Richard Stevens, Dennice Gayme, Charles Meneveau
Department of Mechanical Engineering, Environment, Energy,
Sustainability and Health Institute (E2SHI) & Center for Environmental and
Applied Fluid Mechanics (CEAFM), Johns Hopkins University, Baltimore, MD
Background:
A known drawback of existing fluid mechanics models commonly employed for wind farm design is that they tend to underestimate the strength of interactions due to turbine wake effects, especially in the case of large wind farms. This underestimation can lead to so-called “wind farm under-performance”, which refers to a common phenomenon in which wind farm installations generate less average power than preconstruction predictions. An important aspect missing in traditional wake models is the coupling between the wind farm and the atmospheric flow in the so-called atmospheric boundary layer. Such a coupling is particularly significant for large offshore wind farms because they occupy a significant fraction of the boundary layer height. This modeling challenge has motivated our research into improving the models of both turbulent wakes behind wind turbines and their interactions with the atmospheric winds. New physical insights have been deduced from a suite of large-eddy simulations (LES) based computer models that resolve the turbulent air motions down to several meters.
We use our in-house, high performance computer research code (the JHU-LES code) that can be run efficiently on the Maryland Advanced Research Computing Center (MARCC) cluster. The LES-generated insights [1] have been used to develop the Coupled Wake Boundary Layer (CWBL) model for wind farms[2]. The CWBL model has been compared to field measurements from the Horns Rev and Nysted off-shore wind-farms operating over a wide range of wind inflow directions, showing excellent results. The model has been used to determine economically optimal wind turbine spacings in off-shore wind farms, taking into account both cabling and maintenance costs[3]. We also use large eddy simulations to elucidate the characteristics of wind velocity fluctuations and variability of wind farm power[4,5]. Such knowledge is important to better deal with the inherently variable and intermittent nature of wind energy.
Research objective:
Develop accurate but practical prediction methods for wind farm power output levels over a wide range of flow conditions, based on high fidelity simulation (LES) methods and first principles. Use such models for wind farm layout optimization and resource usage maximization.
Methods:
Large Eddy Simulations based on actuator disk and actuator line models are performed on various percomputers, such as “Bluecrab” at the Maryland Advanced Research Com-puting Center (MARCC); and analysis of simulation results to determine flow properties and power reduction at turbines across wind farm as function of various turbine placements.
Results:
Based on insights from LES, we have developed the new engineering CWBL model, which iteratively couples Jensen wake model with top-down atmospheric model. The model is based on equating wind speeds at turbine location.
Conclusions:
• We have shown how systematic studies of wind farms based on LES can be used to develop an engineering model (the CWBL wind farm model).
• The CWBL model can be evaluated very efficiently, e.g. on Personal Computers.
• Comparison with data: CWBL model is more accurate than prior models
• Such models can be used to find optimal layouts$^3$ and spacings between turbines in off-shore wind farms and thus lower the COE (Cost of Energy).
References:
[1] Stevens, R.J.A.M., Gayme, D.F. and Meneveau, C. (2014), “Large Eddy Simulation studies of the effects of alignment and wind farm length”, J. Sust. Renew. Energy 6, 023105.
[2] Stevens, R.J.A.M., Gayme, D.F. and Meneveau, C. (2015), “Coupled wake boundary layer model of wind-farms”, J. Sust. Renew. Energy 7, 023155.
[3] Stevens, R.J.A.M. Hobbs, B.F., Ramos, A. and Meneveau, C. (2017), “Combining economic and fluid dynamic models to determine the optimal spacing in very large wind farms”, Wind Energy 20, 465-477.
[4] Stevens, R.J.A.M. and Meneveau, C. (2014), “Temporal structure of aggregate power fluctuations in large-eddy simulations of extended wind-farm”, J. Sust. Renew. Energy 6, 0431002.
[5] Wilczek, M., Stevens, R.J.A.M. and Meneveau, C. (2015), “Spatio-temporal spectra in the logarithmic layer of wall turbulence: large-eddy simulations and simple models”, J. Fluid Mech. 769, R1.