Some of you may know that I have written some criticisms over the use of a technique called Hubbert Linearization to predict Peak Oil. I maintain that it is essentially useless for trying to predict an oil production peak. I have attempted to validate the model for a number of countries to see if it could have predicted a production peak in real time, and in the vast majority of cases the answer is no. And if you have a model that is wrong most of the time, you can’t have any confidence in the output for any particular case. But it is being used as evidence that Saudi Arabian oil production has peaked, and I think that’s useless. I wrote as much in an article for The Oil Drum:
Does the Hubbert Linearization Ever Work?
The main proponent for using the HL to forecast a Saudi peak is petroleum geologist Jeffrey Brown, aka Westexas at The Oil Drum. We have sparred quite a bit over the nature of the technique, of the concept of falsification, and over evidence in general. My impression is that he has a filter that does not allow contrary evidence to enter the picture, only highlighting evidence that seems to corroborate the story. This is not an uncommon trait; we all do it to some extent. But I think that by doing this with the HL, a lot of people have been convinced that it has more utility than it really does.
Now, despite our disagreements on the HL, I believe in giving credit where it is due. So, I want to focus on an area in which I think he nailed it.
Export Land
The Export Land Model is straightforward. In fact, the first time I read it, I thought “Yes, that makes perfect sense.” You can read about it firsthand at the two links below (the second one was featured on The Oil Drum yesterday), but I will summarize my take below.
Net Oil Exports and the “Iron Triangle”
Declining net oil exports–a temporary decline or a long term trend?
In brief, the model states that when oil production peaks, not only do we have to worry about declining production, but we have to worry about increasing domestic consumption from oil exporting nations. If Saudi oil production falls by 3% in a year, but domestic consumption in Saudi increases by 3% in that same year, exports are going to fall off at a much steeper rate. Very straightforward, nothing controversial there. And it is worth noting that among the top 5 oil producers, production has fallen over the past couple of years and domestic consumption has in fact risen.
I will add my $0.02 here, as I don’t think world oil production has yet peaked. But it doesn’t have to, for this situation to present a problem. If you have a situation in which oil production is growing slowly, but domestic consumption in major oil-producing nations is growing rapidly, you are setting up a situation in which exports are slashed even though production is growing. This is very similar in concept to Peak Lite, where demand growth is growing faster than production growth and that leads to a pseudo-peak. The export model can also lead to a pseudo-peak, hitting major oil importers the hardest.
As they have begun to recognize the Peak Lite concept, the mainstream media has also finally recognized the export problem:
America’s top oil suppliers to slash exports by 2012
Six of the largest oil suppliers to the U.S. are poised to significantly cut exports by 2012, ramping up pressure on supply and price, and intensifying the focus on one of the last great deposits open to private investment: Canada’s oil sands.
The forecasted cuts by Mexico, Saudi Arabia, Venezuela, Nigeria, Algeria and Russia are the subject of a keynote address that Jeff Rubin, chief market strategist and chief economist at CIBC World Markets will deliver at the firm’s Industrial Conference Oct. 2 in New York City. In his remarks, Mr. Rubin will share his latest research on the global oil supply/demand balance, with specific focus on the size and scope of the oil supply crunch facing the U.S. over the next five years.
And this next section sounds very familiar:
In recent reports and at a major oil and gas conference in Ireland this month, Mr. Rubin explained that surging domestic demand is eating into the export capacity of the world’s leading oil-producing nations. With production likely to plateau or decline in these countries, he expects global oil exports to fall by seven per cent, or 2.5 million barrels a day by 2010.
Of course if this plays out like this – and I think it probably will even if production does increase somewhat – we will continue to see major price pressure on oil and gas. Political leaders have got to wake up to this situation, and get very serious about implementing conservation measures. I think if people knew that gasoline prices were only going higher, they would start to buy more fuel efficient vehicles, carpool, and have second thoughts about moving into a home 30 miles from their job. But as long as they hold out hope that prices will fall – and false promises from political leaders only bolster those hopes – we won’t see major shifts in consumption.
I like the simplicity of the “produce less, consume more” conundrum faced by many key exporters as a way to reinforce Peak Lite observations. There is another layer of complexity to this, however, that I wonder about. That is the matter of internal fuel subsidies and power gen fuel-switching.
