Monday, March 26, 2007

George W. Bush, Meet M. King Hubbert

By: Jeffrey J. Brown

George W. Bush needs no introduction. Increasingly, neither does M. King Hubbert, a renowned geoscientist who used some mathematical modeling techniques to predict, in 1956, that US Lower 48 crude oil production would peak between 1966 and 1971. In 1956, he estimated that world crude oil production would peak no later than 2006.

In this article, I will attempt to put some of the crude oil production and consumption numbers during the first term of the Bush administration in the context of the Peak Oil debate.

I will make three key points: (1) During George Bush’s first term, the world used about 10% of all crude oil that has ever been consumed; (2) Based on our mathematical modeling, at our current rate of consumption, during the second Bush term the world will use about 10% of all remaining conventional crude oil reserves and (3) Net oil exports are falling much faster than overall world crude oil production is declining.

I also have some recommendations for actions on an individual basis.

First, a quick review of mathematical modeling techniques.

The “Hubbert” Methods

Dr. Hubbert used some mathematical modeling to predict that if Lower 48 Ultimate Recoverable Reserves (URR) were 150 billion barrels (150 Gb), the Lower 48 would peak in 1966. If the Lower 48 URR were 200 Gb, the Lower 48 would peak in 1971. Of course, the Lower 48 peaked in 1970. Dr. Hubbert’s (accurate) premise was that Lower 48 crude oil production would show a parabolic (bell shaped) curve. The area under a production rate versus time curve is the URR for a region.

Recently, Kenneth Deffeyes outlined a simplified way to estimate the URR for a region. Note that this applies to conventional crude oil reservoirs. One simply plots the ratio of annual production (P) to cumulative production to date (Q), or P/Q, versus Q. While I think that the method can be applied to get useful estimates in regions with constrained production, such as Texas and Saudi Arabia, it certainly does work better on regions with simpler aggregate producing histories, such as the Lower 48. This method is now commonly referred to as Hubbert Linearization (HL).

When a HL plot starts showing a strong linear pattern, generally with a P/Q intercept between 5% and 10%, one can extrapolate the line down to where it hits the horizontal axis, where annual production is effectively zero. This is Qt, a mathematical estimate of the URR for a region. The assumption is that regions tend to peak or plateau, and then start declining, in the vicinity of 50% of Qt. The following regions have all shown peaks or plateaus in close proximity to their respective 50% of Qt marks: Lower 48; Russia; North Sea and Mexico (for more information, see this article).

Kenneth Deffeyes, using the HL method, predicted that the world would cross the 50% of Qt mark in 2005. He predicted that world crude oil production would start declining between 2005 and 2009, most likely in 2006. The following graph was constructed based on the assumption that the world (in blue), in 2005, was at the same stage of depletion that the Lower 48 (in black) was at in 1970. (Note that the world data are for crude + condensate + natural gas liquids).

World and Lower 48

I would particularly like to draw your attention to four years, January, 2001 to January, 2005, the first term of George Bush’s presidency.

George Bush’s First Term: January, 2001 to January, 2005

In January, 2001, the US produced 5.9 million barrels per day (mbpd) of crude oil, and we imported 9.1 mbpd of crude oil.

Four years later, in January, 2005, the US produced 5.4 mbpd, an annual decline rate of 2.2% per year, and the US imported 10 mbpd of crude oil, an annual increase of 2.4% per year. (EIA data)

With flat to increasing product consumption, and with declining domestic production, our total crude oil plus product imports have been steadily increasing. The conventional wisdom is that this trend can continue indefinitely.

In this four year period, the world consumed about 100 billion barrels (100 Gb) of crude oil (crude oil = crude + condensate). So, during the first term of George Bush’s presidency, the world consumed about 10% of all crude oil that has been consumed to date.

Based on the Hubbert methods, this rate of consumption is unsustainable, since Deffeyes puts remaining conventional crude oil reserves at about 1,000 Gb as of January, 2006. So, based on this estimate we would consume about 10% of all remaining world conventional crude oil reserves in the second term of the Bush presidency, at this rate of consumption.

In fact, as Deffeyes predicted, world crude oil production has been trending downward since May, 2005, while average Brent monthly crude oil prices (at $62 per barrel) have been about two-thirds higher after May, 2005, than in the 20 months preceding May, 2005 ($38 per barrel).

However, the critical problem for importing countries is that world oil exports appear, as widely expected (because of generally declining production and rapidly increasing consumption in exporting countries), to be declining much faster than overall world crude oil production is declining.

So, where are we now?

The First Quarter of 2007

US product inventories are the lowest they have been since May, 2004, down 116 million barrels since October, 2006, while US product consumption is up about 3.3% year over year.

