Book Notes: Hubbert’s Peak: The Impending World Oil Shortage
Review and Notes on: Hubbert’s Peak: The Impending World Oil Shortage
Kenneth Deffeyes, 2001
Read September 2006
The main point of this book is to argue that the world is approaching peak oil. Deffeyes argues that the peak will certainly hit before 2010.
Summary
All opinions stated are that of the author.
Chapter 1: Overview.
Chapter 2: The Origin of Oil – background on oil and where it is found.
Chapter 3: Oil Reservoirs and Oil Traps – more geological background on oil fields and their properties.
Chapter 4: Finding It – Discussion of oil exploration methods.
Chapter 5: Drilling methods – background on oil drilling
Chapter 6: Size and Discoverability of Oil Fields
This is where it gets interesting in terms of peak oil.
Deffeyes starts this question by asking how what part of the discovery of oil is by chance. He takes Kansas as a case study; as it has been highly, if not thoroughly, explored, the geology is well understood, and the historical records are good. For Kansas, he suggests that if you had drilled at random points, your discovery rate would have matched the actual.
The discussion then skips to “Zipf’s Law”, which states that for many phenomena, if you rank measurements in order, and then multiply the measurement value by the rank, the result is a constant number. The example given is that of populations of cities in Belgium:
Deffeyes notes that the same law applies to the size of oil fields. His conclusions here are that finding a few new supergiant fields can be consistent with Zipf’s law. Ditto for finding many new smaller fields. But to be useful, or at least, interesting to major oil companies, new oil fields have to be very large. I honestly can’t follow the logic flow here, but a little later, he presents his conclusion: “The conclusion for future oil supplies is that oil fields bigger than 2 billion barrels contain more than half of the existing oil. … If there is a future oil supply, it has to be anchored by big oil fields.”
Some discussion about where big oil fields have been found in the past, and undiscovered oil in the Middle East.
One interesting point: the North Sea oil fields have been discovered and produced much more rapidly than others. “The North Sea province will have a lifetime, from discovery to depletion, about half as long as those of other major petroleum provinces.” This is for two reasons: seismic methods work better at sea than on land, and the fields were more homogeneous than most.
Chapter 7: Hubbert Revisited – rundown of Hubbert’s prediction and his methods.
See my blog article for more details: http://northern-flicker.blogspot.com/2006/10/hubberts-method.html
Chapter 8: Rate Plots – looking at production data in terms of logistic curves.
Chapter 9: The Future of Fossil Fuels – Heavy oil, oil shale, natural gas, and other topics.
This and the rest of the book are somewhat perfunctory.
Chapter 10: Alternative Energy Sources – geothermal, nuclear, solar, wind
Chapter 11: A New Outlook
Comments and Review
This book reminds me a lot of some of my college professors, and not in a good way. Their lectures consisted of one part basic material and one part “application”, which generally was some personal experience or research that was only tangentially related, and didn’t really illustrate any basic principles. The lectures didn’t fit together; it was rarely clear what the professor was trying to get across, and the whole enterprise was a rambling, disjointed affair.
This book follows that pattern. The background material can certainly be interesting at times, but there are numerous digressions and asides that don’t contribute much of anything. Worse, there is no sense of how pieces fit together. That all comes to a head in chapters 6-8, where Deffeyes argues the state of the world’s oil supply. Seriously – there is nothing in the first five chapters that are used by the later analysis.
The analysis itself is similarly frustrating. I’ve read and reread Chapter 6 (Size and Discovery of Oil Fields) many times trying to make sense of it, and I can only dimly make sense of a logical argument. Going into a little detail: Deffeyes starts by asking if random exploratory drilling is be more or less effective than something more planned. Great question, and the pages spent here are interesting. But then Deffeyes abruptly switches gear to Zipf distributions, and that first part of the chapter is not referred to again in the rest of the book.
It’s hard to see what he’s trying to prove with the Zipf stuff, or that he’s particularly successful. Zipf distributions are descriptive, not predictive, and I can’t see how you can make any kind of confident prediction using those grounds. (An aside of my own: the idea that city populations follow Zipf distributions is interesting; a fun quiz is to ask yourself which American states do or do not fit the pattern.)
Then Zipf is dropped abruptly in favor of discussion of Gaussian distributions, where he makes the one categorical conclusion (quoted above) that I could find in the chapter. Then more bouncing around.
The chapter closes on two tangents. The penultimate paragraph starts with “I have long been curious about the world record; the most productive single oil well”. The final paragraph: “We are left with an ‘honor roll’ of wells larger than 100,000 barrels per day. Notice that any one of these wells would generate a cash flow larger than the gross national product of some United Nations countries.” Huh? What does that have to do with anything? Is that any way to end your train of thought? Like I said, I read and reread the chapter before finally realizing that Deffeyes is just rambling and not putting together anything like a coherent narrative.
