Sometimes, being a roadgeek hurts

[route56]

[NOTE TO ADMIN: This could possibly belong in the Central States or even the off-topic forum]

http://www2.ljworld.com/news/2011/apr/16/multiple-helicopters-being-dispatched-accident-k-1/

On April 16, there was a crossover crash on the K-10 freeway near Eudora. Two people, including a 5 year old boy, were killed in the wreck.

I never met the little boy, but I do know his mother.

In the comments section of the Journal-World forums, there are many commenters who say there have been too many crossover collisions on K-10, and that there needs to be cable barriers. (one person even suggested there needs to be fewer drivers - seriously)

This isn't the first time someone I know has lost a friend/loved one in a car crash, and I know it won't be the last. However, although I have seen where accidents can lead to change, I also know that sometimes it's better to take a step back. There have been multiple stories of crossover collisions and statements saying that K-10 is a dangerous highway; however, I do not know if that perception is a reality - and if I find that it's not, the families involved may take it as insensitive.

Still, no matter what, my friend has lost her son, and there's nothing anybody can do to change that.

[NE2]

Not to be callous, but I don't see where being a roadgeek comes in...

[J N Winkler]

Not to be callous, but I don't see where being a roadgeek comes in...

I think it is probably this paragraph:

This isn't the first time someone I know has lost a friend/loved one in a car crash, and I know it won't be the last. However, although I have seen where accidents can lead to change, I also know that sometimes it's better to take a step back. There have been multiple stories of crossover collisions and statements saying that K-10 is a dangerous highway; however, I do not know if that perception is a reality - and if I find that it's not, the families involved may take it as insensitive.

For what it is worth, my personal view is that any economic evaluation of provision of median cable barrier on K-10 (or, for that matter, any divided state highway in Kansas) needs to be separated from the grief associated with individual incidents.  For the family that lost the six-year-old girl, that loss is the fact that is most important to them, and their grief has to be respected for its own sake.  I don't think they even need to have the pros and cons of median cable barrier explained to them.

My understanding is that the Green Book is currently under revision and AASHTO is moving toward fixing 60' as the minimum width (ETW-to-ETW) of a median for which barriers are not required to be provided.  ISTR that it is also KDOT's intention to provide median cable barrier at least on all Interstates built to the old standard width of 45'.  For this reason I expect soon to see median cable barrier on I-235 in Wichita, which has had serious crossover accidents, though not frequently enough to become notable for them.

I don't know that KDOT has any cutoffs for AADT, crossover crash incidence, or crossover crash rate which would result in some lengths of divided highway not being equipped with median cable barrier.  An economic case can be made for such cutoffs but I rather suspect there aren't any.  In civil engineering in general, and highway engineering in particular, there is often a tradeoff between building to a consistent standard and subjecting every aspect of a particular design to cost-benefit analysis.

[Duke87]

Alright, well... it clearly looks dangerous. That median is too flat and too narrow to not have a barrier of any sort.

[Michael]

I looked at the highway in Street View, and the only thing that stands out to me is the narrow right shoulder.  I went to an overpass near Duke87's link, and the median doesn't look much narrower than the NY Thruway to me.  Are the crossovers No U-Turn areas or allowed crossovers?  Also, is it 55 or 65 MPH in that area?

P.S.: I was expecting to see something about people watching the cars go by based on the title of this thread.

[route56]

Also, is it 55 or 65 MPH in that area?

Actually, the speed limit is 70 MPH. This is Kansas, after all.

Also, an update on the accident has just been posted:

http://www2.ljworld.com/news/2011/apr/18/kansas-highway-patrol-investigates-whether-marijua/

An autopsy on the other fatality, the driver of the vehicle that crossed the median, indicates the presence of marijuana in his system. Also, his license may have been suspended/expired.

