, the HCM, and the other guidelines specify the taper
length as a ratio of 8:1 and 15:1 for design speeds up to 30 mph and up to 50 mph, respectively.
Based on this recommendation, the length of the taper should be in the range from 96 to 180 ft
for low-speed and high-speed roadways, respectively (the width of the turning lane is considered
12 ft). The use of taper length, such as 100 ft for a single-turn lane and 150 ft for a dual-turn lane
for urban streets, is mentioned in the AASHTO guideline. Provided that the deceleration distance
is determined in the high-speed conditions, the recommendation of the shorter taper design from
100 ft to 130 ft is mentioned in Traffic Engineering Handbook (1999). We use this
recommendation in our calculation.
Figure 13. Components of turn lane
The estimation of the deceleration lane length is presented in the Traffic Engineering Handbook
in tabular form. Table 11 and Table 12 demonstrate the distance to decelerate from through
traffic to a stop while moving laterally according to the design speeds in desirable conditions and
in limiting conditions. To follow the AASHTO recommendations, the limiting conditions of the
perception-reaction time would be considered in our calculations. In addition, we will consider
the deceleration distances for right and left turns separately, although they look almost identical.
We will not consider the distance traveled during perception-reaction time that starts before
approaching the taper as part of the deceleration movement.
36
Table 11. Deceleration length for right-turning traffic
Design Deceleration Distance (ft)

Speed
Traffic Engineering Handbook AASHTO***
(mph)
Desirable* Conditions Limiting** Conditions Limiting Conditions
30 225 170 170
35 295 220
40 375 275 275
45 465 340 340
50 565 410 410
55 675 485 485
60 785 565
* Desirable conditions (2.0 sec perception-reaction time; average deceleration 3.5 fps2 laterally into turn lane and 6.0
fps2 thereafter; speed differential <10mph)
** Limiting conditions (1.0 sec perception-reaction time; average deceleration 4.5 fps2 laterally into turn lane and
9.0 fps2 thereafter; speed differential <10mph )
*** The lengths are based on grades less than 3 percent.
Source: Traffic Engineering Handbook, 1999, (p. 328) and AASHTO, 2004, (p. 714)
Table 12. Deceleration length for left-turning traffic
Design Speed Deceleration Distance (ft)

(mph) Traffic Engineering Handbook AASHTO
30 170 170
40 275 275
45 340 340
50 410 410
55 485 485
Source: Traffic Engineering Handbook, 1999 and AASHTO, 2004
37
Thus, according to the Traffic Engineering Handbook assumption, the vehicle starts deceleration
in the taper (lateral movement) and fully decelerates in the deceleration segment of the turn lane
(Figure 14).
Source: Traffic Engineering Design, 1999, (p. 328)
Figure 14. Right-Turn Deceleration Bay
Estimation of the storage length for the Signalized and Unsignalized intersections is important to
predict the overflow or blockage of the turn lane. The existing methodologies (HCM, Traffic
Engineering Handbook) and practices offer to design the storage length taking into account at
least a 95% probability of storing all turning vehicles during the peak hour. The 95th percentile
queue is the queue length that has only a 5 percent probability of being exceeded during the
analysis period. For sample data from our simulation models, refer to Appendix G. Although it is
not typical for average estimation, using the 95th percentile queue length will maintain better
performance of the intersection. Thus, for design purposes, it is more reasonable to analyze the
95th percentile queue length. The coefficients of determination of the regression analysis for the
95th percentile queue length as a function of independent variables are given in Table 13.
38
Table 13. Coefficient of determination of the multivariate regression analysis of the 95th
percentile queue lengths for each group of intersections and their combinations
LEFT RIGHT
Prot(S) Perm(S) Yield(U) Perm(S) Free(S) Free(U) Yield(U)
83.3% 86.4% 82.3% 75.0% 81.1% 80.2% 80.2%
69.1% 58.9%
78.5% 63.0%
65.7% 79.8%
25.5% *
* Regression analysis of the combination of Permissive (Signalized) and Yield
(Unsignalized) intersections.
To maintain the “safety” and “accuracy” factors, we would suggest using the “first level” of our
classification—grouping by the operational modes for Signalized intersections and type of sign
for Unsignalized intersections. The coefficient of determination for those groups is not less than
75%, although the combination of Free and Yield right turns for Unsignalized intersections and
the combination of Permissive left turn for Signalized and Yield left turn Unsignalized show
almost 80% accuracy. The whole list of equations is presented in Appendix F.
5.1 Left Turn Lane Length (Taper + Deceleration + Storage)

Left turn lane length is a sum of three components presented in Table 14. The equations for
calculating the storage bay for the left turn lane, which is the 95th percentile queue length
derived from multivariate regression analysis (parameters are defined in Section 4.1 ), are
presented below:
For Protected left turns on Signalized intersections
LTprot = 35.3 + 0.0203*TV + 1.14*LTV - 0.171*Sp – 6.75*HVT+ 1.32*HVL - 0.16*Gr
T P
Const. 4.71 0.000
TV 3.10 0.002
LTV 52.84 0.000
Sp -1.31 0.190
HVT -10.08 0.000
HVL 4.77 0.000
Gr -0.15 0.884
S = 37.05; R2 = 83.5%; R2 (adjusted) = 83.3%
39
For Permitted left turns on Signalized intersections
LTperm = − 45.2 - 0.00953*TV + 0.0406*OV + 0.610*LTV + 0.348*Sp + 0.812*HVT+

1.76*HVL + 0.35*Gr
T P
Const. -6.09 0.000
TV -2.10 0.037
OV 10.47 0.000
LTV 42.96 0.000
Sp 2.76 0.006
HVT 1.69 0.091
HVL 7.14 0.000
Gr 0.22 0.823
S = 22.16; R2 = 86.6%; R2 (adjusted) = 86.4%
For Yield left turns on Unsignalized intersections
LTyield = 0.00 + 0.00315*TV + 0.0332*OV + 0.345*LTV - 0.149*Sp + 0.224*HVT+

0.629*HVL - 0.080*Gr
T P
Const. 0.00 0.999
TV 1.23 0.220
OV 12.88 0.000
LTV 30.85 0.000
Sp -2.21 0.028
HVT 0.74 0.460
HVL 2.93 0.004
Gr -0.09 0.932
S = 10.23; R2 = 82.8%; R2 (adjusted) = 82.3%
40
Table 14. Three components of the left turn lane length
LEFT TURN LANE LENGTH

Speed
Deceleration Taper Storage Length
30 170 100
LTprot = 35.3 + 0.0203*TV + 1.14*LTV - 0.171*Sp –
35 170 100 6.75*HVT+ 1.32*HVL - 0.16*Gr
40 275 130
45 340 130 LTperm = − 45.2 - 0.00953*TV + 0.0406*OV +
0.610*LTV + 0.348*Sp + 0.812*HVT+ 1.76*HVL +
50 410 130
0.35*Gr
55 485 130
60 485 130 LTyield = 0.00 + 0.00315*TV + 0.0332*OV +
65 485 130 0.345*LTV - 0.149*Sp + 0.224*HVT+ 0.629*HVL -
0.080*Gr
70 485 130
41
5.2 Right Turn Lane Length (Taper + Deceleration + Storage)
Right turn lane length as a sum of three components is presented in Table 15. The formulas to
calculate the storage bay for the right turn lane, which is the 95th percentile queue length derived
from multivariate regression analysis, are presented below:
For Permissive right turns on Signalized intersections
RTperm = 65.5 + 0.0323*TV + 0.0533*CV + 0.186*RTV - 0.829*Sp - 1.50*HVT + 0.818*HVR +

0.107*Gr
Predictor T P
Constant 10.28 0.000
TV 6.28 0.000
CV 4.99 0.000
RTV 15.62 0.000
Sp -9.62 0.000
HVT -6.05 0.000
HVR 3.02 0.003
Gr 0.13 0.899
S = 12.98 R2 = 75.9% R2 (adjusted) = 75.0%
For Free right turns on Signalized intersections
RTfree_S = - 61.5 + 0.0551*TV + 0.265*CV + 0.317*RTV - 0.527*Sp + 0.041*HVT + 1.04*HVR

+ 0.47*Gr
Predictor T P
Constant -9.05 0.000
TV 8.36 0.000
CV 20.22 0.000
RTV 24.78 0.000
Sp -6.35 0.000
HVT 0.07 0.941
HVR 3.72 0.000
Gr 0.34 0.735
S = 18.50 R2 = 81.5% R2 (adjusted) = 81.1%
42
For Yield right turns on Unsignalized intersections
RTyield = 16.0 - 0.00497*TV + 0.0235*CV + 0.0957*RTV - 0.115*Sp - 0.038*HVT + 0.274*HVR

+ 0.223*Gr
Predictor T P
Constant 5.12 0.000
TV -1.54 0.125
CV 6.23 0.000
RTV 19.38 0.000
Sp -2.87 0.005
HVT -0.34 0.732
HVR 5.51 0.000
Gr 0.57 0.569
S = 4.689 R2 = 81.0% R2 (adjusted) = 80.2%
For Free right turns on Unsignalized intersections
RTfree_U = 13.2 − 0.00021*TV + 0.0176*CV + 0.0876*RTV - 0.0615*Sp - 0.094*HVT +

