Will the U.S. overtake Saudi Arabia as the world's top oil producer by 2020?
That's what every headline blared after the release of the International Energy Agency's World Energy Outlook (IEA WEO) on November 12.
Ever eager to report an upbeat story on energy, energy journalists everywhere seized on that part of IEA's forecast, while generally ignoring the caveats. Most of the increase in U.S. production is expected to come from additional "tight oil" production from shale formations like the Bakken in North Dakota and the Eagle Ford in Texas. But, as IEA's chief economist Fatih Birol noted, "Light, tight oil resources are poorly known ... If no new resources are discovered (after 2020) and plus, if the prices are not as high as today, then we may see Saudi Arabia coming back and being the first producer again."
Further, IEA warned that in order to limit global warming to 2°C above pre-industrial levels -- the generally accepted target to stabilize carbon emissions and avoid dangerous climate change -- no more than one-third of the world's remaining proven fossil fuel reserves can be burned, and massive global investments in energy efficiency and low-carbon energy technologies will be needed.
But why let dirty details like those get in the way of a good story? Tell us more about how we'll beat Saudi Arabia and become energy self-sufficient by 2035! Let's hear about North America becoming a net oil exporter by 2030! By all means, tell us how this report means that the "peak oil idea has gone up in flames!"
As always, I prefer to look at the data.
The IEA scenario
The core of the IEA's forecast for U.S. production is expressed in this chart:
U.S. tight oil production data is still notoriously hard to come by, as output is still growing rapidly and the data is reported by the states with various lag times. The most recent figure I could obtain from the U.S. Energy Information Administration (EIA) was 950,000 barrels per day as of April 2012, as cited in an August 2 presentation to the U.S. House of Representatives by EIA chief Adam Sieminski. Further muddying the waters, some analysts lump natural gas liquids in with their tight oil production figures, for a total of 1.2 million barrels per day (mb/d). Others, like the energy consultancy Wood Mackenzie, put U.S. tight oil production at 1.5 mb/d this year. For our purposes today, I'll call current tight oil production a round 1 mb/d.
In the IEA WEO, tight oil production is expected to reach over 3.2 mb/d by 2025 in their New Policies Scenario, implying an increase of around 2.2 mb/d over current levels.
According to the EIA's most recent data (in the October Monthly Energy Review), the U.S. produced an average 6.225 mb/d of crude oil plus condensates (natural gas liquids which are naturally associated and produced along with the crude) in the first seven months of 2012, including tight oil. Over the same period, Saudi Arabia produced 9.926 mb/d of crude plus condensates.
Adding the anticipated 2.2 mb/d increase in U.S. tight oil would bring the U.S. crude plus condensate total to 8.425 mb/d, which is less than Saudi Arabia's current production. Under the IEA's forecast that Saudi Arabia's production will be roughly the same in 2025 as it is today, the U.S. will not overtake it in oil production.
So how did the IEA come up with the figures suggesting that the U.S. would overtake Saudi Arabia? By including stuff that is not oil, mainly natural gas liquids (NGLs). Including these additional liquids would bring the U.S. total to 11.1 mb/d in 2020, versus Saudi Arabia's 10.6 mb/d.
There are just a few problems with this scenario.
First, as I detailed in May ("Fuel to Byrne"), natural gas liquids are not equivalent to oil. They contain only about two-thirds of the energy content of oil; they are far less versatile in their applications; and only about 19 percent of a barrel of natural gas liquids is actually usable as vehicular fuel. If the implication is that the U.S. is going to be able to drive around on its own domestically-produced fuel, then that implication is simply wrong.
Second, even in the IEA's model, the U.S. only surpasses Saudi Arabia for about five years, and only by about 0.5 mb/d at the peak. Australian energy blogger Matt Mushalik helpfully charted this brief, shining moment in his excellent summary of the IEA report:
Source: Crude Oil Peak
This also implies a reduced outlook for Saudi production, which the EIA projected last year would rise to 13.9 mb/d by 2025. Even if this optimistic scenario pans out for the U.S., it's hardly good news from a global standpoint.
Third, the IEA's forecast for the U.S. through 2035 depends on about 2 mb/d of fields "yet-to-be found" and "yet-to-be developed." This has become a regular feature of the IEA's forecasts, and it remains as uncertain as ever. Energy journalist Mason Inman referred to it as the "wedge of hope" in his discussion of the IEA's WEO 2010. If that wedge of hope fails to materialize, then U.S. production in 2035 would be around 7.2 mb/d, below the roughly 8 mb/d of crude and NGLs the U.S. now produces according to the EIA's AEO2012 Reference case. (Including biofuels and refinery gains, U.S. production is now 10.21 mb/d in the AEO2012 Reference case.)
