Why would you buy an oilsands lease, or if you had one, why might you chose to invest billions of dollars of your money, up front, to produce oil for the next 40-50 years? The answer is pretty simple – given your view of future oil prices, the costs of building and operating the plant, and the share of your revenues that the government will take in royalties and taxes, you’d have to be confident that you could make a rate of return on your capital equal to or greater than what you could earn on it in a similarly risky investment somewhere else. If not, why do it?
The same is true for the decision to invest in an upgrader or refinery. Upgraders and refineries make money when the value of the output is high enough, relative to the value of the inputs, to earn a competitive rate of return on capital – they are spread bets. For an upgrader, what you’re really interested in is the expected future spread between heavy oil or bitumen and light or synthetic crude oil. If you look at the figure below, bitumen had an implied average price of $65.50/bbl in 2011, whereas a barrel of lighter, higher value synthetic crude sold for an average of just over $102/bbl – a premium of $36.50/bbl – let’s call that the coke spread, since most upgraders or integrated refineries will employ a coker to strip out the heavier ends of a barrel of bitumen. Looking forward, according to Sproule Associates, the average spread is expected to be a little lower than that, at about $32.20/bbl over the next 10 years.
Source: Data from Sproule Associates (January 31, 2012)
So, can you make money on an upgrader in Alberta with a $32.20/bbl spread? As with anything in economics, the answer is it depends. It depends on how much it costs you to build and operate the upgrader, what tax incentives you might receive, as well as whether you can profit from any synergies between extraction and upgrading. Without tax incentives or significant co-benefits, the short answer is likely no.
Capital costs for building an upgrader in Alberta are likely well-over $60,000 per flowing barrel (per barrel/day of capacity), as this was Suncor’s 2008 estimate for building the 200,000 barrel per day Voyageur project. Even at a more conservative $50,000 per barrel/day of capacity, if you want a 10% rate of return on your invested capital over a 40 year time horizon (yes, I know these are too conservative as well), you’d need to net $15.56/bbl. A more realistic 13% rate of return over 25 years brings the capital costs up to $20.76/bbl, and you haven’t paid to operate the upgrader yet.
Upgraders use natural gas, electricity and chemicals/catalysts as their major inputs. Estimates of fuel and non-fuel operating costs for an upgrader vary significantly, but a low-end is about $7/bbl, with a median estimate of $8-9/bbl, and these costs will increase over time.
There’s one other hitch to this. If your upgrader uses a coker, some of your feed will be converted to low-value coke, which you could either burn or bury. If you bury the coke, which many operators do, you’d lose about 14% of your feed volume, so to produce 100,000 barrels per day of synthetic crude, you’d need about 116,000 barrels per day of bitumen coming into your refinery. That 14% loss of the feed hurts, since that brings your $32.20/bbl spread down to an effective $21.57/bbl of synthetic crude produced. Burning or selling the coke won’t change the deal much.
Given all the assumptions above, the currently expected $21.57 spread between the total cost of feedstock and the value of the end product means that your upgrader will earn 6.6% rate of return on billions of capital initially at risk. At those kind of rates, leverage won’t help you much because you won’t be able to secure debt financing much cheaper than 6.6%. Given that heavy-light differentials are highly volatile (i.e. you can’t guarantee that spread), the likelihood that someone would put this sort of capital at risk unless there were significant process and/or tax/royalty advantages to doing so seems slim. That likely explains why the Northwest upgrading project was only possible with government cost-of service guarantees and why you don’t see other merchant upgraders operating in Alberta.
