IEA Revisions--The 2006 Forecast

Every month, the IEA publishes its Oil Market Report. Today, Adam Porter (thanks, Peaknik) of Resource Investor reported that
The International Energy Agency (IEA) however has slightly less excuse. Energy is their raison d'etre, not political power. Or at least that is what they say. They predicted that in 2005 non-OPEC output growth would be 1.38 mbpd. That figure has now been revised. It now stands at 0.1 mbpd. That in itself may yet be revised, downwards.

So currently the IEA's non-OPEC output growth forecast is only just out. If you could say that an error of 92.75% is `just out'....

Reassuringly the IEA have revised upwards their non-OPEC output growth figure for 2006. Next year they assure us it will now be 1.39 mbpd, not 1.32 mbpd that they had already called. Mind you they do not have to go far to beat this year's prediction.

American `demand destruction' was also in vogue just a few weeks ago. Yes, the IEA told us it was happening. Yet a few weeks later American demand for gasoline rose by 600,000 barrels in one week, to stand at 22.156 mbpd. That is not just a lot. That is a record high, ever.
partially based on IEA's December 13th report. Since IEA publishes monthly, they constantly revise their reports to reflect their current view of reality. But as far as longer term (yearly) forecasting goes in recent years, there seem to be two major discrepancies the IEA makes as the forecast period actually arrives in time versus the IEA predictions made a year or more earlier.
  1. Forecast demand revisions go upward year-on-year
  2. Forecast non-OPEC supply revisions go downward year-on -year
Stuart plotted IEA raw data versus IEA revisions. Let's take a closer look at trend #2, the IEA year-on-year estimates for non-OPEC liquids supply.
First, let's look at the IEA's current December forecast (link above).

Figure 1 -- OPEC versus non-OPEC (all liquids) Supply

Observe the flat non-OPEC supply curve from 4th quarter 2003 through fall of 2005. Now, look at Figure 2.

Figure 2 -- Click to enlarge

Note that as reported by Porter and here by Forbes, the IEA has got non-OPEC production up 1.4/mbd in 2006 over 2005 levels. Given the history from late 2003, is this large forecast increase in 2006 credible?

Using the December 2004 forecast, the IEA reported at that time that the non-OPEC oil (not all liquids) supply was 50.13/mbd and this is confirmed in their latest numbers for that year in Figure 2. As the latest IEA December 2005 report and Porter confirms, non-OPEC oil supplies are just 0.1/mbd over 2004--these were forecast to be +1.02/mbd in December 2004. The highest IEA forecast for 2005 I found in 2004 was +1.31/mbd (in October 2004), so I can't quite verify Porter's 1.38/mbd figure. Nonetheless, there is a large revision downwards of forecast non-OPEC oil supply numbers from 2004 to 2005.

Browsing around, an interesting article called The Crude Story, Part I came up written by a stocks analyst named Brian Trumbore President/Editor of In this story, Trumbore [published 7/23/2004] found some
...information from a recent internal report generated at a large money management firm. I am not able to identify the author, but for those of you who are energy junkies it's terrific research.
Obviously, I can not vouch for the source but there's some very interesting information in this article claiming that downward revisions of non-OPEC supply numbers has been going since the year 2000.
Most oil analysts base their supply-demand models on data generated by the International Energy Agency (IEA), the organization set up by the industrialized economies in the mid- 1970s to monitor global oil markets in the wake of the 1973 oil crisis. Since 2000, the IEA has forecasted that future non-OPEC supply would increase faster than global demand. However...the IEA data has consistently underestimated global oil demand and overestimated non-OPEC oil supply....

The IEA's biggest problem is consistent overestimation of non- OPEC, and particularly non-OPEC / non-FSU (Former Soviet Union) oil production. Over the last four years, the IEA's downward revisions of non-OPEC / non-FSU production from its original estimates have been impressive.

The IEA's original 2001 estimate was revised down by 500,000 barrels per day, and its 2002 estimate was revised downward by 400,000 b/d. Its 2003 estimate was revised downward by 900,000 b/d, and its 2004 estimate has already been revised downward by 700,000 b/d. On a four-year cumulative basis, the IEA has overestimated non-OPEC / non-FSU production by almost 3.7 billion barrels....
It is necessary to go back through previous IEA OMR documents to confirm all this--which I am in the process of doing--but it would seem on the face of it that this is a consistent and erroneous trend in IEA's non-OPEC data reporting. As always, I recommend reading Trumbore's original article and looking at the monthly IEA reports.

