A new rule will allow us to see just how much financial risk oil trains are foisting onto the taxpayers.
Shipping crude oil by train is an extraordinarily dangerous enterprise. Notoriously prone to leaks and spills, recent history has shown that railcars can explode catastrophically when the fuel comes in contact with an ignition source. The damages can be profound. To make matters worse, the railroads that run the oil trains—and that are legally liable for damaging incidents—are chronically and severely underinsured. In fact, Sightline has documented extensively for the public what is something of an open secret in the industry: that even the biggest railroads do not carry insurance proportional to the risks of their cargo.
In the worst oil train incident to date, a July 2013 derailment in Quebec, the resulting inferno killed 47 people and did roughly $3 billion in damage to the small town. The railroad responsible, the Montreal, Maine and Atlantic Railway, which carried a scant $25 million in insurance, filed for bankruptcy almost immediately, sticking Canadian taxpayers with the tab. It’s a risk that looms large for the public in states like Washington, which host to oil trains every day. But now, a new rule will at least allow us to see just how much financial risk oil trains are foisting onto the taxpayers.
Defining a “reasonable worst case spill”
Following on the heels of an oil transport safety bill signed by Governor Inslee in May 2015, the Washington Utilities and Transportation Commission (UTC) adopted new rules in early 2016 to increase the safety of oil train transportation. The rules also require that any railroad transporting crude oil in the state must include financial information in their annual reports to the UTC to show that the company could pay the costs of cleaning up a “reasonable worst case spill.”
But what does it mean to prepare for a “reasonable worst case spill”? The new rules require oil-hauling railroads to show that they can pay the costs, whether through insurance, reserve accounts, letters of credit, or other financial instruments or resources on which the company might rely. But to know how much that might cost, one first must define a “reasonable worst case spill.”
The worst case is widely considered to be a real-life event: the Quebec derailment that, along with killing dozens of people, spilled 1.6 million gallons of oil. That was a starting point for the UTC’s estimates, but the commission scaled down its estimates from there, looking to states like California that have adopted similar rules and also a similar analysis done by the US Pipeline and Hazardous Materials Safety Administration (PHMSA), the federal regulatory agency that oversees oil train safety standards. California, for example, defines a “reasonable worst case spill” as the loss of 20 percent of the oil cargo that a railroad can transport on a single train. The UTC took its cues from a scaled-down methodology created by PHMSA.
Cleanup calculations
To calculate the “reasonable worst case” amount of oil that might spill, the UTC proposed a simple mathematical formula: take the top speed in miles per hour of oil trains operated by the railroad, divide by 65 mph (the speed of the train when it derailed in Quebec), and then square that number to factor in kinetic energy. For example, a railroad that operates oil trains at a maximum speed of 45 mph would divide that speed by 65 and then square the result to conclude that a “reasonable worst case” derailment could result in spilling 48 percent of the oil cargo on a single train.
The next step is to calculate the cost of cleaning up a reasonable worst case spill. The UTC set the minimum cost at of cleanup at $400 per gallon. So to come up with the cost one would multiply $400 by the reasonable worst case spill percentage (calculated on the largest oil train the railroad moved in the past year).
Let’s do the math. Some oil trains have as many as 120 cars, and each car usually carries about 30,000 gallons of oil, so a 120-car train could carry up to 3.6 million gallons of oil. Assuming a top speed of 45 mph, a railroad should be prepared to clean up 48 percent of its cargo, or 1.73 million gallons of oil. At $400 per gallon, the railroad should be able to pay for $691 million in cleanup costs.
Still not enough—by a long shot
Yet even these hundreds of millions of dollars in potential damages are far too low, as the UTC acknowledges. It does not take into account loss of human life, property damage, or other factors. In fact, the UTC admits that the $400 per gallon cost does not capture the “full comprehensive societal damage” that results from an oil train crash. The costs of the Quebec derailment, hardly a theoretical exercise, were more than four times as high.
By contrast, the Genesee and Wyoming railroad that serves Grays Harbor, Washington—and that aspired to host as many as 17 oil trains each week to proposed port terminals—carries at most $500 million in insurance. Tacoma Rail, which delivers 4 oil trains each week through Tacoma’s busy industrial port, has less than $100 million in coverage.
And what can the UTC do with the financial responsibility information when rail companies’ first reports arrive to them next month? Not a lot. The same bill that enabled the UTC to write the rules also prohibited the agency from using the information as a basis for economic regulation or penalization of the railroad.
The new rules will provide state taxpayers with valuable information: for the first time we’ll be able to see just how exposed we are to the financial risks foisted upon us by oil trains.
