Revisiting Laterolog LL7 Data Inversion for Improved Resistivity Reconstruction and Boundary Detection in Multiple Thin-Bed Reservoirs

Summary Reanalysis of Laterolog-7 (LL7) data in thin-bed sequences suffering from severe resistivity suppression caused by shoulder-bed averaging and mud filtrate invasion remains important. We developed a forward modelling and inversion workflow based on the Finite Difference Method (FDM) in cylindrical coordinates (r, z), reducing the problem to an efficient two-dimensional solver. The LL7 bucking current focusing condition is satisfied at every forward step, ensuring correct reproduction of the dynamic instrument response under high resistivity contrasts. The bucking current ratio log (n1/n2) serves as a boundary-detection indicator to initialize the layered-earth model and reduce non-uniqueness. Formation resistivity profiles are recovered through an Inexact Gauss-Newton (IGN) scheme that avoids explicit Jacobian formation, maintaining computational efficiency while guaranteeing a valid descent direction. Synthetic validation demonstrates recovery of the resistivity structure of a single-bed model (0.6 m bed of 30 Ωm with 5 Ωm invaded zone). The inversion successfully recovers the true resistivity of 30 Ωm and background of 10 Ωm. A multi-bed test shows the inversion resolves bed boundaries at decimeter scale, recovering layers as thin as 20 cm from a smeared raw log. Application of the inversion to a real LL7 dataset from the Cuu Long Basin, Vietnam, confirms field applicability. The inversion reconstructs sharper resistive peaks throughout the thin-bed interval, providing an improved representation of true formation resistivity.