Numerical Simulations of Hydrokinetics in the Roza Canal, Yakima Washington

/, Energy, Renewable Energy, Water Power/Numerical Simulations of Hydrokinetics in the Roza Canal, Yakima Washington

Numerical Simulations of Hydrokinetics in the Roza Canal, Yakima Washington

By | 2016-12-02T18:48:25+00:00 December 3rd, 2014|Computational Modeling & Simulation, Energy, Renewable Energy, Water Power|Comments Off on Numerical Simulations of Hydrokinetics in the Roza Canal, Yakima Washington

Grids used in the refinement study showing the bottom elevation of Roza Canal for models with 19,777, 4,956, and 1,237 cells (left to right).

Grids used in the refinement study showing the bottom elevation of Roza Canal for models with 19,777, 4,956, and 1,237 cells (left to right).

Sandia completed two performance-testing studies for hydrokinetic canal effects at Roza Canal in Yakima, Washington. Sandia National Laboratories’ Environmental Fluid Dynamics Code (SNL-EFDC) was used to model the canal and the current-energy-capture device. The first was an evaluation of different grid sizes and refinements to determine an appropriate grid for future modeling efforts. The second study was a set of parameter sweeps to evaluate the impact of important model parameters on the wake characteristics using the optimum grid determined from the first study.

Vertical velocity profiles directly upstream and downstream of the turbine for αmd = 1. As the partial-blockage coefficient (CPB) increases, flow is forced above and below the turbine and the change in water surface elevation, H, increases.

Vertical velocity profiles directly upstream and downstream of the turbine for αmd = 1. As the partial-blockage coefficient (CPB) increases, flow is forced above and below the turbine and the change in water surface elevation, H, increases.

For the grid-sensitivity study, three grids were generated, all of which use the same bathymetric data for the Roza Canal. Converged results were achieved with the 19,777-cell grid. Because high resolution is not needed throughout the entire model domain for the second study, the 19,777-cell grid was coarsened in portions of the canal away from the turbine location to improve computational efficiency without sacrificing accuracy.

  • Based on results from other models calibrated to scale-flume experiments, the model’s two most sensitive parameters appear to be horizontal momentum diffusion and the partial-blockage coefficient, so a parametric study was undertaken to examine their impacts on wake characteristics. As the partial-blockage coefficient increases, the change in water-surface elevation between upstream and downstream of the turbine increases—reflecting the fact that much of the energy a turbine draws from fluid flows comes from potential energy change across the device.