Crash prone 'modern roundabouts'

[kalvado]

Compared to national average traffic fatality rate of about 12 per 100,000 people, Carmel's fatality rate is at two per 100,000.   Nearby Indianapolis has a rate of over 11 per 100,000.   The difference is the roundabouts.  You should focus on the amazing improvement in the overall fatality rate and not the exception caused by an impaired speeding driver.

https://usa.streetsblog.org/2022/09/22/we-couldnt-take-a-roundabout-out-if-we-wanted-to-an-interview-with-jim-brainerd-mayor-of-carmel-ind/

https://www.smartcitiesdive.com/news/roundabouts-carmel-in-mayor-jim-brainard-uscm/571074/

Per-capita fatality rate isn't a good metric, especially when a single suburb is concerned. Looks like it isn't a good one even on the state level.
Normalize per vehicle-miles traveled as it is commonly done; take urban-suburban-rural into account; compare with similar towns in the area - now we're talking. For one, Indianapolis isn't very similar as it has quite a bit of interstate mileage with a lot of through traffic, while Carmel has only a small stretch in town (and there may be further fine print, like town services don't respond to crashes on  that stretch, and they are excluded from statistics)

For the record Carmel responds to accidents on I 465 many times each week, as part of I 465 is within Carmel's boundaries.  The US 31 freeway and Keystone Parkway are also inside Carmel's boundaries and generate their fair share of responses.

Well, still interstate crashes do not dominate Carmel statistics. Fishers seem to get a bigger share.
Overall, this map doesn't show Carmel as a black hole in terms of serious events. Looks pretty much like other suburbs for me:
https://indympo.maps.arcgis.com/apps/dashboards/83d778fc586a4a43aba848a494b1cda3

[cjw2001]

Compared to national average traffic fatality rate of about 12 per 100,000 people, Carmel's fatality rate is at two per 100,000.   Nearby Indianapolis has a rate of over 11 per 100,000.   The difference is the roundabouts.  You should focus on the amazing improvement in the overall fatality rate and not the exception caused by an impaired speeding driver.

https://usa.streetsblog.org/2022/09/22/we-couldnt-take-a-roundabout-out-if-we-wanted-to-an-interview-with-jim-brainerd-mayor-of-carmel-ind/

https://www.smartcitiesdive.com/news/roundabouts-carmel-in-mayor-jim-brainard-uscm/571074/

Per-capita fatality rate isn't a good metric, especially when a single suburb is concerned. Looks like it isn't a good one even on the state level.
Normalize per vehicle-miles traveled as it is commonly done; take urban-suburban-rural into account; compare with similar towns in the area - now we're talking. For one, Indianapolis isn't very similar as it has quite a bit of interstate mileage with a lot of through traffic, while Carmel has only a small stretch in town (and there may be further fine print, like town services don't respond to crashes on  that stretch, and they are excluded from statistics)

For the record Carmel responds to accidents on I 465 many times each week, as part of I 465 is within Carmel's boundaries.  The US 31 freeway and Keystone Parkway are also inside Carmel's boundaries and generate their fair share of responses.

Well, still interstate crashes do not dominate Carmel statistics. Fishers seem to get a bigger share.
Overall, this map doesn't show Carmel as a black hole in terms of serious events. Looks pretty much like other suburbs for me:
https://indympo.maps.arcgis.com/apps/dashboards/83d778fc586a4a43aba848a494b1cda3

Yes I 69 through Fishers is very congested and generates a lot of accidents. 

