A typical method to sidetrack from a cased well is to use a milling assembly to traverse an anchored whipstock to create a window, and then continue to drill a rathole deviating from the original well. A sidetrack can also be conducted in an open, uncased wellbore. The operation usually utilizes a directional BHA to traverse anchored or cemented whipstock and drill ahead while steering away from the original wellbore.
To achieve a successful wellbore departure, it is critical to minimize shock/vibration to avoid failure of the milling assembly and drillstring components. The milled window should allow sufficient clearance for the subsequent BHA to pass through without issues. The rathole also needs to deviate away to avoid collision into the original wellbore. Modeling with the capability to evaluate wellbore departure is valuable to plan the operation and avoid unexpected malfunctioning.
The authors describe how a time-based modeling system was used to simulate drillstring dynamics and the cutting action of casing mills. Extensive laboratory tests were done to capture the interaction between the cutting elements/casing, whipstock, cement and formation. The data and the parameters from the specific application were then fed into the model for simulation. The results include milled window profile, rathole wellpath, BHA shock/vibration, and drillstring bending moment/stresses. By studying different scenarios, the milling assembly, BHA and drilling parameters were optimized to reduce the potential for vibration and chance of breakage, and to ensure appropriate window size and rathole length.
Gantry milling tests were conducted in the laboratory. Milling assemblies were used to cut windows in casings cemented in steel containers. Drilling dynamics data including applied weight, RPM and flow rate were collected and analyzed. The geometry of the windows and the profiles of the whipstock were measured after the tests. This information was used to validate the model and good agreement was observed between laboratory tests and modeling results. Validation was also conducted with results from actual field milling runs/operations. The authors will present actual field examples to show the effectiveness of the model to help diagnose and prevent component failure and provide operating parameters recommendations for optimal results.