Morten

Hi Lino, The IEC61400-12-1 standard states that the ideal distance between the reference mast and the tested turbine is 2.5 times the rotor diameter, 2.5D. I believe this is a compromise between good correlation between the free wind at mast and turbine site and reasonably low effect of flow blockage by the operating turbine. The standard also specify the allowable range to be 2-4D, so unless your turbine is very big, 600m will be a little too far. Sorry for the delay, by mistake I did not subscribe to this topic. Cheers, Morten

Dear Bepi, The previous question by HPJ concerned flow inclination angles by the WAsP Engineering program. To see a map for a specific wind direction you click Insert> New wind from the WAsP Engineering main menu, select Maps and sites> Wind Grid maps, right-click Flow inclination grid and select Open in new spatial view from the popup menu. WAsP Engineering will display maps of individual velocity components in a similar way. Flow inclination angles are included as fields in the WAsP resource grids. These results are only valid when using the WAsP CFD flow model, not the standard WAsP flow model called IBZ. I think the result for all directions still are calculated as the worst local inclination angle for any wind direction, as defined in previous editions of the IEC 61400-1 standard. To see the inclination angles defined in the current IEC 61400-1 Edition 4 - which is an energy-weighted average over all wind directions - you can either use tools> WAsP CFD results viewer from the WAsP main menu or the Windfarm Assessment Tool. Chapter 11.2 of the IEC 61400-1 standard presents a method involving planes fitted to the terrain elevation surface. The purpose of this is just to assess the terrain complexity, not flow inclination angles. This assessment is implemented in the WAsP CFD results viewer and the Windfarm Assessment Tool, not directly in WAsP or WAsP Engineering. Cheers, Morten

I can think of two related problems: The recommendation for WAsP AEP predictions, before the windfarm construction, is to select a reference mast position, which is as representative as possible, e.g. with a ruggedness index close to the average RIX number of the turbines. Another problem is verification of power performance after wind farm construction, where the IEC 61400-12-1 specifies a procedure for power curve verification. For this, the manufacturer and wind farm operator agrees on test turbine and mast position, where the terrain is not too complex, and only applies data from a sector where measurement are not disturbed by wake effects. WAT can help with the IEC 61400-12-1 terrain assessments. However, your application seems different. I am not sure that I understand what you mean by calibrating the wind speed, but you probably need a reliable wind signal for a wind farm in operation, perhaps with curtailments and other complicating factors. I do not think that there is a standard for this, and the solution may on the application. Things to consider: Are there significant terrain effects? Is it possible to estimate speedup effects by flow model or measurements? Should you avoid wake effects, e.g. by installing multiple reference masts for winds from different directions?

It you deactivate a turbine site in the main WAT object hierarchy, it is treated as if it does not exist in the WF layout. You will also find some check boxes in the additional tree view next to the WF power curve. The production form deselected turbine sites does not contribute directly to the WF power curve, but there is an indirect effect since their wakes affect the selected turbines. You can use it if you want to know the AEP a sub-group of turbines, while other turbine groups are still in operation. The thrust-coefficient curve used in WAsP and WAT has a so-called static thrust coefficient (see the WAsP turbine editor), which is applied outside the wind-speed range of turbine operation, i.e. where the rotor is parked or idling. It is usually quite small and should not affect the AEP integrated over all wind conditions much.

WAT tries to simplify the user interface by hiding options which makes no difference for the current. The options for Performance measurement sectors is a good example of this. as it is only shown when you select at turbine site with an site calibration mast as child object.

I will not claim that WAT has an integrated reporting tool, but you can select a plot or table and copy this to the Windows clipboard via Ctrl+C on the keyboard and then paste it into Excel or Word. This will copy what you currently see on the screen, so you may have to repeat with different modelling options. You can also export tables and plots via the tools under edit>export from the main menu, and some of these will iterate over turbine sites. When implementing this, I imagined that the user was writing a report for a client and just needed to insert key results from WAT. I did not attempt to automate a complete report.

The limit on the length of the local time series is exactly three years. I am not sure why the developers introduced this unnecessary limitation - maybe they were simply proud that the method could make realistic 50-year extreme wind estimates from short time series. The spectral correction method actually applies a 21 year long time series of CFDDA reanalysis data with correction by a short time time series of local observations. The idea is that the local data will add fast variations missing in the CFDDA data. The corrected series (or actually spectrum) will then provide more realistic extreme values. We do not recommend to substitute local observations by mesoscale data as they typically have too little high frequency variation.

You are right. Sometimes the software enables you to calculate EWC and U50 estimates even with sparse data and you wonder about the accuracy. Both annual-maximum and peak-over-threshold methods have associated uncertainty estimates. In WAsP Engineering, these uncertainties are indicated by the curves around the extreme wind versus return period plot. This statistical uncertainty depends on sample size and Gumbel alpha parameter. However, there is even uncertainty on the uncertainty estimate itself, so take care with very small samples. The IEC 61400-1 standard includes an annex on measure-correlate-predict methods based on correlation with a reference station. It also discusses extreme-wind estimates based on the MCP long-term-corrected series, and it recommends at least seven years for the reference station. For extreme winds in tropical storms, it recommends methods based on satellite-tracking. I think you need something similar if you only have local data, but it also depends on the uncertainty you are willing to accept. In WAsP Engineering, I recommend to use the spectral correction method, whenever possible. This combines local information with a database using 21 years of modelled data with a short time series of local observations, e.g. only one year long. Unfortunately, the database does not yet cover Australia.