0.169*HVR + 0.164*Gr
Predictor T P
Constant 5.15 0.000
TV -0.08 0.938
CV 5.86 0.000
RTV 21.20 0.000
Sp -1.85 0.066
HVT -0.94 0.349
HVR 4.02 0.000
Gr 0.50 0.616
S = 3.927 R2= 81.0% R2 (adjusted) = 80.2%
43
Table 15. Three components of the right turn lane length
RIGHT TURN LANE LENGTH

Speed
Deceleration Taper Storage Length
30 170 100 RTperm = 65.5 + 0.0323*TV + 0.0533*CV + 0.186*RTV -
0.829*Sp - 1.50*HVT + 0.818*HVR + 0.107*Gr
35 220 100
40 275 130 RTfree_S = - 61.5 + 0.0551*TV + 0.265*CV + 0.317*RTV -
45 340 130 0.527*Sp + 0.041*HVT + 1.04*HVR + 0.47*Gr
50 410 130 RTyield = 16.0 - 0.00497*TV + 0.0235*CV + 0.0957*RTV -
55 485 130 0.115*Sp - 0.038*HVT + 0.274*HVR + 0.223*Gr
60 565 130
RTfree_U = 13.2 − 0.00021*TV + 0.0176*CV + 0.0876*RTV
65 565 130
- 0.0615*Sp - 0.094*HVT + 0.169*HVR + 0.164*Gr
70 565 130
44
6. DISCUSSION AND CONCLUSIONS
The total length of the turn lane is a sum of three components: taper, deceleration length, and
storage length. The equations to calculate the storage length were derived by means of regression
analysis of the data generated through exercising the simulation models of selected intersections.
The other two components of the turn lane—taper and deceleration length—were added
according to the design guidelines.
The general characteristics of the intersections that were chosen are:
− 4-lane and 2-way on arterial road and one turn lane for each turn
− 2-lane and 2-way on crossing road and one turn lane for each turn
− with or without channelization design on right turns
According to our classification, we get the correspondences of storage length as a dependent

variable from the independent variables for each type of intersection. The mentioned results are
applicable for the following range of parameters:
− Volume per Through lane – up to 500 vph
− Turn Volume – up to 250 vph
− Heavy Vehicle Percent – up to 20-25%
− Speed – up to 70 mph
− Grade – for range -4% to +4%
The abovementioned ranges are correlated with the HCM (2000) guidelines, which state that
enough main roadway lanes should be provided to prevent the total of the through plus right-turn
volume (plus left-turn volume, if present) from exceeding 450 veh/h/ln. Higher volumes can be
accommodated on major approaches if a substantial portion of available green time can be
allocated to the subject approach. Where left turn volumes exceed 300 vph, provision of a double
left turn lane should be considered. Thus, when the through and turning volumes exceed the
mentioned values, the recommendation could be an additional lane design consideration or
changes in signal timing.
The influence of each independent variable within the ranges that we investigated is as follows:
For Left Turns:

1. The most important and significant factors are Through Volume, Opposing Volume, Left
Turn Volume, Heavy Vehicle Left Turn Percent, and Speed.
significant for the Permitted turns on Signalized and Yield turns on Unsignalized
intersections.
45
3. The influence of Speed factor exists but does not appear to be significant for Protected
turns on Signalized intersections.
4. The Grade factor does not appear to affect the left turn queue length.
For Right Turns:

1. The most important and significant factors are Through Volume, Crossing Volume,
Right Turn Volume, Heavy Vehicle Right Turn Percent, and Speed.
significant for Free turns on both types of intersections and Yield turns on Unsignalized
intersections.
3. The influence of Through Volume on Free right turns on Unsignalized intersections does
not appear to affect the right turn queue length.
4. The Grade factor does not appear to affect the right turn queue length.
We tried to improve the performance of regression models by removing the factors that have a
weak influence. The coefficient of determination decreased, which means that even weak
influence is still important for accurate performance. To calculate the total length of the turn
lane, the taper and deceleration lengths that match the design speed should be added.
The whole procedure of calculations is divided into the following steps:
− Identification of the type of turn: Left or Right
− Identification of the type of Intersection: Signalized or Unsignalized
− Identification of the type of mode or sign control: Protected, Permitted, Yield, or Free
− Calculation of Storage length for defined type of turn lane
− Identification of the Speed value to define the sum of taper and deceleration length
− Calculation of the total length of turn lane by adding the defined storage length to the
sum of the taper and deceleration length
The flow chart of the turn lane length evaluation process is presented in Figure 15. As an
example of the calculations, the values of the total lengths for different through and turning
volumes and turning vehicles mixture (Heavy Vehicles Through percent is fixed and equals 5,
Grade is 0) are presented in tabular form in Appendix G.
The turn lane design practice uses standard turning length, which works well in many cases: the
turn lane length is equal to 300’ plus 180’ taper length. In our calculations, we applied the
guidelines to determine a deceleration distance that exceeds the abovementioned value for high
speeds. It could be possible to reduce the deceleration distance taking into consideration that
− during the peak period, speeds are considerably less than in the off-peak period, and the
deceleration length could be less than suggested values;
46
− there is an assumption that some deceleration takes place within the through lane prior
entering the turn lane.
At any rate, there is no recommendation on how much the deceleration length could be reduced.
It could happen that the intersection would not have a queue most of the time. This was obtained
in two cases of the selected intersections, and the assumption could be that the storage bay is not
required. Our models are not based on single cases. To generate the regression models, at least
two different intersections from the same category were used. Our guideline would indicate one
car length worth of storage. Incidentally, the recommendation of many guidelines is that the
minimum storage length should be enough to store two cars. In those cases, engineering
judgment needs to be used if the engineer feels that even one-car storage is too much.
The limitation of the results is that the data for this study were collected in the same geographic
area. The characteristics of driver behavior and heavy vehicle mixture that were chosen for the
calibration of generated models could vary for other regions. Another constraint that should be
taken into account is that our results are based on the four-lane arterial road model. We exercised
the same generic model by varying parameters on the minor two-lane road of the SR-169 & CR-
4 intersection for Protected left turn. The regression analysis shows a good fit to the multivariate
equation for the queue length for the arterial four-lane road model with a coefficient of
determination of 83.9%. However, we do not have enough data to predict with accuracy the
queue length on two-lane arterial roads.
47
INTERSECTION
Identify No
Left Right
Turn
Yes Yes
Identify No No
Intersection Signalized Unsignalized Signalized Unsignalized
Yes Yes
Yes
Identify Use LT prot form ula Protected Free Use RT free_ U form ula
Mode/Sign Yes
No
No
Use L T perm form ula Perm itted Yield Use RT yield form ula
Use LT yield form ula Yield Yes

Perm itted Use RT perm form ula
Identify Sp =< 30 add 270 No

Speed
Free Use RT free_ S form ula
30 < Sp =< 40 add 405
add 270 Sp =< 30

40 < Sp =< 45 add 470
add 320 30 < Sp =< 35

45 < Sp =< 50 add 540
add 405 35 < Sp =< 40

Sp > 50 add 615
add 470 40 < Sp =< 45
add 540 45 < Sp =< 50
add 615 50 < Sp =< 55
add 695 Sp > 55
Figure 15. The flow chart of the turn lane length evaluation procedure
48
REFERENCES
Chakroborty P., Kikuchi S., and Luszcz M., “Lengths of Left-turn Lanes at Unsignalized
Intersections”, Transportation Research Record, Vol. 1500, 1995; pp. 193-201.
Dorothy P.W., Maleck T.L., and Nolf S.E., “Operational Aspects of Michigan Design for
Divided Highways”, Transportation Research Record, Vol. 1579, 1997; pp. 18-26.
Harmelink M., “Volume Warrants for Left-Turn Storage Lanes at Unsignalized Grade
Intersections,” Highway Research Record, Vol. 211, 1967; pp. 1-18.
Kikuchi S., Chakroborty P., “Analysis of Left-Turn Lane Warrants at Unsignalized T-

intersections on Two-lane Roadways”, Transportation Research Record, Vol. 1327,
1991, pp. 80-88.
Kikuchi S., Chakroborty P., Vukadinovic K., “Lengths of Left-Turn Lanes at Signalized
Intersections”, Transportation Research Record, Vol. 1385, 1993, pp. 162-171.
Lertworawanich P. and Elefteriadou L., “Determination of Storage Lengths of Left-Turn

Lanes at Unsignalized Intersections Using M/G2/1 Queuing”. Transportation Research
Record, Vol. 1847, 2003, pp. 20-28.
Long G., “Intervehicle Spacings and Queue Characteristics”, Transportation Research

Record, Vol. 1796, Paper No 02-3953, 2002, pp. 86-96.
McCoy P.T., Bonneson J.A., Ataullan S., and Eitel D.S, “Guidelines for Right-Turn
Lanes on Urban Roadways”, Transportation Research Record 1445, 1994, pp. 130-137
Misaghi P. and Hassan A., “Modeling Operating speed and Speed Differential on Two-
Lane Rural Roads”, Journal of Transportation Engineering, June 2005, pp. 408-417.
Oppenlander J.C., Oppenlander J.E., “Design Lengths of Left- or Right-Turn Lanes With
Separate Signal Phases”, ITE Journal, Vol. 59(7), 1989, pp. 23-26.
American Association of State Highway & Transportation Officials. A Policy on the

Geometric Design of Highways and Streets. Washington, DC, 2001.
American Association of State Highway & Transportation Officials. A Policy on the

Geometric Design of Highways and Streets. Washington, DC, 2004.
Minnesota Department of Transportation. Roadway Design Manual, May 2001,

http://www.dot.state.mn.us/tecsup/rdm/index.html
Transportation Research Board. Highway Capacity Manual. Washington, DC, 2000.
49
Institute of Transportation Engineers. Traffic Engineering Handbook, 5th edition,
Washington, DC, 1999.
50
APPENDIX A: QUESTIONNAIRE
A. QUESTIONNAIRE
Signalized Left Turns
1. How does your agency warrant an exclusive left-turn lane at signalized

intersections? (Check all that apply)
Driver complaints
Volume of left-turning vehicles
Other (please specify)
2. Once warranted, how do you decide on the length of a left-turn lane at a

signalized intersection? (Check all that apply)
AASHTO Guidelines
State Guidelines (where can it be found)?
Guidelines Outlined by Harmelink (1967)
Other (please specify):
3. Which of the following parameters does your agency consider in determining the
length of a signalized left-turn lane? (Check all that apply)
Volume of left-turn vehicles

Volume of opposing vehicles
ADT
Grade
Signal Timing Plan
Vehicle mixture
Gap size between vehicles
4. How do you warrant multiple exclusive (left-only) turn lanes at an intersection?

