Simulating scaling-prone scenarios in downhole tools using Rocky-Fluent two-way coupling
Share this
Published on: October 6, 2020
This blog post follows up on the one we published in 2019 that details how Rocky DEM simulated solids deposition in downhole tools subjected to scale-prone environments. Results obtained by two-way coupling of Rocky DEM to Ansys Fluent indicated fouling hotspots in a sliding-sleeve valve (SSV) as well as a pressure loss uptrend due to the fluid flow impairment by the deposited particles.
Inorganic scaling comprises a number of solid-fluid phenomena involving crystallization, crystal growth, and agglomerate build-up by crystal adhesion, which might end up as solid deposition on valve surfaces. Crystallization and agglomeration follow changes imposed on the pressure, temperature, or chemical composition of the system. For instance, mixing chemically incompatible brines in water-assisted EOR operations may result in barium sulfate precipitation, or the production output may lead to a CO2 flash and calcium carbonate formation. Scaling is a major concern for flow assurance, reliability, and safety, as the solid accumulation damages moving parts, and valves may become permanently shut (or open).
Under a Euler-Lagrange approach, the poly-dispersed system of particles represents precipitated crystals that interact continuously through adhesive forces. Crystals also interact with the fluid flow through fluid-particle forces like drag and dispersion. The present analysis focuses on the influence of scaling severity and particle size distribution. Simulations have been performed in the Research Center for Rheology and Non-Newtonian Fluids (CERNN) at the Federal University of Technology – Parana (UTFPR) with Rocky DEM 4.0 two-way coupled to Ansys Fluent 2019 R3.
As discussed in the previous blog post, the sliding sleeve valve is a completion tool widely employed in wells with selective completion architecture. The SSV geometry is notably complex, since the fluid accelerates as it passes through the holes and the groove constrictions. The geometry is shown from an isometric perspective in Figure 1, indicating the main parts of the valve assembly, such as the outer sleeve with the circular holes, the inner sleeve with the grooves, and the production tubing. The fluid inlets through the upper and lower annular surfaces, flows through the holes, through the grooves and, finally, flows upward through the production tubing. Particles are released from two surfaces situated in the annular region.
Oil companies frequently sample produced fluids from wells to track the production site evolution. Such samples undergo characterization by a PVT expansion test, which allows assessing downhole properties of the fluids and estimating the equivalent viscosity and specific mass of the water-in-oil emulsion. Produced-water chemical analysis results in a chemically detailed composition, which is an input for thermodynamic software to compute properties at liquid-solid equilibrium, such as the precipitation rate. Furthermore, testing super-saturated solutions containing precipitated crystals in granulometry equipment allows engineers to obtain properties like the zeta potential, which is related to adhesion, and the crystal size distribution. Such readily available information is the foundation for setting up and calibrating the Rocky DEM simulation in order to study, for instance, the influence of particle size distribution. Indeed, the particle size distribution depends on the system’s chemical composition, temperature, pressure, and pH. Moreover, there are also studies focused on applying an electromagnetic field to affect the process of building particle agglomerates. Results in terms of the particle size distribution are shown in Figure 2 for a constant solid mass flow rate. The size distribution is a normal curve (Gaussian) with the diameter ranging from 0.1 to 1.0 mm with 75,022 particles released for Case 1. The monodisperse scenario, Case 2, introduces 74,867 particles of 0.5 mm. In Case 3, the particles’ diameter ranges from 0.1 to 0.5 mm, with 311,400 particles released. Particles colored by the magnitude of the translational velocity indicate a distinct adhesion pattern for each one of the size distributions.
Figure 2. Solids adhesion on the SSV as a function of particle size distribution; particles’ color refers to the translational velocity magnitude.
In the top view of the SSV, with particles colored according to the diameter in Figure 3, the deposition occurs mainly adjacent to the groove entrance and also on the outer sleeve of the valve. The increase in particle numbers supports the process of building an agglomerate of tiny particles, which result in a higher pressure increase uptrend, as shown in Figure 4 by the transient curve of the dimensionless pressure (with reference to the one-phase fluid flow pressure). Outcomes are discussed in the conference paper “Numerical Simulation to Study Scale Formation on SSV Valves” presented at the 2020 SPE Virtual International Oilfield Scale Conference and Exhibition.
Figure 3. SSV top view showing the formation of agglomerates in the vicinity of hole-groove interface; particles’ color refers to the diameter.
Figure 4. Pressure drop across the valve as a function of particle size distribution obtained by the two-way Rocky-DEM and Ansys Fluent coupling.
Another issue of interest in the inorganic scaling process is the severity of the scaling, which is measured by the Scale Index or the Saturation Index. An increase in saturation leads to a higher precipitation rate, which is configured in Rocky DEM as solid mass flow rate at an inlet. Results varying the solid mass flow rate considering a monodisperse system of 0.5 mm diameter are discussed in a publication to be presented at the 2020 ATCE conference.
In Figure 5, as the particles’ colors refer to the translational velocity magnitude, the increase in the solid mass flow rate results in more particles in the domain, supporting the formation of agglomerates that adhere to the inner sleeve, constricting the available area for the fluid flow. The dimensionless pressure uptrend shown in Figure 6 demonstrates a nearly linear dependency with the solid mass flow rate.
Figure 5. Solids adhesion on the SSV as a function of concentration; particles’ color refers to the translational velocity magnitude.
Figure 6. Pressure drop across the valve as a function of solids’ concentration obtained by the two-way Rocky-DEM and Ansys Fluent coupling.
Results discussed in the present blog post detail the power of Rocky DEM to simulate poly-dispersed particulate systems, including the effect of adhesive forces immersed in a viscous fluid-flow. The application of the present work in the study of completion tools subjected to inorganic incrustation formation aims to predict the influence of the particle size distribution, as well as the concentration of solids in the pressure drop.
Vinicius Poletto
Research engineer of the Research Center for Rheology and Non-Newtonian Fluids (CERNN) at UTFPR
Vinicius Poletto holds a M.Sc. in Mechanical Engineering and B.E. (Mechanical) degree, from the Federal University of Technology – Parana (UTFPR). He is a research engineer at the Laboratory of Porous Media (LaMP) of the Research Center for Rheology and Non-Newtonian Fluids (CERNN) at UTFPR. Mr. Poletto has been working with CFD-DEM techniques aiming the simulation of the liquid-solid flow for applications in the oil and gas industry, like flow assurance, hydraulics of drilling wells and convection in porous media.