How taxpayers will end up paying for the costs of a worst case oil train derailment.
EM
Great article, and I’m happy to Sightline doing some public education on this issue. However, having studied and written on these issues at length, I would point out a couple of things that I think Sightline could do a better job of addressing:
1) The PHMSA ‘scale down’ methodology references the change in kinetic energy (translated to disaster potential) of a train going 45 mph versus one going 65 mph, rather than the amount of oil likely to spill. It’s representative of the exponential decrease to risk based on speed. Trains carrying crude are not legally allowed to travel faster than 45 mph in Washington, so a lower level of decreased risk is justified.
2) The fifty-percent of a train load as a worst-case spill was a precedent first set by the Washington State legislature. HB 1449 said a “‘Reasonable worst case spill’ from a railroad company means fifty percent of the largest train load of crude oil, as measured in barrels, moved by that company in the previous calendar year.”
3) Because the rule is based on the largest train load of crude actually moved, one could argue that the bill and rule actually incentivize transporting less oil in bulk (fewer cars per train). This would also reduce the potential amount that could be spilled.
4) $400 per gallon was the high end for what PHMSA thought was an appropriate estimated cost for cleaning up a gallon of oil. The studies that PHMSA and the UTC referred to estimated $200 per gallon and around $350 per gallon, if I remember correctly. So $400 per gallon was a conservative per-gallon cost estimate.
5) Tacoma Rail operates at maximum speeds of 10mph (according to the comments submitted to UTC). At that speed, using the UTC’s equation, a train on the Tacoma Rail line that encountered an issue would have only TWO PERCENT of the kinetic energy of the Lac Megantic incident. Also using the new rule, that would mean a ‘worst case spill’ would have clean up costs of $28.8 million. So a $100 million insurance policy seems reasonable.
6) Saying that the bill prohibited the agency from using the info to regulate or penalize the railroads doesn’t tell the whole story. The bill nearly doubled the regulatory fees that railroads pay, based on their gross state operating revenues, to go into a state cleanup account, which is a state-operated fund but is NOT taxpayer dollars.
7) That brings us to the last point, and it’s a fundamental one. The basis of this article is that Washington taxpayers foot the bill if a railroad can’t pay. But the regional railroad that was operating and declared bankruptcy in Lac Megantic is NOT BNSF or Union Pacific, the two main oil transporting rail lines operating in Washington and the two largest freight railroads on the CONTINENT. When I googled it, BNSF made over $23 BILLION in revenue in 2014, and is backed by Berkshire Hathaway, an even larger company that would be financially responsible. Union Pacific had nearly $22 BILLION in 2013 revenue and has $49 BILLION in assets. On top of that, the UTC fee / cleanup account kicks in if and when a railroad can’t pay. While the emotional, physical, and environmental damage of such an accident would be catastrophic, I don’t think it’s clear that taxpayers would be the first ones with financial risks “foisted” upon them in the event of an accident.
Hindsight is always 20/20, and hopefully we’ll never be in a position of looking back and saying “IF ONLY.” But if people really want something to be concerned about, if people are truly concerned about oil train catastrophes, they should be calling Congress and telling it to fund railroad infrastructure upgrades. Degraded tracks are by FAR the leading cause of train accidents, according to several reports by the Federal Railroad Administration.
We take risks when we get into our cars for the convenience of travelling quickly. We also enjoy the comforts of a modern society that is powered by energy, and energy production, transport, and development have risks inherent to them as well. We can all agree that oil trains are dangerous and that companies should continue to do what they can to mitigate the risks. But until people stop driving altogether, I am not convinced that one-dimensional opposition to oil and oil trains is justified.
Samir Junejo and Eric de Place
EM, thanks for taking the time to comment on our analysis. In what follows, we respond to your notes sequentially.
1) The PHMSA ‘scale down’ methodology.
You’re right that the effect of it is to reduce the theoretical risk because train speeds are typically lower than 65mph. You seem to think that we are objecting to PHMSA’s methodology, but we never said that in the post. We’re simply describing it. (It is worth noting, however, that trains do sometimes exceed speed limits, as in the runaway unmanned train at Lac-Megantic that was traveling an estimated 65mph when it derailed in an area with a 10mph speed limit.)
2) “‘Reasonable worst case spill’ means fifty percent.
We don’t agree. ESHB 1449, the bill as signed into law, does not define “reasonable worst case spill.” In Section 10 it merely says: “reasonable worst case spill of oil, as calculated by multiplying the reasonable per barrel cleanup and damage cost of spilled oil times the reasonable worst case spill volume as measured in barrels.” The definition of “reasonable worst case spill” that you quoted comes from an August 2015 draft of the UTC rules, but that language did not make it into the final rule. Rather, the final rule defines “reasonable worst case spill” using the calculation that we described in the article.