And the whole book is like that. Any paragraph in isolation is generally fine, but the dots aren’t connected, and key assumptions are glossed over. Needless to say, it’s not a very compelling advocate for the peak oil case. And this guy is a professor at freaking Princeton. I don’t want to dismiss either peak oil or his arguments out of hand, but come on, you have to try harder than that.
Some scattered notes and observations:
It is a little sobering to see how early on most oil fields were discovered. Oil was discovered on the Alaskan Arctic slope in 1923. And technology – the “oil window” is between 7000 and 15,000 feet below the surface. The first rig that could drill 15,000 feet was developed by 1938. Economics – supposedly as the price of oil increases, there is more exploration and consequently more discovery. Consider this: the peak year for oil discovery in the U.S. was in the 1930’s, when we were in the middle of the Great Depression and oil was dirt cheap.
Some people have argued against Hubbert’s methodology saying it doesn’t track U.S. production after the peak in the early 1970’s. Deffeyes protests this, saying that the methodology takes the total amount of the oil in the ground (including undiscovered reserves) as an input. Hubbert’s predictions were off because his estimates of reserves were incorrect, not because the methodology was bad.
Well . . . ok, I guess I can buy that, but like I suggested before, if the method depends on knowing the total amount of oil, then knowing the method is a lot less important than having an accurate estimate of reserves. Deffeyes coverage of that is weak.
Last paragraph of Chapter 7. “This much is certain: no initiative put in place starting today can have a substantial effect on the peak production year. No Caspian Sea exploration, no drilling in the South China Sea, no SUV replacements, no renewable energy projects can be brought on at a sufficient rate to avoid a bidding war for the remaining oil.”
Well, that last sentence is true enough, but then as the economists would put it, there always has been and always will be a bidding war for the remaining oil. As for the first sentence: really? Are you saying that if the world shifted its priorities on a dime, and focused its massive attention and resources solely on oil exploration and improved means of extraction, then there would be no impact on the oil production curve? I find that hard to believe.
In summary – I can’t tell from this book if the peak oil case is weak, or if Deffeyes is simply ineffective at presenting it. The book itself, while it probably has solid technical sections (not totally sure because I’m not qualified to judge), is a mess in terms of presenting coherent lines of thought. Deffeyes is scornful towards the economists, and I don’t always trust them either, but they sure argue their case better.
When I don’t understand someone’s case, I always wonder if it’s because of my limitations, or if it’s the presentation of the case, or the case itself. I think here it’s the middle option.
Kenneth Deffeyes, 2001
Read September 2006
The main point of this book is to argue that the world is approaching peak oil. Deffeyes argues that the peak will certainly hit before 2010.
Summary
All opinions stated are that of the author.
Chapter 1: Overview.
Chapter 2: The Origin of Oil – background on oil and where it is found.
Chapter 3: Oil Reservoirs and Oil Traps – more geological background on oil fields and their properties.
Chapter 4: Finding It – Discussion of oil exploration methods.
Chapter 5: Drilling methods – background on oil drilling
Chapter 6: Size and Discoverability of Oil Fields
This is where it gets interesting in terms of peak oil.
Deffeyes starts this question by asking how what part of the discovery of oil is by chance. He takes Kansas as a case study; as it has been highly, if not thoroughly, explored, the geology is well understood, and the historical records are good. For Kansas, he suggests that if you had drilled at random points, your discovery rate would have matched the actual.
The discussion then skips to “Zipf’s Law”, which states that for many phenomena, if you rank measurements in order, and then multiply the measurement value by the rank, the result is a constant number. The example given is that of populations of cities in Belgium:
| City | Rank | Population (,000) | Rank * Population (,000) |
| Brussels | 1 | 953 | 953 |
| Antwerp | 2 | 450 | 900 |
| Ghent | 3 | 225 | 674 |
| Charleroi | 4 | 204 | 815 |
| … | |||
| Lier | 27 | 32 | 859 |
Deffeyes notes that the same law applies to the size of oil fields. His conclusions here are that finding a few new supergiant fields can be consistent with Zipf’s law. Ditto for finding many new smaller fields. But to be useful, or at least, interesting to major oil companies, new oil fields have to be very large. I honestly can’t follow the logic flow here, but a little later, he presents his conclusion: “The conclusion for future oil supplies is that oil fields bigger than 2 billion barrels contain more than half of the existing oil. … If there is a future oil supply, it has to be anchored by big oil fields.”
Some discussion about where big oil fields have been found in the past, and undiscovered oil in the Middle East.
One interesting point: the North Sea oil fields have been discovered and produced much more rapidly than others. “The North Sea province will have a lifetime, from discovery to depletion, about half as long as those of other major petroleum provinces.” This is for two reasons: seismic methods work better at sea than on land, and the fields were more homogeneous than most.
Chapter 7: Hubbert Revisited – rundown of Hubbert’s prediction and his methods.
See my blog article for more details: http://northern-flicker.blogspot.com/2006/10/hubberts-method.html
Chapter 8: Rate Plots – looking at production data in terms of logistic curves.
