Optimization of the “mature” fields development in machine learning algorithms is one of the urgent problems nowadays. The task is set to extend the effective operation of wells, optimize production management at the late stage of field development. Based on the task set, the article provides an overview of possible solutions in waterflooding management problems. Production management technology is considered as an alternative to intensification of operation, which is associated with an increase in the produciton rate and involves finding solutions aimed at reducing the water cut of well production. The practical implementation of the “Neural technologies for production improvement” includes the following steps: evaluation, selection, predictive analytics. The result is a digital technological regime of wells that corresponds to the set goal and the solution of the optimization problem in artificial intelligence algorithms using the software and hardware complex “Atlas – Waterflood Management”.
“Neural technologies for production improvement” have been successfully tested at the pilot project site of the productive formation of the Vatyeganskoe field. The article provides a thorough and detailed analysis of the work performed, describes the algorithms and calculation results of the proxy model using the example of the pilot area, as well as the integration of the “Atlas – Waterflood Management” and the organization of the workflow with the field professionals of the Territorial Production Enterprise Povkhneftegaz.

The paper studies the wetting effects on the characteristics of two-phase fluid flows in porous media. The originality of paper is a study of displacement under dynamic conditions when the action of viscous forces is significant. As a research tool, the methods of mathematical modeling are used – the lattice Boltzmann equations in a combination with a color-gradient model that describes interphacial interactions. Numerical experiments are carried out in a digital model of a porous medium characterized by a high degree of the pore space heterogeneity. In this work, a map of flow regimes in the coordinates “capillarity number – contact angle” is performed. The identification of four crossover modes between flows with capillary, viscous fingers and with a stable displacement front is carried out. Special attention is paid to the study of the influence of wetting effects on the specific length of the “injected fluid – skeleton” interface.

To analyze the waterflooding system of an oil reservoir and predict the effectiveness of geological and technical measures, information is required on the distribution of injection rate between the reacting production wells and the reservoir boundary. The most reliable methods for calculating these characteristics are methods based on hydrodynamic modeling of flow. Modern commercial software implement algorithms for these purposes based on the construction and analysis of streamlines. At the same time, there are no reliable estimates of the accuracy of these algorithms and recommendations for choosing the optimal parameters in the available literature.
In this paper, we propose an algorithm for calculating the proportions of the distribution of the total well flow rate between the surrounding wells and the reservoir boundary using streamlines. Streamlines are constructed on the basis of a finite element solution to the flow problem averaged over the formation thickness and determine the boundaries of the streamtubes connecting the corresponding wells. The flow rate through the flow tubes is calculated by numerically integrating the Darcy velocity field of the indicated two-dimensional problem. The algorithm was tested on idealized examples of waterflooding elements of typical well placement schemes, when the exact distribution of the proportions of fluid injected into the formation is known, and on the example of comparison with the solution of the problem of simulating the injection of a tracer into the reservoir. Recommendations for the selection of starting points for tracing streamlines are presented, which allow achieving a minimum level of error in determining the mutual influence of wells in a wide range of the computational grid resolution of the flow model.
A more general application of the described method without significant changes is to equip the high resolution flow model along fixed stream tubes with their rate characteristics.

The salt bearing thrust belts are among the most prolific petroleum provinces in the world, including the fold belts flanking the Persian Gulf Basin, the Tarim Basin, the Terek-Sunzha Zone of the Eastern PreCaucasus etc. They are characterized by peculiar structural styles of deformation, large oil and gas pool sizes, significant flow rates. The Chernyshev Swell is among prospective salt bearing thrust belts. It is located in the Cis-Uralian part of the Timan Pechora Basin. Several exploration wells have been drilled in the area so far failed to deliver the anticipated oil and gas deposits. Post mortem study indicated that the main reason for the past failures was the mismatch between the complexities of the geological settings and the used exploration technologies. Integrated study of the area, conducted by the North Uralian Petroleum Company, using modern technologies, including 3D seismic surveying and magnetotelluric sounding, has enabled a far better understanding of this area’s structure and petroleum habitat. It is shown that high petroleum potential is attributed to the transition zone between the Chernyshev Swell and the Kosyu Rogov Foredeep Basin. This study indicated that the area of the Povarnitsa High is of priority interest for explration. This area provides supplementary opportunities in addition to the commercialization of the petroleum exploration potential. It includes production of rare earth and metals in brine of the evaporate deposits. The subsalt reservoirs of the Chernyshev Swell could be used for the storage of gas and CO₂ sequestration. Integrated development of the resource potential of the Povarnitsa High may become a key driver of the economic development of the north-eastern part of the Republic of Komi.

