GCC: ExxonMobil's Energy Outlook 2005

Each year, ExxonMobil publishes its view of the long-term trends of worldwide economic growth and energy demand: The Outlook for Energy. In the newly-released 2005 version, the energy giant's basic conclusion--that through the year 2030, traditional fossil fuels will continue to supply the vast majority of energy needs--remains the same, although the mix and sources are adjusted slightly.
Lots of good stuff from Mike over at GCC...(link)
The report is a joke.  Not dissimilar to the EIA International Energy Outlook 2001 which was wildly optimistic about some non-OPEC provinces about which it has subsequently been shown to be completely wrong.

Or the UK government aviation white paper from Dec 03 which predicts a doubling in air traffic by 2030... a forecast based on indefinitely stabilised oil prices at $25.

These kind of reports are hopelessly optimistic, come on where is OPEC going to find an additional 17mbpd, pumping an extra 300bn barrels over the next 25 years with rates as high as 47mbpd?

Exxon has not done any analysis on OPEC (how could they the books are closed).  They have a demand model; they have some optimistic non-OPEC supply projections and then just use the OPEC fudge factor to make it all add up.

Exxon has not done any analysis on OPEC (how could they the books are closed).  They have a demand model; they have some optimistic non-OPEC supply projections and then just use the OPEC fudge factor to make it all add up.

So it's basically a demand prediction model and not a production prediction model! it's a lot easier because demand will always go up and oil is an infinite ressource!
Re: The IEA being entirely wrong about Norway--as you all know, it's sometimes very hard to get the real scoop. But then I found this article--
which made everything clear.
Norsk olje-produksjon i unnabakke
Espen Bjerke

I juni var norsk oljeproduksjon den laveste på ti år. Nedgangen kan markere begynnelsen på slutten for norsk olje.
Den norske oljeproduksjonen har så langt i år vært langt lavere enn på samme tid ifjor. Oljedirektoratets prognose er at oljeproduksjonen vil ligge på 2,8 millioner fat per dag i år.

Men så langt har ikke oljeproduksjonen klart det målet i en eneste måned i år.... [and further down]

I juni ser situasjonen enda verre ut. Mens oljeproduksjonen i årets første fem måneder var på mellom 2,5 og 2,7 millioner fat, falt produksjonen i juni til under 2,3 millioner fat viser foreløpige tall fra Oljedirektoratet.

Laveste på ti år
- 2,3 millioner fat per dag er lavt. Vi har ikke hatt lavere månedstall siden 1994, sier seniorrådgiver Johannes Kjøde i Oljedirektoratet til dn.no....
So, it obviously it looks bad there despite previous IEA predictions. Here's another 2005 report. Finally, the IEA is reduced to begging--
The increased dependency on just a few nations in unstaable regions may be reduced if Norway could manage to maintain its level of oil production, or at least reduce the expected decline, says IEA chief economist Fatih Birol to Aftenposten.
We introduced the MobilExxon Outlook to our compilation of Peak Oil Depletion Scenarios on Sept 22nd.  With a peak of 117-mbd in 2030 and an inferred URR of 4-Tb, it placed third on our "optimism" scale.  It is compared to the others graphically at http://TrendLines.ca/economic.htm

Yesterday we posted our December USA Energy Reserves Report.  It's on the same page.  Global Crude is shown to be regaining "Spare Capacity" in our EIA graph which bodes well for softer prices in 2006.  Global Production has hit a new record of 85-mbd and our new forecast for the trough in working gas is 1050-Bcf.

The EIA graph shows 1mbpd or some more spare capacity in 2006 than 2005.  However, 1mbpd is significantly less than our uncertainty in the decline rate of fields in production, and I don't believe the EIA knows it much better than we do, since it shows little sign of serious study of the issue.  So I don't buy that we can have any confidence in this number (I'm not saying it definitely won't be so, just that there's no way to really tell next year's spare capacity with anything like that precision).
Of our 11 scenarios, only ASPO/Campbell predicts that we will be held to the 85-mbd present capacity.  Colin says we'll be at this level 'til 2010.  And i respect his work.  But overwhelming outlooks that say we are still going up the ladder cannot be ignored.  The TOD call of a July top in extraction was folly and is easily challenged by the rate of new production records thru 2004 & 2005.  Granted there is a cornucopia of negative post lately at TOD, but that sentiment does not jive with the evidence at hand nor the analyst opinions.

