I have a recently done some calculations where I have been using turbsim to generate wind profiles with different atmospheric stability. However, when I know look at the .sum file they all seem to have neutral profiles. I have used the Richardson number to takecstability into account, and applied the smooth spectra.
Is there something I have done wrong? Or is it the output file .sum that is giving the wrong mean profile.
To maybe make it a bit more obvious what I mean I have attached parts of the .sum file. As you can see the gradient Richardson number is included in the file, but when looking at the mean wind profile it does not seem to be a neutral profile.
Another question is how is the Monin-Obhukov length scale, L, computed from Richardson number ,Ri, by turbSim? According to DNV-RP-C205 the relationship in unstable air should be: L = z/Ri, and for stable air L=z*(1-5*Ri)/Ri. The conditions analyzed in the summary file is stable air, while the relationship between the L and Ri seems to be that of unstable air.
Turbulence Simulation Scaling Parameter Summary:
Turbulence model used = RISO Smooth Terrain
Gradient Richardson number = -3.800
Monin-Obukhov (M-O) z/L parameter = -3.800
Monin-Obukhov (M-O) length scale = -23.829 m
Mean wind speed at hub height = 11.000 m/s
Wind profile type = Logarithmic
Equivalent power law exponent across rotor disk = 0.068
Mean shear across rotor disk = 0.011 (m/s)/m
Assumed rotor diameter = 140.000 m
Surface roughness length = 0.000 m
Number of time steps in the FFT = 36000
Number of time steps output = 28510
For the SMOOTH model, TurbSim calculates the M-O z/L parameter using the relationships
(z/L) = RICH_NO in unstable air (RICH_NO <= 0), and
(z/L) = RICH_NO/(1-5*RICH_NO) in stable air (RICH_NO > 0).
The M-O parameter L is then
L = HubHt/(z/L).
If you do the algebra, that is equivalent to the values in the DNV standard that you stated. However, for numerical reasons, TurbSim does not allow z/L to be greater than 1.
The excerpt of the summary file you included is for unstable air (RICH_NO is negative), and the relationship between the Ri, z/L and L parameters are what I would expect in that case. I’m not sure I understand by your statement about the wind profile seeming to be neutral. Could you clarify what you mean? The profile is calculated using the equation in the TurbSim manual; the psi_M stability parameter used in it is zero for neutral profiles (but it is not monotonic or linear–it has 2 zeros).
Sorry for late response. It was me who had mixed up some numbers in my calculations, I will recheck my calculations a bit better before posting the next time.
I have been looking into the details of how you have included the stability parameter, psi, in TurbSim, and wondered if you have any reference documentation for your equation for psi. I am also wondering why you have chosen to set the parameter ZL to zero for Richardson numbers greater than 0.1667.
I tried to look for the information in the TurbSim guide, but could not find anything specific related to this.
I believe this is documented in
Panofsky, H.A.; Dutton, J.A. (1984). Atmospheric Turbulence: Models and Methods for Engineering Applications. New York: Wiley-Interscience; 397 pp.
Kaimal, J.C; Finnigan, J.J. (1994). Atmospheric Boundary Layer Flows. New York: Oxford University Press.
Businger et al (“Flux-Profile Relationships in the Atmospheric Surface Layer.” Journal of the Atmospheric Sciences, 1971) found the relationship between Richardson Number and z/L for -2<= z/L <= 1. If you look at the equation for stable air, (z/L) is 1 when RICH_NO is 1/6 (approximately 0.1667). TurbSim sets ZL to 1.0 (not 0) for RICH_NO values larger than this. (Also note that this equation goes to infinity as RICH_NO approaches 1/5 so there must be some numerical cutoff point.)