Using my 1.31/mbd downward revision (from October 2004) of IEA's non-OPEC supply for the year 2005 and the numbers quoted in Trumbore's article by an anonymous source, we see that the IEA has overestimated non-OPEC oil supply by 4.18 billion barrels over the last 5 years. As for upward revisions of demand trends (also discussed in the Trumbore article), this calls for further research.

So, these trends in IEA forecasting tend to cast doubt on their current 2006 non-OPEC number of 51.6/mbd (Figure 2) and their all liquids forecast (Figure 1). It appears right now that we can feel somewhat confident that this number will be revised downward as the year 2006 actually goes along. At least, that's the trend we've seen. In the 2005 world, supply & demand have remained on a razor-thin edge. If recent history is any guide, numbers from the IEA give us little reason to believe that excess capacity will become available to change that trend in 2006.
Dec. 28 (Bloomberg) -- Wall Street oil analysts have given up calling for lower prices after missing the rallies of the past four years.


I don't have time to do this now, but we need to calculate hurricane effect on 2005 figures to get a better idea of how far off IEA really was. That said, we also need to remember that this is a real effect and could easily repeat in 2006 and into the future, at least to some degree, given the hotter Gulf temperatures these days.
A rough approximation of the impact of the two hurricanes in terms of decreasing global production is about 250,000 BOPD average for all of 2005.  This assumes about 750,000 BOPD shut in (on average) since September.  This is probably an overestimate of pure crude, but I am not sure how much condensate or NGL's come from the Gulf, and a lot of natural gas is still shut in.  

This rough forecast also does not take into account any additional production that would have been brought on if the hurricanes had not happened (e.g. starting up Thunderhorse).

As for your "rough approximation", Bubba, I'll buy that and it seems to be reflected in IEA's Figure 1 in my post. It seemed too hard (in other words, I was too lazy) to figure out the actual effect on world non-OPEC production resulting from the hurricanes.

However, I think this relatively small "oil shock" hardly changes my analysis of the IEA forecasts and revisions over the last 5 years. And, as peakearl implies, perhaps the IEA ought to factor in the risk of such Gulf of Mexico production disruptions each year in making their forecasts.
Thanks. Seems pretty reasonable to me.
Adam didn't report it today, he reported it December 23.

You also left off that he took a few digs at Peak Oilers:

Peak Oil `believers' have also been polishing their oily crystal balls. This year seems to have been the year for `peak freak'. The fascist British National Party arrived at a depletion conference in Scotland, replete with heavies. Eco-primitivists tell us that 50 years from now we will all be riding oxen and wearing homemade smocks, if we have not died off.

Racists see that `peak oil' will mean a wiping out of immigrants as society collapses into Nazism. Survivalists are polishing their weapon collection. Communists are awaiting the rise of a new Chairman Mao to march us away from our addiction to oil. Liberals believe it will usher in an era of collective responsibility, solar power and homegrown vegi-burgers.

Overall it is a colorful bit of curmudgeonry but IMO not that helpful except to remind us that predictions on all sides have a pretty bad track record in this field.

There are whacko extremists who believe in evolution and quantum mechanics too. Does that make either of those topics less real just because some whacko extremists believe them? I have no idea why you quoted that part of the article. I can make guesses but none are nice so I will refrain. How about we stick to the facts though? Wasn't that Dave's original intent?

As for IEA predictions, which is the topic of this thread, yes the IEA has a pretty bad track record but it is consistently bad in one direction only. It's not random noise where they are too optimistic one year and too pessimistic another. There is a specific trend here that can be shown from observing their predictions against actuals for several years. And that trend reinforces the concerns of peak oilers while it also discredits cornucopians who rely on those same IEA optimistic predictions that everything will be ok.

I'm sure Halfin can defend himself just fine without my help, but I would like to jump in here and point out that he said explicitly that hose comments were "a colorful bit of curmudgeonry but IMO not that helpful except to remind us that predictions on all sides have a pretty bad track record in this field."  

If you want to slap around someone for the content of those comments, why not drop Adam Porter an e-mail?  From what little contact I've had with him he seems to be a level-headed guy, and very open to communication with readers/listeners.

