Over 7000 roundabouts in America

[tradephoric]

Oh, so despite my very routine trips up and down those roundabouts, you think all the crashes I saw are just happening at the same rate at different times of the day?  That is less powerful, actually.  At least I have my observations.  You don't even have that in your rebuttal regarding NY 85.

For all i know the response time at clearing vehicles off the roadway after a roundabout crash has improved significantly, reducing the likelihood of you observing the aftermath of a crash.  After all they have plenty of experience dealing with crashes at the roundabout.  I've personally witnessed a vehicle flipped over at one of the crash prone roundabouts in Michigan, with EMS tending to the driver as he was bleeding profusely from his arm.  That's my personal anecdotal experience.  But me observing one injury accident doesn't tell me much about how the roundabout performs overall over the course of an entire year.  That's why years worth of crash data is useful for analysis. 

I also suspect that since you were able to find crash statistics on the NY 85/NY 140 shortly after the roundabout opened and you touted it's high crash rate that your refusal to look into that particular data again may be evidence of data cherry picking to support your crusade.

You are the one making the argument that i only focus on the accident-prone transition period (ie. the initial spike in accidents after a roundabout is first completed) and ignore when accidents drop off.  Yet I've cited the 18 1/2 Mile & Van Dyke roundabout several times in the past for its high crash rate... and that roundabout is nearly 15 years old.  The transitional period is long over. 

If there was a huge drop off in crashes at the NY 85/NY 140 roundabout then great.  All i know is i have yet to find a 2x2 roundabout with a crash rate below 1.0 MEV whether it's 2 years old or 20.  A crash rate of 1.0 MEV is higher than the average crash rate of signalized intersections... so you would think there would at least be a few multi-lane roundabouts with crash rates below that.  There are now literally hundreds of 2x2 or 2x3 roundabouts in America.  Just cite one with a crash rate below 1.0 MEV (and i don't care about how old it is).  A 2x2 roundabout with a low crash rate is like the mythical unicorn or bigfoot... maybe it exists.

[tradephoric]

^Think of it this way.  A crash rate of 1 MEV equates to 1 accident per million entering vehicles.  Just watch 20 cars go through a complex roundabout and you'll probably already witness a near miss.  You really think a million vehicles are going to navigate the roundabout without crashing into each other at least once?   There are too many ways for drivers to screw up at these complex roundabouts.  I'm not hiding anything from you.

[BrianP]

Just cite one with a crash rate below 1.0 MEV (and i don't care about how old it is). 

Does this one count?
MD450@MD435/MD387
https://www.google.com/maps/place/MD-435,+Annapolis,+MD/@38.976196,-76.5051311,114m/data=!3m1!1e3!4m5!3m4!1s0x89b7f6561b3a54a5:0x10eaff58740622a3!8m2!3d38.9831055!4d-76.5054642

From figure 3: ( I think this column has the value you're referring to, which is accident rate / MVE)
before: 1.52
after: 0.66
https://safety.fhwa.dot.gov/intersection/other_topics/fhwasa09027/resources/Marylands%20Roundabouts%20Accident%20Experience%20and%20Economic%20Evaluation.pdf
I wish there was data for more recent years.

I think taking severity of a crash into account should be looked at more than just a raw accident count.    Having more fender benders is worth it when significantly reducing the number of severe crashes.  I think that's the trade off with roundabouts which is worth it.  The worst crashes that I see are usually the left turns at non-roundabouts.

That said I think Maryland agrees with you about larger roundabouts.  There are no 3x roundabouts that I've found.  And even 2x are rare.  They mostly do 1x or half 2x.  I'd like to see data for this roundabout, which is the most 2x like roundabout I know of in MD:
https://www.google.com/maps/@39.2517845,-76.7143174,159m/data=!3m1!1e3
Even that may not meet a strict interpretation of 2x. Since the inner lane of the roundabout never 'exits'.  It doesn't 'exit' since you have to cross a dotted line to exit from the inner lane.  That makes it seem like they prefer that you change to the outer lane before exiting to avoid the problem covered above. 

[kphoger]

Looks like a 2x2 roundabout to me.  Every approach is two lanes, and the entire circulating roadway is two lanes wide (exact striping varies around the circle but true nonetheless).

[english si]

Actual roundabouts, or just little bitty islands in the middle of the intersections?

While travelling in Ireland last year there were roundabouts everywhere, as you would expect, but there was one intersection that was so small there was simply a 2 foot diameter circle painted in the middle of an otherwise normal looking American intersection.  I still drove around it though.

It's known as a mini-roundabout, and they're pretty common in the UK.  I think the principle is "yield to whoever is in the intersection before proceeding," since there's really no way to traverse the intersection in a circular fashion.