Hi Cristi, I contacted our programmer who replied: Hi Morten, your advice on the forum is correct. There is nothing we can do to make it work, as far as I know. The drivers for the dongles were provided by the company who made the hardware dongles and there is no upgrade for modern Windows OS. Sorry. Best regards, Morten

Hi Cristi, I think the license system of WAsP 10 was the old one based on hardware dongles, and my guess is that the software drivers of those dongle are incompatible with Windows 11. We had reports from Windows 10 users with similar problems. Unfortunately, we no longer support WAsP 10, so my advice is to upgrade to the present WAsP version. Best regards, Morten

In the context of fitting a Weibull distribution to an empirical distribution, I would call the wind speeds above the mean wind speed the high wind speed range. Try to open an observed wind climate in WAsP or WACA and compare the two distributions for high wind speeds and for low ones, and I think you will see what I mean. I do not recommend any correction to the model. We should remember that the AEP is found by an integral of the mean wind speed probability distribution and power curve, so we are not concerned about PDF model errors at wind speeds with zero or very low production.

In WAsP, the WTG object refer to the wind turbine generator, specifying power and thrust-coefficient curves, so I guess that you are actually referring to a turbine site placed at the met mast position. In this way you are making a self-prediction test, and wonder about discrepancies in the predicted mean wind speed. There are many ways to fit a Weibull distribution and in WAsP the fit is designed to match the mean of the cube of the observed wind speeds and the probability of winds higher than the empirical mean wind speed. This usually results in a good fit to data in the high-wind range but a less accurate fit at lower wind speeds. This is a deliberate choice, as wind power production mostly depends on accurate modelling in the high wind-speed range. However, unless the observed wind speed distribution is a perfect Weibull distribution, there is no guarantee that the fitted distribution will match the mean wind speed of observation. This is the main reason for the discrepancy. In addition, there can be small errors in sector-wise frequencies and Weibull distributions, due to a rotation of the wind rose when converting the observed wind climate to the generalized wind climate and a reverse rotation back to a (self-)predicted wind climate. These rotations shifts probability mass between sectors, and the finite sector width leads to discretization errors. This type of error increases for wind climates with a large variation between neighboring sectors.

WAsP applies the wind atlas method. The traditional input is 10-min time series of wind speed and direction measured by a Lidar or anemometer on a mast near the site of application. First step in the analysis is to remove effects of local terrain around the reference meteorological station and produce mean wind climate statistics for the resulting generalized winds, i.e. winds over flat uniform terrain. At the site of application, WAsP will model terrain effects near every wind turbine site and predict local mean wind climates based on speed-up factors and the generalized mean wind climate. Now, reanalysis data are predicted by global circulation models with simplified terrain models capturing only large-scale flow effects. To use them in a WAsP context we need to calculate the generalized wind climate by a method we call downscaling, see https://orbit.dtu.dk/en/publications/mesoscale-and-microscale-downscaling-for-the-wind-atlas-of-south-. We have not done so with the MERRA2 data as input, but the Global Wind Atlas made a downscaling with the WRF (mesoscale) and WAsP (microscale) flow models based on the ECWMF ERA5 2008-2017 reanalysis dataset, see https://globalwindatlas.info/about/method for a quick read. It is possible to download the GWA generalized wind climates from https://globalwindatlas.info/ and use them in WAsP for preliminary wind resource estimates when local data are unavailable. We still recommend local measurements for bankable results.

Hi Ines, The profile between lower and upper tip height is what matters for site assessment, so I would focus on heights in that range. It is also important to use a selection of data with simultaneous measurements from all the heights used for the profile. You may want to skip a height if it has lots of missing observations. As I remember, WindPro can plot the fitted profile and the reference points together. This is a good quality check and might help you deciding which heights to include. Best regards, Morten

Hi Ines, There are lots of internal processing in a lidar, which results in time series of horizontal speed and direction for different heights. Presumably, the data are also screened for measurement errors, which are reported by status codes. If lots of data are missing for some heights, e.g. at the top, you might want to disregard those heights. Basically, you can only measure the wind component along the laser beam, so a single lidar deduces the horizontal wind by several measurements along slanted beams with different azimuth angles. This deduction usually rely on an assumption of zero vertical wind velocity, so results for the lowest heights might be compromised in complex terrain. It is possible to corrects for effects of non-zero wind velocity by a flow model, as I think we mention this in the WAsP Engineering course. I have heard that some manufactures offers this correction as an optional service inside instrument. Basically it is a correction table, which can be calculated based on the orientation of the instrument and the terrain around the measurement position. Hope that this helps a little. We have a large group of lidar specialist here at DTU Wind Energy working on advanced techniques, like multiple coordinated lidars, dedicated scanning patterns and new data processing for turbulence measurements. I must confess that compared to them, I know very little on the topic. With best regards, Morten

