CSX oil train derailment

[ROBERT WILLISON]

Mr. Dining car... Silly silliy comparison.  Terrorism to transportation safety and regulation issue.  Really

[BroadwayLion]

LIONS do not reply to PRIMATES

 

ROAR

[Paul_D_North_Jr]

diningcar

Mr. Willison, you wouldn't be thinking about the threat of bombs being detonated in our shopping malls would you. If one, or more, of these occur a train derailment will be relagated to page 15 next to Ann Landers.

 Well, the Lac Megantic derailment and fire in July 2013 was the next thing to a bomb in shopping mall - a downtown shopping area instead, and resulted in 47 killed, about the same range of death toll as from many terrorist bombs.  It hasn't been relegated to page 15, either - more like on page 1 or 2 in every article about an oil train derailment since then. 

The Chatsworth, Calif. passenger train wreck that resulted in the PTC legislation killed about 25, after the NS Granite, SC chlorine car wreck that killed about 9. 

The risk is clearly out there.  To ignore or minimize it by thinking, saying, or acting like all such future incidents will occur out in the middle of nowhere is contrary to recent experience, and hence foolhardy.   

- Paul North.     

[Euclid]

Fred Frailey has some good information and discussion about the odds of certain magnitudes of oil train wrecks occurring. 

http://cs.trains.com/trn/b/fred-frailey/archive/2015/02/22/more-thoughts-on-exploding-trains.aspx

 

In an earlier piece, he calculated the odds of a major oil train derailment.  He used some statistical methodology that seemed reliable, and it showed the probable frequency of predicted high speed derailments. 

In my opinion, the numbers were large enough to indicate a serious problem for the industry.  It reminds us of a widely perceived vision that domestic oil by rail traffic is in ascendency, and the crowded railroads will get even more crowded with more and more oil trains.

However, as this oil traffic increases, there will need to be a concomitant capacity expansion or else the rate of oil train wrecks will likely increase over time.  

Right now, these oil train wrecks do not seem to be killing people, but each one does demonstrate enormous killing power, and by extension, the realization that the killing power can suddenly strike anywhere that an oil trail rolls.  That is the problem.  Nobody wants to take that chance.  It is a no-brainer among the public. 

Even if you can predict the number of derailments based on probabilities, you can’t predict where those derailments will happen.  They can happen in wide open undeveloped areas, or they can happen in the middle of a good size town.   

Oil trains will occasionally derail, and sometimes at high speed.  The will derail on high fills and in deep cuts.  They will derail on bridges and in tunnels.  They will collide with other trains.  They will collide with vehicles at grade crossings.  Any of these causes could easily pile up 30 tank cars, many of which will be ruptured, and begin leaking.

The media has been priming the pump for a blockbuster story proving that oil by rail is not in the public interest.   They need an accident with fatalities.  This is a dangerous time for the industry.    

[diningcar]

Ah, now that the 'pump has been primed' with our somewhat not too serious comparisons why don't we build pipelines?? Ask our President!!

[schlimm]

Euclid

Fred Frailey has some good information and discussion about the odds of certain magnitudes of oil train wrecks occurring. 

You may have read that the Canadian government is requiring rail shippers of crude to take out insurance against the risk of derailment damage. You can bet that the Marsh & McClellan actuaries are trying to calculate the risk of another OMG incident - a Lac-Megantic or that Amtrak train plowing into the train of crude oil splayed across the right of way.

Let's see: A year or so ago in a blog, at someone's suggestion, I plowed through FRA accident stats and concluded we should expect six noticeable derailments of crude oil trains a year. Well, that's about what is happening.

What I don't know how to calculate the odds of a truly catastrophic incident, a la Amtrak or Lac-Megantic. But this is really a form of Russian Roulette. Spin the chamber often enough -- derail enough crude oil trains over time -- and a bullet will someday be in the chamber.

Your odds of having that day dawn diminish when you decrease the rate of derailments. We need to 1) seriously slow these trains down (the CSX train in West Virginia was doing 33 mph, and six days later the fire is still burning) or 2) take better care of our track and equipment.--Fred W. Frailey

[kgbw49]

Today's tank cars are really just updated versions of the original ones - metal tanks on a platform mounted above the wheels with a very high center of gravity. Even the latest versions look like modifications of the oil tank cars that rode the rails in WW2. What about some outside the box thinking? What about double hulled vessels mounted low to the ground "between" the trucks, much like a well car for containers. In fact, the trucks could be equipped with an aerodynamic faring above them that would contour with the vessel and serve as a deflector to a coupler coming toward the vessel. Perhaps a lower center of gravity and double hulls, with larger capacity vessels to make up for some of the added tare, would result in a more stable ride and perhaps a dragging to a stop in a derailment instead of folding up like an accordion. Also, wouldn't DPUs help reduce some of the accordion effect if they would go into emergency upon breaking of the train line? It seems like all the thinking is "just do the same thing that is going on now, but tweak it". And so far the tweaking doesn't seem to be making any market improvements in the safety.