Chapter 9: The Future of Fossil Fuels – Heavy oil, oil shale, natural gas, and other topics.
This and the rest of the book are somewhat perfunctory.
Chapter 10: Alternative Energy Sources – geothermal, nuclear, solar, wind
Chapter 11: A New Outlook
Comments and Review
This book reminds me a lot of some of my college professors, and not in a good way. Their lectures consisted of one part basic material and one part “application”, which generally was some personal experience or research that was only tangentially related, and didn’t really illustrate any basic principles. The lectures didn’t fit together; it was rarely clear what the professor was trying to get across, and the whole enterprise was a rambling, disjointed affair.
This book follows that pattern. The background material can certainly be interesting at times, but there are numerous digressions and asides that don’t contribute much of anything. Worse, there is no sense of how pieces fit together. That all comes to a head in chapters 6-8, where Deffeyes argues the state of the world’s oil supply. Seriously – there is nothing in the first five chapters that are used by the later analysis.
The analysis itself is similarly frustrating. I’ve read and reread Chapter 6 (Size and Discovery of Oil Fields) many times trying to make sense of it, and I can only dimly make sense of a logical argument. Going into a little detail: Deffeyes starts by asking if random exploratory drilling is be more or less effective than something more planned. Great question, and the pages spent here are interesting. But then Deffeyes abruptly switches gear to Zipf distributions, and that first part of the chapter is not referred to again in the rest of the book.
It’s hard to see what he’s trying to prove with the Zipf stuff, or that he’s particularly successful. Zipf distributions are descriptive, not predictive, and I can’t see how you can make any kind of confident prediction using those grounds. (An aside of my own: the idea that city populations follow Zipf distributions is interesting; a fun quiz is to ask yourself which American states do or do not fit the pattern.)
Then Zipf is dropped abruptly in favor of discussion of Gaussian distributions, where he makes the one categorical conclusion (quoted above) that I could find in the chapter. Then more bouncing around.
The chapter closes on two tangents. The penultimate paragraph starts with “I have long been curious about the world record; the most productive single oil well”. The final paragraph: “We are left with an ‘honor roll’ of wells larger than 100,000 barrels per day. Notice that any one of these wells would generate a cash flow larger than the gross national product of some United Nations countries.” Huh? What does that have to do with anything? Is that any way to end your train of thought? Like I said, I read and reread the chapter before finally realizing that Deffeyes is just rambling and not putting together anything like a coherent narrative.
And the whole book is like that. Any paragraph in isolation is generally fine, but the dots aren’t connected, and key assumptions are glossed over. Needless to say, it’s not a very compelling advocate for the peak oil case. And this guy is a professor at freaking Princeton. I don’t want to dismiss either peak oil or his arguments out of hand, but come on, you have to try harder than that.
Some scattered notes and observations:
It is a little sobering to see how early on most oil fields were discovered. Oil was discovered on the Alaskan Arctic slope in 1923. And technology – the “oil window” is between 7000 and 15,000 feet below the surface. The first rig that could drill 15,000 feet was developed by 1938. Economics – supposedly as the price of oil increases, there is more exploration and consequently more discovery. Consider this: the peak year for oil discovery in the U.S. was in the 1930’s, when we were in the middle of the Great Depression and oil was dirt cheap.
Some people have argued against Hubbert’s methodology saying it doesn’t track U.S. production after the peak in the early 1970’s. Deffeyes protests this, saying that the methodology takes the total amount of the oil in the ground (including undiscovered reserves) as an input. Hubbert’s predictions were off because his estimates of reserves were incorrect, not because the methodology was bad.
Well . . . ok, I guess I can buy that, but like I suggested before, if the method depends on knowing the total amount of oil, then knowing the method is a lot less important than having an accurate estimate of reserves. Deffeyes coverage of that is weak.
Last paragraph of Chapter 7. “This much is certain: no initiative put in place starting today can have a substantial effect on the peak production year. No Caspian Sea exploration, no drilling in the South China Sea, no SUV replacements, no renewable energy projects can be brought on at a sufficient rate to avoid a bidding war for the remaining oil.”
Well, that last sentence is true enough, but then as the economists would put it, there always has been and always will be a bidding war for the remaining oil. As for the first sentence: really? Are you saying that if the world shifted its priorities on a dime, and focused its massive attention and resources solely on oil exploration and improved means of extraction, then there would be no impact on the oil production curve? I find that hard to believe.
In summary – I can’t tell from this book if the peak oil case is weak, or if Deffeyes is simply ineffective at presenting it. The book itself, while it probably has solid technical sections (not totally sure because I’m not qualified to judge), is a mess in terms of presenting coherent lines of thought. Deffeyes is scornful towards the economists, and I don’t always trust them either, but they sure argue their case better.
When I don’t understand someone’s case, I always wonder if it’s because of my limitations, or if it’s the presentation of the case, or the case itself. I think here it’s the middle option.
posted by Jim at
11:06 AM
0 Comments:
Post a Comment
Subscribe to Post Comments [Atom]
<< Home