ABSTRACT: In the field of geosteering, azimuthally-sensitive, electromagnetic tools are regarded as one of the major advances in gaining a detailed picture of the formations away from the wellbore. In this area, inversion techniques are commonly deployed to reduce the interpretation overhead for operators and service providers. Inversion processing for distance to boundaries is often regarded as a black-box solution, with uncertainty surrounding the results that bring the validity of the inversion process into question. This paper will present a comparison of a well with well-defined boundaries mapped using conventional and azimuthal-resistivity methods to a completely unconstrained inversion. A further review will be examined by comparing these results to a partially constrained inversion using the same input dataset. The functional differences between the two inversion results and their implications to real-world geosteering operations will also be discussed. Finally, the paper will demonstrate the results of the inversion against the main lateral of a well that was geosteered using azimuthal resistivity and azimuthal density, with a comparison of the results and discussion of the ambiguities of geological interpretation against the inversion results.
INTRODUCTION In drilling a well, only limited information regarding the exact nature of the formations is known owing to inaccuracies of remote sensing over large distances (e.g., via seismic or offset well interpretation). As the well is being drilled, logging devices in the drillstring provide valuable information that can continuously improve the geological model of formations. Combined with geosteering, a well path can be adjusted on the fly to a position that optimizes recovery of hydrocarbons (Stockhausen et al. 2008; Pitcher et al. 2011). Until the mid-2000s, logging devices to guide geosteering consisted of non-azimuthal resistivity, gamma ray, and neutron tools, with azimuthal density tools to provide structural control.