The NDP put forth a motion in the House last week which states that, “the Keystone XL pipeline would intensify the export of unprocessed raw bitumen and would export more than 40,000 well-paying Canadian jobs, and is therefore not in Canada’s best interest.”
This motion provided me with the motivation to dig into a question – if you had a given amount of capital to spend in the oil sands, would an oil sands mine alone or an integrated project with an upgrader generate the largest value-added return on investment, including total wages, royalties, taxes, and profits, and how would these be distributed?
To tackle this question, I ran two iterations of an oil sands project model based loosely on Suncor’s Fort Hills project combined with upgrader assumptions based on Suncor’s now-cancelled Voyageur project.
I’ve made some pricing assumptions as follows:
First, suppose that oil prices follow the Sproule Associates forecast for October 31, 2013, with West Texas Intermediate oil prices at $87 per barrel plus inflation, and a discount for Canadian heavy oil of 16% along with additional discounts and premia reported below.
| Pricing Assumptions ($2013) | |
|---|---|
| $CDN/$US | 0.98 |
| Edmonton Par discount to WTI (%) | 0% |
| WCS discount to Edmonton Par (%) | 16% |
| SCO premium to Edmonton Par (%) | 7% |
| Condensate premium to Edmonton Par (%) | 6% |
| Diesel premium to Edmonton Par (%) | 25% |
| Project diluted bitumen discount to WCS (%) | 0% |
Next, suppose you can build an oil sands mine for $84,000 per barrel per day of capacity, roughly what Suncor says it will cost to build the Fort Hills project. Further, suppose you could add an upgrader to the project for an additional $50,000 per barrel per day of capacity. I’ve set the upgrader cost really low, given that Suncor’s Voyageur upgrader, which would have been about the scale we are talking about here, was slated to cost at least twice significantly more than that amount. (thanks for catching this holdover from a previous calculation) Further, assume that you can operate the mine for an average cost of $20 per barrel of bitumen production, with an additional cost of $9.80 per barrel for the upgrader (including both energy and labour costs). Sustaining capital costs are $3 per barrel for the mine, with an additional $2 per barrel for the upgrader. The total ($2013) costs for the project, including royalties and taxes conditional on the assumed prices are as follows:
| Total Costs ($2013) | Standalone Mine | Integrated Mine and Upgrader | |
|---|---|---|---|
| Capital and Debt Costs | $mm | 42,080 | 42,080 |
| Operating costs | $mm | 100,521 | 95,677 |
| GHG compliance costs | $mm | 170 | 186 |
| Royalties | $mm | 71,665 | 46,063 |
| Taxes | $mm | 25,719 | 14,430 |
As you can see from the table above, the projects are equivalent in terms of total project capital expenditure. Finally, for tax purposes, I assume that the projects use Class 41 capital cost allowance for 95% of capital expenditures, with the balance attributable to Canadian Development Expense deductions. The Alberta Royalty Regime is applied, and the projects both pay royalties on bitumen subject to the bitumen valuation methodology, and the payout calculation is done with a long term bond rate assumed at 3%. Debt is assumed to cover 50% of the up-front capital costs of the projects, at a rate of 5.5%. GHG emissions policies are assumed to follow the Alberta Specified Gas Emitters Regulation in perpetuity.
| Project Economics | Standalone Mine | Integrated Mine and Upgrader | |
|---|---|---|---|
| Capacity | bbl/d | 280,044 | 180,000 |
| Cumulative Bitumen Production | mmbbl | 4992 | 3208 |
| WTI-equivalent supply cost ($2013) | $/bbl | 72.20 | 78.24 |
| IRR (%) | % | 12.30% | 11.70% |
| Free Cash Flow, $2013 | $/bbl | 14.9 | 18.95 |
| Free Cash Flow, $2013 | $mm | 74,365 | 60,799 |
| NPV (10%) | $mm | 14,221 | 11,256 |
The after-tax `economics’ for the two projects for the investor, shown above, look fairly similar at first glance. The standalone mine provides an after-tax rate of return of 12.3% and earns a break-even 10% rate of return on capital at a WTI price of $72.20 per barrel. The upgrader and mine combined earns a slightly smaller rate of return, at 11.7%, and requires an oil price of $78.24 to make a 10% rate of return. Subtle differences, but not huge and likely not large enough that you could say that one would proceed while the other would not (although, keep in mind that my upgrader costs are likely too low).
On a per-barrel basis, the numbers are equally ambiguous – in fact, the upgrader looks better. Revenues per barrel are, of course, higher with the upgrader with an average of $80.80 per barrel of produced bitumen vs. $62.93 for the mining project alone. Average costs (capital, debt, and operating costs combined) are higher for the integrated project, at $43 per barrel versus $29.15 for the mine, while royalties and taxes are similar at around $19 per barrel of produced bitumen in both cases. The result is that the upgrader earns higher cumulative cash flows, by $4.10, per barrel of bitumen produced.
Keep in mind though that the project with an upgrader earns higher cash flow per barrel, but for an equivalent initial investment (and for roughly equivalent total labour supply over the life of the project) it produces far fewer barrels – 1.7 billion fewer barrels, to be exact. So, over the life of the two projects, the total royalties and taxes collected from bitumen extraction versus mining and upgrading combined would be lower by $36.6 billion, while the profits to the producer would be lower by $13.4 billion. Combined, for a similar capital investment and with similar associated jobs, the bitumen extraction project returns $50 billion more in royalties, taxes, and profits.