2005 was another year of merely crying wolf.  Any supply/price problems were based on distribution and weather related events and not global supply.  2005Q1 & Q2 saw record excesses going into stock building and reserves as the speculators dropped positions.  There is no indication that this did not continue into Q4.  If extraction does not grow, it will because of lack of parallel refining capacity and not potential production.

We didn't call a July top. Several of us noted in November that the then production peak was in April or May depending on time series, and explored what it might take for production to increase or not next year.  I certainly made no conclusive claim, though I do think the balance of the evidence has been shifting towards a nearer peak, rather than a further one.  Some analysts also think that (eg Henry Groppe).

I agree that lack of refining capacity may constrain peak somewhat - as peak gets closer, no-one will want to build more refineries, and this will have a flattening effect on the peak.

What do you believe is the global decline rate in fields in production, and why?

Yes Stuart, it was this implied May date "Happy Peak Oil Day" post that i was thinkin' about.  Your valid question about field or country decline rates is better asked of TOD member and TrendLInes scenarioist, Rembrandt Koppelaar.  Myself, i just watch the net annual gains in my work and not the two components.  And my premise for those annual gains to be prolonged is based on my belief that the Peak will accompany the Hubbert Peak in "all liquids".  In that sense, dramatic growth in URR by all analysts over the last five years tell me that we are very far from the 2-Tb URR that would indicate an extraction Peak in 2005-2010.

The avg URR in our Scenarios is presently 3.1-Tb and if i incl OGJ & API, it is 2.95-Tb.  This implies that we have 0.5-Tb to consume 'til Peak, or 16 yrs at present demand rates and assuming 1-Tb has been consumed to this point.  This points to a peak in 2020 ... if URR does not grow.  But it is clearly growing in leaps and bounds.  And an avg contract price of $50/barrel should push redefinition of much of present reserves historically not feasible over to the URR category.

We indeed have a paradox in refining.  If stakeholders believe "u" and pause on new refining capacity because the top is near, then the extraction peak will become self fulfilling.  OTOH, it they see efficiencies of scale and market share wisdom in building new plants, the product will certainly be there regardless, as old plants are shuttered for whatever reason.  

I'm certain Stuart could defend his work here but may prefer not to. So, I'll be a bit presumptious on his behalf here. When you say "Your valid question about field or country decline rates is better asked of TOD member and TrendLInes scenarioist, Rembrandt Koppelaar".

Nonsense. It is the IEA that does not appear to take into account realistic, empirically discoverable decline rates in those cases where good data is available. When such data is available, these rates appear to be larger than those presumed by the IEA (eg. Norway--see clv101's first post in this thread concerning IEA's pie-in-the-sky predictions--made in 2001--about Norway's post-peak production which show them producing in the year 2020 what they actually managed to produce in 2004 after peaking in the 2000/2001 period). Rembrandt is only one modeller here and so is Stuart and the other well known usual suspects (peak oil people or not) you show at trendlines. Decline rates are everything, the sine qua non of predicted world production curves. This is almost the whole peak oil argument.

If you want to just put up curves showing everybody's predictions, that's fine with me. But a deeper analysis is called for to evaluate their accuracy.
Dave, i am sure u have alotta passion about the peak, but the truth is that the only one predicting a 85-mbd peak is Colin Campbell.  Yet we see at http://trendlines.ca/economic.htm#ASPO that he has continually had to push out and up his predictions.  He backed off his 2002-4 projection this year but in the end he will find that it was spot on and both 2002 & 2005 will be revised upwards.

You cast doubt on the IEA ability to predict, and on a country by country basis, perhaps there are a couple of errors, but on the whole, their global extraction record is exemplary.  The third graph above the aforementioned link shows a 1991 graph by Lynch that illustrates that back then the IEA and EIA for that matter were looking forward to 79-mbd (actual 84-mbd) in 2005 whilst Campbell foresaw 52-mbd this year.  The records speak for themselves regardless of the conspiracy theories of hidden stats and inuendo's that are commonplace here.

In short, Dave, if i was a betting man ...

Re: "IEA ability to predict, and on a country by country basis, perhaps there are a couple of errors...."

No shit. Like Norway, as I said. And perhaps Saudi Arabia, as Matt Simmons claims.