Quartz samples taken from the ore veins of the Darasun gold deposit (Eastern Transbaikalia, Russia) were studied using electron paramagnetic resonance (EPR) and laser ablation (LA) methods. The purpose of the research was to clarify the behavior of the Al impurity during quartz crystallization and its subsequent recrystallization. The results of the research were used to determine the genetic informativeness of the Al impurities concentration in ore quartz.
A separate study of the regularities of the distribution of Al impurity in the regions of crystalline structure and in the zones of crystal lattice distortions was carried out. In the regions of crystalline structure, the Al impurity concentration N_AL was determined by the concentration of paramagnetic Al–O– -centers associated with the presence of substitutional Al³⁺ ions. The content of the Al impurity in the zones of crystal lattice distortions was judged by the difference between the gross concentrations of aluminum C_Al and the values of N_AL. It was taken into account that the intensity and direction of the studied processes can be influenced by the temperature of quartz formation and the degree of its recrystallization. The temperature of quartz formation was estimated by the values of Ti impurity concentration, and the degree of recrystallization was estimated by the content of Li impurity C_Li in the mineral.
It was found that the amount of Al impurity localized in the zones of crystal lattice distortions is an order of magnitude higher than that present in the zones of crystalline structure. It is shown that this phenomenon is explained by the high ability of the zones of crystal lattice distortions to capture of Al impurity during quartz crystallization. The amount of Al impurity trapped by quartz increases sharply with an increase in the temperature of the mineral formation. In areas of regions of crystalline structure, a different picture is observed − impurity capture during crystallization proceeds with low intensity and is not so critically dependent on temperature.
It was found that the Al impurity in the considered zones behaves differently during quartz recrystallization. If Al impurity capture continues in the regions of crystalline structure, then Al impurity removal from quartz can occur in the zones of crystal lattice distortions at high C_Al values.
Based on the data obtained, the areas of use of aluminum impurity contents as a typomorphic feature of ore quartz were determined. The Al impurity concentrations proved to be suitable for use in cases of low ore formation temperatures, and its gross C_Al contents have the prospect of wider use. It is noted that a characteristic feature for quartz from high productivity zones is a negative angle of slope of the C_Al(C_Li) dependence plot.

In the paper data of morphology, textural and structural features of chromitites from deposits of southeast part of Kempirsay massif (South Urals, Kazakhstan) are summarized. It is showed that formation of unique chromium deposits is closely related with formation processes of wall dunite-harzburgite association and that chromitite localization occur abidingly in olivine monomineralic rock – dunite. Superimposed low-T processes altered primary mineralogical composition of wall peridotites completely but these affected weakly their structure on the micro and macro scale. Mesh serpentine replaced olivine and pyroxene grains but pseudomorphosis of both are survive. Addition, significant displacements of mineral aggregates in the massive peridotite blocks are not observed and it allow to study textural and structural characteristic of chromitites and primary wall ultramafic rocks. We have found some major features of building of ore-bearing associations as follow: (i) increasing chromite grain size according to increasing concentration of chromite, (ii) widespread of deformational structures – ore folding and boudinage, extrusion of solid dunite into massive chromitite, break of ore veinlets. We have performed retrospective analysis of papers about Kempirsay chromitite which in present day are not available for wide readers. Based on this analysis and our observations, we propose a modified dynamic model of chromitite formation as result rheomorphic differentiation of upper mantle matter during its upwelling from deep zone of rift structure with later transformation in the upper mantle of fore-arc setting.

The Larino granite-gneiss dome is located on the border of the Aramil-Sukhtelya and Magnitogorsk megazones and is a two-headed anticlinal structure, the nuclear parts of which are composed of granitoids of the Pervomai and Larino massifs, and their framing – apovolcanic amphibolites, garnetmica-tins, garnet-amphislans siliceous-carbonaceous schists of the Bulatovo Series (S₁ –D₁ bl). A series of concentric highgradient zones of metamorphism are observed from granitoids to shales. Mineral parageneses of the amphibolite facies at a distance of several kilometers are replaced by associations of epidote-amphibolite and greenschist facies.
It has been proven that granitoids belong to the granitemigmatite formation, are close to trondhjemites of the inner parts of the continents (granites, leucogranites) and continental margins (gneisized, migmatized granitoids) and were formed in intraplate and collisional geodynamic settings.
Crystallization of granites took place at temperatures of 780–840°С and a pressure of 2.1–3.7 kbar, plagiogranites – 880–940°С and 2.9–3.3 kbar, which corresponds to the abyssal depth zone (7–11 km). For granitoids formed in the process of metasomatic granitization (gneissized and migmatized rocks), formation temperatures are noticeably lower – 650– 680°С, and the pressure is higher – 6.6–7.0 kbar (almandine amphibolite facies).

According to the data of remote sensing of the Earth, the accuracy of the spatial prediction of soil indicators determining sorption properties in relation to pollutants was compared. To build spatial maps of changes in soil properties, machine learning methods based on support vector regression models (SVMr) and random forest (RF) were used. It was shown that the methods of machine modeling using remote sensing can be successfully used for spatial prediction of the content of particle size fractions, organic matter, pH and the capacity of cation exchange of soils in small areas. It is shown that the spatial prediction of the content of silt fraction is best modeled using the RF algorithm, while the other properties of soils that can determine their sorption potential in relation to pollutants are better modeled using the SVMr method. In general, both machine learning methods have similar spatial prediction results.