The use of logging-while-drilling (LWD imaging tools in both post drilling analysis and real-time decision making applications is becoming more prevalent. These applications provide benefits in a variety of disciplines ranging from drilling engineering to petrophysics. Because LWD images are acquired by rotating measurement sensors with the drill string, certain benefits over wireline pad tools, such as 360 degree borehole coverage and the ability to use additional sensor types such as density, resistivity and gamma-ray, can be realized. However, these benefits are not without their corresponding limitations. Artifacts and processing errors due to drilling parameters, borehole environment effects, image acquisition methods, and the inherent measurement physics can create problems in image interpretation. In particular standard density images tend to amplify artifacts from small borehole irregularities, and cause borehole shape and geological features to overwhelm each other.
This paper presents processing methods to turn these limitations to our advantage for two density image applications: wellbore shape characterization and thin bed analysis. The new processing accomplishes the following: a reduction in image artifacts using sensor depth and resolution matching techniques; an improved borehole shape computation that accounts for non-linear density measurement behavior; and an image resolution enhancement based on relying more heavily on gamma counts from short spacing detectors.