Effective and Ineffective Strategies for Mud Removal and Cement Slurry Design

Uncontrolled flows of reservoir fluids behind the casing are relatively common. In worst cases these can lead to blow out, leakage at surface, destruction of subsurface ecology and potential freshwater contamination. Often, safe abandonment of such wells is not possible. A significant cause of flows behind the casing is ineffective mud removal during primary cementing. The ideal situation is that drilling mud is displaced all around the annulus and that the displacement front advances steadily up the well at the pumping velocity. Even better is that the wide and narrow sides of the front advance at the same speed. Conversely, if the fluid on the narrow side of the annulus does not move, or moves very slowly, a longitudinal mud channel can result. Although the possibility of a narrow side mud channel and benefits of a steady state displacement have been recognised since the mid-1960s, there is still little quantitative understanding of when steady state displacements occur. In this paper we present new results on the displacement of cementing fluids along eccentric annuli. We show that for certain combinations of the physical properties there will be a steady state displacement front. Furthermore, we are able to give an analytical expression for the shape of the front and indications of how the shape changes with the key physical parameters of the cementing process. These results are novel and have interesting implications for effective mud removal and complete zonal isolation during primary cementing. Introduction Primary cementing is a critically important operation in the construction of any oil well, (Ref. 1). Apart from providing structural integrity to the well, the chief purpose of the operation is to provide a continuous impermeable hydraulic seal in the annulus, preventing uncontrolled flow of reservoir fluids behind the casing. Serious problems may arise from such flows. Gas or oil may flow to surface causing a blowout, with consequent environmental damage and possible loss of life. Fluid migration into subsurface aquifers can cause contamination of drinking water, or can affect near-wellbore ecology. Even if surface casing vent flows are contained within the annulus, the fact of having pressure at surface prevents a well from being permanently abandoned, i.e. safely, at the end of its lifetime. Instead, these wells become permanently shut-in and remain an environmental risk. Financial consequences of poor zonal isolation on reservoir production are of course well known. A widely cited industry figure, (Ref. 2), is that 15% of primary cementing jobs carried out in the US fail and that about 1/3 of these failures are due to gas or fluid migration. This problem exists worldwide, (e.g. about 9000 wells are suspended or temporarily abandoned in the U.S. Gulf Coast region). It is also particularly acute in Western Canada, where around 34,000 wells are currently shut-in and suspended, (Ref. 3), unable to be permanently abandoned due to gas pressures at surface, between the casing and formation, in the cement. Local variations in this problem can be extreme. For example, field survey results reported in Ref. 4, from Tangleflags, Wildmere and Abbey, (3 areas in Eastern Alberta), reveal that over a number of years, 0-12% (Tangleflags), 0-15% (Wildmere), and 80% (Abbey) of wells have been leaking in these regions. One known cause of surface casing vent flows is that the cement, which is placed in the annulus between the outside of the casing and the inside of the hole, fails to fully displace the drilling mud that initially occupies this space. Incomplete mud removal can manifest in different ways. First, a coherent mud channel can form on the narrow side of the eccentric annulus. Second, residual mud or spacer can contaminate the cement as it sets. Third, mud may remain static in layers stuck to the inner and outer walls of the annulus. In the first or last case, the residual mud dehydrates as the cement sets and allows a porous conduit to develop in the annulus. In the case of severe contamination, the cement may not properly set. These problems and the consequences described, provide the main motivation for this paper. This paper focuses on avoiding all three above cases, by ensuring that the initial displacement is steady. Unless there are significant losses during primary cementing, a steady stable displacement front advancing at the mean pumping speed all around the annulus is sufficient to ensure that the displacement is effective. Therefore, we consider the important question of whether these displacement fronts can be found in a realistic mathematical model of the process. Certainly, physical intuition tells us that certain simple cases should have steady displacement profiles, e.g. a heavy viscous fluid displacing a lighter less viscous fluid in a SPE 80999 Effective and Ineffective Strategies for Mud Removal and Cement Slurry Design I.A. Frigaard, SPE, University of British Columbia, and S. Pelipenko, Schlumberger