Modeling Fluid Interfaces During Cementing Using a 3D Mud Displacement Simulator

Abstract In completion of oil and gas wells, cementing operations are employed to provide zonal isolation, a means to prevent wellbore fluids from contaminating sensitive zones such as freshwater aquifers. Perhaps the most important factor engineers and operators should consider for successful cementing is adequate drilling-fluid removal, or "mud displacement." To help optimize mud removal, the primary technique used is to pump a spacer fluid with modified rheology that creates a favorable fluid-fluid interface to enhance mud displacement. In many instances, it is highly desirable to monitor how this interface evolves over time. Fluid intermingling may inhibit the ability of a fluid to perform its intended purpose, for example, intermixing of spacer fluid with cement slurry may lead to contamination of the cement. This contamination may cause an undesirable failure of the setting of the cement and, consequently, a significant increase in cost because of increased wait time or remedial repair. Therefore, a three-dimensional (3-D) simulator modeling the intermixing of wellbore fluids in a highly eccentric annulus with casing reciprocation and rotation has been developed. The computational system is formulated on a general curvilinear coordinate system whose boundaries can conform to irregular boreholes such as those with washouts. Unlike existing models limited to weakly eccentric annuli without casing movement, the present simulator handles multiple real-world effects and efficiently performs trade-off studies that can enable more economical and effective cementing jobs. The finite difference model provides visual output useful in prejob design and post-job analysis. Among these outputs are 3-D color plots illustrating axial velocity, concentration, viscosity, and density evolution. Introduction Efficient mud displacement is perhaps the most important factor in providing a successful cement job. The primary technique used today is to pump a spacer fluid ahead of the cement slurry. Several other factors that directly impact mud displacement are also known, including wellbore geometry, mud conditioning, casing movement via reciprocation and rotation, casing centralization, and optimizing the pump rate.1,2 However, it is often unknown the extent to which these variables affect mud displacement, especially when applied in combination with one another. Even a relatively straightforward cementing operation can quickly become a challenging scenario with multiple variables. The industry has conducted numerous large-scale physical studies3–8 over the last half-century to empirically evaluate the importance of these factors on displacement efficiency. More recently, however, a number of studies have taken advantage of computational numerical methods to describe the different aspects of the mud displacement process in annular geometries. Tehrani et al.9 discuss combined theoretical and experimental studies of laminar displacement in inclined eccentric annuli. The authors appropriately couple the momentum equation with the concentration equation suggested earlier by Landau and Lifshitz.10 Cui and Liu11 address helical flow in eccentric annuli based on the bipolar coordinate system. Pelipenko and Frigaard12 examine fluidfluid displacement in a two-dimensional (2-D) "narrow annuli" without casing reciprocation or rotation. The well known model discussed by Escudier et al.13,14 considers non- Newtonian viscous helical flow in eccentric annuli for a single fluid.