And for that matter, I'm with Halfin 100% on this one.  Porter's comments weren't what I would consider helpful, but it's certainly true that you can find a LOT of wildly off-base predictions on all sides of the peak oil debate.  

Naturally, I read Porter's comments about "believers" who are "polishing their oily chrystal balls". He was all over the map in that December 23rd article (sorry about my mistake about the date), for example
Meanwhile neo-liberal fundamentalists like author Peter Odell tell us that oil will be drilled forever and waste is a good thing. Their party ideology seems to be based on the same logic as the Japanese whaling industry. The more you consume the more you will consume. If you believe that eat all the food in your fridge now, the market will send you more courtesy of the magic oil pixies.
Fundamentally, I agree about the futility of specific forecasts--this is one reason I wrote this post. Now, the IEA is in the "prediction business" but GreyZone has gotten my point that their forecasts seem to go in one direction only and further that they constantly make the same kind of mistakes (and therefore revisions) every year. You would think that they might wake up and smell the coffee but they never do. However, I do believe in the existence of historical trends. IEA's consistently mistaken forecasts seem to ignore supply & demand trends that we should all be concerned about. Extrapolating such trends is risky too but, hey, we can't just pretend they don't exist.

I get the sense that some making comments here think TOD is in the prediction business too (not thinking of you, Halfin). We're not. This tends to be a news, information and analysis site--I think this approach is its principal virtue. To my knowledge, no editor or contributor (including me) at TOD has made any specific prediction about the timeframe for peak oil and the demise of civilization as we know it. Speaking personally, I vacillate on this all the time. There are lots of other peak oil websites you can visit if you want dire predictions of imminent doom but I won't be naming names here. However, TOD is not one of them.

Dave writes about "IEA's consistently mistaken forecasts"; Greko says that "IEA estimations and statistics are always wrong...".

This is exaggerated, as far as I can tell. Have look at IEA's 1996 world oil production forecasts. They were/are far more accurate than those of the peak oil community. Check the following figures, copied from Freddy Hutter's `Trendlines':

Freddy's original contribution is to be found here:

Perhaps I am mistaken about this and have made some awful gaffe  -- if so, enlightenment most welcome!

It is true and important that Colin Campbell has called peak early on a number of occasions and had to revise his predictions repeatedly.  Michael Lynch I believe was the first to analyze this with some care, and reading Lynch's papers has had a big influence on me.  Personally, this would make me wary of accepting any of Campbell's predictions simply on the basis of his authority without having really understood the issues myself (something I'm in the process of doing).  This is not to say I don't admire a lot of things Campbell has done.  However, Colin Campbell is not the only prominent person calling a peak.  For example, Matt Simmons track record at calling peaks and depletion issues in the face of official optimism has been quite good (North Sea, North American NG).  Of course, past performance is no guarantee....

I personally think the state of understanding of bottom-up projection, on both sides of the debate, is marred by a very inadequate picture of decline rates.  Only as we are able to reduce that uncertainty will we be in a position to make better predictions.

You are a terrier on this, decline rates, Stuart, and are very right to be. It is the key to what happens next. Of all the variables it is the one that most determines the future (sudden human commonsense and appropriate action notwithstanding).

I don't scare easily but I am fearful that decline rates will be significantly higher than expected on the major FIP, given the recovery techniques used lately. The initial cliff off PO might be surprisingly steeper than expected. I'm hoping that I have 4 or 5 more years to get organised but there is too good a chance it could hit sooner and fast.

When I look into the future I'm finding a bifurcation around late March 2006 (yes, being 'strange' on this). It's like there are two main probable futures, one where the shit hits fan bigtime, one where it poodles on as presently. They seem about 40% / 60% so far, it will get clearer as that time approaches. I've never felt this extremity of divergence before and what I 'see' frankly scares me.  Things seem to come back together about 4 to 5 years hence but that isn't positive, they come together on the worse path. Seems if we dodge the bad outcome in a few months time it will impose itself later. I've said too much already.

I wouldn't count on getting your 4-5 years to prepare Agric. I doubt that the financial markets will hold up far into 2006 as the credit bubble is about to burst, taking the real estate market in the US, UK and elsewhere with it.

I'm betting we're in for a liquidity crunch and a deflationary depression that will make peak oil preparations far more difficult. On the upside, if there can be said to be one, demand destruction might bring energy (and other) prices down and delay the peak. The downside is that people's ability to afford energy may fall even more quickly than prices, meaning that prices would still rise for most in real terms.

