Crash prone 'modern roundabouts'

[kalvado]

If the IIHS was smart, they'd start a new study with different variables and at different speed limits.

IIHS (like everyone else) also have their agenda. And it is not clear for me what drives their logic in different situations. For example, they are adamantly opposed to any speed limit increases or higher speeds on the roads - including support for automated speed enforcement. However it appears matter is far from black and white, and such adamant position doesn't mean advocating safety (but does mean advocating higher premiums - they are insurance institute, after all). Their research team is psychology-heavy, which makes some sense; and crash tests and related issues are a big part of their work. I am really not sure if traffic control is their strong point...

[tradephoric]

https://www.youtube.com/watch?v=u4oltdv0i9w
^Apart from the RV blowing through the middle of that Worthington Minnesota roundabout, there has been two other pretty major events at the roundabout over the past couple weeks.

Tanker hauling cream overturns in Worthington roundabout

http://www.dglobe.com/news/accidents/4356792-tanker-hauling-cream-overturns-worthington-roundabout


These little piggies almost didn't make it to market when a Minn. hog truck rolled

http://www.agweek.com/news/4350581-these-little-piggies-almost-didnt-make-it-market-when-minn-hog-truck-rolled

Cream trucks rolling over, hogs running around the highway, some RV driver thinking they are Bo Duke... these roundabouts really give the good people of Worthington Minnesota something to talk about!

Another vehicle vaulted over the Worthington roundabout on Sunday.  The driver sustained non-life threatening injuries.  This comes a month after an RV was videotaped launching over the roundabout in early November.

Minivan vaults over roundabout Sunday
https://www.dglobe.com/news/accidents/4375814-minivan-vaults-over-roundabout-sunday

There have been 2 truck tip-over crashes, an RV flying through the middle of the roundabout captured on video, a driver sustaining non life-sustaining injuries after his min-van flew through the middle of the roundabout, and now 4 reported crashes over a few hour period on Friday.  The roundabouts are doing a stellar job at reducing crashes in Worthington.. or maybe they are doing a worthless job.

Snowpacked roundabouts result in four crashes reported Friday
https://www.dglobe.com/news/4380839-snowpacked-roundabouts-result-four-crashes-reported-friday

[billpa]

https://www.youtube.com/watch?v=u4oltdv0i9w
^Apart from the RV blowing through the middle of that Worthington Minnesota roundabout, there has been two other pretty major events at the roundabout over the past couple weeks.

Tanker hauling cream overturns in Worthington roundabout

http://www.dglobe.com/news/accidents/4356792-tanker-hauling-cream-overturns-worthington-roundabout


These little piggies almost didn't make it to market when a Minn. hog truck rolled

http://www.agweek.com/news/4350581-these-little-piggies-almost-didnt-make-it-market-when-minn-hog-truck-rolled

Cream trucks rolling over, hogs running around the highway, some RV driver thinking they are Bo Duke... these roundabouts really give the good people of Worthington Minnesota something to talk about!

Another vehicle vaulted over the Worthington roundabout on Sunday.  The driver sustained non-life threatening injuries.  This comes a month after an RV was videotaped launching over the roundabout in early November.

Minivan vaults over roundabout Sunday
https://www.dglobe.com/news/accidents/4375814-minivan-vaults-over-roundabout-sunday

There have been 2 truck tip-over crashes, an RV flying through the middle of the roundabout captured on video, a driver sustaining non life-sustaining injuries after his min-van flew through the middle of the roundabout, and now 4 reported crashes over a few hour period on Friday.  The roundabouts are doing a stellar job at reducing crashes in Worthington.. or maybe they are doing a worthless job.

Snowpacked roundabouts result in four crashes reported Friday
https://www.dglobe.com/news/4380839-snowpacked-roundabouts-result-four-crashes-reported-friday

Here's another crash.

https://youtu.be/tUYl1FqeBlA

SM-T230NU

[jeffandnicole]

Here's another crash.

https://youtu.be/tUYl1FqeBlA

SM-T230NU

OK...but this has nothing to do with a roundabout.

[billpa]

No, it has to do with showing random crashes that prove absolutely nothing about the engineering of an intersection.

SM-T230NU

[jeffandnicole]

No, it has to do with showing random crashes that prove absolutely nothing about the engineering of an intersection.

SM-T230NU

Very true.

[MNHighwayMan]

snip

I don't think the idiocy of morons is enough to rule against the installation of roundabouts.

