Machine Learning Model for Drilling Equipment Recommender System for Improved Decision Making and Optimum Performance

Abstract

The oil industry, in its constant strive to maximize gains out of operational data is constantly exploring new horizons where to combine the latest advances in data science and digitalization, into the areas where key decisions to drive economical and operational decisions reside with an aim at optimizing the capital expenditure through sound decision making. High volume operational data has been recognized as hiding many opportunities where the captured details these repositories that include real time logs and bit run summaries, provide a clear opportunity where to extract insights to support optimized decisions in terms of equipment selection to achieve the desired operational objectives. Current possibilities within data science have opened the possibilities through viable solutions, which in this case, aims at providing advise on which equipment in terms of BHA and Bits to select, that would yield the desired outcome for a drilling run. The whole exercise being based on evidence gathered from previous runs where the details for the equipment, the relevant well characteristics, and the observed rates of penetration and the used parameters, are taken into consideration to provide the optimum combination to be implemented in new runs. The present study describes the methodology in terms of data utilization, data science method development and solution deployment, with the associated issues that had to be addressed in order to provide a viable solution in terms of data utilization, technical validity and final user utilization, as well as a series of recommendations to be addressed within any such endeavors to assure the value addition.