Casing Design Methodology For Casing While Drilling

Abstract

In the current plans of Delft Aardwarmte Project (DAP), it is considered to perform the drilling operation by using pipes which remain in the well after drilling thus acting as a casing, the so-called “Casing While Drilling” (CwD) technique. Due to the absence of drill pipe tripping prior to casing the well, this technique results in reduced drilling time compared to conventional drilling. Additionally, potential downhole problems due to drill pipe tripping are precluded. This thesis presents a simulation based approach to selecting casing steel of suitable grade capable of withstanding typical loads encountered while drilling of the well and during its producing life. The developed algorithm is then used in the design of the casing string for the proposed DAP geothermal producer well. The algorithm first considers the effect of uni-axial stresses on casing due to defined burst and collapse pressure loads encountered due to loss of well control while drilling or in the production phase to make a preliminary selection. The effect of axial stress due to buoyed weight of casing and the bending stress due to wellbore curvature is then used to re-evaluate the design against the same collapse and burst loads. This is performed by using a bi-axial approach for the former and the Von-Mises triaxial stress criteria for the latter. A ‘Johancsik’ torque and drag model developed in MATLAB is used to predict drag values during tripping. The bi-axial and Von-Mises stress analysis approach is repeated to include the effect of the computed pull-out drag forces. The associated torque values are used to compute torsional stresses and to identify casing connections of appropriate torque capacity. The final step in the algorithm is to simulate the loads occurring during drilling and calculate the equivalent Von Mises stress values throughout the casing string. Typical drilling loads considered include torque and casing lateral vibration which induce torsional and bending stresses respectively. It was identified that rather than bending stress due to whirling or buckling, torsional stress was more likely to cause casing string failure. This is due to its relatively higher magnitude and the weaker maximum torque capacities of conventional casing connections. Additionally, the MATLAB tools developed for analysing buckling and whirling are used to compute the critical load for inducing sinusoidal buckling as a function of wellbore inclination and the critical rotary speed at surface to induce lateral vibration for varying weight on bit (WOB) respectively. From the generated mode shapes, the bending stress magnitude at each node and therefore the points of maximum stress occurrence for bucking and whirling are also identified.