skip to main content

→ Top Stories:
Clean Power plan
Safe Chemicals

Anthony Swift’s Blog

Tar sands pipeline risks - examining the facts

Anthony Swift

Posted March 30, 2013

, , , , , , , , , , , ,
Share | | |

In early 2011, NRDC raised concerns that an influx of tar sands on the U.S. pipeline network posed greater risks to pipeline integrity, challenges for leak detection systems and significantly increased impacts to sensitive water resources when spilled. Observing a lack of due diligence by industry as it flooded the aging U.S. pipeline system with thick, heavy diluted bitumen tar sands and proposed a major expansion of tar sands transport on new pipelines like Keystone XL, NRDC called on government regulators to identify risks associated with tar sands pipelines and develop safety regulations to address those risks. Since then, information has continued to pile up confirming many of the concerns raised by NRDC – information showing that pipelines moving tar sands are more likely to leak, that leak detection systems are unlikely to detect tar sands spills when they happen, that tar sands spills are significantly more damaging than conventional spills, and that conventional spills response measures are inadequate for containing and cleaning tar sands spills. However, despite the mounting evidence for concern, the tar sands pipeline industry continues to press ahead with their reckless expansion plans while investing in a campaign to avoid due diligence or improved safety standards for tar sands pipelines.


Tar sands spilling from Exxon pipeline, March 29, 2013, courtesy of KARK 4 News

More than two years afte we published Tar Sands Pipeline Safety Risks, here is what we know – and many things we still don’t – about the risks associated with tar sands diluted bitumen pipelines:

Pipelines in the U.S. with longest history moving tar sands diluted bitumen also have worst spill record. Diluted bitumen has only been moved on the U.S. pipeline system since the late 90s and federal regulators still don't provide data with the specificity to evaluate the safety record of pipelines moving tar sands. But a close look at pipeline incident data from states in the northern Midwest, which have seen the greatest volumes of tar sands diluted bitumen over the longest time period, is alarming. Pipelines in North Dakota, Minnesota, Wisconsin and Michigan spilled 3.6 times as much crude per mile than the national average between 2010 and 2012. [1]  

High temperature tar sands pipelines are at greater risk of leaks. Tar sands pipelines operate at higher temperatures than conventional pipelines and high temperature pipelines are more likely to spill due to external corrosion. We know that high temperature pipelines are more likely to rupture due to external corrosion because a small network of pipelines in southern California has provided us with an on point case study. Pipelines serving the Kern River field in California have transported thick heavy crudes to nearby refineries for several decades. In a ten year study of its pipeline network, California regulators found:

“Operating temperature had a significant effect on leak incident rates. Generally, the higher the operating temperature, the higher the resulting incident rate.” California State Fire Marshalls, Pipeline Risk Assessment, 1993. [2]

This study showed that pipelines operating in the range of 130°F to 159°F were nearly 24 times more likely to leak due to external corrosion and six times more likely to leak from any cause than pipelines operating under 70°F. The California study took into account other factors and found that regardless of pipeline age, coating, or pipeline materials, pipelines with higher temperatures had more spills due to external corrosion. Simply stated, the chemical reactions that cause external corrosion occur at a faster rate at higher temperatures. 

And that happens to be the temperature range at which pipelines moving diluted bitumen tar sands operate. TransCanada’s first Keystone pipeline was permitted to operate at 158°F. In its draft Supplemental Environmental Impact Statement (SEIS), State indicated that Keystone XL will operate at a temperature range between 130°F and 150°F.

This is not a new issue. Enbridge’s tar sands spill into the Kalamazoo River in 2010, resulting in the largest and most expensive onshore pipeline accident in U.S. history, was caused by external corrosion.  Moreover, much of Enbridge’s line 6B, which was one of the first pipelines to move significant volumes of tar sands diluted bitumen into the United States, had to be replaced due to hundreds of corrosion abnormalities. NRDC highlighted the risk of external corrosion on high temperature diluted bitumen tar sands pipelines in comments to U.S. pipeline regulators in early 2011.  And yet, industry’s silence on the general risk of high temperature tar sands pipelines and external corrosion speaks volumes.