[tree68]

I would opine (rather obviously) that the danger is relative, which makes it hard to determine a method of rating the danger.

The damages resulting from a 100 car pile-up in a cornfield in Iowa, miles from the nearest residence are going to be far different from those resulting from Lac Megantic.  It might be necessary to piecemeal such ratings.  That might result in some rerouting, but I doubt that most folks would be happy with knowing that their lives and property are worth less that someone elses... 

 

[jeffhergert]

ROBERT WILLISON

Jeff, I think you hit it on the head. The general public understand and accepts reasonable risk. Nothing can be made 100% risk free. 

New tank cars can be  and will be built to higher standards. Thier are too many stake holders here including the railroads, shippers and the end users of the product not to develop a better car.

Most can remember when an air bag was consider an unneeded and costly government regulation. Every one knows  the air bag won't save you if drive over a cliff. Over time they have saved many a life.

A better tank car will be built to meet the needs of all the stakeholders in this issue.  

But time is ticking, and needs to be done in responsible time frame.

 

I think some are missing one of the points of the article.  The answer isn't just building stronger cars, but making the crude less volatile.  You probably can build a car that will withstand almost everything, but it will probably be so inefficient to be rendered useless.  That's not to say a safer car can't or shouldn't be built, but maybe the shippers need to do all they can to make the oil less explosive.

Of course that might mean the oil companies make a few cents less.  No more catastrophic derailments would take the wind out of the sails for those who either want to see no oil extracted out of the ground or have a vested interest in pipelines.

Jeff (who has a vested interest in freight moving by rail)

[Euclid]

The article is about making Bakken oil less volatile.  It offers that as the ultimate challenge.  But it comes to that conclusion because of a startling admission that the problem cannot be solved by stronger tank cars.  So the only recourse it to throw the responsibility of solving the problem back onto the oil producers to make Bakken oil less volatile. 

However, the railroad industry has assured us that they are moving ahead with stronger tank cars, and they have clearly implied that will solve the problem.  So I am amazed to hear the comment in the article:

“Carbuilders themselves recently warned that no amount of extra metal or sophisticated engineering will protect against breaches and fires in high-energy derailments.”

Did they just figure that out?  Or have the carbuilders been dragged along in the industry’s assertion that stronger cars will solve the problem?  If so, it appears that the carbuilders want to set that record straight so as to not be blamed when stronger cars prove inadequate to solve the problem.      

The article strikes me as being in denial and back tracking about what was implied by the call for stronger tank cars.  Now they say that the newer CPC-1232 tank cars were only intended to be safer than the DOT-111 cars in slow speed derailments and rollovers.  In fact, the following statement from the article seems to vindicate the previously maligned DOT-111 cars after the long running consensus that the 111 car was indeed the problem:

“Both the CSX oil train disaster in West Virginia and the nearly concurrent CN derailment and explosions in Ontario involved the newest CPC-1232 tank cars, underscoring the fact that the overwhelming risk factor is the explosivity of untreated crude oil, not the inadequacy of the disgraced DOT-111.”

Oddly, the article concludes that failures of the 1232 tank cars to perform as promised proves that the volatility of Bakken oil is the problem.  That strikes me as childlike reasoning.  We all knew that the oil burns.  The point of stronger tank cars was to prevent spilling the oil.   

[jeffhergert]

Silly me.  I thought we were more concerned with wiping out a town when an accident does happen.  I guess it's better to have all that explosive oil burn off in a big fireball than to just puddle on the ground when the car is breached. 

While both situations are bad, I think one (the explosive one) is worse than the other.  Some things may burn, but won't necessarily explode in most situations.  Since we can only go so far in the design and construction of the cars, I would think other options should be looked at, too.  Doesn't mean you can't improve the tank cars, just that no matter what there will eventually be another accident where a car will be breached.

I wonder what the public reaction would be if it was said that the risk could be reduced by removing some of the volatility of the oil.  All they ever report about is the design and construction of the cars.  Of course, most of the media love the headline inspiring fireball photos.  Many are also on the green/no fossil fuel bandwagon.  Bringing risk to an acceptable level for most people might impact them adversely.  And it's low hanging fruit to just blame the railroad (seems to be a reoccuring theme by some on here,also) even though most cars aren't railroad owned.  It would also require more indepth investigation and reporting.    

Jeff

[tree68]

Jeff - I agree.  I would opine that there is some underlying reason for, if you will, deflecting the blame for these incidents from the volatility of the crude itself to the transport medium (in this case, trains).

The usual mantra is to "follow the money."

