Title
The Game Changing Impact of Data and Data-Driven Solutions in the Upstream Oil and Gas Industry
Outline
- The Data Revolution
- Journey of Data in the Our Industry
- Confusion:
- Digital Energy & Petroleum Data Driven Analytics
- Data Driven Solutions in the Upstream
- Petroleum Data Science
- Some Final Thoughts
- Roles, Players, Objectives, Academia, Startups
- The Way Forward (Road to Adoption)
Description:
The buzz regarding data and what it can do for business today is everywhere. It has captured imaginations. Some imaginations have turned into reality, giving rise to real-time translation tools and self-driving cars. More exciting products will hit the market soon. The excitement that has been stirred by this new trend has made managers and engineers in the oil and gas industry excited.
Autonomous, self-drilling rigs, completely automated, smart completion and hydraulic fracturing, building comprehensive, full field, and fast reservoir models based on facts (field measurements), building smart proxy of complex, multi-million cell, numerical reservoir simulation models that run in fractions of a second without compromising physics or space-time resolution, and coupling subsurface to wellbore and to the surface facilities, in real-time modeling and optimization, are among the rising applications of data driven solutions in the upstream oil and gas industry.
Data-Driven Analytics have already made important contributions to the oil and gas industry. In situations where our understanding of the physics is still in the developmental phase and the number of unknowns are overwhelming (such as production from shale), data-driven analytics have proven to be valuable in understanding the complex nature of the production process, and to help us optimize completion design and production. Some operators are taking advantage of existing data-driven know-how in the industry while others are contemplating the possibilities, and yet others have gone stray with disappointing outcomes.
The journey of data analytics in our industry has not been without pain and confusion. However, it seems that the long battle against traditionalists that view the pathway of starting with first principle physics in explaining nature more as a religion than a scientific endeavor has been largely victorious. But some still find it hard to move into the new millennium. On the other hand, amongst the enthusiasts of this new technology, the overlap between IT with Engineering and Geo-sciences has caused much confusion. Attempts to clarify such confusion have resulted in two separate SPE Technical Sections each dedicated to certain aspects of including data in our everyday operations. However, latest activities in these Technical Sections point to the fact that some confusion still remains.
Large numbers of start-ups, many from areas with minimal historical relevance to the oil industry, are examining their luck and other’s investments in the upstream. This has resulted in large number of claims being made with slick marketing techniques. What most of these start-ups seem to lack is actual domain expertise which has proven to be most crucial in successful implementation of data-driven solutions in our industry.
Now operators are asking what their strategy should be, to take maximum advantage of this technology. Should they develop in-house expertise or should they outsource all or part of their data-driven analytics work? Should they develop or buy the required tools? Do they need statistical and visualization tools or comprehensive workflows and solutions? Should they hire statisticians and teach them petroleum engineering, or train their existing petroleum professionals the art and science of machine learning and pattern recognition? Do we need data scientists in our workforce? Do we need “Petroleum Data Scientists,” or will anyone with a background in statistics do? What is the definition of a “Petroleum Data Scientist”? The perspective on these issues that this presentation provides has been formed through more than two decades of direct experience in petroleum data-driven analytics research and development.
------------------------------
Shahab Mohaghegh
Professor
West Virginia University
MorgantownWV
------------------------------