In the UK, it's 'give way to right', the same as big roundabouts (Ireland is, I believe, the same - though they do love to be different just to be different from the UK - eg diamond warning signs). And you are meant to drive around them, but you can drive over them.

[abefroman329]

In the UK, it's 'give way to right'

That's what I thought.

Apparently in France, traffic in the roundabout yields to traffic entering the roundabout, which makes no sense at all.

[kphoger]

Apparently in France, traffic in the roundabout yields to traffic entering the roundabout, which makes no sense at all.

I believe that depends.  At older traffic circles in urban areas, traffic in the circle yields to entering traffic.  However, France has a zillion newer roundabouts where entering traffic yields to traffic in the circle.  Basically, it's similar to the USA:  it all depends on the location.

[jakeroot]

Just cite one with a crash rate below 1.0 MEV (and i don't care about how old it is). 

Does this one count?
MD450@MD435/MD387
https://www.google.com/maps/place/MD-435,+Annapolis,+MD/@38.976196,-76.5051311,114m/data=!3m1!1e3!4m5!3m4!1s0x89b7f6561b3a54a5:0x10eaff58740622a3!8m2!3d38.9831055!4d-76.5054642

From figure 3: ( I think this column has the value you're referring to, which is accident rate / MVE)
before: 1.52
after: 0.66
https://safety.fhwa.dot.gov/intersection/other_topics/fhwasa09027/resources/Marylands%20Roundabouts%20Accident%20Experience%20and%20Economic%20Evaluation.pdf
I wish there was data for more recent years.

I think taking severity of a crash into account should be looked at more than just a raw accident count.    Having more fender benders is worth it when significantly reducing the number of severe crashes.  I think that's the trade off with roundabouts which is worth it.  The worst crashes that I see are usually the left turns at non-roundabouts.

That said I think Maryland agrees with you about larger roundabouts.  There are no 3x roundabouts that I've found.  And even 2x are rare.  They mostly do 1x or half 2x.  I'd like to see data for this roundabout, which is the most 2x like roundabout I know of in MD:
https://www.google.com/maps/@39.2517845,-76.7143174,159m/data=!3m1!1e3
Even that may not meet a strict interpretation of 2x. Since the inner lane of the roundabout never 'exits'.  It doesn't 'exit' since you have to cross a dotted line to exit from the inner lane.  That makes it seem like they prefer that you change to the outer lane before exiting to avoid the problem covered above. 

Looks like a 2x2 roundabout to me.  Every approach is two lanes, and the entire circulating roadway is two lanes wide (exact striping varies around the circle but true nonetheless).

The West/Spa/Taylor roundabout in Annapolis is a 2x1 roundabout. The West St approaches are two lanes, and both may proceed through. But, the Taylor and Spa approaches are, while two lane, only single lane through movements (both right lanes on those approaches are right-only).

In tradephoric's language, this does not make it a complex multi-lane roundabout.

The MD-45 Towson oval roundabout is also on that list, and crashes went up quite a bit there. Thing is, the study's from 2004, before it was modified. But even before, I don't think it was a 2x2 roundabout (only ever 2x1, if that).

[kphoger]

But, the Taylor and Spa approaches are, while two lane, only single lane through movements (both right lanes on those approaches are right-only).

Yep, I missed that.

[jakeroot]

But, the Taylor and Spa approaches are, while two lane, only single lane through movements (both right lanes on those approaches are right-only).

Yep, I missed that.

Though it is interesting how many roundabouts in Maryland were apparently built to accommodate two lanes around the entirety of the circle, but are mostly fed only by one lane on two of the approaches, and two on the other two.

[tradephoric]

Just cite one with a crash rate below 1.0 MEV (and i don't care about how old it is). 

Does this one count?
MD450@MD435/MD387
https://www.google.com/maps/place/MD-435,+Annapolis,+MD/@38.976196,-76.5051311,114m/data=!3m1!1e3!4m5!3m4!1s0x89b7f6561b3a54a5:0x10eaff58740622a3!8m2!3d38.9831055!4d-76.5054642

From figure 3: ( I think this column has the value you're referring to, which is accident rate / MVE)
before: 1.52
after: 0.66
https://safety.fhwa.dot.gov/intersection/other_topics/fhwasa09027/resources/Marylands%20Roundabouts%20Accident%20Experience%20and%20Economic%20Evaluation.pdf
I wish there was data for more recent years.

Traffic in the right most lane is striped to exit onto West St (so technically only one lane of traffic circulates around the circle along that leg).  But really the only thing preventing it from being a full 2x2 roundabout is the pavement markings.  The real point i wanted to make regarding that circle is the upstream geometry.  The roadway only widens out to two lanes in each direction just before the roundabout.  Otherwise drivers are driving along on a two-way street signed for 25 mph.  There is another multi-lane roundabout in Figure 3 that supposedly has a 0.0 crash rate but i can't even find it.... where is MD139@ Bellona Ave?

