U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
Agencies are confronted with two fundamental types of risk when dealing with design exceptions. The first involves the risk of the solution not performing as expected. The second involves the risk concerning the agency’s ability to defend itself against potential legal actions as a result of its decisions. Most States incur some risk of tort lawsuits arising from crashes alleged to be associated with a design or other problem created by the agency. Design exceptions in particular may represent a potential future risk to the agency if not handled properly.
When designing highways in areas with difficult site constraints, designers should first acknowledge that the inability to meet design criteria may increase the risk of safety and/or operational problems. The degree of risk of these problems should be evaluated before moving forward with a design exception. This is primarily a technical process involving knowledge and tools (such as the IHSDM) that help designers understand operational and safety implications of varying design conditions. The questions below are fundamental to this evaluation and should be looked at in combination because one can have an effect on others and the level of risk as a whole.
Exposure to traffic is one of the most critical factors in measuring the safety risk of any highway element or feature. The more traffic to which the location is exposed, the greater the risk of a crash and/or measurable traffic operational problems. A designer may reasonably accept a design exception for curvature on a two-lane rural highway with low traffic, but be less inclined to do so in a geometrically or physically comparable context with significantly higher volumes. The composition of traffic is also a consideration. For example, there will be a higher level of risk for narrowed lane widths on a highway with a high percentage of large trucks than a highway that carries predominantly passenger vehicles.
The speed or anticipated speed (for proposed designs) is another factor that influences risk. Particularly in terms of substantive safety, the probability of severe crashes will increase as speeds increase.
How much a proposed design exception deviates from the design criteria is one measure for evaluating risk. The probability of safety or operational problems developing may increase as the deviation from design criteria increases. For example, the ability to provide 450 feet of stopping sight distance when 500 feet is specified may be acceptable, but providing only 250 feet may not be. Designers should be able to translate variable dimensions to meaningful operational or substantive safety measures to help make these judgments.
Another factor that influences risk is the presence of two or more design exceptions at a particular location interacting with each other. There is research to support the view that the presence of multiple geometric problems represents particular risk to drivers. For example, one might expect that the risk associated with a horizontal curve that does not meet criteria for curvature and superelevation will increase if horizontal stopping sight distance is also less than the minimum value. Other combinations of design exceptions may function independently and have no effect on each other—for example, vertical clearance and horizontal alignment. The nature of the design elements involved influences whether there is an interaction or cumulative effect that might increase risk.
The length of highway affected by the design exception influences the degree of risk. Length is another fundamental measure of exposure. The extent of this influence depends on many factors, including the magnitude of variance of the design exception.
Design exceptions may occur at just a point location or for a very limited length—for example, a short bridge that does not meet bridge width criteria. Another example would be stopping sight distance at a curve. In these instances, the section of roadway affected by the design exception is relatively limited and so the designer may expect the operational or safety risk to be somewhat limited.
In other cases, a design exception may extend for several miles. An example would be an area with constrained cross-sectional width where narrower lane and/or shoulder widths are used over an extended segment of the highway. Designers should recognize that the presence of a significant design exception over an extended length of highway greatly increases the risk of operational or safety problems to drivers exposed to it.
Is the design exception expected to be permanent? Or is there a reasonable expectation that other planned improvements in the near future may remove or lessen the non-standard condition?
Another important consideration is other highway elements (not necessarily design exceptions) that may have an interaction with the design element being evaluated. A good example of this is a crest vertical curve where there are intersections within the curve or just beyond the crest. The safety risk of non-standard stopping sight distance is greater at such a location compared to a curve where there are no intersections present.
Knowledge of the past safety performance at the location is essential for evaluating risk. Both the crash history and the types of crashes will be needed for this evaluation because the crash types of primary interest would be those with a possible relationship to the design element that does not meet criteria. In addition, the designer needs full knowledge of the expected substantive safety performance of this location. Designers should not expect or promise zero crashes. What they need to understand is how well (or poorly) a location appears to perform compared with others similar to it.
There are tools, methods and published studies that enable formal evaluation of the substantive safety of a condition or location. Designers need to incorporate the use of this knowledge base in their risk evaluations. A location exhibiting acceptable, long-term safety performance relative to expectation, despite having design features that do not meet current criteria, may indicate a lower level of risk. Conversely, designers should resist employing a design exception at a location that is fully in compliance with design criteria but known to be a high crash location.
Care should be taken in relying on historical crash data for locations where significant changes are expected. For example, significant changes in land use and traffic or nearby geometric changes to intersections and interchanges may change how the location functions in the future. So, a high level of substantive safety based on crash data alone does not necessarily mean that the design element should be maintained in its existing condition. The safety performance of the existing roadway may change, particularly if other conditions change. For these situations, the models for predicting expected safety performance are particularly valuable.