Nowhere in my remarks did I predict, nor has Stuart, that 85/mbd is the peak. I regret inaccurate predictions by Colin Campbell in 1991 based on insufficient data. But really--
The records speak for themselves regardless of the conspiracy theories of hidden stats and inuendo's that are commonplace here [at TOD].
As the Japanese say, sayonara.
Well, obviously, much is going to depend on the definition.  The Thanksgiving day post showed data from the Oil and Gas Journal, which does not included NGL, ethanol, GTL, CTL, etc.  It is of interest to know when the black stuff peaks, and that is what I was trying to get a grip on in that post, and in the follow up.  I did not make any categorical statement that oil had peaked.  Instead I noted that there seemed to be something of a plateau, despite the fact that the world economy would obviously like more oil if it could get it (as expressed through the prices it was paying), and looked at who might be able to produce more.  It appeared to me then, as it appears to me now, that there is significant uncertainty as to whether production of black crude can go a whole lot higher.  I'm not saying it definitely cannot - it largely appears to depend on things in Russia and Saudi Arabia that are rather inscrutable.  However, as we explored here, the IOCs have not been able to increase production in the last few years, even though they planned to (Exxon's planned growth of 3% was typical).  Obviously, the decline rate of existing production is both hard to estimate (causing IOCs to miss their growth targets), and pretty large.  As you can see here, the implied decline rate in Exxon's production if you subtract out their new projects is both large and fluctuating.

Hence my comment to you that the EIA's projection of an extra 1mbpd next year of space capacity should be taken with a large grain of salt.  We don't know what the global decline of existing production will be between this year and next with nearly enough precision to say what next year's spare capacity will be and thus project next year's prices.

Now, if you want to talk about "all-liquids" -- certainly a valid topic but not the one I was addressing -- then we have to talk about coal (which obviously dwarfs everything else in potential and will make the URR enormous) biofuels, etc.  But that's a different kind of conversation, I think.  The question then becomes about how hard it is to ramp up very large amounts of CTL and what a global economy and climate running on CTL would look like, what lots of biofuels would do to food prices, and so on.  I certainly mean to get to those topics over time...

Perhaps then it would be helpful to use a somewhat less inflammatory subject line as "Happy Peak Oil Day", eh!  For all intents and purposes, oil is "all liquids".  When i drew attention to the Conventional Oil Hubbert Peak last year, it was an academic exercise only.  Nobody cares.  Where it comes from (conv vs non conv) doesn't matter to price or to supply.  Oil is oil.  Crude or by-products.  It is no use struggling over definitions that are lost in the marketplace.  That is for purists and idealists.  It took a very long time for Campbell to come to terms with that.  His disciples should follow suit.
The title was a tongue-in-cheek reference to Professor Deffeye's famous prediction.  I try to scrupulous about noting whether NGL is included or not in any given set of numbers I graph - that post footnoted the exclusion of NGL (which is the largest non-crude contribution to liquids at present).  As to whether it's important in the marketplace, it all depends who you are.  If you're a motorist, you only care about whether there's some liquid that will burn in your engine available at the gas pump.  If you are considering investing in a CTL plant or a biodiesel company, I think you should be deeply interested in what's going to happen to the conventional crude supply.  
Tell me would you also add in the EXXON CEO's predictions of a few time frames ago, predicting that the URR is really 14 Trillion Barrels of OIL ( USING your "everything that we get is really oil" statement ).  Taking his predictions into account, since you seem to take other folk's of equal standing predictions, you might average out to 5 or 6 Trillion Barrels of URR. Pushing the peak to the 2050 or beyond range.


 From all that I have read of your comments both here at TOD and on the Yahoo Energy Groups, you seem to want to debunk Peak Oil.  If so please step forward and show yourself.  I don't mind really, but I do dislike folks hiding their true agendas behind their backs.

P.S.  My agenda is Learn as much as I can to help me live a better life.

Hi Fred.
Where it comes from (conv vs non conv) doesn't matter to price or to supply.
That's quite a naive statement. There's a big difference between regular oil with an EROI of 1:10 and corn ethanol witn an EROI of 1:1.3 at the best.

Imagine I tell you this:
"Yesterday a backery store closed, today wheat is more expensive"
You'd probably think I'm not that smart.

The prices rise since mid 2004 can only be explained by an higher dependence on lower EROI liquids (other than con Oil). I hope you can realize the rest.

Freddy you left out the question mark. Happy Peak Oil Day?
I would only add that URR world production predictions--say, about 2350 Gb barrels, a number that Stuart has used lately--are completely meaningless when we are looking at production in the next 10 years. That is the peak oil time frame and that's what we need to care about. This is the crisis period in which large economic dislocations may occur. The potential worldwide economic "recession" based on shortfalls in conventional production due to declines in this period may have longer term impacts on what supply is available later.