[kalvado]

Compared to national average traffic fatality rate of about 12 per 100,000 people, Carmel's fatality rate is at two per 100,000.   Nearby Indianapolis has a rate of over 11 per 100,000.   The difference is the roundabouts. You should focus on the amazing improvement in the overall fatality rate and not the exception caused by an impaired speeding driver.

https://usa.streetsblog.org/2022/09/22/we-couldnt-take-a-roundabout-out-if-we-wanted-to-an-interview-with-jim-brainerd-mayor-of-carmel-ind/

https://www.smartcitiesdive.com/news/roundabouts-carmel-in-mayor-jim-brainard-uscm/571074/

Per-capita fatality rate isn't a good metric, especially when a single suburb is concerned. Looks like it isn't a good one even on the state level.
Normalize per vehicle-miles traveled as it is commonly done; take urban-suburban-rural into account; compare with similar towns in the area - now we're talking. For one, Indianapolis isn't very similar as it has quite a bit of interstate mileage with a lot of through traffic, while Carmel has only a small stretch in town (and there may be further fine print, like town services don't respond to crashes on  that stretch, and they are excluded from statistics)

For the record Carmel responds to accidents on I 465 many times each week, as part of I 465 is within Carmel's boundaries.  The US 31 freeway and Keystone Parkway are also inside Carmel's boundaries and generate their fair share of responses.

Well, still interstate crashes do not dominate Carmel statistics. Fishers seem to get a bigger share.
Overall, this map doesn't show Carmel as a black hole in terms of serious events. Looks pretty much like other suburbs for me:
https://indympo.maps.arcgis.com/apps/dashboards/83d778fc586a4a43aba848a494b1cda3

Yes I 69 through Fishers is very congested and generates a lot of accidents.

And that directly contradicts highlighted statement above.

[tradephoric]

It's confusing when CJW asks us to focus on the overall fatality rates and not the exception caused by an impaired speeding driver... when the overall fatality rate include crashes involving impaired speeding drivers.

[Scott5114]

It's confusing when CJW asks us to focus on the overall fatality rates and not the exception caused by an impaired speeding driver.

No, it's not.

[UCFKnights]

I mean, in the most recent instance, it's not guaranteed that having the roundabout replaced with a flat cross intersection would have had a better outcome. If the guy was, for whatever reason, not interested in stopping before hitting a non-moving object, there's no guarantee he'd be interested in stopping before hitting another car in the intersection. At least if he hits a fountain or whatever he doesn't hurt anyone else.

But we don't have a "man hits other car in intersection" thread, because dog bites man is not news.

There isn't really a type of intersection that is safe for a drunk driver going 100 mph. It sounds callous, but I'd rather have that driver dying by hitting the concrete barrier in the middle of the roundabout than that driver killing an innocent person.

Roundabouts are an effective design that reduces fatal crashes by 90%, yet they are also effective at killing off drunk drivers?  That's an impressive design if they can manage to do both.

Its not absurd though. If drunk driving causes people not to notice an intersection is there, and the intersection design is such that if you see the intersection and attempt to navigate it, its much safer, but if you fail to put any input into your vehicle and keep driving straight, its much less safe as the crash is 100% guaranteed.... then thats not far from expectations. And thats what we got with roundabouts.

When I used to drive around a lot at night, I saw many people completely fail to notice intersections, including traffic lights and stop signs. When they didn't notice them, they were always completely fine as it was a low traffic time, and going straight kept you on the road. If it were a roundabout, the crashes would have been guaranteed

[SD Mapman]

Here is a chart that lists the number of fatal crashes at the most deadly intersection in each state over a 20 year period (2000-2019).



The roundabout at 96th and Westfield Blvd in Carmel doesn't have 20 full years of crash data (opened in 2005) but already has been the site of 3 fatal crashes resulting in 4 fatalities.  It would be the most deadly intersection in a number of states including Maine, New Hampshire, Vermont, Massachusetts, Rhode Island, Connecticut, Delaware, Virginia, South Dakota, Nebraska, Utah, Montana, Alaska, and Hawaii.  I think many would be surprised to hear that an innocuous looking single-lane roundabout in Carmel, Indiana (the poster-child for safe intersections) would be the site of so many fatal crashes.  Looking deeper into why people are dying here, every so often you get someone that fails to negotiate the roundabout and crashes straight into the retaining wall in the central island.  As long as that retaining wall exists, there will be more deaths at this roundabout in the future. 

