The Keystone Pipeline’s XL-ish Spill
Externalities from shipping crude by rail are still disproportionally larger than those from pipeline transport.
My twitter feed exploded late last week. And no, I am not talking about the annoying doubling of the character limit, but the fact that there was a spill of what is currently estimated to be 210,000 gallons (5,000 barrels) of oil from TransCanada’s operational Keystone pipeline in South Dakota. Is this a big spill? From what I can tell, out of the 1542 crude oil pipeline spills the federal government reported since 2010, this ranks as the 13th worst spill. For comparison, the Exxon Valdez spill was 11,000,000 gallons – 50 times larger. Transcanada is also the company behind the proposed controversial Keystone XL pipeline.
The snark was off the charts! See! Pipelines are terrible! We always knew it! Boo Pipelines. (I have blogged about the necessity for the Keystone XL pipeline before and came out against it, for different reasons). The economic and engineering problem, of course, is that oil does not come out of the ground where it is being consumed. So you have to ship it from the well to refineries and from refineries to the final consumer. Your choices for long distance shipping are loud Diesel powered choo choo trains or electricity powered and panther quiet pipelines. Which mode of transport is worse for society? Well, says this annoying economist, let’s back up a second and see what the science says. Isn’t that a novel idea? Scientists know stuff (science comes from the Latin word scientia which literally means knowledge)! And it turns out the science is pretty recent, yet clear on this.
Assume for a moment that you would like to ship a certain quantity of oil from point A to point B. If you (the owner of the oil) ship it via rail, you pay for the cost of operating the train and loading and unloading of the oil. If the train crashes somewhere and there is a spill or an explosion, the local residents are negatively affected by this, which is an external cost you do not have to pay for fully (assuming no recklessness). Same goes for pipelines. If there is a spill, the consequences of the spill are an externality. The literature so far has focused on the consequences of crashes and spills, which is important.
Enter a highly useful paper by Karen Clay, Akshaya Jha, Nick Muller and Randy Walsh. They note that spills and explosions are only one part of the external costs. Diesel powered trains emit a significant amount of local and global air pollution in the process of shipping oil, which are bad for human and polar bear health. Pipelines also consume some energy and generate pollution as pumps need to run to maintain pressure to make what’s in the pipeline flow. Assuming the pumps are electric, this pollution is generated not at the pipeline location, but at the power plant that produces the electricity used. This means the air pollution generated by pipeline operation is not borne by people living near pipelines, but by people living near power plants supplying the electricity. So when we decide whether to ship oil via pipeline or rail as a society, we should take into account the full costs – private cost of shipping and external costs (spills, crashes, local and global air pollution damages).
So what’s the issue here? It’s people. The more people near pollution, the bigger the damages. Railroads tend to go near and through population centers, while pipelines tend to be further away from population centers. Also power plants emit local pollutants higher up (smokestacks), whereas trains emit the pollution at ground level, where the people are.
The paper collects a tremendous amount of data on diesel consumption by trains, routes of rails, location and emissions of power plants. They then calculate the damages from this pollution for all of the oil shipped out of North Dakota in the year 2014 – half of which went by rail and then other half by pipeline. To make a long story short, the authors find that the pollution externalities per barrel mile from rail are almost twice as high as those from pipelines. Maybe more importantly, they find that for rail the damages from spills/crashes are only one-half of the air pollution damages and just one-eighth for pipelines!
What does this mean? As a society if we are going to ship a barrel of oil from point A to B, we are better off shipping it by pipeline than shipping it by rail. However, the externalities from shipping fossil fuels are only a small part of the overall externalities that arise when we combust them to e.g. mine for Bitcoins. The bigger question will always be how we reduce our reliance on fossil fuels. Growing populations, incomes and desires for energy consuming durables will lead to a growing need for electricity, which we hope will come mostly from shiny clean renewables before too long.
So it’s Thanksgiving. I would like to end this post by stating something I am grateful for. I am grateful for what looks to be an insane bonkers below the charts $20.57/MWh auction result for solar in Mexico. There literally are rays of sunshine on the horizon. And to the haters that point to the relatively small penetration of new renewables, I point out to you that adoption of new technologies happens in an S-shaped fashion. Once something new becomes the best game in town everyone wants it. Ten years ago, nobody had a smartphone or saw the need for one. I bet some of you are reading this on your Iphone X screen. Mic Drop.
Categories
Maximilian Auffhammer View All
Maximilian Auffhammer is the George Pardee Professor of International Sustainable Development at the University of California Berkeley. His fields of expertise are environmental and energy economics, with a specific focus on the impacts and regulation of climate change and air pollution.
Energy Institute Blog 
Amen!
What if Keystone XL spills oil that contaminates the aquifers in our heartland? Is TransCanada going to made us whole for that damage? Most likely not; it will just declare bankruptcy. As a precondition to approving construction of that pipeline the federal government should have required Canada to provide a sovereign guarantee to pay for any residual damages that TransCanada did not pay. Maybe then Canada would have prohibited TransCanada from shipping Tar Sands oil to the US, or better yet – prohibit its production. That crap should be left in the ground.
