Crash prone 'modern roundabouts'

[jakeroot]

What's your thoughts on the Long Beach/Los Alamitos Circle (if you haven't addressed it before)? It doesn't really look like anything else in the US.

According to this LA Times article from 2006, lane lines were considered, but eventually dropped, to more closely conform to the British standards of the time: https://goo.gl/n99389

Fascinating. I'd never even thought about having circles/roundabouts without lane lines. Could that be a better, less-confusing solution for problematic two/three-lane roundabouts?

My suspicion is that it might reduce one problem and augment another.  But that's just a hunch.

'Tis the case for all traffic control devices. But it's a matter of figuring out which device is the least problematic, assuming all traffic control devices can, in some way, be interpreted improperly.

The famous one in Towson (MD) used to have no lines, but that's no longer the case.


(Google imagery here)

I suspect the Towson circle was changed for the sake of change. A lot of roundabouts in the US, especially those in New England, had "modern roundabout" markings grafted on top of them, which in theory should improve the flow and reduce collisions. But I believe that, barring a few examples, some/many have seen an increase in collisions. Perhaps with all the markings, people were starting to let their guard down? It's said that the Long Beach Circle works because of the heightened sense of awareness that comes along with no markings.

[kphoger]

I suspect the Towson circle was changed for the sake of change.

The Towson roundabout has gone through periodic tweaking for a good number of years, now.  Lane width, striping, curbing, and other stuff.  Basically, they've experimented with it and figured out what works and what doesn't for their particular situation.  That's not to say everything from Towson would work at another location, but I don't see Towson doing things just for the sake of change.

[D-Dey65]

FDOT #7 wants to put a new one up on US 98 and Hernando CR 491 in Deschamp's Corner. I think it's a bad idea

[tradephoric]

Looking at the Los Alamitos Circle on Google imagery you can tell it carries a massive amount of traffic (3 of the 4 legs at the roundabout are 6-lane roads downstream).  Also looking through the historical imagery it doesn't appear the circle has been touched for the past 25 years.  That said i don't know what the crash data is like.  Does it have 150 crashes per year?  If not, it's got fewer crashes than the highest crash roundabouts discussed so far.

[kphoger]

Looking at the Los Alamitos Circle on Google imagery you can tell it carries a massive amount of traffic (3 of the 4 legs at the roundabout are 6-lane roads downstream).  Also looking through the historical imagery it doesn't appear the circle has been touched for the past 25 years.  That said i don't know what the crash data is like.  Does it have 150 crashes per year?  If not, it's got fewer crashes than the highest crash roundabouts discussed so far.

An article from 2013 cites the Long Beach Engineering Bureau as saying the circle tops the list for accidents in the city, but doesn't say what the rate is.  That kind of data would require a call to the bureau.

[tradephoric]

In other news, a 71 year old pedestrian was killed at the Fifth Avenue South and University Drive roundabout in Saint Cloud on Tuesday.

St. Cloud pedestrian dies after being hit at university roundabout
http://www.sctimes.com/story/news/local/2017/10/25/st-cloud-pedestrian-dies-after-being-hit-university-roundabout/799620001/

In yet other news, a deadly intersection in Nevada is going to get a roundabout.  I wonder when a deadly roundabout is going to get an intersection   The Westfield Blvd and 96th Street roundabout in Carmel has been the site of 3 fatalities since it was built (in each fatal crash the driver blew through the middle of the roundabout and hit a retaining wall in the central island).  The perception is roundabouts are safe and there is no public outcry to tear the roundabout down even as 3 people have lost their lives. 

Deadly intersection to get roundabout
http://www.recordcourier.com/news/local/deadly-intersection-to-get-roundabout/

[kalvado]

I've never driven through a 3-lane roundabout, but I can easily imagine making a dangerous maneuver at one.  And I'm a relatively young guy who likes roundabouts; first-time drivers, the elderly, people distracted by something else...  I tell you what, I'm coming around more and more to the turbo roundabout concept.  I wasn't thrilled at first, but I think I'm starting to get it.

There are many great concepts floating around which never got past the concept stage, and quite a few which failed in production and were successfully abandoned.
Nuclear powered cars and planes come to mind as most relevant to this forum - although nuclear warships are out there.. Fuel cell powered everything as a more recent example. I even touched fuel cell powered car (I could get a ride if I wanted) - but didn't see any lately.

Turbo roundabouts exist in reality, though, and have since the turn of the century.

Planes with nuclear power plant did fly, and fuel cell cars did hit the road.  FUel cell laptops were demonstrated. Didn't make it viable in a long run, though

[MNHighwayMan]

In other news, a 71 year old pedestrian was killed at the Fifth Avenue South and University Drive roundabout in Saint Cloud on Tuesday.