[kalvado]

snip

I don't think the idiocy of morons is enough to rule against the installation of roundabouts.

Of course not. But when those morons have jobs in DOT and "certified engineer" seals, things become a bit worse.

[MNHighwayMan]

snip

I don't think the idiocy of morons is enough to rule against the installation of roundabouts.

Of course not. But when those morons have jobs in DOT and "certified engineer" seals, things become a bit worse.

Sure. Roundabouts get installed in the wrong places. I'll give you that, but it's the same with any traffic control device: it gets used and misused. What tradephoric argues is way over the line (not to mention cherry-picked) with what reality brings.

[DaBigE]

If the IIHS was smart, they'd start a new study with different variables and at different speed limits.

IIHS (like everyone else) also have their agenda. And it is not clear for me what drives their logic in different situations. For example, they are adamantly opposed to any speed limit increases or higher speeds on the roads - including support for automated speed enforcement. However it appears matter is far from black and white, and such adamant position doesn't mean advocating safety (but does mean advocating higher premiums - they are insurance institute, after all). Their research team is psychology-heavy, which makes some sense; and crash tests and related issues are a big part of their work. I am really not sure if traffic control is their strong point...

Seems like they follow a pattern. IIHS is opposed to anything that increases speed. Higher speeds = greater risk of death. They're all about anything that lowers high-cost risks. High-speed roadways = higher insurance payouts. Similar with roundabouts. It takes a lot of PDO crashes to equal one fatality payout.

[tradephoric]

Sure. Roundabouts get installed in the wrong places. I'll give you that, but it's the same with any traffic control device: it gets used and misused. What tradephoric argues is way over the line (not to mention cherry-picked) with what reality brings.

Roundabout proponents say that roundabouts reduce injury crashes by 76% but that just doesn't apply when looking at the crash statistics of complex 2x2 roundabouts.  That's the reality.  A large sample size of complex roundabouts in Minnesota found that there was a 6% increase in injury crashes... a lot different than a 76% reduction.  A lot of mindless people will continue to believe that roundabouts reduce injuries at these complex roundabouts simply because that is what the IIHS told us 17 years ago.  But how accurate is that IIHS study?  Of the 9 multi-lane roundabouts analyzed, 4 of them didn't even have before injury crash data to analyze.  The 5 remaining double-lane roundabouts with before/after injury crash data were all along a 2500 foot section of Avon Road in Avon, Colorado.  Good thing the "national" IIHS roundabout study determined the safety statistics of double-lane roundabouts in America based on a 2500 foot section of Avon Road in Colorado (signed for 25 mph).  GMAFB.

[MNHighwayMan]

Sure. Roundabouts get installed in the wrong places. I'll give you that, but it's the same with any traffic control device: it gets used and misused. What tradephoric argues is way over the line (not to mention cherry-picked) with what reality brings.

Roundabout proponents say that roundabouts reduce injury crashes by 76% but that just doesn't apply when looking at the crash statistics of complex 2x2 roundabouts.  That's the reality.  A large sample size of complex roundabouts in Minnesota found that there was a 6% increase in injury crashes... a lot different than a 76% reduction.  A lot of mindless people will continue to believe that roundabouts reduce injuries at these complex roundabouts simply because that is what the IIHS told us 17 years ago.  But how accurate is that IIHS study?  Of the 9 multi-lane roundabouts analyzed, 4 of them didn't even have before injury crash data to analyze.  The 5 remaining double-lane roundabouts with before/after injury crash data were all along a 2500 foot section of Avon Road in Avon, Colorado.  Good thing the "national" IIHS roundabout study determined the safety statistics of double-lane roundabouts in America based on a 2500 foot section of Avon Road in Colorado (signed for 25 mph).  GMAFB.

Well, good thing you simplified it for me! I'm clearly too stupid to figure it out on my own without your own complex analysis making it simple for me to understand!

[jeffandnicole]

Sure. Roundabouts get installed in the wrong places. I'll give you that, but it's the same with any traffic control device: it gets used and misused. What tradephoric argues is way over the line (not to mention cherry-picked) with what reality brings.

