50.

are examples of provident technology which created great new reserves, and these reserves subsequently have been further extended by economic improvements in mining and processing methods. Concurrent decreases in energy costs played no small part in keeping production and processing costs down as the grade of ore decreased.

In recent years, however, the real costs (in constant monetary units) of producing metallic aluminum and copper have started to rise, partly because of increases in energy costs, but largely because limits on the efficiency of extraction and processing are being reached and no new technologic break through has occurred. Consequently, further decreases in the work expended per unit of metal produced have not been possible and net work profit, after decades of increase, is starting to decline. For most mineral commodities, technology appears no longer able to keep ahead of increasingly adverse geologic parameters. On the other hand, because the known deposits of aluminum, iron, and copper ore are large, and many have gradational boundaries, real costs of these metals can be expected to advance slowly rather than swiftly as would be the case with materials with less abundant reserves (in relation to demand) and whose deposits have sharp boundaries.

An example of the latter is natural gas, where demand already has outrun the supplies available at costs near those of recent record. In fact, it is the natural gas situation that seems to offer the most persuasive argument against the utility of the economic approach to estimation of the availability of a nonrenewable resource. It is highly improbable that complete abandonment of price controls on natural gas in the United States, or even the creation of price supports for domestic production at a level five times the recent controlled wellhead price,

51.

would result in domestic production sufficient to close the "unsatisfied demand" gap shown on the U.S. gas supply-demand projections (Fig. 15) published recently by the Federal Power Commission.

There may well be sufficient gas "in place" to satisfy the projected demand, but the exponential imperative that skyrockets production costs with increasing depth and with increasing "tightness" of reservoir rocks puts a very real limit on the amount producible at prices which could be justified by most uses of the fuel. Natural-gas reserves in the United States are running out, and estimates of demand based on untenable assumptions of availability, as well as optimistic estimates of gas-inplace, are of relatively little use in planning for the conversion to other fuels that must now take place. Stimulation of domestic production and imports of liquefied natural gas can only buy, at substantial expanse, the time required for such conversions.

The geologic-analogy method of calculating potential resources, although it implicitly recognizes physical limits as the overriding control on ultimate recovery, is based on three assumptions that are debatable if not downright untenable. The first is that oil exists in unexplored basins and other areas in the same ratio to the volume of the host rocks as it has been found to exist in the basins and areas of production. Very few sedimentary basins are unexplored simply because no one has thought to go there or because it was impossible to get there and carry out exploration; only basins covered by deep water, thick alluvium or an ice cap fit this criterion. All other basins are unexplored at least in part because of lack of evidence of oil being there. Only if

50

Figure 15

1990

53.

exploration and development over the past hundred years had progressed outward as a wave front from one or more points chosen at random within each continent would this first assumption of the geologic-analogy method appear valid.

The second debatable or untenable assumption is implicit in the method, and it is that the price of oil will increase to meet the greater costs of producing oil from deeper horizons, from more hostile environments, and from reservoirs more distant from the centers of use.

A third questionable assumption implicit in the geologic-analogy method is that the sum of all technological and political impacts on reserves during the period of use of the potential reserves will not be negative. As a matter of logic, a technologic breakthrough in substitution technology (in the production of synthetic crude oil, for example) would have a strongly negative effect on petroleum reserves; indeed, the moment it costs more to find and produce a barrel of "new" petroleum than it does to produce a barrel of synthetic crude from coal, oil shale, or tar sands, there will be no more "potential" petroleum reserves and the question of how much petroleum remains to be found will be of no further importance. Political decisions to extend the geographic areas in which petroleum exploration and production are prohibited would also have a negative impact on potential reserves.

The geologic-analogy method has led to two practices which deserve mention. One is misleading and unjustifiable; the other is simply misleading. The first is the practice of adding together subjective estimates of differing magnitudes and ranges of uncertainty to produce a single estimate of "potential reserves". As noted previously in this paper,

54.

such practice is logically unsupportable; more importantly, as Hubbert has pointed out (1971, p.10), "it is possible that governmental and industrial policies based on such estimates could within a matter of decades prove to have been seriously misguided."

The second practice that can mislead the unwary involves the use of oil-in-place estimates, modified or unmodified by ultimate-recovery ratios. The National Petroleum Council's 1971 estimate for total oilin-place discoverable (in the U.S.) of 727.4 billion barrels appeared very reassuring if one made the mistake of comparing it with the "oil produced" total of about 90 billion, or with the 1971 domestic production of 3.5 billion. Only when one noted in the NPC estimates that there was less oil-in-place remaining to be discovered (332.0 billion barrels) than had already been discovered (395.5 billion barrels) and when one calculated a 30.4 percent cumulative recovery efficiency from the NPC estimate of the oil that will ultimately be recovered from the found portion, did one begin to realize that the NPC estimate was much more conservative than it appeared at first glance. In fact, application of the NPC estimate of 37 percent cumulative recovery efficiency (to be attained in 1985) to the total estimate of oil-in-place discoverable and addition of the result to the NPC estimate of total oil to be recovered from already-found oil, results in a figure of 242.5 billion barrels for ultimately recoverable U.S. petroleum, far below the estimate of the same year by the American Association of Petroleum Geologists of 432 billion barrels (based on the wildly optimistic assumption of 60 percent ultimate recovery of discovered oil-in-place) and only 27.6 percent above Dr. Hubbert's estimate of 190 million barrels, which had been called unrealistically pessimistic by many in the oil industry.