Horizontal Well Array Induction Logging Data Processing and Interpretation Based on Parameterized Modeling and Inversion

Logging array induction tools in horizontal wells have become a standard practice in the realm of oilfield exploration and production, despite encountering formidable challenges due to the intricate geological structures and limitations of current processing technologies. These complexities often render induction logging curves difficult to interpret, hindering their practical utility. As horizontal well drilling proliferates, the issue of anomalous array induction logging responses has become increasingly prominent. Formation inhomogeneity-including steep borehole trajectory-formation intersections, multiple traversals, and interbedding is identified as the primary source of these complexities. To address this, we adopt dimensionality reduction techniques, simplifying the complex three-dimensional problem into a manageable one-dimensional framework. We then develop an adaptive method that aligns stratigraphy with borehole trajectories and introduce a parameterized modeling and inversion scheme tailored specifically for horizontal wells. Leveraging the Fast Hankel Transform Method, we have developed an advanced array induction simulation program for horizontal wells based on the PM model. This tool allows for a profound analysis of how complex trajectories and inhomogeneous stratigraphy impact array induction logging responses. Furthermore, by adhering to the Born approximation principle, we decompose the logging response into background and local components, enabling their independent processing. This approach markedly improves the accuracy of array induction logging interpretation, making it more suitable for horizontal well application. Importantly, this method has been successfully implemented and applied in real horizontal wells across various oil-producing regions in western China, offering novel insights and methodologies to advance horizontal well array induction logging technology.