Uncertainty of Oil in Place in Unconsolidated Sand Reservoirs - A Case History

Abstract

Reinterpretation of core analysis of an unconsolidated sand containing 8,000-cp oil indicates that the actual porosity of the sand in place is much lower than that arrived at even by special measurements on core samples recompacted under high confining pressure! The density log and sonic log from the same well support this conclusion. Introduction Accurate determination of oil in place in a reservoir is important when decisions are being made regarding development of a field; it is even more important later when decisions are made regarding installation of fluid injection projects when less of the oil remains; and it is extremely important in considering the recovery of additional oil by tertiary methods. Basically, volumetric estimates involving core analyses, fluid samples, well tests, well logs, and other geologic information are the only direct measure of oil initially in place. Frequently, interpreting pressure-production performance of the reservoir through material balance techniques performance of the reservoir through material balance techniques helps to establish the reliability of volumetric estimates. In some very heterogeneous reservoir rocks or in some reservoirs of limited areal extent, a material balance estimate is superior to the volumetric estimate. Uncertainties exist in all factors involved in both types of estimates. Because of alteration of cores during coring, handling, and analysis, volumetric estimates of oil in place in unconsolidated sand reservoirs are subject to added uncertainties. Unusual circumstances in the basal Oil Creek sand in the S.E. Pauls Valley field, Garvin County, Okla., provide some quantitative insight into these uncertainties. The reservoir rock is a completely unconsolidated, clean, well sorted, silica sand. It contains 10 degrees API gravity oil having a viscosity of about 8,000 cp at reservoir temperature. The oil has essentially no gas in solution. Thus, alteration of oil saturation by invasion of water-base mud filtrate and by release of pressure in bringing the core to the surface was minimized. In the example well drilled in 1966, 80 ft of oil-saturated core was recovered with a rubber-sleeve core barrel. This sand was so soft that it slumped by its own weight when the rubber sleeve was cut away from the core. One core sample per foot had fluid content measured by conventional, routine per foot had fluid content measured by conventional, routine analysis by a commercial laboratory. Porosity was determined by the sum-of-fluids technique. Eight samples, still in the rubber sleeve, were recompacted under simulated overburden pressure before porosity and fluid saturations were determined. pressure before porosity and fluid saturations were determined. Open-hole logs obtained from the same well included a dual induction-laterolog, a compensated formation density logs, and a sonic-gamma ray log. This paper reviews the various estimates of oil in place that might have been made if only parts of the core analysis and log data had been available. Conventional Core Analysis For those who are unfamiliar with the laboratory procedure, the sum-of-fluids porosity technique involves placing a weighed sample of core in a mercury porosimeter in which the bulk volume of the sample is first determined by displacement of mercury. Then mercury is injected into the core under high pressure to determine the volume of gas-filled space in the core sample. A weighed companion sample, or in some cases the mercury-impregnated sample, is retorted to determine the quantities of oil and water present. JPT P. 1315