(Check all that apply)
AASHTO Guidelines
5. In your opinion, do the guidelines that your agency use provide efficient turn
lane lengths at all signalized intersections? (Please Explain)
A-1
Unsignalized Left Turns
6. How does your agency warrant an exclusive left-turn lane at an unsignalized

intersection? (Check all that apply)
Driver complaints
Volume of left-turning vehicles
7. Once warranted, how does your agency decide on the length of a left-turn lane at
an unsignalized intersection? (Check all that apply)
AASHTO Guidelines
length of an unsignalized left-turn lane? (Check all that apply)

Volume of opposing vehicles
ADT
Grade
Signal Timing Plan
Vehicle mixture
9. In your opinion, do the guidelines that your agency use provide efficient turn lane
lengths at all unsignalized intersections? (Please Explain)
General Comments:
A-2
Signalized Right Turns
1. How does your agency warrant an exclusive right-turn lane at signalized

intersections? (Check all that apply)
Driver complaints
Volume of right-turning vehicles
2. Once warranted, how do you decide on the length of a right-turn lane at a

signalized intersection? (Check all that apply)
AASHTO Guidelines
length of a signalized right-turn lane? (Check all that apply)

Volume of crossing vehicles
ADT
Grade
Signal Timing Plan
Vehicle mixture
4. How do you warrant multiple exclusive (right-only) turn lanes at an intersection?

(Check all that apply)
AASHTO Guidelines
State Guidelines (where can it be found?)
Comments:
5. In your opinion, do the guidelines that your agency use provide efficient turn
lane lengths at all signalized intersections? (Please Explain)
A-3
Unsignalized Right Turns
6. How does your agency warrant an exclusive right-turn lane at an unsignalized

intersection? (Check all that apply)
Driver complaints
Volume of right-turning vehicles
7. Once warranted, how does your agency decide on the length of a right-turn lane
at an unsignalized intersection? (Check all that apply)
AASHTO Guidelines
length of an unsignalized right-turn lane? (Check all that apply)
Volume of right-turn vehicles

Volume of crossing vehicles
ADT
Grade
Signal Timing Plan
Vehicle mixture
9. In your opinion, do the guidelines that your agency uses provide efficient turn
lane lengths at all unsignalized intersections? (Please Explain)
General Comments:
Thank you for your time!
A-4
APPENDIX B: MAPS
B. MAPS
B.1. Maps of North of St. Paul Area
Figure B.1. 1 North of St. Paul Area
B-1
Figure B.1. 2. SR-169 & CR-4 Figure B.1. 3. SR 65 & SR 5
Figure B.1. 4. TH 10 at Big Lake Industrial Park Figure B.1. 5. TH 169 & CR 25
B-2
B.2. Maps of Brainard/Baxter Area
Figure B.2. 1. Brainard/Baxter Area
Figure B.2. 2. SR-371 & CR-125 Figure B.2. 3. SR-371 & CR77-CR49
B-3
B.3. Maps of St. Paul Metro Area
Figure B.3.1. St. Paul Metro Area
Figure B.3. 2. SR-7 & Louisiana Avenue Figure B.3.3. SR-7 & CR-73 (17th Avenue)
B-4
FigureB.3.4. TH 3 (Robert ST)& Marie Ave.
Figure B.3. 5. TH 36 & SR120 (Century Ave)
B-5
APPENDIX C: DIAGRAMS
C. DIAGRAMS
190
N
110
225
315
275 250
90
110 150
50 200 275 225
455
975
50
150
Figure C.1. SR-169 & CR-4
C-1
N
150
150
285
285
645 120
200 60
420
975
190
200
Figure C.2. SR-65 & SR-5 (Isanti)
C-2
N
350
150 450
100
Figure C.3. US 10 – Industrial Dr.
C-3
100
200
154
Figure C.4. TH 169 – CR 25
C-4
N
150
100
300
250
120
300
110
150
Figure C.5. SR 371 - CR 125
C-5
N
140
140
175
260 265 125

