Office of Operations
21st Century Operations Using 21st Century Technologies

Decision Support Framework and Parameters for Dynamic Part-Time Shoulder Use:
Considerations for Opening Freeway Shoulders for Travel as a Traffic Management Strategy

Printable version [PDF 6.7 MB]
You may need the Adobe® Reader® to view the PDFs on this page.
Contact Information: Operations Feedback at OperationsFeedback@dot.gov

U.S. Department of Transportation - Federal Highway Administration (logo)

U.S. Department of Transportation
Federal Highway Administration
Office of Operations
1200 New Jersey Avenue, SE
Washington, DC 20590
ops.fhwa.dot.gov

November 2019
FHWA-HOP-19-029


Table of Contents

[ Notice and Quality Assurance Statement ] [ Technical Report Documentation Page ] [ SI Modern Metric Conversion Factors ] [ Acronyms ]

Executive Summary
Chapter 1. Introduction
No Value Choosing Part-time Shoulder Use
No Value Relationship with Other Active Traffic Management Treatments
No Value Network Considerations
No Value Purpose, Scope, and Target Audience
No Value Organization of Report
Chapter 2. What is Dynamic Part-Time Shoulder Use?
No Value Dynamic Part-Time Shoulder Use and Operations
No Value Decision to Open/Close Dynamic Part-Time Shoulder
No Value Known Deployments
No Value Dynamic Part-Time Shoulder Use Research
No Value Advantages and Challenges of D-PTSU Over S-PTSU
No Value Cost of Dynamic Part-Time Shoulder Use
Chapter 3. Decision Support Framework for Dynamic Shoulder Use Operations
No Value Systems Engineering
No Value Developing the Concept of Operations
No Value Candidate Part-Time Shoulder Use Facilities
No Value Selecting the Level of Dynamic Part-Time Shoulder Use
No Value Selecting Shoulder Operations Decision Parameters
No Value Developing the Decision Support Framework
No Value Considerations for Permanent Shoulder Conversion
Chapter 4. Decision Parameters for Opening the Shoulder
No Value Methods for Selecting Shoulder Use Type and Decision Parameters
No Value Use Cases for Shoulder Use and Decision Parameter Selction
No Value Bottleneck Identification
No Value Method I: Demand-to-Capacity Patterns
No Value Method II: Empirical Performance Data
No Value Method III: Macroscopic Decision Parameter Optimization
No Value Method IV: Microscopic Decision Parameter Refinement
No Value Method V: Monitoring and Adjustment
No Value Conclusions
Chapter 5. Closing the Shoulder
No Value Realtime and Predicted Traffic Conditions
No Value Maintenance, Incidents, and Emergency Response
No Value Safety Considerations
Appendix A. Part-Time Shoulder Use Questions
Appendix B. Dynamic Part-time Shoulder Use Applications Fact Sheets
Appendix C. Decision Parameter Development Methods
Appendix D. Generalized Thresholds for Opening Shoulder
Appendix E. Additional Resources
Acknowledgments
References

