Rubin, oil sands, and the bitumen bubble

A piece in Corporate Knights today from Jeff Rubin cites what has become a common metric in discussions of oil sands: supply costs. Understanding what these supply costs are and are not, and how they interact with bitumen discounts is crucial to understanding oil sands investments.

Rubin states that a recent report by Carbon Tracker estimates these supply costs, or “break-even oil prices” for new oil sands projects at $80 to $100 per barrel, “based on reports from Goldman Sachs, Wood Mackenzie and IHS CERA.” You could add my own research and recent reports from Canada’s National Energy Board and the Canadian Energy Research Institute to that mix as well—we all get basically the same numbers.  Rubin notes that, “Western Canadian Select, the benchmark price for Alberta’s heavy crude, ranks among the cheapest oil globally. It trades as much as $40 a barrel below the U.S. benchmark price, West Texas Intermediate, and at times over $50 less than the global benchmark price, Brent.” He then argues that the supply costs noted above are, “challenging enough on their own, but when you consider that producers only get a fraction of global oil prices, the picture starts looking pretty grim.”  To set the argument up as he does double-counts the discount and ignores what’s actually taken into account in a supply cost calculation.

First, what is the definition of supply cost?  It’s the value of a global price benchmark (West Texas Intermediate, or WTI, is usually used for oil sands supply costs) which you’d need to see in order for an oil sands project to earn an acceptable (usually 10-12%) rate of return on capital. In this context, it’s not a production cost—supply cost will always be larger than production costs by a large margin—nor is it a true break-even cost in that significant profit is still earned at those prices.

Let me give you a quick example. Take a typical new oil sands development, using in situ technology (oil sands produced through wells using steam injection as opposed to mining) which costs $2.25 billion to build and is expected to product 50,000 barrels per day for 30 years.  Assume that the project uses three barrels of steam per barrel of oil produced. Gas and labour costs combined would imply operating costs of about $22 dollars per barrel.  Revenues would come from selling diluted bitumen, made up of 30% natural gas liquids and 70% bitumen from the facility, which will sell at a discount to light oil of (for now) $15, while the natural gas liquids used to dilute the bitumen would sell at a premium to light oil of (again, for now) 7%. Finally, let’s assume that the Canadian dollar exchange rate is stable at $US 0.90 per $CAD over the life of the project.

For those assumptions, using my own model, I calculate a WTI supply cost of $US 62.50—that’s the West Texas Intermediate oil price which, if maintained through the life of the project, would allow the owner to make a 10% after-tax and after-royalty rate of return on capital deployed.  The diluted bitumen price at which the project would earn the same rate of return is $CAD 54.44 per barrel, while the implied bitumen price would be $46.50 per barrel.  At these prices, over the life of the project, profits would exceed $2.9 billion dollars, or $6 per barrel produced—hardly break-even in the sense you might imagine.  Importantly, the fact that oil sands producers only receive a fraction of the WTI price for their bitumen is already built-in to the supply cost, not something which should be thought of in addition to supply cost numbers.

Changes in underlying assumptions will affect the supply cost.  For example, if the bitumen discount were to remain stable at the levels cited by Rubin of $40 per barrel,  supply costs increase markedly, to WTI prices of $US 97.20 per barrel. If the Canadian dollar returned to parity with the US, the same project would break even at $US 69.60 per barrel for WTI (assuming discounts for diluted bitumen are back to the original $CAD 15 case). Importantly, in both of these cases, actual production costs haven’t changed. If natural gas prices increased to $8/GJ, supply cost would increase to $6.30 per barrel, while production costs would increase by $5.35 per barrel.

Supply costs are a way to relate oil sands investment decisions to world oil prices.  If you expect world oil prices to be below current estimates of oil sands supply costs, then you should also expect to see a downturn in investment.  Oil sands supply costs have increased markedly in recent years, mostly because of operating and capital cost inflation in Alberta, and that’s something we should be talking about.  Let’s make sure we get the meaning of the numbers right while we’re doing so.