skip to main content

→ Top Stories:
Clean Power plan
Safe Chemicals

Danielle Droitsch’s Blog

Report: Keystone XL tar sands pipeline more of an economic liability than benefit

Danielle Droitsch

Posted March 12, 2012

, , , , ,
Share | | |

A new report from the Cornell University's Global Labor Institute shows how the Keystone XL tar sands pipeline is an economic liability with the potential to cause significant job losses from a major tar sands spill.  Because tar sands oil is more corrosive and toxic than conventional oil, it can increase the frequency of pipeline spills.  Moreover, a  tar sands spill causes far more damage than a conventional oil spill. Take, for example, the 1.2 million gallon tar sands spill on the Kalamazoo River in Marshall Michigan in 2010 where the clean up costs have been 10 times higher than a typical conventional oil spill.  While there has been a lot of attention to the possible jobs created from the Keystone XL pipeline – far less than what proponents claim – there has been very little attention to jobs that could be lost from a tar sands spill. Keystone XL is expected experience up to 91 significant spills over a 50-year period.    Which jobs are at risk?  Hundreds of thousands of workers in the agricultural and tourism sectors contribute ten of billions of dollars to the economy in the Keystone XL pipeline states.  The Cornell report helps illustrate yet one more reason why the Keystone XL tar sands pipeline should be rejected.   


Here are some of the key findings from the report:

Tar sands spills more likely

Tar sands oil is more corrosive and toxic than conventional oil and can therefore increase the frequency of pipeline spills.  According to the Cornell report, “Between 2007 and 2010, pipelines transporting tar sands oil in the northern Midwest have spilled three times more per mile than the U.S. national average for conventional crude.”

Keystone XL likely to experience significant spills

An independent analysis conducted by the University of Nebraska concluded that Keystone XL over a 50-year period is expected to experience 91 significant spills (greater than 50 barrels).  In fact, the University of Nebraska study found Keystone XL could spill as much as 6.9 million gallons of raw tar sands crude oil at the Yellowstone River crossing.  In just its first year of operation, the first Keystone pipeline operated by TransCanada has spilled 35 times in the United States and Canada in 2010.  This spill frequency is 100 times higher than forecast by TransCanada.

A spill from Keystone XL threatens jobs and the economy in pipeline states

While tar sands spills can have a tremendous impact on the environment, a tar sands spill on the Keystone XL pipeline through America’s agricultural heartland could cause significant economic damage and job losses.  The farming, ranching, and tourism sectors are major sources of employment along the pipeline’s route employing 571,000 workers with an output of $76 billion.  The pipeline will also cross over 90 miles of recreational lands in the pipeline state including state parks, national historic trails, and wildlife refuges.  

“Despite TransCanada’s assurances, we know there will be leaks and spills…It is not a matter of it, it is a matter of when, how often, and how much leakage there will be…When a leak happens, it will be [the farmers’] drinking water, their livestock water supply, and their irrigation supply that will be contaminated.  Their economic well-being is directly impacted by spills and leaks.” Nebraska Farmers Union

Tar sands spills more devastating than conventional oil spills

The Cornell report also looked closely at the largest tar sands spill in U.S. history on the Kalamazoo River in Michigan in 2010 where the costs have escalated to $750 million – 10 times as much per litre as conventional crude.  The clean up of the Kalamazoo river spill which has lasted almost two years has been especially difficult because conventional oil response techniques have been ineffective according to the EPA.   Today, 20 months since the tar sands spill, the entire length of the river (35+ miles) remains closed.   While conventional oil floats on the surface, tar sands is thick and heavy and sinks in water making it very difficult to clean up.

“Enbridge compensated us for the initial shutdown of our business, but we are concerned about the long-term impact that the spill has had on our business…one and a half years later our business is still suffering financially…" Debra Miller, Carpet Story Owner

An accurate assessment of the economic risks still needed

Ultimately, the report said that while there has been significant attention to the Keystone XL pipeline’s potential to create jobs, “scant attention has been given to how existing jobs and economic sector would be impacted from Keystone XL leaks and spills.”  The report said a more detailed risk assessment of the Keystone XL pipeline – one that considers job losses and economic harms from one or more tar sands spills – has not been completed.

Until such as assessment is completed and a full accounting of potential job losses from pipeline spills are considered, the Obama administration should not issue any approvals allowing TransCanada to move ahead with construction of the pipeline.    

