The following review was posted to the Viridian List and appeared in Note 00316. Eric also forwarded it to me for posting here, since I am not on the Viridian List.

Hubbert's Peak, The Impending World Oil Shortage
by Kenneth S. Deffeyes
Princeton University Press, 2001

Reviewed by Eric Hughes

One of my favorite Matt Groening cartoons is near the beginning of the Love is Hell series, where he recommends, next time you are think about doing something shameless, just consider "how long you're going to be dead". Below that is a timeline, stretching to infinity in both directions, with a small black dot in the middle labeled "you are here."

The same illustration might well have been captioned "how long you're not going to be mining oil from the ground."

This last weekend I had the great pleasure of reading Hubbert's Peak, a book on the global petroleum industry. The author is a recently-retired geologist who taught at Princeton for thirty years, worked in the oil industry before that, and grew up in the middle of Kansas oil fields. The man is pure oil-intelligentsia, a category I was not previously familiar with. The book contains a wealth of detail, not exhaustive and dry detail, but selective and illuminating detail.

The purpose of the book is to make and justify a prediction about the year of maximum world oil production. His precedent is a 1956 prediction by M. King Hubbert (a former colleague of his) that U.S. oil production would peak in the early 1970's. The actual peak was 1970.

Let's cut to the chase. Deffeyes's prediction is 2004.7, plus or minus a couple of years. As he says immediately thereafter: "There is nothing plausible that could postpone the peak to 2009. Get used to it."

I'm not going to summarize the argument; that's the content of the book. I will summarize, though, the subject elements that lead up to the argument: initial oil formation, the geology of its deposits, prospecting, drilling and extraction, and the statistics of oil fields and their discovery. He then spends a couple of chapters on Hubbert's argument, and another couple on the future. It's a relatively slender book, only about 200 pages. It's plenty to make his point, without overflowing into boredom.

One of the delights of the book is the author's sense of humor. At one point in the text he presents a particular way of doing a graph that he's proud of. Fifty pages later, in an endnote, he makes the following comment: "Here in the back of the book, where my editor isn't likely to look, we can derive the equation." This combination of enthusiasm and restraint in the exposition is one of the charms of the book.

Another appreciation I got from this book is the difference in the dynamics between oil and gas production. They are intertwined in terms of their discovery and production techniques, but their drilling and markets are divergent.

Oil is found primarily in the "oil window" of 7,500 and 15,000 feet; gas is found below those depths. Both are sources of combustion energy, but they aren't direct replacements. In order to switch, capital has to be invested, which takes a while. The oil peak really matters, since it's not possible to immediately transition to natural gas or any other energy source.

It is easy to question individual assumptions and intermediate conclusions in the argument. That's not really the point, though. The conclusion that Deffeyes draws is both very narrow and quite robust. His only point is to estimate the year in which the most oil will be taken from the ground worldwide. He is not making a prediction about oil prices, about how many years of oil there are left, nor of any number of other questions about the dynamics. What's interesting is that under many variations of the model and permutations of the assumptions, the peak year doesn't really budge.

The model is robust against variations in reserve estimates. As recently observed on this list, total reserves depend a lot on the prevailing oil price. This observation, though, doesn't change the estimate of the peak year of production. Discoveries lead production by about 11 years. This reflects the capital investment cycle and the time needed to get into production.

As oil prices rise, existing reserves come into economic viability, but still with a lag time, and that lag time is longer because many of these sources require technology transfer in addition to the capital spending. As a result, there may well be some local upswings in production after the peak, but total world production will never reach its maximum again.

Along the lines of other critiques, Thomas Gold's idea of deeply buried hydrocarbons does not change this prediction of the oil peak. First, Gold is speaking primarily of methane, the principal component of natural gas. Even the partial replenishment referenced in Viridian Note 00314 was "very light oil and gas." The oil window goes down only to 15,000 feet because below that, it's too hot for long-chain hydrocarbons to be stable. So even if Gold is right about oil and gas being produced by microbes from raw organic materials embedded in the earth, that doesn't change the timing of the oil peak. Deep drilling for oil is poppycock; deep drilling gets you gas, not oil.

This is a science book, not an economic or political tract. Deffeyes makes no political predictions, but he does observe that political disruption is inevitable. The U.S. oil crisis of the early 1970's was precipitated by the peak of U.S. oil production in 1970. Prior to that, fluctuations in U.S. demand could be satisfied by simply pumping more American oil. After the American peak, the excess demand could only be met with overseas oil. Once OPEC realized this, there was a "crisis".

The disruptions that will happen after a world oil peak will be even more interesting, as in the curse about "interesting times." Excess demand is going to be met by rationing; some will simply have to go without. Previously, demand could be met by shifting the source of supply. This time there is no more source of supply anywhere. The few years after the peak will be particularly disruptive, since during that time many profound assumptions about the way the world works will be proven wrong. Bad investment decisions will be revealed in hindsight, and one should expect an orgy of finger-pointing.

Deffeyes mentions a potential flashpoint: "You guessed it; several islands stick up in the middle of the South China Sea, and the drilling rights are claimed by six different countries." These, as I recall, are the Spratly Islands. The claimants, just to give an idea of the touchiness of the situation, are China, Taiwan, Vietnam, Malaysia, Brunei, and the Philippines. Control of the islands is currently divided between the six disputants.

After a world oil peak, oil prices will go inevitably upward for some years, continuously increasing the stakes. This is an unstable situation that may well lead to war. If this sounds alarmist, it's worth considering the Middle East, the single largest oil region in the world. Ask yourself if the U.S. would have much political interest in a Middle East without oil fields.

I have painted a pessimistic picture of the medium-term future. The long-term future, while not worked out, has no such negative necessities. For one, there's plenty of oil left after the peak of production; it's just going to get rarer and more expensive. The taps won't all turn off at once. There is plenty of time to develop a successor energy infrastructure. Note that I didn't say "alternative energy infrastructure." There won't be anything alternative about the successor energy regime.

Right now everything is an alternative to oil. Soon these other sources will become the main ball game. I was gratified to read about the large remaining natural gas reserves. Reformation of natural gas is the best immediate source of fuel hydrogen, and there's plenty of it. Making the transition to a hydrogen economy will cost, but it will certainly be possible.

On balance, I'd consider this book a Viridian must- read. It's a wake-up call, a particular Viridian competence. Every important opinion needs enthusiasts as promoters. I'll be promoting the lesson of Hubbert's peak; please do likewise.

Eric Hughes

Page first created Friday, October 11, 2002
Page last modified Friday, October 11, 2002 11:04 AM
Go back to index page.