I can no longer abide talk from IEA, EIA, USGS, IHS Energy/CERA and others that there is plenty of potentially producible hydrocarbon liquids--3 trillion, 4 trillion, hey, sky's the limit! Reserves growth--not based on new discoveries, of course, but based on better "knowledge" of what actually lies under the ground and is recoverable. There may be lots of potentially recoverable hydrocarbons out there (leaving aside the EROEI) but as far as "conventional oil" goes, which we have not weaned ourselves of, these estimates make no difference whatsoever in any timeframe we care about. Tar sands in Canada provide the best example--the best estimates I've seen put production of these at about 4.0/mbd by 2020. But by that time, declines in major producers (Russia, Indonesia, China, Mexico, Norway, the UK, Quatar, Kuwait, Saudi Arabia?) will have swamped the increases. So, let's get real here. Now would be a good time to do that.

We are concerned about available supply year-to-year to meet demand--that's it--in a timeframe that makes sense. Millions of barrels per day, year to year--2006, 2007, 2008 and so on. That's what we should care about, not proven & probable reserve estimates that indicate trillions of barrels of available supply in the longterm.
It seems we do have a refinery bottleneck, preventing climbing crude stocks from being converted to products and thereby reducing product prices. However, quite a few refiners are expanding as fast as possible, not least Valero, which will profit handily on low cost heavy/sour crudes whether we peak or not. We will then see if the world can produce enough to continue the current high rate of refinery utilization.
"Any supply/price problems were based on distribution and weather related events and not global supply."

Complexity rears it's ugly head.  And any solution-aside from collapse as economizing- requires more complexity and thus more energy.

Venezuela's energy minister said the government was preparing to take over an oil operation of US-based ExxonMobil and Spain's Repsol on January 1 if the US firm refuses to accept new terms for foreign oil companies.
"Of course we will take control of it," Energy Minister Rafael Ramirez, who also serves as president of state-owned Petroleos de Venezuela, told reporter in response to a question on the Quiamare-La Ceiba oil field


Bloomberg-122306-Venezuela Confiscates Exxon/Repsol Oil Field

Freddy, where did you get the 85-mbd figure for 2005?  I'm curious because I believe that Chris Skrebowski's mega-projects update predicted only 80.3 mbd, and I want to confirm the accuracy of this.
Cynus, each month the IEA does a press release on current production and amends past reports with more current data.  The EIA & OPEC have a similar process and there is seldom discrepancy between the data of the three independent agencies.  To summarize recent published data:
2005 June - 84.1 mbd
2005 July - 84.3 mbd
2005 Aug  - 84.7 mbd
2005 Sept - 83.8 mbd (hurricane related)
2005 Oct  - 84.4 mbd
2005 Nov  - 85.0 mbd
Are these the IEA numbers?  Is there a way to get a past monthly series of what they say without paying $000s?
Well, here's the last two reports (for October & November numbers)

November (full report)

December (highlights only)

IEA Oil Market Reports All 2005
Sorry, all monthly supply reports are available back to 1990 here.
I can't find a reasonable way to get a monthly series for total world production from your link.  It will let me make graphs of individual countries, and it will let me have quarterly balances in PDF.  I want raw monthly data as a CSV or an XLS.  There's a monthly data service, but it costs 5500 euros for a single user license.
And since EIA gives their data away, and OGJ charges $50, the IEA price seems rather steep.
Ok - one can get it from


but I still have to do it one month at a time.  Grrrr.  How hard can it be to put an XLS of the data on the site.

Yes they are IEA, but not much different than the other two.  Our first depletion graphs did not go very far into the future, but for perspective we have pushed the range far into the 22nd Century.  By doing that we lost the resolution of Y2k to 2010 that many of us were concerned about as we feared an imminent peak or one that had passed (ASPO driven).  Perhaps i could graph the post Y2k extraction to illustrate the present situation.  If not, i'll try to post links to or the data series themselves.

On the matter of global unused capacity, imho, the EIA has always been conservative with in their graphs.  The IEA was showing 2-mbd spare capacity as of Sept 30th 2005.

