Eliminating human error during drilling: the first step toward automating optimized geothermal drilling

Rate of penetration (ROP) optimization during drilling reduces costs during geothermal exploration and development of geothermal fields.One of the major bottlenecks for ROP optimization is bad zeroing, which depends on hookload and weight on bit (WOB).WOB is a calculated parameter that depends on the weight of the drill string and the hook load.Because the string weight is manually measured after a new stand has been added and before the drill bit is on hole bottom, it is subject to human error and can cause the WOB to deviate from the correct calculation by up to 100\%.The primary purpose of this study is to help drillers avoid bad zero thus preparing the dataset for ROP optimization.We used the 2020 Utah FORGE drilling dataset from well 16A(78)-32 1Hz data to identify when a new stand is added to the drill string and calculated the change of string weight in the dataset once the new stand has been added.This identifies the instances where string weight was incorrectly updated.We demonstrate how our algorithm helps eliminate human error by identifying the window in which the hook load can be updated or by automatically correcting the string weight.This leads to a correct WOB calculation thus effectively optimizing ROP.