A Study of Core-Analyses Data from Flooded-Out Oil Sands

Abstract

Abstract Flood-pot tests have been used for many years to determine the minimum residual oil saturation which would occur with 100 per cent water flood recovery efficiency. To evaluate the validity of such measurements, a study was made of core-analyses data from wells drilled prior to flooding and from wells drilled late in the flood life when operations were at or near the economic limit. The 19 wells in the latter category included in this study were limited to sands where "flushing of cores" was negligible. A comparison of oil saturations measured on fresh core samples in watered-out areas and companion flood-pot samples is shown. While the data indicate that average reservoir residual oil saturation exceeds the flood-pot residual value by 5 to 9 per cent of pore space, they also show that 5 per cent of the samples analyzed from watered-out areas contained no mobile oil and that 15 per cent of the samples contained only 1 to 2 per cent (pore volume) mobile oil. Correlations verifying these results are included and involve 165 cored wells from the Bartlesville formation in Oklahoma and Kansas, and the Bradford formation in Pennsylvania. The data further indicate that the commonly accepted concepts of the effects of permeability distribution on vertical conformance and the "dead areas" on sweep efficiency in pattern water flooding may not be applicable. Introduction In predicting the reserves and performance of a waterflood operation by both analytical and empirical methods, the minimum residual oil saturation after flooding must be determined. Core-analysis data often are used in obtaining this factor. One of two procedures can be followed:consider the oil saturation found by conventional core analysis after correction for shrinkage as the residual oil saturation to be expected from flooding with water, orperform flood-pot test to measure the residual and mobile oil saturations. In the latter, the mobile oil saturations determined from cores for certain areas are accepted as precisely those in the reservoir, although in a majority of pools it is recognized that some flushing of the core takes place and allowance is made for this through empirical factors. Thus, two opposite views are expressed; obviously, there is some merit to both. Spencer and Harding claim that, for determining residual oil saturations for conditions before and after flooding, reliability of core analysis depends on the properties and depth of the rock. These authors believe that during coring there is a tendency for the drilling-fluid filtrate to penetrate the formations ahead of the bit and displace oil. This action is supposed to be a minimum for formations having low vertical permeability and high reservoir pressures. However, the writers have observed that in pressure-depleted formations, or in formations containing oil with little or no gas in solution, little or no flushing occurs. For example, oil saturations greater than 50 per cent have been found in hundreds of cored wells in fields subjected to vacuum production for many years. It is believed that, in most productive formations, gas evolving from solution while the core is removed from the well is the principal reason for loss of oil rather than filtrate invasion during drilling. The following questions arise:Can a representative minimum residual oil saturation (immobile) after flooding be determined from conventional core-analysis liquid saturation tests, or should flood-pot tests be used?Are flood-pot test data usable in predicting waterflood reserves and performance? It was the object of this paper to answer these questions through a study of core-analyses data from watered-out areas. The study also includes data on wells drilled in the theoretically determined dead-area, or area supposedly not contacted by the injected water. It will be recalled that this concept originated from model and other research work performed by Muskat, Wyckoff, Botset and Muskat, and others. JPT P. 114^