Railroad safety and oil

[buzz456]

http://online.wsj.com/news/articles/SB1 ... _whatsNews

Duh,
When do you think the geniuses in DC are going to figure out that pipelines are a safer way to move crude?

[BNSFdude]

http://online.wsj.com/news/articles/SB10001424052702303880604579405412862500426?mod=WSJ_business_whatsNews

Duh,
When do you think the geniuses in DC are going to figure out that pipelines are a safer way to move crude?

They aren't really much safer. The only difference between the two is a pipeline can directly poison water tables, while trains can blow up an entire town. It's lose-lose, really, but I'd rather see the railroads profit than big oil getting gov't subsidies for land, when they plan on just exporting the oil anyways.

[5292nate]

Only reason I'd take trains over the pipeline, is because it's the only reason I get cool power! lol!

[buzz456]

http://online.wsj.com/news/articles/SB10001424052702303880604579405412862500426?mod=WSJ_business_whatsNews

Duh,
When do you think the geniuses in DC are going to figure out that pipelines are a safer way to move crude?

They aren't really much safer. The only difference between the two is a pipeline can directly poison water tables, while trains can blow up an entire town. It's lose-lose, really, but I'd rather see the railroads profit than big oil getting gov't subsidies for land, when they plan on just exporting the oil anyways.

You are just dead wrong. There are very very few poisoned water tables from pipelines. First time one of these trains blows up in a major urban area we can re-visit this. The rest of what you have to say is just populist bs.

[buzz456]

Petroleum.jpg

Having said that the railroads do a pretty good job, however it's only going to take one bad one here in the states.

[ErikGorbiHamilton]

And yet they still ship HCL in tankcars...

[buzz456]

And all kinds of other highly flammable stuff. It's usually not a whole train other than we have 20 or more cars of ethanol going by here all of the time.

[up_8677]

Which may still qualify it as a so-called key train.

And also, Lac-Megantic was, to my understanding, more of an issue of a second rate railroad running one man crews. That system is a disaster waiting to happen, and guess what...it happened. People "in the know" definitely agree that it is likely that the engineer simply didn't tie down enough hand brakes...but doing it by yourself is extremely difficult. As I understand it, one process is that the conductor ties down some brakes, and then radios the engineer to work some power against them; when the locomotives will no longer move the train, that's enough handbrakes. But doing that while by yourself is obviously really hard...

BNSF is also moving ahead with acquiring it's own fleet of safer oil cans, too. Well ahead of any FRA mandates...

http://www.reuters.com/article/2014/02/ ... BE20140220

[BNSFdude]

As I understand it, one process is that the conductor ties down some brakes, and then radios the engineer to work some power against them; when the locomotives will no longer move the train, that's enough handbrakes. But doing that while by yourself is obviously really hard...

Now it is a rule you have to do it on all trains because of EO28.

[Chacal]

The problem with figures like in the table above is they present a chosen aspect of reality.
In this case, it presents the number of incidents, which says nothing about the severity of incidents.
It should present risk instead.

Risk depends on two factors: impact and probability.
The table above presents only probability, which is easily quantifiable.
Impact is more difficult to quantify and verify.
Material losses, life losses, liability, cost of clean-up, pollution of water table (easily the worst impact), etc.

The risk is probably about the same in both rail and pipeline transport.
For rail transport, probability is higher but impact is less for most incidents. You don't get a Lac Megantic often, and the amount of spill is usually limited, so pollution of water table is not frequent.
For pipeline transport, probability is lower (although incidents are easier to hide and are often not reported), but impact is higher because of the volume of spills.

Whatever the mean of transport, risk must be reduced by reducing both impact and probability. Stricter rules, better cars, better quality control for pipelines, etc. This will drive up the price of oil, which is good because it will reduce oil consumption, which is the only true way to lower the risk to an acceptable level.

[philmoberg]

... Having said that the railroads do a pretty good job, however it's only going to take one bad one here in the states.

