Drillstring Mechanical Modeling and Real-time Monitoring
Research output: Thesis › Master's Thesis › Research
For torque and drag modeling, numerous approaches with different levels of sophistication are available. However, these models were originally created for the planning phase of a well. When implementing the models in real time data monitoring, only a selection is suitable for this purpose due to the limited amount of input data available. This thesis therefore presents different soft- and stiff string models and compares them in terms of their potential for an application in real time monitoring at the rig. One stiff string model, which proves most suitable for this purpose, as well as the industry standard soft string model are both put into practice using MATLAB software in order to evaluate, if the higher degree of sophistication associated with the stiff string model, provides a significant gain in accuracy compared to the soft string approach. The percentage differences of the computed data points for a total of 32 wells are analyzed statistically. Additionally, a Wilcoxon rank-sum test is conducted to investigate if the calculated trends match. Another aspect is the future need to integrate the measurements of multiple torque and drag downhole sensors into a torque and drag model. An algorithm is developed, which is able to account for multiple downhole sensor readings by calculating iteratively a set of wellbore friction factors for various hole sections. The results are presented on the one hand by plotting surface hookload (and surface torque respectively) over depth, and also to visualize the soft- and stiff string trends. On the other hand a graphical user interface is introduced that makes use of the simulated data to display weight on bit reserve until buckling, tensional/compressional forces within the drillstring and the associated wellbore contact forces. Additionally, various limitations of the existing torque and drag models and further requirements for the drillstring mechanical modeling software created are discussed in this thesis.
|Award date||24 Oct 2014|
|State||Published - 2014|