Jump to content


Photo

Lac-Mégantic, Quebec - Oil Train Derailment


  • Please log in to reply
155 replies to this topic

#21 The Dutchman

The Dutchman

    Member

  • Members
  • PipPip
  • 276 posts
  • Location:Essex, Connecticut
  • Interests:self-propelled railcar engineering<br />NTSB Reports

Posted 14 July 2013 - 09:23 PM

I have spent the last week intensively studying the various links in the chain that led to the wreck, and speaking with top management, and would report my observations as follows.
The Railroad: The original line was built by CP as a direct link from Montreal to the ocean port of St. John, New Brunswick. This gave the CP a deep-water port that was ice free in winter. It also gave onward access to the rest of the Eastern maritime Provinces, otherwise dominated by the CN. In time, it was parceled up and sold, and eventually it became (in part) the Bangor and Aroostock RR. The RR went under as its main business, timber haulage, declined. The current owners bought it out of the bankruptcy court and set about to do two things: (1) cut operating costs, by automation, and (2) try to find new business. In the automation end, the train crew was cut to one man, the engineer, and it and an iron-ore RR on the North Shore of the St. Lawrence were the only two RR granted special permission to run 1-man trains. Oil haulage was not in the picture.

the load: the marketing dept secured a contract to start hauling Bakken light crude from North Dakota to the refinery in St. John. The refinery was previously buying "offshore oil" and paid a premium, especially for Brent Light Crude. When Bakken came on-line a trial load was shipped out and it worked fine. So Irving Oil started looking for ways to ship Bakken. It made a deal with CP to ship by railcar direct to the Port of Albany, NY on the Hudson river, then to be transloaded onto barges for tugboat tow to the refinery. However, there are costs attendant to the trans-ship and it was slow, so inventory is tied up in transit. therefore, there was an opportunity for two New England RR properties, the MMA and the Pan Am RR (formerly, Guilford Transportation) to capture a chunk of the business. MMA had the most direct route so it got substantial loads.
In its raw state, crude oil is NOT classified as a Hazmat Load; it is considered a raw mineral commodity in its natural state. What was not clear to management, nor revealed apparently to Transport Canada, (and which I have independently figured out) is that the oil was being diluted with benzine, which is a light aromatic that is quite volatile, and the vapors ignite easily and are explosive if confined or pressurized. I do not think the RR managers understood this. The benzine was being used as part of the fracking process, and while unconfirmed I also suspect it was added to increase the flow rate. In my view, if benzine was added, it becomes a Hazmat Load, as it is no longer a natural crude oil but is a man-made, or manufactured, synthetic oil.
The oil ignites readily, and is shipped in its crude form direct to West Coast terminals for use as fuel oil in large marine diesels. Thus, freighters (containerships and wood carriers) engaged in the China Trade can bunker direct with Bakken straight, and not have to purchase high-priced diesel in Singapore. The oil to the freighter market is all by rail tanker.

the Run: The traincars are turned over by CP to Montreal Maine and Atlantic [MMA] in a yard in Farnham, Quebec, just East of Montreal. They started hauling bulk oil to Irving Refinery last September. Although the one-man crew permit was issued prior to MMA involvement with oil, neither the Company nor Transport Canada stopped to consider the implications of this cargo. Also, some 85% of the tank cars used in Canada for crude shipments are of type DOT-111, which are made from thin plate, are single-hull, and have a history of cracking and folding up in collision (and leaking cargo).

It is one day shift for the run from Farnham to Nantes, which lies some 7 miles West of lac-Megantic. The distance is about 191 km, or roughly 95 miles. Assuming the shift is 12 hours, that implies an average speed of 8 mph for the consist. It is not a fast train. At Nantes, the crews shift over. The Nantes change consists of a siding track with switches at both ends, and the main and siding are sloped downward on the Eastern end at least at a grade of about 1.2%. There is no stub track or derail switch. Usually, the trains are put on the siding for layover, but this time the engineer simply left it on the main. It is all downhill from Nantes to lac-Megantic.