The problem with figures like in the table above is they present a chosen aspect of reality. ... Whatever the mean of transport, risk must be reduced by reducing both impact and probability. ...

Having read more accident reports, and having reviewed and commented on more legislation and regulation than I care to remember, and having worked (however briefly) in train service, I'll add a few comments to the discussion. Indeed you don't get a Lac Megantic often, these days, certainly not as often as we got them back in the '70s. That's primarily a function of much better track maintenance, improved equipment and more diligent operating practices. IMHO, Lack Megantic shouldn't have happened in the first place: it may be a pain in the neck, but if you leave a train of any sort of volatile or explosive material stopped on any grade and walk away from it, you should have had every handbrake would up tight. It wouldn't have hurt to chock a few wheels wheels they were at it. Good operating practice demands no less a standard of conscientious behavior than that.

Invariably, this sort of thing will result in additional legislation and regulation, primarily of the don't-just-stand-there-do-something culture the elected sector are operating in (Note: why they are operating in this sort of culture is a discussion I won't address). They will probably throw a lot more administrative process and possibly some improved technology at the situation. This may or may not help appreciably, based on what I've seen. If the last forty years of North American railroading have taught us anything with respect to safety, it is that the human factor is the biggest factor in any calculation of risk. Where operators have been willing to conscientiously maintain their infrastructure a state of good repair and hold themselves to good operating practices, their safety record has been excellent. The same can be said of pipeline operators.

I'm not sure what the effect of the additional regulatory costs will ultimately be, but experience suggests that they will have a negligible effect on pump prices: the cost of transporting petroleum products, by whatever means, is only a very small component of the cost to the consumer. That said, I'm barely old enough to remember when large quantities of high-octane gasoline moved by rail with a safety record my [former] colleagues would scarcely find believable. Many of these shipments reached their destination moving through heavily urbanized areas, indeed, through city streets in a couple of cases that come to mind, with a safety record so good that people who lived next to it took it for granted. It is possible to reduce the risk to the point that technological failure becomes the biggest risk, because we once did so.

[byelen]

"philmoberg", you make a very good point. The "root" cause of the Lac Megantic disaster was the inadequate manner in which the train was "tied down". The explosions and fire would not have occured if the train hadn't broken loose, ran away, and eventually derailed in the middle of town. In a sane rational environment (which always goes out the window when government gets involved) two major things must happen: (1) - All trains, especially those carring hazardous materials must be properly secured to avoid this type of accident to begin with. The costs for this solution are those brought on by the probable need to run with appropriately sized crews. (2) - What can be done to "harden" the cars invovled so that in a future accident, they are better able to withstand the forces involved in a wreck. This will be very expensive.

Past history will show that rollingstock can be redesigned to better withstand this type of wrecks. A good case in point is the evolution of the railroad passenger car in the United States and Canada. In the 1870's for example, passenger cars were made of wood. In the event of a wreck, the biggest cause of passenger deaths was the destruction caused by one car "telescoping" into another. This was largely eliminated by going to all steel construction with heavy frames and reinforcing at the ends of the car. Look at the differences between North America and some of the international countries when a serious passenger wreck occurs.

- Bruce

[philmoberg]

The "strongbox theory," as it used to be called a few decades back, is a valid argument as far as it goes. The problem is that you get into a cycle of diminishing returns as the weight of the vehicle goes up. This happens for a several reasons, the first of which is physics: force equals mass times half the square of the velocity. This the heavier the vehicles get, the greater the impact when they hit. Some of our New York area members may be familiar with the ghoulish photo that graced the cover of the Daily News, about half a century ago when tow trains of MP54s hit head on at speed, and the two lead cars completely telescoped, the one on the bottom having been sheared-off at the belt rail. These were all-steel cars built to the Pennsy's take on heavyweight construction, and were well designed for their type. The cause of the accident was human error: IIRC one of the trains was wrongly routed at an interlocking. Similarly, many of us will be familiar with the Chase, MD wreck, in which a Conrail train ran a red block and ended up in front of a passenger train at speed. The lead unit of the passenger train was effectively vaporized, according to the report. The engineer of the Conrail train was intoxicated. Most of the hideous wrecks involving large death tolls were due to negligence and/or irresponsibility, even in the wooden car era. Improved communications and operating discipline did more to alleviate the problem than everything else combined.

