AARoads Forum

Please login or register.

Login with username, password and session length
Advanced search  

News:

The AARoads Wiki is live! Come check it out!

Author Topic: Crash prone 'modern roundabouts'  (Read 640482 times)

tradephoric

  • *
  • Offline Offline

  • Posts: 2435
  • Last Login: March 18, 2024, 11:18:58 AM
Re: Crash prone 'modern roundabouts'
« Reply #50 on: May 22, 2015, 01:27:05 PM »

(297%) Canal & 25th Street (Milwaulkee, Wicsonsin):
https://www.google.com/maps/@43.0311071,-87.9434376,92m/data=!3m1!1e3

I have to wonder if the number of collisions at this intersection where one or more operator was intoxicated is significantly higher.  It's an exit route for fans leaving Brewer games and for patrons dumping off money at the Potowatomi Casino.  Two situations where drinking is encouraged.
It is said that drunks are more prone to side-swipe collisions and roundabouts are purposefully designed so if there is a crash, it will be low-angle and hopefully non-injurious.

Maybe, maybe not.  Livernois & Hamlin looks pretty innocuous yet that multi-lane roundabout saw over a 300% increase in total crashes. 
Logged

jakeroot

  • *
  • Offline Offline

  • Posts: 15874
  • 日本標準時

  • Age: 28
  • Location: Uruma-shi, Japan
  • Last Login: Today at 04:53:27 AM
    • Flickr
Re: Crash prone 'modern roundabouts'
« Reply #51 on: May 22, 2015, 02:50:18 PM »

Another crash prone roundabout is at 14th Street & Superior in Lincoln, Nebraska.  This triple-lane roundabout was completed in November 2012.  In the first 11 months of operations, crashes quadrupled from 27.3 to 119.8.  In October 2013, the roundabout was converted to a two-lane roundabout.  Since the conversion, annualized crashes have dropped to 40.4 (which is still 47% higher than it was before the roundabout was constructed).

They only gave it 11 months? Jesus. So immediate top-tier performance or bust? I think crashes would drop over time, no need to act so hastily.

I think roundabouts with three entry lanes are growing out of favor.  Roundabouts with three entry lanes were constructed as part of the US41 project in Wisconsin.  From my understanding, Wisconsin has no plans to design new roundabouts with three entry lanes moving forward.  DaBigE might have more information regarding this based on some posts I've read of his.

Here are some videos of crashes that occurred at Maple & Drake (before the roundabout was converted to two entry lanes).  This roundabout had three entry lanes for several years before it was converted to two entry lanes: 

I'm fine with reducing the amount of lanes at roundabouts, but only if it's related to "too much capacity" instead of uneducated drivers (Bloomington Hills for example). Basic math suggests that more collisions occur at three lane roundabouts because three lane roundabouts can process more cars.

Quillz

  • *
  • Offline Offline

  • Posts: 3315
  • A Talking Pokémon

  • Age: 35
  • Location: Los Angeles, CA
  • Last Login: March 15, 2024, 10:13:27 PM
Re: Crash prone 'modern roundabouts'
« Reply #52 on: May 22, 2015, 03:04:24 PM »

Speaks more to the (in)flexibility of Michigan drivers than it does any inherent safety risk with roundabouts.  New roundabouts elsewhere have seen major drops in the number of crashes.
Yup, I'm actually writing a paper on roundabouts right now and this has largely been my conclusion, too. Driver behavior (i.e. being unfamiliar with how roundabouts work) is the main cause of accidents within a roundabout. The design in of itself has been consistently shown to be safer than most traditional signalized or stop controlled intersections.
Logged

6a

  • *
  • Offline Offline

  • Posts: 965
  • Age: 48
  • Location: Hilliard, Ohio
  • Last Login: March 13, 2024, 02:33:17 AM
Crash prone 'modern roundabouts'
« Reply #53 on: May 22, 2015, 05:44:28 PM »

That's exactly what's being argued though.  Listen to what Jeannie Willis has to say about the expected safety benefits of a 3-lane roundabout currently under construction in Dublin, Ohio (6A touched on this roundabout previously in this thread).  What are the residents of Dublin to think if her predictions don't come to pass?  At that point, who cares right?  It will already be constructed.