The Layover: the train, with 5 locos and 72 oil cars, arrived at about 2300 hrs. A taxi was waiting for the engineer, to take him to a favorite hotel in lac-Megantic, the L'eau Berge Hotel. What as so importantly waiting for him at that hotel to cause him to rush off is not detailed. The lead engine had the Independent Brake lever set to full brake, and the other locos set to trail position. There is no evidence that the car air brakes were engaged, apparently because there would be a time lapse to pump them back up by the next trainman. There is no evidence that a single hand brake was engaged. So the entire 8,000-ton train was hanging on the 20 axles of engine brakes. all engaged by the one compressor; the 4 trail locomotives were shut down. Apparently the theory was: you have air, you have the independent brakes holding the train, the replacement guy is here in two hours, so what else do you need? And the driver screws off. This is my take on it. The Transport Safety Board (Canadian equivalent of the NTSB) has run into this before; they call this behavior an "adaptation." That is to say: there is the Official Rule, Rule 112, that determines the number or railcars that must be hand braked, the train has to be chained, a derail fitted, chocked, push-pull test, whatever, and then you have the specific Company Rules on top of that - and it all gets ignored out in the field at 2300 on a gloomy Friday night in the middle of nowhere. It also implies that this has become the "stealth practice" or norm on the RR, done whenever the bosses are not around, because these guys view not putting their chum to "extra work" unchaining a train as a form of professional courtesy. Again, that is my take on it. So in my view it is a behavioral issue, not a technical issue.

The Fire: not in the above calculations was "the fire." Turns out the lead loco developed a fire in that engine compartment, and the local volunteer fire guys showed up to put it out. They also shut down the engine, and that shut down the replacement source of compressed air. The fire guys had no idea of the consequences of shutting off the engine (and the air) as they are not train guys and have no clue. So now the engine is off, the fire is out, and the fire guys screw off; the train is left in the gloom. It is now midnight and another hour to go before the replacement engineer is to show up. Once the tank air in the engines bled out, the train started its roll. By this time, nobody was around (about 0100). On a 1.2% grade with 8,000 tons and no brakes, it does not take long to get up some speed. At the bottom, there is a switchyard with leading-point switches and frog points, the cars jump the tracks, and you know the rest. The locos were found many more miles down the track (they made it through the yard, probably due to their weight on the rails).

My conclusion: the engineer did not chain the cars, out of a desire for a quick departure, an indifference to the consequences that he did not foresee, and as a "courtesy" to the next guy who would have to un-chain them. He did not set the car brakes as that would have required time to pump them up, and that was another of those "courtesy" ideas. All he did was set the Independent engine brake, turning that lever some 60 degrees - and walking away from the train, with the idea - born of years of such practice, no doubt - that it would be just fine. Field adaptations to the Rules. Guess what: the Rules are there because over time, their inviolable nature to avoid disaster have been proven. Ignore them at everyone else's peril.

I would be grateful for the comments of the highly experienced rail men on this Forum to the above analysis.

#22 EdFindlay

EdFindlay

    Member

  • Members
  • PipPip
  • 410 posts
  • Location:Chelsea, MA
  • Interests:Trains, Trains, Trains

Posted 15 July 2013 - 05:03 PM

An eye witness to the explosion posted this video that the BBC picked up on, just seconds after the initial explosion: http://www.bbc.co.uk...canada-23252281

#23 The Dutchman

The Dutchman

    Member

  • Members
  • PipPip
  • 276 posts
  • Location:Essex, Connecticut
  • Interests:self-propelled railcar engineering<br />NTSB Reports

Posted 15 July 2013 - 05:48 PM

An eye witness to the explosion posted this video that the BBC picked up on, just seconds after the initial explosion: http://www.bbc.co.uk...canada-23252281

What you see is a very rapid flame-front propagation, at one point virtually instantaneous. You do not get that type of burn from crude oil. Crude is a gooey mineral slime. If you take a bucket of oil and stick a lit match or piece of wood into it, the flame will extinguish. Mostly does that even with straight diesel (without the winter additives to increase volatility and cetane). What has been happening is that, in the oil extraction process, another oil compound has been injected into the ground ("fracking fluid") which in turn is composed of benzine, toluene, xylene, and various compounds including diesel and ethanols. This witches' brew is calculated to transmit hydraulic pressure readily, and thus is not viscous but quite fluid. It is this additive, in large quantities, that is creating the volatility and consequent fast-moving flame propagation front that you see in that BBC video.