[english si]

Though it is interesting how many roundabouts in Maryland were apparently built to accommodate two lanes around the entirety of the circle, but are mostly fed only by one lane on two of the approaches, and two on the other two.

In Britain we have a lot of roundabouts that can accommodate two lanes with all the approaches being one lane until just before the roundabout.

[kphoger]

There is another multi-lane roundabout in Figure 3 that supposedly has a 0.0 crash rate but i can't even find it.... where is MD139@ Bellona Ave?

http://roundabout.kittelson.com/Roundabouts/Details/82
Status = removed

[froggie]

139 at Bellona was never a full 2x2 roundabout.  Westbound-to-southbound was the only part of the roundabout that had 2 lanes.  It was removed ca. 2011-12 when the ramp to 695 was separated from the leg of Bellona west of Charles St.

[mrsman]

139 at Bellona was never a full 2x2 roundabout.  Westbound-to-southbound was the only part of the roundabout that had 2 lanes.  It was removed ca. 2011-12 when the ramp to 695 was separated from the leg of Bellona west of Charles St.

Here's a picture:

https://www.google.com/maps/@39.4180381,-76.629021,3a,75y,217.34h,73.94t/data=!3m7!1e1!3m5!1swkk5EobPLecXSnMi5xUhhg!2e0!6s%2F%2Fgeo1.ggpht.com%2Fcbk%3Fpanoid%3Dwkk5EobPLecXSnMi5xUhhg%26output%3Dthumbnail%26cb_client%3Dmaps_sv.tactile.gps%26thumb%3D2%26w%3D203%26h%3D100%26yaw%3D208.5173%26pitch%3D0%26thumbfov%3D100!7i13312!8i6656


What seems extremely complicated is that there seems to be 3 (or 4) places near Baltimore where Ballona and Charles intersect.

But I think what they did to this intersection is a big improvement.  The roundabout was not helpful.

[BrianP]

But, the Taylor and Spa approaches are, while two lane, only single lane through movements (both right lanes on those approaches are right-only).

Yep, I missed that.

Thanks.  Usually when it's a 2x1 it's easier to tell from the ground markings like this one:
https://www.google.com/maps/@39.1456596,-76.9039883,91m/data=!3m1!1e3

[jakeroot]

But, the Taylor and Spa approaches are, while two lane, only single lane through movements (both right lanes on those approaches are right-only).

Yep, I missed that.

Thanks.  Usually when it's a 2x1 it's easier to tell from the ground markings like this one:
https://www.google.com/maps/@39.1456596,-76.9039883,91m/data=!3m1!1e3

Haha I spent an inordinate amount of time studying that interchange last week, trying to see if any of the roundabouts operated in 2x2 configuration. As far as I can tell, though they used to lack striping, none of the roundabouts through there ever had four 2-lane entrances.

[tradephoric]

Saw the title of the thread changed numbers, let me know when it's over 9,000. 

Depending how strict of a definition you want to use for "modern roundabout", i bet you we are close to 9,000 today.  Just think of how many little subdivision roundabouts like this haven't been documented:

https://www.google.com/maps/@33.1405037,-96.8720605,110m/data=!3m1!1e3

[D-Dey65]

Today, I was looking up sites in Johnston County, North Carolina, because I made two Wikimedia commons categories for towns within that county, and I wanted to see how many more images I could add to those categories, or at least somewhere else. What does that have to do with roundabouts, you ask? Well one of those sites is the Bentonville Battlefield and the nearby Harper House, and the nearest major highway to that is US 701. That site is north of Newton Grove, North Carolina, which has an otherwise six-way intersection with US 701, US 13, and NC 50 and 55, all of which are obstructed by a traffic circle/roundabout.

I don't know when that was built, but if there's any example of one that shouldn't be torn down, it's probably that one.

[mgk920]

Are there any freeway roundabouts? I could see this concept work in rural areas with lots of room for a wide radius to keep speed. Would save on overpasses and flyovers.

This triple-lane roundabout in Wisconsin replaced a trumpet interchange about 10 years ago:

https://www.google.com/maps/@44.9210291,-91.3805333,156m/data=!3m1!1e3

That trumpt had a *nasty* reducing radius curve on the loop ramp from WB WI 29 to SB WI 124, too - a perfect textbook example of what not to do when designing an interchange.  When it was rebuilt I said "Good riddance!" on that interchange - that roundabout was the ideal solution there when the WI 29 east-west Chippewa Falls/Eau Claire bypass freeway opened in the early-mid 00s.

Mike

[mgk920]

I guess this would kinda count as a freeway roundabout?

https://www.google.com/maps/@46.0974443,-64.7697549,15.72z

Howabout I-87/587 in New York?

Mike