Predicting oil train accidents

Posted by Fred Frailey
on Saturday, January 18, 2014

I’ve been wondering, as you may have: How many serious incidents involving unit oil trains should you expect? The goal may be none, obviously, but we know a lot about railway accidents, and statistics provide answers. Something I read on the RBN Energy web site, in which a subscriber did his own probability analysis, sent me to the Federal Railroad Administration’s accident database. There, I wasn’t able to replicate what this fellow did, but came up with my own answer to the question.

In the first ten months of 2013, there were 1,747 reported railroad accidents. Subtract the 990 that occurred in yards, and you are left were 757 accidents. In that same period, 548 million train miles rolled off outside of yards. That works out to a rate of 1.38 accidents per million train miles, rounded to 1.4.

The Association of American Railroads estimates that railroads carried about 400,000 crude oil loads in 2013. Let’s make a few assumptions, because we must: Assume all those loads were carried in unit trains (most were), that each train carried 100 tank cars, and that the average trip length was 1,000 miles, each way, very conservative in that most unit trains head for the Atlantic, Pacific, or Gulf coasts, which are 1,000 to 1,800 miles from the oil fields of North Dakota.

These numbers yield 8 million train miles by crude oil unit trains. Now multiply that by the rate of accidents (1.4) per million train miles, and you get a statistical probability of 11.2 accidents involving such trains annually. Finally, cut that number in half because those involving empty trains probably won’t threaten public safety.

So there it is: the statistical probability of five or six possibly very serious oil train accidents per year. I think the number in 2014 could be higher, if nothing is done to reduce the probabilities. That’s because railroads will probably handle substantially more oil loads this year than last. Plus, average trip length may be longer than the 1,000 miles I estimate.

You can look at five or six accidents two ways. One is that it’s only five or six rolls of the dice, perhaps five or six minor derailments. The other is that railroads can look forward to five or six Lac-Megantics or Casseltons every year if they don’t do something, which is how I see things. In other words, a call to action if I’ve ever seen one. — Fred W. Frailey

Comments
To leave a comment you must be a member of our community.
Login to your account now, or register for an account to start participating.
No one has commented yet.

Join our Community!

Our community is FREE to join. To participate you must either login or register for an account.

Search the Community

Newsletter Sign-Up

By signing up you may also receive occasional reader surveys and special offers from Trains magazine.Please view our privacy policy