MINERAL RESOURCE ESTIMATES AND PUBLIC POLICY 



15 



drift, seawater, or other materials that conceal pos- 

 sible mineral-bearing rocks or structures assures us 

 that undiscovered deposits are still to be found. 

 Qualitatively, at least, we know something about 

 the distribution of minerals with respect to other 

 geologic phenomena and, if this is so, we have a 

 chance of developing quantitative relations that will 

 give us at least a start. 



Two principal approaches to the problem have 

 been taken thus far. One is to extrapolate observa- 

 tions related to rate of industrial activity, such as 

 annual production of the commodity ; the other is to 

 extrapolate observations that relate to the abun- 

 dance of the mineral in the geologic environment 

 in which it is found. 



The first of these methods has been utilized by 

 M. K. Hubbert (1969), C. L. Moore (1966), and 

 M. A. ElHott and H. R. Linden (1968) in estimat- 

 ing ultimate reserves of petroleum. The essential 

 features of this approach are to analyze the growth 

 in production, proved reserves, and discovery per 

 foot of drilling over time and to project these rate 

 phenomena to terminal values in order to predict 

 ultimate production. Hubbert has used the logistic 

 curve for his projections, and Moore has utilized the 

 Gompertz curve, with results more than twice as 

 high as those of Hubbert. As Hubbert has pointed 

 out, these methods utilize the most reliable informa- 

 tion collected on the petroleum industry; modern 

 records on production, proved reserves, number of 

 wells drilled, and similar activities are both rela- 

 tively complete and accurate, at least as compared 

 with quantitative knowledge about geologic features 

 that affect the distribution of petroleum. 



The rate methods, however, have an inherent 

 weakness in that the phenomena they analyze reflect 

 human activities that are strongly influenced by 

 economic, political, and other factors that bear no 

 relation to the amount of oil or other material that 

 lies in the ground. Moreover, they make no allow- 

 ance for major breakthroughs that might transform 

 extensive paramarginal or submarginal resources 

 into recoverable reserves, nor do they provide a 

 means of estimating the potential resources of un- 

 explored regions. Such projections have some value 

 in indicating what will happen over the short term 

 if recent trends continue, but they can have only 

 limited success in appraising potential resources. 



Even the goal of such projections, namely the 

 prediction of ultimate production, is not a useful 

 one. Not only is it impossible to predict the quanti- 

 tative effects of man's future activities but the con- 

 cept implies that the activities of the past are a 

 part of an inexorable process with only one pos- 



sible outcome. Far more useful, in my opinion, are 

 estimates of the amounts of various kinds of ma- 

 terials that are in the ground in various environ- 

 ments ; such estimates establish targets for both the 

 explorer and the technologist, and they give us a 

 basis for choosing among alternative ways of meet- 

 ing our needs for mineral supplies. 



The second principal approach taken thus far to 

 the estimation of undiscovered resources involves 

 the extrapolation of data on the abundance of min- 

 eral deposits from explored to unexplored ground 

 on the basis of either the area or the volume of 

 broadly favorable rocks. In the field of metalliferous 

 deposits, Nolan (1950) pioneered in extrapolation 

 on the basis of area in his study of the spatial and 

 size distribution of mineral deposits in the Boulder 

 Dam region and in his conclusion that a similar 

 distribution should prevail in adjacent concealed 

 and unexplored areas. Weeks (1958, 1965) and Pratt 

 (1950) played similar roles with respect to the esti- 

 mation of petroleum resources — Weeks extrapolat- 

 ing on the basis of oil per unit volume of sediment 

 and Pratt on the basis of oil per unit area. Many 

 of the estimates of crude oil that went into the 

 NPC study were made by the volumetric method, 

 utilizing locally appropriate factors on the amount 

 of oil expected per cubic mile of sediment. Olson and 

 Overstreet (1964) have since used the area method 

 to estimate the magnitude of world thorium re- 

 sources as a function of the size of areas of igneous 

 and metamorphic rocks as compared with India and 

 the United States, and A. P. Butler (written com- 

 mun., 1958) used the magnitude of sandstone ura- 

 nium ore reserves exposed in outcrop as a basis for 

 estimating the area in back of the outcrop that is 

 similarly mineralized. 



Several years ago, Zapp (1962) and Hendricks 

 (1965) introduced another approach, based on the 

 amount of drilling required to explore adequately 

 the ground favorable for exploration and the re- 

 serves discovered by the footage already drilled — 

 a procedure usable in combination with either the 

 volumetric or areal approach. Recently J. B. Zim- 

 merman and F. L. Long (cited in "Oil and Gas 

 Journal, 1969") applied this approach to the esti- 

 mation of gas resources in the Delaware-Val Verde 

 basins of west Texas and southeastern New Mexico ; 

 and Haun, Barlow, and Hallinger (1970) used it to 

 estimate potential natural gas resources in the 

 Rocky Mountain region. In the field of metals, 

 Lowell (1970) has estimated the number of undis- 

 covered porphyry copper deposits in the southwest- 

 ern United States, Chile and Peru, and British Co- 

 lumbia as a function of the proportion of the 



