Optimization of GIS modules for processing data of gravity monitoring of subsoil in the Republic of Kazakhstan

Abstract

Summary Purpose. The work of a GIS shell with a web interface is analyzed aimed to support oil-and-gas reservoirs during its development by processing the data from gravity variations monitoring. The shell is implemented based on Post¬GIS and tested on the example of processing monitoring data for one oil and gas reservoir. Method. Data processing involves a number of procedures, such as data averaging along the measurement profile by the moving average technique with the calculation of residual local anomalies, the analytical approximation of the gravity curve by a trend, the calculation of the residual field and differentiation (i.e. calculation of its horizontal and vertical gradients). Data modeling means the solution of direct gravimetry problems by the fitting. To optimize the model, mathematical models of the medium were calculated using the simulated annealing technique. Results. The GIS shell GravGIS has been developed to solve applied problems of processing monitoring data for oil-and-gas reservoirs of Kazakhstan. It provides for the calculation of gravity transformants such as first and second derivatives, horizontal and vertical gradients, mathematical norms, etc. For modeling the oil-and-gas reservoir in this shell, GeoM computing unit is introduced for solution of direct problems of gravimetry. A 1-D solution to the problem by the method of simple iteration was developed. Solutions of direct problems of gravimetry by the simulated annea¬ling technique are tested for gravitating objects specified by elementary approximating bodies (sphere, ledge, hori¬zontal and vertical prism, etc.) and are consistent with solutions of these problems by other methods. Practical value. The system allows combining data from monitoring gravity and other geophysical methods. Inte¬gration of these data and gravity inversion with the constraints of a priori parameters of the geological media allows making optimal decisions.