WHEN WILL THE WORLD run out of oil? The question has huge implications for political policy and business planning. And in 1997 it had a variety of answers.
According to one set of views on the subject, depletion was in sight for the world`s endowment of conventional oil. Supporters of this position stressed the finite nature of natural resources in general and conventional petroleum resources in particular. Production of depletable resources, they pointed out, follows a predictable pattern. And that pattern indicates that worldwide production of conventional crude oil and natural gas will reach its peak level early in the 21st Century and decline rapidly afterward.
Supporters of the imminent-depletion argument included serious earth scientists as well as pressure groups eager to promote nonfossil energy for environmental or commercial reasons.
In 1997, however, depletion exerted little discernible influence on policy-making. Even in the U.S., which in the late 1970s had based energy policy on official pessimism-soon discredited-about energy resources, fears about running out of oil and gas held much less sway with legislators and regulators than did the environmental consequences of burning hydrocarbons.
Rightly so, argued voices on the question`s other side. In their view, the limiting factor was not the absolute volume of petroleum in nature but rather knowledge about size of the endowment and how to make use of it. The more the oil and gas industry learned about how and where to explore, the more petroleum it found, sometimes in places once thought not to be prospective for hydrocarbons. And the more the industry produced from a specific basin or reservoir, the more it learned and, consequently, the more oil and gas it found itself able to produce. The phenomenon of reserves growth, the tendency of ultimate recovery to exceed original estimates of reserves-often many times over-lent support to the latter view.
In 1997, as worldwide oil production climbed once again while estimates of global reserves held steady, the issue was far from settled. Its importance, however, would only grow with time.
The Hubbert Curve
Central to most discussions of oil depletion is what geologists know as the Hubbert Curve.
Named for geologist M. King Hubbert, who in 1956 and years afterward challenged optimistic expectations about U.S. oil supply, the curve postulates the depletion pattern for any nonrenewable resource. It says that production of such a resource rises from zero to a peak, then falls back to zero along a slope mirroring that of the rise.
Hubbert challenged official estimates of the U.S. petroleum resource as unreasonably optimistic. He predicted ultimate production from the U.S. Lower 48 would total 150-200 billion bbl, which although low by modern estimates was one fourth to one third of what most analysts believed at the time. And he correctly predicted that U.S. production of conventional crude oil would peak in the late 1960s or early 1970s. The peak came in 1970.
If the bell-shaped Hubbert Curve pattern is inevitable, the longstanding decline of conventional oil production in the U.S. should be in an acceleration phase. In 1996 and 1997, the rate of decline slowed. Two years` worth of change is hardly conclusive, however, where long-term trends are concerned.
Pessimists about oil supply tend to concentrate on the accuracy of Hubbert`s prediction about timing of the peak in Lower 48 oil production. Yet other aspects of the Hubbert Curve`s predictive powers have been subjected to question.
In a December 1995 paper, Edward Porter of the American Petroleum Institute noted that estimates of ultimate oil recovery implied by the Hubbert Curve depend heavily on the data period used to estimate parameters of the curve`s mathematical function.
Actual production data for the period 1980-93 have strayed from the curve on the positive side (Fig. 1). By 1993, Porter observed, Hubbert Curve prediction of Lower 48 production was nearly 30% below actual levels, and the disparity appeared to be growing.
Porter said the Hubbert Curve lacks any geologic or economic theoretical justification, is subject to questions about statistical reliability since estimates about recoverable resources are uncertain, and postulates a symmetry in the pattern of production that hasn`t held true.
The curve nevertheless provides a standard proposition about depletion: that the midpoint of depletion occurs when the production rate reaches its peak. The question then becomes how to treat, within the context of the Hubbert Curve, the apparent tendency of reserves to grow over time-or whether the curve should be applied at all.
Backdating
One outspoken observer cautions against making too much of reserves growth. C.J. Campbell, a geologist and petroleum consultant in Millhac, France, points out that companies nearly always understate estimates of reserves at the time of discovery. It is only natural, then, that calculations of a field`s ultimate recovery climb toward a median probability value-or the best estimate of what actually will be produced-as production advances.
Campbell, who outlined his theories in a 1997 book entitled The Coming Oil Crisis and in an Oil & Gas Journal special report article the same year, says reserves growth of that type is really an "artifact of definition and reporting procedure." It must be distinguished from growth resulting from economic and technological factors.
Campbell points out that country reserves estimates vary in their accuracy, some of them being known exaggerations. And he argues that inferences about future discoveries drawn from past trends require that reserves revisions be backdated to the time of discovery. Backdating puts revisions in perspective relative to the standard depletion pattern and, applied to the world total, shows the minor contribution that discoveries make to current reserves additions.
At present, Campbell says, worldwide discoveries total less than 6 billion bbl/year in a downward trend, and production totals about 24 billion bbl/year in an upward trend.
To estimate median probability value of reserves, Campbell adjusts country estimates to remove what he calls "spurious" values, then applies a factor to convert the adjusted number (Table 1). By this reckoning, the world had produced 784 billion bbl of conventional oil by yearend 1996 and had reserves of 836 billion bbl. The sum of the values, 1.6 trillion bbl, is the total discovered. Extrapolation of the discovery trend yields a yet-to-find total of 180 billion bbl and an ultimate total of 1.8 trillion bbl.
