Assessing crash rates on the region's highway network is important to help determine underlying problems at particular locations. As one of its performance management initiatives, CMAP previously developed "scans" of weekday congestion on expressways in the Chicago region by visualizing speed at five-minute intervals throughout a typical day. Now CMAP has expanded this concept by visualizing crash frequency, producing similar "crash scans" that show the number of crashes that occurred per 100 million vehicle miles traveled (VMT) between 2008 and 2012 at points one-tenth of a mile apart along almost all of the expressways and tollways in the region. This work also broadens similar analysis done by the Illinois Department of Transportation (IDOT) for its engineering study on the Eisenhower Expressway, now allowing different expressway segments to be compared against one another quickly and intuitively.  Major findings from the analysis are that crash frequency and congestion on expressways are closely related, although the relationship is complex, and that crash rates on expressways in general are much lower than on arterials.

Crash frequency on the region's expressways and tollways

With their high traffic volumes, the Stevenson, Kennedy, Eisenhower, and Dan Ryan Expressways all have much higher crash rates than outlying expressways. At the same time, design clearly plays a role. For instance, the scans indicate that crashes tend to occur around interchanges and merge/diverge locations more generally. Returning again to the example of crashes on the Eisenhower Expressway, the lane drop between Central Avenue and Austin Boulevard and the left-hand off-ramps at Austin and IL 43 are clearly associated with higher crash rates than elsewhere on the westbound Eisenhower. Addressing these problems is an important reason why GO TO 2040 includes reconstructing and adding an express toll lane to the Eisenhower as a major capital project.

Another expressway involving a proposed GO TO 2040 major capital project is the I-55 Stevenson Expressway. Congestion and crash frequency are also closely related on the Stevenson, but at the same time several locations are uncongested yet have relatively high crash rates. Reconstruction of the road as part of the major capital project provides an opportunity to study and address any underlying geometric deficiencies that may contribute to crashes.

The link between congestion and crash rates

The relationship between congestion and crashes can also be visualized directly, as in the following figure. The travel time index on the x-axis measures the amount of time a trip takes in the peak period relative to uncongested conditions. For example, if a trip takes 20 minutes in light traffic and 30 minutes during the evening rush hour, then the travel time index is 30 ÷ 20 = 1.5.

The figure below shows the average crash rates on all expressway segments that had the same travel time index value. It suggests that a 1-unit increase in the travel time index on expressways is associated with an increase of 77 crashes per 100 million VMT.  (For reference, 100 million VMT is an amount of travel that occurs in about two days on the Chicago region expressway system.) Note, however, that this is just a correlation and that confounding factors are involved. For instance, the scans indicate that crashes tend to cluster near ramps, but expressway segments with higher congestion and higher volumes also tend to have more closely spaced ramps (that is, more ramps per expressway segment). Additional analysis may provide more insight on which factors influence crashes on expressways in the Chicago region.

The connection between crashes and congestion is an area of current research interest, with some researchers finding results similar to those shown here. For example, a recent study of expressways in five U.S. cities sponsored by the Strategic Highway Research Program found a "J-shaped" relationship in which crash rates were lowest at moderate traffic volumes and slightly higher at very low traffic volumes. As traffic densities increased and congestion ensued, there was a continuous increase in the crash rate. By contrast, CMAP's study did not examine crash rates when traffic volumes were very low. A literature review by Dutch researchers found mixed results, with one study finding that higher congestion resulted in higher crash rates but less severe crashes, another finding no relationship, and yet another finding that crash rates decrease at high traffic density. In one British study, researchers similarly hypothesized an inverse relationship between congestion and crash rates but ultimately found that congestion had little impact on accident frequency, at least on the roadway involved in the study. 

Crash frequency on expressways versus arterials

While the crash scans focus on the expressway system, expressways in general have a much lower crash rate than roadways that are not access-controlled. The difference is mostly due to the presence of at-grade intersections on arterials, which lead to more stopping and starting and cause potential conflicts between turning vehicles. Thus, projects that encourage more traffic to use the expressway system rather than arterials generally reduce crash rates (although initiatives to shift auto trips to transit, biking, or walking would have the largest safety benefit by reducing driving altogether).

Because of higher speeds, the fraction of fatal and serious injury crashes on expressways is higher than on arterials. Even then fatal and serious injury crash rates are lower on expressways than on arterials: The rate on expressways between 2008 and 2012 was 3.2 per 100 million VMT, while on arterials it was 13.3 per 100 million VMT. The latter value is an average of crash rates on principal arterials, minor arterials, and collectors. Local roads and streets are not included (they cannot be directly compared because of changes in the availability of VMT information during the 2008-12 period). Serious injuries are any incapacitating injury that prevents the injured person from walking, driving, or normally continuing the activities the person was capable of performing before the injury occurred.

The data used to compile these crash scans originated with IDOT, which makes the data available through its Safety Portal. The data are anonymized to omit all details of a crash beyond its type and location. CMAP downloaded the data in tabular form, mapped the crashes in a Geographic Information System (GIS), and then assigned the crashes to tenth-mile intervals along each expressway. The resulting dataset is available on CMAP's Data Hub. CMAP intends to update the crash scans to include more recent data at a later time.

Looking ahead

GO TO 2040 identifies public safety as a crucial goal. This includes developing a transportation system that reduces the risk of crashes. GO TO 2040 supports strategies by transportation agencies to reduce crashes through project design, enforcement, education, and other approaches. CMAP also works with partner agencies through CMAP's Advanced Technology Task Force and Regional Transportation Operations Coalition to identify, finance, and implement needed technology and highway operational improvements to reduce the number of crashes and respond to them quickly when they do occur.  In its own work, CMAP has used IDOT's identification of "5-percent locations" (locations that are the worst 5 percent on the highway network in terms of the frequency and severity of crashes) to help guide programming of highway projects.  Under new federal performance measurement requirements, CMAP will also need to coordinate with IDOT to track and set targets for reducing fatal and serious injury crashes on public roads in northeastern Illinois, and establish regional strategies for doing so.