A Look at Canadian Oil Sands

A vocal debate about an early peak in the global capacity to produce conventional crude oil has been going on for some time now. To some, the dramatic run-up in oil prices – close to a doubling in inflation-adjusted terms over the last three years – has given added weight to the notion of a long-run supply constraint, never mind that short-term factors, such as the unexpectedly sharp acceleration in demand by China, India, and other fast-growing economies was clearly among the precipitating causes.

But even if worldwide oil productive capacity poses little likelihood of early decline, a steadily rising share of total output will most likely originate in regions posing geopolitical disruption risks as well as able to exercise market power in world oil. Translating that dual prospect into a future that may be subject to high and volatile oil prices makes it worth taking another look at liquid fuels for their abundance and reliability. Here I will focus on the prospective role of Canadian oil sands. (The outlook for a viable U.S. coal-to-liquids industry is far more problematic.)

Oil sands deserve attention for two reasons: their underlying resource base is vast, and they are being profitably produced in large amounts. At the same time, however, their long-term viability may depend on success in managing the significant carbon dioxide (CO2) emissions inherent in their production.

Oil Sands Facts

Canada’s proven recoverable reserves of some 180 billion barrels—exceeded only by Saudi Arabia’s conventional oil reserves—are concentrated in the Athabascan region of Northern Alberta. Oil sands, valued for their hydrocarbon content (called bitumen), occur as a near-solid, tar-like substance whose overall volume is a huge multiple of its energy content—thus creating a major waste-management burden.

Oil sands extraction takes place by one of two techniques: surface mining (not unlike open-pit coal mining) or underground (in situ) extraction.  For now, oil sands production is dominated by mining. But, because overall reserves occur predominantly in deep deposits, in situ recovery is likely to dominate over the long run.

When mined, the stripped overburden—removed by giant shovels—must be upgraded by a complex, multi-stage chemical transformation process to yield a conventional petroleum-equivalent product. In situ extraction typically involves, as a prior step, the injection of steam to make the bitumen less viscous and capable of being forced to the surface for upgrading.

In either case, conversion is an energy-intensive process that accounts for one of its most problematic features—significant CO2 release. The CO2 emissions associated with oil sands, compared to those associated with conventional crude oil, exceed the latter by about 20 percent on a life cycle or—in more catchy terms—“well-to-wheel” basis.

Trends and Projections

Oil sands production currently amounts to well over a million barrels a day—a significant proportion of Canada’s total oil production of around 3.5 million barrels a day. Over the next decade, a ramp-up in oil sands output to over five million barrels a day is widely foreseen. Because Canada is the leading source of U.S. oil imports—with a rising share of those imports derived from oil sands—that prospect is both reassuring, in that Canadian imports are certainly more secure, and worrisome because of the CO2 implications just described.

The fact that, absent CO2 emissions restrictions, oil sands production is currently competitive with conventional crude oil provides little comfort about the situation in a CO2-constrained regime. A major thrust of a recently released RAND report (see Suggested Readings) was an effort to consider how that competitive status might play out with severe CO2 restrictions, whether met by adoption of carbon capture and sequestration (CCS) technology or by purchase of carbon credits. Such credits can be represented by a “shadow price” of CO2, reflecting, say, payment of a carbon tax or purchase of cap-and-trade permits.

The RAND report provides a cautiously favorable picture of the long-term ability of oil sands to remain commercially attractive even while obliged to comply with formidable CO2 restrictions. A good way of encapsulating that key outcome is by means of the accompanying figure, depicting an estimated range of basic relationships (expressed in constant 2005 U.S. dollars) for the year 2025. Here are several noteworthy conclusions evident in the figure: Even with CCS, oil sands production costs in 2025 are competitive with conventional crude at near the $60/barrel world oil price, projected in the Department of Energy/Energy Information Administration (DOE/EIA) “reference case” (published in the 2007 Annual Energy Outlook). A DOE/EIA “high oil price” projection, close to the recent $100/barrel, makes the competitive advantage of oil sands still more robust.

Estimated Unit Production Cost in 2025 of Mined Oil Sands, With/Without CCS, Relative to Shadow Price of Carbon and Conventional Crude Oil

• That advantage would prevail at a shadow carbon price from zero all the way to around $100/ton of CO2. (By way of context, the price has hovered around U.S. $33/ton in the current E.U. carbon market. Up to a shadow carbon price of around $60 per ton of CO2, the economics favor paying the shadow price rather than installing CCS. Beyond that point of “indifference,” CCS becomes progressively more attractive.
• Oil sands extraction and upgrading currently relies principally on use of natural gas; although not shown in the figure, variations in natural gas prices can therefore signify lower or higher overall unit production costs.

In its wide-ranging scope, the analysis underlying the figure lends considerable credibility to these findings. Nonetheless, we are dealing with a number of unprecedented technological and environmental challenges whose ultimate success cannot simply be taken for granted but requires a sustained commitment to research and re-evaluation as experience dictates.

Consider just one elusive goal being pursued in a major research effort in Saskatchewan: CO2 sequestration that promises long-term stability and integrity. It is frequently observed that CO2 has routinely been injected into operating oil reservoirs so as to achieve enhanced oil recovery. But there is no assurance that such CO2 will remain locked in place and not seep into the atmosphere over the long-term future.

Of course, oil sands operations also involve numerous non-carbon environmental challenges. Companies must comply with regulations governing land reclamation, water-use management, and extended monitoring of tailing ponds containing mining spoils. In principle, such costs are embodied in unit production costs. But unforeseen externalities have a habit of arising in many natural resource development projects.

Even with oil sands production rising to a level of over five million barrels a day, with a significant share of that increment destined for the U.S. market, it’s useful to place that number in the wider perspective of the world oil market. True, in security terms, a marginal barrel of oil originating in Canada trumps the alternative of that marginal barrel from a politically problematic source in the Eastern Hemisphere. All the same, even an oil-sands contribution in excess of five million barrels a day has to be seen in relation to world oil demand of 100 million barrels a day a decade or so from now. In that sense, to the extent that the U.S. energy system remains significantly oil-based, relief provided by Canadian oil sands—whether in economic or security terms—may be meager. Indeed, it is one—but only one—element within the broad-based energy strategy that is in this country’s interest.

Author’s Acknowledgment:

I want to thank Mike Toman and my other co-authors on the RAND report (see below) for their collegial partnership throughout that effort. Of course, observations in this commentary are entirely my own responsibility. I also want to thank the Canadian Embassy in Washington for sponsoring a series of instructive site visits in Alberta and Saskatchewan.