Liming Yu

Determination of relative angles and anisotropic resistivity using multicomponent induction logging data

We have developed a new algorithm that retrieves information about relative dip angle, relative azimuth angle, vertical resistivity, and horizontal resistivity from multicomponent EM induction logging data. To investigate how relative dip and azimuth angles affect multicomponent induction logging data, we performed a sensitivity analysis using an anisotropic whole space model. Based upon the sensitivity analysis, we designed a two-step procedure to recover relative dip, relative azimuth, horizontal resistivity, and vertical resistivity. In the first step, the observed data are transformed into a new data set independent of the azimuth angle; a simultaneous inversion method recovers relative dip angle, vertical resistivity, and horizontal resistivity. In the second step, a 1D line search is performed to decide relative azimuth angle. Synthetic and field data tests indicate that the new inversion algorithm can extract information about relative dip and azimuth angles as well as the anisotropic resistivity structure from multicomponent induction logging data.

Multicomponent induction response in a borehole environment

Field and numerically simulated data show that coplanar induction measurements (σxx and σyy) are affected by both conductive and resistive muds. The effect is, for a resistivity contrast of 10 or less between the formation and mud, approximately proportional to the resistivity contrast. The borehole effect is generally stronger for a decentralized tool. In the perpendicular eccentricity mode, the eccentricity effect on σxx or σyy can be greater than 100% of the true formation responses. We demonstrate that dual‐frequency processing can reduce the eccentricity effect by an order of magnitude. We investigate the effect of invasion for both circular and elliptic invasions. For circular invasion, the effect on σxx, σyy, or the coaxial measurement (σzz) is characterized by a critical invasion thickness. For invasion thinner than the critical thickness, the effect is minimal. The effect increases rapidly as invasion thickness increases beyond the critical thickness. For elliptic invasion, both the σxx and σzz r...

A Fast Inversion Method For Multicomponent Induction Log Data

Summary A new multicomponent induction tool has been developed for detecting resistivity anisotropy of the formation. This new tool comprises of three mutually orthogonal transmitter-receiver coil configurations measuring all components of the magnetic field tensor. To provide the horizontal and vertical resistivities of the formation at well site, we developed a fast, robust, and hands-off inversion method. The method describes the layered formation using equally thick thin layers. A 1-D vertically layered forward modeling code is used to generate the synthetic data. The formation resistivities are updated iteratively by the difference between the synthetic data and the measured data without computing a sensitivity matrix. In a vertical well, the horizontal resistivity is inverted from the Hzz * component first, and then the vertical resistivity is inverted from the Hxx or Hyy component. In a deviated well, we use the same methodology, but we apply a constraint to derive horizontal and vertical resistivities simultaneously. We test the method on synthetic examples and we will show that the algorithm is fast and robust.