Kriging Is All You Need to Obtain Rock Mechanical Properties From Sparse Spatial Data

Summary Over the last decade, the number of horizontal and multilateral wells has continuously grown. One can also spot the significant fraction of horizontal well sections completed with multistage fracturing right after drilling in past years. It happens as the target geological intervals are becoming more and more complex in terms of geometry and transport/storage properties. The complex drilling and completion are rather expensive and risky in terms of potential failures associated with the mechanical properties of the drilled/fractured rock. To avoid borehole-stability issues, stuck pipe events, and ineffective fracturing, one needs to have a proper tool for accurate planning of these technological operations. To perform the planning, it is essential to have proper input of rock mechanical data. We present a workflow for estimating the mechanical properties of subsurface rock using drilling data. We apply kriging mapping of 1D arrays of mechanical properties along the nonvertical boreholes to the entire 3D domain of a reservoir. The actual field data is used to validate and compare the different scenarios of the 3D mapping application. The mapping workflow developed in this paper is applicable for real-time updating of 3D geomechanical property maps with each new well section drilled in the field. The test on the real field data shows that the proposed workflow allows to reach an accuracy for property mapping higher than 0.65 in terms of the R-squared metric, while adding each new well in a newly drilled reservoir can increase the overall accuracy by a factor of 2–10 even without reference to any nondrilling measurements such as seismic.