Modeling displacement flow inside a full-length casing string for well cementing

Abstract

While computer modeling of annular displacement efficiency is widely applied in cementing engineering, modeling the displacement flow inside a casing or drill string for cementing operations has received less attention. Although predicting displacement efficiency inside a full-length pipe is desired by cementing engineers, the attempt of developing a model with both efficiency and accuracy faces challenges. Access to computer simulators for this purpose is limited. Compared with annular flow, the displacement flow inside pipe, although within a simpler geometry and without eccentricity effect, is not simpler in physics, modelling strategy and predictability, because a variety of flow patterns and flow instabilities can develop to create complicated fluid interfaces. In this paper, we present an integrated numerical model developed to simulate displacement flows inside a full-length pipe, which connects an existing annulus model to enable complete displacement simulations of cementing jobs. The model uses three-dimensional grid to solve fluid concentrations with degrees of mixing, and incorporates flow instability detection and flow regime determination. Applied in cementing, the model accounts for effects of pumping rate, well inclination, pipe rotation, fluid densities, rheological parameters and more. This computationally efficient model does not rely on high-resolution mesh as often required by conventional Computational Fluid Dynamics models, thus it is suitable to be implemented in a cementing software for daily use by well cementing engineers. The methodology of the model is discussed in detail in this paper. To validate the model, we examine simulation results against experimental results obtained in our laboratory tests and CFD simulations; acceptable agreement is found under different testing conditions. We also presented two case studies of real cementing jobs with cement evaluation logs compared to simulation results, showing that the model can predict consistent displacement efficiency results.