Novel Logging While Drilling Azimuthal Laterolog Resistivity Instrument Design for Oil-Based Mud

This paper proposes a novel structure for the Logging While Drilling (LWD) azimuthal laterolog resistivity instrument specifically designed for oil-based mud (OBM). The instrument utilizes capacitive coupling to quantitatively evaluate formation resistivity, permittivity, and the standoff between the measuring electrode and the wellbore wall. This innovation addresses the challenges associated with azimuthal resistivity measurement in highly inclined and horizontal wells, while also offering a deeper depth of investigation (DOI). The tool response characteristics and data inversion methodology were analyzed through numerical calculations. Initially, a 3D numerical simulation model of finite element method (FEM). that incorporates the instrument structure was developed and its accuracy was verified. By adjusting the electrode size, its DOI was optimized to over ten centimeters. The measurement impedance was examined in relation to varying measurement frequencies, the standoff and the electrical parameters (resistivity, permittivity) of both the mud and the formation. Then, suitable instrument parameters and effects of formation and OBM were clarified. The simulation results show that the optimum frequency range is 0.1 MHz and 100 MHz. To reduce the impact of standoff, the diameter of the drill collar should be larger than that of the conventional tool. Finally, a deep learning workflow was established, encompassing deep neural network (DNN) training, parameter optimization, and testing. The accuracy and generalizability of the inversion method were validated using two complex synthetic stratigraphic models. The results of the inversed resistivity and standoff have good consistency with the set values of the two models. System numerical analysis demonstrates that the proposed instrument scheme can effectively address the drilling and formation evaluation requirements for ultra-deep wells and unconventional oil and gas resources, thereby providing a new option for data inversion.