Well, I can't speak for the specific roundabout being built; it's entirely possible that the existing intersection is so dangerous that it will reduce both the absolute frequency and severity of crashes (the data suggest some roundabouts do lead to less crashes than the intersections they replaced, and without systematically distinguishing between what the previous intersection type was it's hard to predict which will reduce crashes and which will lead to more crashes).

Assuming this is the intersection in question, given the weird angle of the existing intersection I suspect the multilane roundabout will be an improvement on both scores.

That is indeed the intersection. It's hard to tell from that view obviously, but SB traffic goes up a somewhat sharp incline immediately before the light. I've mentioned that left turns are banned on the E-W route, and even with that and the SB bypass road it's the third worst crash intersection in the city. That's why I said...


Uh... Yes. 

Lordsutch said "I don't think anyone ever argued that roundabouts reduce collisions overall."  Jeannie Willis, the engineering manager for the City of Dublin, argued that "the frequency of the crashes will be reduced".


Here are some videos of crashes that occurred at Maple & Drake (before the roundabout was converted to two entry lanes).  This roundabout had three entry lanes for several years before it was converted to two entry lanes.

...I really do wonder if that will be the case in this instance. Call me an eternal optimist on this one. However, with the area already being quite well educated in roundabouts and I can only assume the quirky hill at the intersection being addressed, can it really get worse than #3 in the city? (As I type this my curiosity is piqued as to what the others are and how that one relates. Off to the bowels of the Internet...)

In those videos it appears the problem in that layout is three *through* lanes. The Dublin one is only three lanes in a quarter of the design (the SB bypass is being retained.) but only two of those are through lanes. The innermost lane will be restricted to left-turning traffic. If the mockup I posted is correct as to striping, it would appear the problem of a driver in the inner lane trying to go straight is addressed. Of course we all know lines don't keep cars in their place, as Mr. FedEx learned in your last video.

So, time will tell I suppose. I do admit to a raised eyebrow when I first saw the design and, if anything, it should have enough traffic (and our eyes) from the beginning to judge whether this design is genius or garbage.

Edit: here is the current view SB (speed limit 45)


And NB. Note each direction gets an exclusive green phase.
« Last Edit: May 22, 2015, 05:52:21 PM by 6a »
Logged

tradephoric

  • *
  • Offline Offline

  • Posts: 2435
  • Last Login: March 18, 2024, 11:18:58 AM
Re: Crash prone 'modern roundabouts'
« Reply #54 on: May 22, 2015, 07:19:12 PM »

I'm fine with reducing the amount of lanes at roundabouts, but only if it's related to "too much capacity" instead of uneducated drivers (Bloomington Hills for example). Basic math suggests that more collisions occur at three lane roundabouts because three lane roundabouts can process more cars.

Traffic volumes at 14th Street and Superior is down from the 2008 peak.  The 4 fold increase in crashes at this roundabout was not because more vehicles were navigating through the intersection.  Similarly, traffic volumes didn't drop overnight when the roundabout was converted from three lanes to two.
Logged

jakeroot

  • *
  • Offline Offline

  • Posts: 15874
  • 日本標準時

  • Age: 28
  • Location: Uruma-shi, Japan
  • Last Login: Today at 04:53:27 AM
    • Flickr
Re: Crash prone 'modern roundabouts'
« Reply #55 on: May 22, 2015, 09:42:46 PM »

I'm fine with reducing the amount of lanes at roundabouts, but only if it's related to "too much capacity" instead of uneducated drivers (Bloomington Hills for example). Basic math suggests that more collisions occur at three lane roundabouts because three lane roundabouts can process more cars.

Traffic volumes at 14th Street and Superior is down from the 2008 peak.  The 4 fold increase in crashes at this roundabout was not because more vehicles were navigating through the intersection.  Similarly, traffic volumes didn't drop overnight when the roundabout was converted from three lanes to two.

It's not pure volume. It's throughput. The intersection is pushing a whole bunch more cars through every second compared to a smaller roundabout. There's basically a higher likelihood of a collision because there's more cars that can collide. You can avoid this by making the roundabout smaller, but the intersection capacity is likewise reduced.

tradephoric

  • *
  • Offline Offline

  • Posts: 2435
  • Last Login: March 18, 2024, 11:18:58 AM
Re: Crash prone 'modern roundabouts'
« Reply #56 on: May 22, 2015, 10:33:01 PM »

It's not pure volume. It's throughput. The intersection is pushing a whole bunch more cars through every second compared to a smaller roundabout. There's basically a higher likelihood of a collision because there's more cars that can collide. You can avoid this by making the roundabout smaller, but the intersection capacity is likewise reduced.

