Digital graphic monitoring of energy condition of oil reservoirs

Abstract

Relevance. Control of energy state of reservoirs is an integral part of the overall system for monitoring the development of hydrocarbon deposits. The traditional way to control the energy state of reservoirs is to build isobar maps, while the input data are the materials of well tests in unsteady conditions. In the current technical and economic conditions, it should be considered impossible even conditionally simultaneous shutdown of the entire well stock for the actual determination of reservoir pressure. This shortcoming is devoid of indirect methods for determining reservoir pressure. In this regard, it seems relevant to compare direct and indirect methods for determining reservoir pressure when using their data to analyze the energy state of hydrocarbon deposits. Aim. Comparative assessment of direct and indirect methods for determining reservoir pressure in the analysis of the energy state of deposits (when constructing isobar maps). Object. Tournaisian-Famenian carbonate deposits of oil from the fields of the Perm Krai. Methods. Well tests, analysis of production history by wells (module Topaze (Kappa Workstation)), machine learning methods (modular service Data Stream Analytics (DSA)), mapping, correlation analysis. Results. Well tests carried out at different times do not allow a reliable assessment of the current energy state of reservoirs, in contrast to indirect methods for determining reservoir pressure, the practical implementation of which allows obtaining the desired value for any date. However, with conditionally the same high predictive ability of indirect methods, the considered methods of machine learning should be considered a priority. This is due to their advantageous characteristics, such as low duration of computational operations, a minimum set of initial data, an integrated mapping service.

Keywords

• formation pressure
• well tests
• production analysis
• machine learning methods
• isobar map