Roundabout proponents say that roundabouts reduce injury crashes by 76% but that just doesn't apply when looking at the crash statistics of complex 2x2 roundabouts.  That's the reality.  A large sample size of complex roundabouts in Minnesota found that there was a 6% increase in injury crashes... a lot different than a 76% reduction.  A lot of mindless people will continue to believe that roundabouts reduce injuries at these complex roundabouts simply because that is what the IIHS told us 17 years ago.  But how accurate is that IIHS study?  Of the 9 multi-lane roundabouts analyzed, 4 of them didn't even have before injury crash data to analyze.  The 5 remaining double-lane roundabouts with before/after injury crash data were all along a 2500 foot section of Avon Road in Avon, Colorado.  Good thing the "national" IIHS roundabout study determined the safety statistics of double-lane roundabouts in America based on a 2500 foot section of Avon Road in Colorado (signed for 25 mph).  GMAFB.

You complain about a national study consisting of limited data in Colorado, then argue that against your national study consisting of limited data in Minnesota.

[tradephoric]

^I understand the point you are making J&N.  But the Minnesota research analyzes 3X more complex roundabouts than the IIHS study.  Also, the Minnesota research analyzes complex roundabouts on roadways with higher than 25 mph speed limits.  Not to mention the Minnesota research isn't analyzing crash data from the previous century.  And the Minnesota study does exactly what it claims.. it analyzes roundabouts in the state of Minnesota.  I'm not going to fault Minnesota for not analyzing double-lane roundabouts in Texas.  OTOH, the national IIHS study only analyzes double-lane roundabouts in Colorado.  Unless if we are living in the United States of Colorado, that doesn't sound too national to me.

[kalvado]

snip

I don't think the idiocy of morons is enough to rule against the installation of roundabouts.

Of course not. But when those morons have jobs in DOT and "certified engineer" seals, things become a bit worse.

Sure. Roundabouts get installed in the wrong places. I'll give you that, but it's the same with any traffic control device: it gets used and misused. What tradephoric argues is way over the line (not to mention cherry-picked) with what reality brings.

Yes, and you don't want to give complex tool to morons. They - at least at NYSDOT - struggle with regular traffic lights, and fail at synchronized traffic light stage.
Roundabouts are complex devices which need to be used with some understanding - which is currently missing. You end up with wrong people using wrong kind of stuff - and being perfectly happy about results.

[kphoger]

Roundabouts are complex devices

Just the geometry, really.

[kalvado]

Roundabouts are complex devices

Just the geometry, really.

Sure, just the geometry. But many geometrical problems are more difficult that they sound. Classic ones - squaring the circle, doubling the cube - are plainly unsolvable..

[MNHighwayMan]

Roundabouts are complex devices

Just the geometry, really.

Classic ones - squaring the circle

Isn't that what we're doing with roundabouts?

[kalvado]

Roundabouts are complex devices

Just the geometry, really.

Classic ones - squaring the circle

Isn't that what we're doing with roundabouts?

Sure, that is called one-way streets... Look at NYC, downtown manhattan is a bunch of squared and packed, traffic light controlled roundabouts...

[jakeroot]

In the 1950s, the City of Edmonton, Alberta built 12, two-lane traffic circles along major arterial roads. Seven were subsequently replaced with intersections (two with grade-separated junctions) but five remain, mostly untouched (except for modern markings). Lucky for us, the city compiles their crash data throughout the year, and puts it into a report (read here). Here are the top ten intersections in the city for number of collisions:

1) 107 Ave @ 142 St (a traffic circle) -- 134
2) Yellowhead Trail @ 127 St -- 83
3) Yellowhead Trail @ 149 St -- 76
4) Yellowhead Trail @ 121 St -- 72
5) 118 Ave @ Groat Road (a signalised, squared-off traffic circle -- see below) -- 62
6) 34 Ave @ Calgary Trail (NB) -- 60
7) 167 Ave @ 97 St -- 58
8) 34 Ave @ 91 St -- 58
9) 137 Ave @ 97 St -- 57
10) 34 Ave @ Calgary Trail (SB) -- 56

Of the 23,139 collisions in the city in 2016, the top causes of collisions were following too closely (38.6%, 8,928 collisions); struck parked vehicle (13.0%, 3,019); changing lanes improperly (10.8%, 2,497); left turn across path (6.9%, 1,593); and ran off road (6.4%, 1,483).

Interestingly, only two of the original five traffic circles in the city make the top ten. The rest are all (what I assume to be) very high-volume signalised intersections. Though it's possible the other three traffic circles make up spots 11-13.

Number 5 on the list above is a signalised, squared-off traffic circle (https://goo.gl/BLK3vM). There's no way to know what's causing the collisions at this intersection, nor what the numbers were prior to signalisation, but it still doesn't seem to be performing that well, even with the decidedly square layout.

