NOVEL EULERIAN APPROACH FOR THE NUMERICAL SIMULATION OF THE PERFORMANCES OF HYDRAULIC MACHINERIES IN SEDIMENTS – LADEN FLOW CONDITIONS AND THE ESTIMATION OF THE EROSION PATTERNS

The sediment management strategy of a hydropower plant is one of the key aspects affecting the plant environmental impact, since both completely interrupting the sediment flow through the plant and letting uncontrolled streams of sediments flow through the power plant can cause counterproductive results in terms of environmental impact and erosion damages respectively. Therefore, over the past decades, different strategies (e.g sluicing, flushing, coatings application…) have been put in place to try maintaining the continuous flux of sediments through the power plant. However, to progress in this direction, it is mandatory, nowadays, to develop accurate tools capable of predicting changes in performance, erosion hotspot locations and possibly suggesting new design criteria to minimize erosion damages inside fluid machines. In such a context, CFD certainly plays a significant role in simulating sediment laden flows through hydraulic turbines. One option to simulate the particle motion in a fluid is represented by the Eulerian – Lagrangian approach, which accurately reproduces erosion spots, but is also extremely time and CPU consuming. These characteristics make this approach hardly appropriate for industrial applications. The other possible way to simulate sediment flow inside fluid machineries is the Eulerian approach: this approach is much faster than the above-mentioned Eulerian-Lagrangian approach, but it requires a proper tuning to provide accurate results due to the lack of a specific erosion model. Working in such direction, the present work aims to provide and validate a novel CFD model based on the inhomogeneous Eulerian multi-phase approach to study sediments flows occurring inside hydraulic machineries, in particular hydraulic turbines. As a first step, the validation of the proposed CFD model has been performed considering a simple case involving dense slurry flow through a pipe. The inhomogeneous Eulerian approach has been selected and combined with the Kinetic Theory of Granular Flows. Particular attention has been paid to the interaction between pipe walls and the solid phase, imposing a no-slip condition for the sediments. During the validation process, two parameters emerged as strongly influencing the system: in particular, the solid phase dynamic viscosity resulted to heavily influence the slurry pressure drops through the pipe, while the solid phase concentration profile resulted to be highly affected by the Turbulence Dispersion Coefficient. Both parameters have been subjected to a tuning process, finally determining correlations for the selection of their optimal value in order to obtain results in good agreement with the experimental data. The obtained correlations have been then used to determine the solid phase dynamic viscosity values to be set for the simulation of a Pelton nozzle operating in high sediments content conditions for different opening configurations. The three phases flow has been simulated using the previously validated CFD model and the resulting injector performances have been compared with the results obtained in pure water conditions. The comparison gave the possibility to observe a general decrease in the discharge coefficient in sediments – laden flow conditions, highlighting the presence of sediments to cause higher pressure losses compared to the corresponding pure water configurations. In addition, several scalar distributions deriving from the Eulerian simulations have been qualitatively compared with real erosion patterns reported in the literature for the Pelton nozzle case, finding good correlations between the sediments wall shear stress distributions and the areas usually subjected to erosion damages. This analysis made it possible to take a first step toward the proposal and creation of a new erosion model suitable for the prediction of erosion patterns in combination with CFD techniques based on the application of Eulerian multiphase approaches.