Share | | |


John F. DunbarMar 13 2012 11:42 AM

The jobs argument mounted by fossil fuel proponents is completely false, especially at this time. For instance, most of the jobs created by building a pipeline are short term construction jobs and would be immediately replaced by construction of alternative energy infrastructure. Longer term jobs required for operating and maintaining this new infrastructure would more than replace the jobs required to clean up after pipeline disasters. Looking at the still longer term, workers in the plastics, pharmaceutical, paint, cosmetics, lubricants and other industries would lose their jobs earlier than necessary as the raw material for their products is depleted by burning, an absolute waste.

Don BaumhefnerMar 13 2012 01:46 PM

I would like to know about the uses of tar sands products. Can it actually be refined to make it useful for cars? Also, how much more expensive is it to produce than oil out of a drilling rig in America? Thanks, Don

BSMar 13 2012 02:20 PM

Don--Yes, oil sands can be used to make the same petroleum products you get from other types of oil. Processing the oil is more costly (breaking down the size of some of the heavier molecules), but that is more than offset by the fact that the oil itself is cheaper.

John Dunbar--Plastics only accounts for a very small % of oil use. Also, a lot of plastics come from natural gas liquids which are really not even present in tar sands. So using tar sands technically has no impact on the plastics industry.

BSMar 13 2012 02:47 PM

So, do you really expect anyone to think that report from Cornell was even a little bit objective? It read more like an NRDC report than something that would have a major university's name on it.

The report from the University of Nebraska at least seemed to make an attempt to actually be objective. However, it was still written by someone who does not understand pipelines.

For example, he cites some report saying that a 1.5% leak could go undetected for 90 days. First, that was from a 2006 report and had nothing to do with Keystone. Second, measurement standards on pipelines today require accuracy to within +/-0.5%. A 1.5% leak would be detectable by that measurement and would not take anywhere near 90 days to find.

There were a number of errors like that in his report.

Lastly, the Nebraska report made use of the phrase "higher temperatures and pressures" to describe how the Canadian oil is pumped. This phrase is patently false and suggests that there is also bias in this report.

The Keystone lines are designed to operate at the same pressures as just about any other liquid pipeline system in the US. The elevated temperatures are simply due to laminar flow which reduces heat loss. This reduction in heat loss actually makes the fluid easier to pump, which is a good thing.

Also, the elevated temperature is something like 80F. The ground temperature in the southern US hits 80F every year. So these high temperatures are not unheard of. And lastly, some crudes, including heavy California curde, are shipped at temperatures as high as 180F without any problems.

QuijoteMar 13 2012 06:23 PM

How can such BS be published without any scientific data to back it up. Keystone pipeline has never LEAKED there have been small fitting leaks at pump stations and valve sites big difference! Also keep in mind energy companies have to report every drop they spill and clean it up, they are good corporate citizens too bad the general pubic doesn't clean up all their messes instead of washing it down the sewer.

Carrie La SeurMar 13 2012 06:28 PM

Where do you get the 35 spill number for Keystone I? Can you substantiate that?

For what it's worth, the Canadian stretch of pipeline wasn't new. TC built Keystone I where it is because there was an existing pipeline across the prairie provinces.

And I'm surprised that neither Prof. Stansbury nor you mention that KXL would cross the Yellowstone with horizontal directional drilling for KXL, i.e., they'd drill through bedrock under the river. It seems to make a difference for worst impact spill calculations. Have you read the EIS?

BSMar 13 2012 09:59 PM

"Have you read the EIS?"

LOL. I think we know the answer to that question. I'll give them credit. They are trying a little harder and making it slightly harder to refute their BS than just a couple of months ago. But not by much.

Woody PfisterMar 14 2012 08:31 AM

NRDC and their minion, President Obama are blocking $billions of shovel ready jobs! Plus causing working people to pay $5+ per gallon for gas.

The only one who benefits from blocking the Keystone XL pipeline is the Obama puppetteer WARREN BUFFETT and his BNSF Railway.

Anthony SwiftMar 14 2012 10:41 AM

Hi Carrie –

The 35 spill number for Keystone I includes 14 spills in the U.S. portion (which have been well documented and are included in the EIS) and 21 on the Canadian side is from an access to information request by Mike De Souza of Post Media News, reported here:

And you’re right, much of the Canadian stretch of the pipeline was a natural gas line that was converted for use as a liquid pipelines.

