Status and Prospect of Artificial Intelligence Application in Fracturing Technology

With the rapid development of artificial intelligence (AI) theory and computer technology, intelligence and digitalization have become important forces to promote the fracturing technology of reservoirs. In terms of the intelligent development of the fracturing technology, the research progress and application of AI technology in the prediction of geological parameters, optimization design of fracturing parameters, real-time diagnosis and control of fracturing construction, and development of fracturing tools and materials were introduced. The main problems existing in the development of the intelligent fracturing technology as well as the key development direction were analyzed. It suggests that the intelligent fracturing technology was still in the stage of exploration and trial. Foreign countries have taken the lead in the intelligent identification of sweet spots, optimization of fracturing parameters, and intelligent control of field construction, and they have obtained successful applications in fracturing services in several fields in North America. China has only carried out early exploration in fracturing big data machine learning, intelligent fracturing materials, and there was no significant progress in intelligent fracturing equipment and tools, real-time monitoring and diagnosis, and intelligent field control. Therefore, there is a gap between China and other countries. The key problems affecting the development of intelligent fracturing technology were proposed, including the poor reliability of data samples, the lack of integrated intelligent fracturing methods and equipment, and the shortage of interdisciplinary talents. It was also predicted that with the development of the Internet of Everything technology, an intelligent completion fracturing system would be developed, which could complete reservoir evaluation, the sweet spot identification, optimization design of fracturing parameters, field control, post-fracture, etc. without human intervention. At that point, the integrated intelligent reservoir stimulation could be truly realized.