A computed profile from the Wolfcamp B using Factor Drive. The dotted red line is the wellbore trajectory. The blue zone is the most probable geologic structure, while the red-orange field indicates uncertainty about that answer. The good agreement in the type log correlations at left and predicted LWD along the top confirm the plausibility of this structure.
Geosteering innovations continue to have dramatic breakthroughs in accuracy, efficiency, and real-time monitoring capabilities, thanks to pioneering work in analytics with new algorithms and approaches to machine learning. Welcome to an interview with Hugh Winkler, who will also be presenting at AAPG's U-Pitch New Technology Showcase, at IMAGE.
What is your name and your company?
Hugh Winkler, of Factor Technology.
What is the name of your new product, service, or process?
Factor Drive Computed Geosteering.
Why was it developed? What main problem is it designed to solve?
Geosteering, the science of landing a horizontal well in the target zone and keeping it there, wasn't being done scientifically. Geologists interpreted while-drilling data using art, intuition, and heuristics. They had desktop software to assist them drawing trial geologic structures, and evaluating the plausibility of those trial structures. But they had no way to explore the whole space of possible structures and simply compute the best one. They filled in the gaps using their geologic knowledge and intuition, and they did a good job. But they could be disastrously wrong, and even when working well, the trial-and-error approach doesn't scale. It's exhausting!
I published a Bayesian network solution to the geosteering problem in Geophysics; then I co-founded Factor, and we began commercializing that technology, because it changes the whole well positioning field. You pour your data in, you tell it some constraints, you punch a button, and out comes your answer, along with some quantitative guidance on the uncertainty.
What would you like to accomplish by participating in U-Pitch?
We'd like to develop awareness with technology scouts for oil companies. We want to expand our base of early adopters. And we'd like to meet investors excited about using the potential of technologies like ours to drill and complete wells cheaper, and make them produce more.
What are your plans for the future?
In five years, everyone will compute geosteering interpretations. We're preparing to be the leader in that market, and to expand into adjacent spaces where our Bayesian network technology gives us a unique advantage.
Please name a person, book, or program that has influenced you in a positive way.
The University of Texas Jackson School and the U.T. Institute for Geophysics tolerated me as a grad student, beyond all reason, and the things I'm working on now are the direct application of inverse problem work I studied there long ago. My thesis supervisor, Paul Stoffa, hooked me on computational seismology, and showed by example how to think originally and back it up with theory and code.