Quantifying Safety of Design Alternatives

Design Alternatives refer to the choices made by planners and engineers about what transportation infrastructure to provide. Options can be related to cross-section design, vertical curvature, roadside design, traffic control and other measures (e.g., provision of warning signs).

Common examples of design alternatives:

• Intersection traffic control: should an intersection be signalized, stop-controlled or a roundabout?
• Cross-section: should a roadway have a raised median, and if so, what should the median width be?
• FacilitiesInfrastructure provided for road user movements, e.g. roads, bicycle lanes, sidewalks: should a roadway have a separated bike lane and narrower vehicle lanes?

Consideration of Design Alternatives is relevant for new infrastructure or existing infrastructure that will be undergoing changes.

Back to top

The safety of design alternatives refers to a quantitative measure of the long-term expected crash frequency. Safety is ideally measured by crash type and crash severity, considering that crashes of different types are more or less severe, and that reductions in fatal and serious injury crashes should be prioritized.

It should be noted that there is no ‘safe’ or ‘unsafe’ design. All infrastructure carries a level of risk.

The relative safety of design alternatives can be compared by assigning a societal cost of crashes to the estimated long-term expected crash frequencies. If smaller jurisdictions do not have their own cost of crash estimates to use, they can adopt those used at the federal or provincial/state level of government.

Assigning crash costs also allows for safety to be quantitatively evaluated alongside other transportation performance measures such as traffic operations, environmental impacts, and construction costs. This provides a rational basis for making and documenting decisions.

Back to top

The number of expected crashes to occur at a location can be estimated through Safety Performance Functions (SPFs) and Crash Modification Factors (CMFs). For locations with an observed crash history, these data can also be used to refine the estimates.

An SPF is a mathematical model that predicts the mean crash frequency for similar locations with the same traffic volume. The following equations are examples of SPFs for road segments and intersections.

For road segments:
Collisions per year = α(segment length)(AADT)β

For intersections:
Collisions per year = α(Major road entering AADT) β1(Minor road entering AADT)β2

Where the AADTs are traffic volumes and α, β, β1 and β2 are numbers estimated during the SPF development.

A CMF is a multiplicative factor that adjusts the SPF estimate when the location under consideration differs from the locations used to estimate the SPF. For example, the SPF may be based on straight segments of roadway and a CMF applied to account for horizontal curvature. A CMF greater than 1.0 indicates an expected increase in crashes, while a value less than 1.0 indicates an expected reduction in crashes.

When crash data are available, these can be combined with the SPF and CMFs to better estimate the long-term expected number of crashes. This combining of information is superior because crash counts alone are not reliable due to the randomness of crashes and the SPFs/CMFs represent average conditions, rather than for a specific location.

Back to top

Evaluating the safety effects of design alternatives requires police-reported and/or self-reported crash data and vehicle traffic volume data as a minimum. Other required data will depend on what factors are required by the SPFs and CMFs being applied. Examples of possible data requirements include:

• Roadway data
  • Number of lanes
  • Median presence and type
  • Shoulder type and width
  • Lane width
• Traffic control data
  • Control type (i.e., signalized, two-way stop-controlled)
  • Signal phasing
  • Presence of signage, pavement markings
• Exposure data
  • Average annual daily traffic
  • Commercial vehicle volumes
  • Bicycle volumes
  • Pedestrian volumes
• Crash prediction models (Safety Performance Functions) relevant to the site and design
• alternatives being considered
• βCrash Modification Factors (CMFs) relevant to the site and design alternatives being considered

Back to top

Evaluating the safety effects of design alternatives requires persons with a general knowledge and understanding of roadway design, traffic control, crash data as well as an introductory level of statistics.

Back to top

There are several sources of information and tools related to methods for quantifying the safety effects of design alternatives.

1. FHWA Safety Data and Analysis YouTube Channel
This YouTube channel provides five videos related to applying the predictive method for crashes included in AASHTO’s Highway Safety Manual (HSM). These include: 1) The Predictive Method; 2) Observed, Predicted and Expected Crashes; 3) Application of CMFs; 4) Selecting a Method to Analyze Multiple CMFs; 5) Applying a Method to Analyze Multiple CMFs.

2. An Introduction to the Highway Safety Manual
This report provides an overview of AASHTO’s Highway Safety Manual (HSM), including the Part C Predictive Method that is used for quantifying the safety of design alternatives with an example.

3. Overview of Interactive Highway Safety Design Model (IHSDM)
IHSDM is a suite of software analysis tools used to evaluate the safety and operational effects of geometric design decisions on highways. The IHSDM—which supports the Federal Highway Administration’s (FHWA’s) Data-Driven Safety Analysis (DDSA) initiative—includes five evaluation modules (i.e., Crash Prediction, Design Consistency, Policy Review, Traffic Analysis, and Driver/Vehicle), as well as an Economic Analyses Tool. This website provides a link to download the software, summarizes the capabilities and applications of the IHSDM evaluation modules, and provides a library of the research reports documenting their development.

4. Guidelines for the Development and Application of Crash Modification Factors (CMFs)
This report provides guidance to practitioners for situations when applying a CMF is not straightforward. These include 1) applying CMFs when site characteristics may not match sites used to develop the CMF, and 2) applying multiple CMFs at a single location.

Back to top

Back to top