Fourth, the forecast of 3.2 mb/d by 2025 in the IEA WEO is significantly higher than even the most optimistic EIA scenarios published four months earlier, as another of Mushalik's helpful charts shows:
Source: Crude Oil Peak
Fifth, the IEA's forecast is also substantially higher than a bottom-up reckoning of U.S. tight oil production suggests. A new analysis of the HPDI database -- one of the few comprehensive well-by-well databases in existence -- by veteran Canadian geologist David Hughes calculated the average decline rates for recent tight oil wells against the number of remaining prospective drilling locations estimated by the EIA. He found that at current rates of drilling, U.S. tight oil would peak at around 2.23 mb/d in 2016 -- a full 1 mb/d below the IEA's scenario. At a slower rate of drilling, U.S. tight oil would peak at around 2 mb/d in 2018. If Hughes is correct, then the entire US tight oil boom will have crested and begun its decline by 2020, the year in which IEA expects the US to overtake Saudi Arabia. (Hughes' report is scheduled for released in January 2013 by the Post Carbon Institute.)
A recent analysis by Bernstein Research matches Hughes' forecast quite closely, seeing production from crude, condensates and NGLs of 10.5 mb/d by 2015, a 2 mb/d increase over their 8.5 mb/d estimate for this year. After that, they expect U.S. production to fall back to 9 mb/d by 2020.
Sixth, the IEA's scenario depends on what appears to be a simplistic assumption for future oil prices. Their base price scenario depends on a crude oil import price of $120/bbl (as compared with an average $108/bbl in 2011), a highly uncertain assumption for a 23-year forecast period. In his note to clients about the IEA report, Deutsche Bank analyst Paul Sankey wrote that the anticipated increase in tight oil would overwhelm U.S. refiners' already-falling demand for light sweet crude, driving prices and profits below the threshold needed by U.S. producers to sustain such an increase in drilling. Profit margins could decline to $10 - $20/bbl and stifle the appetite of investors in a highly debt-driven business. "In short, a prediction that the US will become the world's largest oil producer is a self-defeating proposition," Sankey summarized.
While I agree with Sankey's point, I'll reiterate that there is also a limit to what consumers will pay for gasoline and diesel. As I detailed in June, the current ceiling for oil prices appears to be around $125/bbl for the global Brent benchmark (or $105/bbl for the WTI U.S. benchmark). The IEA's average price assumption is just below that ceiling, implying that prices would certainly exceed it repeatedly over their 23-year scenario, killing demand and further weakening investor appetites each time it did. Indeed, energy analyst Gail Tverberg calculates that on current trends, the real cost of a barrel would be $169 in 2020 and $467 in 2035. The IEA does not recognize a price floor nor a ceiling in its scenario. If global crude production begins its long decline around 2015, as I (and others, including former IEA forecaster Olivier Rech) expect, then the stable, modest prices IEA assumes through 2035 are simply off the table.
More dubious assumptions
To complete the picture of U.S. energy self-sufficiency, IEA assumes that oil demand can be cut by 5 mb/d, a 26 percent drop from the current consumption level of 19 mb/d. This would reduce net U.S. oil imports to 3.4 mb/d.
The vast majority of the reduction would come from "demand-side efficiency," which I assume means more efficient vehicles. As I explained in September, the majority of the decline in U.S. oil consumption in recent years owes more to the recession than to improving vehicle efficiency, so I am skeptical of this claim.
IEA also foresees a significant increase in biofuel usage, a particularly dubious proposition considering that corn ethanol production is no longer subsidized and is more likely to decline than increase from here. Apparently IEA hopes that alternative biofuels made from crop waste, switchgrass, and other biomass (some of it imported) can supply this wedge, but after many years of watching such sunny forecasts wreck on the rocks of real-world price and scalability issues, I am inclined to disregard this portion of their forecast as well. Only their tiny expectation for increased natural gas transport looks reasonable, and in fact may be too pessimistic. By 2035, I think switching transport trucks to natural gas could save as much as 1 mb/d, at least from a technical and economical standpoint.
The remainder of the IEA's forecast, in which the U.S. "becomes all but self-sufficient in net terms by 2035" [emphasis mine] presumably relies on additional exports of coal and natural gas. But even with their optimistic assumptions, they do not appear to have eliminated U.S. oil imports by 2035.
In summary, the IEA's forecast shows the U.S. exceeding Saudi Arabia in liquids (not oil) production for a brief period of time, under highly optimistic assumptions. The U.S. might become energy self-sufficient on a net BTU basis, after accounting for increased exports, but it would not eliminate oil imports. And the prospect of North America becoming a net oil exporter by 2030 depends on a host of additional uncertainties in Canada and Mexico; even if that headline-grabbing milestone were achieved, which I think is unlikely, it would not result in lower fuel prices for North Americans.
Finally, a puny 3.2 mb/d of U.S. tight oil production by 2035 hardly dismisses the prospect of peak oil in a world that burns over 90 mb/d with a background decline rate of over 5 percent per year, according to the IEA's own calculations. Indeed, in the new report the IEA forecasts "peak demand" before 2020 in several scenarios, which longtime observers will instantly recognize as code for "oil supply is going to be a problem, but we'll get around it through better efficiency and fuel-switching without uttering the dread words peak oil."