НЕФТЯНАЯ ПРОВИНЦИЯ

Abstract

​

This paper is devoted to the applicability analysis of the neural network usage for automatic pay intervals detection. Machine learning methods allow the fastest way to process large data arrays, as well as to identify the necessary signs and relationships. The problem of this work is to find the optimal neural network, which will most accurately determine the pay intervals using well logs data. To obtain an accurate result, one of the most significant aspects is the preparation of data for the study. Preprocessing of data is a prerequisite for any method of machine learning. The results obtained were compared with the results of geoscientist`s interpretation. The selected algorithm allows automating the process of pay zones detection.

Key words:

​

Machine learning, neural network, pay intervals detection, sequence analysis, data preprocessing

References

 

1. François Chollet Glubokoe obuchenie na Python [Deep Learning with Python]. St. Petersburg: Piter Publ., 2018 (translated into Russian). 400 p.

2. Hamada G., Ahmed E., Chao N. Neirosetevoi raschet poristosti i vodonasyshchennosti dlya peschano-glinistyh kollektorov [Neural network prediction of porosity and water saturation for sand shale reservoirs] Dostizheniya v oblasti prikladnyh nauchnyh issledovanii, No.8, 2018. pp. 26-31.

3. Tsvetkovich M. Velik J., Malvik T. Primenenie nejronnyh setej dlya opredeleniya litologii kollektorskih gornyh porod i rascheta nasyshchennosti [Neural network for determination of reservoir rock lithology and prediction of saturation]. Geologiya Horvatii, No. 62, 2009. pp. 115-121.

4. M. Mardi, H. Nurozi, S. Edalatkhah Raschet vodonasyshchennosti s primeneniem nejronnyh setej i issledovanie vliyaniya izmeneniya koefficienta cementacii i pokazatelya nasyshchennosti na Iranskoj neftyanoj skvazhine [A water saturation prediction using artificial neural networks and an investigation on cementation factors and saturation exponent variations in an Iranian oil well]. Neftegazovaya nauka i tekhnologii, No. 30, 2012. pp. 42-434 (translated from English)

5. H. Chicheng, S. Misra, P. Srinivasan, M. Shuhiang Kogda petrofizika stalkivaetsya s ogromnym massivom dannyh: Chto mozhet mashinnoe obuchenie? [When petrophysics meets big data: what can machine do?]. SPE Middle East Oil and Gas Conference, 2019. 24 p.

6. Zhu L., Li H., Yang Zh., Li Ch., Ao Yi. Uluchshennaya litologicheskaya interpretaciya karotazhnyh dannyh s primeneniem svertochnykh neironnykh setei [Intelligent logging lithological interpretation with convolution neural networks]. Petrophysics, Vol. 59, Iss. 06, December 2018. pp. 799-810.

​

Authors

​

I.S. Kanaev, LLC «Tyumen Petroleum Research Center»

79/1, Osipenko st., Tyumen, 625002, Russian Federation

E-mail: iskanaev@tnnc.rosneft.ru

​

For citation:

​

I.S. Kanaev Nejrosetevoe detektirovanie produktivnyh intervalov na primere ob#ekta BV10 Samotlorskogo neftegazokondensatnogo mestorozhdenija [Pay intervals detection by neural network on the example of the BV10 reservoir of the Samotlor oil field]. Neftyanaya Provintsiya, No. 4(20), 2019. pp. 157-171. https://doi.org/10.25689/NP.2019.4.157-171 (in Russian)