Previous long-term government, industry oil forecasts badly overestimated supply; why should we listen now?

[I]f you're still operating under the assumption that the earth's petroleum--or at least the cheap stuff--is about to run out, you're not going to thrive in the new oil era. Technology is making it possible to find, produce, and refine oil so efficiently that its supply, at least for practical purposes, is basically unlimited.

--Businessweek, December 14, 1998

The writer of the above sentences was reacting to oil prices hovering around $11 a barrel. He could not have known then that we were about to embark on a bull market that would take oil to its highest price ever--even adjusted for inflation--just 10 years later. And so, after oil's run, it's all the more astonishing that as Brent Crude--now the true worldwide benchmark price--stands above $100 a barrel, we are hearing a similar message about the future of oil both from official agencies and the oil industry.

The reverence accorded each new forecast of future energy supplies from international and government agencies and from major oil companies seems to go far beyond that accorded to the oracle of Delphi in ancient Greece. That oracle's record may be lost in the mists of time, but we can check the record for these modern energy oracles.

The U.S. Energy Information Administration (the statistical arm of the U.S. Department of Energy), the Paris-based International Energy Agency (a consortium of 28 countries), the National Intelligence Council (an advisory body to the U.S. Director of National Intelligence) and the oil giant ExxonMobil, all regularly release long-term forecasts for world energy supplies. The last three have released their latest forecasts this fall. The U.S. EIA updated its world projections in 2011.

Looking back at forecasts made in the year 2000 by the U.S. EIA, the IEA, and the NIC, it becomes obvious that drawing an upward line on a chart does not make an oil forecast magically come true. All were considerably off the mark. ExxonMobil's oldest forecast available online dates back to 2006. It, too, has proved wide of the mark.

First, let's see by how much each forecast missed. Reports issued in the year 2000 by the U.S. EIA and the IEA contained similar projections. The U.S. EIA forecast that total world liquid fuel supplies would reach 93.2 million barrels per day (mbpd) in 2010. The IEA forecast 95.8 mbpd. Though the NIC report did not provide an explicit forecast for 2010, the implied forecast was around 92 mbpd. All those numbers include not only crude oil and lease condensate which constitute the proper definition of oil, but also natural gas plant liquids (only a fraction of which can be substituted for oil) and refinery processing gain (which is the result of applying energy to break oil into its components, causing the final volume to expand). We can now check those numbers. Actual total worldwide liquid fuel production for 2010 was 87.1 mbpd.

All three groups overestimated production by a considerable margin. This helps to explain the colossal miss on prices. The U.S. EIA report included a price projection for crude oil of about $28 a barrel for 2010 (adjusted for inflation). The actual average price for oil traded on the New York Mercantile Exchange in 2010 was $79.61. The 2000 IEA report forecast an inflation-adjusted price for oil in 2010 only 25 cents higher than the U.S. EIA forecast. The NIC report did not provide an explicit price forecast for oil, but did say this: "Meeting the increase in demand for energy will pose neither a major supply challenge nor lead to substantial price increases in real terms." All three groups failed to anticipate the plateau in worldwide crude production that began in 2005. All failed to gauge properly the pace of growth in oil demand in Asia, particularly China and India, which put upward pressure on prices.

Of course, it's easy to pick apart long-term forecasts when the actual data become available. But, the point here is not that the forecasts were wrong, but that they were all wrong in the same direction, namely, overestimating actual production. Taking an average of all three would still have resulted in a substantial overestimate. That's a serious concern because the forecasts provided by these groups are used worldwide for government and corporate planning and policy purposes. They are extremely influential. And, yet experience should have taught us by now that long-term energy forecasts by anyone--even people whose job it is to study energy markets and supplies--are a poor guide to policy and planning.

There is a basic asymmetry in the effects of energy supply forecasts. If an oil production forecast promises a business-as-usual future (i.e., continually growing production) as all three forecasts mentioned above did and that forecast turns out to be too low, the mistake is benign for most people. Extra supply means lower prices and therefore more money available for other things. If, however, such a forecast turns out to be too high, the consequences can be severe because the global system we now have is acutely sensitive to changes in the price and supply of energy, especially oil. We have seen just how sensitive it can be as we've watch oil prices reach historic highs in the last decade and remain high. Negative supply surprises have the potential to undermine the very stability of our global system, and the only way to prevent that is to prepare for scenarios that these official reports refuse to contemplate.

Now keep in mind that the comparisons made here between forecast and actual production and prices are for a 10-year period. The latest U.S. EIA and IEA world forecasts extend out 23 years to 2035. The newest NIC forecast purports to know the state of the world in 2030. The accuracy of any forecast deteriorates rapidly the further it goes into the future, and these forecasts go out about twice as far. We are taking many more times the risk if we rely on them. This is because energy transitions can take up to 50 years. Waiting until the last minute to begin the inevitable transition away from fossil fuels could cause terrible discontinuity and possibly disaster.

Imagine how different government energy policy and corporate planning would have been had all three forecasting groups predicted in 2000 that oil prices would rise above $100 a barrel by late in the decade. Imagine if all three groups had predicted a plateau in the worldwide rate of production of crude oil proper (defined as crude oil including lease condensate) starting in 2005. Imagine further if all three groups had predicted that global net exports of petroleum liquids--the petroleum fuels available for import by such importers as the United States, China, India, Japan and most of Europe--were going to decline consistently starting in 2006, leading to intense competition for supplies among importing nations.

