DDOT released a study report of bicycle and pedestrian crash locations which can give some idea about where facility or enforcement changes can have the largest benefit.
The information obtained from TARAS contained a ...total of 754 locations involving bicycles crashes. Since the total numbers were a broad amount, a narrowed list was selected based on the locations with a higher crash occurrence.
This report defines the top 5% high hazard locations using the available data from crash reports. DDOT established three categories of crash statistics to define high hazard intersections: the total number of crashes, the number of injury crashes, and the total number of injuries. Intersections are ranked from most to least severe in all three categories.
As verbalized in the Annual Report, “Intersections featuring twelve or more crashes per year, four or more injury crashes per year, and five or more injuries per year are the established thresholds for identifying hazardous intersections”.
In addition the study relied on conclusions from the 2012 report on cycletracks that I reported on at the time here.
The intersections with the most injuries (though none of the fatalities) were:
- 14th and U NW
- 14th and Columbia NW
- 14th and V NW
- 16th and New Hampshire NW
The report had information on these and 10 other intersections. Of those 14, 6 are in some level of review or redesign. with 14th and U currently in construction. Three of these were included because they had only one injury crash but it was a fatality. None of these are highly used by cyclists as near as I can tell.
They show an analysis of "causes" of the injury crashes and by my count, drivers were at fault in 14, cyclists in 7* and pedestrians in 2, with the others unknown. Not that you can draw many conclusions from that, but there you go. Report is below the jump
I hate when these studies show the "most dangerous intersections for cyclists", when in fact they simply show the places with the highest absolute number of collisions. The most dangerous intersections are the ones which have the highest RATE of collisions.
For example, if 10,000 cyclists/year pass through an intersection and 100 crash (1% crash), surely this is safer than an intersection where only 500 cyclists/year pass through and 20 crash (4% crash), even though there are only 1/5 the number of total crashes.
Posted by: Jacob | February 15, 2013 at 09:53 AM
I find it problematic that the DDOT suggests putting in a sign that says "No Left Turn" at 7th & H when there are NO TURNS of any kind allowed at 7th and H, yet drivers make illegal turns there all the time despite the 8 or more signs at the intersection and on all the streets before the intersection. Lack of enforcement seems to be the biggest problem there.
Posted by: CycleGrrl | February 15, 2013 at 11:01 AM
Sorry, @Jacob, but collisions does not equal danger. Danger is serious injury and death. Death (and serious injury) is not always coorelated with collision rate. In fact, this trick has been used to show that cycle tracks are dangerous b/c there are more collisions.
What they don't tell you is that most are harmless, few send people to the hospital.
Contrast this to the roads that I am forced to ride on, daily, where a single collision has an 80% chance of death.
Speed limit is the only factor that coorelates highly with injury and death for cyclists not helmets, not education, not motorist awareness campaigns, and not bright colors. We see deaths of all the above all the time, but rarely in places where motorists drive 15 MPH or less.
I have had a few minor accidents, and I believe that they actually make you safer as pain is a great teacher. If you never suffer any kind of minor collision, you start to believe that you have some kind of "skill" that is keeping you safe while it's really your environmental circumstances.
Posted by: Fred | February 15, 2013 at 11:27 AM
Just going to copy/paste what I said on DCist- An official traffic engineering report with no mention of collision rates is a useless document.
Fred also raises a good point about discarding crash severity, which is also typically included in such metrics.
Posted by: Bossi | February 15, 2013 at 11:39 AM
Jacob, I don't think they're saying that these are the most dangerous intersections, just the crashiest, which for DDOT is relevant.
Imaagine an intersection (A) that is relatively safe, but has high traffic so that there is 1000 crashes per year.
Then, another intersection that is twice as dangerous with 1/10th the traffic so that there are 200 crashes per year.
Fixing either intersection might be roughly the same cost, but a 50% drop in crashes at the A is likely to result in a better return than at B.
But I agree that crashiest is not most dangerous. And that we need to weight the crashes, as well as the cost of improvements to determine where to target our resources.
But how many resources do we want to spend on targeting. Sometimes brute force is the cheapest path [I tell myself all the time when coding]. So for DDOT to say "These intersections have the most crashes and we're going to focus on them" is not an unreasonable path forward.
Posted by: washcycle | February 15, 2013 at 12:29 PM