Wind Energy Modeling and Simulation: Atmosphere and plant

Cover
Contents
Disclaimer
Preface
List of acronyms
1 Looking forward: the promise and challenge of exascale computing
1.1 Introduction
1.1.1 Future wind plant technology
1.1.2 Physical scales driving HFM and HPC
1.1.3 Turbine technology changes requiring HFM and HPC
1.1.4 Wind plant performance
1.2 Mathematical and numerical modelling pathways
1.3 Challenges at petascale and the need for exascale
1.4 The challenge of exascale computing
1.5 Concluding remarks
Acknowledgements
References
2 Blade-resolved modeling with fluid–structure interaction
2.1 The extraordinary range of length and time scales relevant to wind-turbine operation
2.1.1 Impacts of atmospheric "microscale" turbulence
2.1.2 The rotor and blade-boundary-layer response length and time scales
2.1.3 The wake response length and time scales
2.1.4 Influences from atmospheric mesoscales and related weather events
2.1.5 Concluding discussion
2.2 Essential numerical and modeling elements in blade-resolved simulation of wind turbines
2.2.1 CAD model and mesh generation
2.2.2 CFD solver
2.2.2.1 Incompressible-flow solvers
2.2.2.2 Compressible-flow solvers
2.2.3 Turbulence modeling
2.2.4 Fluid–structure interaction
2.3 Practical issues in performing bladeboundary-layer-resolved simulations
2.3.1 Mesh generation
2.3.2 Mesh quality
2.3.3 Convergence and time step
2.3.4 Verification
2.3.5 Validation
2.4 Conclusions and challenges for future advancement in the state-of-the-art
Acknowledgments
References
3 Mesoscale modeling of the atmosphere
3.1 Introduction to meteorology for wind energy modeling
3.1.1 Forces and the general circulation of the atmosphere
3.1.2 Scales and phenomena in the atmosphere
3.1.3 Atmospheric energetics
3.1.4 The chaotic nature of atmospheric flow
3.2 Basics of atmospheric modeling
3.2.1 Historical perspective
3.2.2 Governing equations for flows in the atmosphere
3.2.3 Numerical resolution requirements
3.2.4 Reynolds averaged Navier–Stokes simulation methodology
3.2.5 Discretizations
3.2.6 Forcing physics and parameterizations
3.3 Initial conditions and data assimilation
3.3.1 Nudging
3.3.2 Variational DA
3.3.3 Ensemble Kalman filters
3.3.4 EnVar and hybrid DA
3.4 Boundary conditions
3.4.1 Forcing from global models
3.4.2 Top boundary
3.4.3 Bottom boundary
3.4.4 Coupled models
3.5 Using NWP for wind power
3.5.1 Resource assessment
3.5.2 Forecasting
3.5.3 Turbine wake parameterization
3.5.4 Postprocessing
3.5.5 Assessment
3.6 Uncertainty quantification
3.6.1 Quantifying parametric uncertainty
3.6.2 Quantifying structural uncertainty—ensemble methods
3.6.3 Calibrating ensembles
3.6.4 Analog ensembles
3.7 Looking ahead
3.7.1 Storm-scale prediction
3.7.2 Scale-aware models
3.7.3 Blended global/mesoscale models
3.7.4 Seasonal to subseasonal prediction
3.7.5 Regime-dependent corrections
3.8 Summary and conclusions
References
4 Mesoscale to microscale coupling for high-fidelity wind plant simulation
4.1 Introduction
4.1.1 Overview of atmospheric simulation at meso and microscales
4.2 Large-eddy simulation of the atmospheric boundary layer
4.2.1 ABL LES setup
4.2.1.1 Forcing
4.2.1.2 Mesh spacing
4.2.1.3 Turbulence generation
4.2.2 LES assessment
4.2.3 Unsteady conditions
4.2.4 Stable conditions
4.2.4.1 Nocturnal low-level jets
4.2.4.2 Lateral boundary conditions
4.3 Enabling multiscale simulation
4.3.1 Methods to extend the applicability of periodic LES
4.3.2 Coupling LES to mesoscale model output at lateral boundaries
4.3.2.1 Turbulence generation: mesoscale to LES
4.3.2.2 The terra incognita
4.3.2.3 Terrain-following coordinates
4.3.3 Online versus offline coupled simulations
4.3.3.1 Top and bottom boundary conditions
4.4 Additional challenges facing high-fidelity multiscale simulation
4.4.1 LES SFS models
4.4.2 Flow transition at coarse-to-fine LES refinement
4.4.3 Bottom boundary condition
4.4.4 Data assimilation
References
5 Atmospheric turbulence modelling, synthesis, and simulation