 Leak detection systems miss 19 out of 20 spills. In Tar Sands Safety Risks, NRDC identified a higher risk of false alarms for leak detection systems in pipelines moving diluted bitumen tar sands. And indeed, the National Transportation Safety Board’s (NTSB) investigation of the Kalamazoo tar sands spill found that a seventeen hour delay from the time of the rupture and their final shutoff of the pipeline was due to the belief by Enbridge’s control center that the leak detection system was giving a false alarm.  

However, several new reports suggest that pipeline leak detection systems are far blunter instruments than many operators care to admit. An investigation of pipeline accident reports from the last ten years shows that leak detection systems miss 19 out of 20 spills. This problem isn’t limited to small spills – these systems also miss 4 out of 5 spills greater than 42,000 bpd. A Congressionally mandated study of leak detection systems by federal regulators identified major gaps in leak detection systems and U.S. regulations. 

Communities have a right to be concerned by the poor state of leak detection technology as they face industry proposals to move tar sands in new or aging pipelines - particularly ones that transverse sensitive water resource. After all, when a spill happens, odds are they’re going to be the ones to find it.

Tar sands diluted bitumen spills are more damaging and difficult to clean. The 2010 Enbridge tar sands spill into the Kalamazoo River highlighted an industry that was unprepared to address the unique challenges associated with tar sands diluted bitumen spills. Nearly three years after Enbridge spilled a million gallons of tar sands crude into the Kalamazoo River watershed and almost a billion dollars has been spent on cleanup, and 38 miles of that river are still contaminated.

Tar sands diluted bitumen is a mixture of very light petrochemicals and very heavy bitumen. Once spilled in a waterbody, the light petrochemicals – including toxins such as benzene and toluene - gas off, leaving the heavy bitumen to sink. As Inside Climate covered in Dilbit Disaster: Inside the Biggest Oil Spill You’ve Never Heard Of, during the Enbridge tar sands spill in Kalamazoo, Michigan, significant heavy crude sank below the water’s surface and traveled along the river bed. EPA’s on-site spill coordinator Mark Durno summed it up:

“Where we thought we might be winding down our piece of the response, we’re actually ramping back up. The submerged oil is a real story -- it’s a real eye-opener. … In larger spills we’ve dealt with before, we haven’t seen nearly this footprint of submerged oil, if we’ve seen any at all.”

In another press interview:

"This was the first time the EPA or anyone has done a submerged cleanup of this magnitude. I would never have expected... that we would have spent two or three times longer working on the submerged oil than surface oil. I don't think anyone at the EPA anticipated that, I don't think anyone at the state level anticipated that, I don't think anyone in industry anticipated that."

One could argue that companies planning to move billions of barrels of tar sands across sensitive water resources by pipeline should have done due diligence before moving ahead. It is much harder to defend the fact that over two years after the Kalamazoo tar sands spill, neither industry nor regulators have evaluated the risks posed by diluted bitumen spills to the environment or developed measures to mitigate those risks.

Conventional spill response methods have proven ineffective for tar sands diluted bitumen spills. During the Kalamazoo tar sands spill, conventional cleanup methods failed, and in some cases made the spill worse. EPA officials were forced to improvise, using extreme measures to recover oil from riverbeds and the nearby Morrow Lake. The spill cleanup continues, but now EPA officials have focused on ensure new areas are not contaminated, concluding that it would be too damaging to fully clean the nearly 40 miles of the Kalamazoo River that are already contaminated by tar sands.   

Over two years ago, NRDC called for an evaluation of the risks of tar sands spills and improved spill response planning for diluted bitumen spills in close consolation with locate emergency response teams and community. Unfortunately, neither regulators nor industry has made progress in evaluating or addressing the risks caused by tar sands spills. The extent of damage done to the region’s watershed may not be known for years to come. Michigan State University Biologist Stephen Hamilton concluded:

"This kind of crude oil is a complex mix of hundreds of compounds—some known to be toxic—that has not been studied much. We just don't understand the consequences well enough."