List of Figures

Figure 1. Diagram. Considerations in choosing part-time shoulder use.
Figure 2. Photo. Yellow dashed lines divide the left shoulder (used for part-time travel) from the general purpose lanes in Colorado.
Figure 3. Photo. A lane-use control sign (on the far right side) indicates whether the shoulder is open or closed to traffic.
Figure 4. Photo. A left shoulder is available for travel on a dynamic part-time shoulder use facility in Denmark.
Figure 5. Photo. A right shoulder is available for travel on a dynamic part-time shoulder use facility in Denmark.
Figure 6. Diagram. Systems Engineering V diagram for intelligent transportation systems projects.
Figure 7. Diagram. Considerations in choosing part-time shoulder use.
Figure 8. Diagram. Decision parameters for opening a shoulder to travel based on predicting breakdown.
Figure 9. Diagram. Decision parameters for opening a shoulder to travel based on an observed breakdown.
Figure 10. Diagram. Events preceding the opening of a dynamic shoulder.
Figure 11. Diagram. Example shoulder opening decision tree.
Figure 12. Diagram. Example shoulder closing decision tree.
Figure 13. Diagram. Example application of speed and volume decision parameters.
Figure 14. Chart. Example of freeway sensor metering due to congestion.
Figure 15. Photo. Gantry with dynamic lane use signs I-66 in Northern Virginia.
Figure 16. Chart. Speed heat map for eastbound I-66 analysis segment from probe data.
Figure 17. Charts. Compound figure depicts speed band comparison for I-66 eastbound.
Figure 18. Charts. Compound figure depicts sample Product Limit Method analysis for the morning and afternoon peak on a freeway in California.
Figure 19. Charts. Compound figure depicts temporal distribution of breakdown events.
Figure 20. Chart. Using Highway Capacity Manual speed-flow curves to select thresholds.
Figure 21. Chart. Identification of threshold for congested speeds.
Figure 22. Chart. Probability of breakdown in next 15 minutes.
Figure 23. Chart. Example a.m. peak period speed-flow data for one day.
Figure 24. Chart. Annual percent of time shoulder is open or closed in a.m. peak.
Figure 25. Diagrams. Compound figure illustrates ramp-freeway junction types in FREEVAL experiment.
Figure 26. Chart. Demand profiles for three-lane facility.
Figure 27. Chart. Use of realtime and historical data for part-time shoulder use decisionmaking.
Figure 28. Chart. Effects of varying peak bottleneck d/c ratios for a three-lane, Type A facility geometry and demand increase slope (offset = 30, medium).
Figure 29. Chart. Comparison of vehicle hours delay and number of periods the shoulder is open for a three-lane Type A merge facility for a peak demand-to-capacity ratio of 1.06 and no slope offset.
Figure 30. Chart. Comparison of vehicle hours delay and number of periods the shoulder is open for a three-lane Type A merge facility for a peak demand-to-capacity ratio of 1.06 and a 50-minute slope offset.
Figure 31. Charts. Compound figure depicts network delay comparison under various decision parameter variations and maximum demand to capacity ratios.
Figure 32. Charts. Compound figure compares delay reduction and shoulder open duration under various decision parameter variations for different maximum demand to capacity ratios and slope offsets.
Figure 33. Charts. Compound figure depicts vehicle throughput for slope offset = 0, maximum d/c = 1.04 under speed decision parameter = 55 mi/h and volume decision parameter = 0.8*d/c scenarios.
Figure 34. Map. Dynamic part-time shoulder use in Colorado.
Figure 35. Photo. Example of dynamic part-time shoulder use open.
Figure 36. Photo. Example of dynamic part-time shoulder use closed.
Figure 37. Map. Dynamic part-time shoulder use in Michigan.
Figure 38. Photo. Dynamic shoulder lane in construction.
Figure 39. Photo. Rendering showing dynamic shoulder lane open.
Figure 40. Map. Dynamic part-time shoulder use in Minneapolis, Minnesota.
Figure 41. Photo. Dynamic shoulder lane open.
Figure 42. Photo. Dynamic shoulder lane closed.
Figure 43. Map. Dynamic part-time shoulder use in Virginia.
Figure 44. Photo. Dynamic shoulder lane open.
Figure 45. Photo. Dynamic shoulder lane closed.
Figure 46. Map. Dynamic part-time shoulder use in Washington.
Figure 47. Photo. Dynamic shoulder lane open.
Figure 48. Photo. Dynamic shoulder lane closed.
Figure 49. Photo. Dynamic shoulder lane open.
Figure 50. Photo. Dynamic shoulder lane open on the right.
Figure 51. Photo. Dynamic shoulder lane open on the left.
Figure 52. Photo. Dynamic shoulder lane open.
Figure 53. Photo. Dynamic shoulder lane being monitored.
Figure 54. Equation. The capacity distribution function.
Figure 55. Equation. The product-limit estimator for the capacity or breakdown probability distribution.
Figure 56. Equation. The product-limit estimator for the probability of observed breakdown volume being greater than the observed volume.
Figure 57. Equation. Product-limit estimator for observed volume that causes a breakdown but is considered separately.
Figure 58. Equation. The likelihood function for capacity analysis.

List of Tables

Table 1. Advantages and Challenges of D-PTSU.
Table 2. Dynamic part-time shoulder use facilities in the United States.
Table 3. International dynamic part-time shoulder use facilities.
Table 4. Cost component considerations for dynamic part-time shoulder use and static part-time shoulder use.
Table 5. Estimated cost ranges for key components.
Table 6. Levels of part-time shoulder use.
Table 7. Highway Capacity Manual summary of bottleneck capacity estimated in vehicles (per hour per lane).
Table 8. Target demand-to-capacity ratios for part-time shoulder use viability.
Table 9. Parameter variations of experiments and total scenarios analyzed in FREEVAL.
Table 10. Different thresholds for the two decision parameter types.
Table 11. Example of appendix D table—minutes until capacity is reached when per lane capacity is 1,900 veh/h/ln.
Table 12. Parameter variations of experiments analyzed in microsimulation.
Table 13. Minutes until Capacity is reached when per lane capacity is 2,100 veh/h/ln.
Table 14. Minutes until Capacity is reached when per lane capacity is 2,000 veh/h/ln.
Table 15. Minutes until Capacity is reached when per lane capacity is 1,900 veh/h/ln.
Table 16. Minutes until Capacity is reached when per lane capacity is 1,800 veh/h/ln.
Table 17. Minutes until Capacity is reached when per lane capacity is 1,700 veh/h/ln.
Table 18. Minutes until Capacity is reached when per lane capacity is 1,600 veh/h/ln.
Table 19. Minutes until Capacity is reached when per lane capacity is 1,500 veh/h/ln.

next

Office of Operations