I got what you're saying now Jake.  You make a good point.
Logged

colinstu

  • *
  • Offline Offline

  • Posts: 397
  • Last Login: August 10, 2023, 12:09:09 PM
Re: Crash prone 'modern roundabouts'
« Reply #57 on: May 23, 2015, 10:02:50 AM »

(297%) Canal & 25th Street (Milwaulkee, Wicsonsin):
https://www.google.com/maps/@43.0311071,-87.9434376,92m/data=!3m1!1e3

I have to wonder if the number of collisions at this intersection where one or more operator was intoxicated is significantly higher.  It's an exit route for fans leaving Brewer games and for patrons dumping off money at the Potowatomi Casino.  Two situations where drinking is encouraged.
It is said that drunks are more prone to side-swipe collisions and roundabouts are purposefully designed so if there is a crash, it will be low-angle and hopefully non-injurious.

Umm... guys? I don't know where that 297% number is coming from but check out historical aerial imagery in Google Earth.

An intersection didn't even EXIST in that location before the roundabout, it was a a curved thru-road with a dirt road going over train tracks coming off the middle of it for the heavy industry that was present in the valley.
Canal St didn't even exist between the Brewers Parking lots and 25th St.

It was that way up until 2005 when Canal St between those two locations began as well as the roundabout taking shape. It wasn't completed until 2006, and more people probably didn't figure out that completed road existed until even later.
Logged

tradephoric

  • *
  • Offline Offline

  • Posts: 2435
  • Last Login: March 18, 2024, 11:18:58 AM
Re: Crash prone 'modern roundabouts'
« Reply #58 on: May 23, 2015, 11:05:47 AM »

Umm... guys? I don't know where that 297% number is coming from but check out historical aerial imagery in Google Earth.

The number came from a study funded by the Wisconsin Department of Transportation.  Looking at the historical aerials, I wouldn’t have included 25th Street & Canal in the analysis since the geometry of the before and after conditions is quite a bit different.  Recently, a three entry leg roundabout was constructed in Michigan that has seen a significant increase in total crashes.  I haven’t mentioned the location of the roundabout in this thread since it added another leg to the intersection that wasn’t previously there (and analyzing the before and after crash data wouldn’t be fair IMO).

Am I surprised that the Wisconsin Department of Transportation funded a report that analyzed a roundabout with dissimilar before and after conditions?  No.  This is the same agency who designed these crash prone multi-lane roundabouts to begin with.  To be fair, several of the roundabouts included in the report have similar before and after conditions.  If you take out the results of the 25th Street & Canal roundabout, the remaining multi-lane roundabouts saw an overall increase in crashes. 
Logged

DaBigE

  • *
  • Offline Offline

  • Posts: 1970
  • Following the LEGO® brick road

  • Age: 40
  • Location: Southcentral Wisconsin
  • Last Login: April 23, 2020, 04:51:41 PM
Re: Crash prone 'modern roundabouts'
« Reply #59 on: May 25, 2015, 01:52:02 AM »

I think roundabouts with three entry lanes are growing out of favor.  Roundabouts with three entry lanes were constructed as part of the US41 project in Wisconsin.  From my understanding, Wisconsin has no plans to design new roundabouts with three entry lanes moving forward.  DaBigE might have more information regarding this based on some posts I've read of his.

As far as I know, we're designing the last roundabout in Wisconsin that will have 2, three-lane entries. Yes, Wisconsin does now have a phobia/moratorium of any additional roundabouts with three-lane entries (with the exception of one region). Other states seem to be sharing the trepidation as well. We've worked on a few three-lane designs for other states, but many have ended up being scaled back. In many cases, three lane entries aren't needed for a long time, and depending on the analysis software/skill of the analyst, may never be needed. Attempting to design for 20-year forecasts can be very tricky/dangerous, especially for new developments.

There is a growing belief that many roundabouts are being over-designed, which leads to larger than necessary designs. The problem in the US is lack of reliable data (number of years of available of data), and correction factors that vary depending on what portion of the country you're in. If you read the WisDOT FDM section on roundabout design (FDM 11-26 if you're interested), Wisconsin uses it's own critical headways and follow-up headway values for capacity analysis.