5.1 Introduction
5.1.1 Notation and ensemble averaging
5.1.2 Defining the notion of turbulence simulations
5.2 Simulating turbulence for wind turbine applications
5.3 Turbulence in the atmospheric boundary layer
5.3.1 Surface-layer scaling and Monin–Obukhov similarity theory
5.3.2 Above the surface layer: typical wind turbine rotor heights
5.4 Which characteristics of turbulence affect wind turbines?
5.5 Synthetic turbulence and standard industrial approach
5.5.1 Statistical attempts
5.5.2 Standard spectral models
5.5.2.1 Kaimal spectra with exponential coherence model
5.5.2.2 Mann model: rapid-distortion theory with eddy lifetime
5.5.3 Extensions of the spectral-tensor model
5.5.3.1 Turbulence synthesis
5.6 Large eddy simulation
5.6.1 The fundamentals
5.6.2 SGS models
5.6.2.1 Smagorinsky models: first-order and O(1.5) closure
5.6.2.2 Advanced Smagorinsky-type closures
5.6.2.3 Higher order SGS closures
5.6.2.4 Boundary conditions
5.6.3 Numerical approach
5.7 Final remarks
References
6 Modeling and simulation of wind-farm flows
6.1 Introduction
6.2 Why simulate the flow through wind plants?
6.2.1 Improved physical understanding
6.2.2 Design
6.2.3 Wind-farm control
6.2.4 Special cases of interest and forensic analysis
6.2.5 Design of experiments
6.3 Simulation approaches
6.3.1 Noncomputational-fluid-dynamics-based approaches
6.3.1.1 Inflow wind generation
6.3.1.2 Wake modeling
6.3.2 Computational-fluid-dynamics-based approaches
6.3.2.1 Choice of equation set
6.3.2.2 Treatment of turbulence
6.3.2.3 Choice of numerical methods
6.3.2.4 Inflow wind generation
6.3.2.5 Wake generation
6.3.2.6 Putting the components together to use CFD to simulate the wind farm
6.4 Validation efforts
6.5 Future development
Acknowledgment
References
7 Wind-plant-controller design
7.1 Introduction
7.1.1 Structure of the chapter
7.1.2 Current practice in wind farm operation
7.1.3 Degrees of freedom in the wind farm control problem
7.1.4 Objectives of wind farm control
7.1.4.1 Maximization of the farm's annual energy production
7.1.4.2 Minimization of the turbines' structural degradation and fatigue
7.1.4.3 Provision of ancillary services for the electricity grid
7.2 A classification of wind farm control algorithms
7.2.1 Current practice; greedy operation
7.2.2 Open-loop model-based controller synthesis
7.2.3 Closed-loop model-based controller synthesis
7.2.4 Closed-loop model-free controller synthesis
7.3 Control-oriented modeling
7.3.1 Steady-state surrogate models
7.3.2 Control-oriented dynamical surrogate models
7.4 Examples
7.4.1 Steady-state wind farm model: FLORIS
7.4.1.1 Model description
7.4.1.2 Real-time wind farm optimization using FLORIS
7.4.2 Dynamical wind farm model: WFSim
7.4.2.1 Model description
7.4.2.2 Real-time model adaptation for the WFSim model
7.5 Software architecture
7.5.1 Centralized vs. distributed control
7.5.2 Communication with other simulation submodels
7.6 Conclusion
Acknowledgment
References
8 Forecasting wind power production for grid operations
8.1 The role of wind-power forecasting
8.2 Sense: gathering and ingestion of predictive information
8.2.1 Area of influence
8.2.2 Observation targeting
8.3 Model: translating predictive information into a forecast
8.3.1 Physics-based techniques
8.3.2 Statistical approaches
8.3.2.1 Methods
8.3.2.2 Applications
8.3.3 Power output models
8.3.4 Integrated forecast system
8.4 Communicate: inform the user for decision-making
8.4.1 Deterministic versus probabilistic
8.4.2 Time series versus event-based
8.5 Assess: evaluation of forecast performance
References
9 Cost of wind energy modeling
9.1 Introduction
9.2 Levelized cost of energy (LCOE)
9.3 Overview of cost of energy modeling
9.4 Modeling investment costs
9.5 Modeling energy production
9.6 Modeling operational expenditures
9.7 Modeling cost of capital
9.8 Calculating cost of energy
9.9 Estimating future cost of wind energy
9.10 Considering the value of wind energy
9.11 Conclusion
Acknowledgment
References
Index
Back Cover