NEPA on Crude-by-Rail: A Paper Tiger?

The trouble with "costs" is that they're so often opportunities, too.

Lac Mégantic clean-up crews (July 2013)

Railroads have scored big as more crude oil has been transported by rail over the last five years.  The Association of American Railroads (AAR) knows it has a problem, though.  As the devastating derailment in Lac Mégantic, Quebec warned in 2013, most of the tank car fleet carrying this crude isn’t up to task.  Lac Mégantic was unique only for its death toll: derailments and tank car failures have grown alarmingly common as more oil-by-rail traffic has emerged.

The DOT 111 Legacy

The maker of the cars that derailed and exploded at Lac Mégantic, killing 47 people and burning down most of the town, toured the site and found that almost every car that derailed was punctured, allowing its cargo to flow out, fuel the blaze, or run off into nearby waters. {See Nat. Geographic’s piece.}

Enter the U.S. Department of Transportation which, unlike our National Transportation Safety Board, can actually regulate these cars and the railroads that use them.  (NTSB’s suggested guidelines in the wake of another tragic derailment in Cherry Valley, Ill., in 2009 went largely unheeded.)

Most of the tank cars in service today are so-called “DOT 111s.”  DOT’s July 2014 proposed rule aims at a phase-out of these legacy cars by October 2017 unless they are operated under speed restrictions or retrofitted to add thermal protection systems, thicker steel plates at the ends, outer steel jackets, and a re-engineered bottom outlet valve to ensure it does not break off and leak during a derailment.

How fast to phase out the outmoded, insufficiently armored 111’s and replace them with safer cars able to withstand more of a crash?  Being an agency governed by Executive Orders 12866 and 13563, DOT prepared a Regulatory Impact Assessment (RIA) for this question, finding that its preferred policy—a phase-out by 2017 with a mix of train operation safety improvements (braking systems, speed limits, etc.)—was risk- and cost-justified.  But some disagree, challenging DOT’s data, its underlying assumptions and its predictions about how the market will respond to a quick phase-out of DOT 111s.

What is 0.002%, anyway?

The Railway Supply Institute, a trade group, estimates that in 2014 there were about 102,000 DOT 111 cars in service transporting oil, ethanol, and other flammable fuels.  U.S. carloards of oil were around 11,000 in 2009.  In 2013, it was about 415,000.  That is a massively increased threat in a short time.

But wait, the American Petroleum Institute (API) and AAR argue.  Oil-by-rail works perfectly 99.998% of the time.  The problem is so tiny!  We shouldn’t over-estimate the risks of DOT 111s.  Indeed, these two teamed-up to offer their own proposed car enhancements.  Theirs would come with a shell that is about a ½ inch thick—which is about ⅛ of an inch thinner than what DOT proposed.  That might seem like a niggling difference, but the cost of hauling that much more steel rather than crude system-wide is quite significant.  And their preferred phase-out would be ten, twelve, or fifteen years—not two.

Maybe it’s worth pondering a point API/AAR argue: with that much more steel to haul in a thicker walled update, an “unintended consequence” would be the need for more train trips, “negat[ing] any additional safety benefits by requiring more trains to pull the same volume of crude.”

It also might not be worth pondering.  The market forces and factors that combine to select train sizing, routing, and inter-modal loading, I gather, are complex and dynamic.  The idea that gross weight differences like this could predictably shift these dynamics is more faith than fact.

But this tank wall thickness debate is especially interesting from within the practice of risk/benefit balancing.  Ideally, finding some hypothetical “sweet spot” of thick-enough-but-not-too-thick is possible with the right data, marginal analysis, and understanding of market forces. {See, for example, this Bloomberg piece.} In practice, it is incredibly squishy work—one that risk governance professionals rightly do with a healthy dose of skepticism.

A quick phase out of the DOT 111 would be costly (if not “infeasible” as API says).  This is always the case when environmental priorities demand a turn away from prevailing technology.  The API’s consultants even put the number at around $45B in “costs” to consumers.  Those cost figures are calculated assuming that the “shop capacity” to build the substitute cars doesn’t exist.  But that is something car manufacturers dispute—and who knows better about this number: the manufacturers or API?  Indeed, for an economic opportunity of such immensity, it is hard to believe that future market actors won’t move and move quickly to benefit from such regulation-mandated “costs” by filling a need and thereby lowering costs through the invisible hand of demand, supply and competition.  Only in the land of opposing regulatory precautions are some market forces suspended while others are amplified in the name of prudence (not profit!).  On its best days, NEPA hardens agency decision-makers against that very failing in our system.  In our next post we’ll dig into the differences between DOT’s “Environmental Assessment” and its RIA pursuant to Executive Orders 12866/13563.


I teach environmental, natural resources, and administrative law at Penn State Law. Before teaching I was an enforcement lawyer at U.S. EPA. Along the way I've done work for environmental nonprofits and written a fair bit about NEPA.
No Comment

Leave a Reply