There's also big debate over how to handle trucks (semis - WB-62s, WB-67s): should they stay in-lane, and if so, where? Just on the entry (aka Case II)? Throughout the roundabout movement (Case III)? Neither (Case I)? That can also lead to larger than necessary designs > faster operating speeds > more crashes.

Am I surprised that the Wisconsin Department of Transportation funded a report that analyzed a roundabout with dissimilar before and after conditions?  No.  This is the same agency who designed these crash prone multi-lane roundabouts to begin with.  To be fair, several of the roundabouts included in the report have similar before and after conditions.  If you take out the results of the 25th Street & Canal roundabout, the remaining multi-lane roundabouts saw an overall increase in crashes.

All I am going to say on this is there's a lot of flaws that can be found in WisDOT's roundabout crash studies (Phase III is set to be released later this summer).

The take-home message has been severe injury/fatal crashes have seen significant decreases. Fender-benders (aka PDO crashes) have gone up significantly in multilane roundabouts. Having read through many crash narratives, the majority have been attributed to failure to yield or improper lane usage). The stupid maneuvers people try to pull...it's almost so common it ceases to amaze me any more.
Logged
"We gotta find this road, it's like Bob's road!" - Rabbit, Twister

froggie

  • *
  • Offline Offline

  • Posts: 12911
  • Location: Greensboro, VT
  • Last Login: March 18, 2024, 11:08:21 PM
    • Froggie's Place
Re: Crash prone 'modern roundabouts'
« Reply #60 on: May 25, 2015, 12:18:13 PM »

DaBigE (and tradephoric too for that matter), have you looked into the research MnDOT completed last year on signage and striping with multilane roundabouts?  This is just one study of one multilane roundabout, but they found tweaking signage and striping led to significant decreases in the number of crashes caused by yield failures and improper lane usage:

http://www.dot.state.mn.us/research/TS/2014/201404TS.pdf (summary)

http://www.dot.state.mn.us/research/TS/2014/201404.pdf (study)
Logged

DaBigE

  • *
  • Offline Offline

  • Posts: 1970
  • Following the LEGO® brick road

  • Age: 40
  • Location: Southcentral Wisconsin
  • Last Login: April 23, 2020, 04:51:41 PM
Re: Crash prone 'modern roundabouts'
« Reply #61 on: May 25, 2015, 11:48:50 PM »

DaBigE (and tradephoric too for that matter), have you looked into the research MnDOT completed last year on signage and striping with multilane roundabouts?  This is just one study of one multilane roundabout, but they found tweaking signage and striping led to significant decreases in the number of crashes caused by yield failures and improper lane usage:

http://www.dot.state.mn.us/research/TS/2014/201404TS.pdf (summary)

http://www.dot.state.mn.us/research/TS/2014/201404.pdf (study)

Yes, I am fairly familiar with that study. My biggest issue with the study is that it's impossible to tell what had the most effect on driver behavior since they implemented two relatively significant changes at the same time. My gut says it was the change in signing, based driver observations on obeying markings as well as changes Madison, WI made to this roundabout, after it was in the top 5 worst intersections in the city for crashes for several years. It has since dropped off the radar/top ten. Several signs were added; no pavement markings were changed. Unfortunately Google StreetView doesn't show them, but the city added a few lane reminder signs in the outside corners (one significant problem was crossing crashes at the exits), as well as a standard NO RIGHT TURN with a LEFT LANE plaque on the eastbound approach. Similar to my roundabout size comment, many [designers] think this roundabout is sized a bit too large, as speeds are quite high.

