Since President Dwight D. Eisenhower signed the Federal-Aid Highway Act of 1956, the national Interstate System has served an essential role in America’s prosperity and way of life. It, among other influences, connects diverse geographical areas, provides access to employment, and decreases driving distances and travel times. Thus, it is of national interest to continually preserve and enhance the system by assuring it provides the highest level of service in terms of safety and mobility for all users. The Interstate System also influences the mobility and safety of people and goods by providing access to local highways and networks of public streets. Complete control of access along the Interstate mainline and ramps and control of access on the crossroad at interchanges is critical to providing the highest level of service.

The Federal Highway Administration provides the requirements for the justification and documentation necessary to substantiate any proposed changes in access to the Interstate System. Changes in access to the Interstate System may involve proposing a new interchange or proposing modifications to an existing interchange. FHWA’s decision to approve new or revised access points to the Interstate System—under Title 23, United States Code, Section 111—-is made after considering an analysis of expected operational and safety performance impacts.

Typically, substantiated information comes by way of Interchange Justification Reports (IJRs). IJRs are written early in the project planning process, with details generally consistent with conceptual design and descriptions of potential safety performance impacts.

Historic crash data shows that high traffic facilities such as freeway interchanges—like the one at Interstate 75 (I–75) and University Parkway in Sarasota, FL—exhibit stable annual traffic crash patterns over time; therefore, their annual safety performance should be predictable. “When an existing interchange or intersection exhibits a history of safety and/or operational problems and is being proposed for conversion into a new interchange or configuration, it helps to know the expected safety performance,” says Brian Cronin, former director of the FHWA Office of Safety and Operations Research and Development. “Although analysts can identify many causal factors leading to traffic crashes, so far, no simple and reliable predictive methodology has been developed that can accurately predict an interchange’s safety performance in terms of total annual crashes and severity distribution,” Cronin continues.

ISAT and ISATe: The Initial Predictive Tools

In 2007, FHWA completed the Interchange Safety Assessment Tool (ISAT), a spreadsheet-based tool for performing safety assessments of interchanges and adjacent roadway segments and intersections. This original tool used a building block approach and focused on traffic volumes, the number of lanes on each interchange roadway segment, ramp configurations, and ramp terminal control. ISAT also provided limited opportunity to consider design features and the differences in interchange configurations. ISAT was developed as an intermediate tool to meet immediate needs, providing crash estimates of only three typical interchange configurations—diamond, partial cloverleaf, and full cloverleaf.

In 2009, FHWA began collaborating with the Transportation Research Board to develop the Enhanced Interchange Safety Analysis Tool (ISATe), an improved prediction methodology and safety analysis tool for corridor and site-specific analysis. ISATe also provides information about the relationship between roadway geometric design features and safety. This version of the tool automates a safety prediction method consisting of algorithms and equations. The lead researchers of ISAT and ISATe worked as the lead researchers in the development of the Highway Safety Manual (HSM). Hence, many of the methodologies used in the interchange safety tools are adopted in HSM.

Eventually, in 2014, ISATe became a supplemental component of HSM, which provides detailed descriptions of the underlying data needs and predictive models implemented in ISATe and serves as the primary means for evaluating safety performance for freeway segments, ramp segments, and ramp terminals. To predict crashes for these facilities, the ISATe and HSM predictive models require detailed sets of inputs—down to specific design characteristics—for features such as lane width and shoulder width as well as specific locations and design of barriers and curves for freeway facilities. Ramp connection locations and ramp terminal design details (e.g., channelization and turn-phasing) are also required. These details are not commonly known during the planning phase, leaving agencies unable to use the HSM models for purposes of safety performance assessment. In addition, even for instances where the model inputs are available during the planning phase, aggregating component-by-component predictions (e.g., individual ramp and freeway segment) may only partially capture the safety performance impacts when considering a project location. Finally, some interchange configurations, such as diverging diamond interchanges, cannot yet be evaluated using HSM predictive models.

As an alternative approach to using ISATe during the planning phase, particularly for interchange configurations not considered in the existing tool, analysts can use a single crash modification factor (CMF) to assess the difference in expected safety performance from one interchange configuration to another. Unlike HSM models, a single CMF fails to capture complex interactions taking place and cannot be used to evaluate factors within an interchange configuration that may impact safety performance. However, this is a reasonable approach given what is known at the planning stage of a project.

Developing the Next Best Tool

While ISATe filled a critical need, analysts identified several difficulties in utilizing the tool. When funding became available in fiscal year 2020 to support the need for substantive safety performance assessments for IJRs, FHWA began developing planning-level models and tools to predict crash frequency and severity for an existing or proposed interchange. Published in early 2023, these planning-level models now allow analysts to compare the potential safety performance effects of freeway access and interchange design decisions earlier in the project development process without knowing the geometric details.

