Challenges of Modeling Drilling Systems For the Purposes of Automation and Control

Abstract To fully model every aspect of the process of drilling a borehole is still in the realms of research. Great strides are being made to develop high-fidelity models of well-defined domains such as the rig systems, drillstring, rock-bit interaction, fluid control systems and the Earth. Bringing all these models together in any unified manner and proposing a unified control solution to fully automate the whole process is still an exploratory venture. The uncertainty prevailing over the magnitude and spatio-temporal distribution of disturbances to be controlled or rejected by systems best described by non-linear partial differential equations rather than linear approximations, makes for a very challenging control problem. This uncertainty also raises interesting questions on how detailed the models need to be and how this might change our approach to modeling in the future. However technology is never static and certain developments are currently in play that will dramatically improve our capacity to model and control processes which are currently considered too complex to control.