What's interesting about that MnDOT study is many of the changes they implemented closely mimic WisDOT standards for roundabout signing/marking (although we tend to favor overhead signing when possible, rather than ground-mount). I've worked on a couple Minnesota roundabouts and trying to implement overhead signs is a PITA.
« Last Edit: May 26, 2015, 01:36:58 AM by DaBigE »
Logged
"We gotta find this road, it's like Bob's road!" - Rabbit, Twister

jakeroot

  • *
  • Offline Offline

  • Posts: 15874
  • 日本標準時

  • Age: 28
  • Location: Uruma-shi, Japan
  • Last Login: Today at 04:53:27 AM
    • Flickr
Re: Crash prone 'modern roundabouts'
« Reply #62 on: May 26, 2015, 01:26:22 AM »

Not sure what kind of signing might be helpful, but two additions that I think we could adopt from the UK:

1) Pavers indicating movement direction:



2) larger chevrons indicating movement direction:



While I'm not entirely certain how relevant left turns are at roundabouts these days (most studies, which I'm sure cover the stats, confuse me to no end), I would think these practices would put a stop to it. I know number 1 has been used before (couldn't quote any examples, I'm just certain I've seen it) but number 2, while we use chevrons, I don't think we've ever used it to such an extent. Perhaps while we're at it, we ought to bring back the circular one-way sign.

DaBigE, despite my suggestions above, do you think there is such a thing as signage overload? If so, do you think it holds some relevance to this here conversation?

tradephoric

  • *
  • Offline Offline

  • Posts: 2435
  • Last Login: March 18, 2024, 11:18:58 AM
Re: Crash prone 'modern roundabouts'
« Reply #63 on: May 26, 2015, 11:56:11 AM »

There's also big debate over how to handle trucks (semis - WB-62s, WB-67s): should they stay in-lane, and if so, where? Just on the entry (aka Case II)? Throughout the roundabout movement (Case III)? Neither (Case I)? That can also lead to larger than necessary designs > faster operating speeds > more crashes.

I’m very interested in this comment DaBigE.  Assuming there is an adequate entry deflection angle to slow the speed of traffic, wouldn’t larger roundabouts potentially result in fewer crashes and not more crashes?  With larger roundabouts, it’s easier to physically stay in your lane (likely resulting in fewer side-swipe crashes).  In addition, drivers might be able to judge available gaps in traffic better, since there is greater distance between legs.  With smaller roundabouts, entering drivers seem to be guessing what circulating vehicles intent is (since there isn’t as much distance between legs).   

The multi-lane roundabouts in SE Michigan that have seen significant increase in total crashes all seem to have smaller footprints (IE. central island diameters of 100 feet or less).  The roundabout at Ellsworth & State (that saw nearly a 10x increase in crashes) has a central island diameter of 100 feet.  Conversely, there are several examples of larger roundabouts in SE Michigan with central island diameters of greater than 150 feet that haven’t seen significant increases in crashes (with some seeing significant crash reductions).
Logged

DaBigE

  • *
  • Offline Offline

  • Posts: 1970
  • Following the LEGO® brick road

  • Age: 40
  • Location: Southcentral Wisconsin
  • Last Login: April 23, 2020, 04:51:41 PM
Re: Crash prone 'modern roundabouts'
« Reply #64 on: May 26, 2015, 07:05:03 PM »

There's also big debate over how to handle trucks (semis - WB-62s, WB-67s): should they stay in-lane, and if so, where? Just on the entry (aka Case II)? Throughout the roundabout movement (Case III)? Neither (Case I)? That can also lead to larger than necessary designs > faster operating speeds > more crashes.

I’m very interested in this comment DaBigE.  Assuming there is an adequate entry deflection angle to slow the speed of traffic, wouldn’t larger roundabouts potentially result in fewer crashes and not more crashes?  With larger roundabouts, it’s easier to physically stay in your lane (likely resulting in fewer side-swipe crashes).  In addition, drivers might be able to judge available gaps in traffic better, since there is greater distance between legs.  With smaller roundabouts, entering drivers seem to be guessing what circulating vehicles intent is (since there isn’t as much distance between legs).

While the amount of deflection is important, where and how the deflection ties into the circulatory is just as/more important. Many of the designs I've seen implemented lately are what we refer to as "hockey stick" designs... where the R1 entry curve that is too far upstream which results in a very flat entry at the circulatory (long tangent). These long tangents allow drivers to accelerate as they approach the yield point where they instead should be slowing down. Combine that with a larger roundabout diameter (ICD*), the faster the vehicle speeds though the roundabout (R1 & R2 speeds). The larger roundabouts may work safer when circulating volumes are higher or when the roundabout is operating closer to its capacity. However, the majority of crash reports I've seen occur during the off-peak times. During off-peak times, you might as well not paint any lines, as drivers tend to follow the fast-path though the roundabout. The faster the roundabout operates, the more willing drivers are to take smaller gaps, and crash doing so because they misjudged the speed of the circulating vehicle.