KXL will cross the Yellowstone river using horizontal directional drilling (HDD) – it’ll be below the riverbed in silt and sandstone layer. This will protect it from anchors and flooding. However, when there are problems, detection is far more difficult - keep in mind than Prof. Stansbury estimated the worst case spill on Keystone XL to be one that fell below the pipeline’s leak detection threshold and relied on a biweekly flyover for discovery. The length of his worst case scenario was a spill that continued for 14 days. However, the problem with HDD is leaks can take significantly longer to detect.

And regarding Keystone XL’s spill detection – here’s what the Final EIS has to say:

“Software based volume balance systems would monitor receipt and delivery volumes and would detect leaks down to approximately 5 percent of pipeline flow rate. Computational Pipeline Monitoring or model based leak detection systems would monitor small pipeline segments on a mass balance basis. These systems would detect leaks down to approximately 1.5 to 2 percent of pipeline flow rate. Computer based, non-real-time, accumulated gain/loss volume trending would assist in identifying seepage releases below the 1.5 to 2 percent by volume detection thresholds.”

Keystone XL FEIS, 3.13.60

Moreover, Keystone XL’s non-real-time accumulated gain/loss system is governed by Canadian standards (both because the spill response center is in Calgary and one of the “special conditions” in the EIS notes that Canada’s standards will be used). Those set a threshold for non-real time leak detection at 2% per week or 1% per month. By comparison, Prof. Stanbury’s worst case estimates were conservative.

As far as temperature goes, the Keystone XL pipeline will operate between 130 and 150 degrees F – take a look at the figures on L-4 to L-18 in the EIS’s pipeline temperature effect study and you’ll see the pipeline is running in that range. Also note that another special condition that was placed on Keystone XL was that it not operate continuously above 150 F without a waiver from PHMSA. If the pipeline was operating a few degrees above ambient temperatures, as many pipelines do, there would be little point to the condition (which was billed as making the pipeline safer).

Keystone Final EIS Spill Temperature Effects Study, L-4 to L-18

As far as the leaks on Keystone 1, one was approximately 21,000 gallons, or 500 barrels. At high pressures, a small rupture can result in a big leak – this one was called in by a farmer that saw a 60’ geyser of oil over the cottonwoods trees.

BSMar 19 2012 08:59 AM

Anthony, as a supposed lawyer, I'd expect you to try harder.

Please click on the link you posted after your last paragraph above which describes the 500 barrel leak:

That one link you posted basically proves everything else you said to be wrong. Trans Canada's leak detection system detected the leak which was much, much less than 1.5% of the pipeline flowrate. It was more like a pinhole leak. And while a farmer did call it in, the pipeline was already being shut down by the time the farmer called.

Based on the report that you yourself posted, do you still want to claim that a leak of less than 1.5% would go a long time without being detected?

With respect to the temperature issue, I'm suprised to see that the temperature will be so high. However, the point I'm making is that NRDC is being misleading to it's readers when they claim that the bitumen must be shipped at "high temperatures and pressures" as if it must be heated in order to flow and that it will be shipped at pressures higher than what other pipelines are normally designed for. This is not true.

The temperature rise is due heat gain from the friction losses in the pipe and heat gain from pumps (because pumps are not 100% efficient, so they produce some heat). This happens in all pipelines. The difference with Keystone is that the flow is not turbulent, which allows the oil to hold on to that heat better (vs releasing it to the ground). This actually significantly reduces the amount of energy required to pump the oil by making it less viscous.

However, just like your link to the report on the 500bbl leak, you have also misrepresented this temperature report. The pipeline will not "operate between 130 and 150 degrees F". According to Figure 1, the pipeline fluid temperature would exceed 130F only on the last 600 miles (30% of the pipeline). And even then, it would only be 130-135F on the hottest days of the summer. The other 90% of the year, the temperatures would be lower.

You can argue with me all you want, but you're never going to get anywhere as long as you continue to distort the facts.

Florian SchachMar 19 2012 05:23 PM

While it is perhaps controversial in its nature I believe that there’s two sides to every project and with this, there is something that we as a nation can gain from it both now and in the future. Right now the United States is still facing a labor crisis and according to recent data is experiencing lagging and not upward growth, this needs to be rectified through a series of projects and initiatives of which Keystone happens to be a part ( The project would employ people directly and immediately which would help add to the number of jobs in the economy overall. The pipeline should also be seen a s a way to bridge the gap between our current needs and our future desires not as a totally negative vein from which nothing good can come.

