A Tale of Four Predictions – Hubbert, Deffeyes, Yergin & Jackson

[pepe]

In 1956, M. King Hubbert, a well known geoscientist, predicted that US Lower-48 oil production would peak and start an irreversible decline between 1966 and 1971. Lower 48 production peaked in 1970, 14 years after Hubbert’s prediction.

In a book published in 2001, Kenneth Deffeyes, an associate of M. King Hubbert, predicted that world oil production would peak between 2004 and 2008. He later stated that the most likely peak was late 2005. Two measures of world oil production, Crude + Condensate and Crude + Condensate + Natural Gas Liquids (NGLs), are both down or flat relative to December, 2005, according to the US Energy Information Administration (EIA). Total Liquids, which is all of the above plus such things as refinery gains, bitumen/water blends and ethanol, are up slightly from December, 2005.

In a column published in Forbes Magazine in November, 2004, Daniel Yergin, an historian and chairman and co-founder of Cambridge Energy Research Associates (CERA), in response to a question regarding future world oil production and oil prices, predicted that world oil production would surge, driving oil prices down to $38 per barrel by November, 2005. In fact, oil prices have traded in a range about 50% to 100% higher than Yergin’s predicted long term index price, as flat to falling oil production has forced oil prices higher in order to equalize supply and demand.

Last week, Peter Jackson, an associate of Daniel Yergin, offered a critique of the Peak Oil theories outlined by Hubbert and Deffeyes and, like Yergin before him, predicted rising world oil production, with the world not showing any real decline until the 2040 to 2050 time frame.

In the past, Hubbert was right and Yergin was wrong. Now, their respective associates are making similar predictions, using similar methods.
Deffeyes uses a method that is now commonly referred to as Hubbert Linearization (HL), which involves plotting annual production (P) divided by cumulative production to date (Q) versus Q to estimate the Ultimate Recoverable Reserves (URR) for a region, which Deffeyes calls Qt. Regions, in the absence of political and/or technical problems, tend to peak and start declining shortly after reaching the point at which they have produced 50% of Qt, i.e., half of their recoverable reserves.

The following regions have shown lower production after crossing the 50% of Qt point: Texas; Lower 48; Total US (after a secondary lower peak following the beginning of North Slope production in Alaska); Russia; North Sea; Saudi Arabia; Mexico and most recently the world (except for Total Liquids).

I should be clear that the HL method applies to conventional oil production, which I define as oil production that will move to a wellbore without the application of heat energy. The two largest concentrations of unconventional deposits are the large bitumen deposits in Canada and Venezuela. There is also considerable research being done on oil shales, which are really kerogen deposits, a precursor to bitumen. Deffeyes’ opinion is that unconventional sources of oil will most likely serve to slow, but not reverse the decline in aggregate world oil production. Recent reports from Canada and Venezuela support Deffeyes’ view. In any case, the bottom line is that all of the unconventional sources of oil are hugely expensive, energy intensive and are very slow to ramp up production rates.

Jackson is asserting that better technology and the exploitation of unconventional sources of oil (plus gas-related liquids, which aren’t considered here) will permit the world to have several decades of rising, or at worst, flat production.

First, consider the Lower 48, where the industry has tried virtually every new technological innovation known to the industry, and production has fallen fairly steadily, now down more than 50% since peaking in 1970.

What about more recently developed regions? Haven't they done better than the Lower 48? Let’s consider the North Sea, which peaked in 1999 (crude + condensate) and started a very rapid decline. It is compelling that two vastly different producing regions -- the Lower 48 and the North Sea, with the North Sea being developed with vastly better technology than the Lower 48 —- peaked at the same 50% point, relative to their Qt estimates (for both crude and condensate.)

The basic premise of the HL method is that the first half of the production for a region is a good predictor of the second half of production. “Khebab,” a contributor on The Oil Drum blog, has demonstrated this mathematically. He took the production data only through the 50% of Qt mark for the Lower 48 and Russia (1970 and 1984 respectively) and predicted the post-50% of Qt cumulative production for the two regions, again using only production data through 1970 and 1984 to generate the model. The post-50% of Qt cumulative production through 2004 for the Lower 48 was 99% what the HL model predicted, and the post-50% of Qt cumulative production through 2004 for Russia was 95% of what the HL model predicted.

Today, we have the same amount of production data for the world that resulted in the highly accurate post-50% cumulative production predictions for the Lower 48 and Russia.

Peter Jackson is asking us to believe that we are going to see what we have never seen before: conventional oil production rising for decades after crossing the 50% of Qt mark. This prediction is especially remarkable given the near certainty that all four of the current super-giant oil fields producing one million barrels/day or more are in decline or crashing, while there is only one new super giant field being developed, the problematic Kashagan Field that won’t reach peak production, at the earliest, until 2020.

In conclusion, according to the EIA the world through August 2006 has produced roughly 100 million fewer barrels of crude + condensate than if we had simply maintained the December 2005 production level. This is consistent with the Hubbert/Deffeyes model. It is not consistent with the Yergin/Jackson model. The early data suggest that Deffeyes is correct and that Jackson is wrong, but we can't yet say with certainty who is correct.

Regardless of whether the Hubbert/Deffyes or the Yergin/Jackson model is correct, we need to start as soon as possible to fundamentally change the way we use energy in the United States. My personal opinion is that we need to tax energy consumption to fund Social Security/Medicare, offset by cutting or eliminating the Payroll Tax.

Jeffrey J. Brown is an independent petroleum geologist in the Dallas, Texas area. His e-mail is westexas at aol.com