This is not to say that small roundabouts are the best. Like Goldilocks, there is a "just right" size. The general consensus regarding ICD is an inverse relationship between size and safety due to the direct relationship between size and speed.

The multi-lane roundabouts in SE Michigan that have seen significant increase in total crashes all seem to have smaller footprints (IE. central island diameters of 100 feet or less).  The roundabout at Ellsworth & State (that saw nearly a 10x increase in crashes) has a central island diameter of 100 feet.  Conversely, there are several examples of larger roundabouts in SE Michigan with central island diameters of greater than 150 feet that haven’t seen significant increases in crashes (with some seeing significant crash reductions).

A central island diameter of 100-feet is way to small even for a single lane roundabout. A typical minimum ICD for a multilane roundabout is approximately 160-feet. Assuming approximately 28-feet of pavement for the circulatory, that would net you a central island of 132-feet 104-feet (truck apron inclusive).

*NOTE: designers generally refer to the size of the roundabout based on the diameter to the face of the outside curb of the circulatory, or the inscribed circle diameter

**Edited to correct math error
« Last Edit: May 26, 2015, 11:17:46 PM by DaBigE »
Logged
"We gotta find this road, it's like Bob's road!" - Rabbit, Twister

DaBigE

  • *
  • Offline Offline

  • Posts: 1970
  • Following the LEGO® brick road

  • Age: 40
  • Location: Southcentral Wisconsin
  • Last Login: April 23, 2020, 04:51:41 PM
Re: Crash prone 'modern roundabouts'
« Reply #65 on: May 26, 2015, 07:30:38 PM »

DaBigE, despite my suggestions above, do you think there is such a thing as signage overload? If so, do you think it holds some relevance to this here conversation?

YES. That is precisely why there is minimum recommended spacing between signs in general and maximum recommended amounts of legend in a guide sign. I cringe a little every time I have to deal with an intersection of more than two or three routes. Post too many signs, and drivers will ignore them. This is why I don't like the Ped Xing sign being posted at every. single. crosswalk. For every sign you post, there is a perception-reaction time associated with it. Despite what some think, posting another sign will not correct an overall design flaw or lack of education.

I am a firm believer in less is more - to a point. The key is getting the important information to the driver at the right time. That is why I personally prefer overhead lane signage to post-mounted, especially in urban areas. Getting drivers into the appropriate lane early when approaching a roundabout is critical.

Not sure what kind of signing might be helpful, but two additions that I think we could adopt from the UK:

1) Pavers indicating movement direction:


This has been tried once or twice before, and the end result was a PITA to install and maintain. It's too subtle for many US drivers to comprehend. With so many central islands being used for public artwork, many think it's just part of the design. IMO, it's a waste of money.

Quote
2) larger chevrons indicating movement direction:


In my experiences, wrong-way movements with regards to a roundabout appears to be a shrinking problem. Referencing the discussion with tradephoric, properly executed deflection should render these signs virtually useless. Improper yielding is the biggest concern, which is why WisDOT uses the plaque TO TRAFFIC FROM LEFT, with other states and municipalities using variants of that.

Perhaps while we're at it, we ought to bring back the circular one-way sign.

Testing may prove me wrong, but I don't think another style of sign is going to gain compliance. Only a vehicle with lights and siren seem to work.
Logged
"We gotta find this road, it's like Bob's road!" - Rabbit, Twister

tradephoric

  • *
  • Offline Offline

  • Posts: 2435
  • Last Login: March 18, 2024, 11:18:58 AM
Re: Crash prone 'modern roundabouts'
« Reply #66 on: May 26, 2015, 10:05:53 PM »

A central island diameter of 100-feet is way to small even for a single lane roundabout. A typical minimum ICD for a multilane roundabout is approximately 160-feet. Assuming approximately 28-feet of pavement for the circulatory, that would net you a central island of 132-feet (truck apron inclusive).

