One of the interesting features of the climate change debate is that many of the highest profile academics in the field are European and many of the papers have what, to outsiders such as myself, might appear to be unconscious European biases. I previously noted this in a blog post discussing the famous Nature article titled “The geographical distribution of fossil fuels unused when limiting global warming to 2°C” authored by McGlade and Ekins. Well today another paper (in this journal short articles are called “letters”) came out titled: Macroeconomic impact of stranded fossil fuel assets by Mercure et al. Now let’s point out that Dr. Mercure is actually a Canadian, but the paper itself, as I will discuss, really misses out the importance that geography plays in the North American energy markets and in doing so I think it misses the mark. This letter is already being used by opponents of the Trans Mountain Expansion (TMX) project as a justification to not complete the project. This blog post will examine some of the strange issues I have with this letter that demonstrate why I believe it shouldn’t be considered a useful resource in discussing whether the TMX project is built or not.
Now one of the complaints about my posts is they are too long (TL:DR) so this time I am going to try something new. I am going to go with an Executive Summary approach using bullets that summarize the points I will address in detail in this blog. Here are the main reasons I have serious reservations about this letter:
- The letter assumes that the US and Canada will simply roll over and abandon their fossil fuel industries without imposing any import controls or trade protection measures.
- The letter deliberately excludes the logistical challenges associated with getting crude oil into continental North America and “minimises” crude oil transportation costs.
- The letter ignores the difference between heavy and light oils and assumes that all the oil produced in the world will be OPEC lights, irrespective of the refining capacity and the needs of the chemical industries for heavy oils.
- The model used to calculate stranded fossil fuel asset (SFFA) impacts comes to some strange conclusions about ho that stranding will affect other seemingly unrelated parts of the Canadian economy including:
- The models somehow concludes SFFA will result in a substantial decrease Canadian agricultural production.
- The models somehow concludes SFFA will result in a substantial reduction in the Canadian metals extraction industry even though those metals will be absolutely necessary in the production of the renewable energy technologies (including the co-production of critical elements as a by-product of the fossil fuel extraction) .
Looking at this list it is hard to believe that this letter would be used for any serious policy discussion about whether the TMX project is built or not.. Now let’s look at some details.
To begin I will note that I was a child during the 1979 Energy Crisis which I remember vividly. One of the big outcomes of that crisis was a political decision that the United States would never again allow itself to become dependent on Middle Eastern oil for its economic survival. The US even went so far as to ban the export of crude oil to ensure that its domestic supply was reserved and available for domestic use. It was only with the most recent shale boom that the US government reversed these policies. Yet in this study Mercure et al. assume that the United States and Canada would not impose any import controls or trade protection measures in response to the decrease in demand for fossil fuels. Mercure et al. assume the United States would allow themselves to become utterly dependent on the mullahs in Riyadh and Tehran for the fuels that run their economy and their military. I’m sorry but no political leader in the United States would return the US to dependence on Saudi Arabia and Iran for their economic lifelines.
Another assumption I take issue with is the cost factor. In the letter the authors are incredibly careful to consider the costs of production of fossil fuels but appear to completely ignore the costs to transport those materials. One of the stated reasons for building pipelines in Canada is the decrease in price we currently obtain for our crude because of the costs involved in transporting the crude to markets. Yet in the letter the authors specifically ignore how these costs would affect the decisions to shutter our energy projects. As the authors state in the supplemental material it is instead assumed that the available supplies are matched to demands in an efficient manner with transportation costs minimised. I personally believe this is an example of that unconscious European bias I mentioned above. Europe is so completely interconnected that from there it is easy for forget how hard it is to get fuel to a place like Yellowknife or Price George. Perhaps the authors of the paper forgot that North America is a huge continent with serious geographical challenges. Getting crude oil from Saudi Arabia to the Midwest refineries is not a cost-free measure, but that is how the authors treat it. As for Canada, in order for that Saudi crude to make it to the refineries around Edmonton we would need to re-configure several thousand kilometers of pipelines including the addition of a massive number of pump stations. You can’t just reverse the Trans Mountain pipeline. The pump stations were designed to take advantage of the elevation drops, they aren’t located to push material up the coastal mountains and over the Rockies in an eastward direction. Ultimately, the costs presented by the authors don’t incorporate the billions necessary to get that crude to the interior of North America where all those refineries are found. Once those costs are incorporated North American crude becomes very cost-competitive to the Saudi crude. A factor the authors don’t really consider.
