LWD PROPAGATION RESISTIVITY INVASION PROCESSING BASED ON DIELECTRIC-INDEPENDENT RESISTIVITIES OR ON ALTERNATIVELY-PARAMETERIZED RESISTIVITIES

摘要

Logging-While-Drilling (LWD) invasion processing based on standard phase and/or attenuation resistivity values often leads to substantial underestimates of the true un-invaded formation resistivity when dielectric effects are present on the data. The current paper discusses the advantages offered by the use of dielectric-independent resistivity values in the invasion processing algorithm instead of the standard resistivities. Substantial increases in the un-invaded formation resistivity are noted in resistive formations such as the pay-zones. T R Dielectric-independent resistivities evaluated jointly from phase-shift and attenuation measurements but without a dielectric assumption were not previously used for the purpose of invasion processing since they were thought only to be valid in homogeneous, un-invaded formations. Recently, it was shown that the dielectric-independent data are valid even for invaded or inhomogeneous formations. This innovation lead to the development of the new invasion processing algorithm proposed in the current paper. Comparisons against conventional results for cases where a commonly used dielectric assumption is violated, illustrate that this procedure is more accurate and runs faster. It was also shown that standard phase and attenuation resistivities were not valid in inhomogeneous or invaded cases, even if the formation satisfies the dielectric assumption. To overcome this issue, one must use alternative phase and attenuation resistivities, which are computed jointly from both measurements by permuting the dielectric assumption. Correspondingly, the invasion processing based on the alternative resistivities is also analyzed. It was noticed that if the dielectric assumption is met, the computation is highly efficient in terms of both accuracy and reduced CPU-time. Field and synthetic data examples are provided in addition to theoretical discussions. Since the methodology is suitable to multiple frequency type data, it can be correlated with frequency dispersion correction algorithms, accounting for effects caused by geological formations having the electrical response governed by multiple time constants.