Challenges and Solutions to the Correct Interpretation of Drilling Friction Tests

Abstract

Abstract During drilling operations, it is important to monitor carefully that cuttings are transported out of the hole and that the open hole formations do not collapse. A friction test is a simple procedure involving measurement of pickup weight (PUW), slack off weight (SOW) and free rotating torque (FRT) that provides information about possible changes in mechanical friction between the drill-string and the borehole while drilling deviated wells. It is not uncommon to observe very large variations in measured values between consecutive tests, even if there are no obvious reasons why the downhole conditions should have changed. The inherent problem with the interpretation of friction tests is that many factors influence the observations. First, the drilling fluid circulation conditions have an impact on the forces applied to the drill-string. If friction tests are performed with different flowrates, the results are necessarily dissimilar. Second, as the drill-string is elastic, the transitions from pure rotation, to upward axial movement and then to lowering are transient and require some time before reaching steady state movement. Since the elasticity of the drill-string changes with the bit depth, the minimum distance that is necessary to attain stable measurements varies throughout the drilling operation. A dynamic drill-string mechanical model coupled with a transient hydraulic model has been used to analyze a series of high quality real-time measurements taken during the drilling of a medium complexity well in the North Sea. Half of the friction tests were taken with a totally automated procedure while the other half were performed manually by the drillers. The automatic friction test procedure resulted in greatly improved consistency in test measurements compared to the manually executed tests. The results of this investigation show that many of the variations observed in measurements of the PUW, SOW and FRT were in fact related to slight dissimilarities in the way the manual friction tests were executed by the drillers. This study clearly demonstrates the value of using transient mechanical and hydraulic models to interpret friction tests. Especially in cases where the friction tests are performed manually, these models enable much more reliable indications of whether the downhole conditions are changing or not. Furthermore, the new method has shown that it is possible to even further improve the automatic test procedure by reducing execution time without sacrificing the result quality.