So we're going to be able to test this theory in real time over the next few years; the intersection of US 77 and NE 109 on the north side of Wahoo, NE had 5 fatal crashes in a year, and NDOT is planning to put a roundabout in.

https://journalstar.com/news/local/wahoo-roundabout-delayed-two-years-but-deadly-intersection-tamed-by-temporary-changes/article_af3322e6-2444-57eb-bc18-bf8161274cbe.html

As someone who was a family friend of the three kids that died there (my wife's from the town), I hope you're wrong about big roundabout safety data, but we'll have to wait and see.

[kalvado]

Here is a chart that lists the number of fatal crashes at the most deadly intersection in each state over a 20 year period (2000-2019).



The roundabout at 96th and Westfield Blvd in Carmel doesn't have 20 full years of crash data (opened in 2005) but already has been the site of 3 fatal crashes resulting in 4 fatalities.  It would be the most deadly intersection in a number of states including Maine, New Hampshire, Vermont, Massachusetts, Rhode Island, Connecticut, Delaware, Virginia, South Dakota, Nebraska, Utah, Montana, Alaska, and Hawaii.  I think many would be surprised to hear that an innocuous looking single-lane roundabout in Carmel, Indiana (the poster-child for safe intersections) would be the site of so many fatal crashes.  Looking deeper into why people are dying here, every so often you get someone that fails to negotiate the roundabout and crashes straight into the retaining wall in the central island.  As long as that retaining wall exists, there will be more deaths at this roundabout in the future. 

So we're going to be able to test this theory in real time over the next few years; the intersection of US 77 and NE 109 on the north side of Wahoo, NE had 5 fatal crashes in a year, and NDOT is planning to put a roundabout in.

https://journalstar.com/news/local/wahoo-roundabout-delayed-two-years-but-deadly-intersection-tamed-by-temporary-changes/article_af3322e6-2444-57eb-bc18-bf8161274cbe.html

As someone who was a family friend of the three kids that died there (my wife's from the town), I hope you're wrong about big roundabout safety data, but we'll have to wait and see.

Roundabouts are not intrinsically good or bad. It's lack of engineering skill and excess of heavy handed regulation that makes them bad. It's hard to say without looking at the map, traffic data, and actual design - and even then there are surprises.

[kphoger]

There isn't really a type of intersection that is safe for a drunk driver going 100 mph. It sounds callous, but I'd rather have that driver dying by hitting the concrete barrier in the middle of the roundabout than that driver killing an innocent person.

The catch in your argument, as has been pointed out up-thread:  a drunk driver going 100 mph straight through a roundabout has a 100% of hitting the island (and a non-zero chance of hiting an inncocent person nearby), whereas that same drunk driver going 100mph straight through a red light has a less-than-100% chance of hitting another vehicle.

The counter-argument to which, of course, is that the center island of a roundabout has a 0% chance of being an innocent person.

[tradephoric]

So we're going to be able to test this theory in real time over the next few years; the intersection of US 77 and NE 109 on the north side of Wahoo, NE had 5 fatal crashes in a year, and NDOT is planning to put a roundabout in.

https://journalstar.com/news/local/wahoo-roundabout-delayed-two-years-but-deadly-intersection-tamed-by-temporary-changes/article_af3322e6-2444-57eb-bc18-bf8161274cbe.html

As someone who was a family friend of the three kids that died there (my wife's from the town), I hope you're wrong about big roundabout safety data, but we'll have to wait and see.

The conceptual roundabout design for US 77 and NE 109 is showing a simple single-lane roundabout.  These are usually not too concerning in regards to total number of crashes.  Ideally the final design doesn't call for a bunch of fixed objects in the central island of the roundabout (rock walls, statues, retaining walls, etc.) as inevitably at some point a driver along Hwy 77 (which is a high-speed arterial) will misjudge the roundabout and fly right through the middle of the thing.  If there's a bunch of fixed objects to hit in the central island that's the scenario that can lead to a fatal roundabout crash. 

[kalvado]

There isn't really a type of intersection that is safe for a drunk driver going 100 mph. It sounds callous, but I'd rather have that driver dying by hitting the concrete barrier in the middle of the roundabout than that driver killing an innocent person.