The top 5 highest crash frequency intersections in Oshkosh, Wisconsin are at multi-lane roundabouts with central island diameters of 100 feet or less.  Here is a table of the Oshkosh roundabouts with the construction year highlighted in yellow:


http://www2.ci.oshkosh.wi.us/WebLink8/0/doc/711055/Electronic.aspx

I’m starting to see a theme...  multi-lane roundabouts with central island diameters of 100 feet or less have high crash frequencies (especially at roundabouts with high AADT).  Nearly every roundabout cited in this thread (apart from the triple lane roundabout from Lincoln, Nebreska) has had a central island diameter of roughly 100 feet.
Logged

jeffandnicole

  • *
  • Offline Offline

  • Posts: 14811
  • Age: 49
  • Location: South Jersey
  • Last Login: March 18, 2024, 10:52:41 PM
Re: Crash prone 'modern roundabouts'
« Reply #67 on: May 26, 2015, 10:35:31 PM »

A central island diameter of 100-feet is way to small even for a single lane roundabout. A typical minimum ICD for a multilane roundabout is approximately 160-feet. Assuming approximately 28-feet of pavement for the circulatory, that would net you a central island of 132-feet (truck apron inclusive).

The top 5 highest crash frequency intersections in Oshkosh, Wisconsin are at multi-lane roundabouts with central island diameters of 100 feet or less.  Here is a table of the Oshkosh roundabouts with the construction year highlighted in yellow:


http://www2.ci.oshkosh.wi.us/WebLink8/0/doc/711055/Electronic.aspx

I’m starting to see a theme...  multi-lane roundabouts with central island diameters of 100 feet or less have high crash frequencies (especially at roundabouts with high AADT).

It would probably be fair to see a longer list, which will show both 100' diameter roundabouts that don't have high crash frequencies, and roundabouts with larger diameters and how they rank. 
Logged

DaBigE

  • *
  • Offline Offline

  • Posts: 1970
  • Following the LEGO® brick road

  • Age: 40
  • Location: Southcentral Wisconsin
  • Last Login: April 23, 2020, 04:51:41 PM
Re: Crash prone 'modern roundabouts'
« Reply #68 on: May 26, 2015, 11:28:07 PM »

A central island diameter of 100-feet is way to small even for a single lane roundabout. A typical minimum ICD for a multilane roundabout is approximately 160-feet. Assuming approximately 28-feet of pavement for the circulatory, that would net you a central island of 132-feet (truck apron inclusive).

The top 5 highest crash frequency intersections in Oshkosh, Wisconsin are at multi-lane roundabouts with central island diameters of 100 feet or less.  Here is a table of the Oshkosh roundabouts with the construction year highlighted in yellow:


http://www2.ci.oshkosh.wi.us/WebLink8/0/doc/711055/Electronic.aspx

I’m starting to see a theme...  multi-lane roundabouts with central island diameters of 100 feet or less have high crash frequencies (especially at roundabouts with high AADT).  Nearly every roundabout cited in this thread (apart from the triple lane roundabout from Lincoln, Nebreska) has had a central island diameter of roughly 100 feet.

And after correcting my math error (forgot I was working with a diameter, not a radius :pan:), the 100-foot central island falls right at the prescribed minimum for a multilane roundabout size - at least according to WisDOT and NCHRP Report 679. In fact, Report 679 suggests a minimum ICD as low as 150-feet. IIRC, Europe has even more compact roundabouts, yet not nearly the crash problem.
Logged
"We gotta find this road, it's like Bob's road!" - Rabbit, Twister

DaBigE

  • *
  • Offline Offline

  • Posts: 1970
  • Following the LEGO® brick road

  • Age: 40
  • Location: Southcentral Wisconsin
  • Last Login: April 23, 2020, 04:51:41 PM
Re: Crash prone 'modern roundabouts'
« Reply #69 on: May 26, 2015, 11:45:23 PM »

A central island diameter of 100-feet is way to small even for a single lane roundabout. A typical minimum ICD for a multilane roundabout is approximately 160-feet. Assuming approximately 28-feet of pavement for the circulatory, that would net you a central island of 132-feet (truck apron inclusive).

The top 5 highest crash frequency intersections in Oshkosh, Wisconsin are at multi-lane roundabouts with central island diameters of 100 feet or less.  Here is a table of the Oshkosh roundabouts with the construction year highlighted in yellow:


http://www2.ci.oshkosh.wi.us/WebLink8/0/doc/711055/Electronic.aspx

I’m starting to see a theme...  multi-lane roundabouts with central island diameters of 100 feet or less have high crash frequencies (especially at roundabouts with high AADT).