Consider countries that sell fuel internally (like motor fuels or even fuels for cooking like kerosene or LPG) at a discount to citizens. These subsidies are typically funded with, you guessed it, oil export revenues. Historically, it’s a classic redistribution of wealth transferring part of export revenues back to citizens. But as exports decline, there’s less money to pay subsidies, so at some point internal prices rise and demand drops. Or you experience a revolution, but that’s another matter.
Also, what about cheap crude fueling power plants across the Middle East? I think we’re already seeing some switching to gas (often wastefully flared at present) allowing the more valuable crude to be exported.
I’m not sure of the magnitude of these affects, but would be interested to learn more of their contributions to the export picture.
RR_
My thoughts exactly on both Hubbert’s curve and the Export Land Model.
I posted several times on TOD that North American fossil crude production should actually hit an all-time high a few years from now, thanks to Canada and the new Chevron strike in the Gulf, and also to stable domestic (lower 48) production. If it was possible to noose somene digitally, they woud have done it. So maybe 40 years after Hubbert’s Peak, we get to an new all-time high in crude production, for North America.
Now, on exports from thug states, there are worrisome trends. They are using more themselves. But the bigger factor is sheer imcompetence, boobery, thugginess and corruption. Oh, and horrible repression and theocracy.
So, we have Venezuela, Mexico, Libya, Iran, Iraq, Russia and former SU states, Nigeria all effectively “cutting” production, probably cumulative 5 mbd to 10 mbd a day, compared to what free markets would have obtained. Toss in KSA (they are so nice) cutting production, and you have the current price regime.
Without thug statism, it is arguable that we would now have a glut.
This is a far different picture from Peak Oil due to geological conditions.
The other factor widely ignored by doomsters is that demand is slackening rapidly, and new fields in KSA, Kuwait, Qatar, former SU and elsewhere are opening up, although there have been snags aplenty.
I suspect even with all the problems of thug states, and declining output in older fields, and oil-exporting nations consuming more, we still hit softer prices later this decade.
Best of all, I am hopeful there is no civilization-ending crisis ahead. The US could get by on half the BTUs we do now, and going to PHEVs and an electrical grid juiced by solar, wind, geothermal and nukes is probably a great way to raise living standards while gaining a better environment.
If oi prices go high enough, I suspect that is our future, and it is a good one.
All concerns about peak oil and wrangling about the biofuels implicitly assume that we are going to continue to use the inefficient internal combustion engine forever.
The electric motor is 85% efficient, the ICE only 25% efficient. At present it costs about 12 cents a mile to drive a gas vehicle and about 2 cents a mile to drive an electric vehicle.
Nearly 50,000 electric taxis are being readied by the Chinese for the coming Olympics. There are about 70 million electric vehicle of all types on the roads worldwide, everything from electric bicycles to electric taxis to electric buses and 7 1/2 ton trucks.
The French are planing to convert their fleet of 55,000 mail trucks to electrics. The 600 vehicle taxi fleet of Katmandu, Nepal are all electric vehicles. Santa Barbara, CA has a fleet of 30 electric buses.
So when is this electric age supposed to start happening ? …. It already has.
It is interesting that this idea that a predictive model should be tested (that it can be tested) finds such low traction in Peak Oil circles.
Interesting, but hardly news to me.
Indeed, the introspective peak oiler might pause to consider how important untested predictions are is to the movement.
Want to know how the world ends? Maybe Peak Oil is for you.
—
I get that models are useful even when they are not 100% accurate. They can give us useful information even when they are merely “sometimes” accurate. That’s fine if we treat them as fuzzy and unreliable signals.
It seems though that fuzzy signals, and partial confidence, are not in the Peak Oil mainstream. There, the tea leaves are read again every day, in an endless (and empty) pursuit of certainty.
—
So sure, the Export Land Model might have more guts to it than the various Linearizations, but I think the story is in the first part, in the hold that those Linearizations have on the public Peak Oil mind.
Robert,
Unfortunately, I did not see until today your story about the “Hubble Linearization” in TOD and so I cannot post my comments there.
In my opinion, we should view and treat HL as a “feature” of a dynamical model describing discovery, production and depletion of a finite resource and not as a data fit exercise as it is almost always done.
The logistic ODE is a beautiful two-parameter dynamical model that can be nicely used to demonstrate the concept of peak production rates and to predict trends. Predicting the time of the peak (even post facto) is not its strong suit.
I hope we can discuss this further.