Wellbore instability accounts for a significant amount of nonproductive time. Continuous improvement of today's technologies enables real-time analysis of acquired data to identify and remediate potential drilling hazards. Efforts toward the automation of relevant steps and processes for real-time wellbore stability modeling, monitoring, and calibration are considered highly important. Automation adds expertise to the rigsite engineers and enables simultaneous supervision of more than one well for remote operating engineers.
The complexity of the data flow, the reliability of the models, and the personnel required for monitoring pose great challenges for automation. Take pore pressure analysis as an example; some methods often require manual input from the user for placing normal compaction trend lines. Data used for pore pressure calibration (e.g., gas data) are more qualitative indicators rather than precise quantities. Important data are obtained only at discrete locations, such as leak-off tests for fracture gradient estimation. These examples illustrate the diversity of relevant data and information, which have to be properly managed.
In this paper we present some examples of the automation of single modules of a geomechanical analysis workflow. These include automatic setting of a normal compaction trend line and determination of the onset of overpressure using statistical methods, image interpretation algorithms to identify depth and width of breakouts, and automatic shale discrimination at the bit with drilling vibrations. These modules eventually become the building blocks of a complete automation system for wellbore stability.