[Murphy Siding]

     At what point does the carrier- rail or truck- insist that the oil  carried meet a certain *safety*(?) level, or charge a higher rate for shipment, or ask that the shipper provide some sort of insurance policy?

[Euclid]

jeffhergert

Silly me.  I thought we were more concerned with wiping out a town when an accident does happen.  I guess it's better to have all that explosive oil burn off in a big fireball than to just puddle on the ground when the car is breached. 

While both situations are bad, I think one (the explosive one) is worse than the other. 

 

Jeff,

[mudchicken]

Murphy: Since Graniteville, the railroads have been asking that question and have yet to get  a proper response. (Including from STB, where the shippers (mostly chemical shippers) are claiming unfair rates and conditions inhibiting interstate commerce. (NOR 42117,lots of charges and countercharges, no real solution  - so far everybody just agrees to disagree.)  

[WilliamKiesel]

Is there an engineer or physicist who can help? It appears from looking at the accident sites on Google maps at for the the bakken accidents at Mt Carbon, WV 381496N 81.293108, Casselton, ND 46.901619N 97.283389W, Aliceville, AL 33.095494N 88.146056W, Lynchburg VA 37.415672N 79.138683W. I also looked at the New Brighton, PA 40.738831N 80.131896W derailment involving ehtanol on 2006. All of these sites are either straight or moderate curves. Waht is the factor of cargo sloshing in the tank cars? What are the fluid dynamics for a strtified liquid such as bakken crude? How does the distributed inertia and viscosity of bakken crude coontributing to unstable vehicle dynamics and rail/ wheen interaction? Might  skin friction result in the tank due to sloshing creating localized heating, strees and failure? Are boiling liquid expanding vapor events causing the derailments?

Are the tank cars full when then leave the bakken terminal? A full container lessens slosh. Are the cars losing contents through pressure relief valves in transit sufficient to contribute to slosh? Is the braking dynamics of ordinary air brakes contributing to slosh and a series of events leading to failure? Is there are optimal speed of operation for bakken crude? Is there a maximum number of tank cars not to be exceeded?

Given that the cars at Mt Carbon WV were 1232 types, it is apparent that preventing a catastrophe is much more important.

The railroad companies merely accepting bakken crude now amounts to negligent behavior. The Railroad company must analyze and assure that critical components in the bakken crude are not part of the lading. The railroad company has a duty to refuse.

While Casselton was a collision with another freight train, was the series of events of the derailments due to sloshing then a collision? Lynchburg, Alliceville and Mt Carbon happened on straight and / or gentle curves. Higher speed was not a factor.

Looking at the 2006 ethanol derailment at New Brighton, PA, the NTSB found a rail failure at the point where the r.o.w. ended at the end of a bridge and returning to earth and ballast. Was derailment due to a rail failure? Or, was the derailment due to sloshing seeting the events in place for vehicle wheel separation and the energy required to contribute to rail failure?

In looking at the Vandergrift, PA derailment the news photos showed tank cars that stringlined ahead of an apparent rail weld failure. The cars behind the gap in the rail derailed. Was that tank car derailment due to sloshing that created the rail weld failure? In that tank car derailment bakken was not the crude oil being carried. No explosion. No leakage.  The Conemaugh of the NS line has remarkable curvature and is operated at slow speed. The combination of curvature and  / or normal braking: Could sloshing have contributed to the February 14, 2014 accident.

So, is sloshing the problem? If so, what is the appropriate speed. What is the best manner for operation? What is the optimum number of tank cars? Should spacer cars with new draft gear and dampening design be interspered through the train? already suggested, might an electric controlled air brake system be part of he answer?

[BaltACD]

One trueism -

There is no perfect form of transportation.

Accidents/incidents will happen in any and all forms of transportation that will cause loss of commodity and/or loss of life.

Highway, railroad, water, pipeline and air all have their incidents - they always have and they always will.

Moving anything subjects it to an accidental risk.  Just moving one's own person around the house subjects you to the risk of a slip, trip or fall.

[carnej1]

Regarding the extraction of volatiles from Bakken crude and utilization of the resulting gasses (rather than flaring);there is some interest in the industry in shipping gas in liquified form in tank cars (we've had at least one thread about the proposal) but that O.C will also generate some controversy regarding..

 

[oltmannd]

kgbw49

Today's tank cars are really just updated versions of the original ones - metal tanks on a platform mounted above the wheels with a very high center of gravity. Even the latest versions look like modifications of the oil tank cars that rode the rails in WW2. What about some outside the box thinking? What about double hulled vessels mounted low to the ground "between" the trucks, much like a well car for containers. In fact, the trucks could be equipped with an aerodynamic faring above them that would contour with the vessel and serve as a deflector to a coupler coming toward the vessel. Perhaps a lower center of gravity and double hulls, with larger capacity vessels to make up for some of the added tare, would result in a more stable ride and perhaps a dragging to a stop in a derailment instead of folding up like an accordion. Also, wouldn't DPUs help reduce some of the accordion effect if they would go into emergency upon breaking of the train line? It seems like all the thinking is "just do the same thing that is going on now, but tweak it". And so far the tweaking doesn't seem to be making any market improvements in the safety.