[tradephoric]

^Good info Jake.  By far the most crash prone intersection in Edmonton is the un-signalized 2x2 traffic circle at 107 Ave & 142 Street.  It experienced 134 crashes in 2016, a full 51 more crashes than the next crash prone intersection in the city.  Edmonton drivers have had 50 years to learn how to drive through the traffic circle, and it still experienced well over 100 crashes in 2016.  Here is GE imagery of the #1 and #2 most crash prone intersections in Edmonton:

#1: 132 crashes (107 @ 142 St) traffic circle: 
https://www.google.com/maps/@53.5510288,-113.5656331,265m/data=!3m1!1e3

#2: 83 crashes (Yellowhead Trail @ 127 St) signalized intersection: 
https://www.google.com/maps/@53.5807075,-113.5412952,229m/data=!3m1!1e3



Regards to 118 Ave & Groat Rd, the squared off corners makes it resemble a Town Center Intersection to me.  However, it's a very tight TCI and the curves through the intersection may be leading to increases in crashes.  Of course 62 crashes is a lot less than 134 crashes.... 

[tradephoric]

Seems like they follow a pattern. IIHS is opposed to anything that increases speed. Higher speeds = greater risk of death. They're all about anything that lowers high-cost risks. High-speed roadways = higher insurance payouts. Similar with roundabouts. It takes a lot of PDO crashes to equal one fatality payout.

If we use FHWA's comprehensive crash costs by injury severity level, the social costs of a fatal crash is estimated to be $4,008,900 and the cost of a PDO crash is $7,400.  It would take 541 PDO crashes to equal the cost of 1 fatal crash.  But what we aren't considering is the probability of a fatal crash at an intersection.  The FHWA reported there were 2,924 fatal crashes at the roughly 300,000 signalized intersections in 2007.  Extrapolating those numbers, the average signalized intersection in America averages a fatal crash about every 100 years.  Looking at a more localized example, the top 20 most crash prone intersections in SE Michigan experienced 3 fatalities over the past 10 years.  Extrapolating those numbers, that equates to a fatality every 66.7 years (200 years of crash data / 3 fatalities).   Let's assume a signalized intersection averages a fatal crash every 50 years.  The average "fatal"  payout per year would be $4,008,900/50 years... or $80,178.  Now it only takes 11 PDO crashes at $7,400 a pop to equal the average "fatal"  crash payout. 

When doing a cost/benefit analysis for the 2x2 roundabouts analyzed in the Minnesota study, the 2x2 roundabouts have a higher crash cost than the signalized intersections they replaced (by $1,664,300).  Remember it only takes 11 PDO crashes to equal the average "fatal" crash payout, yet the roundabouts in Minnesota had a whopping 270 more PDO crashes.

MnDOT releases a detailed study claiming massive decreases in both fatal and non-fatal incidents at intersections where roundabouts have been installed (this includes both state and non-state-built projects).

http://www.dot.state.mn.us/trafficeng/safety/docs/roundaboutstudy.pdf

There were 6 full dual-lane roundabouts analyzed in that Minnesota study (ie. 2x2 roundabouts).  Here is a table of the dual-lane roundabouts with before construction and after construction crash data based on crash severity:



If we use FHWA's comprehensive crash costs by injury severity level to perform a cost/benefit analysis, we see that these dual-lane Minnesota roundabouts increased crash costs by $1,664,300.  Even though there were 3 less A-level crashes at the roundabouts, there were 7 more C-level crashes and a whopping 270 more PDO crashes.

BEFORE ROUNDABOUT CRASH COSTS = $4,882,600
Fatality (K)           0 X $4,008,900 = $0
Disabling Injury (A)   3 X  $216,000 = $648,000
Evident Injury (B)   15 X $79,000 = $1,185,000
Possible Injury (C)   46 X $44,900 = $2,065,400
PDO (O)          133 X $7,400 =  $984,200

AFTER ROUNDABOUT CRASH COSTS = $6,546,900
Fatality (K)           0 X $4,008,900 = $0
Disabling Injury (A)   0 X  $216,000 = $0
Evident Injury (B)   15 X $79,000 = $1,185,000
Possible Injury (C)   53 X $44,900 = $2,379,700
PDO (O)          403 X $7,400 =  $2,982,200

[tradephoric]

News reports for fatal roundabout crashes that occurred in 2017. 

