Real-time Estimation of Surge and Swab During Running of Casing
MetadataVis full innførsel
The importance of ECD prediction was recognized at an early stage of the drilling history. Narrow pressure windows make the ECD prediction even more important.The variation in ECD with respect to surge and swab pressures is more or less dependent on the tripping velocity of the pipe string. When tripping into a hole the pipe movement causes positive pressures know as surge pressures, while the swab pressures are negative pressures which normally occur during tripping out of the hole. However swab pressures can occur when the string is decelerated during tripping into the hole, likewise surge pressures can occur when tripping out of the hole. According to Burkhardt, there are mainly three effects which can lead to surge pressures; pressure generated by breaking of gelled mud, pressure generated by inertia of mud column and pressure generated from viscous drag of mud column .Clearly the viscous drag of the mud column is the most important contribution to the surge pressure, and also the most complicated one. Burkhardt developed a complicated method which involved finding the viscous drag component of the surge pressure by utilizing a friction factor in the fanning equation. This method was found to complicated for Matlab implementation since it needed several values read from graphs. However approximated Burkhardt equations for laminar and turbulent flow were found useful. The surge pressure could also be found by approximating equations to the computed Bingham pressure loss model for variation in Vp. These equations were in turn corrected with a correction factor for each mud parameter and geometry, by variation of mud parameter, giving a set of New Formulas valid for a range of mud parameters. The two models were implemented in a Matlab script together with the Bingham pressure loss model meant to compute the surge pressure with real-time parameters for a case study.Since the models only evaluated the viscous drag of the mud column, they were considered as steady-state models. The steady-state model did not give good results compared to the measured surge pressure for the case study. However steady-state models were found useful when identifying the peak surge pressure at peak velocity. Addition of pressure effects from inertia of mud column and pressure generated by gelled mud gave good results compared to the measured pressure for the case study. However the two models used requires further field validation before they are accepted as a good surge and swab predictors. It is also recommended that compressibility and tubular elasticity are included to achieve full transformation to a dynamic surge model.