The Science of Making Torque from Wind (TORQUE 2018)

Authors:
Jason Jonkman, Puala Doubrawa, Jennifer Annoni

Abstract:
FAST.Farm is a new midfidelity multiphysics engineering tool for modeling the power performance and structural loads of wind turbines within a wind farm, including wake and array effects. Previous work performed to calibrate the tuneable model parameters of FAST.Farm has shown that its prediction of wake dynamics for a single wind turbine across different atmospheric stability conditions and nacelle-yaw errors matches well with high-fidelity large-eddy simulation, at a small fraction of the computational expense. This paper presents the validation of FAST.Farm against large-eddy simulation for a series of cases—independent from those used to support the calibration—considering single turbine and small wind farm scenarios, both subject to variations in inflow and control.

Authors:
Bossuyt Juliaan, Charles Meneveau, Johan Meyers

Abstract:
A common challenge for the validation of wind farm simulations with field data is the complex, variable and not precisely defined inflow conditions. In this study, we make use of the first published wind tunnel data of a scaled farm with 100 instrumented porous disk models, and for highly controlled wind conditions, to validate a high-fidelity Large Eddy Simulation code. We present a simple methodology to simulate the spatially-developing turbulent boundary layer from the experiment. The method makes it possible to match the vertical profiles of mean velocity and turbulence intensity in front of the farm. The mean-row power from the simulations shows a good agreement with the experimental results. We find that simulating the spatially-developing boundary layer in a channel with the same height as the wind tunnel, instead of the common approach of a fully developed half-channel flow, is essential to match the experimental results. It is concluded that experiments with a naturally developing boundary layer and a clean flow above, without the use of static turbulence grids, provide well-defined conditions that can be simulated accurately in Large Eddy Simulations and are useful for the validation of wind farm simulations.

Authors:
Chengyu Wang

Abstract:
The aim of this paper is to validate large-eddy simulations (LES) of scaled wind turbines operating in a boundary layer wind tunnel. The final goal of this effort is to develop a validated digital copy of the experimental facility, in support of wind farm control research. The numerical simulations are compared with experiments where scaled wind turbine models are arranged in different waked configurations and yaw misalignment conditions. In the experiments, the wind turbine response is measured in terms of various operational parameters, while the flow is measured with two scanning Lidars. The wind turbine blades are modelled with the aeroelastic code FAST, while nacelle and tower are modelled with an immersed boundary method. Both in the experiments and the simulations, the wind turbines are controlled by the same collective pitch and torque controller. The wind tunnel turbulent flow is obtained with a precursor simulation that models the tunnel inlet and the turbulence generating devices placed downstream from it. Simulation and experimental results are compared in terms of flow characteristics, turbine states and wake behaviour.

Authors:
Kisorthman Vimalakanthan, Marco Caboni, Gerard Schepers, Evgeniy Pechenik, Paul Williams

Abstract:
Besides classical horizontal and vertical axis wind turbines, alternative ways to harvest wind energy are currently being explored. One promising concept is represented by airborne wind energy systems. Different airborne wind energy concepts have been ideated and investigated over the past few decades. The work reported here focuses on the concept being developed by Ampyx Power that basically generates power using a tethered airplane which drives a generator on the ground. The aim of the work reported here is to develop and compare design calculations for such power plane. In this study a comparison was made between steady and unsteady RANS CFD calculations from SU2 and OpenFOAM. Subsequently, using the data from the CFD calculation two different power prediction models were developed and compared, one using 2D CFD data with the free vortex code AWSM and the other using the same data with a nonlinear lifting line approach to model the power plane in its figure eight flying trajectory. Overall comparison show that for all angles of attack at different flap angles 2D CFD results agree extremely well between both solvers. The 3D CFD results comparison showed <20% differences on the computed total forces, despite achieving good agreement on the relative force contributions across the power plane. A very good agreement (5%) has been established between the power prediction models.

Authors:
Mone Lunacek, M. Jason Fields, Anna Craig, Joseph Lee, Ryan King

Abstract:
The pre-construction energy generation of a wind farm (P50) is difficult to estimate and evaluate. This paper presents a methodology to measure the    accuracy of the p50 prediction, which we call the Historical Validation Survey (HVS), for several wind farms in the continental United States.  Our results indicate that there is a significant bias between predicted and measured energy, even when controlling for factors like commission date. We also find that our results depend on the assumptions we make during analysis, which we quantify with a sensitivity analysis. This method allows the estimation of uncertainty we have in our findings. When we account for reasonable ranges of model assumptions, we find that, in the best case, there is still a $-5.5\%$ bias when estimating pre-construction energy generation.

Authors:
Christian Santoni, Edgardo García-Cartagena, Umberto Ciri, Giacomo Valerio Iungo, Stefano Leonardi

Abstract:
Numerical simulations are performed for the study of wake interactions in a real wind farm in the North Texas region. The Weather Research and Forecasting model is used to obtain realistic wind conditions over the wind farm from the mesoscale to microscale through five nested domains. An additional offline domain using a high-fidelity code is solved to obtain a higher accuracy in the wake of the turbines modeled as rotating actuator disks. An assessment of the numerical results of mesoscale simulation showed good agreement to the meteorological and LiDAR field measurements.