Upload
trinhphuc
View
217
Download
3
Embed Size (px)
Citation preview
Universitetet i Stavanger
uis.no
Ingrid K. Feyling, Siri M. Kalvig & Knut Erik Giljarhus
The Fifth Symposium on OpenFOAM in
Wind Energy (SOWE 2017)
Research Network for Sustainable Energy at UiS & IRIS
26-28 April 2017
Pamplona, Spain
Overview
Ingrid K. Feyling 2
Introduction
Motivation
Wave-wind interactions
Previous work
Methods
Preliminary comparison
Conclusions and future work
Introduction
5
• The wave’s effect on the wind profiles usually ignored in offshore wind turbine engineering
• Recent research has shown that the MABL is influenced by waves in much higher levels than
previously thought (e.g. S. Kalvig, 2014)
• Waves will influence the power production as well as turbulence levels and wave decay
Challenge:
How can we model the effect the waves have on the wind profile? And
will the wave effect be significant for the performance of a wind
turbine?
Wind wave interaction investigated with two different CFD methods for
wave representation - Solid surface and moving mesh vs. VOF
Ingrid K. Feyling
6
Motivation
Need a better link between
wave models and atmospheric
models
Offshore wind energy – great
potential! Høg-jæren vindpark, photo:
Norsk Vindenergi
0
20
40
60
80
100
0 5 10 15
Heig
ht
(m)
Mean wind speed (m/s)
Photo: Lene Eliassen
Atmospheric stratification and wave
effects are two major factors affecting
wind conditions over sea versus land
Wind and wave
misalignment is quite
common
S.A Hassan (2017)Earth Nullschool
Ingrid K. Feyling
Wave-wind interactions
Wind sea and swell influences the atmosphere different
Wind sea - waves generated by local wind
Swell - long period waves generated by distant storms
Ingrid K. Feyling 8
The marine atmospheric boundary layer (MABL): part of the atmosphere that
is directly influenced by the ocean
One of the major problems in understanding the dynamics of the wind in the
surface layer is the difficulty to get experimental data at spatial scales from
few meters to few kilometers
Parameterization of the wind-waves interaction: complex challenge due
to time-evolving wave fields CFD simulations and comparison to
experimental measurements to evaluate the wave effect on the MABL
Marine Atmospheric Boundary Layer (MABL)
9
Previous work
Grand Valley State University
Need a new boundary condition that take into account the
sinusoidal movement of the “ground”
Kalvig & Manger (2014)
Waves + Actuator Line (SOWFA) FAST
WIWiTS
R.Kverneland (2012)
VOF method:
volume fraction of a
fluid occupying each
element in the
computational domain
is defined by F where
(0 ≤ F ≤ 1)
10
OpenFOAM case set-up:
Boundary Layer test case: pimpleFoam
Volume of Fluid:
interFoam:air/water
waves2foam toolbox
(developed by N.G Jacobsen et al. 2012)
Solid surface and moving mesh:PimpleDyWFoam solver
(developed by Kalvig & Manger 2014)
Methods
Ingrid K. Feyling
Ingrid K. Feyling 11
Based on OpenFOAM solver: pimpleDyMFoam
Wind: ABL inflow
Waves: Stoke’s 1st order (Airy)
PimpleDyWFoam solver (1)
Approximations used:
Newtonian fluid and incompressible flow
Coriolis force and buoyancy not included
Turbulent closure: k-ε (Reynold’s stresses proportional to the mean rates of deformation)
Water depth: set to ‘infinite’ to exclude the effects from the sea bottom
Waves seen as a solid – no deformation due to the wind (!)
