News:

Thanks to everyone for the feedback on what errors you encountered from the forum database changes made in Fall 2023. Let us know if you discover anymore.

Main Menu

Crash prone 'modern roundabouts'

Started by tradephoric, May 18, 2015, 02:51:37 PM

Previous topic - Next topic

tradephoric

It's not just total crashes that are a problem at the Homer Watson roundabout.  According to the record, injuries have more than doubled since the roundabout opened in 2011.  This is another example of a large multi-lane roundabout seeing a spike in total crashes with no reduction in injury crashes (a lose-lose).   The argument has been that crashes that occur at a roundabout are low impact "fender benders" .  If that is true, how did 40 people get injured at the Homer Watson roundabout in just the first 3 years of operation?   

QuoteIts safety performance is not as poor as the dangerous roundabout it overtook, Homer Watson Boulevard at Block Line Road. After opening in 2011, the Kitchener intersection saw its collisions soar to 282 from 44, while injuries more than doubled to 40 people hurt from 16 hurt. This compares its first three-plus years as a roundabout to its last three-plus years as a traffic light. The second-busiest roundabout shed seven per cent of its traffic in 2014 to fall to second place.

http://m.therecord.com/news-story/6218115-86-crashes-in-first-17-months-collisions-soar-at-new-cambridge-roundabout


tradephoric

Quote from: 6a on May 20, 2015, 03:52:13 PM

Quote from: jakeroot on May 20, 2015, 03:38:09 AM
Quote from: KEK Inc. on May 20, 2015, 03:26:57 AM
I think they're a bit too close to each other.  I also went during rush hour.

I'd rather have two closely spaced roundabouts than two closely spaced signals. Then again, they don't seem that close?

These work just fine. A bit slow at rush hour but compared to the old signals it's a big improvement.



These two roundabouts in Hilliard were the #1 & #3 most crash prone intersections in the entire state of Ohio for the years 2013-2015.  There were 218 crashes at the Cemetery/Main roundabout and 119 crashes at the Main/Scioto Darby roundabout.  Hell, even the Google streetview car captured a crash at the roundabout:





http://branlawfirm.com/most-dangerous-intersections-in-ohio/

johndoe

https://www.google.com/maps/@43.4183801,-80.470737,110m/data=!3m1!1e3

The Homer Watson roundabout appears to have an ICD of about 200', so it's not undersized.  Do you see any patterns in the crash reports?  You can see they striped out the third circulating lane for the SB approach, I wonder if crashes were reduced after they made that change.

jakeroot

Quote from: tradephoric on March 08, 2016, 12:11:59 PM
It's not just total crashes that are a problem at the Homer Watson roundabout.  According to the record, injuries have more than doubled since the roundabout opened in 2011.  This is another example of a large multi-lane roundabout seeing a spike in total crashes with no reduction in injury crashes (a lose-lose).   The argument has been that crashes that occur at a roundabout are low impact "fender benders" .  If that is true, how did 40 people get injured at the Homer Watson roundabout in just the first 3 years of operation?

What constitutes an injury? Technically, both breaking your neck and straining your neck are injuries. A t-bone collision could result in several broken bones, but no deaths, and would be labelled an "injury collision". Equally, being rear-ended and bumping your head on the steering wheel is also an "injury collision" (presuming one sought medical attention, reporting the injury to insurance).

As we've previously discussed, there are multiple levels of injuries. Injuries may have increased, but it's the severity that matters. It's ridiculous to support an intersection design that (perhaps) puts people in hospital for months, but oppose an intersection design that results in several visits to a chiropractor, simply because there's fewer injuries overall.

If one were to put a price tag on "emotional distress" caused by collisions (as many lawyers do, in order to offset the loss of income from being in hospital), my guess is that 6 or 7 really bad injury crashes is still much pricier than 40 or 50 head-bumps and neck-strains.

tradephoric

Quote from: johndoe on March 08, 2016, 07:53:48 PMThe Homer Watson roundabout appears to have an ICD of about 200', so it's not undersized.  Do you see any patterns in the crash reports?  You can see they striped out the third circulating lane for the SB approach, I wonder if crashes were reduced after they made that change.

