Release:
2019, Vol. 5. №4 (20)About the authors:
Anatoliy A. Kislitsin, Dr. Sci. (Phys.-Math.), Professor, Department of Applied and Technical Physics, University of Tyumen; a.a.kislicyn@utmn.ru; ORCID: 0000-0003-3863-0510Abstract:
This article presents the problem of determining the conditions of intensive flooding of horizontal wells for complicated geological structure strata, such as the Pokur formation at the Vostochno-Messoyakhskoye oil field. It corresponds to alluvial continental planes or to coastal-sea conditions of sedimentary rocks accumulation. The principal peculiarity of the geological structure of these strata is the high lateral heterogeneity, which is connected with riverbed migration by sedimentary rocks accumulation. Using the neural network method, the authors have developed an algorithm that allows explaining the different dynamics of displacement characteristics for the wells with identical geological and technological indicators. Having analyzed the dynamics and causes of water cut of 125 wells at East-Messoyahskoe oil field, the authors show that the geo-statistical methods do not apply to the task of describing continental accumulation objects with compound construction. However, the results of seismic data interpretation provide the basic volume of information about the inter-well space. The authors have developed an algorithm for complete regression analysis for the adaptation of the hydrodynamic model, which includes the method for constructing a cube of sandiness based on neural network modeling. It follows the basic factors, those exert influence on dynamics of water cut. They include distance at well’s tube to water-oil contact, and presence of impenetrable or semi penetrable interlayer between tube and water-oil contact. The neural network algorithm (Genetic Inversion) allowed performing the test calculations on one of group wells most operated. The suggested approaches in the construction of the reservoir distribution in the inter-well space allow achieving better integral convergence of the dynamics of water cut at the first iterations of the full-scale hydrodynamic model.Keywords:
References:
Barsky A. B. 2004. Neural Networks: Recognition, Management, Decision Making. Moscow: Finance and Statistics. [In Russian]
Borovikova V. P. 2008. Neural Networks STATISTICA Neural Networks: Methodology and technology of present-day analyses of data. Moscow: Goryachaya liniya — Telecom. [In Russian]
Demyanov V. V., Savelyeva E. A. 2010. Geo-Statistics: Theory and Practice. Moscow: Nauka. [In Russian]
Dinariev O. Yu., Svitelman V. S. 2013. “Geostatistical analysis of rocks microtomograms: some new approaches and results”. Izvestiya Vuzov. Oil and Gas, no 2, pp. 16-21. [In Russian]
Zhemchugova V. A. 2012. Reservoir Sedimentology. Lecture Course. Moscow: EAGE Geomodel. (In Russian)
Pisarev A. D. 2019. “Energy efficient biomorphic pulse information coding in electronic neurons for the entrance unit of the neuroprocessor”. Tyumen State University Herald. Physical and Mathematical Modeling. Oil, Gas, Energy, vol. 5, no 3, pp. 186-212. DOI: 10.21684/2411-7978-2019-5-3-186-212 [In Russian]
Podnebesnykh A. V., Malyshevskaya K. A., Malyshevskaya T. S., Ovchinnikov V. P. 2014. “Integrated approach to investigation of gas cap in the formation PK1-3”. Izvestiya Vuzov. Oil and Gas, no 6, pp. 13-18. [In Russian]
Priezhev I. I. 2009. “The method of seismic inversion using a genetic algorithm with the subsequent use of the inversion results in modeling reservoir properties of the reservoir”. Seismic technology, no 2, pp. 18-23. [In Russian]
Khusainov A. T. 2013. “Innovative analytical method for operation estimation of oil fluids”. Izvestiya vuzov. Oil and Gas, no 3, pp. 61-63. [In Russian]
Chan K. S. 1995. “Water control diagnostic plots”. SPE Annual Technical Conference and Exhibition (22-25 October, Dallas, Texas). Article No SPE-30775-M DOI: 10.2118/30775-MS
