Evaluation of Waterflood Residual Oil Saturations Using Log-Inject-Log Procedures

Abstract

The log-inject-log procedure provides more accurate reservoir saturation values by eliminating the need for some of the basic parameters required in the conventional analyses of a single log. Twelve examples of the utility and limitations of the technique are presented. Introduction The petroleum literature of the last decade has provided evidence of a continually increasing interest in improving recovery of oil from known reservoirs. Tests to evaluate and develop tertiary oil-recovery methods have been numerous. These efforts have emphasized the need to define quantitatively the volume and distribution of oil remaining after waterflooding that might be at least partially recoverable by some "tertiary" process. There is partially recoverable by some "tertiary" process. There is serious concern that the oil remaining after waterflooding is distributed in a manner that will make further, or tertiary, displacement economically unattractive. Just how is the oil distributed in a reservoir that is being or has been waterflooded? Obviously, there are intervals that have been flushed free of mobile oil and, thus, contain an immobile residual value. Core tests historically teach that these residuals may range from 15 to 50 percent of pore space. One can also expect to find intervals that have not been contacted by water for reasons relating either to the reservoir heterogeneity or the injection-production well pattern. The oil in these intervals is mobile depending if it can be contacted. Between these limits there are likely intervals that have been flushed only partially and still contain some mobile oil. Thus, at any point in a reservoir, a well drilled to evaluate the remaining oil saturation and its vertical distribution might find all three types of intervals. The technique used to define the oil saturation and its distribution within the reservoir after waterflooding must be definitive for each type of interval. The log-inject-log procedure, or modifications of this procedure, has proved procedure, or modifications of this procedure, has proved to be effective for this definition. This paper presents the results of 12 different applications of this procedure that provide evidence of relatively significant amounts of oil provide evidence of relatively significant amounts of oil remaining in waterflooded reservoirs. Also presented are limited comparisons with other measurement methods that, in general, support the results of the log-inject-log procedure. procedure. Log-Inject-Log Procedures As discussed in an earlier paper, a logging procedure that involves the use of two or more logging surveys run with the same tool before and after an induced change in formation conditions can provide measures of reservoir saturations that are more reliable than can be computed from the response of a single log. These procedures, referred to as log-inject-log or time-lapse logging, can be applied using both electric and pulsed neutron logs to measure formation water saturations and using neutron logs to measure formation gas saturations. Log-inject-log procedures have been developed to measure oil, gas, procedures have been developed to measure oil, gas, and water saturations at any time during the producing history of a reservoir. The log-inject-log procedure used to measure the residual oil saturations in the field examples that had been flooded by water injection involved logging the well with a pulsed neutron log after injecting fresh water and again after injecting salt water. The equation used to compute the residual oil saturation from these logs is .................(1) JPT P. 178