(Disengaging logic mode, returning to grief mode)

[J N Winkler]

It is worth noting that even when a median barrier is present, it is still possible to have crossover fatalities.  There was an incident in Wichita about two months ago when two men (both of whom had histories of license suspensions due to DUI, and one of whom did not have a valid license) got into a drag race on westbound Kellogg between Seneca and Meridian.  Kellogg at this point is a full freeway with a center Jersey barrier and a 60 MPH speed limit.  One of the vehicles involved in the drag race managed to jump the center barrier, into the eastbound lanes, where it ran head-on into a car carrying a woman and her young daughter.  The daughter (who was secured in a child seat) was killed instantly and the mother was taken to the hospital in critical condition.

Reform of the DUI laws in Kansas has been under study for a couple of years now, and a blue-ribbon commission has sent its report with recommendations to the Legislature.  Action has been postponed, however, because with the ongoing budget difficulties, the Legislature does not want to create fresh demand for prison beds.  Personally, I support penal regimes focused on rehabilitation rather than punishment, but I firmly believe that it needs to become much more difficult for convicted DUI offenders to gain access to vehicles they can drive while drunk, even if that means confiscating their vehicles and selling them at auction.

[hbelkins]

I understand where you are coming from. Seems that anytime there is some sort of tragedy, no matter how isolated it may be, there is always a cry of "We gotta DO SOMETHING!!!!"

Case in point: Here in eastern Kentucky there is a narrow stretch of KY 451 in Perry County that is squeezed in between a river and the railroad. This route carries only about 400 VPD and is a local road. Really, the only people who are going to be traveling this route are those who live along it. Last year, a vehicle overturned and slid into the river and the driver died. This is the only fatality accident that has been recorded in this area, but residents started a clamor to "DO SOMETHING!!!" and erroneously made the claim that this route has a high accident history. The newly elected state representative took this up as a pet project, demanding that guardrail be installed. This section of roadway ranked very low in our guardrail program due to low traffic counts, lack of accident history, etc. Yet the residents and the politicians persisted. This wasn't a typical guardrail installation, either. Because there was no shoulder, we had to drill steel and backfill to create a shoulder so we could install guardrail. Despite this not meeting any criteria for immediate work, and ranking very low in the guardrail program, the politician succeeded in having the transportation secretary allocate $73,000 in discretionary funding to install guardrail. IMHO this was a waste of money. Yes, the fatal wreck was tragic, but the response was overblown. You can't go around installing guardrails or cable barriers at the site of every fatal wreck.

[myosh_tino]

I understand where you are coming from. Seems that anytime there is some sort of tragedy, no matter how isolated it may be, there is always a cry of "We gotta DO SOMETHING!!!!"

There was a similar situation when the CA-85 freeway was built through Cupertino, Saratoga and San Jose back in 1994.  Initially there was no center median barrier (guardrail or jersey barrier) because Caltrans only required a barrier if the median was less than 45 feet wide and CA-85's median was 46-50 feet wide (I can't say if this was done intentionally to cut costs but it does look a little fishy IMO).  Within the first year, one person died in a head-on collision and immediately there was talk of why there was no median barrier on this urban freeway. Nothing was done until 1997 when 6 people died in head-on collisions in a single year and Caltrans was forced to re-evaluate their policy on median barriers.

The result was the installation of a post-and-metal (essentially a double-sided guardrail) barrier which is standard on most California freeways.  Caltrans also amended their median barrier policy for high-volume roads by requiring medians less than 75 feet wide to have a barrier of some type installed.

[oscar]

Case in point: Here in eastern Kentucky there is a narrow stretch of KY 451 in Perry County that is squeezed in between a river and the railroad. This route carries only about 400 VPD and is a local road. Really, the only people who are going to be traveling this route are those who live along it. Last year, a vehicle overturned and slid into the river and the driver died. This is the only fatality accident that has been recorded in this area, but residents started a clamor to "DO SOMETHING!!!" and erroneously made the claim that this route has a high accident history. The newly elected state representative took this up as a pet project, demanding that guardrail be installed. This section of roadway ranked very low in our guardrail program due to low traffic counts, lack of accident history, etc. Yet the residents and the politicians persisted. This wasn't a typical guardrail installation, either. Because there was no shoulder, we had to drill steel and backfill to create a shoulder so we could install guardrail. Despite this not meeting any criteria for immediate work, and ranking very low in the guardrail program, the politician succeeded in having the transportation secretary allocate $73,000 in discretionary funding to install guardrail. IMHO this was a waste of money. Yes, the fatal wreck was tragic, but the response was overblown. You can't go around installing guardrails or cable barriers at the site of every fatal wreck.