There were voices warning that such things might happen. But the entrenched institutional prejudices in all three groups prevented them from contemplating these outcomes. So deep are those prejudices that none of the three seems yet to have picked up on the issue of shrinking global net exports even though it is now clear in the data. Of course, it is safer to be wrong when the vast majority are wrong with you. That way you can say, "Nobody could have seen it coming."

As the realities of constrained worldwide oil supply have become apparent, all three groups have gradually lowered their long-term oil supply forecasts. But, all three continue to believe that supply will be there to match projected demand, a dubious assumption given the experience of the last decade.

It's true that forecasts can miss by being too pessimistic as well. None of the three groups foresaw the shale gas phenomenon back in 2000. It was assumed that North America would be importing a considerable amount of liquefied natural gas (LNG) by now as domestic supplies declined. Having missed the rise in gas production, it is possible that all three groups are now simply following the trend and projecting it forward with little skepticism--much as they did for oil in 2000.

One thing they all seem to be missing is that production of large amounts of shale gas will depend on much higher prices as drillers move from the easy-to-get gas--which is currently flooding the North American market at prices that are below the cost of production--to the difficult-to-get gas that will flow at slower rates and be much more costly to extract. They also seem to be missing the fact that high decline rates for such wells mean that rig counts and infrastructure will have to expand almost geometrically to keep supplies growing. That expansion will hit a wall at some point when the price of natural gas rises to reflect this reality and forces some consumers to cut back on natural gas use. Already U.S. natural gas production has been essentially flat since December 2011 as prices have vaulted from multiyear lows. This comes after years of persistent growth in supplies from the low seen after Hurricane Katrina in 2005.

Because ExxonMobil has recently released its 2013 Outlook for Energy with projections to 2040, I had hoped to find the company's report from 2000. The oldest I could find came in the form of a slide show from 2006. Even at this late date, ExxonMobil's forecast was predicting consistent, uninterrupted growth in the worldwide production of crude oil proper and assuming that North America would need considerable LNG imports in the coming decade. The report shows that the conventional wisdom remained intact well into the period when underlying events made them no longer tenable.

ExxonMobil's latest report, not surprisingly, concludes that fossil fuels will continue to provide the bulk of the world's energy well into the future and that there will be plenty of them. With the media repeating what will inevitably turn out to be a flawed forecast, they are forgetting to point out the obvious. The company has an interest in convincing consumers and shareholders that oil and natural gas will be the dominant fuels of this century--which is all the more reason to be skeptical about the company's projections.

When the U.S. EIA, the IEA and the NIC made their long-term forecasts in 2000, supposed game-changing technology was going to make it possible to extract oil in the Arctic and in deepwater "at improbably low costs." The NIC even prophesied that natural gas from methane hydrates, essentially methane trapped inside ice crystals in the deep ocean, would become an increasing part of the natural gas supply. No commercial production of methane from methane hydrates has so far taken place.

And, with regard to oil, even though prices have risen from an average $30.26 in 2000 to an average of $94.60 this year on the New York Mercantile Exchange, we are told again by all three groups that a new miracle technology called hydraulic fracturing is going to make future oil supplies plentiful. (By the way, that technology is 60 years old.) Given their record on such pronouncements, we would be wise to be cautious.

Long-term forecasts are inherently unreliable. In the case of the U.S. EIA, the agency does provide three forecasts based on various price assumptions. But price is not the only variable, and among those forecasts, even the most conservative in 2000 was still too optimistic about supplies and wildly wrong about prices. At the very least, all long-term forecasts should have wide error bands around them. Those error bands would aid us in understanding the risks. No one can know the future. So, we are left with evaluating scenarios that help define the risks we face.

When it comes to policy, it is not the benign energy supply scenario which should guide us, but the most severe because it has the potential to undermine the very stability of global society. It may be for political reasons that statisticians who plot the data for such projections choose to leave out the error bands which they know ought to be there. If policymakers and planners understood just how big the uncertainty about future fossil fuel supplies is, they might panic. But, they might also do something else; they might quickly wean society off finite fossil fuels wherever possible in favor of energy sources such as wind and solar which won't be running out any time in the next few billion years. And, they might also require deep reductions in energy use starting now to guard against the day when fossil fuels decline.

It comes down to whether it is wise to continue with a system of energy, the stability of which is entirely dependent on highly uncertain long-term forecasts for fossil fuel supplies. I repeat: History tells us that it can take up to 50 years to complete an energy transition. Previous transitions gradually moved us to new fuels having increasing energy densities--wood to coal, coal to oil and natural gas, oil and natural gas to uranium. But coal, oil, natural gas and uranium are all finite, and we will someday--perhaps very soon in the case of oil--find that their supply cannot grow any more and will even begin to decline.

When long-term forecasts promise energy that is both cheap and plentiful as the U.S. EIA, the IEA, and the NIC reports did in 2000, governments, businesses and individuals do little to prepare for scarcity. Wouldn't it be wiser to build an energy system which would free us from the inherently risky and unreliable racket of long-term forecasts? Wouldn't it be wiser to build an energy system that is forecast-proof because the energy that powers it is constantly renewed?

Kurt Cobb is an author, speaker, and columnist focusing on energy and the environment. He is a regular contributor to the Energy Voices section of The Christian Science Monitor and author of the peak-oil-themed novel Prelude. In addition, he writes columns for the Paris-based science news site Scitizen, and his work has been featured on Energy Bulletin, The Oil Drum, OilPrice.com, Econ Matters, Peak Oil Review, 321energy, Common Dreams, Le Monde Diplomatique and many other sites. He maintains a blog called Resource Insights and can be contacted at kurtcobb2001@yahoo.com.