For many Americans, railroad transportation rarely appears on our mental radar, and when it does, it often comes with a negative context: either we are stopped at a railroad crossing while running late for work or we come across a news article with shocking images of smoke and flames accompanied by reports of exploding tank cars.
Months go by with no thought of railroad transportation, until another derailment occurs, and we again associate trains with fire and danger.
With US crude oil production nearing all-time highs – averaging over 8.5 million barrels per day in 2014 – many are expressing fears about the potential of a crude oil spill in their community. And last week, the National Transportation Safety Board released “urgent” recommendations to promote the safety of shipping crude oil, ethanol and other flammable materials by rail.
What’s behind this rapid rise in oil-by-rail transport? How dangerous is it and can new technology make it safer? To answer these questions, we’ve prepared two articles on transporting oil by rail in the US. The first explores the economic drivers and assesses the rail industry’s record on safety; the second evaluates the technology, research and railroad operating practices that can lead to the greatest level of public good.
By better understanding the underlying issues, we can have a meaningful dialogue and take action towards the common goals of improved safety, security and economic stability.
Economics of oil by rail
Oil production in the US is booming. Last year, for the first time since 1987, annual US field production of crude oil topped three billion barrels, a 170% increase since 2008.
Technological advances such as hydraulic fracturing, or “fracking,” and horizontal drilling allowed for increased production, notably in the Bakken formation in North Dakota. But rail has been integral to the domestic oil surge. It was the availability of low-priced, flexible transportation that allowed crude oil to be shipped to US coastal refineries, creating the market for Bakken oil.
As pipelines quickly reached capacity, oil shippers turned to the railroads, which provided multiple incentives, including: flexibility in shipping options and contract timelines, shorter transit times to the refineries (five to seven days by rail compared with 40 days by pipeline), and the ability to choose which refineries to use. While pipelines allow for higher volumes to be transported, the higher speed afforded by rail results in reduced transit time for long distances.
Association of American Railroads, Author provided
As a result, Bakken oil production increased from 81,000 barrels per day in 2003 to more than one million barrels by mid-2014, with more than three-quarters of those barrels moving daily out of North Dakota by rail. While carloads of crude oil increased dramatically, on the whole it still comprises a relatively small portion of total railroad shipments – only about 1.6% of all carloads for US Class I railroads.
Because of increased domestic production and increased imports from Canada, 66% of US oil demand is now sourced from North America, a shift that’s lowered imports and will create billions of dollars in economic activity over the next several decades.
Assessing the risk
While there exists no universally accepted definition of risk, it is widely accepted that the risk associated with transportation of hazardous materials must factor in both the probability of a release of the hazardous material as well as the magnitude of the consequences of that release.
Statistically, the probability of an oil train derailment is very low and lower than other forms of transportation (see figure, below). But the potential undesirable consequences are relatively high, including damage to human life, property and the environment.
A worst-case scenario occurred in the Lac-Mégantic accident of 2013 in Canada, which resulted in 47 fatalities, another 2,000 people evacuated from their homes, almost 1.6 million gallons of crude oil released and millions of dollars in property damage.
Since 2013, three other notable oil train derailments have occurred in Canada, including recent derailments in Ontario, and seven in the US, including the recent derailments in West Virginia in February and Illinois in March. Using data available from the FRA Office of Safety Analysis, here is a summary of statistics for US crude oil train derailments from 2013 to 2014:
eight derailments were reported involving a crude oil release
two of these derailments resulted in a release exceeding 450,000 gallons
two of these derailments resulted in a release between 15,000 and 30,000 gallons
the remaining four derailments resulted in a release of 5,000 gallons or less
injuries were reported in only two derailments, resulting in four total injuries
no fatalities were reported in any derailment
people were evacuated in three of these derailments, with the number of people affected ranging from 16 to about 1,000 people
track and equipment damage exceeded $1 million for all derailments, with only one derailment resulting in more than $5 million in damages (damage to private property or depreciation of property values not included).
In terms of hazardous materials risk, the consequence to human life was very low in these US incidents, with zero fatalities and only four reported injuries.
Environmental and economic impacts, however, were substantial. Recent reports have noted that the amount of oil spilled in 2013 alone from train derailments, at more than 1.1 million gallons, was greater than the total amount of oil spilled from 1975 to 2012. As demand for crude oil shipments has increased, railroads have shifted to using “unit trains” in which nearly every car carries oil instead of the variety of railcar types found on a manifest train. For unit oil trains, the only cars that are not tank cars are the “buffer cars”, typically located in the front and rear of the train to provide an added level of safety for the train crew in the event of an accident.
The use of unit oil trains has resulted in larger amounts of oil being spilled in a single derailment. For example, the majority of oil released in 2013 resulted from only two derailments, occurring in Aliceville, Alabama, in November of 2013 and Casselton, North Dakota, in December of 2013. The recent accident in West Virginia on February 16 of this year likely resulted in a release of similar magnitude to the 2013 spills. The accident in Galena, Illinois on March 5th of this year resulted in a spill of over 200,000 gallons of crude oil released from seven tank cars.
The other side of the coin for risk assessment is the probability of release, which is extremely low when compared with other transportation modes. In 2013, which was the worst year to-date for oil train derailments, about 28,000 barrels of oil were released from railroad tank cars out of the approximately 300 million barrels of oil delivered by rail.
In other words, less than one hundredth of 1% of the volume of oil transported by rail in 2013 in the US was released into the environment. According to an analysis of US oil spillage, the amount of oil spilled by railroads per billion ton-mile transported declined by approximately 85% throughout the 1990s and 2000s. By comparison, pipelines experienced closer to a 40% decline in oil spilled per billion ton-miles over the same period. While this report does not include the recent increase in unit oil trains, it does provide a valuable comparison across transportation modes.
Due to changes in safety culture and numerous technological advances, railroads have continued to improve safety over the last decade, with accident rates reaching all-time lows in 2014 at only 2.24 train accidents per million train miles. The industry has been clear about its goal to continue to use new technologies and improved operating practices to drive accident rates even lower, asserting that “No accident, big or small, is acceptable."
In our next piece, we’ll look at some technologies that can improve safety.
Bryan W Schlake does not work for, consult to, own shares in or receive funding from any company or organisation that would benefit from this article, and has no relevant affiliations.
Authors: The Conversation