This week, the question of whether or not and, if so, how, the Government of Alberta should encourage upgrading and/or refining of bitumen in the province is back on the front page. Much of this coverage is due to backlash over the Government’s decision to not proceed with the Alberta First Nations Energy Center (AFNEC) under the Bitumen Royalty in Kind (BRIK) program. There are as many myths as ever bouncing around this, and so I’ve spent the last little while trying to untangle them for myself. Here are some thoughts and, as always, your comments and clarifications are welcome.
What was the proposed project? The project would have processed 125,000 barrels of bitumen per day, producing diesel fuel, jet fuel, gasoline, and other refined products. The price tag for the project has been published at $6.6 billion, or $52,800 per flowing barrel of production. For what it’s worth, that figure seems staggeringly low given the price tag for the Northwest Upgrading integrated upgrader-refinery is estimated at a little under $5 billion, or $100,000 per flowing barrel, for a facility less than half the size. It’s likely that uncertainty over these capital costs was one of the stumbling blocks in finalizing the deal – more on that later.
Would this facility have made money? Well, that’s hard to say – if I could predict oil price spreads, I’d be rich. If they could meet their capital cost estimates, and guarantee last year’s spreads between the value of refined products and the value of bitumen, they almost certainly would. Even at NWU’s capital costs, the historically high spreads seen last year make a refinery project look very attractive – the question is whether the government should be betting on those spreads.
As you can see from the Figure below, price spreads in regions where prices are set based on WTI (the benchmark against which Alberta crudes are generally priced) have been historically high over the past 2 years (see here for a longer history of light:heavy differentials in Alberta), and this is compounded in Alberta due to discounts applied to our crudes relative to WTI, and larger discounts applied to Alberta heavy oil and bitumen relative to light-heavy differentials on the Gulf Coast. The bitumen spreads are positively massive, estimated at over $70/bbl in the previous month, and that’s simply a difference to a barrel of Brent Crude – the added value of the refined products would add another $10/bbl.
This graph should make every single Albertan very angry – your natural resources are being sold at prices up to $40/bbl below comparable world prices. Unfortunately, anger does not always make for the best decisions. You can likely build a refinery anywhere on the planet and make money with an $80/bbl bitumen:diesel spread, but that doesn’t mean we should build more refineries in Alberta today. A refinery is not a short term investment, so you can’t make a decision based on today’s spreads.
To get a sense of where that spread would have to be in the long term in order for these operations to make money, you have to start with the capital cost. If we assume that you can build the facility for $52,800 per flowing barrel, as the published AFNEC claims suggest, you’d need to earn $15.94/bbl based on a 25 year amortization. Add another $2.41/bbl for each additional billion dollars in capital cost, again based on the AFNEC 125,000 barrel per day facility. You’d then need to cover energy and non-energy operating costs as well as any sustaining capital expenditures. We don’t really have a perfect benchmark for AFNEC operating costs, but assuming they are close to comparable integrated facilities such as Cenovus’ Wood River and Borger refineries, $7-9/bbl in operating costs is likely a lower-bound. With an assumption of another $5 in sustaining capital expenses (someone please fill in the blanks with a reliable number I should cite for this), you are breaking even at $30-32/bbl, assuming a 10% cost of capital throughout.
Where have spreads been historically? Since 2006, the average 3:2:1 crack spread, which is an approximation for the gross return to a refinery’s output, has been US$10.39/bbl on the Gulf Coast, and a little higher than that in Chicago. Add to that, the average price at which heavy oil has traded at Edmonton relative to light oil of CDN$17.14/bbl, and you start to see why the investment is risky. For all intents and purposes, if the Edmonton bitumen to global gasoline price spread averages what it has for the last 7-10 years, you’d be barely in the money with a commercial upgrader/refinery in Alberta, assuming your costs stay competitive.
The last sentence above is crucial, since trends and policies suggest that neither will be the case. The WTI-Brent spread is at historic levels, and so projects both in Canada and in the US including but not limited to the Keystone XL and Northern Gateway pipelines are aimed at narrowing it. Most forecasts expect the WTI-Brent spread to be under $10/bbl by the end of the decade. As the light-oil differential narrows, so too will the discount to Alberta bitumen relative to other crudes. The heavy discount in the Midwest will also narrow as new coking capacity is added at Wood River and Whiting, among others, but will tend to expand as more bitumen production comes on-line in Alberta if pipelines to new markets are not built. Of course, it’s also difficult to assume that operating costs in Alberta will not increase rapidly, or that capital costs are not going to be subject to escalation.
The combined price and cost risk explains why the NWU contract was not simply a contract to provide Alberta bitumen via the BRIK program, but rather a cost-of-capital and cost-of-service processing agreement. Under the BRIK program, NWU does not take significant spread risk – they will be a regulated utility processing bitumen. The same would have likely been true for AFNEC, although the specific details of the agreement have not been released. What does that mean? Under the terms of the contract with NWU, the Government guaranteed that the proponent would receive a 10% return per annum on prior capital costs of $329 million plus facility construction costs of up to $5 billion, and that they would be reimbursed by the Crown for operating and sustaining capital costs as long as those costs do not exceed specific benchmarks. In other words, the downside risk that the spread collapses is borne by the government, not by the refiner, since they would still be paid to process the bitumen even if they were doing so at a loss. It seems to me that this type of long-run oil price arbitrage is not where the Government should be. We’re already leveraged to oil prices – why double-down on a spread?
I’d be happy to see more refining done in Canada, either in Alberta or elsewhere, if the comparative advantage is there but I think we need to look at it like any other industry. In the same way as I would be hesitant to say that Ontario should agree to guarantee the rate of return and operating costs of an automobile assembly plant, I don’t think Alberta should be doing so for a refinery. If the government were proposing to hire workers at market rates, and sub them out to a refinery or any other industry at a discount, people would likely scream and yell. Somehow, the same reaction is not present when the conversation turns to the government providing bitumen at below market rates, or taking on downside risk to make a business work. In the budget, at the end of the year, it amounts to exactly the same thing.
I do think Alberta should take a hard look at why were are selling our natural resources at far below their global market price and ask whether this is entirely due to infrastructure constraints, but that’s not an excuse to decide that we should instead look to sell our natural resources at below the global market price to encourage domestic processing. We should look to get the maximum value for our natural resources, not look to apply the maximum processing before export or look to offer a hometown discount.