IEA estimations and statistics are always wrong, because IEA has a political function. It always tries to persuade OPEC that oil production is not enough.

See the critique here:

Thanks for the link. Henry Groppe spoke eloquently at ASPO-USA. Check out the summary of their July 2005 report The Oil Business (pdf) from the site. Here's their non-OPEC summary.

Note the large differences with the IEA numbers. Groppe, Long & Littell is one of the oldest and most respected consulting services in the Oil & Gas business. With discrepancies as large as these, what's a poor boy to do?
Groppe has a far better track recoord than IEA, or anybody else, for that matter. Best to look carefully at his tables. If true, we get maybe 1/2% more/year for five years, then 3/4% less/year for five years, then maybe not so good. At today's price the world seems to want 1-2% more/year, so looks like a firm price for quite a while, or until ctl takes up the slack. Maybe we have arrived at a plateau, which might not satisfy wannabe drivers, with new production more or less countering old field declines.

In spite of this, I think prices will be soft at least until summer.

One thing I'd be curious to see with the IEA is some data from how well they did earlier than 5 years ago.

Starting around 1999-2000 the oil market made a dramatic reversal. Prior to that there was a decade-long (two decades if you squint) declining trend in oil prices. Since then oil has of course been on a fairly steady and strong upward trend.

Since the market trend reversed at that time, I wonder if the IEA's behavior also changed then. How accurate were their production estimates during the 1990s? Were they overestimating back then like today? Or perhaps they have a general bias towards "stability" and overpredict production during rising prices, underpredicting during falling prices? It would be interesting to see.

The December 19th issue of the Oil and Gas Journal has a good article called "Global Reserves, oil production show small increases for 2005". Included with this is their "Worldwide Look and Reserves and Production".  This focuses on crude (to the exclusion of NGL's and gas condensate) and shows 0.1% decline in Non-OPEC production from 2004-2005 along with a 2% increase in OPEC production (primarily from Saudi Arabia) to give a worldwide change in crude oil production increase of 0.8% from 2004 to 2005.  

(As a clarifying note, the full year 2005 numbers are only estimates at this point)

Another interesting tidbit is that the report lists the number of producing wells for each country, OPEC, and the world as a whole.  As a rough appoximation, a producing well in the US averages 10 BOPD, for OPEC it averages a little under 1000 BOPD (800-ish) and for the world as a whole each well averages a little under 100 BOPD.

OPEC seems pretty confident that non-OPEC will increase production:
From Rueters
Growing expectations that producers cartel OPEC will cut production when it meets at the end of January added support.

Iran viewed a one million barrel per day cut by OPEC as a "good figure" when the cartel meets on Jan 31, the country's oil minister said in a newspaper interview on Wednesday.

OPEC is likely to decide to cut production from the second quarter of 2006 on expectations the call on its oil will decline by 2 million barrels per day, OPEC president Sheikh Ahmad al-Fahd al-Sabah said on Monday.

1 million?
That ain't messin around.

Hi everyone, Adam Porter here.

Oil Drum is a great site, the debates are really good.

I hope no one gets annoyed by what i said about peak oil. It's a really interesting subject.

I was trying (no doubt poorly) to make the point that this year many different political groups have attached themselves to peak oil. All of them seem to think that peak oil will bring about the stuff they already believe in.

No doubt if there were a group who thought that soon we will be ruled by super intelligent Gibbons they would also think that peak oil will bring this about. :)

It doesn't detract from the debate around the subject, it just means there is often quite a lot of drivel around one has to wade through to get to the interesting bits.

Secondly I try very hard not to give away too many opinions that I have about any subject I cover. Otherwise that becomes writing not journalism. However this piece did sit a bit between the two, that is true.

The piece in question just looked at how futile it is making forecasts. I mean, I could make a few now, but they would be about as valuable as my cat`s biscuits. In fact maybe less so. The IEA don't really have such an excuse (agh! an opinion!)

Making forecasts is and exciting. Predicting the future always is. But one thing about oil is that we don't really know where we are now. So that makes predictions even harder, as i said like trying to predict the weather tomorrow from inside a windowless box.

I was also trying to be fairly light hearted and so on. And i missed this debate as I was on mi hols!

all the best

PS: if you ever want to shout at me you can get me at oilcast at gmail dot com, i even reply to people who hate me (yes that's you Mum ;) ).