The behavior of horizontal solid−liquid (slurry) pipeline flows was predicted using a transient three-dimensional (3D) hydrodynamic model based on the kinetic theory of granular flows. Computational fluid dynamics (CFD) simulation results, obtained using a commercial CFD software package, ANSYS-CFX, were compared with a number of experimental data sets available in the literature. The simulations were carried out to investigate the effect of in situ solids volume concentration (8 to 45%), particle size (90 to 500 μm), mixture velocity (1.5 to 5.5 m/s), and pipe diameter (50 to 500 mm) on local, time-averaged solids concentration profiles, particle and liquid velocity profiles, and frictional pressure loss. Excellent agreement between the model predictions and the experimental data was obtained. The experimental and simulated results indicate that the particles are asymmetrically distributed in the vertical plane with the degree of asymmetry increasing with increasing particle size. Once the particles are sufficiently large, concentration profiles are dependent only on the in situ solids volume fraction. The present CFD model requires no experimentally determined slurry pipeline flow data for parameter tuning, and thus can be considered to be superior to commonly used, correlation-based empirical models.

No abstract

In this paper, the results of a multi-year research project to develop reliable engineering scale-up models of water-assisted pipeline transport of heavy oils and bitumen are described. Empirical correlations currently in use do not properly account for the effects of flow rate and pipe diameter on friction losses. They account not at all for effects of water cut, temperature, oil viscosity or sand concentration. Additionally, sand accumulation in operating pipelines is a concern because no accurate method of predicting the conditions under which sand can be transported is available.In water-assisted pipeline transport, water present in the production fluid can form a layer that separates the oil-rich core from the pipe wall, thereby drastically reducing the energy required to transport the mixture. Alternately, small amounts of water can be added to provide the lubricating effect.A multi-year project to explore water-assisted flow regimes was sponsored by Husky Energy, Nexen Inc., Shell Canada Energy and four other heavy oil and/or oil sands producers. An extensive experimental test program was carried out in SRC's 50, 100 and 260 mm (diameter) pipeline flow loops, using oil/water/sand mixtures containing heavy oil, bitumen or a viscous lube oil. Measurements collected during the tests included the frictional pressure drop, thickness of the oil wall fouling layer and solids concentration distribution. The friction loss model developed as part of this project assumes that the flow is only partially lubricated by the water layer so that oil-oil contact at the pipe wall becomes more important as the water cut, superficial mixture velocity and/or ratio of oil-to-water viscosity decreases. The sand transport criterion developed here compares the particle terminal settling velocity to the friction velocity of the turbulent water layer.The models developed here provide accurate predictions for the scale-up, design and operation of water-assisted pipeline flow technology, which has significant potential to reduce the costs and environmental impact associated with heavy oil production and transportation.

The dynamics and wall collision of inertial particles were investigated in non-isotropic turbulence of a horizontal liquid channel flow. The inertial particles were $125~\unicode[STIX]{x03BC}\text{m}$ glass beads at a volumetric concentration of 0.03 %. The bead-laden flow and the unladen base case had the same volumetric flow rates, with a shear Reynolds number, $Re_{\unicode[STIX]{x1D70F}}$, of the unladen flow equal to 410 based on the half-channel height and friction velocity. Lagrangian measurements of three-dimensional trajectories of both fluid tracers and glass beads were obtained using time-resolved particle tracking velocimetry based on the shake-the-box algorithm of Schanz et al. (Exp. Fluids, vol. 57, no. 5, 2016, pp. 1–27). The analysis showed that on average the near-wall glass beads decelerate in the streamwise direction, while farther away from the wall, the streamwise acceleration of the glass beads became positive. The ejection motions provided a local maximum streamwise acceleration above the buffer layer by transporting glass beads to high velocity layers and exposing them to a high drag force in the streamwise direction. Conversely, the sweep motion made the maximum contribution to the average streamwise deceleration of glass beads in the near-wall region. The wall-normal acceleration of the beads was positive in the vicinity of the wall, and it became negative farther from the wall. The investigation showed that the glass beads with sweeping motion had the maximum momentum, streamwise deceleration, and wall-normal acceleration among all the beads close to the wall and these values increased with increasing their trajectory angle. The investigation of the beads that collided with the wall showed that those with shallow impact angles (less than $1.5^{\circ }$) typically slide along the wall. The sliding beads had a small streamwise momentum exchange of ${\sim}5\,\%$ during these events. The duration of their sliding motion could be as much as five times the inner time scale of the unladen flow. The wall-normal velocity of these beads after sliding was greater than their wall-normal velocity before sliding, and was associated with the rotation induced lift force. Beads with impact angles greater than $1.5^{\circ }$ had shorter interaction times with the wall and smaller streamwise and wall-normal restitution ratios.