This paper aims to study the applicability of machine-learning algorithms, specifically neural networks, for forecasting the effectiveness of Improved recovery methods. Radial jet drilling is the case operation in this study. Understanding changes in reservoir flow properties and their effect on liquid flow rate is essential to evaluate the radial jet drilling effectiveness. Therefore, liquid flow rate after radial jet drilling is the target variable, while geological and process parameters have been taken as features. The effect of various network parameters on learning quality has been assessed. As a result, conclusions on the applicability of neural networks to evaluate the radial jet drilling potential of wells in various geological conditions of carbonate reservoirs have been made.

Current methods of oil and gas field development design rely on reservoir simulation modeling. A reservoir simulation model is a tool to reproduce field development processes and forecast production data. Reservoir permeability is one of the basic properties that determines fluid flow. From existing approaches, the porosity and permeability values should be consistent with petrophysical correlations obtained from core sample tests in the course of development of an absolute permeability cube in the reservoir simulation model. For carbonate reservoirs with complex pore space structure and fractures, the petrophysical correlations are often unstable. To factor in the fluid flow in a fractured rock system, dual-medium models are developed, allowing for matrix and fracture components. Yet in this case, the degree of uncertainty only increases with the introduction of a new parameter: a cross-flow index of fluid migration from matrix to fracture, which is only determined indirectly by results of fluid flow studies conducted in the initial development period, and therefore most often is adaptive. Clearly, for well-studied fields there is an extensive data pool drawn on research findings: core studies, well logging, well flow testing, flowmetry, special well-logging methods (FMI, Sonic Scanner, etc.); the dual-medium model development for such reservoirs is fairly well-founded and supported by actual studies. However, at the start of the field development, the data are incomplete, which renders qualitative dual-medium modeling impossible. This paper proposes an approach to factor in the target’s permeability anisotropy at an early development stage through the integration of well, core and 3D seismic surveys. The reservoir was classified into pore space types, to which different petrophysical correlations were assigned to develop a permeability array, and relative phase permeabilities were studied. The fluid flow model was history-matched with allowance for permeability anisotropy and rock types. Comparative calculations were conducted on the resulting model to select the optimum development strategy for the target.