This activity aims to clarify the impact turbulence has on cuttings transport and hole cleaning by obtaining a better quantitative description of turbulence in Non Newtonian Fluids and transport mechanisms of solid particles.
Turbulence impacts strongly the wall friction, the transport of cuttings, and pressure drop through the system. As the drilling mud is a Non-Newtonian Fluid the turbulence is impacted by the rheology (Rudman and Blackburn 2006). In addition, the cuttings effect the turbulence structure and thereby the transportation properties of the flow (Deubelbeiss, Kaus et al. 2010; Moysey, Rama Rao et al. 2013). The basic mechanisms responsible for transport are generally not well known. An additional challenge is that the solids (cuttings) fraction may be quite high. Hence, the cuttings may both dissipate and enhance turbulence (Rani, Winkler et al. 2004; Yeo, Dong et al. 2010). Therefore it is important that the effects of turbulence are well understood.
Previously, we have investigated the impact of turbulence on well cleaning. Adding drag reducers to increase velocity is believed to improve hole cleaning. However, if this restricts turbulence then the cleaning efficiency may be significantly reduced.
Direct numerical simulations will be used for understanding the turbulence structure. The particle flow, including the cuttings bed motion, will be simulated by a particle in cell method (Cloete, Johansen et al. 2012), coupling a discrete particle element model (Moysey, Rama Rao et al. 2013) with the flow. The model will include particle rotation. A full DNS, including the particle boundary layers will not be attempted. Instead, the drag and lift forces in Non-Newtonian flows may have to be studied in individual studies, resulting in linearized drag and lift forces. When good quality models and data are found from the literature, these will be adopted.
The motivation is to learn so much about the flow that simplified RANS /U-RANS models can be developed.
The fundamental modeling work will be supported by experiments. For the experiments, there will be performed pipe flow experiments of velocity and stress profiles for selected rheologies and Reynolds numbers. Using index-matched particles, the velocity profiles and Reynolds stresses will be measured for given bed heights and solids mass flow rates.
In parallel, development of simplified flow models, appropriate for the pressure drop and cuttings transport in 1D flow models will be developed. These models will be based on combinations of analytical solutions, 1D numerical solutions in the flow cross section (Johansen, Skalle et al. 2003) and empirical correlations. The models will describe the flow both in drill-pipe and annulus, including drill-pipe rotation and eccentricity. The development here will to a large external be based on multi-dimensional RANS /U-RANS models (Johansen, Wu et al. 2004; Wu, Shyy et al. 2004) which we will develop. These models will use the learning from the DNS work and experiments to arrive at good closures which can give acceptable accuracy. Similarly, we will develop the solids transport model for Non-Newtonian fluids. Modeling challenges here may be significant, as the particle response to the fluid depends on the local fluid strain and possibly strain history.
The simplified flow models will be developed such that they can be adopted in simulators for drilling operation and troubled shooting.
The multi-dimensional RANS /U-RANS models will be available in a simulation tool which allows two-phase flows of cuttings and drilling fluid. The model will allow particle fluid interaction and solids build-up due to wall friction. This will aid in understanding avalanches related to critical conditions for cuttings slides and the understanding the flow of cuttings and fluids during tripping.