Journal of the Korean Society of Mineral and Energy Resources Engineers ISSN:2288-0291(Print) 2288-2790(Online) 한국자원공학회지

This paper describes a new approach to automated interwell log correlation using artificial intelligence and principal component analysis. The approach to correlate wireline logging data is on the basis of a large set of subjective rules that are intended to represent human logical processes. The data processed are mainly the qualitative informations such as the characteristics of the shapes extracted along log traces. The apparent geologic zones are identified by pattern recognition for the specific characteristics of log trace collected as a set of objects by object oriented programming. The correlation of zones between wells is made by rule-based inference program. The reliable correlation can be established from the first principal component logs derived from both the important informations around wellbore and the largest common part of variances of all available well log data. Correlation with field log data shows that this approach can make interwell log correlation more reliable and accurate.