Elmod-Putting Electromagnetic Modeling To Work To Improve Resistivity Log Interpretation

Forward modeling is the computation of the log that a tool would produce in a defined formation. Recent advances in numerical techniques and high-speed computing have led to the development of modeling codes for resistivity tools that are now sophisticated enough to compute accurate tool response in actual logging situations. In order to better integrate this improved modeling capability wit h everyday log interpretation, we have made some of our most frequently used electromagnetic modeling codes available for use by both our own and oil company log analysts. These codes are organized in an Electromagnetic Modeling package (ELMOD), which consists of a set of VAX-based command files and programs. Using ELMOD, a log analyst can define a formation and compute and display induction tool response (both conventional and Phasor*) in a format identical to that of field logs. Future releases of ELMOD will add modeling codes for other electromagnetic tools, such as laterologs and the Electromagnetic Propagation Tool (EPT*), to the induction software. In order to illustrate how log analysts can use forward modeling to aid in log interpretation, ELMOD is used to analyze two case studies. The first is a field log where the inverted order of Dual Induction-SFL* curves indicates the possibility of an annulus. As a first step, correction charts and preliminary modeling are used to eliminate borehole effect, shoulder effect and sondeerror adjustments as contributory factors. Next, other data, including nuclear and microresistivity logs, information from log headings, and cores are used along with the resistivity logs to set up a trial model. Resistivities and invasion and annulus dimensions are then systematically varied within physically realistic limits, until a computed log that is identical to the field log is generated. The computed log verifies the presence of an annulus. Water saturation is derived from the final resistivities. In the second case, a similar procedure is used to interpret induction response in dipping beds in order to determine Rt and bed thickness. A trial formation is set up using published examples of dip effect to characterize basic tool response and EPT logs to help locate bed boundaries. Resistivity levels and bed boundary locations are adjusted in a series of ELMOD runs until asquared resistivity profile that reproduces the field log is found. The squared resistivity profile gives a value of & corrected for dip, which is used to obtain better values of water saturation.