282 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1916. 



this quantity possibly 10 per cent to 15 per cent is lost in production 

 or used for fuel, so that of the total supply but 34 per cent to 54 per 

 cent ordinarily is marketed. Many years may be required to make 

 even this recovery. 



It is the writer's belief that estimates based on the saturation method 

 are much less reliable and satisfactory than those worked out through 

 the production-curve method, but the former must be used for new 

 or poorly developed fields and will be briefly described. 



The thickness of producing oil sands or oil zones varies from 2 feet 

 in the Illinois field to over 200 feet in the California field. Total 

 supply or saturation as marked by the porosity varies from a trace, 

 in sands, up to 50 per cent in some exceedingly porous dintomaceous 

 shale from California. Between 5 and 15 per cent is the average for 

 sands, some, however, going as high as 30 per cent. An acre of 

 ground covered with oil a foot deep (1 acre-foot) contains 7,758 bar- 

 rels. This would be complete saturation for the 43,560 cubic feet. 

 Assmning an average of 10 per cent saturation would give 775.8 bar- 

 rels per acre-foot for normal conditions. On this basis a 5-foot sand 

 would contain 3,879 barrels per acre, and a 50-foot sand 38,790 bar- 

 rels. Actual yields of over 100,000 barrels per acre are known. 

 Estimates of the average production per acre for the various States 

 are given in Table III. Most of these figures are based primarily 

 on the production-curve method, but a few are based on or checked 

 by the saturation method. 



, PRODUCTION-CURVE TVIETHOD. 



General statement. — Estimating future production or supply by 

 a plotting of hypothetical curves, based on actual figures in well- 

 known areas or fields, is the safest method, as it involves factors 

 which it is possible to obtain. Another thing in its favor from the 

 standpoint of the producer and marketer of oil is that it is based 

 on and has to do with actual " net " oil figures, instead of theoretical 

 quantities. 



Basis of theoretical curve. — The theoretical curve shown in the 

 diagram accompanying this article is based primarily on the yearly 

 total production figures of New York and Pennsylvania. These 

 figures cover a period of productivity of 54 years, longer by far 

 than that of any other field in the United States. Furthermore, over 

 this period this field has been subject to all of the vicissitudes from 

 both natural and artificial causes that beset oil fields in general. The 

 area involved in the Pennsylvania and New York field is greater 

 than that in any other field in the United States, which is still 

 another reason why the result should be conservative. 



The interesting part of the production curve is that following the 

 period of maximum yield. In some instances it is fairly safe to 