I've gotten the same reaction about a low-volume backcountry road, where people sometimes miss a turn, veer into a shallow creek only a few inches deep and a few feet below roadway level, and occasionally turn over -- never a fatality.  Whenever I'm with a passenger on that road, the question of installing guardrail often comes up.  I have to remind them that guardrail isn't as cheap as people think (not so much the railing itself, as the anchors and their installation), and the county is doing pretty well just putting pavement on that road.

[JREwing78]

I've seen many a freeway with narrower medians without any kind of barrier whatsoever. As highways go, this one appears to be quite safe.

[J N Winkler]

It would be hard to tell the median width without actually pacing it out, or getting the original construction plans from KDOT (a major undertaking in itself).  However, the K-10 freeway was built in the mid-1970's (using official state maps scanned and available on the KDOT website:  not present 1970, shown 1980, "under construction" linestyle in 1975), which would make it the same vintage as US 54 in western Sedgwick County.  So my guess would be 70' dish median, 8' inside shoulders (so 54' of the median would be vegetated), two 12' travel lanes in each direction for a total 24' traveled way per direction, and 10' outside shoulders, all accommodated within a 300' right of way.

I have heard of median cable barrier initiatives for 45' medians with partial vegetation but Myosh_tino's example is the only one I have heard of which takes median cable barrier provision up to 75' median width with partial vegetation.  Of course there is a payoff point for providing median cable barrier on medians that wide or even wider, but I don't know if the traffic volumes on K-10 reach that level.

[route56]

It would be hard to tell the median width without actually pacing it out, or getting the original construction plans from KDOT (a major undertaking in itself).  However, the K-10 freeway was built in the mid-1970's (using official state maps scanned and available on the KDOT website:  not present 1970, shown 1980, "under construction" linestyle in 1975), which would make it the same vintage as US 54 in western Sedgwick County.  So my guess would be 70' dish median, 8' inside shoulders (so 54' of the median would be vegetated), two 12' travel lanes in each direction for a total 24' traveled way per direction, and 10' outside shoulders, all accommodated within a 300' right of way.

Construction began in 1974, Desoto to K-7 completed in 76, Lawrence to DeSoto in 78. The K-7 to 435 section began construction in 1980, opened in 1984. You could have asked me.

Also, mention has been made about a 2008 KDOT study that showed that US 75 in Topeka and K-96 near Wichita were considered better candidates for median cables.

http://www2.ljworld.com/news/2011/apr/19/kdot-cites-2008-study-about-cable-median-barriers-/

[much of the comments of this article are in reply to a suggestion that the median should go "all natural"]

[realjd]

Median barriers  (or cables) seem like they should be a standard safety feature on all freeways IMO.

At least they're calling for something that would actually save lives instead of blaming it on the speed limits being too high.

[J N Winkler]

Construction began in 1974, Desoto to K-7 completed in 76, Lawrence to DeSoto in 78. The K-7 to 435 section began construction in 1980, opened in 1984. You could have asked me.

Thanks for this information.  I figured you would have the exact dates, but it was easier to let my fingers do the walking.

Also, mention has been made about a 2008 KDOT study that showed that US 75 in Topeka and K-96 near Wichita were considered better candidates for median cables.

Yup.  I believe it is this study:

http://www.nlc.state.ne.us/epubs/R6000/B016.0133-2008.pdf

Despite the word "guidelines" in the title, which tends to imply it is a sort of manual, it is actually a research report (based on Kansas crash data) which attempts to refine warrants for cable median barrier installation based on climate (cross-median crashes in winter tend to involve lower speeds and lesser crash severities), median cross-section (including presence of steep foreslopes which prevent recovery), frequency and degree of curves, geometry and spacing of interchanges, etc.  Some interesting facts:

*  60' is the dominant median width on barrierless Kansas freeways (761 miles in all):  88% of this freeway mileage is classified as having a 60' median.