To be honest I'm not sure when the third lane was striped out at the Homer Watson roundabout.  Here's a youtube video posted April 2013 where the SB approach was still 3-lanes. 
https://www.youtube.com/watch?v=B6yYJYgdNTA

In my opinion this is a properly sized roundabout.  It is large enough where there is separation between legs of traffic yet small enough where drivers aren't circulating around at high speeds.  Yet there were 40 injuries in the first 3 years of operations.  One thing you notice in the video is that a pedestrian crossing the roundabout can lead to "˜gridlock'.  Perhaps the large number of crashes and injuries can be attributed to the high pedestrian volumes at this roundabout.  This is just a guess though since I haven't gone through the crash reports.  Before this roundabout was built, residents voiced concerns that a busy roundabout shouldn't be constructed next to a high school with heavy pedestrian traffic.    In October 20011, shortly after the roundabout opened, a 16-year old St. Mary's High School student was struck by a bus while she was walking in the crosswalk.   She experienced significant injuries.  In the news reports you could see the indentation her head made with the buses windshield.   This accident placed a negative light on the roundabout in the community and the city has been trying to make the roundabout safer ever since (ie. reducing a circulating lane at one of the legs, placing stopping guards during school times, etc).
   
Quote from: johndoe on March 08, 2016, 07:53:48 PMAs we've previously discussed, there are multiple levels of injuries. Injuries may have increased, but it's the severity that matters. It's ridiculous to support an intersection design that (perhaps) puts people in hospital for months, but oppose an intersection design that results in several visits to a chiropractor, simply because there's fewer injuries overall.

Roundabouts can be the site of serious, life altering accidents.  Just ask Cassi Lam who was the 16-year old girl struck at the Homer Watson roundabout.  The article I cited touches on another roundabout in the Waterloo region.  The roundabout at Hespeler Road and Queen Street "...went from 15 total collisions that hurt nine people to 86 total collisions that hurt six people. This compares its last 17 months as a traffic light to its first 17 months as a roundabout. The average severity of injury did not change and the roundabout had one more injury-causing collision than the traffic signal, although fewer people were hurt."  

PurdueBill

Things like the above are what has me wondering about the wisdom of the push for a roundabout at State Street and River Road (former IN 26 and US 231) in West Lafayette.  The push is to make State Street so much more pedestrian and bike friendly, increasing foot traffic from Chauncey Hill down to the Levee, but through a roundabout? With tipsy students heading to and from bars at night? Between multilane streets that carry significant commuter traffic to and from campus from and to Lafayette? Seems like asking for trouble.

tradephoric

It sounds like they are axing their plans to build a roundabout at State Street and River Road.   The roundabout footprint would encroach onto private property making it cost prohibitive.  Even if it wasn't cost prohibitive they shouldn't build it because it's just not the right location for a roundabout (for the reasons mentioned by PurdueBill).

http://www.purdueexponent.org/city/article_e57dcfac-825e-5df8-9ae8-08b2c8a4f343.html#comments


PurdueBill

Quote from: tradephoric on March 10, 2016, 12:54:57 PM
It sounds like they are axing their plans to build a roundabout at State Street and River Road.   The roundabout footprint would encroach onto private property making it cost prohibitive.  Even if it wasn't cost prohibitive they shouldn't build it because it's just not the right location for a roundabout (for the reasons mentioned by PurdueBill).

http://www.purdueexponent.org/city/article_e57dcfac-825e-5df8-9ae8-08b2c8a4f343.html#comments



That hopefully will stick.  Can you imagine asking McDonald's to close again after the old one was shut down and demolished for the River Road work that rerouted US 231?  There is too much pedestrian and bicycle traffic for the roundabout to work there.  It seems that West Lafayette has been going kinda wild with stuff on roads that INDOT left to them; the signals at Northwestern and Stadium were in fine working order and not that old but needed to be replaced because they weren't pretty enough--same for Northwestern and Grant.  The roundabouts at Northwestern and Yeager and at new and old SR 25 have their issues without adding pedestrians and bicycles to the mix in large quantity.

jakeroot

Why not signalize the pedestrian crossings at roundabouts? Some sort of pole-mounted HAWK setup might work well.

