TX 75083-3836, U.S.A., fax 01-972-952-9435.
AbstractThe evaluation process for oil and gas reservoirs comprises the accurate estimation of underground formation resistivities. A substantial amount of Russian and old electrical log resistivity data has been acquired in the past that can be used in an enhanced interpretation process. Old electrical devices continue to be used in various parts of the world. Due to the characteristics of these instruments, which do not provide symmetric responses (unfocused logs), the data are difficult to interpret and require a large degree of expertise when conventional interpretation techniques are used (e.g., charttype processing). In addition, conventional interpretation does not allow for extracting all information contained in the data, for which oil and gas bearing formations may be overlooked.Modeling and inversion methods represent an excellent way to interpret these types of resistivity logs. These methods allow for simultaneously interpreting the entire suite of available logs, incorporating the physics of the measurements, taking into account borehole, invasion, and shoulder-bed effects, and involving data uncertainties. Inversion methods provide reliable estimates of formation parameters and quality control indicators to assess the accuracy of the interpretation results.A new philosophy for interpreting these resistivity logs, based on two-dimensional modeling and inversion, has been developed and a strategy implemented, allowing the log analyst to perform the full-spectrum of data processing required by the interpretation. The complete process involves the following major steps: data depth matching, mud resistivity estimation, log calibration, environmental corrections, utilization of prior information, and estimation of formation parameters.The inversion technology has been implemented on both scalar and parallel machines, the latter providing a fast turn around time. The proposed scheme was evaluated on synthetic cases and applied to a set of Russian field data acquired in Western Siberia. The results are presented showing the proposed methodology's ability to obtain improved estimates of formation parameters and a consistent earth model explaining the entire suite of logs.