It would probably be fair to see a longer list, which will show both 100' diameter roundabouts that don't have high crash frequencies, and roundabouts with larger diameters and how they rank.

It would also be valuable to note that Washburn & 9th, Koeller & 9th are part of a corridor of 4 closely-spaced roundabouts. Frankly, I'm surprised to not see the roundabouts at Wis 21/I-41/Washburn/Koeller appear high on the list, as those are also similarly designed roundabouts, a few with 3-lane entries. Maybe they show up further down on the list...can't tell right now because I get a server error when attempting to use the link tradephoric provided. However, IIRC, Wis 21 wasn't opened until 2012 or early 2013.

Does anyone have crash data for Carmel, Indiana?
Logged
"We gotta find this road, it's like Bob's road!" - Rabbit, Twister

tradephoric

  • *
  • Offline Offline

  • Posts: 2435
  • Last Login: March 18, 2024, 11:18:58 AM
Re: Crash prone 'modern roundabouts'
« Reply #70 on: May 27, 2015, 03:05:59 PM »

It would probably be fair to see a longer list, which will show both 100' diameter roundabouts that don't have high crash frequencies, and roundabouts with larger diameters and how they rank. 

Here is a list of 40 multi-lane roundabouts that I think could be meaningful to this conversation.  This was queried from a database of over 5,100 modern roundabouts.  This was the criteria used to query out the list:

-roundabouts constructed within the last 10 years
-all approaches have 2 entry lanes
-roundabouts have 4-legs
-main & side streets are major state or county routes (ie. likely high AADT roundabouts)
-interchange roundabouts (and frontage road roundabouts) not included



Now the hard part is finding accurate before/after crash data to make any type of analysis.  Here is a google KMZ file that includes the 40 roundabouts in the chart above:

http://www.mediafire.com/download/cd570rarros2c3g/Multi-Lane+Roundabouts+%28with+high+AADT%29.kmz

Logged

tradephoric

  • *
  • Offline Offline

  • Posts: 2435
  • Last Login: March 18, 2024, 11:18:58 AM
Re: Crash prone 'modern roundabouts'
« Reply #71 on: May 27, 2015, 04:49:38 PM »

Crashes up at multi-lane roundabout in Cheyenne, Wyoming:

http://www.wyomingnews.com/articles/2014/08/15/news/19local_08-15-14.txt#.VWYsqEYmlBI
Logged

tradephoric

  • *
  • Offline Offline

  • Posts: 2435
  • Last Login: March 18, 2024, 11:18:58 AM
Re: Crash prone 'modern roundabouts'
« Reply #72 on: May 27, 2015, 04:50:41 PM »

Logged

colinstu

  • *
  • Offline Offline

  • Posts: 397
  • Last Login: August 10, 2023, 12:09:09 PM
Re: Crash prone 'modern roundabouts'
« Reply #73 on: May 27, 2015, 05:08:07 PM »

I think one important factor that needs to be considered is the severity of the crashes, not the number of them. Assigning weights to different kinds of crashes to how commonly they occur to more accurately compare them. What also needs to be considered are traffic counts. If traffic at the intersection doubles... Well no doubt there will be more collisions. All these variables need to be leveled out to make any kind of meaningful comparison.

Roundabouts (properly designed) vs 4-way intersection aren't going to have head-on or t-bone collisions. While they may have more fender benders (and thus more "crashes") ... The severity of those (and risk of injury / death) is much lower. I'd rather have more safe accidents then less deadly accidents.
Logged

DaBigE

  • *
  • Offline Offline

  • Posts: 1970
  • Following the LEGO® brick road

  • Age: 40
  • Location: Southcentral Wisconsin
  • Last Login: April 23, 2020, 04:51:41 PM
Re: Crash prone 'modern roundabouts'
« Reply #74 on: May 27, 2015, 05:26:35 PM »

Three of Madison's most crash prone intersections are roundabouts:

http://www.channel3000.com/news/Three-of-Madison-s-most-crash-prone-intersections-are-roundabouts/16170598

That article is from 3 years ago. Virtually all of Madison's roundabouts are out of the top ten list in the report from last year.
Logged
"We gotta find this road, it's like Bob's road!" - Rabbit, Twister

 


Opinions expressed here on belong solely to the poster and do not represent or reflect the opinions or beliefs of AARoads, its creators and/or associates.