US EIA numbers are:
2005 Jan 83750
2005 Feb 84260
2005 Mar 84540
2005 Apr 84660
2005 May 84650
2005 Jun 84290
2005 Jul 84065
2005 Aug 84184
2005 Sep 83230

Rather different, go figure. IEA are estimates and will change; Petrologistics just reduced their estimate of Nov OPEC production by 300,000/day. EIA came out with its Nov figures before the countries themselves even announced them (e.g. Mexican numbers came out just two days ago). I am suspicious of IEA numbers both for this reason and that the hurricane recovery was not as fast as they would imply.

Did you notice that Exxon/Mobil was totally wrong about even their own short-term production estimate for the ensuing 5 years starting in 2000 as illustrated by Stuart on TOD before. If they can't guess their own short-term production from their own fields over a short time span, how do we trust long-term predictions on fields they will never see?
All those curves going up over 100 mbd. Wow. I thought the APSO curve was on the optimistic side with 85 mbd out to 2010. I guess I was wrong. Call me chicken little, I guess.

As I am not an oil person, I have only recently become exposed to the various estimates of production rates, size of reserves, decline rates, etc.  And as a layman, I must say that I am very surprised and a bit appalled that  there is still considerable argument over what is down there, how fast we can get it out, and how much is left.  Oil easily has got to be the most important resourse in the world, yet the numbers surrounding oil appear to have less firmness than the numbers surrounding the population of the snail darter. Now why is that?

Either the people that know are keeping the info secret, or the people that should know really don't. The latter possiblity is the more likely and the more scary of the two. Considering the importance of oil, we should have a good handle on every thimble-full of the stuff and how soon it will be gone, but instead we have funny numbers coming from every direction.

The paranoid side of me gets very suspicious that the powers that be are harboring a deep dark secret.

DENIAL is powerful for most of us. In some cases I suspect this is the case. But for those concerned asking the right question of the right people I am concerned your paranoid side is right; this would be hiding "peak" I presume you are thinking.   This is as someone has said our no. 1 security question.
The worldwide data situation is shocking. It moves all of our discussions farther away from science, and closer to conjecture. It's no way to run a world economy.

As for the deep, dark secret: yes, it could be there, and there are certainly reasons that politicians and oil producers don't want to discuss the peak. Talking about problems to come is not a good way to become popular. Prophets are still without honor in their own lands.

But consider Occam's Razor: the idea that given two equally predictive theories, the simpler is preferred. The two theories: 1) oil data is bad because someone is carefully, systematically, and successfully hiding the truth from us or 2) data is bad because it is generated and held in different ways by thousands of uncoordinated people around the globe.

Personally, I'd bet on theory 2. But is is entirely possible that both theories have some truth.  

I believe the confusion stems from lack of understanding of URR.  There are estimates of up to 7-11-Tb of all oils down below.  As the price goes up, and technology advances, we have the ability to draw out more volume feasibly.  In Canada, our non conventional oil was thought to be decades away with its $19/barrel breakeven point.  Now that looks like a bargain, and billions in investments are pouring into that sector.  And we have the billions.  As the prez of ExxonMobil said recently, at sixty bucks, crude has a twenty dollar fear component.  In Canada, we don't know what to do with all the obscene profits and royalties.  We're sending a few hundred back to the taxpayers, our gov't surpluses are sinful and the industry stocks are phenom.  This bodes well for new exploration and infrastructure.  And there is no sign that the dividend will end soon.  The Peak Oil lobby has convinced the consumer that oil will disappear.  And thus we have the situation that price is fear driven instead of demand or costs.  Speculation is rampant and amateurs use the spot market as their playground.  At times it was 20% higher than contract prices in 2005.
There are estimates of up to 7-11-Tb of all oils down below.  As the price goes up, and technology advances, we have the ability to draw out more volume feasibly.

Freddy, there are a couple of issues in your above statements.  First 7-11 Tb sounds more like OIP than URR and I assume that you include all forms of oil including oil sands and oil shale.  Second, you provide no evidence that indicates that enough capital and technology will be available to extract oil at the rate at which it is demanded.

In my mind, from the supply side, the three variables to be considered are 1)The total volume of oil available or OIP which is an essentially fixed but unknown amount, 2) the amount that will ultimately be recovered (URR) which is a variable amount based on technology and capital available to extract that oil and 3) the rate at which the oil can be extracted in barrels per day (bpd).

It seems to me that optimists tend to focus on OIP and URR whereas the key focus of pessimists about Peak Oil is the maximum rate at which the oil can be extracted.  The constraints that will cause a near-term peak are a combination of economics (amount of capital investment available, the price of oil, etc.) and geology ( URR, ease of extraction, etc.).

