Data-Driven Preprocessing of Gravity Data in Oilfield Gis Monitoring System in Kazakhstandata-Driven Preprocessing of Gravity Data in Oilfield Gis Monitoring System in Kazakhstan

Abstract

Summary A necessary prerequisite for correct subsoil use is the prevention of pollution of the subsoil and the surrounding environment. An effective tool for this aim is the instant monitoring of negative phenomena by set of geophysical fields data. The least resource-consuming and most informative in this set is the monitoring of the gravity field and the corresponding transforms in order to maximize the derivation of useful information from data. Large bulks of initial data of the survey, the complexity of combining heterogeneous a priori information, and decision-making time constraints dictate the choice of a monitoring tool as a kind of spatial database of attributive data, combined with filtering procedures, primary data processing, forward and inverse modeling of surveyed fields. These requirements are partially met by proprietary GIS packages. In the first approximation, an information system for monitoring the state of an oil and gas field has been developed, containing a module for solving the forward problem of gravimetry using global optimization techniques for a set of simple mathematical models of the geological media, concluded in a priori known range of depths in an oil and gas field. This system has a data-driven character and allows for: the processing of primary data from field studies of variations in the gravity field; filling the database interconnected with other data; import of a priori data; statistical data processing; stable solution of the forward problem for three models of the geological media (homogeneous sphere, horizontal prism, vertical ledge) within the GeoM module, where suitable parameters are fitted for given models by minimizing the residual of gravity variations with global optimization approaches; the solution control by simulated annealing algorithm and genetic algorithm; the displaying post-processing data in the form of a pivot table and appropriate charts.