The definition is two consecutive years of 3% declines in production. In retrospect that may have been two strict - we could see something that historians would eventually recognize as a peak without seeing these kinds of declines until much later. Or it could be too weak - we could have a severe two-year recession/depression and have such declines, before bouncing back to higher production than ever. It's tough to come up with a clear definition. Like the fall of the roman empire, it's something that can really only be declared in hindsight.

See the claim at http://www.ideosphere.com/fx-bin/Claim?claim=pkyr20 . Trading prices predict a peak (as defined above) in 2010. (That is, 2011 production will be 3% less than 2010, and 2012 will be 3% less than 2011.)

(http://www.newsfutures.com/pdf/Does_money_matter.pdf provides evidence that play-money prediction markets can be as accurate as real-money ones.)

For instance, if the housing bubble (which Robert Schiller and The Economist both say is the largest financial bubble in human history) bursts, popping the larger debt bubbles, it is quite possible for demand to shrink by enough to cause an retroactive "peak" while dropping oil prices back to levels where it is uneconomic to pursue alternatives (like tar sands). A prolonged, deep recession could easily mean cheap oil prices and reduced production until the eventual recovery meets up with the downside of Hubbert's curve.

Conversely, political troubles in the middle east could send production plummetting, in which case "peak" would be reached immediately. Such an energy crisis would probably pop the bubbles and produce the results mentioned above.  

In either of these cases, which are very possible (check out Prudent Bear.com), then our hair pulling and bickering will seem rather irrelevant.

http://www.forbes.com/business/2006/07/27/oil-natural-gas_cz_mag092_0727shell.html
Shell Shocked
Christopher Helman 07.27.06, 3:00 PM ET

Excerpt:

In physical terms Shell is weak. Over the past three years it has replaced only 38% of the oil it has pumped. It is desperately searching for new pools of crude while trying to make up for their scarcity by shifting its focus toward natural gas. Gas is plentiful but hard to move--hence, in most cases, much less valuable as it comes out of the ground. Shell has suffered disarray in giant projects in Nigeria and on Russia's Sakhalin Island, and pricey delays in developing several prospects, including South Mars.

All oil companies confront the fact that petroleum is getting hard to find in politically stable parts of the globe. But Shell is in more of a crisis than the others because of a problem of its own doing: For years it exaggerated the size of its reserves. In January 2004 it confessed to the problem, eventually slicing 4.5 billion barrels of oil-and-gas equivalents off proved reserves, which now stand at 11.5 billion barrels. Counting only crude oil, reserves are down from 6.6 billion barrels in January 2003 to 4.6 billion now. This is more than a bookkeeping matter. Shell's production of oil has been dwindling at a rate of 7% a year.

Estimating reserves is as much art--or politics--as science. Shell has been in Nigeria for 50 years, and there's a lot of oil there. But how much can it get its hands on? Some of its Nigerian fields had leases set to expire before the company could possibly recover the oil it claimed as proved reserves. And relations with the host government are not good. The company's enemies in that country accuse it of destroying the environment, bribing officials, arming security forces and working behind the scenes to silence anti-Shell protesters. The company says it is a responsible corporate citizen and respects human rights.

Following is an excerpt from an Amazon.com review of "Hubbert's Peak," published in 2001.  Note that his updated work, in his second book, put 50% of Qt mark right in the middle of his original range of 2004-2008.

From Amazon:

"In Hubbert's Peak, Deffeyes writes with good humor about the oil business, but he delivers a sobering message: the 100-year petroleum era is nearly over. Global oil production will peak sometime between 2004 and 2008, and the world's production of crude oil "will fall, never to rise again." If Deffeyes is right--and if nothing is done to reduce the increasing global thirst for oil--energy prices will soar and economies will be plunged into recession as they desperately search for alternatives."