The oil and gas industry recently has transformed itself into a data-intensive industry with the addition of artificial intelligence, intelligent or cognitive computing, machine learning, the Internet of things, and robotics. With such growth and expansion, various exponential technologies can carry the energy industry’s transformation forward in the new era of cyberautomation.
Exponential technology and its adoption is based on my growth criterion, defined as
Exponential Growth =
(Exponential Technology)(Exponential Thinking)
Exponential technologies in other industries drive the industry after the bifurcation point, depending on how the knowledge is being assimilated. Exponential thinking is called the “exponential surprise factor.” These underpinnings are observed on the tubular-mechanics side also through data analytics, machine learning, artificial intelligence, and cognitive processes.
Even though we have made some inroads, we have to deal with a complex high-aspect-ratio system of length to diameter, connected from surface to bottom by way of drillstring, casing string, and tubing string that is also highly dynamic in nature. We are making advancements by changing the tubulars from dumb iron to intelligent tubulars. The complexity of tubular modeling is very specific to our industry; we deal with many uncertainties and coupled variables and not observed effects. Because of the inherently exquisite difficulties of turning modeling to action, we need a strong coupling between the data and physics-based engineering models. As we move to automation, we need a strong coupling with the tubular engineering and manufacturing through data analytics in the form of adaptive engineering summarization to control the bit from surface. This is a combination of white-box models, which are perfectly specified by previous knowledge and physical insight through laboratory tests, reliability studies, and black-box models, which are models in the absence of a priori physical knowledge or insight that will result in the downhole state. In this way, paradoxes of the tubular behavior are neutralized, speculations are suppressed, conundrums are counteracted, and engineering principles are honored so that we have empirically verifiable and practically dependable models.
I have selected a few papers that are structured toward this hybrid approach that includes data mining, design improvement, and reliability.
This Month's Technical Papers
Electromagnetic-Based Tool Allows In-Situ Inspection of Multiple Metallic Tubulars
New Analytical Model Enhances Understanding of Connection Strengths
Data Mining Effective for Casing-Failure Prediction and Prevention
Recommended Additional Reading
SPE 195203 Numerical Study on Casing Integrity During Hydraulic Fracturing Shale Formation by Xiaye Wu, University of Oklahoma, et al.
SPE 194155 Frictional Heating of Casing Due to Drillstring Rotation—Finite Element and CFD Simulations by Wissam Assaad, Shell, et al.
| Robello Samuel, SPE, is a technology fellow at Halliburton based in Houston. He has more than 34 years of multidisciplinary experience in domestic and international oil and gas drilling and, for the past 14 years, has held concurrent adjunct-professor appointments at the University of Houston and the University of Southern California. Samuel has published 13 books and more than 180 technical papers and holds 66 patents. In 2013, he received the SPE Gulf Coast Section Drilling Engineering Award, and, the following year, he was named an SPE Distinguished Lecturer. Samuel holds BS and MS degrees in mechanical engineering and MS and PhD degrees in petroleum engineering from The University of Tulsa and is a member of the JPT Editorial Committee. He can be reached at robello.samuel@halliburton.com. |