The more interesting question is: did they disclose all this to the RR? I don't think so. The compounds used in fracking fluid is a closely-guarded internal secret; they think of that as being a trade secret that gives the producers some imagined comparative advantage. Meanwhile, the RR is not exactly climbing up to the top of each railcar, drawing out a sample, and doing a chromatographic analysis of the stuff. The RR is nothing more than a glorified UPS outfit; it relies on the representations made by the shipper. The RR was blind-sided.

Nobody is going to transport this stuff by DOT-111 railcar with this volatility. That is not credible. Either the shipper removes the fracking oils at the wellhead, or the stuff goes into special double-hulled tanks, or the load is refused entirely. It is not going to be shipped as if it were a load of plywood. Ain't gonna happen. Nobody does that. And it would not be moved in a 72-car train. It would go in a 20-car train, and special precautions including a compressor car and a caboose and a ton of guys hanging out watching like hawks, and the shipper gets to pay extra.

Fir those of you who do not believe that crude oil is non-volatile, I invite you to look over the published results of the notorious Torrey Canyon wreck off the UK some 40 years ago. The ship grounded with 120,000 tons of crude, and the UK Government made the decision to burn it off (to prevent the oil fouling the beaches). The Royal Air Force bombed the wreck with 42 1,000-lb bombs, and dropped incendiaries and napalm all over the wreck. Result: no fire. The oil finally spilled out and created a huge mess.

Well, if the RAF cannot set crude on fire with napalm and aviation gasoline drops and incendiary bombs, then why would this rail load go blow up? Reason: it was no longer crude; it was "something else".

#24 KevinKorell

KevinKorell

    Board Leader

  • Sr. Admin
  • PipPip
  • 82316 posts
  • Gender:Male
  • Location:Lakewood, NJ
  • Interests:Making America TRAIN again!

Posted 17 July 2013 - 07:45 AM

Los Angeles, CA Times, 7/16/13:

Engineer in Canada rail disaster unfairly blamed, lawyer says


The train engineer on duty during the July 6 Canadian rail disaster that killed at least 50 people and incinerated the Quebec town of Lac-Megantic is devastated and in seclusion, his lawyer told Canadian journalists Tuesday.

Article


Kevin Korell


OTOL Board Leader


Lakewood, NJ


#25 KevinKorell

KevinKorell

    Board Leader

  • Sr. Admin
  • PipPip
  • 82316 posts
  • Gender:Male
  • Location:Lakewood, NJ
  • Interests:Making America TRAIN again!

Posted 17 July 2013 - 07:51 AM

Progressive Railroading, 7/15/13:

TSB will perform 'top-to-bottom' investigation of Quebec derailment, Tadros says


Transportation Safety Board of Canada (TSB) Chairwoman Wendy Tadros on Friday visited the site of the Montreal, Maine & Atlantic Railway derailment in Lac-Mégantic, Quebec, and stressed that the agency has launched a full investigation into the deadly accident that occurred on July 6.

More here.


Kevin Korell


OTOL Board Leader


Lakewood, NJ


#26 Sloan

Sloan

    Member

  • Global Moderator
  • PipPip
  • 12851 posts
  • Gender:Male
  • Location:Harrisburg, PA
  • Interests:Light Rail &amp; Commuter Rail&lt;br&gt;Passenger Rail Stations

Posted 17 July 2013 - 11:10 AM

Ed Burkhardt talks with Railway Age on Lac-Mégantic

Responding to a blog posted on this website by Editor-in-Chief William C. Vantuono, Rail World President and CEO Ed Burkhardt had this to say to Railway Age about the tragic accident at Lac-Mégantic, Quebec, on the Montreal, Maine & Atlantic Railway:


http://www.railwayag...nt=Full Article

#27 EvergreenRailfan

EvergreenRailfan

    Member

  • Members
  • PipPip
  • 2870 posts
  • Location:Seattle, Wa.

