Last week in Portland a group of senators, city and state officials and local emergency responders discussed the threat of oil train accidents in their communities.
They’re worried about a problem that happens .0023 percent of the time, according to the rail industry.
In the discussion of oil by rail, railroads and industry groups often cite that “99.9977 percent” of all rail shipments of hazardous material reach their destination “without a release caused by a train accident.”
Where does that number come from? And if it’s true, why does it feel like I read about an oil train derailment every other week?
The Association of American Railroads calculates this number based on comparing its own data and data from the Federal Railroad Administration. The AAR keeps numbers on how many car loads move each year, and the FRA knows how many car loads reached their destination without releasing material.
Divide one by the other, and you get AAR’s 99.9977 percent figure.
It’s based on some of the best data available. It’s not inaccurate. But there’s more to the story.
For one, the AAR figure is only as recent as 2012 and focuses on hazardous materials, not specifically crude oil. It doesn’t cover last year’s boom, in which railroads shipped more than 11 billion gallons of crude oil. More than 1.15 million gallons of oil spilled from those rail shipments. (McClatchy News in DC reported last week that’s more than in 1975-2012 combined.)
So in 2013, close to .01 percent of oil in rail shipments spilled. Still a tiny percent, but also more than four times the industry line.
In past years, that small percentage represented one or two incidents per year. But the surge in shipments means that even a fraction of oil trains releasing crude amounted to more than 100 rail incidents with crude spilled in 2013, according to data from the Pipeline and Hazardous Materials Administration.
The seemingly small incident rate leaves room for massive oil spills. Last year, the 748,000 gallons that spilled from tanker cars in Aliceville, Ala., put that derailment among the largest on-shore spills in the U.S.
All of these figures also are based on car loads, or actual volume of material, rather than the number of actual trains. That’s because trains vary in size and because it’s “more meaningful to count the freight units — measured in either carloads or intermodal units,” because it is closer to the true volume, AAR spokeswoman Holly Arthur said in an email.
It’s meaningful and certainly much more meaningful to the rail industry to count car loads, but if oil trains are passing through your town there’s a good chance you care more about the likelihood one of those trains will have an accident rather than how many individual car loads reach their destination without a release.
“If anything goes wrong, we would be very, very very vulnerable,” Steve Massey, a city councilman in Rainier, Ore., said at last week’s meeting in Portland. The Portland & Western Railroad tracks cut directly through the heart of his town. “We hate to even think about anything going wrong and we hope that there’s a moratorium put on these trains until we get these cars refitted.”
He’s referring to the DOT-111, a dangerous tank car model used to transport crude oil. An oil train can contain more than 100 tank cars, meaning that while the rate of accidents for oil trains is still small, it’s much higher than .0023 percent.
With oil shipments increasing on Northwest rail lines, what’s the likelihood that one will spill or catch fire? I haven’t seen that specific question addressed yet.
It’s difficult to draw comparisons between rail and other forms of transportation, because their total volume is estimated by difference sources, and what each industry has to report is different. Pipelines, for instance, only have to report spills larger than five gallons or spills that result in injury, death or damages over $50,000.
I’m sifting through data sets showing spills and total volume moved by mode of transportation, and I’m looking through some previous studies that have been done on the topic. Some indicate rail is safer than pipelines and much safer than trucks. Expect more on that soon.
— Tony Schick