To be fair, it is St. Cloud we're talking about. Not exactly a stellar example of the best of humanity. (Note that I am talking about the driver, not the pedestrian.)

[Brian556]

What's your thoughts on the Long Beach/Los Alamitos Circle (if you haven't addressed it before)? It doesn't really look like anything else in the US.

According to this LA Times article from 2006, lane lines were considered, but eventually dropped, to more closely conform to the British standards of the time: https://goo.gl/n99389

Fascinating. I'd never even thought about having circles/roundabouts without lane lines. Could that be a better, less-confusing solution for problematic two/three-lane roundabouts?

My suspicion is that it might reduce one problem and augment another.  But that's just a hunch.

The famous one in Towson (MD) used to have no lines, but that's no longer the case.


(Google imagery here)

What the heck is that? An ovalabout?

[jakeroot]

I suspect the Towson circle was changed for the sake of change.

The Towson roundabout has gone through periodic tweaking for a good number of years, now.  Lane width, striping, curbing, and other stuff.  Basically, they've experimented with it and figured out what works and what doesn't for their particular situation.  That's not to say everything from Towson would work at another location, but I don't see Towson doing things just for the sake of change.

"Change for the sake of change" was the wrong phrasing. They changed things with the understanding that, whatever the "situation" was, would improve. I assume they wouldn't change things just because they can. I would hope there was a reason. As we can see in this thread, a lot of roundabouts are built to reduce collisions, although the opposite usually occurs. I can't help but wonder if crashes increased after the modern roundabout markings were first installed.

I really want to know the stats on the Towson Circle from its construction in the late 90s, to when they made the first changes in 2007. That's a long time without changes if crashes were a problem, especially if, like you say, the city continues to modify the circle to figure out what works "best".

[jakeroot]

The famous one in Towson (MD) used to have no lines, but that's no longer the case.

http://councilmandavidmarks.com/wp-content/uploads/2013/06/towsontraffic.jpg

What the heck is that? An ovalabout?

There's a couple other roundabouts with similar shapes:

US-395 @ Hawthorne & Glenn Ave's, Colville, WA: https://goo.gl/wdWF45
US-101 @ Milpas St, Santa Barbara, CA: https://goo.gl/TMwxG5

I seem to recall them just being referred to as "squeezed roundabouts" but I'm pretty sure there's an official name.

[MNHighwayMan]

Maybe I've missed it, given that I haven't thoroughly read the entire thread, but I feel like there's an important nuance that this discussion seems to miss: the severity of accidents. While there sure seems to be a lot of talk about how accident rates have increased at these roundabouts, I think a key factor that seems to be overlooked is the severity. I'd much rather have, for example, 100 side-swipe accidents, which usually result in few or no injuries, than 20 head-on or side-impact collisions that could potentially result in serious injuries.

Again, maybe I missed the discussion about this because I haven't read this thread fully. Still, I think it's something to keep in mind while discussing this topic.

[jakeroot]

We have discussed severity, and it's a point of contention amongst sides here. Tradephoric, and those that lean his direction, maintain that severe collisions at normal intersections, while technically possible, are rather uncommon, so the low crash rate of most normal intersections is considered more desirable, if choosing between the two

The other side believes that roundabouts don't produce as-severe collisions due to the angle of impact. The problem is that not all drivers are paying attention when approaching a roundabout, and if you so choose to, you can drive straight over the roundabout, and straight into whatever cars are passing through the circle. The angle of impact can be 90-degrees in these situations, and can be just as severe of a collision as a T-bone at a normal signal. Then, you have all the fender benders and side-swipes from people cocking about while driving through the roundabout. After a year, you might only have one death in a roundabout, but you also have way more collisions overall.

As a percentage of collisions, do signals have more severe collisions? Maybe. But, crashes in general are pretty uncommon, so that's not saying much.

[kalvado]

Maybe I've missed it, given that I haven't thoroughly read the entire thread, but I feel like there's an important nuance that this discussion seems to miss: the severity of accidents. While there sure seems to be a lot of talk about how accident rates have increased at these roundabouts, I think a key factor that seems to be overlooked is the severity. I'd much rather have, for example, 100 side-swipe accidents, which usually result in few or no injuries, than 20 head-on or side-impact collisions that could potentially result in serious injuries.

Again, maybe I missed the discussion about this because I haven't read this thread fully. Still, I think it's something to keep in mind while discussing this topic.

You need to start assigning weights  and do fine accounting to get  required "improvement"... WHich basically means concept already falls apart.
Anecdotal data: I have  huge roundabout  300 feet from my desk.  Project data: 0 fatal accidents in prior 10 years.  Since constructed ~10 years ago: at least 1 fatal accident. But this is a small price to pay in the name of safe... coolness, right?

Simple estimate: 1 death cost is about equal to cost of 100-500  sidesweep accidents. (yes, human life has a cost)  How many intersections are there with annual fatal accidents to justify 100 crashes/year on replacement roundabout?

