Agglomeration and transport of drilling generated particles in the oil well

Abstract

The challenge presented by IRIS was to develop a better understanding of agglomeration and transport of cuttings generated during the drilling of an oil well. There is a myriad of challenges involved in the mathematical modelling of this complex process. For this reason, two complementary approaches are combined in this report. First, we derive a one-dimensional fully-developed model based on the Phillips-type shear induced diffusive migration. We start by assuming a Newtonian nature of the suspending drilling mud, but an extension to account for a non-Newtonian rheology is proposed herein. Second, a direct numerical simulation approach is employed to predict settling rates of drilling cuts when the assumptions made in Part 1 break down and when complex geometries are taken into account. When drilling an oil well, rock cuttings are generated and must be transported to the surface for disposal. For this purpose, drilling fluid (or 'mud') is pumped down inside the drill pipe and exits at the drill bit. There, the drilling fluid combines with the rock cuttings whereupon both are transported through the annulus of the well back to the surface. The cuttings are then separated from the mud as required. This basic process is made complicated by a number of factors. First, the rheology of the drilling fluid is non- Newtonian. Although drilling fluid is normally oil-based or water-based, particles are added to give it a non-Newtonian (shear-thinning) rheology, so as to enhance the transport of the mixture of drilling fluid and cuttings to the surface. Secondly, an oilwell may have significant inclination ('directional wells'), meaning that the cuttings can 'settle' at the bottom of the annulus, forming a 'cuttings bed' potentially leading to clogging of the well. Finally, the drilling is a transient operation characterized by frequent start-ups and shutdowns: typically, the drilling is halted periodically to enable strands of drill pipe to be added as the well grows longer. Frequent shut-downs promote settling and therefore contribute to cuttings-bed formation. The fundamental problem addressed in this Report is to develop a physics-based understanding of the cuttings bed. From a practical aspect this is crucial, as poor control of cuttings may cause critical situations and a loss of the well. In this work, two complementary approaches are taken. In the first approach ("Theoretical Modelling", Part 1), a continuum theory is formulated on the basis that a mixture of drilling fluid and cuttings can be treated in a fluid-mechanical framework. Existing methods concerning dense suspensions can then be applied (and improved as necessary) to predict settling rates as a function of input parameters. For small cutting sizes, such an approach is justified. However, for larger cutting sizes (e.g. rock cuttings comparable to the size of the annular region of the flow domain), such a continuum theory will inevitably break down. To understand the transport processes in this limit, direct numerical simulation is proposed (Part 2). Sample simulations using the freely available OpenFoam fluid solver are presented and future work to improve and refine the computational model is discussed.