3) The rule incentivizes transporting less oil in bulk.
We agree that the rule creates that incentive. If transporting less oil in bulk means an increase in train frequency then it may not necessarily lead to an overall decrease in the risk of spills. But regardless, yes, smaller trains would reduce the potential amount that could be spilled per accident.
4) $400 per gallon is a high-end estimate.
We disagree. $400 per gallon is a very modest estimate. It is just slightly more than the costs of the oil train derailment in Lynchburg, Virginia, an accident that would have done vastly more damage had the tank cars tipped the other direction (into town instead of into the river). Consider that the Lac-Megantic costs of around $2.7 billion resulted from a “spill” (and fire) of 1.6 million gallons, which means that the actual costs of that accident were about $1,688 per gallon.
In a situation where an oil spill results in property damage or loss of life, the per-gallon costs could easily exceed $400. In fact, Sightline’s review of actual legal settlements from train derailments (e.g. the Cherry Valley ethanol explosion and others), actuarial calculations (e.g. from federal terrorism risk analyses), and risk assessments (e.g. PHMSA’s own methods for estimating the costs in an urban area in its Regulatory Impact Assessment in its hazardous substance train rule) finds that a reasonable worst case scenario derailment, particularly in an urban area, could run up to $5 to $8 billion or beyond.
5) Tacoma Rail.
You’re right that the UTC’s estimate of clean-up costs (based on their modest $400 per gallon estimate) would be $28.8 million, which is less than Tacoma Rail’s insurance policy. But you are overlooking what we said in the article: the UTC’s figures do not take into account loss of human life, property damage, or other factors. In fact, the UTC admits that the $400 per gallon cost does not capture the “full comprehensive societal damage” that results from an oil train crash. (The costs of the Quebec derailment were more than four times as high.) Tacoma Rail operates oil trains in a very busy area with lots of people, lots of high-value industrial property, and lots of additional external factors that exacerbate risk. If the UTC is putting the clean-up costs at roughly $29 million before counting loss of life or property damage or other costs, we think that’s concerning. At minimum, it begs for a much keener analysis.
6) The bill increased the regulatory fees that railroads pay.
If you look at Section 10 of ESHB 1449 you can see that the law clearly states that the UTC cannot use the financial responsibility information to regulate or penalize the railroads. Our article is about the financial responsibility portion of the new regulations, not about regulatory fees or the clean-up account, which are entirely separate from the statutory prohibitions on penalization and economic regulation. As for the areas of the bill you mention, yes the law increases regulatory fees. You seem to be talking about the oil spill response account and oil spill prevention account. The bill did change the funding of those accounts, so the oil spill response and oil spill administration taxes now cover transportation of oil by rail, whereas previously it was just oil transported by marine vessels and barges (see: Section 14 of ESHB 1449). That’s interesting as far as it goes, but it’s not germane to our article, which is about the financial responsibility portion of the UTC’s rule.
7) Taxpayers won’t pay for cleanup costs.
It’s hard to know where to begin with this portion of your comment. First off, neither BNSF nor UP have anywhere near the level of insurance they would need to cover the costs of a worst-case scenario derailment, a fact that is widely known in the insurance industry. Second, while it is true that BNSF (the nation’s biggest hauler of crude oil) generates huge revenues this has no relevance to what they would pay for cleanup and, if anything, it only makes their chronic insurance shortfall the more galling. Third, it is absolutely false that Berkshire Hathaway provides BNSF with some sort of financial insurance backstop. Fourth, the cleanup account you reference is far too small to cover the potential shortfalls. Fifth, in Tacoma, taxpayers are directly on the hook because Tacoma Rail itself is a publicly-owned entity, so even if it were liquidated to pay for cleanup that would be a cost to taxpayers.
M.Lee Dodge
This is a perfect case of corporations Privatizing profit and socializing Risk. As such the public needs to recognize that they are the ones that will pay for what the corporation either refuses to cover-by limiting their insurance- or flees from paying thru bankruptcy which is even more costly for the taxpaying public. The question searching for an answer is: what is the best plan to wake up the public to how they are being exploited ?
This is a very complicated effort but it’s a keystone to limiting the public’s exposure in what, without a massive public outcry, the politicians can’t “hear”.
I’m an engineer, not the best expertise to address this important plan and I’m prepared to do what I can to contribute to the process.