*  If there is snow and ice out, you are two to two and a half times as likely to cross the median or have a cross-median crash, but one-third as likely to die as a result of it and one-half as likely to either die or receive a disabling injury, compared to events that occur outside winter driving conditions.

*  The imputed cost of a cross-median crash, averaged over all weather conditions, is about $1 million.  (The imputed cost of a fatality by itself is just over $3 million.)  But the imputed cost of a cross-median crash involving ice is a little under $350,000 while one involving slush is just under $900,000.  On the other hand, when it rains, the imputed cost is about $1.9 million.  (People think they can drive at 70 in the rain until they hydroplane and then discover wet grass has no traction . . .)

*  The researchers say the cost of a collision with median cable barrier is just under $50,000.  Their treatment of Missouri crash cost data ($38,000 adjusted downward to $26,000 when unreported accidents are taken into account) is (to my eye) somewhat evasive.

*  The researchers suggest that median cable barrier not be considered (due to the potential for underestimation of cost of median cable barrier crashes) unless the benefit-to-cost ratio is at least 2.  They also note that in the Midwestern states, resurfacing projects tend to have BCRs around 4.  They suggest that median cable barriers begin to be funded when BCR is between 2 and 4, which corresponds to AADTs between 57,700 and 69,200 for freeways with 60' medians.

These values are about 20,000 to 30,000 VPD lower for facilities with median widths between 40' and 50', corresponding to I-235 in Wichita.  In fact I-235 meets the suggested AADT thresholds for median cable barrier provision throughout its entire length, as do all other Wichita-area freeways with the lone exception of the Kansas Turnpike.  Only Kellogg, I-135, and the Turnpike do in fact have median barriers.  Meanwhile, K-10 (which it seems safe to say has a 60' median in rural areas) does not meet the AADT criteria in any of its rural lengths.  I would guess that it does between K-7 and I-435, depending on how wide the median is in that location.  (Assuming 60' width, it definitely does between I-435 and the next two exits west.  It probably also does further west of there if the median is narrower than 60'.)  Moreover, most of the rural lengths have AADTs about half those warranting provision of cable median barrier according to the report's criteria--which suggests most of K-10 would not even reach the breakeven point, let alone the suggested BCR floor of 2.

So, if this report is to be taken as gospel, it seems that KDOT has the difficult problem of explaining why it is a waste of money to provide cable median barrier on a road where a little kid died, when the mayor of Eudora has already sent Brownback an open letter asking him to make sure KDOT is forced to provide barriers.  I predict that KDOT will resist getting drawn into the debate (though with less of the "Shut up, you know nothing" attitude of the Carlson years) and that matters will slowly descend into inanity.  This outcome is less than I would like to see but, to be fair to KDOT, understanding the issues involved in median barrier provision will seem too much like real work for most people, and for government departments in general the safest approach is always to stay at three or four removes from any demagoguing that goes on around a particular issue.

[route56]

It's been pointed out that the recent K-10 study indicated that the rural section of K-10 had a higher accident rate than the state average. Certainly, that will play a factor.

It may be morbidly strange, however, since you mentioned 235 in Wichita; I have to wonder - which highway has the higher rate of crossover accidents. Although the Nebraska study utilized AADT and median width, accident rate did not appear to be a factor.

