The Use of Multiple Linear Regression Analysis to Calculate Formation Water Saturation from Logs

$10.00

Jay G. Patchett, Amoco Production Company

6th Formation Evaluation Symposium of the
Canadian Well Logging Society in Calgary, October 24, 25 & 26, 1977

1977

Abstract

SPWLA EIGHTEENTH ANNUAL LOGGING SYMPOSIUM, JUNE 5-8, 1977 THE USE OF MULTIPLE LINEAR REGRESSION ANALYSIS TO THE USE OF MULTIPLE LINEAR REGRESSION ANALYSIS TO CALCULATE FORMATION WATER SATURATION FROM LOGS CALCULATE FORMATION WATER SATURATION FROM LOGS by by Jay G. Patchett Jay G. Patchett Amoco Production Company Amoco Production Company Tulsa, Oklahoma Tulsa, Oklahoma SUMMARY SUMMARY Casual observers of well log interpretation are often impressed by the Casual observers of well log interpretation are often impressed by the apparent simplicity of the log interpretation process. The basic equa- apparent simplicity of the log interpretation process. The basic equa tions used appear to be uncomplicated. When the water saturation equation tions used appear to be uncomplicated. When the water saturation equation is expanded, however, the number of logging measurements, parameters and is expanded, however, the number of logging measurements, parameters and environmental constants necessary to calculate an accurate water satura- environmental constants necessary to calculate an accurate water satura tion can exceed a dozen. Few if any of these are known explicitly. They tion can exceed a dozen. Few if any of these are known explicitly. They are subject to measurement errors and misestimates which are not necessar- are subject to measurement errors and misestimates which are not necessar ily symmetrical about their expected value. However, there is usually a ily symmetrical about their expected value. However, there is usually a great deal of pertinent information relating to these necessary data great deal of pertinent information relating to these necessary data within the logs themselves. Plotting of observed log data has evolved as within the logs themselves. Plotting of observed log data has evolved as one of the most valuable interpretation tools available to the log analyst.