Anthony SwiftMar 19 2012 05:32 PM

I don't necessarily expect to get anywhere with you. But I will just throw a couple of points out for consideration:

First, Keystone 1’s SD leak was well above the 1.5% threshold of TransCanada’s leak detection system – the leak lasted from 3:51 am to 4:35 am, or 44 minutes. In that time it spilled 500 barrels of crude. That’s a spill rate of 680 barrels an hour – or 3.2% of the flowrate of a 500,000 barrel per day pipeline like Keystone 1 at full capacity. And Keystone wasn’t at full capacity, so the percentage of oil flowrate spilling was even higher. This was the kind of spill that we expected TransCanada to catch.

That said, it’s unclear that TransCanada did catch it. The pipeline’s shutdown procedures take nine minutes.
According to the investigation, the landowner called at 4:26 am, shutdown was started immediately and the pipeline was completely shutdown by 4:35 am. Given that abnormalities show the spill began at 3:51 am, it’s hard to say when the company would have initiated its shutdown procedures – according to the landowner, TransCanada was not aware of a problem when he called. One of the technical issues with diluted bitumen pipelines is the varying viscosities, pressures and temperatures in the pipeline create alot of noise. There are frequently pressure abnormalities – and frequently they are not due to leaks. That makes detecting real leaks all the more difficult.

But the larger concerns are leaks less than 1.5% on these pipelines. On a 830,000 bpd pipeline like Keystone XL, that would allow a 12,400 barrel a day leak to go undetected – over half a million gallons. And that is legitimate cause for concern.

The issue we have with pipeline temperature is simply that these pipelines are operating at higher temperatures – and that those temperatures affect some of the more problematic properties of diluted bitumen and have potential long term impacts on pipeline integrity that have not been considered. TransCanada has responded to this legitimate concern with the red herring that it “doesn’t heat” the pipeline. That’s all well and good – our contention was always that heat was generated by friction caused by high viscosity (thickness) of tar sands and pressures. Whether that’s heating the pipeline in an active sense or not is immaterial – the important issue is that these pipelines are hot.

In California, where there are also relatively hot pipelines, studies have shown a direct relationship between higher temperatures and leakage rates – particularly due to external corrosion. Meanwhile, U.S. pipeline safety regulations don’t differentiate standards based on temperature. And that’s a problem.

BSMar 19 2012 09:51 PM

Anthony, you are making a number of assumptions about the leak.

You are assuming that it started at 03:51. An imbalance was detected at that time meaning the leak would have started some time earlier than that.

You are assuming the leak was at a constant flowrate. A leak is more likely to start smaller and grow.

You are assuming the leak stopped as soon as the line was shut down. In reality, the leak would continue until the pressure in the line was insufficient to force oil out of the opening.

You are assuming that the top rate of a pipeline is based on pumping 24hrs/day. In reality, a pipeline's stated capacity is based on pumping 22 or 23 hours/day to allow for planned and unplanned downtime. So the top flowrate is a little higher than you have assumed.

Of course, the pipeline may not have been running at top rate when the leak occured. That assumption is in your favor. But then again, the stated capacity in May 2011 was (and still is) 590Mbpd, not 500Mbpd.

With respect to the temperature/corrosion issue, pipeline companies are required to routinely conduct internal inspections of their pipelines for metal loss due to corrosion. If metal loss is found, the pipeline's operating pressure is derated until the problem is corrected. There are also many steps that a company takes to prevent corrosion. Given a multi-billion dollar investment with a low rate of return, TransCanada has a vested interest in keeping their pipeline operating safely.

Do you agree that the NRDC was in error when it claimed that dilbit has to be shipped at higher pressures than other types of oil?

BSMar 20 2012 09:12 AM

Also on the subject of temperatures, you neglect that the temperature study has assumed a worst-case scenario of all heavy crude. In reality, Keystone XL would be shipping very light Bakken crude as well as the Canadian crude. As a common-carrier pipeline, it may ship other varieties of much lighter domestic crudes as well. These crudes require much less energy to pump, meaning much less heat due to friction.

So in practice, the true temperature profile is likely to be lower than what is shown in the environmental report.

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: Danielle Droitsch’s blog

Feeds: Stay Plugged In