85
150 140
115 160
50 395
305
305
150
150
Figure C.6. SR 371 – CR 77/49
C-6
65
180
120
105
Figure C.7. SR 7 – CR 73
C-7
310
N
120
275
400 160
550 164
150 350
170 270
130
130
200
Figure C.8. SR 7 – Louisiana Ave
C-8
N
200 45
120 80
Figure C.9. Robert St – Marie Ave.
C-9
0
12
80
Figure C.10. SR 36 – SR 120 (Century)
C-10
APPENDIX D: INTERSECTION DATA SUMMARY
D. INTERSECTION DATA SUMMARY
Direction SB NB WB EB
Description SL ST SR NL NT NR WL WT WR EL ET ER
Total Volume 64 527 130 185 773 174 91 170 36 133 192 108
Heavy Vehicles 20% 5% 2% 4% 4% 8% 18% 2% 20% 2% 4% 5%
Peak Hour Factor
Growth Factor
Lane Width 12 12 12 12 12 12 12 12 12 12 12 12
Storage Length 315 225 975 455 275 275 150 200
Storage Length adj. 378 262 1000 475 350 300 187 217
Storage Lanes 1 1 1 1 1 1 1 1
Turning Speed 20 18 20 18 20 18 20 18
Link Speed 45 45 30 30
Cycle Length 300
Right Turn on Red yes yes yes yes
Leading Detector 50 475 50 5 475 50 15 125 15 125 50
Trailing Detector 2 2 2 2
Headway Factor
Minimum Green 5 15 5 15 5 10 5 10
Vehicle extension 3 5.5 3 5.5 3 3.5 3 3.5
Yellow Time 3 6 3 6 3 3 3 3
All-Red Time 1 1.5 1 1.5 0.5 2 0.5 2
Max Green 1 60 120 60 120 40 38 35 60
Max Green 2 60 150 60 150 40 38 35 60
Maximum Split 1 64 127.5 64 127.5 43.5 43 38.5 65
Maximum Split 2 64 157.5 64 157.5 43.5 43 38.5 65
Minimum Split 9 22.5 9 22.5 8.5 15 8.5 15
Minimum Gap 4.5 4.5
Time bef. Reduce 30 30
Time to Reduce 30 30
Ped Walk 7 7 7 7
Ped Don't Walk 16 16 32 32
Turn Type Prot Free Prot Free Prot Prot Free
Total Split 1 66 129.5 66 129.5 45.5 45 40.5 67
Table D.1. SR-169 & CR-4
D-1
Direction EB WB NB SB
Description EL ET ER WL WT WR NL NT NR SL ST SR
Total Volume 234 108 155 62 95 45 304 604 83 221 498 36
Heavy Vehicles 2% 9% 5% 10% 9% 15% 6% 2% 2% 7% 5% 3%
Peak Hour Factor
Growth Factor
Lane Width 12 14 12 12 12 16 12 12 12 12 11 12
Storage Length 250 50 200 645 975 420 285 285
Turning Speed
Right Turn on Red yes yes yes yes
Leading Detector 120 120 120 120 0 625 625 0 625
Headway Factor
Cycle Length 300
Minimum Green 5 10 5 10 5 15 5 15
Vehicle extension 3 3 3 3 3 6 3 6
Yellow Time 3 3 3 3 3 6 3 6
All-Red Time 0.5 2 0.5 2 1 1.5 1 1.5
Max Green 1 20 60 20 60 80 90 60 90
Max Green 2 40 40 40 40 40 40 60 40
Max Green 3 20 60 20 60 80 150 60 150
Max Split 1 23.5 65 0 23.5 65 0 84 97.5 0 64 97.5 0
Max Split 2 43.5 45 0 43.5 45 0 44 47.5 0 64 47.5 0
Max Split 3 23.5 65 0 23.5 65 0 84 158 0 64 158 0
Min Split 8.5 15 0 8.5 15 0 9 22.5 0 9 22.5 0
Minimum Initial
Minimum Gap 4 4
Ped Walk
Ped Don't Walk
Turn Type Prot+ Perm Prot+ Perm Prot Perm Prot Perm
Perm Perm
Total Split
Table D.2. SR-65 & SR-5 (Isanti)
D-2
Direction SB NB WT EB
Total Volume 19 2 7 30 3 49 15 831 23 9 897 7
Heavy Vehicles 5% 0% 0% 0% 0% 4% 73% 11% 13% 0% 6% 14%
Peak Hour Factor
Growth Factor
Lane Width 14 12 16 14 12 16 12
Storage Length 200 150 100 50 350 350 450 450
Turning Speed 20 18 20 18 20 18 20 18
Right Turn on Red
Leading Detector
Trailing Detector
Headway Factor
Phasing Unsignalized Unsignalized Unsignalized Unsignalized
Table D.3. TH 10 & Industrial Dr.
Direction SB NB WT EB
Total Volume 0 714 7 88 1525 0 10 0 44
Heavy Vehicles 0% 4% 10% 4% 2% 0% 2% 0% 6%
Peak Hour Factor 0.92 0.78 0.83 0.93 0.78 0.78
Growth Factor
Lane Width 12 11 16 12 11 11
Storage Length 200 154
Storage Lanes 1 1 1 1
Turning Speed 9 15 15 9
Right Turn on Red
Leading Detector
Trailing Detector
Headway Factor
Link Speed 45 45 30
Phasing Unsignalized Unsignalized Unsignalized Unsignalized
Table D.4. TH 169 & CR25
D-3
Direction SB NB WB EB
Total Volume 0 602 38 73 515 73 0 2 2 35 0 50
Heavy Vehicles 0% 4% 0% 0% 4% 0% 0% 0% 0% 0% 0% 0%
Peak Hour Factor
Growth Factor
Lane Width 14 12 14 14 12 14 14 12 14 14 12 14
Storage Length 300 250 300 120 300 300 300 300
Turning Speed 15 9 15 9 15 9 15 9
Right Turn on Red
Leading Detector
Trailing Detector
Headway Factor
Phasing Unsignalized Unsignalized Stop sign Stop sign
Table D.5. SR-371 & CR-125
D-4
Direction SEB NWB ENB WSB
Description SEL SET SER NWL NWT NWR ENL ENT ENR WSL WST WSR
Total Volume 67 639 17 45 382 19 37 36 75 37 30 68
Heavy Vehicles 4% 4% 12% 11% 7% 5% 16% 6% 1% 8% 7% 4%
Peak Hour Factor
Growth Factor
Lane Width 13 12 21 13 12 26 12 12 21 12 12 22
Storage Length 260 175 305 305 160 395 150 265
Turning Speed 15 9 15 9 15 9 15 9
Right Turn on Red
Leading Detector 10 300 0 300 45 525 45 525
Trailing Detector 10 10 10 0 0 0
Headway Factor
Cycle Length 110
Minimum Green 5 14 5 14 10 10
Vehicle extension 4 7 4 7 4.5 4.5
Yellow Time 3 5.5 3 5.5 4 4
All-Red Time 2 1.5 2 1.5 2 2
Minimum Split 10 21 10 21 16 16
Minimum Gap 4 4
Time before Reduce 40 40
Ped Walk 7 7 7 7
Turn Type Prot Prot Perm Perm
Max Green 1 45 120 45 120 45 45
Max Green 2 150 150
Max Green 3 180 180
Total Split 1 50 127 50 127 51 51
Total Split 2 157 157
Total Split 3 187 187
Table D.6. SR-371 & CR77-CR49
D-5
Total Volume 141 581 58 79 731 171 85 154 65 174 184 124
Heavy Vehicles % 4% 4% 7% 4% 2% 4% 2% 1% 8% 5% 1% 5%
Peak Hour Factor
Growth Factor
Lane Width 12 12 17.5 12 12 20 12.5 12 15 12 12 13
Storage Length 225 302 680 280 120 55 180 Infin.
Storage length adj. 260 359 753 400 158 83 202
Storage Lanes
Turning Speed 15 9 15 9 15 9 15 9
Right Turn on Red Yes Yes Yes Yes
Grade 0 0 0 0 0 0 0 0 0 0 0 0
Headway Factor
Cycle Length 243
Minimum Green 7 20 7 20 7 10 7 10
Vehicle extension 4 4.5 3 3 3 3 3 3
Yellow Time 3 4.5 3 4.5 3 4 3 4
All-Red Time 2 2 2 2 2 2.5 2 2.5
Minimum Split 12 26.5 12 26.5 12 16.5 12 16.5
Minimum Gap 4.5 4.5
Ped Walk 7 7 7 7
Turn Type Prot Prot Prot Prot
Protected Phases Lead Lag Lead Lead
Permitted Phases
Max green 35 120 35 90 25 40 25 40
Total Split 40 126.5 40 96.5 30 46.5 30 46.5
Table D.7. SR-7 & CR-73
D-6
Total Volume 172 1306 243 204 866 94 133 215 157 196 179 100
Heavy Vehicles 2% 2% 4% 9% 3% 4% 6% 4% 6% 3% 4% 7%
Peak Hour Factor
Growth Factor
Lane Width 14 12 21 14 12 22 12 12 21 12 12 22
Storage Length 350 270 550 400 130 200 275 310
Turning Speed 15 9 15 9 15 9 15 9
Headway Factor
Cycle Length 182
Minimum Green 7 20 7 20 7 10 7 10
Vehicle extension 3 6 3 6 3 4 3 3
Yellow Time 3 4.5 3 4.5 3 3.5 3 3.5
All-Red Time 2 2 2 2 2 2.5 2 2.5
Minimum Split 12 26.5 0 12 26.5 0 12 16 0 12 16
Minimum Gap 4.5 4.5
Ped Walk 7 7 7 7
Turn Type Prot Free Prot Free Prot Free Prot Free
Max green1 30 80 30 80 30 30 30 30
Max green 2 25 60 25 60 25 25 25 25
Max Green 3 30 80 20 80 20 30 30 20
Total Split 1 35 86.5 35 86.5 35 36 35 36
Total Split 2 30 66.5 0 30 66.5 0 30 31 0 30 31
Total Split 3 35 86.5 0 25 86.5 0 25 36 0 35 26
Table D.8. SR-7 & Louisiana Ave
D-7
Cars 157 122 80 80 91 50 53 767 36 64 793 79
Total Volume 158 125 81 81 93 52 54 774 37 67 799 80
Heavy Vehicles 1% 2% 1% 1% 2% 4% 2% 1% 3% 4% 1% 1%
Peak Hour Factor
Growth Factor
Lane Width 13 12 12 13 12 12 13 12 15 13 12 12
Storage Length 80 80 200 125 80 80 200 200
Turning Speed 15 9 15 9 15 9 15 9
Headway Factor
Cycle Length 105
Minimum Green 5 10 5 10 5 15 5 15
Vehicle extension 3 3 3 3 2 4.5 2 3.5
Yellow Time 3 3.5 3 3.5 3 4 3 4
All-Red Time 2 2 2 2 2 1.5 2 1.5
Minimum Split 10 15.5 10 15.5 10 20.5 10 20.5
Minimum Gap 2 2 2 2.5 2 2.5
Time bef. Reduce 10 10 20 0 15
Time to Reduce 10 10 15 15
Ped Walk 9 9 7 7
Ped Don't Walk
Turn Type Perm Perm Perm Perm Prot+ Perm Prot+ Perm
perm perm
Max green1 15 30 15 30 15 45 15 45
Max green 2 15 15 10 30 10 30
Total Split1 20 35.5 20 35.5 20 50.5 20 50.5
Total Split 2 5 20.5 5 20.5 15 35.5 15 35.5
Table D.9. TH 3 (Robert ST) & Marie Ave
D-8
Direction NB SB NEB SWB
Description NL NT NR SL ST SR NEL NET NER SWL SWT SWR
Total Volume 178 231 87 59 246 118 130 706 155 87 721 52
Heavy Vehicles 5% 3% 2% 2% 3% 1% 3% 5% 6% 1% 4% 0%
Peak Hour Factor
Growth Factor
Lane Width 12 12 13 12 12 14 12 12 30 10 12 19.2
Storage Length 80 350 250 245 320 350 375 285
Turning Speed 20 9 20 9 20 9 20
Leading Detector 180 180 180 180 180 180 400 400 400 400
Trailing Detector 10 10 0 10 10 0 5 5
Headway Factor
Cycle Length 101
Minimum Green 5 10 5 10 7 20 7 20
Vehicle extension 2 3.5 2 3.5 2.5 6 2.5 6
Yellow Time 3 5 3 5 3 5 3 5
All-Red Time 2 3 2 3 2 2 2 2
Minimum Split 10 18 10 18 12 27 12 27
Minimum Gap 4 4
Time before Reduce 20 20
Ped Walk 16 16 6 6
Prot+ Prot+
Prot Prot
Turn Type Perm Perm
Max green1 26 33 26 33 17 75 17 75
Max green2 12 30 12 30 25 65 25 65
Max green3 14 28 14 28 14 45 14 45
Total Split 1 31 41 31 41 22 82 22 82
Total Split 2 17 38 17 38 30 72 30 72
Total Split 3 19 36 19 36 19 52 19 52
Table D.10. SR36 & SR120 (Century Ave)
D-9
APPENDIX E: SUMMARY DATA FOR THE
SIMULATION QUEUE LENGTHS
E. SUMMARY DATA FOR THE SIMULATION QUEUE
LENGTHS
Intersection Directions EB WB NB SB
North of St. Paul EL ER WL WR NL-out NL-in NR SL-out SL-in SR
SR-169 & Av. QL in cars 4.8 0.6 2.8 0.5 7.4 0.6 2.4 1.2
CR-4 Max QL in cars 12.0 4.0 6.0 3.0 13.0 6.0 8.0 6.0
95% QL in cars 8.1 2.1 5.4 1.8 13.0 3.7 5.4 4.2
Av. QL in ft 120.0 15.0 70.0 12.5 185.0 15.0 60.0 30.0
95% QL in ft 202.5 52.1 136.0 45.5 325.3 93.4 134.3 104.3
SR-65 & Av. QL in cars 5.8 0.6 1.2 0.1 8.5 0.2 0.7 0.4
SR-5 (Isanti) Max QL in cars 13.0 3.0 4.0 1.0 16.0 1.0 2.0 3.0
95% QL in cars 11.5 2.1 3.0 0.6 14.3 0.9 19.0 1.7
Av. QL in ft 145.0 15.0 30.0 2.5 212.5 5.0 17.5 10.0
95% QL in ft 287.5 52.5 75.0 15.0 357.5 22.5 475.0 42.5
TH 10 & Av. QL in cars 0.2 0.2 0.3 0.5 0.9 0.7 0.5 0.3
Industrial Dr. Max QL in cars 1.0 1.0 2.0 2.0 5.0 1.0 3.0 1.0
95% QL in cars 0.9 0.9 1.1 1.5 2.9 1.5 1.8 1.1
Av. QL in ft 5.0 5.0 7.5 12.5 22.5 17.5 12.5 7.5
95% QL in ft 22.5 22.5 27.5 37.5 72.5 37.5 45.0 27.5
TH 169 & Av. QL in cars 1.0 1.3 0.9
CR25 Max QL in cars 2.0 4.0 4.0
95% QL in cars 2.2 4.6 2.7
Av. QL in ft 25.0 32.5 22.5
95% QL in ft 55.0 115.0 67.5
Near Brainard/Baxter
SR-371 & Av. QL in cars 1.4 1.1 0.0 1.4 1.2 0.8
CR-125 Max QL in cars 6.0 2.0 1.0 4.0 5.0 2.0
(Gull Dam 95% QL in cars 2.4 1.8 0.1 2.0 3.2 1.1
Road) Av. QL in ft 35.0 27.5 0.0 35.0 30.0 20.0
95% QL in ft 60.0 44.0 2.5 50.0 79.5 27.5
SR-371 & Av. QL in cars 1.2 1.1 1.0 1.4 2.3 0.3 1.6 0.6
CR77-CR49 Max QL in cars 4.0 3.0 3.0 4.0 7.0 1.0 6.0 2.0
95% QL in cars 2.5 2.3 2.3 3.1 5.1 1.1 4.4 1.6
Av. QL in ft 30.0 27.5 25.0 35.0 57.5 7.5 40.0 15.0
95% QL in ft 62.5 57.5 57.5 77.5 127.5 27.5 110.0 40.0
Table E.1 Total data for the Queue Lengths
E-1
Intersection Directions EB WB NB SB
St. Paul Metro Area EL ER WL WR NL NL-in NR SL-out SL-in SR
SR-7 & Av. QL in cars 7.3 1.7 3.4 6.0 4.8 2.2 7.8 4.0
CR-73 Max QL in cars 15.0 5.0 8.0 10.0 8.0 5.0 13.0 11.0
95% QL in cars 12.8 4.2 6.4 9.3 8.3 4.7 12.6 7.6
Av. QL in ft 182.5 42.5 85.0 150.0 120.0 55.0 195.0 100.0
95% QL in ft 320.0 105.0 160.0 232.5 207.5 117.5 315.0 190.0
SR-7 & Av. QL in cars 8.0 6.5 10.0 2.6 3.4 4.0 4.3 4.8 6.2 2.7
Louisiana Ave Max QL in cars 16.0 19.0 16.0 8.0 7.0 8.0 8.0 11.0 14.0 6.0
95% QL in cars 14.3 12.9 15.4 5.6 5.9 10.3 7.4 8.6 11.0 5.2
Av. QL in ft 200.0 162.5 250.0 65.0 85.0 100.0 107.5 120.0 155.0 67.5
95% QL in ft 357.5 322.5 385.0 140.0 147.5 257.5 185.0 215.0 275.0 130.0
TH 3 Av. QL in cars 3.0 1.2 1.7 1.2 1.3 1.9
(Robert ST) & Max QL in cars 7.0 3.0 5.0 3.0 4.0 4.0
Marie Ave 95% QL in cars 5.1 2.2 3.8 2.5 2.8 3.6
Av. QL in ft 75.0 30.0 42.5 30.0 32.5 47.5
95% QL in ft 127.5 55.0 95.0 62.5 70.0 90.0
SR36 & Av. QL in cars 3.4 2.1 2.8 1.0 4.9 1.4 1.6 2.1
SR120 Max QL in cars 8.0 4.0 6.0 3.0 10.0 4.0 5.0 7.0
(Century Ave) 95% QL in cars 7.2 3.8 5.6 2.2 7.9 2.9 3.6 4.7
Av. QL in ft 85.0 52.5 70.0 25.0 122.5 35.0 40.0 52.5
95% QL in ft 180.0 95.0 140.0 55.0 197.5 72.5 90.0 117.5
Table E.1. Continued
E-2
APPENDIX F: REGRESSION EQUATIONS
F. REGRESSION EQUATIONS
Right Turns
The regression equations for 95th percentile queue storage length for right-turn lanes are
the following:
Permitted right turns on Signalized intersections
RTperm = 65.5 + 0.0323*TV + 0.0533*CV + 0.186*RTV - 0.829*Sp - 1.50*HVT +