The catch in your argument, as has been pointed out up-thread:  a drunk driver going 100 mph straight through a roundabout has a 100% of hitting the island (and a non-zero chance of hiting an inncocent person nearby), whereas that same drunk driver going 100mph straight through a red light has a less-than-100% chance of hitting another vehicle.

The counter-argument to which, of course, is that the center island of a roundabout has a 0% chance of being an innocent person.

Counter counter argument is soft sand in the island may reduce crash severity...

[jamess]

If people keep crashing into a roundabout at high speeds, then it would make sense to add a speed bump or rumble strips approaching the roundabout.

That being said, a design where a drunk driver is killed but no innocent victims are harmed is a success.

[kphoger]

That being said, a design where a drunk driver is killed but no innocent victims are harmed is a success.

Again... counter-argument... a design where neither person is harmed is a bigger success.  And a stoplight intersection has a greater chance of that happening.

[Scott5114]

We should create an intersection where, if a drunk driver passes straight through, they get teleported to a pocket dimension where they are forced to read this thread in its entirety, over and over.

[kalvado]

We should create an intersection where, if a drunk driver passes straight through, they get teleported to a pocket dimension where they are forced to read this thread in its entirety, over and over.

I have seen one such spot on masspike. It is signed as "runaway truck ramp", but any vehicle with a drunk driver would get stuck until help arrives....

[jamess]

  And a stoplight intersection has a greater chance of that happening.

Citation?

[SD Mapman]

So we're going to be able to test this theory in real time over the next few years; the intersection of US 77 and NE 109 on the north side of Wahoo, NE had 5 fatal crashes in a year, and NDOT is planning to put a roundabout in.

https://journalstar.com/news/local/wahoo-roundabout-delayed-two-years-but-deadly-intersection-tamed-by-temporary-changes/article_af3322e6-2444-57eb-bc18-bf8161274cbe.html

As someone who was a family friend of the three kids that died there (my wife's from the town), I hope you're wrong about big roundabout safety data, but we'll have to wait and see.

The conceptual roundabout design for US 77 and NE 109 is showing a simple single-lane roundabout.  These are usually not too concerning in regards to total number of crashes.  Ideally the final design doesn't call for a bunch of fixed objects in the central island of the roundabout (rock walls, statues, retaining walls, etc.) as inevitably at some point a driver along Hwy 77 (which is a high-speed arterial) will misjudge the roundabout and fly right through the middle of the thing.  If there's a bunch of fixed objects to hit in the central island that's the scenario that can lead to a fatal roundabout crash. 

We'll see what they actually build (I think that was the plastic plan that never got built), the plan now is to build it as part of the 77 widening project in a couple years. If I remember right they modified it to neck you down to one lane on 77 before 109 (it's two-lane east of Wahoo), but when there's four lanes on both sides of the intersection I don't know if they'll neck it down or not. This will be the first roundabout on the Nebraska Expressway system to my knowledge, so there's going to be a high percentage of through traffic.

[kphoger]

That being said, a design where a drunk driver is killed but no innocent victims are harmed is a success.

Again... counter-argument... a design where neither person is harmed is a bigger success.  And a stoplight intersection has a greater chance of that happening.

Citation?

I thought it was obvious.

If no other vehicle is actually in front of the driver, then he simply blows through the red light at 100 mph with no incident.  And if this happens at, say, 2:30 AM, then those are pretty good odds.

If no other vehicle is actually in front of the driver, then he nevertheless plows straight into the roundabout's central island.  It doesn't matter if it's 2:30 PM or 2:30 AM–there's going to be an injury.

[jamess]

That being said, a design where a drunk driver is killed but no innocent victims are harmed is a success.

Again... counter-argument... a design where neither person is harmed is a bigger success.  And a stoplight intersection has a greater chance of that happening.

Citation?

I thought it was obvious.

If no other vehicle is actually in front of the driver, then he simply blows through the red light at 100 mph with no incident.  And if this happens at, say, 2:30 AM, then those are pretty good odds.