So, if you had $42 billion to invest in the oil sands and had the land base, I think you’d have a hard time convincing your board to build an upgrader rather than a larger mine. From a labour perspective, it’s likely close to a wash. For governments, the answer is clear – if labour and capital are constrained, there’s a distinct preference for extraction over integration.
Note that the story may change when one considers the delivery of a barrel to a US consumption point or to the Far East.
The heavy oil will require diluent of at least 25% to 50% of the volume of diluent. At some point that diluent will either be returned (at cost) or consumed.
One diluent is light crude or the output of the upgrader.
Try your calculations with delivering light crude to market (100%) with the other option of 50% heavy oil and 50% of light crude. Also note that the energy to push the diluted heavy oil through the pipeline is greater.
Finally, when you observe Gulf Coast producers buying heavy oil from Canada, using resources to deliver at the Gulf Coast and refining it for products either used in the US or exported, then is it logical that this is possible without Canada giving up a lot of rent to the US refiners?
Mike,
Thanks for reading and commenting. I’m taking account of the diluent requirements in my pricing assumptions, which use Hardisty-based values for diluted bitumen and SCO, as well as for diluent. I am using a 33% blending assumptions for bitumen, and pipeline tolls which are 25% higher for heavy. The analysis doesn’t change if you use US-based netbacks since those are already in the prices at Hardisty.
Nothing in my analysis says anything about what’s possible – it simply looks at where the margins are, and what’s paying for the jobs. My analysis suggests that, where you’re doing a larger share of upgrading, you’ll have a lower share of tax and royalty revenues, and lower profits. You’re dissipating profits to wages and capital, but you’re dissipating royalty and tax revenues to wages and capital at a much higher rate.
Happy to chat about this in more detail or to look at more detailed numbers from your study.
Andrew
Hi Andrew, One comment is that your WCS discount to Edmonton Par of 16% results in a relatively high bitumen price — boosts the stand-alone mine and penalizes upgrading. My numbers show that 16% discount is right on for the past 5 years, but averaged 23% over the past decade (ie, was 30% in the 5 years preceding that). Has the supply cost of upgrading declined, or are we in a period of upgraders having overbuilt?
Hi Paul,
I used the 16% figure out of Sproule’s reserve evaluation forecast (which is about the same as GLJ, McDaniel, and others are using). Suncor has been back and forth (they used a larger discount in their Fort Hills numbers, but suggested it would be smaller when cancelling Voyageur. It likely depends on which pipelines go ahead and what the US does in terms of a crude export ban and how that impacts light oil values. You’re right though – the higher the discount, the better the integrated project looks. I’ll run some numbers out on that when I get a chance…maybe for ESNA in December.
Andrew
Your assumption of fixed investment means you also assume oil in the ground has no value. If you instead assume that innovations made in the future can drum up new investment but the quantity of oil in the ground is fixed, then a government might come to different conclusion.
I’m not sure it would be wise for a government to look at each project on the margin, since (I assume) roughly speaking each project drives up extraction costs for all future projects (by claiming the easiest oil).
Yes, different factors held constant would lead to different answers. However, with 170 billion barrels of economically viable reserves, and 2.2. trillion barrels of oil in the ground, the quantity of oil in the ground isn’t the limiting factor. So, yes, if you look at a lease-basis, your point is correct, but it’s difficult to assume that more intensive development on a single lease doesn’t have inflationary effects on other leases, whereas if you make the assumption I did (fixed scale of investment) you can safely ignore more inflation externalities.
I disagree with your conclusion from a government standpoint, Andrew. If the goal is to maximize govt revenue and jobs PER BBL of oil produced, the govt would clearly favour the upgrader. What you seem to be ignoring is that the resource is non-renewable. Slowing down and getting more for Alberta and Canada out of the resource makes more sense than simply allowing companies to maximize short term profit.
I have just started reading your posts, Andrew, and find them very interesting. Not being an economist, I am finding it hard to follow some of your arguments because of your extensive use of abbreviations. If you were to give a legend for them, people outside of your discipline would be able to follow without unnecessary hardship.
Abbreviations I refer to in this post include: Edmonton Par, WCS, SCO, IRR, NPV, mm. You do not have to define them; just tell us what they are short for.
Government revenue is basically a share of profits, so if you believe there is higher gov’t revenue from upgraders, you need to assume one of two things: 1) a stock of workers would magically appear to work in upgraders, but not to work in extraction, so costs would be lower; or 2) profits would be significantly higher, since royalties and taxes are net revenue based, and an upgrader would have a lower average tax rate, so to make up the difference, it would have to be VERY profitable. yes, the resource is technically finite, but the problem is one of managing abundance unless you believe the world is going to be using a lot of oil for a few centuries yet.
I believe that the world should have the resource available for a long time in the future: not necessarily for burning either as there are a lot of other uses for hydrocarbons. Are you planning for the world to end some time soon?
Exploiting the oil sands rapidly in a time of climate change makes no sense. In terms of ecological damage, exploiting it rapidly makes no sense, either. So I see maximizing economic and job benefits with a slowed down exploitation of the resource to make complete sense, and doing so within the limits of the number of workers and capital available.