0.818*HVR + 0.107*Gr
Predictor Coef StDev T P

Constant 65.451 6.368 10.28 0.000
TV 0.032337 0.005147 6.28 0.000
CV 0.05332 0.01069 4.99 0.000
RTV 0.18628 0.01193 15.62 0.000
Sp -0.82860 0.08612 -9.62 0.000
HVT -1.5019 0.2482 -6.05 0.000
HVR 0.8181 0.2713 3.02 0.003
Gr 0.1069 0.8395 0.13 0.899
S = 12.98 R2 = 75.9% R2 (adjusted) = 75.0%
Free right turns on Signalized intersections
RT free_S = - 61.5 + 0.0551*TV + 0.265*CV + 0.317*RTV - 0.527*Sp + 0.041*HVT +

1.04*HVR + 0.47*Gr

Constant -61.514 6.798 -9.05 0.000
TV 0.055069 0.006588 8.36 0.000
CV 0.26516 0.01312 20.22 0.000
RTV 0.31653 0.01277 24.78 0.000
Sp -0.52696 0.08299 -6.35 0.000
HVT 0.0408 0.5520 0.07 0.941
HVR 1.0352 0.2786 3.72 0.000
Gr 0.468 1.382 0.34 0.735
S = 18.50 R2 = 81.5% R2 (adjusted) = 81.1%
F-1
Yield right turns on Unsignalized intersections
RTyield_U = 16.0 - 0.00497*TV + 0.0235*CV + 0.0957*RTV - 0.115*Sp - 0.038*HVT +

0.274*HVR + 0.223*Gr

Constant 15.961 3.115 5.12 0.000
TV -0.004968 0.003221 -1.54 0.125
CV 0.023514 0.003776 6.23 0.000
RTV 0.095743 0.004941 19.38 0.000
Sp -0.11494 0.04010 -2.87 0.005
HVT -0.0376 0.1097 -0.34 0.732
HVR 0.27385 0.04966 5.51 0.000
Gr 0.2228 0.3902 0.57 0.569
S = 4.689 R2 = 81.0% R2 (adjusted) = 80.2%
Free right turns on Unsignalized intersections
RTfree_U = 13.2 - 0.00021*TV + 0.0176*CV + 0.0876*RTV - 0.0615*Sp - 0.094*HVT +

0.169*HVR + 0.164*Gr

Constant 13.165 2.556 5.15 0.000
TV -0.000206 0.002660 -0.08 0.938
CV 0.017576 0.002999 5.86 0.000
RTV 0.087636 0.004133 21.20 0.000
Sp -0.06151 0.03322 -1.85 0.066
HVT -0.0945 0.1007 -0.94 0.349
HVR 0.16896 0.04204 4.02 0.000
Gr 0.1640 0.3266 0.50 0.616
S = 3.927 R2= 81.0% R2 (adjusted) = 80.2%
F-2
The combination of Free and Permitted right turns on Signalized ntersections
RTfree_perm_S = 5.32 + 0.0341*TV + 0.144*CV + 0.249*RTV - 0.563*Sp + 0.025*HVT +

0.747*HVR + 0.42*Gr

Constant 5.321 6.659 0.80 0.425
TV 0.034063 0.006032 5.65 0.000
CV 0.14415 0.01222 11.80 0.000
RTV 0.24904 0.01282 19.43 0.000
Sp -0.56258 0.08601 -6.54 0.000
HVT 0.0250 0.3563 0.07 0.944
HVR 0.7466 0.2800 2.67 0.008
Gr 0.416 1.148 0.36 0.717
S = 23.49 R2= 59.5% R2 (adjusted) = 58.9%
The combination of Free right turns on Signalized and Unsignalized intersections
RTfree_S_U= - 55.0 + 0.0751*TV + 0.178*CV + 0.255 *RTV - 0.330*Sp - 1.18*HVT+