[US 89]

What's your thoughts on the Long Beach/Los Alamitos Circle (if you haven't addressed it before)? It doesn't really look like anything else in the US.

According to this LA Times article from 2006, lane lines were considered, but eventually dropped, to more closely conform to the British standards of the time: https://goo.gl/n99389

That looks like a recipe for sideswipes and disasters.

The problem with trying to be British is that the drivers who are going to go through this aren't British.

[Truvelo]

I don't see any problems with that Long Beach circle. In fact it looks better than most modern designs. Too many modern roundabouts have excessive deflection on the approaches causing semi trucks to flip over. I don't see Britishness being a problem, especially when American drivers get used to them once they become common.

[MNHighwayMan]

You need to start assigning weights  and do fine accounting to get  required "improvement"... WHich basically means concept already falls apart.
Anecdotal data: I have  huge roundabout  300 feet from my desk.  Project data: 0 fatal accidents in prior 10 years.  Since constructed ~10 years ago: at least 1 fatal accident. But this is a small price to pay in the name of safe... coolness, right?

Simple estimate: 1 death cost is about equal to cost of 100-500  sidesweep accidents. (yes, human life has a cost)  How many intersections are there with annual fatal accidents to justify 100 crashes/year on replacement roundabout?

Anecdotal data is meaningless, and that sample size is so ridiculously small that it makes it even more meaningless. Also, if I have to choose, I'm going with hundreds of dinged up fenders and bumpers over one person's death. It's ridiculous to even equate the two.

[kalvado]

You need to start assigning weights  and do fine accounting to get  required "improvement"... WHich basically means concept already falls apart.
Anecdotal data: I have  huge roundabout  300 feet from my desk.  Project data: 0 fatal accidents in prior 10 years.  Since constructed ~10 years ago: at least 1 fatal accident. But this is a small price to pay in the name of safe... coolness, right?

Simple estimate: 1 death cost is about equal to cost of 100-500  sidesweep accidents. (yes, human life has a cost)  How many intersections are there with annual fatal accidents to justify 100 crashes/year on replacement roundabout?

Anecdotal data is meaningless, and that sample size is so ridiculously small that it makes it even more meaningless. Also, if I have to choose, I'm going with hundreds of dinged up fenders and bumpers over one person's death. It's ridiculous to even equate the two.

You're a young men with noble heart, I am an old fart who is  totally down-to-earth... May I ask how many fender benders are equal to one fatal accident in your world? 1000? 10,000? 1,000,000? Would you be willing to break every vehicle in your city if that would save one life?
Statistically speaking it will work - cities do have their fair share of road fatalities, and breaking all vehicles around should reduce this number - until a new fleet is purchased...

[jakeroot]

I don't see any problems with that Long Beach circle. In fact it looks better than most modern designs. Too many modern roundabouts have excessive deflection on the approaches causing semi trucks to flip over. I don't see Britishness being a problem, especially when American drivers get used to them once they become common.

It seems to work pretty well. The idea of a large circle without any markings is very foreign to us Americans, particularly Southern California drivers. You can watch videos of the circle on YouTube. It seems like drivers will cut other drivers off a lot, enter without yielding, and continue circling even from the outer edge (pretty sure outer lane always exits as a matter of principle). But still doesn't seem to perform all that bad, especially considering the amount of traffic it handles (as tradephoric alluded to just above).

[CtrlAltDel]

You're a young men with noble heart, I am an old fart who is  totally down-to-earth... May I ask how many fender benders are equal to one fatal accident in your world? 1000? 10,000? 1,000,000? Would you be willing to break every vehicle in your city if that would save one life?
Statistically speaking it will work - cities do have their fair share of road fatalities, and breaking all vehicles around should reduce this number - until a new fleet is purchased...

Well, the FAA (admittedly not the purview here) values a human life at 5.8 million dollars. If your average fender bender causes damage in the amount of 1,500 dollars, the tipping point would be 3,867 fender benders. So, that makes a good starting point.

[kalvado]

You're a young men with noble heart, I am an old fart who is  totally down-to-earth... May I ask how many fender benders are equal to one fatal accident in your world? 1000? 10,000? 1,000,000? Would you be willing to break every vehicle in your city if that would save one life?
Statistically speaking it will work - cities do have their fair share of road fatalities, and breaking all vehicles around should reduce this number - until a new fleet is purchased...

Well, the FAA (admittedly not the purview here) values a human life at 5.8 million dollars. If your average fender bender causes damage in the amount of 1,500 dollars, the tipping point would be 3,867 fender benders. So, that makes a good starting point.

$1-2M is a more realistic number given that money come directly from drivers (maybe in the form of insurance).  FAA enforces extremely high standard...
And as we were talking sidesweep, I assumed 3 doors, 2 quarter panels and a headlight - a bit more than $1.5k..

