Logging-while-drilling (LWD resistivity measurements are frequently affected by the presence of shoulder/proximity beds. The traditional approach of making formation resistivity corrections at the center of the bed based on premodeled charts is not practical for horizontal wells, for which the distance to boundary and shoulder resistivities are usually not known. Adding to the difficulty, such corrections must be applied over thousands of meters.

This paper presents a model-based inversion approach to remove the shoulder-bed effect in multispacing LWD propagation resistivity tool measurements in high-angle and horizontal wells by inverting for the true formation resistivity, shoulder resistivity, and the distance to the shoulder at the same measurement depth. Fundamental to the success of the method is the utilization of a three-bed effective model to approximate the effect of formation above and below. This model greatly reduces the complexity of the problem and makes it possible for this method to be used by petrophysicists for practical interpretation. Although the measurements are subject to the influence of all the formation layers above and below, the paper demonstrates that the responses are not sensitive to the fine details of the shoulder-bed formation and that an effective three-bed model is accurate enough to represent complex layered formations for shoulder-bed correction purposes. A heuristic-based strategy is also utilized to select the best solution from multiple possible candidates to improve the consistency of the inversion results and remove ambiguity caused by the existence of multiple solutions. Considerable optimization has also been applied to speed up the inversion so quick interpretation becomes possible.

The algorithm has been tested successfully against synthetic logs as well as actual field logs. Results show that when it is a dominant environment effect, the shoulder-bed effect can be removed and the true formation resistivity can be obtained. In certain situations, the distance to the boundary can be determined when within 3 feet true vertical depth of the boundary. Integration of bed boundary correction with corrections of other single environmental effects will also be presented. Applications of such log-processing techniques to practical interpretation, ranging from simple automated inversion with minimum user input to expert guided inversion to improve the quality of the overall inversion results, will be presented.

This content is only available via PDF.
You can access this article if you purchase or spend a download.