Multivariate Statistical Analysis for Automatic Electrofacies Determination from Well Log Measurements

Abstract

Abstract A quantitative and systematic methodology is presented for the prediction of the lithology using electrofacies determination from well log data. Multivariate statistical techniques are adopted to segment wireline log measurements and group the segments into electrofacies types. To consider corresponding contribution of each log and reduce the computational dimension, multivariate logs are transformed into a single variable through principal components analysis. Resultant principal components logs are segmented using the statistical zonation method to enhance the quality and efficiency of the interpreted results. Hierarchical cluster analysis is then used to group the segments into electrofacies. Optimal number of groups is determined on the basis of the ratio of within-group variance to total variance and core data. This technique is applied to the wells in the Korea Continental Shelf. The results of field application demonstrate float the prediction of lithology based on the electrofacies classification works well with high reliability to the core and cutting data. This methodology for electrofacies determination can be used to define reservoir characterization which is helpful to the reservoir management. Introduction As the variation of petrophysical properties often corresponds to lithologic variation, lithology determination plays an important role for reservoir characterization. Subsurface lithology is traditionally determined from both core and cutting analyses. Cores are generally not continuous and do not provide complete descriptions of formations crossed by a well. Cuttings always have some uncertainties in depth and it can be difficult to restore the components and thickness of lithologic column. As a result, the lithology based on those data is not sufficiently accurate and precise for the quantitative use. On the other hand, well logs have the advantage of providing a continuous record over the entire well and can be obtained in conditions where coring is impossible. Therefore the integration of core and well log data can give a good lithologic description of the formations. An electrofacies is defined as "the set of log responses which characterizes a bed and permits it to be distinguished from the others." The electrofacies derived by selecting, weighting, and combing well log data can be used as an indicator of lithology. Once good correlations between electrofacies and core analysis are established on a local basis, significant geological information can be extracted from well logs alone. In this study, a systematic technique has been developed for the electrofacies determination from well logs using multivariate statistical analysis and applied to the wells in the Korea Continental Shelf. Electrofacies Determination Wolff et al and Moline et al developed the multivariate statistical procedures to determine the electrofacies. In this study, to enhance the quality and efficiency of the interpreted results, the zonation of principal components logs is constructed prior to clustering. The segmented principal logs instead of the original logs are used for cluster analysis. The electrofacies determination procedure used in this study is summarized by the flow chart in Fig. 1. Descriptive Statistical Analysis Data. Data can be revealed their features when they are properly organized. The univariate tools can be used to describe the distribution of individual variable. Histogram is known as one of the most common and useful presentation and organization of data for univariate description. The importance features of most histogram can be captured by the summary statistics: measures of location, measures of spread, and measures of shape. When we analyze a multivariate data set like well logs, the relationships and dependencies among variables are important feature of data. P. 109^