tradephoric

https://www.youtube.com/watch?v=2GQlwg5zfSU

The non-compliance rate at the HAWK is high especially for drivers exiting the roundabout.   A driver exiting the roundabout may not see that the HAWK signal is activated in time or they may be worried about being rear ended if they come to a stop so close to the roundabout exit.  And a lot of the drivers that do come to a stop are stopping for too long.  They treat the alternating flashing red signals like a railroad crossing (wait until the lights stop flashing).  This just leads to more gridlock and drivers backing up through the roundabout.  I find it curious that the HAWK video only demonstrates pedestrians crossing the entry leg of the roundabout.  Why not show what it looks like crossing the exit leg?  That's when you see drivers backing up through the roundabout and causing gridlock.


jakeroot

Quote from: tradephoric on March 11, 2016, 01:42:39 PM
The non-compliance rate at the HAWK is high especially for drivers exiting the roundabout.

How high is the compliance rate for a normal crosswalk? We're discussing the lesser of two evils here.

Quote from: tradephoric on March 11, 2016, 01:42:39 PM
A driver exiting the roundabout may not see that the HAWK signal is activated in time or they may be worried about being rear ended if they come to a stop so close to the roundabout exit.  And a lot of the drivers that do come to a stop are stopping for too long.  They treat the alternating flashing red signals like a railroad crossing (wait until the lights stop flashing).  This just leads to more gridlock and drivers backing up through the roundabout.  I find it curious that the HAWK video only demonstrates pedestrians crossing the entry leg of the roundabout.  Why not show what it looks like crossing the exit leg?  That's when you see drivers backing up through the roundabout and causing gridlock.

All of those things could be said about a normal crosswalk at a roundabout (backing up through the circle, not stopping because of a fear of being rear-ended, etc). I'm simply saying that, legally, crosswalks have to be provided somewhere at every intersection (or at least I'm pretty sure they are) The HAWK may have compliance issues now, but in several years, as drivers come to understand their meaning, I suspect that they'll provide a superior crossing option for roundabouts, especially those with two or more lanes. They force drivers to stop for 8-10 seconds to let pedestrians cross, and then go to a flash mode so drivers can continue, should there not be any more pedestrians.

Pedestrian safety at roundabouts seem to be related to drivers stopping right at the crosswalk, making the pedestrian invisible to subsequent lanes. The pedestrian assumes right-of-way because one car stopped, but the car in the next lane cannot see the pedestrian, and the two collide. HAWK signals force traffic to stop farther back, so visibility isn't hindered as much.

There's also an operational advantage. Rather than having traffic stop every 10-15 seconds to let one person cross, traffic stops every 60-80 seconds to let three or four people cross. Perhaps the parallel leg could also be timed so that the pedestrians can cross both the ingress and egress leg of the roundabout at the same time.

Tarkus

Both the conventional and HAWK situations with pedestrians in roundabouts are problematic, because of that exit point issue, which could potentially lock up the entire circular roadway in the right (wrong) traffic load and travel patterns.  If you're going to get into a situation where the traffic is going to have to stop within the circular roadway, you might as well have the crosswalk go across the circular roadway instead of across the approach.

Or better yet, don't build a roundabout in a spot with high pedestrian volumes, especially a multi-lane one where motorists seem to have enough difficulty processing the information to enter the circular roadway safely, without having to worry about another mode of transportation coming from yet another angle.  With the phased approach of signalized intersections, you don't have pedestrians going one direction locking up an entire intersection.

Theoretically, if you were going through all the trouble of installing a HAWK at a multi-lane roundabout like Homer Watson/Block Line, for the cost, I almost wonder if it may make as much as sense to meter the roundabout approaches like freeway onramps.  After all, the circular roadway in a roundabout is just a really miniature beltway.  It'll eliminate the decision-making of determining when it's safe to enter the multi-lane fray (the decision becomes objective rather than subjective), the red phases could be prolonged as needed when there's pedestrians without the nasty lockup issue, and the conflict points would become less . . .conflicting. 

tradephoric

Quote from: jakeroot on March 11, 2016, 02:44:21 PMThe HAWK may have compliance issues now, but in several years, as drivers come to understand their meaning, I suspect that they'll provide a superior crossing option for roundabouts, especially those with two or more lanes.

