Wind Turbine Wake Modeling
Accurate prediction of wind turbine wakes in the atmosphere is one of the main challenges in state-of-the-art wind energy research. It is well accepted that the interaction of wind turbine wakes with downstream turbines within a wind turbine array (or wind farm) leads to decreased array efficiency and increased blade fatigue. – Our group investigates this difficult problem by modeling the wakes of wind turbines in an Atmospheric Boundary Layer (ABL).
In our research efforts, we are using Actuator-Line Modeling (ALM), which represents wind turbine blades as distributed momentum sources in the governing fluid-dynamics equations, i.e. Reynolds-Averaged Navier-Stokes (RANS) or Large-Eddy Simulation (LES). We are using the OpenFOAM framework for our simulations. We have also identified some shortcomings in current ALM w.r.t. the accuracy at which the blade loads are modeled by the distributed momentum sources.
Our group interacts closely with researchers at the National Wind Technology Center (NWTC) to investigate potential improvements to ALM towards a better prediction of sectional blade loads. Furthermore, we are very interested in the effect of atmospheric turbulence on blade loads, the physics of the wake momentum deficit, and the power production of an array of wind turbines. Our developments are an integral part of the Cyber Wind Facility project at Penn State. Of particular interest is the interaction of wind turbines with turbulent eddies in a Moderately Convective Boundary Layer (MCBL). At present, we are also investigating various turbulent statistics extracted from our high-resolution numerical data that will enhance the physical understanding of the interaction of blade tip vortices with turbulent atmospheric structures, turbulent momentum transport into the rotor disk area of downstream wind turbines, and the generation mechanism of wake meandering.
Icing on Wind Turbines
The success and increased demand for renewable energy produced by wind has led to turbine siting and wind project developments in colder climate regions both in Europe and North America. Colder climate regions are attractive to wind energy because of good wind resources combined with an increased air density of >10% compared to hot and dry climates. However, wind turbines can be subject to several atmospheric icing events during a winter season. While atmospheric icing events may only last for a few hours or days, the adverse effects associated with altered blade aerodynamics and decreased power production due to accreted ice shapes can last for weeks after the event. – In collaboration with the Adverse Environment Rotor Test Stand (AERTS, Dr. Jose Palacios) in the Department of Aerospace Engineering at Penn State we were able to perform the first scaled ice accretion experiments on a rotating wind turbine blade [Han et al., JWEIA 109 (2012)]. A combined aerodynamic- and icing scaling methodology was developed to link the unique dataset to actual operating conditions on full-scale wind turbines.
In addition, our group is working on a Turbine Icing Operation Control System (TIOCS), a coupled methodology of ice accretion modeling, iced airfoil analysis, and wind turbine performance quantification. The objective of TIOCS is to understand how blade aerodynamics changes during an atmospheric icing event, and, more importantly, what strategies can be devised to i) minimize the amount of ice accretion, and ii) safe operation at controlled loads post event, both by means of altered blade pitch and rotor speed control.
We are using the NASA Lewice code for modeling ice accretion in combination/comparison with experimental data obtained in AERTS. Alterations in airfoil characteristics due to accreted ice shapes is quantified by in-house Computational Fluid Dynamics (CFD) simulations as well as correlation tables available in the literature. Our group’s XTurb-PSU code is used for wind turbine performance analysis.
Wind Turbine Blade Design and Optimization
Our group is also working on several projects related to wind turbine blade design and optimization. The wind turbine design and analysis code XTurb-PSU is being actively developed in our group and used primarily for research and class projects in a graduate course on “Wind Turbine Aerodynamics”.
Active developments include a novel stall-delay model and a blade optimization methodology that couples the XTurb-PSU code with an evolutionary strategy based on Covariance Matrix Adaptation (CMA-ES). As opposed to other common genetic algorithms that require tuning of various parameters, the only free parameter in CMA-ES/XTurb-PSU is the population size. We have successfully demonstrated the capability of XTurb-PSU/CMA-ES by designing wind turbine blades under a variety of constraints, e.g. maximum Annual Energy Production (AEP) and minimum root flap bending moment. Further developments and applications are underway.