Data Science Applications in Oil and Gas Industry

MACSPro'2019 - Modeling and Analysis of Complex Systems and Processes, Vienna
21 - 23 March 2019

Dr. Fedor Krasnov

Conference website http://macspro.club/

Website https://exactpro.com/
Linkedin https://www.linkedin.com/company/exactpro-systems-llc
Instagram https://www.instagram.com/exactpro/
Twitter https://twitter.com/exactpro
Facebook https://www.facebook.com/exactpro/
Youtube Channel https://www.youtube.com/c/exactprosystems

Transcript

1. GazpromNeft Science and Technology Center Digital Platform Buildup Principles Fedor Krasnov and Mars Khasanov 21/03/2019
2. Digital platforms are not just a matter of “fashion” for companies but also a reflection of deep changes in the economy. 2 49 %  2% Non tech brands
3. GAZPROM NEFT STC DIGITALIZATION APPROACH
4. STC operates on services market 4 Global growth Global recession Shrinking resource base Alternative energy Degree of influence The demand for the services of STC Demand growth factors Factors of falling demand Rising oil prices Reduction of investment programs Degree of influence Falling oil prices Production decline
5. The balance of supply and demand for R&D 5 STC Academics HQ License renewal Suggestions to improve oil recovery Problems with an oilfield Subsidiaries Innovation institutes
6. ERA:GRAD Digital platform
7. Three principles of digital platform for STC • DATA AUTONOMY • ALGORITHMS STORE • «EVERYTHING IS A SPREADSHEET» 7
8. The principle of data autonomy 8 • Data sources are multi-variant • Data is quasi stationary • Calculations demand a data coherency Cross-functional calculations 30+ different data sources Backward transparency
9. ALGORITHMS STORE 9 Algorithms + Data Structures = Programs David Wolpert
10. «EVERYTHING IS A SPREADSHEET» 10 Spreadsheet Input filters Dataset Result Feedback DataFrame
11. ERA:GRAD embedded in Scientific Ecosystem 11 It does not matter who works as an ERA:GRAD agent: an employee, a student performing laboratory work, a developer of a new algorithm or an Infobot engaged in data quality control. Research organizations Software development companies Academic departments Universities Scientific Engineering in STC Research labs Innovation institutions Research foundations Service company Ministry of science Scientific conferences Professional community Oilfield service companies Oil companies Skolkovo, Rosatom Rusnano Russian Foundation for Basic Research, Russian Science Foundation SPE, SEG
12. Six hypotheses for discussion
13. THE ADDITIVE NATURE OF RESEARCH Using results (“outputs”) of one study as inputs for another study is a key issue to build continuous processes and reduce multiple repetitions of the same actions (to avoid the “every time from scratch” situation). 13
14. AUTOMATION OF ROUTINE INTELLIGENT OPERATIONS The routine operations identification process cannot be carried out “from the top downwards” because any regulation built on the basis of production of corporate process hypotheses is only a simplified model that replaces the routine operations automation tasks with the own relevance declaration tasks. The basis for the routine intellectual operations automation effort is the systemic understanding of the essence of research activities and the development of best practices in direct dialogue with researchers (users). 14 How we do it: Common mistake:
15. THE HYPOTHESIS OF THE ROLL OUT In the eyes of researchers, there is a competition between commercial software products and in-house systems. So it is impossible to neglect the laws of competition when planning a digital platform roll out. Substitution in the implementation should be understood as a two-way displacement of expensive and underutilized options and simple operations from commercial products into the company’s own software systems. 15 Common mistake: How we do it:
16. THE HYPOTHESIS OF MULTI-VARIANT DATA ON RESEARCH SUBJECT A digital platform should provide the ability to track the complete history of data from the moment of their acquisition and version through the use of algorithms, to the final result. 16
17. THE HYPOTHESIS OF ONTOLOGY Any new algorithm must find its place in the existing structure and become its integral part. 17
18. THE HYPOTHESIS OF TECHNOLOGIES From a strategic viewpoint the choice of a particular technology and “religious worship” to its advantages and disadvantages compared to other technologies is not so important. 18
19. Thank you.