Another thing I really noticed in reading the paper is how completely the authors appear to misunderstand the global crude market. A scan of the paper shows no evidence that they distinguish between heavy and light crude oils. I have already written a post discussing why heavy crude oils are a different product than light crude oils and cannot be readily replaced by light crude oils in existing refineries. Apparently the authors believe that all we need is the light crude oil produced by the low-cost OPEC partners but we now know that there are different markets for heavy and light crude oils. Any analysis that ignores the difference between heavy and light crude, and the amount of existing infrastructure tuned to refine heavy crude, is simply incomplete. As I pointed out in my previous post, a heavy crude refinery produces more gasoline and diesel and less waste using heavy crude than it does using light crude and has significantly higher margins running the heavy crude. This means that they can afford to buy slightly more expensive domestic heavy crude and still turn a profit. Once again this upends the narrative presented by the authors that the heavy oils in Alberta will be the first to disappear from the market.
Now the highlight of the paper has to be that terrifying Figure 3 which shows a massive drop in Canadian GDP associated with the SFFA. This caused me to go to the supplementary information to try and figure out where all that GDP went. This info is summarized in Supplementary Table 8 which shows some unexpected conclusions. I downloaded the spreadsheet associated with the report but it was not much help and since I am only doing a quick post I am not going to go through the effort of trying to get the original spreadsheets from the authors. Even absent the spreadsheets it is pretty clear that the underlying model makes some unexpected assumptions. Look at the loss in Canadian agricultural output associated with SFFA. Apparently the decrease in our oil industry will result in over 8% loss in agricultural production by 2035. In the model description they appear to indicate that the drop in fossil fuel assets will have cascading effects on consumer confidence and that might be the case if our agricultural industry was primarily for internal use. The problem is that a significant percentage of Canada’s agricultural production is for export which should not be affected by a drop in domestic demand. Moreover, let’s remember, that the global climate models suggest that Canadian agricultural production will increase with the initial phases of climate change. Thus under any reasonable future scenario we would expect the increase in both the amount of agricultural land in Canada and in demand for that output from other parts of the world (particularly near the equator where they will be seeing decreases in production) this should result in substantial increases in Canadian agricultural output not the drops suggested by the model used in this letter.
More puzzling is the precipitous drop in our extraction sectors. Canada has a lot of primary extraction. According to the Government of Canada website, Canada produced $96.9 billion of exports in “Mining, Quarrying, and Oil and Gas Extraction” in 2016 (last year available) with oil and gas making up $61.2 billion of that (63% of total) and mining and quarrying the remaining $35.7 (37%). Metals make up the lion’s share of that second number. Returning to the letter we see that authors predicting an 81.9% decrease in extraction sectors. That means the authors see the collapse of the oil and gas sector taking with it almost half of our metals mining as well. As a primary producer of base metals and minerals it is not clear why we would expect such a dramatic drop in demand for these raw materials in a scenario where massive growth of the alternative energy technologies is expected. Perhaps the authors expect these metals will simply materialize when required? Alternatively it might be that the model is overly simplistic and doesn’t effectively acknowledge what will happen to the Canadian economy in the situation presented.
Looking at the issues with this letter it is clear that a future drop in demand will not result in a massive abandonment of fossil fuel extraction infrastructure in North America. North American fossil fuel assets will not be stranded in the manner suggested by the authors because North American is huge and has a complex geography that affects the cost to move primary resources. North America also has legacy refining infrastructure that will not be abandoned simply because the Saudis can pump cheap light crude oil into the Arabian Gulf. That OPEC oil is not always the right choice for the refineries and it will cost a lot of money to get it to the interior of North America. Moreover, do you really expect the United States to rely on Iranian and Saudi oil? Finally, it is clear that something is amiss with the model the authors use to establish what the reduction in crude demand will have on the North American economy. The integrated North American economy is a complex one and the model the authors have used simply doesn’t effectively reflect that economy. Under climate change Canadian agricultural production is predicted to surge, not decrease, and demand for critical metals and elements will increase not collapse. Any letter that suggests the alternative is true needs to be read with a healthy dose of skepticism.
Dr. Mercure has kindly commented on this post below and I have replied. In particular Dr. Mercure noted some imprecise language which, in the light of the day, needed to be cleaned up. The downside of late-night blogging is you don’t always get the wording right the first time around and as such I have substantially edited this post since it was first put online. I have not changed the content but have definitely changed the tone. Dr. Mercure’s work represents a solid step advancement on our knowledge-base but like any limited paper was not able to be a be-all-end-all document. I feel I have raised some legitimate concerns why this letter should not serve as the basis for a major critique of the TMX pipeline project.