Daniel Stuckless asked today on Twitter, “how does the value (of bitumen) change with (a requirement for) domestic processing?” Well, Canadian producers (and hypothetical merchant upgraders) are price takers for oil, so work backwards from there using the figures above. If upgraders need to earn a 13% rate of return to find people willing to put that kind of capital at risk, then the expected spread between 1.14 barrels of bitumen and and a barrel of synthetic crude oil has to be at least $30, or about 40% higher than it’s expected to be over the next decade. Since you can’t increase the world price of oil to increase that spread, you have to decrease the capital cost, operating cost, or feedstock cost. That means capital and operating cost tax allowances or subsidies, cost of service guarantees, or discounted bitumen feedstock. If you want more upgrading, it’s likely going to come at the government’s expense, not the oil companies’. Why would the oil companies bear the costs if we’ve passed a law that says the bitumen must be upgraded here? Our bitumen producers would be a captive market.
The real question people should be asking is, “should we be willing to subsidize, either directly or through trade barriers, merchant upgraders and refineries in this country in order to export a higher value end-product?” That boils down to a choice to sell it on the open market and to allow it to be upgraded where costs are lowest, or to sell it to a domestic upgrader for 10% less so that the upgrader can make money producing synthetic oil. Sure, we can have more upgraders and people will work to build and operate them, but we have to be willing to spend the money (or the bitumen) to make water run up hill, so to speak. I simply don’t see why the value from our natural resource should go to subsidizing the processing of our natural resources. In my view, there are far better uses for that value.
Great article!
Sometimes you must feel like quoting Rev Lovejoy from the Simpsons…
“Short answer is “Yes” with an “If”; long answer is “No” with a “But”"
Nice read..wonder if in equilibrium we are indifferent between upgrading and shipping.
That should be the outcome – also explains why you only see upgraders in integrated facilities where they can capture some “external” benefits from being located at the extraction site. I read a couple of other piece which suggested the break-even point for a refinery refit to handle heavy-sour feeds was around $15/bbl – so as long as transportation cost differences are that large, then US refinery refits should come before merchant upgraders here, in a free market.
Very interesting. What happens if you add a cost for “emissions” or “externalities”, be that a price for carbon or another instrument?
Thanks Annette. GHG emissions frm upgrading would depend on where you draw the line between the extraction facility and the upgrader. Brandt (2011) has upgrading emissions at just a little more than mining extraction emissions per barrel. I’d need to go through the numbers in detail to get an exact value, but roughly 0.04t/bbl for upgrading. For every $25 in carbon price, you’d be adding $1/bbl to the average operating cost.
You say $50,000 is more conservative. I would argue it should be termed aggressive. Not sure anyone in the financial world would believe a $50,000 number given the consistent story of cost over-runs, delays, etc. However, you note costs are higher so you actually paint a much rosier picture than is possible. Further, what would provide a better absolute economic benefit: running a mining/in situ operation at a loss for longer and paying lower royalties for longer before jumping to the higher royalty level but having the upgrader built with associated tax revenues remembering the upgrader won’t pay taxes for years as they recoup their investment?
Hi Jeff,
I agree. $50,000 was a low estimate, as was $7/bbl operating cost, as was the 2% annual cost inflation. Point was, even if you low-ball all the costs, you still need a big spread to break even.
Your last point is exactly what I was trying to illustrate. Higher value end-products, when driven by fiat, are not generally net-positive. You’ll be giving up value somewhere else in the supply chain, or using gov’t revenues, to make it work.
Andrew
A thorough and objective analysis. Guess we’ll be seeing lots of you on TV, radio and in the press, again. Great work.
Any thoughts on new technology like paraffinic froth?
http://www.albertaoilmagazine.com/tag/paraffinic-froth-treatment/
Lower oil recovery at start but when combined with a second treatement, the recovery is improved. See Titanium Corporation Inc. http://bit.ly/wCiQlP
Combined, the claimed benefits are impressive. Huge savings in capital costs, produces equivalent of Mayan heavy crude, high oil recovery, oil and mineral recovered from tailings, no tailings ponds, no massive energy hit as in upgrading and hence some reduction in carbon tax exposure.