http://www.sun-sentinel.com/local/broward/fl-boca-man-dies-in-parkland-crash-20170117-story.html
http://www.theledger.com/news/20170208/teen-faces-manslaughter-charge-after-girl-dies-in-crash-following-chase-with-lpd-officer
http://www.wdbj7.com/content/news/60-year-old-Roanoke-County-woman-dies-in-motorcycle-crash-413537653.html
http://www.fox9.com/news/unbelted-man-dies-in-chisago-county-crash-friday-afternoon
http://www.wcsh6.com/news/local/fatal-crash-closes-rte-114-bypass-in-gorham/433810393
http://www.toledoblade.com/Police-Fire/2017/04/30/Alcohol-believed-to-be-factor-in-fatal-Springfield-Twp-crash.html
http://www.gainesville.com/news/20170502/roundabout-crash-kills-two-police-say
http://www.bellinghamherald.com/news/local/article153269789.html
http://www.cecildaily.com/police_and_fire_beat/article_bed6099c-45a6-5be5-b61d-ec91d893631b.html
http://www.wpta21.com/story/35827071/coroner-identifies-motorcyclist-involved-in-thursday-morning-fatal-crash
http://www.prescottenews.com/index.php/news/current-news/item/30613-37-year-old-man-from-page-dies-in-crash-at-roundabout
http://www.startribune.com/driver-in-st-cloud-roundabout-fatally-hits-pedestrian/453113583/
http://www.lasvegasnow.com/news/one-killed-in-overnight-crash/859083995

This was based on about an hour of google searches and based on this limited research, there were at least 13 fatal roundabouts crashes in 2017 resulting in 14 fatalities.   According to the latest 2013 FHWA estimates, there are roughly 3200 roundabouts in the United States.  I personally have compiled a database of over 5200 roundabouts in America, but I don't know the exact methodology used to compile the FHWA database.  Let's split the difference and say there are about 4200 roundabouts in America.   In 2017 each modern roundabout averaged 0.0033 fatal crashes (14 fatalities / 4200 roundabouts).  This ratio is comparable to previous years based on data compiled by the FHWA:

https://safety.fhwa.dot.gov/intersection/innovative/roundabouts/docs/fhwasa15072.pdf

So how does this ratio compare to signalized intersections?  Based on FHWA data there were 2,924 fatalities at approximately 311,000 signalized intersections in 2007.  Each signal averaged 0.0094 fatal crashes (2,924 fatalities / 311,000 signals). 
https://safety.fhwa.dot.gov/intersection/other_topics/fhwasa10005/brief_2.cfm

Now we can use these ratios to get a sense of how much roundabouts are reducing fatal crashes.  Is it really 90% as the IIHS claims?  To me it looks to be closer to 65% (1-(0.0033/0.0094)).  That's still a big reduction, but it's not 90%.  And consider that the IIHS claims that roundabouts reduce injury crashes by 76%.  However, Minnesota research found that complex roundabouts in that state actually saw a 6% increase in injury crashes.  The point is there is potentially a subset of modern roundabouts that are increasing fatal crashes, not reducing them.  I'll tell you one subset of roundabouts that are potentially increasing fatal crashes... roundabouts that have retaining walls in the central island.  Do you think it's a coincidence there have been 3 fatalities in the past 10 years at the 96th and Westfield Blvd roundabout in Carmel?  Every time a driver misjudges the roundabout and drive through the middle of it, they smack into a concrete wall.  Boom!  Dead!

[cjw2001]

Stop blaming the roundabout and start blaming the person driving drunk at high speed.  The Carmel roundabouts have been a wonderful improvement to traffic flow in the area.   A driver going far in excess of speed limits and driving impaired is not an indictment of the roundabout, it's a failure of personal responsibility.  Get over it.

[jakeroot]

Stop blaming the roundabout and start blaming the person driving drunk at high speed.  The Carmel roundabouts have been a wonderful improvement to traffic flow in the area.   A driver going far in excess of speed limits and driving impaired is not an indictment of the roundabout, it's a failure of personal responsibility.  Get over it.

While that's true, and I largely agree that personal responsibility is the problem, we (as a society) have a responsibility to create an environment whereby screwing up doesn't kill you. Roundabouts are great, as long as everyone is looking at the road, and no one has any health problems. But, we both know that both of those things aren't the reality. So, we shouldn't be creating an environment where these things, combined with the operation of a motor vehicle, create a dangerous situation.