In my opinion, a Brent price of $62 was sufficient to cause forced conservation in many poorer regions, such as Africa.

However, in my opinion the upcoming bidding for declining world oil exports will be much tougher, against regions like the EU and China, instead of regions like Africa.

What should we do?

Recommended Actions

While hope springs eternal, I have doubts about any policy changes, at least in the short term, especially with groups like Cambridge Energy Research Associates (CERA) and ExxonMobil telling Americans that we don’t have to worry about Peak Oil for decades to come. By the way, note that oil companies working the Lower 48 and the North Sea have so far been unable to reverse the Lower 48 and North Sea declines, no matter how much technology has been applied. Oil fields are still being found—we just can’t offset the declines from older, larger fields.

George Bush has talked about the US “oil addiction,” and he has talked about curtailing US gasoline consumption and encouraging biofuels production, but the underlying assumption is that we can continue our current lifestyle, perhaps with just more efficient SUV’s. If he were still with us, I suspect that M. King Hubbert would disagree.

I recommend “ELP” on an individual basis.

Economize--Try to live on half or less of your current income.

Localize--Try to reduce the distance between work and home to as close to zero as possible, in much smaller more energy efficient housing, and integrate yourself into the community.

Produce--Try to become, or work for, a provider of essential goods and services.

When policy changes become possible, I would strongly recommend an Energy Consumption Tax, to be primarily used to fund Social Security and Medicare, offset by cutting or eliminating the highly regressive Payroll Tax.

Good luck to us all. We are going to need it.

Jeffrey J. Brown is an independent petroleum geologist in the Dallas, Texas area. His e-mail address is

Wednesday, March 14, 2007

Could Saudi Arabia Be More Than 70% Depleted?

by Jeffrey J. Brown

Texas and US Lower 48 oil production as a model for Saudi Arabia and the world
by Jeffrey J. Brown & "Khebab"
published on 24 May 2006 by GraphOilogy and the EnergyBulletin.

Matt Simmons, in his excellent book, “Twilight in the Desert”, warned that Saudi Arabia probably has far less recoverable oil reserves than they claim to have. Since Saudi Arabia is the world’s largest oil exporter, it is critical that we understand what their true reserve status is.

The Hubbert Linearization (HL) method gives us a way of estimating the Ultimate Recoverable Reserves (URR) for a region, using historical production. Therefore, we can estimate the URR for a region where accurate recoverable reserve estimates are hard to obtain, for example regions like Saudi Arabia.

In the captioned article, I argued, based on “Khebab’s” technical (HL) work, that Saudi Arabia in 2005 was at the same stage of depletion that the prior swing producer, Texas, peaked, in 1972.

Because we have so much data for Texas, we can get a fairly accurate estimate of URR, probably about 62 billion barrels (Gb), or what I will call Probable URR.

In the article, I noted that Texas peaked in 1972 at 57% of probable URR. Recently, Robert Rapier, on The Oil Drum, has raised some questions about what could have been predicted prior to the Texas peak. As a result of the questions that Robert raised, I have gone back and reexamined the data, and I have come up with some troubling conclusions.

Hubbert Linearization on Texas Oil Production
Hubbert Linearization on Texas Oil Production (Crude oil + condensate).

When I first did a HL plot for Texas, I used the data from 1958 through 2004 to get an estimated URR of 66 Gb. Khebab used a slightly different set of data to get an estimated URR of 62 Gb. In any case, the Texas data show a strong linear pattern from 1958 to 1965, with an inflection from 1966 to 1972, where the plot again showed a strong linear pattern.

Qt is an estimate of URR, based on the HL method. If we focus on Khebab’s HL plot, a linear plot using the 1958 to 1965 data would have given us a Qt that is reasonably close to the Probable URR. However, if we use the 1958 to 1972 data, it would give us a Qt of about 85 Gb, about 37% larger than the Probable URR.

Saudi Arabia has shown a strong linear pattern from 1991 to 2002, with an inflection from 2002 to 2005. If we use the data from 1991 to 2005, it suggests that Qt is approximately 186 Gb.

Hubbert Linearization on Texas OilProduction
Hubbert Linearization on Saudi Arabia (Crude oil + condensate).

However, if we use the Texas model, and discount the recent inflection, which is probably an artifact of the swing producer maximizing production prior to the peak, it suggests that the Saudi Qt is only 150 Gb, which in turn suggests that Saudi Arabia is now over 70% depleted, with about 40 Gb in remaining recoverable reserves. This is approximately one-fourth of the most commonly used estimate of remaining recoverable reserves, and it is a very small fraction of the ultimate possible reserves that the Saudis now claim to have.

Jeffrey J. Brown is an independent petroleum geologist in the Dallas, Texas area.