Modeling Real-Time Sag in the Wellbore

Abstract Weight material sag occurs during drilling operations when drilling fluids remain idle in the wellbore for time periods ranging from hours to several days at static or low-shear conditions. This paper presents a novel method of predicting real-time sag behavior in the wellbore. The method also accurately predicts the fluid mud weight collected at the surface as it is circulated out after a sag event, providing a significant advancement in drilling fluids engineering. Weight material sag is broadly viewed as the settling of barite particles in drilling fluids (primarily oil-based) under various mechanisms. Fluid composition (barite size/concentration and oil/water ratio) and fluid properties (i.e., rheology) influence the barite settling rate. The well geometry (i.e., inclination and diameter) also affects sag behavior. In addition, well operating conditions (i.e., temperature, pressure, and time for which fluid is uncirculated) also influence sag severity. A comprehensive computational approach was developed to model the sag behavior in wellbores using fluids composition/properties and wellbore geometry/conditions information. The wellbore sag model predicts information about the changing fluid composition along the wellbore depth as affected by sag, when the fluid in the well is under idle conditions (static or low shear). Specifically, for a wellbore section of a given inclination, the model provides quantitative density/rheology estimates of the barite-depleted zone at the top of the section and the barite-accumulated zone at the bottom. As the sag-affected fluid is circulated out after an idle period, the model also predicts the varied mud weight of the fluid collected at the surface and the corresponding transient bottomhole equivalent circulating density (ECD) as the circulation begins. The wellbore sag model predictions, especially the variation in the mud weight of the fluid collected at the surface after the sag event, were tested on several wells and fluid systems. The model provided good correlations between measured and predicted fluid density, typically less than 0.1 lbm/gal. The model accurately demonstrates how certain fluids show good sag resistance in some wells yet fail in some complex wells. The model also shows excellent sag resistance of certain specialized fluids in complex wells, which is in agreement with the field observation. A method to accurately determine real-time sag in the field has been a long-standing need. The successfully validated wellbore sag model, which captures the combined effect of fluids composition/properties and wellbore geometry/conditions, could serve as a useful tool for mud engineers to evaluate the sag behavior. It could also enable fast decision making at the rig site to optimize fluid formulations and operating conditions for sag management.