The second problem you get into with attempting to harden the rolling stock is that you quickly reach a point at which the infrastructure needs to be rebuilt to accommodate heavier vehicles. A few decades back, we still had the benefit of the foresight of the early civil engineers who built railroad infrastructure with a safety factor of four - that is about four times more rugged than it needed to be for what they were running then. They knew from the then-brief history of railroad operation that locomotives and rolling stock were liable to become much larger and heavier, even though they didn't know quite how that would be accomplished - mechanical stokers, for example, had not yet been invented - so they took their best guess as to what the future would look like. We got a break, beginning in the '20s, as metal alloys became stronger and more durable for a given weight, allowing much of this infrastructure to function with a minimum of maintenance, and many railroads got used to cutting back those expenses first. Several breakthroughs in design and construction began to push freight car design from the 15-ton axle loadings that were still typical of freight cars when I was in high school in the late-60s, and 20- to 25-ton axle loadings quickly became typical and didn't stop there. That's when the deferred maintenance began to catch up with them, as the track structures began to fail under the heavier traffic and the first of a rash of spectacular derailments involving petroleum products and/or chemicals began to happen. Railroading was a low-margin business, at best, in those days, and this resulted in the abandonment of a lot of lines that simply couldn't be rebuilt because their traffic, however heavy (in some cases), couldn't generate the income to pay back the capital investment at market rates.

As of now, a 35.75-ton axle loading is becoming an industry standard. The reason for this is economic: when I was in high school, a typical freight car could carry, at best, about twice its tare weight (take tare weight as empty weight, in this case, although the terms are not strictly equivalent). Currently, a car grossing 143 tons is carrying about 110 tons of payload, two thirds better than the best case, and better than double the typical case of my high school days. This is the primary reason railroads are becoming investment-grade businesses again, but the income isn't sufficient to reconstruct the entire network for substantially heavier rolling stock, nor can they afford to operate the network at the lower speeds necessary to make the strongbox theory workable. An excellent discussion of the problems related to higher axle loadings begins on Page 83 of this document (http://www.ct.gov/dot/lib/dot/documents ... 1-8-12.pdf). It's probably in other State Rail Plans as well, but I'm most familiar with this one because I reviewed an early draft of it back in 2011.

Even assuming they could make it work economically, the problem you get into is one of human nature, in that people tend to become over-reliant on the safety technology and tend to lose operating discipline - intentionally or otherwise - as a result. A study of highway safety measures over the years is very instructive, in this respect, mainly because the database is larger than it is for other modes. Every improvement in headlight technology, for example, has been followed by an increased tendency of drivers to "over-drive" their headlights in conditions of bad visibility. It's one of the reasons I have misgivings about the implementation of PTC: it would not have solved the root problem of the Chatsworth wreck, which was irresponsible behavior in the cab. It would have the unintended consequence of making it possible to get away with the behavior, then leaving the passengers vulnerable to the consequences of dumping the air at speed: for most of them, the experience would only have been nominally better than the collision itself

Based on the evidence I've seen, Chacal's point about risk-based problem solving would appear to be the most valid approach; and the evidence seems to indicate pretty clearly the biggest risk is the human factor. It's less a question of how big the crew is than it is how responsible their behavior is - SNCF has been running one-man crews on freights trains with an enviable record for a long time now, for example. When the cab-forwards were introduced to Donner Pass, one of the engineers was heard to have objected to the possibility of ending up with a caboose in his lap. The Superintendent was quoted as having replied, "Do your job right and you won't have a caboose in your lap!" It's as true now as it was then, and earlier, when my great grandfather organized his local of the Brotherhood of Railway Trainmen, early in the last century.