Since you consider all liquids, let's take the Oil Sands as an example.  OIP is estimated at 1.6 Tb, URR at 311 Bb, enough to meet world demand at current rates for over 10 years.  Sounds good, until you try to figure out the maximum rate at which you can extract it. So, let's be optimistic and assume we can achieve a maximum production rate of 8 Mbpd and the resource will last for over a century. We do have a problem if we require more than 8 Mbpd from that resource to compensate for a decline in production in existing fields.

Ray et al, it is sexy to dwell in the conceptual and muse about the future because there are no facts to defend.  We have seen in the last decade large recalculations of URR (for the reasons that i expressed).  Both OIP & URR are rising dramatically.  The Hubbert Curve worked impecably for Conventional Oil ... peaking by my stats in the Spring of 2004 for both extraction and the inherent half-way point of consumption.  ASPO and friends initially vehemently disagreed with my findings but in the end relented.  Now when i apply the same concept to "all liquids" the same cornucopia of wrong headed arguments surface and will similarly be proved wrong.  There is a correlation between URR and Extraction rates.  But the Peaksters say this time it's different.

The only well documented forecast of a Peak presently occuring is done by Colin Campbell.  The 10 other modelers also provide good data.  But it is Campbell alone that must continually revise upwards in extraction and outwards in time frame.  His credibility sank like a stone when he refused to acknowledge data that refuted his Peak(s).  But he has had an epiphany and now revises his model immediately upon discovering new data whether it moves the peak ahead or back.  That is admirable and i follow his work closer than anybody because he revises it monthly.

When the Peaksters find another ally that will make his work public so that we can judge it as a road map i will be happy to include it in my Scenarios.  I believe that they are there already.  Jean Laherrere and Rembrandt Koppolaar both share their excel data with http://TrendLines.ca but few here acknowledge their excellent work ... because it does fit THE AGENDA.  Those two gentlemen are known to be concerned about Peak Oil and its ramifications.  But they didn't get into a game of data fitting to promote their belief systems.  Rather, their data puts them on a better foundation to speak with perspective and revelance.

So the challenge is put forth.  Tell me who's data is better than Campbell, Laherrere & Koppelaar and i'll be glad to publish it.  But i'm not going to get into the game of playing with monthly or seasonal data.  The history of extraction data is one of cycles of rapid and slow and negative growth.   That someone feels strongly that there is a problem eighteen months out means nothing if it is not in cntext with what infrastructure is coming on stream in 2008-2010.

Stakeholders and decision makers want to see the decadal trends.  Not groundless musings about a Peak that never comes...    

This is somewhat peripheral to this specific topic but goes to the overall point about accuracy of predictions. The truth is that most so-called "expert" predictions are no better than random guesses. Here is a review of a recently published book which discusses the phenomenon:


The book is Philip Tetlock's "Expert Political Judgment: How Good Is It? How Can We Know?"

http://www.amazon.com/gp/product/0691123020/qid=1135215537/sr=2-1/ref=pd_bbs_b_2_1/103-1198668-85022 46?s=books&v=glance&n=283155

The book discusses a long-term experiment by Tetlock in judging the accuracy of expert predictions. From the review:

Tetlock got a statistical handle on his task by putting most of the forecasting questions into a "three possible futures" form. The respondents were asked to rate the probability of three alternative outcomes: the persistence of the status quo, more of something (political freedom, economic growth), or less of something (repression, recession)...

The results were unimpressive..., the experts performed worse than they would have if they had simply assigned an equal probability to all three outcomes--if they had given each possible future a thirty-three-per-cent chance of occurring. Human beings who spend their lives studying the state of the world, in other words, are poorer forecasters than dart-throwing monkeys, who would have distributed their picks evenly over the three choices.

Tetlock also found that specialists are not significantly more reliable than non-specialists in guessing what is going to happen in the region they study... "In this age of academic hyperspecialization, there is no reason for supposing that contributors to top journals--distinguished political scientists, area study specialists, economists, and so on--are any better than journalists or attentive readers of the New York Times in `reading' emerging situations." And the more famous the forecaster the more overblown the forecasts...

People who are not experts in the psychology of expertise are likely (I predict) to find Tetlock's results a surprise and a matter for concern. For psychologists, though, nothing could be less surprising. "Expert Political Judgment" is just one of more than a hundred studies that have pitted experts against statistical or actuarial formulas, and in almost all of those studies the people either do no better than the formulas or do worse...