Thanks. Paraffinic froth is not really new, but this particular implementation seems to be. I’ve just been reading up on it over the last couple of days. Anything that can get partial upgrading of oilsands feeds at low capital and labour costs is a positive for the province. Titanium Corp is an interesting company – they have been on my students’ energy briefing note lists for a couple of years now, and no one has taken them on, which I find odd. Neat business model.
Andrew
First of all, why are you talking about upgraders when everyone is talking about refineries. Second, your numbers are meaningless without a comparison to numbers in areas where the oil is shipped for upgrading (ie. Texas).
Robert, some people are talking about refineries, while others are talking about upgrading which would allow more refineries to handle Canadian crude, including those in Eastern Canada. As always, if you think my analyss is useless or irrelevant, you are most welcome not to read it.
I’m not too sure if I trust the source here. They seem to be saying that Canada shouldn’t be builing refineries because the costs are too high. If the costs are so restrictive how come other countries have figured out how to refine at a profit. Heck, they were even able to refine when prices were below $50 a barrel. There’s something a little smelly about your whole report Andrew.
Hi Cal,
By all means, question the numbers but I thibk that you’ll find the #s line up if you do some research. Two things: first, upgrading and refining are different elements, but one reason that equilibrium spreads are low is that it’s relatively inexpensive to retrofit an existing refinery to process heavy oil into higher value end products, and the US and Canada both have a lot of refinery capacity, including some idled plants. Second, even within Canada and the USn building in Alberta is very expensive since you’re competing for the same labour and service firms as oilsands production. So, building a new facility is a hot labour market to compete w existing facilities in soft labour markets is not usually going to end with profits. Finally, if you look at historic refining spreads, they are generally small with periods of high profits.
Andrew
Andrew
Two questions, rather than comments. First, what’s wrong with using coke to fuel the upgrading process & why is in uneconomic? I remember in Holland using coke to fuel our stoves & central heating systems and it seemed to be quite effective (although it didn’t have lasting overnight qualities; for that we used anthracite. Secondly, what’s wrong with underground, in situ upgrading like Petrobank’s THAI methodology?
Thanks.
Nick
Hi Nick,
Thanks for reading. Some do burn coke, while others bury it, and Nexen is a bit of a hybrid since they run an asphaltine gasifier. The reasons for this I know only superficially, and keep in mind that I’m an economist not a refinery engineer. Coke is generally quite high in impurities and so burning it is a bit of a challenge, especially if you have to meet stringent local air quality regulations. Some of it was also policy, which Alberta reversed in 2008 (http://www.pembina.org/media-release/1783) preventing some operators from burning coke, although the Alberta Specified Gas Emitters Regulation on co2 went into effect in 2007, so that likely negates some of the impact. Here’s a little explainer from the Alberta Government on properties of oilsands pet coke: http://www.energy.alberta.ca/OilSands/792.asp
Re THAI, it’s really taken a kicking from the low gas prices, since a lot of the upside on that was that you’re not using much natural gas to produce it as you’re essentially using the bitumen which would otherwise have remained in the ground to produce bitumen. Petrobank did not really have an in situ upgrading angle to the same degree as Shell or some others did (Shell actually had a Peace River Pilot which effectively produced diesel fuel directly from the ground), but they do get a little bit of in-situ cracking and they also have a pre-commercial pilot for a catalytic casing (CAPRI) that might get them another 7-8 degrees API. Combined, I’ve seen a few different projections, but the most ambitious ones get them to an API 20 final product, equivalent to a conventional heavy like a Lloyd blend. So, they’re getting some upgrading but not to the same degree as a Syncrude, Suncor, Nexen or CNRL which produce in the 30-40 degree API range (Nexen’s is really light at almost 40). So, nothing wrong with it – it’s just less cost-competitive now than when people were expecting $15-$20 natural gas, and it’s not really a substitute for full upgrading.
Andrew
[...] written at length about the first two questions here and here, so that leaves the issue of the long-term spread. A good benchmark for a long-run [...]