0.371*HVR - 0.24*Gr

Constant -55.047 7.654 -7.19 0.000
TV 0.075051 0.007487 10.02 0.000
CV 0.17786 0.01203 14.78 0.000
RTV 0.25524 0.01400 18.23 0.000
Sp -0.33044 0.09722 -3.40 0.001
HVT -1.1821 0.4659 -2.54 0.011
HVR 0.3710 0.2149 1.73 0.085
Gr -0.238 1.373 -0.17 0.862
S = 24.73 R2 = 63.5% R2 (adjusted) = 63.0%
F-3
The combination of Yield and Free right turns on Unsignalized intersections
RTyield_free_U= 14.6 - 0.00251*TV + 0.0206*CV + 0.0918*RTV - 0.0878*Sp -

0.0869*HVT+ 0.216*HVR + 0.178*Gr

Constant 14.630 2.024 7.23 0.000
TV -0.002506 0.002094 -1.20 0.232
CV 0.020569 0.002397 8.58 0.000
RTV 0.091763 0.003253 28.21 0.000
Sp -0.08780 0.02627 -3.34 0.001
HVT -0.08688 0.07478 -1.16 0.246
HVR 0.21572 0.03278 6.58 0.000
Gr 0.1784 0.2570 0.69 0.488
S = 4.370 R2= 80.2% R2 (adjusted) = 79.8%
Left Turns
The Regression Equations for 95th percentile queue storage length for left-turn lane are
following:
Protected left turns on Signalized intersections
LTprot_S= 35.3 + 0.0203*TV + 1.14*LTV - 0.171*Sp - 6.75*HVT + 1.32*HVL - 0.16*Gr

Constant 35.302 7.500 4.71 0.000
TV 0.020264 0.006537 3.10 0.002
LTV 1.14443 0.02166 52.84 0.000
Sp -0.1712 0.1305 -1.31 0.190
HVT -6.7508 0.6695 -10.08 0.000
HVL 1.3159 0.2756 4.77 0.000
Gr -0.157 1.069 -0.15 0.884
S = 37.05 R2 = 83.5% R2 (adjusted) = 83.3%
F-4
Permitted left turns on Signalized intersections
LTperm_S = - 45.2 - 0.00953*TV + 0.0406*OV + 0.610*LTV + 0.348*Sp + 0.812*HVT+

1.76*HVL + 0.35*Gr

Constant -45.212 7.428 -6.09 0.000
TV -0.009530 0.004543 -2.10 0.037
OV 0.040649 0.003882 10.47 0.000
LTV 0.61003 0.01420 42.96 0.000
Sp 0.3476 0.1257 2.76 0.006
HVT 0.8117 0.4789 1.69 0.091
HVL 1.7617 0.2466 7.14 0.000
Gr 0.347 1.549 0.22 0.823
S = 22.16 R2 = 86.6% R2 (adjusted) = 86.4%
Yield left turn on Unsignalized intersections
LTyield_U= 0.00 + 0.00315*TV + 0.0332*OV + 0.345*LTV - 0.149*Sp + 0.224*HVT+

0.629*HVL - 0.080*Gr

Constant 0.004 4.001 0.00 0.999
TV 0.003152 0.002562 1.23 0.220
OV 0.033179 0.002576 12.88 0.000
LTV 0.34507 0.01119 30.85 0.000
Sp -0.14906 0.06747 -2.21 0.028
HVT 0.2242 0.3029 0.74 0.460
HVL 0.6290 0.2146 2.93 0.004
Gr -0.0800 0.9335 -0.09 0.932
S = 10.23 R2 = 82.8% R2 (adjusted) = 82.3%
F-5
The combination of Protected and Permitted left turns on Signalized intersections
LTprot_perm_S = 7.4 + 0.0785*TV - 0.0632*OV + 1.20*LTV - 0.322*Sp - 6.68*HVT +

0.655*HVL - 0.08*Gr

Constant 7.42 10.52 0.71 0.480
TV 0.078469 0.008267 9.49 0.000
OV -0.063188 0.007805 -8.10 0.000
LTV 1.20237 0.02642 45.52 0.000
Sp -0.3221 0.1680 -1.92 0.056
HVT -6.6784 0.8630 -7.74 0.000
HVL 0.6546 0.3424 1.91 0.056
Gr -0.077 1.613 -0.05 0.962
S = 65.42 R2 = 69.3% R2 (adjusted) = 69.1%
The combination of Yield left turns on Unsignalized and Permitted left turns on
Signalized intersections
RTyield_U&perm_S = - 38.5 + 0.00572*TV + 0.0326*OV + 0.649*LTV + 0.297*Sp -

0.512*HVT+ 1.72*HVL + 0.17*Gr

Constant -38.471 8.920 -4.31 0.000
TV 0.005724 0.005633 1.02 0.310
OV 0.032618 0.004727 6.90 0.000
LTV 0.64883 0.01710 37.95 0.000
Sp 0.2970 0.1411 2.11 0.036
HVT -0.5115 0.4878 -1.05 0.295
HVL 1.7218 0.2788 6.18 0.000
Gr 0.171 1.747 0.10 0.922
S = 28.42 R2 = 78.8% R2 (adjusted) = 78.5%
F-6
The combination of Permitted and Protected left turns on Signalized intersections
and Yield left turns on Unsignalized intersections
LTall = 10.1 + 0.0587*TV - 0.0907*OV + 1.10*LTV - 0.039*Sp - 5.34*HVT+

0.919*HVL - 0.00*Gr

Constant 10.074 9.528 1.06 0.291
TV 0.058676 0.007481 7.84 0.000
OV -0.090714 0.005213 -17.40 0.000
LTV 1.10321 0.02540 43.43 0.000
Sp -0.0386 0.1533 -0.25 0.801
HVT -5.3435 0.7981 -6.70 0.000
HVL 0.9194 0.3199 2.87 0.004
Gr -0.002 1.600 -0.001 0.999
S = 67.23 R2 = 65.9% R2 (adjusted) = 65.7%
F-7
APPENDIX G: EXAMPLE
G. EXAMPLE
Permitted Left Turn of Signalized Intersections
Speed TV /OV LTV HVT HVL Storage Decel. Taper Total Length
(mph) (veh/ph) (veh/ph) (%) (%) (ft) (ft) (ft) (ft)
45 300/300 50 5 0 14 340 130 484

50 5 5 23 340 130 493
50 5 10 32 340 130 502
50 5 15 41 340 130 511
50 5 20 50 340 130 520
100 5 0 45 340 130 515
100 5 5 54 340 130 524
100 5 10 62 340 130 532
100 5 15 71 340 130 541
100 5 20 80 340 130 550
150 5 0 75 340 130 545
150 5 5 84 340 130 554
150 5 10 93 340 130 563
150 5 15 102 340 130 572
150 5 20 111 340 130 581
200 5 0 106 340 130 576
200 5 5 115 340 130 585
200 5 10 123 340 130 593
200 5 15 132 340 130 602
200 5 20 141 340 130 611
250 5 0 136 340 130 606
250 5 5 145 340 130 615
250 5 10 154 340 130 624
250 5 15 163 340 130 633
250 5 20 172 340 130 642
G-1
(L_Perm_S)
45 600/600 50 5 0 24 340 130 494

50 5 5 32 340 130 502
50 5 10 41 340 130 511
50 5 15 50 340 130 520
50 5 20 59 340 130 529
100 5 0 54 340 130 524
100 5 5 63 340 130 533
100 5 10 72 340 130 542
100 5 15 81 340 130 551
100 5 20 89 340 130 559
150 5 0 85 340 130 555
150 5 5 93 340 130 563
150 5 10 102 340 130 572
150 5 15 111 340 130 581
150 5 20 120 340 130 590
200 5 0 115 340 130 585
200 5 5 124 340 130 594
200 5 10 133 340 130 603
200 5 15 142 340 130 612
200 5 20 150 340 130 620
250 5 0 146 340 130 616
250 5 5 154 340 130 624
250 5 10 163 340 130 633
250 5 15 172 340 130 642
250 5 20 181 340 130 651
G-2
(L_Perm_S)
45 900/900 50 5 0 33 340 130 503

50 5 5 42 340 130 512
50 5 10 51 340 130 521
50 5 15 59 340 130 529
50 5 20 68 340 130 538
100 5 0 63 340 130 533
100 5 5 72 340 130 542
100 5 10 81 340 130 551
100 5 15 90 340 130 560
100 5 20 99 340 130 569
150 5 0 94 340 130 564
150 5 5 103 340 130 573
150 5 10 112 340 130 582
150 5 15 120 340 130 590
150 5 20 129 340 130 599
200 5 0 124 340 130 594
200 5 5 133 340 130 603
200 5 10 142 340 130 612
200 5 15 151 340 130 621
200 5 20 160 340 130 630
250 5 0 155 340 130 625
250 5 5 164 340 130 634
250 5 10 173 340 130 643
250 5 15 181 340 130 651
250 5 20 190 340 130 660
G-3
(L_Perm_S)
50 300/300 50 5 0 16 410 130 556

50 5 5 25 410 130 565
50 5 10 34 410 130 574
50 5 15 42 410 130 582
50 5 20 51 410 130 591
100 5 0 47 410 130 587
100 5 5 55 410 130 595
100 5 10 64 410 130 604
100 5 15 73 410 130 613
100 5 20 82 410 130 622
150 5 0 77 410 130 617
150 5 5 86 410 130 626
150 5 10 95 410 130 635
150 5 15 103 410 130 643
150 5 20 112 410 130 652
200 5 0 108 410 130 648
200 5 5 116 410 130 656
200 5 10 125 410 130 665
200 5 15 134 410 130 674
200 5 20 143 410 130 683
250 5 0 138 410 130 678
250 5 5 147 410 130 687
250 5 10 156 410 130 696
250 5 15 164 410 130 704
250 5 20 173 410 130 713
G-4
(L_Perm_S)
50 600/600 50 5 0 25 410 130 565

