SPE Young Professionals meeting, November 30, 2017

Thursday, 30 November 2017 Read 952 times
Rate this item
(0 votes)
  • Start Time: Thursday, 30 November 2017, 19:00
  • Event Type: Conference

We invite you to SPE young professionals meeting on "Inversion and Machine Learning in Well Logging data Interpretation".

SPE Young Professionals meeting, November 30, 2017

The presentation will be delivered in English

 

Meeting will take place on November 30,Thursday, at 1900 hours at Pokrovskiy bulvar 3 bld. 1, Moscow.

Meeting is open for everyonenot only young professionals!

Preliminary registration is required!

Please reserve a place and send your full name (including middle name), company and job title to This email address is being protected from spambots. You need JavaScript enabled to view it. not later than 1200 on November 29. You will have a confirmation after you’re registered.

Looking forward to seeing you!

 

More details about the topic:

The standard methods of well logging data interpretation include the steps of qualitative and quantitative analysis of the complex of the well logging tools measurements with the goal of evaluation of geological and petrophysical properties of the analyzed part of target reservoir. The qualitative analysis step includes, as a rule, the answers to conceptual questions of lithological composition, potential reservoirs intervals determination, determination of the interval of distorted and non-informative data. The common quantitative methods include the mathematical procedures of the poro-perm properties evaluation by combination of one or more well logging measurements, for example, in a system of linear equations and the following its solution by the means of the least square approach. The interpretation of the well logging data with the first qualitative, and following with the second quantitative steps may lead to uncertain results due to the following factors.

Firstly, the discrepancies in depth of investigation and vertical resolution of different well logging methods leads to inaccuracy in petrophysical properties determination by combining the methods in a system of lineal equation. This problem is partially solved in industry solved with forward physical modelling of each of the well logging tools responses and joint inversion of the well logging data by the means of one or a complex of the optimization methods in order to solve an ill-posed inverse problem for the physical properties of formations reconstruction. The main issues encountered here are the ill condition of an inverse problem and a big number of equivalent solutions, as well as considerable hardware and time resources required for the inversion procedure.

Second, it is the poro-perm properties estimation, such as porosity and water saturation of porous media, from the well logging tools responses, which requires some relation “Well Log – Petrophysical Parameter” for each of the well logging method. Nowadays the most trustful basis for such a relation are the laboratory measurements of core, however, more investigation results in numerical porous space modelling and digital core experiments become available in science and commercial communities.

The last but not least, is the qualitative interpretation of the results of the first inversion and the second petrophysical steps in order to construct a conceptual geological picture and represents, nowadays, the result of the subjective judgment of an interpreter based on cumulative database and integration all available of geological and geophysical data. Particularly, in well logging the electrofacies analysis is used for the determination of relationships between physical and petrophysical parameters and sedimentary environment (mostly in vertical wells) and reservoir geometry reconstruction (mostly for horizontal wells).

The main approaches and examples for the forward modelling of the well logging tools and inversion of the well logging data, available for practical usage, are considered, machine learning and pattern recognition algorithms for electrofacial analysis and for automatical well logging data interpretation in vertical and horizontal well are observed and discussed.

 

About author:

Krutko Vladislav Vadimovich, Baker Hughes

 

Vladislav started his career as a geoscientist in the European branch of the Baker Hughes B.V. Geoscience team in the Netherlands.

For 3 years he has been responsible for the integrated interpretation of the standard and special well logging methods for the European, UK markets and others around the world. His responsibility covered the planning, real-time support and final report on the results of the well-logging procedures.

Also, his responsibilities included the procuring the common activity between Geoscince interpretation and Research teams in the R&D centers of the company.

At the moment Vladislav is part of the Russian Geoscience team of the Baker Hughes and is responsible for the development of the methodology of the well logging data interpretation in high angle and horizontal wells.

Vladislav has obtained bachelor degree in Physics (Physics of Semiconductors) in Novosibirsk State University, Master of Science in geophysics in Tomsk Polytechnic University and Master of Science in Geology in Heriot -Watt University (Edinburgh, UK).

The main professional interests include the ill-posed problems in geophysics, numerical modelling of physical processes, machine learning.

Leave a comment

Make sure you enter the (*) required information where indicated. HTML code is not allowed.