If no other vehicle is actually in front of the driver, then he nevertheless plows straight into the roundabout's central island.  It doesn't matter if it's 2:30 PM or 2:30 AM–there's going to be an injury.

So your citation is a ridiculous made up example?

What about the time the driver blows through the intersection but there are multiple other cars already in it?

Like this crash that killed 6 at 1:40pm

https://www.youtube.com/watch?v=xvi4-QDRFR8

[kphoger]

So your citation is a ridiculous made up example?

What about the time the driver blows through the intersection but there are multiple other cars already in it?

Like this crash that killed 6 at 1:40pm

I mean, I guess I could post a video about an incident in which nothing happened, but you know what, nobody makes news stories about that.

OK, you want a non-hypothetical, non-made-up example?  I've personally blown through a red light while crossing Michigan Avenue just north of downtown Chicago.  I had grown up in a small town, was a bit overwhelmed by city traffic, and neglected to notice the red light.  You know what happened?  Nothing.  I think a CTA bus honked at me, and that was it.  Forgive me if I can't link to a news story about it.  So yeah, I ran straight through a red light at a busy intersection in one of the nation's most crowded cities, and there were zero injuries.

[kalvado]

That being said, a design where a drunk driver is killed but no innocent victims are harmed is a success.

Again... counter-argument... a design where neither person is harmed is a bigger success.  And a stoplight intersection has a greater chance of that happening.

Citation?

I thought it was obvious.

If no other vehicle is actually in front of the driver, then he simply blows through the red light at 100 mph with no incident.  And if this happens at, say, 2:30 AM, then those are pretty good odds.

If no other vehicle is actually in front of the driver, then he nevertheless plows straight into the roundabout's central island.  It doesn't matter if it's 2:30 PM or 2:30 AM–there's going to be an injury.

So your citation is a ridiculous made up example?

What about the time the driver blows through the intersection but there are multiple other cars already in it?

Like this crash that killed 6 at 1:40pm

https://www.youtube.com/watch?v=xvi4-QDRFR8

1. "Chance" is not certainty.
2. Cars on the side of roundabout where uncontrollable vehicle comes from would still be t-boned, and thrown into a central feature

[jamess]

2. Cars on the side of roundabout where uncontrollable vehicle comes from would still be t-boned, and thrown into a central feature

One car, yes, not multiple.

I mean, I guess I could post a video about an incident in which nothing happened, but you know what, nobody makes news stories about that.

OK, you want a non-hypothetical, non-made-up example?  I've personally blown through a red light while crossing Michigan Avenue just north of downtown Chicago.  I had grown up in a small town, was a bit overwhelmed by city traffic, and neglected to notice the red light.  You know what happened?  Nothing.  I think a CTA bus honked at me, and that was it.  Forgive me if I can't link to a news story about it.  So yeah, I ran straight through a red light at a busy intersection in one of the nation's most crowded cities, and there were zero injuries.

More anecdotes.

Your statement "a stoplight intersection has a greater chance of that happening" implied facts behind it. 

[kphoger]

Your statement "a stoplight intersection has a greater chance of that happening" implied facts behind it. 

All other things being equal...

Fact:  an out-of-control vehicle heading for a roundabout's central island will hit that central island.

Fact:  an out-of-control vehicle heading for a regular intersection will not necessary hit any other vehicle.

[tradephoric]

2. Cars on the side of roundabout where uncontrollable vehicle comes from would still be t-boned, and thrown into a central feature

One car, yes, not multiple.

James knows exactly how many vehicles will get t-boned when a driver blows through a roundabout at 100 mph.  One car, yes, but never two.

[kalvado]

2. Cars on the side of roundabout where uncontrollable vehicle comes from would still be t-boned, and thrown into a central feature

One car, yes, not multiple.

SOmething like this has a great potential for being a multi-vehicle event:
https://www.youtube.com/watch?v=qB3sbNQ1Qt4
https://www.youtube.com/watch?v=aQw9HcvIo_o