Wireline log data provide a multi-dimensional picture of the formations encountered by the borehole. Although each log represents a distinct dimension in the view of the formation it does not necessarily follow that each log contributes original information to the formation evaluation. The technique of principal components analysis is used to establish the amount of original information contained in a given data set. Principal component analysis examines the total variability represented in the data and describes this variability in terms of a set of mutually perpendicular axes. Each axis will account for a proportion of the original variability and will be uncorrelated with all other axes found. Generally, the bulk of the variability is described by fewer principal axes than there are logs in the data set. The principal axes represent the amount of cross-correlation between logs. The use of sophisticated statistical log interpretation procedures can be of great benefit to the log analyst but may also provide ambiguous answers.

You can access this article if you purchase or spend a download.