A 2011 study analyzed the performance of a Pedestrian Hybrid Beacon (PHB, also known as HAWK) at a three-lane roundabout in Oakland County, Michigan.  A link to the 115 page report is below.  The chart on page 49 lists the driver behavior at the HAWK during different signal intervals along the three-lane approach.  During the steady red / walk phase, 94.4% of drivers came to a stop at the entry leg.  However, only 68.9% of drivers came to a stop at the exit leg.  Nearly 1 in 3 drivers exiting the roundabout disregarded the solid red and drove through the crosswalk when a WALK indication was being displayed to the pedestrian.  Drivers appear to be less aware of the HAWK signal on the exit leg.  It's not necessarily that drivers don't know what to do; they just aren't reacting in time. 


http://www.rcocweb.org/Lists/Publications/Attachments/126/HAWK%20Final%20Report%202011.pdf

Ok, but would the HAWK be safer than a roundabout with no signalized crosswalks?  I would argue NO.  Pedestrians get a false sense of security when a "WALK"  indication is displayed.  Personally, my head is on a swivel when crossing the road, but I see so many pedestrians barely paying attention, just assuming it's safe to step into the crosswalk because their "WALK"  came up.  That pedestrian behavior can be dangerous when 31.1% of drivers are disregarding the solid red indication.

jakeroot

Quote from: tradephoric on March 12, 2016, 10:16:49 AM
Ok, but would the HAWK be safer than a roundabout with no signalized crosswalks?  I would argue NO.  Pedestrians get a false sense of security when a "WALK"  indication is displayed.  Personally, my head is on a swivel when crossing the road, but I see so many pedestrians barely paying attention, just assuming it's safe to step into the crosswalk because their "WALK"  came up.  That pedestrian behavior can be dangerous when 31.1% of drivers are disregarding the solid red indication.

That's for sure. I saw the RRFB in that study you linked above. That looks like a promising alternative, though the HAWK looks promising for entry legs, and I suspect exit-leg compliance may increase over time as HAWKs become more commonplace; my suspicion is that driver's have about the same reaction time to the HAWK as they do to the pedestrians themselves. So, I don't think there's any harm in installing HAWK signals. After all, the alternative is signalized junction with the same WALK/DON'T WALK signals, and about the same number of people who are either swiveling their heads around, or not looking at all.

We really ought to consider Pelican phasing, like in the UK. They use the standard R/Y/G signal setup, with solid red during the WALK phase, and flashing amber during the DON'T WALK phase. If the iffy compliance with HAWK signals has anything to do with failure to understand the signals themselves, perhaps compliance would go up when people meet a standard traffic light.

https://www.youtube.com/watch?v=aI3FiOf4rJI

Though I will agree that roundabouts are not a particularly good idea in urban areas (though one-lane roundabouts likely work fine in both urban and rural areas alike), we still have to consider the safety of pedestrians even in the most remote of roundabouts. Unless we plan on dumping them altogether (a bad idea), we need to find a way to make the pedestrian crossings at roundabouts safer.

In terms of the ADA, they may prefer signalized crossings at roundabouts. Those who are blind have trouble telling when it's safe to cross. The Pelican crossing above has a chirping noise for the WALK phase, and may be helpful for those individuals.

english si

^^ there's also a spinning cone under the box with the button as an added feature for sensory impaired people.

jakeroot

Quote from: english si on March 12, 2016, 07:17:31 PM
^^ there's also a spinning cone under the box with the button as an added feature for sensory impaired people.

Tom Scott demonstrated that feature in one of his videos. Genius idea..

http://youtu.be/fRPWlBAgSnw

tradephoric

Another driver was killed at a roundabout in Carmel.   Of the roughly 100 roundabouts in Carmel, three have been the site of a fatal crash since 2014.  Here are news reports of the fatal crashes that have occurred:

Fatal crashes at Carmel roundabouts:
http://fox59.com/2014/08/06/one-dead-in-carmel-accident-passenger-taken-to-hospital/
http://fox59.com/2014/09/26/person-dead-after-motorcycle-accident-in-carmel/
http://www.indystar.com/story/news/crime/2016/03/18/driver-killed-carmel-roundabout-crash/81967984/

jakeroot

Quote from: tradephoric on March 21, 2016, 02:34:43 PM
Another driver was killed at a roundabout in Carmel....http://www.indystar.com/story/news/crime/2016/03/18/driver-killed-carmel-roundabout-crash/81967984/