Two comments (one critical and one Kudos!):  

Current tank cars are quite bit different from those 50 years ago.  Biggest difference?  No center sill anymore!  The current designs all use the the tank itself as the car's structural strength.

If we are going to be involved in "boom and bust" of Bakken crude as the price of oil fluctuates, then maybe container tanks in well cars would be a more flexible way to approach this traffic.

[oltmannd]

carnej1

Regarding the extraction of volatiles from Bakken crude

Apparently, the API says extraction of volatiles won't make much difference....but they're hardly impartial.

[oltmannd]

WilliamKiesel

So, is sloshing the problem? If so, what is the appropriate speed. What is the best manner for operation? What is the optimum number of tank cars? Should spacer cars with new draft gear and dampening design be interspered through the train? already suggested, might an electric controlled air brake system be part of he answer?

These are all good questions.  It's clear that the standard method of doing business is inadequate for these trains since there are so many and the danger is fairly great.  The industry better get busy figuring out the answers.

[BaltACD]

oltmannd

WilliamKiesel

So, is sloshing the problem? If so, what is the appropriate speed. What is the best manner for operation? What is the optimum number of tank cars? Should spacer cars with new draft gear and dampening design be interspered through the train? already suggested, might an electric controlled air brake system be part of he answer?

 

 

These are all good questions.  It's clear that the standard method of doing business is inadequate for these trains since there are so many and the danger is fairly great.  The industry better get busy figuring out the answers.

That I am aware of, train handling or uncontrolled slack action within a oil train has not been identified as a cause of the currently known derailments.

[Murphy Siding]

    Perhaps they could simply take the gas out of the oil, and use that gas to power ethanol plants in eastern North and South Dakota.  What fault could the petroleum industry possibly find with that?

[BaltACD]

I can't believe it - railroad not involved.

http://news.yahoo.com/tanker-truck-9k-gallons-fuel-crashes-catches-fire-164511265.html

[tree68]

BaltACD

I can't believe it - railroad not involved.

http://news.yahoo.com/tanker-truck-9k-gallons-fuel-crashes-catches-fire-164511265.html

If people realized how many of those were on the highway at any given time, they'd, well.....

Most don't realize that one of those pulls into their neighborhood gas station several times a week, often during the wee hours of the morning.  

[Murphy Siding]

     My house is 6 blocks from a rail line, 8 blocks from an Interstate highway.  I'd be more concerned about the highway.

[BaltACD]

And if you have oil heat, you have this threat!

http://baltimore.cbslocal.com/2015/02/23/hazmat-responds-after-heating-oil-truck-overturns/

[blue streak 1]

It seems that the BAKKEN crude has much more volitiles.  At present the producers are not removing these volitiles.  Testing of the crude needs to determine how much shaking of the crude causes these volitiles to separate from the underlying crude.  If it separates then do these cars need the protection that is afforded to propane type tank cars ?

As well how much heat causes these volitiles to separate ?  6 hours before the last explosion may indicate that most cars contained the volitiles and maybe there was no pressure relief ?  If cars were tipped a gas pressure relief valve on the top of car might not have been able to work ? Instead the crude may have blocked it ?

 

Of course the condition of the track or wheels of a car may haave caused these wrecks.  However the behavior of the product after a derailment may be as important ?

 

 

[tree68]

blue streak 1

...maybe there was no pressure relief ?

The new cars do include a pressure relief valve, but I did not spot what it's set for.  It would certainly be enough to prevent any significant overpressure.

The issue with an overturned car and the relief valve is the relative viscosities of a gas (ie, air) and a liquid which is probably "thicker" than water.  We all know that you can blow a much larger quantity of air through a straw in a given time than we can a mouthful of water.  So if something (ie, a fire) is causing the pressure in an overturned car to rise, the relief valve probably won't help a lot.  And that assumes that it will even work on an overturned car.  I can't speak to that.

[schlimm]

oltmannd

 

 

WilliamKiesel

So, is sloshing the problem? If so, what is the appropriate speed. What is the best manner for operation? What is the optimum number of tank cars? Should spacer cars with new draft gear and dampening design be interspered through the train? already suggested, might an electric controlled air brake system be part of he answer?

 

 

These are all good questions.  It's clear that the standard method of doing business is inadequate for these trains since there are so many and the danger is fairly great.  The industry better get busy figuring out the answers.

 

http://hosted2.ap.org/APDEFAULT/f70471f764144b2fab526d39972d37b3/Article_2015-02-22-US--Oil%20Trains-Safety/id-302da298369f4215850e884a3895e211