Tue 18 Nov 2008
At the CNU Transportation Summit 2008, two presentations were especially consequential for the study of street networks. Both reported preliminary findings about the public safety implications of street connectivity. The results deserve attention from planners and transportation engineers.
Norman Garrick and Wesley Marshall, of the University of Connecticut’s Center for Transportation and Urban Planning, investigated the relationships between connectivity, network configuration, density, severe vehicle crashes, and mode choice. Matt Magnasco of the Charlotte (N.C.) Department of Transportation, studied the effect of connectivity on fire station service area and capital facilities planning.
In the extended entry, summary descriptions of both presentations.
Street Network, Safety and Mode Choice
Norman Garrick and Wesley Marshall, of the University of Connecticut’s Center for Transportation and Urban Planning, investigated the relationships between connectivity, network configuration, density, severe vehicle crashes, and mode choice.
Twenty-four California cities were analyzed at the block level; half were classified as “safe cites” (severe/fatal crash rates one-third of the state average), and half as “less safe cities” (severe/fatal crash rates close to the state average). The safe cities had
- an average intersection density of 106/sq mi
- a walking/biking/transit mode share of 16 percent
- a fatality rate per 100,000 people of 3.2 per year.
The less safe cities had
- an average intersection density of 63/sq mi
- a walking/biking/transit mode share of 4 percent
- a fatality rate per 100,000 people of 10.5 per year.
Interestingly, the safe cities were well established prior to 1950; the less safe cities were largely developed after that time. Even within the safe cities, the changes in street network patterns over time were related to big differences in performance. In the example of Davis, CA, the pre-1940s sections of town (intersection density 211/sq mi) had a fatal/severe crash rate that was half the rate of the post-1970 sections of town (intersection density 111-132/sq mi). The walking/biking/transit mode share was 59 percent in the pre-1940 sections of town; in the post-1980 sections of town the walking/biking/transit mode share was 14 percent.
In addition to intersection density, the researchers also investigated street network configuration — grid patterns, cul-de-sac patterns, and everything in between. The results were consistent across the board, with highly connected networks of small blocks exhibiting the best performance in all categories. Garrick’s presentation, Network, Placemaking and Sustainability, is available from the CNU library in slideshow and audio formats.
The Emergency Responder’s Perspective
The other presentation discussed the effect of connectivity on fire station service area and capital facilities planning, based on research by Matt Magnasco of the Charlotte (N.C.) Department of Transportation. The starting point of the study was the city’s standards for response time. An example was, “One fire company arrives in 6 minutes or less, 80% of time.” The standards are used to establish service areas for each fire station.
The key is this: The cost to operate a fire station generally is fixed. The size of the service area and the number of properties served per station don’t really affect operating cost. Therefore, the bigger the service area and the more properties that can be served, the more efficiently the fire department is using taxpayer money. If the fire department can serve more properties with fewer stations while meeting response time standards, it can save taxpayer money.
The study examined eight fire stations in the Charlotte area and found as street connectivity increased, the number of households served by each fire station increased as well. The least-connected service areas served 5,700 to 7,300 households; the most-connected service areas served 20,800 to 25,900 households. That means there are dramatic differences in the fiscal efficiency of individual fire stations. The stations in least-connected areas cost $586 to $740 per capita annually; the stations in most-connected areas cost $159 to $206 per capita annually.
The study also looked at the trend of fire response times over the past 38 years. As the Charlotte region developed with increasingly disconnected street patterns, average fire response times increased from 4.5 to 5.5 minutes. However, in 2001 an ordinance was passed to require connectivity in new subdivisions. Since 2001, average response times have dropped below the 5-minute mark. This has occurred even though the rate of new fire station openings has remained nearly flat.
Finally, the study showed a case study comparison of two stations. Station 15, in an area developed in the 1950s-60s, had a connectivity index of 1.3 and a service area of 13.4 sq mi. Station 31, in an area developed in the 1980s-90s, had a connectivity index of 1.09 and a service area of 8 sq mi. Because Station 31’s service area was so disconnected, an entire neighborhood that was nearby could not be included in the service area, because the travel routes were so circuitous and the response time was too long. Adding a single 300-foot connection cut a mile from the travel distance and increased the size of the service area by 17 percent.
The presentation was given by Danny Pleasant, interim director of the Charlotte Department of Transportation, and is available from the CNU library.
Both of these studies are trailblazing additions to the field of network performance and street connectivity. Their future release in peer-reviewable format is eagerly anticipated.