Posted 17 July 2013 - 04:42 PM

I am beginning to see that many could be at fault, especially the engineer for not setting the hanbrakes, and the oil company for not disclosing what was in the cargo. Still, will have to wait for the investigation to run it's course. Haven't the oil companies been pushing for state and federal laws protecting the trade secrets of the fracking fluids? Although the railroad should have the right to know, for specifically this reason, so they can tell the firefighters what they are up against. I know that this was a small town fire department, without the resources of the SFD or the Montreal FD, though.

#28 The Dutchman

The Dutchman

    Member

  • Members
  • PipPip
  • 276 posts
  • Location:Essex, Connecticut
  • Interests:self-propelled railcar engineering<br />NTSB Reports

Posted 20 July 2013 - 11:00 PM

The population of Lac-Megantic is about 6,000. In a town that size, the fire department is all-volunteer. Ironically, I live in a town of about equal size and also was a volunteer firefighter. Although the guys mean well, they are not trained to handle complex machinery. In any fire of any size, the protocol is to call for "mutual aid" from neighboring towns. In the inferno, volunteers with their trucks came from over 100 miles away, including from within the US (quite common along the Border, where these small towns have symbiotic relationships). When the volunteers came to the locomotive "fire," their reaction was to shut off the engine. They did not stop to think: "Why is the engine running in the first place, unattended? What function is that running engine serving? What are the results, or at least the implications, of the engine being shut off? Should another engine be started up?" It is hard to fault them for that; they are not railroad men. Note also that no portable derail was fitted ahead of the lead locomotive, either, or for that matter anywhere else. the reality is the engineer screwed off and left the train; not a very bright choice.

#29 KevinKorell

KevinKorell

    Board Leader

  • Sr. Admin
  • PipPip
  • 82316 posts
  • Gender:Male
  • Location:Lakewood, NJ
  • Interests:Making America TRAIN again!

Posted 21 July 2013 - 11:07 AM

Reuters, 7/21/13:

Insight: Canadian train disaster a dark turn for rail veteran


A blinding flash of orange light jarred Weyauwega residents awake before dawn on March 4, 1996. An 81-car freight train had been barreling toward the farm town in central Wisconsin when it jumped a broken rail. The train's propane and petroleum cargo had caught fire and exploded.

Article


Kevin Korell


OTOL Board Leader


Lakewood, NJ


#30 CNJRoss

CNJRoss

    Administrator

  • Admin
  • PipPip
  • 43390 posts
  • Gender:Male
  • Location:Fairfax, VA

Posted 21 July 2013 - 02:46 PM

CBC News, 7/18:

Transport Canada mum on rail safety rules
Companies won’t share details following Lac-Mégantic rail disaster


Transport Canada won’t say what the minimum requirements are for making sure a parked train won’t roll away and it won’t disclose the rules set by the rail companies for keeping unattended trains with potentially dangerous goods stationary.

The CBC asked Transport Canada to clarify the rules for tying down a train a few days after the Lac Megantic tragedy. More than two days later, the response ignored the specific request for minimum requirements and referred to the Canadian Rail Operating Rules.

The guidelines in the CROR, however, lack specificity on parking trains.

Section 112, on "Securing Equipment", says: "When equipment is left at any point a sufficient number of hand brakes must be applied to prevent it from moving. Special instructions will indicate the minimum hand brake requirements for all locations where equipment is left."

The section goes on to discuss parking a train on a siding, and to specify that hand brakes need to be fully tested before they can be used to secure equipment.

Continue.




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users