To maximize the applicability of a planning-level interchange configuration safety comparison tool, FHWA began identifying interchange types in the top 75 percent of those considered by State transportation agencies. “Although numerous IJRs are submitted to FHWA for review, the types of interchange configurations evaluated in those IJRs concentrate into mainly just a few categories. If we have a tool that can assess the safety performance of these more commonly studied types of interchanges, it will be a huge relief to practitioners involved in evaluating safety on IJRs,” says Mark Doctor, senior safety and design engineer at FHWA Resource Center.

The research team surveyed FHWA division representatives for all 50 States, plus Washington, D.C., and Puerto Rico, to identify how commonly each interchange configuration is considered or constructed in each State, city, or territory annually. The survey included a graphical representation of potential interchange configurations to choose from, in order of importance or value to the State.

FHWA representatives from 47 divisions responded to the electronic survey, providing detailed descriptions of practices and information on access proposals received over the past few years in their respective locations. The results of the survey indicated the following configurations should be included in the next improved version of the predictive tool for interchanges (in order of most to least responses):

• Standard diamond interchange.
• Diverging diamond interchange.
• Single-point diamond interchange.
• Compressed diamond interchange.
• Tight diamond interchange.
• Partial cloverleaf (Parclo) type A.
• Parclo type B.
• Parclo type AB.

Additionally, this survey revealed the need to differentiate single roundabout interchanges from diamond interchanges with roundabout ramp terminals, otherwise referred to as roundabout diamond interchanges. The development of the new Interchange Configuration Safety Comparison Tool included roundabout diamond interchanges.

Identifying the Scale and Scope of the New Predictive Model

The primary consideration for evaluating interchange safety performance was identifying the analysis area scope for the new predictive model. Due to the variability of an interchange’s influence area (or area of study), a planning-level predictive method based on influence area was impractical. This level of analysis would have included the freeway sections outside of the interchange area—all the way to adjacent interchanges and crossroad sections—to at least the nearest intersections. Therefore, the predictive method focused on the interchange area. The interchange area consists of the freeway mainline, crossroad, and ramp terminals. The freeway mainline includes 457 meters (1,500 feet) upstream and downstream of the entrance and exit ramp gores and 30 meters (100 feet) beyond the curb return on the crossroad, which is consistent with other planning-level safety analysis tools. Additionally, this distance provides consistency with ISATe’s research, which shows a decreasing influence from ramp entrances and exits, with no influence beyond 805 meters (2,640 feet), when there is not an adjacent interchange nearby. Analysts included all crashes observed in the interchange area in the scope for this tool; however, the tool does not predict crash frequency by location within the interchange area, only for the interchange area as a whole.

The method underlying this tool separately predicts fatal, injury, and property damage-only crash frequency. When combined, the tool predicts the total crash frequency for each interchange configuration considered. Moreover, the tool includes inputs to evaluate the relationship between geometric and operational features on the predicted severity level of injury crashes. Analysts can use the tool to evaluate the predicted crash frequency for varying interchange configurations, considering a standard diamond interchange or compressed diamond interchange as the baseline.

The predictive models underlying this tool identify that the relative safety performance among interchange configurations differs as the freeway volume per lane, crossroad volume per lane, or interaction of freeway volume to ramp volume changes. This means that for a given set of conditions, the relative safety performance among interchange configurations is different than an alternate set of conditions (i.e., a single CMF would provide misleading results of relative safety performance).

In addition to traffic volumes, this tool includes crash frequency adjustment factors (like the HSM predictive method) for planning-level geometric and operational considerations, including:

• The number of through lanes on the freeway and crossroad.
• The interchange location (within an urban, suburban, or rural area).
• Closely spaced adjacent interchanges on the freeway mainline.
• The presence of one or more managed lanes on the freeway mainline.
• Skew between the freeway mainline and crossroad.
• The number of left-turn lanes present on the crossroad.
• Consistency in traffic volumes among interchange ramps.

Additionally, this tool includes adjustment factors for planning-level geometric and operational characteristics that will impact injury severity, given that a crash has occurred. Considerations impacting crash severity include:

• Traffic volume on the freeway mainline or crossroad.
• Closely spaced adjacent interchanges on the freeway mainline.
• An adjacent intersection in close proximity to the ramp terminal on the crossroad.
• The posted speed limit on the freeway mainline or crossroad.
• The number of through lanes on the freeway mainline or crossroad.
• The number of pedestrian crossings conflicting with right turns at ramp terminals.