A nice article.
One observation on the last paragraph: There is a tremendous focus on $/MWh price that renewables bid in, and yes they have become quite low, but this does not factor in system integration costs (the externalities caused by renewables) or more simply put the primary reason why there tends to be a significant gap between wholesale and retail electricity prices (transmission congestion pricing, A/S procurement, etc.)
One alternative quick and dirty way to think through this is by considering capacity factors. At a capacity factor of 20%, for example, the price bid by a renewable generator needs to be scaled by 90 / 20 to give a true sense of cost, were the RSE to run as an equivalent to a baseload plant with high capacity factor. Just some food for thought.
Charalampos,
You were on a roll until you suggested that scale-up factor, No, 90/20 does not give us a “true sense of cost” because the capital cost of the renewable resource is fully captured in the $/kWh levelized cost of the energy it produces. If you want to compare the cost renewable energy with that of a base load resource you need to add in the cost of storage needed to “firm up” the renewable energy. As I have repeatedly stated in the past, one cannot directly compare the levelized cost of any intermittent resource with the legalized cost of a base load resource unless you add in the cost of the storage needed to firm up the output of the intermittent resource.
Regarding the cost of integration, what is generally overlooked is that the outputs of renewable resources vary seasonally, e.g., solar output is highest in the summer and lowest in the winter, and are also highly correlated, e.g., wind speeds simultaneously fall off over broad regions as does solar output (like every night). This implies that we will need months of storage capacity to seasonally integrate high levels of renewables to maintain current system reliability. How much will that storage cost?
Mark Jacobson’s study assumes that the needed storage will largely come from hydro resources. However, it is unclear that North America has the potential for developing that much storage hydro.
Robert,
Neither does the levelized cost for baseload generally include the environmental damages, and the increased need for flexible capacity to supplement baseload to meet varying load. As for hydro, in the West the reservoirs could be reoperated to match renewable output patterns instead.
Agreed. The capacity factor goes into the LCOE, but what the LCOE doesn’t capture are the externalities of the resource (in terms of storage and transmission among others). Currently, NG is used as a poor substitute for batteries in order to make up for seasonal and daily variability of renewables. The scale-up factor is by no means a rule of physics or economics, but rather a visualization aid for understanding the effect of the need for firm power to cover the deficiency in theCF of renewables.
Max,
One of the elements you left out of your analysis was the “who” in the scenarios. Fairness dictates that the ones least responsible for a “bad,” should be the ones least affected by its negative consequences. Your purely scientific analysis of the options for shipping oil left out considerations of equity. Who are the majority of people closest to the power plants, and thereby most affected by pollution from them? Who are the people most affected when a Keystone spill contaminates the water supply? They are not the people whose voices are much heard in the debates about how to use or transport oil. Nor are they the people who are proportionately most responsible for the need to extract, refine and burn fossil fuels. They are the poor. Until we rectify the disparity between those who proportionately are a greater element of the cause of the damage, and those who are proportionately most affected, the “scientific” arguments have a hollow ring to them.
I agree on your point that pipelines are a better way to ship crude oil than by rail. What are your thoughts on California since our choice is not pipelines vs rail? It’s tanker ships vs rail if we are not going to produce here in CA. Have you written anything on domestic California crude production that you can point me to that is similar to what you wrote above? Thanks.
Pipelines are the cheapest, least polluting way to transport liquids – even cheaper than supertankers. I think the more relevant question is why are we transporting any of that toxic, heavy Canadian oil to the US? There is a worldwide glut of crude oil today without the Tar Sands crude. Canada should not even be producing that bad stuff.
With regards to Keystone XL, the issue is not with pipelines per se but more the “need” of this facility and the benefit to the US. There was no real need demonstrated for this pipeline in the US. The need is more for the Canadian oil company to export their product out of a warm water port. The other opposition had come from the fact that there was little ancillary benefit aside from construction, e.g. jobs related to running this facility. So the question is, why risk the American environment to transport a foreign “good”?
@Dave Perrino, agree across the board.
Also, in any discussion of Alberta’s oil sands, we only hear about “1 pipeline”, by which all other metrics are based. Reality is, for every barrel of “oil” out of Alberta, there are 3 other pipes required; 1) for the water to “steam” the oil out of the sand, 2) for the natural gas to boil the water for #1, and 3) to supply NGL (natural gas liquids) or ultralight crude to “dilute” the bitumen so it has a chance in hell of actually flowing in the pipeline – which also requires heating too.
Then there’s always the question as to how much the “special” relationship between Alberta & GoM refineries constrains or otherwise negatively impacts US pipeline networks for the delivery of US produced crude to US coastal (primarily Eastern) refineries. If GoM & Midwest refineries reconfigured to lighter grades, US produced crude wouldn’t need to be transported by rail, while also adding to US jobs. It would also mean the end of Alberta’s tar sands other than for Alberta’s “domestic” consumption.