Campbell applies these numbers to a depletion model that assumes that peak production in each country comes at the midpoint of depletion and that demand increases at 2%/year. Middle East countries are assumed to play a swing role. When the swing share exceeds 30% of world supply, Campbell expects a sharp increase in oil price and consequent drop in demand (Fig. 2).
Gauging the limits
A different view, one focusing on problems with the concept of resource limits, leads to a very different conclusion.
M.A. Adelman and Michael C. Lynch of the Massachusetts Institute of Technology argue that forecasts of declining reserves and production fail by treating ultimate recovery as a fixed quantity when it is really a dynamic process dependent upon growth in knowledge (Table 2).
A certain understanding about how much oil exists in nature and can be economically produced requires accurate forecasts of future earth science and technology, Adelman and Lynch say. Such forecasts are impossible to make.
Adelman, an economist and MIT professor emeritus, and Lynch, a political scientist and executive director of MIT`s Working Group on Asian Energy and Security, join API`s Porter in citing problems in the predictive power of the Hubbert Curve. The area under the curve, they point out, is supposed to represent ultimately recoverable resources (URR). Yet URR is the final result of changes in costs caused by changes in knowledge. As knowledge grows, the area in the Hubbert Curve representing the future also grows.
Thus Hubbert in 1974 estimated URR for the U.S. at 170 billion bbl. Cumulative production has already surpassed that value, and the U.S. Geological Survey now puts URR at 250 billion bbl.
Similarly, Hubbert estimated URR for the entire world at 1.25 trillion bbl. In 1994 the USGS estimated the total at 2.4 trillion bbl.
Adelman and Lynch also point to pessimistic forecasts, based on analyses of existing fields, that have not come true. Petroconsultants, which maintains authoritative data bases of country reserves, geology, and production, in 1986 predicted that output could only continue to decline in countries not belonging to the Organization of Petroleum Exporting Countries. Yet non-OPEC production has grown impressively in the 1990s. Other such forecasts have similarly underestimated production potential in specific regions.
The MIT analysts attribute the pessimism of those forecasts to assumptions of unrealistically high production decline rates and to failure to account for branch and satellite development projects around existing fields and for projects to improve recovery.
Production decline curves of most fields, they argue, typically take upward steps because of incremental investments. One type of investment improves a field`s recovery factor-the percentage of oil in place ultimately produced. Another type of investment identifies nearby reservoirs by improving knowledge about geology in and around a field. And a third type of investment makes once-marginal prospects in the area of a field economic for development by making transportation services and supply-and-service centers available.
Adelman and Lynch draw a sharp distinction between proved and other types of "reserves." Proved reserves, they say, are estimated assets, essentially an inventory renewable by constant new investment. Probable, possible, and undiscovered "reserves" are, by contrast, not assets but rather implicit forecasts of investment and, therefore, of future technology.
Predictions about ultimate recovery of oil, therefore, because they must assert a level of investment that is unpredictable, are always wrong.
U.S. field longevity
A study reported in 1997 by consultants David B. Hatcher and Arlon R. Tussing of Mercer Island, Wash., underscored the tendency of reserves to "grow" in the U.S. Lower 48.
Reserves of 24.9 billion bbl estimated in 1977 would have been depleted by production by 1987 if additions hadn`t occurred through new discoveries and other means, which of course they did. In 1987, the Lower 48 reserves estimate was 19.9 billion bbl. Most of the additions came not from new field discoveries, however, but from extensions of existing fields and revisions of reserves estimates.
Hatcher and Tussing examined 15 of the largest Lower 48 fields and concluded that production declines in most cases were far slower than both original expectations and traditional decline models. The fields thus have had longer than anticipated lives, so that most of them probably will produce for more than 100 years.
The consultants also combined basin data from the Energy Information Administration with an analysis of several of the largest producing basins in the Lower 48 and observed that:
- Most if not all established petroleum basins in the continental U.S. were discovered in the first half of the 1900s or earlier.
- The peak decades of discovery of the basins were mostly in the first half of the century.
- Changes to estimated reserves in the basins have been mostly upward and typically multiply original estimates by two or more.
- The likelihood that a reserve change will be negative is comparable to the likelihood that it will increase by a factor of 10.
Hatcher and Tussing noted the historic tendency to underestimate reserves at the time of discovery and attributed the longevity of Lower 48 giant fields to steady improvements in geophysical knowledge about the fields and to advances in technology.
Implications
Observations about resources and reserves tend to occur on an academic level. Yet they have large practical implications, not least of which is size of the remaining opportunity to make money producing oil.
A more subtle but nevertheless important implication is that discovery is no longer the core function of the upstream oil and gas industry. To be sure, discoveries continue to be made. But by far the largest share of total additions to reserves comes from extensions to existing fields and revisions and adjustments of reserves estimates for those fields.
Is the heyday for oil and gas discoveries over? That, like size of ultimate petroleum recovery, remains to be seen. For now, however, the industry seems to be in a period in which, while exploration certainly continues, reserves additions come mainly from learning more about what already has been found and how to produce it.
And all this applies only to conventional oil. When analysis confines itself to conventional oil, strictly defined, pessimistic conclusions about oil supply are easy to draw, especially under static assumptions about future technology and investment.
In fact, an enormous resource of heavy oil, tar sand, and oil shale-unconventional oil-has been discovered and can be developed when the technology and economics are right. In 1997, a surge in investment in heavy oil projects may have indicated the process was well under way.
 |
 |
 |