The experts' trouble in Tetlock's study is exactly the trouble that all human beings have: we fall in love with our hunches, and we really, really hate to be wrong...

Tetlock's experts were also no different from the rest of us when it came to learning from their mistakes. Most people tend to dismiss new information that doesn't fit with what they already believe. Tetlock found that his experts used a double standard: they were much tougher in assessing the validity of information that undercut their theory than they were in crediting information that supported it. The same deficiency leads liberals to read only The Nation and conservatives to read only National Review.

This is a great review and it sounds like a wonderful book. I recommend that people take the time to read the review, which has many examples of the kinds of cognitive errors endemic to the prediction field.

It would be no exaggeration to say that my philosophy of life is based on the kinds of ideas presented here, ideas which I have been fortunate to be exposed to for a number of years (thanks largely to the work of my friend Robin Hanson). It is the foundation for my skepticism on Peak Oil and the certainty which is so rampant on both sides of this issue (and most others). From my very first posting on this board I have emphasized the uncertainty of the situation and the inappropriateness of the dogmatic attitude held by many Peak Oil true believers.

I believe that the forecasting methodologies explored here, based on attempts to acquire expertise about the details of oil field exploration and development, are doomed. The mere fact that every group applying this method gets a different result is prima facie evidence for this. There is simply too much uncertainty and too many ways to bend the results to satisfy our prejudices and expectations. We would do better, like the experts Tetlock studied, to stand back and throw darts at a chart.

Nice link Halfin - I ordered the book.  I agree of course about the difficulty of prediction - it is the hardest thing in the world.  However, in real life we have families, careers, businesses to look after that require making judgement calls, and so make them we must, as best we may.  The state of bottom-up prediction in peak oil is unsatisfactory for the reasons you note, but that is not the same as saying it is impossible to improve it.  Improving it requires greater transparancy of methodology, mutual critique, and actually changing one's methodology when errors are found.  Finally, I salute you for seeking out us TODers to disagree with you :-)
Halfin- You are entirely right about the uncertainty on all of this. Each of us can argue his particular viewpoint on Peak Oil-but none of us has a case that would stand up to legal scrutiny. There is a lot of doubt, and some of it (but not all) is reasonable.

But it makes sense to take a risk analysis approach, considering both probability of occurrence, and consequences of that occurrence.

What is the probability of a peak in oil production? Given a long enough time, it it a certainty (100%). Most of us are arguing not about whether there will be a peak, but when: is it 2005, or 2030, or whatever?

What are the consequences of a peak? Depending on what sort of peak, demand levels, timing, and how good our preparation has been, consequences range from very uncomfortable to devastating.

We often get this horribly wrong, spending vast sums on preparation for improbable events, like stockpiling millions of doses of Cipro for anthrax attacks. Or spending far too little on high probability problems, like the 41 million HIV-positive people in the world who will leave behind orphans, devastated economies, and be a huge reservoir of unpleasant diseases that can infect any of us.

The truly conservative course of action is to hope for the best, but prepare for something closer to the worst. This is what the world is doing for H5N1 avian flu (but not for peak oil, yet). If we invest in becoming more energy efficient, and develop new technologies, it will be good for the world ecosystem and world economy even if the peak is delayed for eons. Conversely, we can simply argue that the data is unclear, so we shouldn't bother with long leadtime preparations. But if we're wrong, we've made a misjudgment that may be irredeemable.

The combination of data uncertainty, but drastic consequences, was memorably summed up by Detective Harry Callaghan:

"I know what you're thinking. "Did he fire six shots or only five?" Well, to tell you the truth, in all this excitement I kind of lost track myself. But being as this is a .44 Magnum, the most powerful handgun in the world, and would blow your head clean off, you've got to ask yourself a question: Do I feel lucky? Well, do ya, punk?"
 Rick  Very good analysis, and best laugh Ive had in a weeek.        

Regarding theory 2. (above) we are talking about the folks tallying data. So I think this bodes for more of there motives "leaking " in thru assumptions/denial.However there seems to direct maneuvering of reasonable data. Anyway thanks for the laugh. Unfortunatly IMO, and laughter the analogy fits.

In a way, "right or wrong" is not the important axis for predictions. Prediction has old religious roots (prophecy, Jesus will return etc.), and is used as a practical tool to persuade people to action. Like Condaleeza Rice's vision of "a mushroom cloud", which helped mobilize the U.S. army. Or an army which believes God has foretold their ultimate victory. Those people can be really stubborn, so prophecy/prediction is good for morale. It doesn't matter whether it is true or not. What matters is whether it works.

