This research addresses the safety of two distinct but related designs for increasing freeway operational performance: part-time shoulder use and High Occupancy Vehicle/High Occupancy Toll (HOV/HOT) lanes.
Part-time shoulder use is an increasingly popular and practical performance-based design strategy for both reducing freeway congestion and improving bus travel time and reliability. In general, there are three types of part-time shoulder use: bus-on-shoulder (BOS), static shoulder use, and dynamic shoulder use. BOS is limited to authorized buses driven by trained drivers. Static shoulder use and dynamic shoulder use allow all or most vehicles to use the shoulder when it is open; some facilities prohibit trucks from using the shoulder. Static shoulder use opens the shoulder to traffic on a fixed schedule, and dynamic shoulder use opens the shoulder to traffic in response to traffic conditions. Sixteen states have one or more part-time shoulder facilities in operation, and many of these have opened within the past five to ten years. FHWA published a guide on part-time shoulder use in February 2016, and the lack of tools for quantifying the safety impact of part-time shoulder use was identified as a major research gap by the guide’s authors.
The current Highway Safety Manual (HSM) freeway models cannot be used to evaluate the potential safety performance of part-time shoulder use. Observed crash data along with other operations aspects of part-time shoulder use cannot be modeled using HSM freeway chapters because these predictive models do not address shoulder use in particular.
While it may not be possible to develop crash prediction models for part-time shoulder use at this time, research is needed to develop a quantitative safety analysis of the part-time shoulder use operations as well as documentation and guidance that provide the profession a better understanding of the potential tradeoffs as well as the elements that create successful (limited collisions) operations. A quantitative safety assessment will highlight the critical data elements needed for a safety analysis—crash types, start and end time of part-time shoulder operation, temporal and spatial characteristics of crashes, distance from the entrance ramp and exit ramp, AADT on the general purpose travel lane, etc.—obtained from existing observed crash data of at least three to five years from different facilities across the country. The analysis will further consider volume and basic geometric information (such as the number of lanes) as crash predictors as well as temporal factors such as density, speed, and lane-by-lane variations of the two. There is evidence these factors impact freeway crash frequency, and facilities with changing lane configuration may be more impacted. This aspect of the research will address Themes 3.3 and 4.4 of the action plan in the current AASHTO SCOHTS strategic safety plan.
HOV/HOT lanes are involved in many recent traffic operational improvements for urban and suburban freeways in the U.S. The term ‘managed lanes’ has been applied to the overall practice and both HOV and HOT lanes are examples of managed lanes. According to the TRB Committee on Managed Lanes database of facilities, the total centerline mileage of electronically tolled HOT facilities in the country is around 1145 lane-miles. The physical separation of the general-purpose and HOV/HOT lanes usually varies from 2 to 22 feet and may make use of a buffer zone or barriers like channelizers and bollards. As more agencies retrofit existing HOV lanes to HOT or introduce a managed lane into an existing corridor, fewer separation barriers are being built. These geometric and operational aspects of HOV/HOT facilities impact safety on the entire facility in urban and suburban freeway sections.
Design elements of freeways with HOV/HOT lanes that influence safety include lane orientation (i.e., contra-flow or concurrent flow), lane access type (i.e., continuous or limited), and lateral separation from general purpose lanes (i.e., buffer or barrier). Research indicates that the addition of HOV lanes can increase crashes, depending on the tradeoffs made between components of the full freeway cross section. Increasing the buffer width can reduce crashes, provided that this additional buffer width does not lead to compromises in the width of the traffic lanes or shoulders. A few states have added flexible tubular delineators to the buffer; however, anecdotal evidence is that safety is degraded by these supplemental devices. The toll facilities associated with HOT lanes represent an area of increased lane changing and queuing. These characteristics are likely to increase crash risk, especially if the area provided for lane changing and queuing is constrained. The safety effect of toll facilities is also likely to be influenced by the percentage of vehicles that use electronic toll tags.
Research is needed to develop a safety prediction method that can be used to estimate the expected crash frequency of a range of freeway facilities with HOV or HOT with various geometric and traffic volume characteristics. The method will provide information useful to the planning and design of these facilities, and, thus, it must be sensitive to a wide range of HOV and HOT lane design elements. This aspect of the research will address Themes 4.4 and 6.6 of the action plan in the current AASHTO SCOHTS strategic safety plan.
The objectives of this research are to quantify and expand the knowledge base of safety performance of part-time shoulder use (Part A) and to develop a predictive methodology to estimate crash frequency and severity for freeway facilities, and associated ramps, with HOV or HOT lanes (Part B).
In Part A, the research shall determine how the safety performance of freeways with part-time shoulder use differs from the safety performance of freeways without shoulder use and how the safety performance of freeways with part-time shoulder use varies based upon traffic volumes and the design and operating features of the shoulder. Further, the research shall distinguish between bus-on-shoulder (BOS) operation and shoulder use open to general purpose traffic (static or dynamic shoulder use). Static and dynamic shoulder use shall be grouped together due to the similar operating characteristics, the existence of static facilities that incorporate some elements of dynamic facilities, and the small number of existing dynamic shoulder use facilities.
This research shall specifically exclude shoulder use in work zones, conversion of a shoulder into a full-time (24 hour a day) lane, and shoulder use on arterials.
In Part B, the research shall address the development of predictive methodology to help highway agencies consider safety in decisions about planning and designing freeways with HOV or HOT lanes on urban freeways. The predictive methods shall include safety performance functions, crash modification factors (CMFs), and calibration factors in a format that is consistent with the predictive methods in the existing HSM Part C. Separate predictive methods shall be developed for freeways and interchange ramps with HOV or HOT components. The methods shall be sensitive to the traffic volumes on the freeways and ramps and to the geometric design elements and traffic control features considered by engineers and planners during the project development process. Coordination of the research is required with the joint AASHTO/TRB activities related to the HSM, FHWA’s Interactive Highway Safety Design Model (IHSDM), AASHTO’s SafetyAnalyst software tools, the 3rd Edition of the Human Factors Guide, and the Towards Zero Deaths strategic highway safety planning initiative.
Note: At the direction of the AASHTO Standing Committee on Research, NCHRP Project 17-89 combines the objectives of problem statements 2018-B-12, “HOV/HOT Crash Prediction Method for the Highway Safety Manual,” and 2018-G-03, “Safety Performance of Part-Time Shoulder Use.” The NCHRP project panel will decide whether to conduct research to accomplish the two objectives as one project or two.