“moving mesh”
PimpleDyWFoam solver (2)
𝝽 𝑥, 𝑡 = 𝑎 𝒂𝒏 𝑠𝑖𝑛 2𝜋(𝑥−𝑐𝑡
𝜆) + 𝒘𝒏 𝑐𝑜𝑠 2𝜋(
𝑥−𝑐𝑡
𝜆)
𝝽 is the total wave surface displacement, 𝒂𝒏 and 𝒘𝒏 are unit vectors, x is the horizontal position at a
given time t, a is the wave amplitude, λ is the wavelength and c is the wave speed
Ingrid K. Feyling 12
Ingrid K. Feyling
Results applying this showed that the wind field is influenced by the wave high above the wave boundary layer (WBL)
PimpleDyWFoam solver (3)
The horizontal component
of the wind speed (upper),
the vertical component of
the wind speed, (middle)
and the turbulent kinetic
energy (lower) over a wave
with:
A=4 m, λ =70 m, c=10.5 m/s
Domain size: 450 m x 400 (close up)
R. Kverneland (2012)
Expanded to pimpleDyWTurbineFoam for
inclusion of a wind turbine in 3D
domain
13
14
Developed by Jacobsen et al. (2012)
Uniform wind inflow
Wave theory library included
Relaxation zone technique
Pre- and post-processing utilities tailored for use for free surface flows
Waves2foam
waveFoam: modified waveFlume case
waveProperties input file: wind and wave definition
Ingrid K. Feyling
Ingrid K. Feyling 15
Volume of Fluid: waves2foam Airy wave, no wind applied, c = 10.5 m/s, λ = 70 m, and a = 2 m
Preliminary Comparison (1)
Ingrid K. Feyling 16
Volume of Fluid: waves2foam Airy wave, uniform wind 1 m/s at inlet, c = 10.5 m/s, λ =70 m and a = 2 m
Preliminary Comparison (2)
Ingrid K. Feyling 17
Volume of Fluid: waves2foam Airy wave, uniform wind 10 m/s at inlet, c = 10.5 m/s, λ = 70 m and a = 2 m
Preliminary Comparison (3)
Ingrid K. Feyling 18
Solid surface & moving mesh: pimpleDyWFoam Airy wave, uniform wind 10m/s at inlet, c = 10.5 m/s, λ = 70 m and a = 2 m
Preliminary Comparison (4)
200 m
t=100 s
Fading wave zonePreliminary wind profiles
Preliminary Comparison (5)
Solid surface & moving mesh: pimpleDyWFoam Airy wave
Logarithmic wind at inlet: U400m = 8 m/s and z0=0.0002 m
Domain: 1200m x 25 m x 400 m
Profiles are sampled from the middle of the domain (x=600 m) for every second between 251-300 seconds of simulations
Wave with a = 4 m, λ = 50 m, c = 8.8 m/s Wave with a = 4 m, λ = 100 m, c = 12.5 m/s
Wind velocity profile sampled from
different cases:
(black) wind without waves
(blue) wind aligned with the waves
(red) wind opposed to the waves
(S.Kalvig, 2014)
Ingrid K. Feyling 20
The flow response over the waves is very different compared with flow
over a flat sea surface
Wave direction relative to wind direction important
The wind speed profile and the turbulent kinetic energy pattern far
above the waves will be different depending on the wave state and wave
direction
Both solid wall/moving mesh and VOF method for wave induced wind
simulations works well when dealing with swell conditions
….However, more testing and validation required
Conclusions
Ingrid K. Feyling 21
Implement atmospheric stability
More wind and wave conditions
Opposing and misaligned wind/wave cases
Investigate parametric relationship between wind and waves (!)
Future Work
Overall:
Improve the understanding of the marine atmospheric boundary layer (MABL) and its turbulence
level in the presence of waves of different orientation and magnitude
Improve the understanding of the influence of wave affected MABL on power production and
dynamic loads
Improve the understanding of the influence of wave affected MABL on wind turbine wake
Definition of wave and wind relative orientation
E-mail: [email protected] 22
[1] https://www.fastcompany.com/3059959/scotland-will-be-home-to-the-worlds-largest-wind-floating-wind-farm
[2] www.uis.no/miljoenergi
[3] S. Kalvig, “On wave-wind interactions and implications for offshore wind turbines” (2014)
4] S. Kalvig, E. Manger, B. Hjertager and J. Jakobsen, "Wave influenced wind and the effect on offshore wind turbine performance, Energy Procedia,
vol. 53, pp. 202-213, 2014
[5] S. Kalvig, E. Manger & R.Kverneland “A method for wave driven wind simulations with CFD” (2013)
[6] P. P. Sullivan, J. C. McWilliams and E. G. Patton, "Large-eddy simulation of marine atmospheric boundary layers above a spectrum of moving waves" (2014)
[7] N.G. Jacobsen, D.R. Fuhrman, J. Fredsøe, “ A wave generation toolbox for the open-source CFD library, OpenFOAM” (2012)
[8] R. Kverneland , “CFD – Simulations of wave-wind interaction” (2012)
[9] S.A. Hassan, UiS MSc .Thesis, unpublished (2017)
[10] K.Okamori, «Fluid Simulation Analysis of Multi-phase Flows by Oka-san 3, VOF method” (MSc. Course)
References
Acknowledgement: Statoil Academia & Research Network for Sustainable Energy at UiS & IRIS