1. Left-Turn Operation When Opposed by Multiple Lanes
April 27, 2012

2. The Problem Given Question
Subject left-turn vehicle opposed by 2+ lanes, one is LT/TH lane
Question
How does a queue in the opposing shared lane affect subject left-turn vehicle?

3. The Problem
Outside Lane QAP
Setup
At what time can subject vehicle filter through opposing vehicles?
Opposing Vehicles
Outside lane
Inside lane
Subject Vehicle
Inside Lane QAP

4. The Problem
Issue 1
How is each opposing lane considered by the subject left-turn drivers while assessing the availability of gaps in the opposing traffic stream?
Does this behavior change if a left-turn vehicle is at the stop line of the opposing inside lane?

5. Results Using Other Methods
2000 HCM Method
F-B. Lin Method
CORSIM

6. Results Using Other Methods
Test Bed Intersection
NB and SB Volume
100 to 1500 veh/h

7. Results Using Other Methods
Approach Control Delay
Northbound approach
20% Left Turns
10% Left Turns

8. Results Using Other Methods
Outside Lane QAP
HCM 2000
gq = duration of gs , which is based on...
50% traffic in outside lane
Outside lane sth = 1800 veh/h/ln
Queue in inside lane not a factor
Outside lane
Inside lane
Subject Vehicle
Inside Lane QAP - not considered

9. Results Using Other Methods
F-B. Lin Method
Vo = f(PL , PLTO , G, C, Va , Vo )
Vo ≈ opposing shared lane volume – opposing left-turn volume – 20
Suggests that opposing lefts that arrive randomly after queue clearance have limited effect on subject left-turn driver’s gap selection

10. Results Using Other Methods
CORSIM
Northbound approach (10% lefts)
Presence of left turns on SB approach increases NB left turn capacity, which reduces delay

11. Findings Findings from Other-Method Evaluations
Left-turn vehicles in the opposing shared left-turn lane appear to increase the capacity of the subject left-turn movement
Similar to single-lane model in 2010 HCM
When first vehicle in opposing shared lane queue is left-turn vehicle, the queue presence does not appear to restrict the subject left-turn driver’s decision to turn
Opposing lefts that arrive randomly after queue clearance have limited effect on subject left-turn driver’s gap selection

12. Recommendations Option 1 Option 2
Update the HCM methodology to explicitly address the case of an opposing multilane approach with a shared permitted left-turn lane
Consider queue in all opposing lanes
Minimize effect of opposing left-turn vehicles in shared lane on subject left-turn movement
Option 2
Interim adjustments to 2010 HCM method
Queue clearance time
Permissive capacity

13. Recommendations Option 2 (continued) Queue clearance time
Improve prediction of lane group volume on opposing approach
Procedure in 2010 HCM intended to balance volume based on equalizing queue service time among lanes
Improvements identified in next slide
Compute for each opposing lane group, excluding any opposing shared left-turn lane
Similar to method in 2000 HCM for multilane approaches
Would need research (w/field data) to calibrate queue service time model for shared left-turn lane, so ignore for interim

14. Recommendations Option 2 (continued)
Saturation flow rate for lane-group volume procedure
Sat. flow rate dictates lane-group volume, which dictates when queue clears
Issue with procedure for predicting lane group volume
Method of addressing sneaker capacity in this procedure is pessimistic such that many through vehicles avoid shared lane
This leads to an unrealistically long queue clearance time in through lane
But, using sneakers = 1 + PL overcompensates
So, equation below is recommended 2
Sneakers PL 1

15. Recommendations Option 2 (continued) Opposing flow rate
Exclude the left-turn volume in a shared lane
Based on two cases
Case 1: Vo = opposing through + right volume
Case 2: Vo = opposing through volume
If opposing right is in a shared lane, use Case 1
If opposing right is in an exclusive lane, analyst must determine whether to use Case 1 or 2 based on local driver behavior, traffic conditions, and intersection geometry

16. Recommendations Option 2 (continued) Follow-Up Headway
Follow-up headway likely is a function of PL
Tf = 2.5 s when PL = 1.0
Tf = 4.5 s when PL = 0.5
Value of Tf most critical as PL  1.0
Use 2.5 s (instead of 4.5 s)

17. Evaluation Both Opposing Approaches are MSPLT 10 percent left turns
Current
Revised

18. Evaluation Both Opposing Approaches are MSPLT 20 percent left turns
Current
Revised

19. Closure
Questions or Comments?