50 5 5 34 410 130 574
50 5 10 43 410 130 583
50 5 15 52 410 130 592
50 5 20 61 410 130 601
100 5 0 56 410 130 596
100 5 5 65 410 130 605
100 5 10 74 410 130 614
100 5 15 82 410 130 622
100 5 20 91 410 130 631
150 5 0 86 410 130 626
150 5 5 95 410 130 635
150 5 10 104 410 130 644
150 5 15 113 410 130 653
150 5 20 122 410 130 662
200 5 0 117 410 130 657
200 5 5 126 410 130 666
200 5 10 135 410 130 675
200 5 15 143 410 130 683
200 5 20 152 410 130 692
250 5 0 147 410 130 687
250 5 5 156 410 130 696
250 5 10 165 410 130 705
250 5 15 174 410 130 714
250 5 20 183 410 130 723
G-5
(L_Perm_S)
50 900/900 50 5 0 35 410 130 575

50 5 5 44 410 130 584
50 5 10 52 410 130 592
50 5 15 61 410 130 601
50 5 20 70 410 130 610
100 5 0 65 410 130 605
100 5 5 74 410 130 614
100 5 10 83 410 130 623
100 5 15 92 410 130 632
100 5 20 100 410 130 640
150 5 0 96 410 130 636
150 5 5 105 410 130 645
150 5 10 113 410 130 653
150 5 15 122 410 130 662
150 5 20 131 410 130 671
200 5 0 126 410 130 666
200 5 5 135 410 130 675
200 5 10 144 410 130 684
200 5 15 153 410 130 693
200 5 20 161 410 130 701
250 5 0 157 410 130 697
250 5 5 166 410 130 706
250 5 10 174 410 130 714
250 5 15 183 410 130 723
250 5 20 192 410 130 732
G-6
(L_Perm_S)
55 300/300 50 5 0 18 485 130 633

50 5 5 27 485 130 642
50 5 10 35 485 130 650
50 5 15 44 485 130 659
50 5 20 53 485 130 668
100 5 0 48 485 130 663
100 5 5 57 485 130 672
100 5 10 66 485 130 681
100 5 15 75 485 130 690
100 5 20 84 485 130 699
150 5 0 79 485 130 694
150 5 5 88 485 130 703
150 5 10 96 485 130 711
150 5 15 105 485 130 720
150 5 20 114 485 130 729
200 5 0 109 485 130 724
200 5 5 118 485 130 733
200 5 10 127 485 130 742
200 5 15 136 485 130 751
200 5 20 145 485 130 760
250 5 0 140 485 130 755
250 5 5 149 485 130 764
250 5 10 157 485 130 772
250 5 15 166 485 130 781
250 5 20 175 485 130 790
G-7
(L_Perm_S)
55 600/600 50 5 0 27 485 130 642

50 5 5 36 485 130 651
50 5 10 45 485 130 660
50 5 15 54 485 130 669
50 5 20 62 485 130 677
100 5 0 58 485 130 673
100 5 5 66 485 130 681
100 5 10 75 485 130 690
100 5 15 84 485 130 699
100 5 20 93 485 130 708
150 5 0 88 485 130 703
150 5 5 97 485 130 712
150 5 10 106 485 130 721
150 5 15 115 485 130 730
150 5 20 123 485 130 738
200 5 0 119 485 130 734
200 5 5 127 485 130 742
200 5 10 136 485 130 751
200 5 15 145 485 130 760
200 5 20 154 485 130 769
250 5 0 149 485 130 764
250 5 5 158 485 130 773
250 5 10 167 485 130 782
250 5 15 176 485 130 791
250 5 20 184 485 130 799
G-8
(L_Perm_S)
55 900/900 50 5 0 36 485 130 651

50 5 5 45 485 130 660
50 5 10 54 485 130 669
50 5 15 63 485 130 678
50 5 20 72 485 130 687
100 5 0 67 485 130 682
100 5 5 76 485 130 691
100 5 10 85 485 130 700
100 5 15 93 485 130 708
100 5 20 102 485 130 717
150 5 0 97 485 130 712
150 5 5 106 485 130 721
150 5 10 115 485 130 730
150 5 15 124 485 130 739
150 5 20 133 485 130 748
200 5 0 128 485 130 743
200 5 5 137 485 130 752
200 5 10 146 485 130 761
200 5 15 154 485 130 769
200 5 20 163 485 130 778
250 5 0 158 485 130 773
250 5 5 167 485 130 782
250 5 10 176 485 130 791
250 5 15 185 485 130 800
250 5 20 194 485 130 809
G-9
(L_Perm_S)
60 300/300 50 5 0 20 485 130 635

50 5 5 28 485 130 643
50 5 10 37 485 130 652
50 5 15 46 485 130 661
50 5 20 55 485 130 670
100 5 0 50 485 130 665
100 5 5 59 485 130 674
100 5 10 68 485 130 683
100 5 15 76 485 130 691
100 5 20 85 485 130 700
150 5 0 81 485 130 696
150 5 5 89 485 130 704
150 5 10 98 485 130 713
150 5 15 107 485 130 722
150 5 20 116 485 130 731
200 5 0 111 485 130 726
200 5 5 120 485 130 735
200 5 10 129 485 130 744
200 5 15 137 485 130 752
200 5 20 146 485 130 761
250 5 0 142 485 130 757
250 5 5 150 485 130 765
250 5 10 159 485 130 774
250 5 15 168 485 130 783
250 5 20 177 485 130 792
G-10
(L_Perm_S)
60 600/600 50 5 0 29 485 130 644

50 5 5 38 485 130 653
50 5 10 46 485 130 661
50 5 15 55 485 130 670
50 5 20 64 485 130 679
100 5 0 59 485 130 674
100 5 5 68 485 130 683
100 5 10 77 485 130 692
100 5 15 86 485 130 701
100 5 20 95 485 130 710
150 5 0 90 485 130 705
150 5 5 99 485 130 714
150 5 10 107 485 130 722
150 5 15 116 485 130 731
150 5 20 125 485 130 740
200 5 0 120 485 130 735
200 5 5 129 485 130 744
200 5 10 138 485 130 753
200 5 15 147 485 130 762
200 5 20 156 485 130 771
250 5 0 151 485 130 766
250 5 5 160 485 130 775
250 5 10 168 485 130 783
250 5 15 177 485 130 792
250 5 20 186 485 130 801
G-11
(L_Perm_S)
60 900/900 50 5 0 38 485 130 653

50 5 5 47 485 130 662
50 5 10 56 485 130 671
50 5 15 65 485 130 680
50 5 20 73 485 130 688
100 5 0 69 485 130 684
100 5 5 78 485 130 693
100 5 10 86 485 130 701
100 5 15 95 485 130 710
100 5 20 104 485 130 719
150 5 0 99 485 130 714
150 5 5 108 485 130 723
150 5 10 117 485 130 732
150 5 15 126 485 130 741
150 5 20 134 485 130 749
200 5 0 130 485 130 745
200 5 5 139 485 130 754
200 5 10 147 485 130 762
200 5 15 156 485 130 771
200 5 20 165 485 130 780
250 5 0 160 485 130 775
250 5 5 169 485 130 784
250 5 10 178 485 130 793
250 5 15 187 485 130 802
250 5 20 195 485 130 810
G-12
(L_Perm_S)
65 300/300 50 5 0 21 485 130 636

50 5 5 30 485 130 645
50 5 10 39 485 130 654
50 5 15 48 485 130 663
50 5 20 57 485 130 672
100 5 0 52 485 130 667
100 5 5 61 485 130 676
100 5 10 69 485 130 684
100 5 15 78 485 130 693
100 5 20 87 485 130 702
150 5 0 82 485 130 697
150 5 5 91 485 130 706
150 5 10 100 485 130 715
150 5 15 109 485 130 724
150 5 20 118 485 130 733
200 5 0 113 485 130 728
200 5 5 122 485 130 737
200 5 10 130 485 130 745
200 5 15 139 485 130 754
200 5 20 148 485 130 763
250 5 0 143 485 130 758
250 5 5 152 485 130 767
250 5 10 161 485 130 776
250 5 15 170 485 130 785
250 5 20 179 485 130 794
G-13
(L_Perm_S)
65 600/600 50 5 0 31 485 130 646

50 5 5 39 485 130 654
50 5 10 48 485 130 663
50 5 15 57 485 130 672
50 5 20 66 485 130 681
100 5 0 61 485 130 676
100 5 5 70 485 130 685
100 5 10 79 485 130 694
100 5 15 88 485 130 703
100 5 20 96 485 130 711
150 5 0 92 485 130 707
150 5 5 100 485 130 715
150 5 10 109 485 130 724
150 5 15 118 485 130 733
150 5 20 127 485 130 742
200 5 0 122 485 130 737
200 5 5 131 485 130 746
200 5 10 140 485 130 755
200 5 15 149 485 130 764
200 5 20 157 485 130 772
250 5 0 153 485 130 768
250 5 5 161 485 130 776
250 5 10 170 485 130 785
250 5 15 179 485 130 794
250 5 20 188 485 130 803
G-14
(L_Perm_S)
65 900/900 50 5 0 40 485 130 655