These inputs—both for crash frequency and severity—provide analysts the flexibility to understand the impacts of geometric design decisions on interchange-level safety without needing detailed design information for each freeway mainline, ramps, and ramp terminals. Further, these planning-level models incorporate the interactions among interchange elements by considering features together rather than independently.

How to Apply the Predictive Model

The FHWA research team developed an implementation spreadsheet based on the predictive model. The implementation spreadsheet allows users to enter data for any or all applicable interchange configurations for simultaneous analysis. Users can enter the exact data for each alternative or enter specific features as needed.

The implementation spreadsheet provides the predicted property damage only, fatal and injury, total crash frequency, and a 95 percent confidence interval for each interchange configuration entered. Additionally, the implementation spreadsheet provides a graphic representation of the outputs for visual analysis. As a companion document, the Interchange Comparison Safety Tool User Guide gives practitioners details on input data requirements and examples of data elements required for applying the predictive model.

Further, analysts can use the implementation spreadsheet to compare the predicted safety performance for alternative interchange configurations, focusing on the planning-level inputs required to use the model. In addition to configuration, analysts can evaluate the impacts of interchange spacing, proximity to adjacent intersections, number of through lanes on the freeway or crossroad, presence of managed lanes, posted speed limits, and number of right turns conflicting with pedestrian crossing movements. The results of the analyses can be used along with other considerations—ones typically considered in a particular region but not explicitly considered in the tool—to evaluate alternatives early in the planning and conceptual design phases as well as provide supporting documentation for decisions made. This tool can be used directly to support interchange access proposals or to support decisionmaking on other freeway interchange projects, on or off the Interstate System.

What’s Next

The development of the Interchange Configuration Safety Comparison Tool bridged a gap, attempting to strike a balance between the need to assess the safety implications of design-level details and those details that are generally known during project concepts and preliminary engineering. Further, this tool focused on considering the interactions between individual project elements, rather than using a building-block approach; however, the safety effects of some interchange components (e.g., ramp terminal control) were difficult to isolate.

This new tool focused on the service interchange configurations most considered in access modification requests and cannot be applied to unique interchange configurations or system interchanges. Future efforts should build on the foundations of this tool to incorporate more locations (to support identifying the safety effects of ramp terminal configuration and traffic control) and to incorporate more interchange configurations considered by agencies as viable alternatives during project planning.

Service interchanges are facilities primarily designed to facilitate vehicular traffic movements, and this tool is developed based on this reality. Although design elements can be implemented at such facilities to better integrate non-motorized traffic, the design thresholds, such as roadway grades, curvatures, and design sight-distances, already adopted for decades at such facilities, imply that there are limits of what can be done to accommodate non-motorized users. At high vehicular traffic facilities like service interchanges, it is better to completely separate non-motorized traffic from motorized traffic using micro tunnels, separated bridge decks, or light-weight overpasses as conduits for moving non-motorized users.

There are two ways of providing better access and safety to non-motorized road users:

1. Mandate safe and equal access for all modes of users at all types of roadway facilities, and
2. Provide safe and equal access for all modes of users at low to moderate speed facilities but plan and design separate facilities for non-motorized users in a way to minimize direct connection with high-speed facilities.

The latter approach requires less change to the current roadway design standard and can also deliver network-wide connectivity to all modes of road users.

Currently, the concepts of Complete Streets, Safe System approach, or safe roads for all users (different ways of providing better safety for non-motorized road users) are getting more attention from top-level decisionmakers. As the Vision Zero policy gets implemented broadly and into real projects, facilities for non-motorized road users will be systemically planned, designed, and built out, and the mix between motorized and non-motorized traffic will gradually shift to a new (more stable and livable) balance. As a result, traffic crashes will also change to new patterns with stable outcomes for all modes of road users. During the transition period, the safe performance prediction models for different types of facilities should be updated periodically to reflect the new paradigm.

Wei Zhang, Ph.D., P.E., is a research highway engineer/intersection safety program manager with FHWA, specializing in innovative intersection designs and safety analysis. He has a doctorate in civil engineering from the University of Minnesota.

Scott Himes, Ph.D., P.E., is a highway safety engineer with an engineering consulting firm in Raleigh, NC, specializing in the development and implementation of HSM methods. He has a doctorate in civil engineering from Pennsylvania State University.

For more information, visit:
https://highways.dot.gov/research/publications/safety/FHWA-Interchange-Safety-Comparison-Tool, or contact Wei Zhang, (202) 493-3317, wei.zhang@dot.gov.