A Novel Well Log Data Quality Prescriptive Framework for Enhancing Well Log Data Quality Interpretation

Abstract

AbstractReservoir formation evaluation in real time has become essential with the rise of logging while drilling solutions, and the strive to achieve greater precision measurements, while reducing the time from drilling to production. Well log data quality represents a crucial part of ensuring that the interpretations are precise, and avoiding erroneous data that may lead to misinterpretations. Prescriptive artificial intelligence focuses on developing trained predictive models to recommend actions, while optimizing looked after outcomes.We present an innovative prescriptive artificial intelligence framework for determining well log data quality to assist in interpretation and decision making. The framework incorporates a decision tree approach to ensure that the various collected well logs are aligned with each other, and then determines which well logs are of sufficient quality, and which have to be adapted in the interpretation framework. The well logs are then adapted in order to ensure that they provide consistent interpretations.The framework was examined on a synthetic reservoir with various well logs. The framework evalues the quality of the well logs, and filters them in order to ensure proper correlation with each other. A sensitivity analysis and adaptation of the well logs ensures proper interpretation.This work represents an innovative approach to autonomous well log data quality determination, and provides a new approach to assist well log interpretation.