50 5 5 49 485 130 664
50 5 10 58 485 130 673
50 5 15 66 485 130 681
50 5 20 75 485 130 690
100 5 0 70 485 130 685
100 5 5 79 485 130 694
100 5 10 88 485 130 703
100 5 15 97 485 130 712
100 5 20 106 485 130 721
150 5 0 101 485 130 716
150 5 5 110 485 130 725
150 5 10 119 485 130 734
150 5 15 127 485 130 742
150 5 20 136 485 130 751
200 5 0 131 485 130 746
200 5 5 140 485 130 755
200 5 10 149 485 130 764
200 5 15 158 485 130 773
200 5 20 167 485 130 782
250 5 0 162 485 130 777
250 5 5 171 485 130 786
250 5 10 180 485 130 795
250 5 15 188 485 130 803
250 5 20 197 485 130 812
G-15
(L_Perm_S)
70 300/300 50 5 0 23 485 130 638

50 5 5 32 485 130 647
50 5 10 41 485 130 656
50 5 15 49 485 130 664
50 5 20 58 485 130 673
100 5 0 54 485 130 669
100 5 5 62 485 130 677
100 5 10 71 485 130 686
100 5 15 80 485 130 695
100 5 20 89 485 130 704
150 5 0 84 485 130 699
150 5 5 93 485 130 708
150 5 10 102 485 130 717
150 5 15 110 485 130 725
150 5 20 119 485 130 734
200 5 0 115 485 130 730
200 5 5 123 485 130 738
200 5 10 132 485 130 747
200 5 15 141 485 130 756
200 5 20 150 485 130 765
250 5 0 145 485 130 760
250 5 5 154 485 130 769
250 5 10 163 485 130 778
250 5 15 171 485 130 786
250 5 20 180 485 130 795
G-16
(L_Perm_S)
70 600/600 50 5 0 32 485 130 647

50 5 5 41 485 130 656
50 5 10 50 485 130 665
50 5 15 59 485 130 674
50 5 20 68 485 130 683
100 5 0 63 485 130 678
100 5 5 72 485 130 687
100 5 10 80 485 130 695
100 5 15 89 485 130 704
100 5 20 98 485 130 713
150 5 0 93 485 130 708
150 5 5 102 485 130 717
150 5 10 111 485 130 726
150 5 15 120 485 130 735
150 5 20 129 485 130 744
200 5 0 124 485 130 739
200 5 5 133 485 130 748
200 5 10 141 485 130 756
200 5 15 150 485 130 765
200 5 20 159 485 130 774
250 5 0 154 485 130 769
250 5 5 163 485 130 778
250 5 10 172 485 130 787
250 5 15 181 485 130 796
250 5 20 190 485 130 805
G-17
(L_Perm_S)
70 900/900 50 5 0 42 485 130 657

50 5 5 50 485 130 665
50 5 10 59 485 130 674
50 5 15 68 485 130 683
50 5 20 77 485 130 692
100 5 0 72 485 130 687
100 5 5 81 485 130 696
100 5 10 90 485 130 705
100 5 15 99 485 130 714
100 5 20 107 485 130 722
150 5 0 103 485 130 718
150 5 5 111 485 130 726
150 5 10 120 485 130 735
150 5 15 129 485 130 744
150 5 20 138 485 130 753
200 5 0 133 485 130 748
200 5 5 142 485 130 757
200 5 10 151 485 130 766
200 5 15 160 485 130 775
200 5 20 168 485 130 783
250 5 0 164 485 130 779
250 5 5 172 485 130 787
250 5 10 181 485 130 796
250 5 15 190 485 130 805
250 5 20 199 485 130 814
G-18

Turn Lane Lengths For Various Speed Roads and Evaluation of Determining Criteria

Uploaded by

Copyright:

Available Formats

Turn Lane Lengths For Various Speed Roads and Evaluation of Determining Criteria

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Turn Lane Lengths For Various Speed Roads and Evaluation of Determining Criteria

Uploaded by

Copyright:

Available Formats

2008-14

Turn Lane Lengths for Various Speed Roads and

17. Document Analysis/Descriptors 18. Availability Statement

Operator Performance Laboratory (OPL)

Minnesota Department of Transportation

Table 1. Comparison of length requirements (in number of vehicles): Proposed

1. Identification of the type of turn: Left or Right;

These recommendations are applicable for the following range of parameters:

1.1 Problem Statement

1.2 Solution Approach

2.1 Simulation and Prediction Tools

2.1.1. Survey of Intersection Design Practices and Standards

60.0% Signalized Left Turn

Figure 2. Source of guidelines for signalized and unsignalized turn length

Figure 3. Guidelines efficiency

Figure 4. Scheme of Unsignalized intersection with turning lanes

Source: Chakroborty et al., 1995

Figure 5. Lane overflow and blockage of lane entrance at a signalized intersection

Table 2. Comparison of NETSIM with proposed model results

Source: Kikuchi et al., 1993

2.3 Existing Guidelines

Table 3. Guide for left-turn lanes on two-lane highways

Source Deceleration Storage Length Taper Length

For evaluation of right-turn lanes, the formula is

Figure 8. Right-Turn Lane Warrants

3.1 Selection and Documentation of Intersections

Figure 9. Overall map of area

Intersection Name Lat/Lon Signal Type Description

3.2 Queue Length Observation in Field

3.3 Analysis of Field Data

for Right Turns

# Intersection Arterial Road Crossroad Number of

Left Turn Right Turn Left Turn Right Turn

Protected Permitted Permitted Free Yield Free Yield

Figure 10. Classification chart

3.4 Simulation Model Calibration

4.1 Simulation Model Parameters

4.2 Simulation Model Data

For Protected left turns on Signalized intersections

Aver = -8.64 + 0.00618*TV + 0.881*LTV - 0.119*Sp - 2.24*HVT + 0.863*HVL + 0.055*Gr

(Parameters are defined in Section 4.1 )

S = 21.51; R2 = 89.8%; R2 (adjusted) = 89.7%

For Permitted left turns on Signalized intersections

Aver = 20.7 - 0.00159*TV + 0.0233*OV + 0.483*LTV - 0.580*Sp - 0.192*HVT + 0.382*HVL -

(Parameters defined in Section 4.1 )

S = 12.70; R2 = 91.5%; R2 (adjusted) = 91.3%

Aver = -9.75 + 0.00179*TV + 0.0216*OV + 0.226*LTV - 0.0626*Sp + 0.249*HVT -

(Parameters are defined in Section 4.1 )

S = 5.987; R2= 85.5%; R2 (adjusted) = 85.0%

Aver = -29.4 + 0.00248*TV + 0.0203*OV + 0.416*LTV + 0.133*Sp - 0.315*HVT + 0.922*HVL

(Parameters are defined in Section 4.1 )

S = 17.82; R2 = 78.7% ; R2 (adjusted) = 78.4%

Aver = -7.69 + 0.0379*TV - 0.0409*OV + 0.859*LTV - 0.298*Sp - 2.92*HVT+ 0.106*HVL +

(Parameters are defined in Section 4.1 )

S = 38.19; R2 = 76.4%; R2 (adjusted) = 76.2%

Aver = -6.24 + 0.0282*TV - 0.0565*OV + 0.799*LTV - 0.101*Sp - 2.78*HVT+ 0.269*HVL +

(Parameters are defined in Section 4.1 )

S = 40.71; R2 = 72.3%; R2 (adjusted) = 72.1%

For Free right turns on Signalized intersections

Aver = -8.64 + 0.00618TV + 0.881LTV - 0.119Sp - 2.24HVT + 0.863HVL + 0.055Gr

Aver = 20.7 - 0.00159TV + 0.0233OV + 0.483LTV - 0.580Sp - 0.192HVT + 0.382HVL -

Aver = -9.75 + 0.00179TV + 0.0216OV + 0.226LTV - 0.0626Sp + 0.249*HVT -

Aver = -29.4 + 0.00248TV + 0.0203OV + 0.416LTV + 0.133Sp - 0.315HVT + 0.922HVL

Aver = -7.69 + 0.0379TV - 0.0409OV + 0.859LTV - 0.298Sp - 2.92HVT+ 0.106HVL +

Aver = -6.24 + 0.0282TV - 0.0565OV + 0.799LTV - 0.101Sp - 2.78HVT+ 0.269HVL +

LTprot = 35.3 + 0.0203TV + 1.14LTV - 0.171Sp – 6.75HVT+ 1.32HVL - 0.16Gr

LTperm = − 45.2 - 0.00953TV + 0.0406OV + 0.610LTV + 0.348Sp + 0.812*HVT+

LTyield = 0.00 + 0.00315TV + 0.0332OV + 0.345LTV - 0.149Sp + 0.224*HVT+

RTperm = 65.5 + 0.0323TV + 0.0533CV + 0.186RTV - 0.829Sp - 1.50HVT + 0.818HVR +

RTfree_S = - 61.5 + 0.0551TV + 0.265CV + 0.317RTV - 0.527Sp + 0.041HVT + 1.04HVR

RTyield = 16.0 - 0.00497TV + 0.0235CV + 0.0957RTV - 0.115Sp - 0.038HVT + 0.274HVR

RTfree_U = 13.2 − 0.00021TV + 0.0176CV + 0.0876RTV - 0.0615Sp - 0.094*HVT +

Volume of left-turn vehicles