Processing borehole imaging tool measurements and interpreting image data involve complex workflows comprising of many computation steps processing gigabytes of high-resolution borehole image data at once. These workflows are among the first computations run after the acquisition of borehole images, and the dips that are interpreted on the images (e.g. beds and fractures) are particularly valuable in time-critical contexts for near real-time decision for example for geosteering. Over the years, these computations have been individually improved to provide better results with fewer manual interactions.

We identified an opportunity to improve the efficiency of these interpretation workflows in the current trend of digital transformation. Setting up and running an interpretation workflow for borehole images has so far remained a sequence of time consuming and intellectual operations coupled with mundane tasks, requiring the interpreter to be trained on the processing of each type of imager and his full attention from the start of the workflow to its end. These requirements are obstacles to early interpretation, especially for locations where the ability to work remotely and transfer image data in a timely manner is impaired due to internet connectivity being slow, very costly to setup, or insecure.

This presentation describes how we built fully automated workflows to process multiphysics borehole image logs and to extract the bed surfaces from them, down to a single click and with no supervision. These automated workflows allow users to automatically generate high-quality borehole images, interpret dips and differentiate between sedimentary and structural features, serving as inputs for further structural reservoir modeling or detailed sedimentary analysis.