If distraction is at play here, he could have just as easily run a stop sign or signal. At least he hit the central concrete decorations instead of another car.

paulthemapguy

I was at that pair of Hilliard roundabouts just last month!  People in Ohio have no idea how to use them, apparently.  It was chaos!  I thought a yield sign at each and every approach would make them pretty self-explanatory...nope!  And if this is kind of like a big 6-way intersection, why not install one big roundabout to accommodate all 3 through-roads.

(sidenote:  there's a really sweet comic/games shop called Packrat Comics right near these roundabouts.  That's why I ended up in that area XD)
Avatar is the last interesting highway I clinched.
My website! http://www.paulacrossamerica.com Now featuring all of Ohio!
My USA Shield Gallery https://flic.kr/s/aHsmHwJRZk
TM Clinches https://bit.ly/2UwRs4O

National collection status: 361/425. Only 64 route markers remain

Tarkus

#369
Quote from: tradephoric on March 21, 2016, 02:34:43 PM
Another driver was killed at a roundabout in Carmel.   Of the roughly 100 roundabouts in Carmel, three have been the site of a fatal crash since 2014.  Here are news reports of the fatal crashes that have occurred:

Fatal crashes at Carmel roundabouts:
http://fox59.com/2014/08/06/one-dead-in-carmel-accident-passenger-taken-to-hospital/
http://fox59.com/2014/09/26/person-dead-after-motorcycle-accident-in-carmel/
http://www.indystar.com/story/news/crime/2016/03/18/driver-killed-carmel-roundabout-crash/81967984/

The interesting thing to note about that latest Carmel crash is that it occurred at the Westfield Blvd and E 96th St roundabout, which is a three-leg single-lane roundabout with one right turn bypass from Westfield NB to 96th EB.  Based on a quickie measurement on Google Maps, the center island is about 120 ft. in diameter.  The crash did occur on St. Patrick's Day at night, so one can't rule out impairment, but there's no indication of that yet in the news reports.

Quote from: jakeroot on March 21, 2016, 03:52:13 PM
If distraction is at play here, he could have just as easily run a stop sign or signal. At least he hit the central concrete decorations instead of another car.

Considering that he got all the way to the middle of the roundabout, I'd say the type of intersection is moot.  If the impact with the central island was severe enough to kill him, reason stands that had there been another car in the roundabout there, it could have been just as ugly as a wreck at a stop or signal-controlled intersection.  There's not much deflection at the Westfield/96th roundabout--this also appears to be the case with the 106th/Pennsylvania and 126th/Hazel Dell roundabouts, which were the sites of the previous accidents.

It's also worth noting that in every case, we're looking at single-vehicle fatal fixed object collisions caused by hitting curbing, and resulting in particularly violent impacts.  I'd be curious to see the stats for these three roundabouts in question.

Edit: Doing a little research (typed in 96th and Westfield accident Carmel), and found there was another fatal accident at this same roundabout in 2007, which killed two.  The car was found in the center island.  Another news story has police saying that the driver was drunk and entered the roundabout going "at least 62mph".

tradephoric

There are roughly 100 modern roundabouts in Carmel, Indiana with 4 known fatal roundabout crashes occurring since their inception.  So how does that compare to other parts of the country?  I decided to look at the fatalities that have occurred at the 100 most crash prone intersections in SE Michigan.  I chose SE Michigan because it has publically accessible data for every intersection in the region dating back a decade.  Of the 100 most crash prone intersections in SE Michigan, there have been 29 fatal crashes (23% of the intersections were the site of a fatal crash and 4% had multiple fatal crashes).


http://semcog.org/data-and-maps/high-frequency-crash-locations

The numbers initially look pretty good for the roundabouts.  There were only 4 fatal crashes at the 100 Carmel roundabouts compared to 29 fatal crashes at the 100 most crash prone intersections in SE Michigan.  This equates to an 86% reduction in fatal crashes.  However, there are some important factors to consider:

#1.  There are roughly 100 roundabouts in Carmel, but back in 2005 - the first year crash data was analyzed for the intersections in SE Michigan - there were only about a half dozen roundabouts in Carmel.  The roundabout craze hadn't taken off yet.   This doesn't make it a fair comparison and underestimates the fatal crashes that may have occurred in Carmel had there been 100 roundabouts back in 2005. 

