An AI-Based Workbench for Knowledge Capture and Integration in Sub-Surface Characterization

Abstract

Nowadays the oil and gas industry is facing a data flooding problem. Each step in the E&P value chain creates terabytes of information, but it is still a challenge to extract meaningful knowledge and insights to support decision-making processes. In this work, we present a workflow based on artificial intelligence (AI) tools that help to combine different sources of knowledge, therefore enabling better analytics tools to ground decisions. This workflow integrates machine learning models, like computer vision and data-driven learning processes, with formal knowledge, like ontologies, rules, and facts structured in a knowledge graph. We also capture former expert’s interactions thus enabling an active/continuous-learning environment.

To illustrate our approach, we describe a real-case scenario that starts from seismic interpretation, going through the process of reservoir characterization and simulation, up to the assessment of different scenarios of oil-production. We capture all critical decisions in this workflow and use that as a documentation of the analysis, enabling the decision maker to investigate the rationale behind the final results. This contextualized historical data can also be used as supporting evidence in future analysis.

AAPG Datapages/Search and Discovery Article #90323 ©2018 AAPG Annual Convention and Exhibition, Salt Lake City, Utah, May 20-23, 2018