Incidentally, a facebook group has been set up to promote the installation of crossover cables on K-10. One of the people who "liked" the page is official KDOT facebook page created by the social media manager (AKA @KDOTHQ on twitter)

[J N Winkler]

The study I found has to be the one cited in the Journal-World article linked to above, because the lead author name (Sicking) and the date both match.  Is there a separate study which focuses on K-10 specifically?  The impression I have had from the coverage I have seen is that KDOT has been taking the 2008 report and plugging in the AADT numbers, which of course do not justify provision of median cable barrier.  But your point about unusually high numbers of accidents on K-10 is a good one.  If the excess is partly or wholly made up of cross-median events, then this alters the calculus in favor of providing barriers.  I don't know if KDOT stratifies crash incidence data to this degree as a matter of course.  When Sicking et al. did the 2008 study, they had to have their graduate students wade through stacks of accident reports in hardcopy in order to identify crashes involving median intrusion.

Community activists could also try hitting KDOT over the head with what NCDOT is doing--cable median barrier provided, no questions asked, regardless of traffic volume, for medians up to 70' width.

Regarding I-235 versus K-10, I have no data on crossovers but I would find it worrying if K-10 had more than I-235 (AADT around 50,000 at the Kellogg cloverleaf, as opposed to 25,000 on rural lengths of K-10).  I am also sort of interested in knowing where the median cable barrier experiments are in Wichita and Topeka.  My guess would be K-96 just east of I-135 (AADT of almost 60,000), although the rural lengths around the Maize/Colwich exit northwest of Wichita would be a better choice if they wanted to examine a marginal case.  (This is in roughly the same area as several well-publicized car/horse-and-buggy crashes, which prompted KDOT to ask for an engineering review which found that while the geometric design technically complied with AASHTO Green Book standards, it had very little "slack" owing to the use of near-minimum values for various criteria.)

[route56]

Is there a separate study which focuses on K-10 specifically?

A commenter in the J-W forum mentioned KDOT's K-10 corridor study (i.e., the one that noted that a full stack at K-7 may be in the cards soon). The study did confirm that K-10's accident rate on it's rural segment was higher than average for rural 4-lane controlled access facilities.

Regarding I-235 versus K-10, I have no data on crossovers but I would find it worrying if K-10 had more than I-235 (AADT around 50,000 at the Kellogg cloverleaf, as opposed to 25,000 on rural lengths of K-10).  I am also sort of interested in knowing where the median cable barrier experiments are in Wichita and Topeka.

I was thinking more in terms of accidents per VMT; obviously, VMT on 235 is significantly higher than 10.

[route56]

UPDATE!

http://www.ksdot.org/PDF_Files/Governor-issues-directive-to-KDOT-regarding-K-10.pdf

Governor Brownback has responded to the Eudora Mayor's message. He has ordered KDOT to re-evaluate the cable barrier study on this segment of K-10, and to start local consultation, starting with the Mayor's office.

He also requested that a project to pave and add rumble strips to the the shoulders of K-10 in Douglas County be expedited. Currently, K-10 in Johnson county has paved shoulders and rumple strips, but the Douglas County portion has gravel shoulders. He would like to see the project completed this fall.

[J N Winkler]

It looks like Deb Miller has her marching orders.

The shoulder paving project is overdue, frankly.  US 54 in eastern Kingman County/western Sedgwick County has between one-third and one-fourth the AADT of rural K-10 but received full-depth reconstruction in Portland cement concrete around 2003, including full-width structural shoulders.  Prior to that it had PCCP in the traveled way with (badly raveled) bituminous shoulders.  I found it a little surprising that K-10 still had gravel shoulders, unless KDOT was waiting to add them as part of full-depth reconstruction.

[route56]

Going back to a point that H.B. made:

In an article written today, KDOT pointed out that the fatal accident rate for K-10 is .56 per 100 million vehicle-miles travelled. The article mentioned that the statewide average is 1.23. I also checked the Accident statistics provided by KDOT - the average for rural 4-lane divided highways with full access control is .577. Thus, statistically speaking, K-10 has a below average fatality rate.

On the other hand, there have been three separate fatal crashes in the past 9 months. Given K-10 current traffic levels, that raises the crash rate significantly, possibly causing it to go above average - and thus coming back to KDOT's immediate attention.

[route56]

A letter to the editor that I wrote regarding Median Barriers on K-10.

http://www2.ljworld.com/news/2011/may/02/cable-barriers/