1. USING PeMS DATA TO EMPIRICALLY DIAGNOSE FREEWAY BOTTLENECK LOCATIONS IN ORANGE COUNTY, CALIFORNIA Robert L. Bertini Portland State University Aaron M. Myton Oregon Department of Transportation TRB 84 th Annual Meeting Washington D.C. January 9-13, 2005

2. Objectives Identify Freeflow  Congested Conditions Determine Bottleneck Locations Measure Queue Discharge Features Conclusions

3. Foundation for Understanding Freeway Operations Queues Cause Delays Loop Detectors Good Data Source Previous Studies –Throughput Increase 2-5% Prior To Queue –Diverge Bottleneck Identified PeMS Provides Opportunity for Detailed Data Analysis Background

4. Culver Dr. University Dr. Sand Canyon Ave. Hwy. 133 Irvine Center Dr. Travel Direction 130 110100 90 80 70 60 50 40 3020 10 Site Map Loop Detector Station Number L E G E N D 110 120 Four Northbound Lanes One Northbound HOV Lane Freeway Ramps Between Stations, Except 100/110 140 N Post-mile

5. Bottleneck Definition Queue upstream Freely-flowing traffic downstream Temporally and spatially variable

6. Data Freeway Performance Monitoring System (PeMS) –http://pems.eecs.berkeley.edu/Publichttp://pems.eecs.berkeley.edu/Public Single Inductive Loops Vehicle Occupancy and Count Data Aggregated in 30-sec. Intervals On-Ramp/Off-Ramp Data Not Available Monday June 1, 1998

7. Methodology Cumulative Curves – Vehicle Count & Occupancy Oblique Plots N and T

9. Analysis Time, t 0.60 1.60 2.60 3.60 4.60 5.60 7:007:308:008:309:00 Post mile Station 140 130 120 110 100 90 80 70 60 50 40 30 20 10

10. SEGMENT # QUEUE NO QUEUE 7:41:30 8130 9762 Q U Culver Dr. University Dr. Sand Canyon Ave. Hwy. 133 Irvine Center Dr. Travel Direction Station 140 130 120 110 100 90 80 70 60 50 40 30 20 10 Station 70 – Loops 1-4 Oblique N and T

11. Time, t 0.60 1.60 2.60 3.60 4.60 5.60 7:007:308:008:309:00 Post mile Station 140 130 120 110 100 90 80 70 60 50 40 30 20 10

12. SEGMENT # QUEUE NO QUEUE Q U Culver Dr. University Dr. Sand Canyon Ave. Hwy. 133 Irvine Center Dr. Travel Direction Station 140 130 120 110 100 90 80 70 60 50 40 30 20 10 Station 60 – Loops 1-4, Oblique N and T

13. Time, t 0.60 1.60 2.60 3.60 4.60 5.60 7:007:308:008:309:00 Post mile Station 140 130 120 110 100 90 80 70 60 50 40 30 20 10

14. Time, t 0.60 1.60 2.60 3.60 4.60 5.60 7:007:308:008:309:00 Post mile Queue 1 Queue 2 Queue 4 Queue 5 Queue 7 Queue 3 Queue 6 Queue 9 Queue 10 Queue 8 Station 140 130 120 110 100 90 80 70 60 50 40 30 20 10 Queue - Activation Queue - Deactivation LEGEND

15. Results

16. Conclusions Ten Queues Formed at Five Locations Flow Reduction of 3-15% No Queue in HOV Lanes Bottlenecks Typically Activated Near Ramps No Speed Threshold Needed

17. Future Research Study site on different days Manual replication of procedure Develop semi-automated method

18. Acknowledgements Prof. Pravin Varaiya and Prof. Alexander Skabardonis, U.C. Berkeley Dr. Karl Petty, Berkeley Transportation Systems Pherak Hay, Vu Mai, and Roger Lindgren

21. SEGMENT # QUEUE NO QUEUE 7:41:30 8130 9762 Q U Station 70 – Loops 1-4 Oblique N and T

22. SEGMENT # QUEUE NO QUEUE Q U Station 60 – Loops 1-4, Oblique N and T