[silverback1065]

What's your thoughts on the Long Beach/Los Alamitos Circle (if you haven't addressed it before)? It doesn't really look like anything else in the US.

According to this LA Times article from 2006, lane lines were considered, but eventually dropped, to more closely conform to the British standards of the time: https://goo.gl/n99389

Fascinating. I'd never even thought about having circles/roundabouts without lane lines. Could that be a better, less-confusing solution for problematic two/three-lane roundabouts?

My suspicion is that it might reduce one problem and augment another.  But that's just a hunch.

'Tis the case for all traffic control devices. But it's a matter of figuring out which device is the least problematic, assuming all traffic control devices can, in some way, be interpreted improperly.

The famous one in Towson (MD) used to have no lines, but that's no longer the case.


(Google imagery here)

I suspect the Towson circle was changed for the sake of change. A lot of roundabouts in the US, especially those in New England, had "modern roundabout" markings grafted on top of them, which in theory should improve the flow and reduce collisions. But I believe that, barring a few examples, some/many have seen an increase in collisions. Perhaps with all the markings, people were starting to let their guard down? It's said that the Long Beach Circle works because of the heightened sense of awareness that comes along with no markings.

Thats more of a turbine square than a roundabout

[tradephoric]

Well, the FAA (admittedly not the purview here) values a human life at 5.8 million dollars. If your average fender bender causes damage in the amount of 1,500 dollars, the tipping point would be 3,867 fender benders. So, that makes a good starting point.

Here is a chart taken from the FHWA website detailing comprehensive crash costs by injury severity level:

Fatality (K)           $4,008,900
Disabling Injury (A)   $216,000
Evident Injury (B)   $79,000
Fatal/Injury (K/A/B)   $158,200
Possible Injury (C)   $44,900
PDO (O)   $7,400

They estimates the comprehensive crash costs of a fatal crash at $4 million.  The thing is fatal crashes are a very rare event.  To put things into perspective I took the most crash prone intersection in Michigan over the past 10 years according to SEMCOG (Haggerty & Ford Road) and broke out the Crashes by Severity:

Fatal = 0 X $4,008,900 = 0
A-Level = 8 X $216,000 = $1,728,000
B-Level = 45 X $79,000 = $3,555,000
C-Level = 131 X $44,900 = $5,881,900
PD0 = 690 X $7,400 = $5,106,000
TOTAL COST = $16,270,900

Of the 874 total crashes at Haggerty and Ford, 690 were PDO crashes and 184 were injury crashes.  There were no fatal crashes at this signalized intersection over the past 10 years.  The estimated cost for all these crashes comes to $16,270,900.  Now suppose they decide to install a roundabout at the intersection Let's assume the roundabout will eliminate all injury crashes but it will increase PDO crashes by 4X.   

Fatal = 0 X $4,008,900 = 0
A-Level = 0 X $216,000 = 0
B-Level = 0 X $79,000 = 0
C-Level = 0 X $44,900 = 0
PD0 = 2760 X $7,400 = $20,424,000
TOTAL COST = $20,424,000

Maybe arguing that the roundabout will see a 4 fold increase in crashes is too high.  But there are plenty of examples of similar increases when complex roundabouts have been built.  And obviously roundabouts don't eliminate all injury crashes.  The point is a compelling case can be made that building a roundabout at Haggerty and Ford would actually increase comprehensive crash costs at the intersection.

[cjw2001]

In other news, a 71 year old pedestrian was killed at the Fifth Avenue South and University Drive roundabout in Saint Cloud on Tuesday.

St. Cloud pedestrian dies after being hit at university roundabout
http://www.sctimes.com/story/news/local/2017/10/25/st-cloud-pedestrian-dies-after-being-hit-university-roundabout/799620001/

In yet other news, a deadly intersection in Nevada is going to get a roundabout.  I wonder when a deadly roundabout is going to get an intersection   The Westfield Blvd and 96th Street roundabout in Carmel has been the site of 3 fatalities since it was built (in each fatal crash the driver blew through the middle of the roundabout and hit a retaining wall in the central island).  The perception is roundabouts are safe and there is no public outcry to tear the roundabout down even as 3 people have lost their lives. 

Deadly intersection to get roundabout
http://www.recordcourier.com/news/local/deadly-intersection-to-get-roundabout/

The roundabout at Westfield and 96th isn't going away.  New artwork just installed. 

https://www.youtube.com/watch?v=u2KeQBtaPLM


Get over it.

[jakeroot]

^^
I really wish cities would just install shrubbery within the center of circles. I don't think people drive through the middle of roundabouts very often (obviously, it would be accidental), but with a sculpture in the middle, the potential for damage and injury is much higher than with some bushes.