Upcoming Events

CFD Models for Wind Farm Calculations

May 9, 2013 3:00PM to 4:00PM
Presenter 
Klaus Vogstad and Vaibhav Bhutoria
Location 
Building 240, Room 1407
Type 
Seminar
Series 
EVS Seminar Series
Abstract:
Wake induced turbulence is one of the most important considerations in design of wind farms. Unfortunately, wake interactions in wind farms have not been fully understood, and a lot of research is devoted to this topic. Moreover, sophisticated CFD methods such as LES are computational expensive and not presently used for wind farm design.

To address these issues we will present:

  1. Turbine wake simulations using OpenFoam CFD: Our participation in the NORCOWE/ NOWITECH blind test.
  2. InstantWind - model reduction for fast CFD computations
  3. Wind farm design - how wind farm layouts are being designed today

Turbine Wake Simulation Using CFD

Wake interaction in wind farms have not been fully understood. We present our contribution to the NORCOWE/NOWITECH blind test on wake models of two turbines. The participants were asked to simulate the wakes of an experimental setup of two turbines in a row (see below picture). We present the results of our OpenFoam simulations with actuator line LES model. The results are compared against wind tunnel measurements. Moreover, the results and approaches of other blind test participants will be discussed. The presentation concludes with the challenges and uncertainties of wake simulation using CFD.

References

Krogstad, P-A and PE Eriksen, "Blind test" Calculations of the performance and wake development for a model wind turbine. Renewable Energy Vol. 50, 2013

Krogstad, P-A et al, "Blind test 2" Calculations for two wind turbines in line (forthcoming)

InstantWind - Model Reduction for Fast CFD Computations

CFD methods for wake representation are too computationally expensive for wind farm layout optimization. We apply model reduction and tiling techniques (POD, SVD) to reduce the order of the CFD model. The model tiles are then coupled together for the whole wind farm. This approach reduces the model order by a factor of 10000, with a loss in precision of about 1%. The model reduction technique shows promising results for the use of detailed CFD models in visualization, optimization and control.

Wind Farm Design - Current Practice W

We present techniques for wind farm design. Markedslabben has developed ParkOptimizer, a wind farm design tool to optimize turbine layouts subject to wind quality constraints using CFD results. The presentation illustrates the optimization problem, and the importance of good representation of turbine wakes.