When a prediction "works", there's no way to evaluate it, so the whole system is kind of self-validating. If I cause a bank run because I predict the bank will fail, was my prediction "correct"? If William Stanton euthanizes the UK down to population 2 million, does that mean the UK would have died-off if he had not done so? Would there have been a mushroom cloud if the U.S. hadn't invaded Iraq?

JD you're sounding much too much like Karl Rove :-)  I think we at TOD (at least the editors) are more into being part of the reality-based community.
Watch what they do, not what they say.
If Exxon thought oil would be that plentiful, that far into the future, they would be investing heavily in refineries, pipelines, drill rigs...and so on.  They are not.  They are investing in stockholder payouts, CEO bonuses and  in liquid natural gas.
Follow the money....
Bingo.  In my opinion this is one of the most compelling pieces of "proof" that companies like EM know exactly what's going on.  Ditto for all the other companies and parts of the energy infrastructure that were strung as tight as piano wire even before the hurricanes stuck the Gulf region.

The only other possible explanations are an immense act of collusion between all the oil companies to push up prices (which I don't believe would hold together for more than a three-day weekend without someone leaking), or monumental, industry-wide stupidity.  

If it's all a hoax, we could pool our money, build a refinery and get rich!

I think President Bush is offering free, unregulated space on military bases.

But nobody seemed to take him up on that ... hmmm ...

Einstein said that only two things are infinite, the universe and human stupidity and he wasn't sure about the universe. Oil execs may have all fallen into groupthink which is a form of collective stupidity. For instance the threat posed by the Soviet Union was a gross overestimate of their true capabilities but anyone privy to this fact and stated such was subject to ridicule. American politicians needed an enemy to justify the 'sacrifices' they imposed on the common man and to keep campaign coffers full. We are now asked to make sacrifices to protect us from "terrorists". If they were serious then the Bush twins would have been pushed into joining the army.
Look at what they do or don't do not what they say.
2005 was another year of merely crying wolf.  Any supply/price problems were based on distribution and weather related events and not global supply.

Freddy, I've been to your site and I like your work and numbers, but on this quotation, you can't be serious.

Look at a graph of the price of oil for say 1988 thru 2005. The only anomalies are 90/91 for the first Gulf War and then 1997-1999 for the Asian financial crisis and aftermath. Everything else comes under "normal." Even if you knew what to look for, you couldn't find Katrina on a long-term graph on price.

Supply from 2003-2005 (3 straight years!) has everything to do with the price (again, for 3 straight years!) and it is global, my friend, IT IS GLOBAL. The global part is key. That is what we are saying. The Iraq stuff, Venezuela, Katrina, weather - weather happens all the time, my friend, - this is bigger. When Nigeria and Angola are our only success stories for the year(2005) - we might want to be a tad more concerned.

This may be the wrong time and place to ask this, but here goes. In an article in Asia Times by E. William Engdahl: China Lays Down Gauntlet in Energy War, where the discussion centers around an oil pipeline to run from Kazakhstan to northwest China, the paragraph below appears. Can someone confirm or deny the accuracy of the reserves claimed? Thanks.


Best public estimates are that Kazakhstan has 35 billion barrels of discovered oil reserves, twice the amount in the North Sea, and may hold about three times more, according to a Kazakh government report released on November 18 in London. German oil engineers have privately reported that recent drilling by Italy's AGIP, the current oil consortium leader for Kashagan, a huge field offshore Kazakhstan southwest of Tengiz, has confirmed enormous oil deposits there

Tengiz and Kashagan are very big fields.  The numbers are not unreasonable.  Bear in mind that the world uses 30gb a year though.
Splitting hairs a little, North Sea will probably have 31 Gb extracted from the UK area and 32 Gb from the Norwegian side. So Kazakhstan doesn't have twice the oil reserves of the North Sea. However, if you say the UK is likely to extract another 5 - 7 Gb and Norway 10 - 12 Gb, then your statement on remaining oil reserves is true. I have also read web site pages which indicate it could be very difficult to extract the oil from some Kazakh oil fields as the geology is not very favourable. I am sure some regular contributors can add their penny's worth of knowledge about how easy/difficult the Kazakh oil fields may become. One reason why British Gas and other firms have sold their stakes in the oil fields (along with political reasons I am sure). Very interesting link by the way. Certainly added a twist to what is printed in the main newspapers.