#2.   Carmel has a population of roughly 85,000 people whereas SE Michigan has a population of 4.7 million people.  Analyzing the top 100 crash prone intersections in a region of 4.7 million people is going to be weighted towards high volume intersections whereas many of the Carmel roundabouts are single-lane with relatively low traffic volumes.  Obviously, intersections with 70,000 vehicles per day (ie. major SE Michigan intersections) are more likely to see a fatal crash than an intersection with 15,000 vehicles per day (ie. single-lane Carmel roundabouts). 

#3.  I don't have an accurate database of the Carmel crashes and am relying on news articles to estimate how many fatal roundabout crashes have actually occurred over the past 10 years.  Potentially, the fatal crashes that have occurred inside Carmel roundabouts are being underestimated.

Based on these 3 factors, the 86% reduction in fatal crashes isn't as good as it sounds.  There is one more caveat though.  This is looking at fatal crashes and not total fatalities.  When looking at total fatalities, the numbers might look better for the Carmel roundabouts (since a fatal crash at an intersection may be more prone to multiple fatalities).  I just didn't take the time to parse through the crash forms to see how many fatalities occurred in each of the 29 fatal crashes.  This can be easily done though.

tradephoric

Let's look at it this way... Over the past 27 months, the 100 roundabouts in Carmel has been the site of 3 fatal crashes.  That equates to an average of 1.33 fatal crashes per year.  The top 100 crash prone SE Michigan intersections have seen on average 2.9 fatal crashes per year. 

(1.33/2.9) -1 = -54%   

Factor #1 is dealt with in this shorter 27 month analysis since we are comparing crash data from roughly 100 roundabouts to 100 SE Michigan intersections.  But we are getting farther and farther away from the 90% reduction in fatal crashes.  The Carmel roundabouts are now only seeing a 54% reduction in fatal crashes.  And we still haven't dealt with Factor #2.  The traffic volumes at Carmel roundabouts aren't as high as the intersections in SE Michigan.  Many of the SE Michigan intersections have traffic volumes double and triple that of the single-lane roundabouts found in Carmel.  If you factor in the traffic volumes, the reduction in fatalities at Carmel roundabouts just hasn't been that impressive.

It's been alluded to that the majority of people dying at roundabouts are drunk.  The safety numbers stated in the literature is that roundabouts reduce fatalities by 90%.  That is the number I'm focusing on and it includes alcohol related fatal crashes.  You can't just ignore a fatality because somebody was drunk.

lordsutch

Heck, if *only* half as many people are dying instead of a tenth as many people, let's just tear all the roundabouts out and let the carnage recommence.

tradephoric

Quote from: lordsutch on March 22, 2016, 10:08:58 PM
Heck, if *only* half as many people are dying instead of a tenth as many people, let's just tear all the roundabouts out and let the carnage recommence.

Having 50% less fatalities in Carmel isn't that impressive when the typical Carmel roundabout has 50% less traffic though.  Over the past 27 months, the fatality rate at Carmel roundabouts has been similar to the fatality rate at non-roundabout communities.  Carmel has had a bad run over the past 27 months when it comes to fatalities at roundabouts. 

jakeroot

Quote from: tradephoric on March 23, 2016, 07:30:58 AM
Quote from: lordsutch on March 22, 2016, 10:08:58 PM
Heck, if *only* half as many people are dying instead of a tenth as many people, let's just tear all the roundabouts out and let the carnage recommence.

Having 50% less fatalities in Carmel isn't that impressive when the typical Carmel roundabout has 50% less traffic though.  Over the past 27 months, the fatality rate at Carmel roundabouts has been similar to the fatality rate at non-roundabout communities.  Carmel has had a bad run over the past 27 months when it comes to fatalities at roundabouts.

Have you compared Carmel's crash stats to other cities of the same population?



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.