Tar sands pipeline industry’s argument against improved safety standards and practices are flawed

Unfortunately, rather than put resources toward badly needed due diligence to evaluate the risks of tar sands pipelines, industry has responded to legitimate public concerns with a public relations campaign founded on flawed studies and vitriolic Op Eds. While good faith research is badly needed, none has been done. The problems with the handful of industry funded studies can be summed up by the fact that they don’t address any of the following:

  • Evaluate the risk of higher leak frequency in pipelines carrying diluted bitumen tar sands at the temperatures at which those pipelines operate. The only study touching on this subject was one that evaluated internal corrosion at low temperatures.
  • Addressing the failure of detection systems to detect and/or locate pipeline leaks. There has yet to be a move by the industry to enact measures to address the major problems the Pipeline and Hazardous Materials Administration (PHMSA) identified in its study of the poor state of the art technology employed to detect pipeline leaks.
  • The behavior of heavier than water bitumen blended with diluents (diluted bitumen) in waterbodies. Enbridge, the company that was responsible for the Kalamazoo tar sands spill, did conduct a study of how lighter than water crude blended with lighter than water crude behaved in a laboratory. The not particularly earth-shattering results were summed up in the InsideClimate News story headline - Dilbit Sinks in Enbridge Oil Spill, but Floats in Its Lab Study.
  • Development of effective methods to clean up tar sands spills when they occur. Over two years after Kalamazoo, neither industry nor regulators have acted to develop measures effectively contain and clean the next tar sands pipeline spill.

Most recently, the Canadian Energy Pipelines Association (CEPA) released with the title “State of the Art Report: Dilbit Corrosivity.” In its press release describing the study, CEPA asserts:

“The Penspen report examined 40 studies addressing the behaviour of diluted bitumen and conventional crude.  In these studies, which spanned over 40 years, the research concluded that diluted bitumen is no more corrosive when compared to conventional crude oil.”

CEPA appears to be using the phrase “study addressing the behavior of diluted bitumen and conventional crude,” as a term of art to mean what the rest of the world refers to as “footnotes.”  And it’s true, the CEPA study has 40 footnotes, including references to an Oil and Gas Journal article, a market forecast by the Canadian Association of Petroleum Producers, a couple presentations by Enbridge and the American Petroleum Institute, and even a couple NRDC reports highlighting greater risks of diluted bitumen pipelines. But this report does not cite 40 studies addressing the risks of diluted bitumen - it merely recount a couple industry studies discussed above, focused entirely on the narrow issue of internal pipeline corrosion at temperatures substantially below those encountered in tar sands diluted bitumen pipelines.  

After the report was released, CEPA’s chief executive cited these 40 “studies” to argue that public scrutiny and further due diligence of the risks associated with tar sands pipelines was unmerited, saying:  

“But, real scientists do not make up myths and they do not take things for granted. They are meticulous. They are thorough.”

This is true. CEPA’s representation of its report is not. More than two years after the Kalamazoo River tar sands spill, industry has yet to take the steps to necessary to evaluate and address the legitimate risks posed by diluted bitumen tar sands pipelines. Over two years after the Kalamazoo tar sands spill and there aren’t many excuses left.  


[1] North Dakota, Minnesota, Wisconsin, and Michigan have 6,416 miles of crude pipeline, or about 12.1 percent of the U.S. total. PHMSA. State Mileage by Commodity Statistics. 2013. Meanwhile, between 2007 and 2010 pipelines in North Dakota, Minnesota, Wisconsin, and Michigan spilled 27,911 barrels of crude in underground leaks, or 40.2% of the 63,987 barrels of crude spilled in the United States from 2010-12. Pipeline and Hazardouns Safety Materials Administration (PHMSA), Data and Statistics, Crude pipelines 2010-2012,

[2] California State Fire Marshalls, Pipeline Risk Assessment, 1993. Pg. 68,

Share | | |


Mike HolmstromMar 30 2013 10:09 PM

I'm hearing conflicting things about the Mayflower AR ExxonMobil Pegasus Pipeline crude spill. One source said it's Wabasca Heavy dilbit, but not from tar sands, another said it is tar sands. ExxonMobil did move in pretty fast to start cleaning.

