Discriminant Analysis as a Technique for Exploitation of Petroleum Reservoirs

Abstract

Abstract Discriminant analysis, a statistical technique, was used to classify oil wells into productivity groups based upon data obtained productivity groups based upon data obtained from geophysical well logs. A calibration run conducted in the northern area of the East Canton oilfield verified that valid forecasts can be obtained by the procedure. Of 20 wells studied, four appeared to have a greater potential than that indicated by actual potential than that indicated by actual production figures. A similar study of the production figures. A similar study of the southern area of the East Canton reservoir also proved discriminant analysis to be effective for predicting productivity, although different variables were significant than those for the north reservoir. Discriminant analysis was indicated to be adequate and potentially helpful in arriving at decisions potentially helpful in arriving at decisions concerning well completion. Introduction Located in northeastern Ohio, the East Canton oilfield is one of the largest ever discovered in the Appalachian basin. The field now comprises over 60,000 acres with more than 1,000 oil wells. Fig. 1 shows the field and illustrates a proposed additional drilling area—one of more than 500 similar locations for new wells in the area. Upwards of 12,000,000 barrels of premium Pennsylvania Grade crude oil have been obtained from the field; current production is about 11,000 barrels per day. Exploiting the East Canton reservoir presents many problems to the petroleum presents many problems to the petroleum engineer, not the least of which is deciding to complete a well. Log data and field experience in offsetting wells can be helpful but evaluating oil wells for porosity and hydrocarbon saturation on the basis of log analysis is especially difficult. Few of the wells have been plugged as "noncommercial" immediately following drilling, but about 20 percent of them could have been plugged percent of them could have been plugged because they will not return drilling and completion costs. Discriminant analysis, which has been shown to be useful in oil and gas exploration and other areas of endeavor, was investigated by the Bureau of Mines as a method of predicting oil well production from the East predicting oil well production from the East Canton field. Current oil well production data and log data serve as input to the discriminant function, and the model is then used to classify new wells into high, medium, and low production as a function of log-determined reservoir properties. Rate of return from a new well can then be estimated, and a decision can be made concerning completion of the well. A unique feature of the concept is that the statistical prediction technique adapts itself to the reservoir.