Heavy crude is in the news, and rightly so. Global oil demand will continue to grow for the next few decades. With most of the "easy" low-cost conventional oil already under production in mature oil fields, new sources will come from unconventional, higher-cost sources.

These include Canadian oil sands, Venezuela's Orinoco Belt crude, offshore heavy oil and onshore shale oil. Because the Canadian oil sands are cost competitive, the energy industry must determine how much potential new supply is feasible, and when it will help meet demand.

Future oil prices, infrastructure and demand will drive the pace of Canadian oil-sands development. Over the next 20 years, two significant new resources will supply the North American market: shale oil and additional oil-sands production. Meanwhile, North American oil demand will remain flat or decrease.

According to Hart's North American Shale Quarterly, shale-oil production will reach 900,000 to 1 million barrels per day (bbl/d) by 2020, and condensate producers will add 300,000 to 400,000 bbl/d more.

Hart Energy analyzed two Canadian oil-sands production scenarios. A short/ medium-term scenario includes existing production and projects that should be developed by 2020. A longer-term scenario includes projects starting after 2020.

In the first scenario, Canadian heavy-crude production doubles from 2010's 1.5 million bbl/d to 3 million bbl/d by 2025 and holds there. The second scenario adds 1.7 million bbl/d by 2030 and 2 million per day by 2035. If this high level is achieved, where will it be processed?

Currently, most (1.1 million bbl/d or 80%) of Canada's crude is processed in U.S. Midwestern PADD 2 refineries having 3.7 million bbl/d of combined capacity. Yet, little Canadian crude reaches the larger, 8.5-million-bbl/d PADD 3 Gulf Coast refiners (half of U.S. capacity), even though they process 2.3 million bbl/d of imported heavy crude.

Ongoing refinery upgrades will incrementally expand (but certainly not double) PADD 2 heavy-crude refining capacity. Increasing oil-sands production will therefore require expanded pipeline delivery to PADD 3 and/or exports elsewhere, since PADD 2 heavy-crude refiners will not keep pace with Canadian producers. Exports via tanker to the U.S. West Coast or Asian refiners may arise via pipeline routes from Alberta to marine terminals along British Columbia's coast.

Multiple southbound pipeline-delivery options with varying timelines under discussion should increase exports to the U.S. Gulf Coast. These include TransCanada Corp.'s proposed Keystone XL expansion and En-bridge Corp.'s proposed Monarch pipeline.

After issuance of a final Keystone XL environmental impact statement, TransCanada awaits a "national interest" determination by the U.S. State Department, since the project crosses an international border. An additional presidential permit granted by President Obama before year-end would keep the project on track to deliver heavy oil to the U.S. Gulf Coast by late 2013.

The Monarch pipeline requires no interest determination or presidential permit, because pipeline assets will not cross U.S. borders. Although environmental permits are needed, Enbridge could deliver Canadian crude to the Gulf Coast through a combination of pipeline expansions, new construction and pipeline-reversal projects. Timing for any Monarch heavy-crude deliveries would likely coincide with the deliveries envisioned by the Keystone XL developers.

For numerous reasons, the U.S. Gulf Coast is the North American market par excellence for Canada's future heavy-crude oil production. Eastern population centers representing essentially half of the U.S. population rely on these complex refineries.

Gulf Coast refiners process 59% of heavy crude imported into the U.S. at prices higher than Canadian heavy crude costs. And they are expanding to refine more cost-advantaged Canadian heavy oil. Currently, only 5% of Gulf Coast imports come from Canada.

Greg Haas, Hart Energy refinery editor, contributed to this article.