Yes, higher temperature will accelerate external corrosion on pipelines, basic chemistry there. The mastic types of coatings on older pipelines already have adhesion problems in the long term, higher temperatures may make it worse. Yet, there's many miles of crude oil pipelines using mastic coatings.

J MarkMar 31 2013 12:07 AM

The NTSB report from Kalamazoo shows that over 330 people were treated for health effects from the dilbit. It not only made the water toxic for a couple of years, the dilutents vaporize, causing immediate problems. This is why over 50 families were evacuated for a couple of months.

In the pictures from AR, no one was wearing protective suits and masks. Why didn't the ExxonMobile notify the residents and public safety personnel of the danger? By the way, the Keystone pipeline would be almost twice as big in diameter.

Secondly, are there any updates on longer term health hazards? The industry keeps the toxic cocktail ingredients secret, callling them "trade secrets". Why isn't the NTSB requiring the exact mix of any crude or crude blend? Clearly, the oil companies should be paying into a fund so that the regulators can do appropriate tests to assess the hazard potential to the public, water supplies and environment.

This is so rediculous. How are any of these communities supposed to retain any value in their properties?

Michael BerndtsonMar 31 2013 09:10 PM

Excellent post and very informative. Here's some thoughts from the internetz so take it for what it's worth:

Dilbit physical property variability: tar sands are being produced by surface mining and in situ (in place) extraction. As of 2013 it appears to be about 50/50 for each technique. Mining would bring to the surface every last molecule into the liquid mix before processing and dilution. In situ may produce a lighter raw material given the below ground heat/stripping phenomena; lighter MW hydrocarbons are produced at a higher fraction than heavier molecule hydrocarbon. Above ground processing before dilution may greatly affect dilbit physical property, which may include simple solids separation and dilution on the simple extreme and hydrotreating (syncrude) on the other.

Diluent: diluent ranges from natural gas liquids to apparently diesel range hydrocarbons. This would greatly affect the overall mixture physical properties and how the mixture separates or speciates in the environment (i.e. transport and fate).

The lab study on dilbit: the lab study cited in the Inside Climate paper was poorly designed. It wasn't mechanistic and scalable. This means they didn't control the experiment by breaking it up to study important factors. They attempted to mimic transport and fate in the environment with sort of a hamfisted attempt at emulating mother nature. There's already many good experiments for data and interpretation to date: THE actual real world spills in Michigan and elsewhere.

Transport and fate in the environment: dilbit will speciate first based on the energy associated with "mixing" followed by evaporation, dissolution, adsorption, etc.. A flowing stream will assist in speciation of separable components, i.e. the bitumen and the diluent. Heavy ends will sink to the stream bottom and medium ends will bob up and down through the water column like a lava lamp. Lighter ends will float on the surface. The other mechanisms of transport and fate such as dissolution and evaporation chemical/biological degradation subsequently take control of the fate of the spill. This is seen on shoreline cleanups as well.

If dilbit spills at or just below the surface, it will speciate due to the chromatographic (adsorption/desorption) effects of the soil column. In other words, as a mixture sinks through the unsaturated and saturated portions of the soil column, it will separate into distinct fractions or phases. Upon reaching the water table, lighter ends will remain floating on the groundwater and heavier ends will sink through the groundwater.

Also important is that water in oil (W/O) emulsions result when hydrocarbon liquids enter the environment. Emulsions in creation are assisted by the colloidal properties of clay and other solids liquids come into contact with.

In my most humble opinion, the best analogy for dilbit seems to be creosote, which is a coal tar derivative that has to be cut (diluted) in order to flow. There are many remediation sites regulated under CERCLA and RCRA that have addressed creosote spills and where data is readily available.

Comments are closed for this post.


Switchboard is the staff blog of the Natural Resources Defense Council, the nation’s most effective environmental group. For more about our work, including in-depth policy documents, action alerts and ways you can contribute, visit

Feeds: Anthony Swift’s blog

Feeds: Stay Plugged In