Hi Inés, I just remembered one more reservation: WAsP applies the geostrophic drag law for conversions between actual and generalized winds. This law works well for wind climate statistics, but you often see deviations from the geostrophic balance when considering 10-min measurement periods and the site of prediction is not close to the reference site BR Morten

Hi Inés, We know that time series of power production has many applications, but we have not yet found a way to support this in WAsP. WAsP is basically designed to work with climate statistics and predict annual energy productions. One might use the flow corrections of WAsP to translate wind conditions for each record in a reference time series to individual turbine sites, and I think our partners at WindPro offers a module, which follow this approach. There are however some general concerns: WAsP operates with wind sector statistics, e.g. effects of roughness changes are modelled as average corrections for each wind sector. Real terrain might be more complicated. The background wind profile in WAsP includes the average effect of atmospheric stability, but not time-dependent variation of atmospheric stability. Wind conditions measured over 10-min periods at reference mast and site of prediction are not perfectly correlated for individual records in the time series. This de-correlation is expected to depend on the distance from the reference site, terrain, wind speed, turbulence, mesoscale convection, etc. Realistic time series simulation may have to include a stochastic element to model these effects. The standard WAsP wake model does not include effects of variable atmospheric stability conditions affecting the wake development. The model also ignores wake meandering causing fluctuation in the wake loss. These concerns may not be relevant for all applications, e.g. you might consider the above-mentioned effects as acceptable uncertainties. Best regards, Morten

Hi Erkan, Both of the attached WindPRO error messages say “project class failed to create a new project instance” which indicates a problem with the installation. I also see a button called “show bug report” and I guess that you will get detailed information when you press that. If that does not help you, I suggest that you contact WindPRO support. If I read your question correctly, you were able to calculate with WEng 4.0, but now are having problems going back to WEng 3.1. I don't know whether the two versions are supposed to co-exist under a WindPRO installation. If you were running our programs directly, I would recommend that you switched to WEng 4.0. Actually, we no longer support WEng 3.1 (latest version is from 2015) so I no longer have it on my PC. Best regards, Morten

Hi Stefan, I remember one detail which used to trouble me with this script: the Lidar and reference masts sites must be given the exact names ‘L’ and ‘M’, so maybe check this first. Apart from this, I have no idea of what might be wrong. As Mark suggest, you better submit a WEng project file illustrating the problem to WAsP support together with possible additional inputs (if required, I cannot remember) for the script. Cheers, Morten

Hi Pedro, I have no personal experience with this, but I believe that roughness values to the left and right side of each polygon or polyline in the SHP file should be stored in an a separate DBF file – with the same name and in the same folder, just using different filename extensions. You can read about this in the Map Editor help file section called ‘Map Formats> Vector maps> Shape files’. I am not sure exactly how you generate the DBF file with ArcGIS as I have never used this tool myself. With best regards, Morten Nielsen

Hi Paul, You are right, the background flow is not irrotational, but I just think Eqn. 14 states that the flow field has no gradient. This is true for the two horizontal dimensions, since the background flow is the same everywhere. It is also true for the vertical direction, since its vertical component is zero. Report Risø-R-900 focussed on roughness-change perturbations. The model for effects of variable terrain elevation was revised in Risø-R-1356. Cheers, Morten References: https://backend.orbit.dtu.dk/ws/portalfiles/portal/7766601/RIS_R_900.pdf https://backend.orbit.dtu.dk/ws/portalfiles/portal/7726842/ris_r_1356.pdf

Alternative power curves could reflects different ambient conditions, like air density, and modes of turbine operation, like reduced tip-speed as a means for noise reduction. The changes are made by the wind turbine controller managing tip speed and blade pitch angle, so yes, you are right, this should both affect power curve and thrust-coefficient curve. However, we normally don’t know the exact strategy of the control system or the detailed rotor aerodynamics, so it is difficult to predict changes in the thrust just from changes in the power. Some manufactures supply sets of power- and thrust-coefficient curves for different conditions, and then you type both curves into the WAsP turbine editor. However, I have also noticed manufactures who supply the same thrust-coefficient curve for different power curves. I think the argument is that the power curve is the most important one, as it influences the AEP more directly. The usually wake loss is only 5-10%, so a modified thrust-coefficient curve will not have as significant an effect on the AEP as a modified power curve. Cheers, Morten

Hi Windfrosch, It is a standard deviation of the extreme wind estimate associated with the uncertainty of the fitted statistical model for the extreme-wind distribution. There are actually alternative models, see the post at http://www.wasptechnical.dk/forum/viewtopic.php?id=1060 When running the script you mention WEng will use the model defined by the selected extreme-wind-climate object, either annual maximum (AM), peak over threshold (POT) or spectral correction method (SC). Cheers, Morten

Dear Sinisa Knezevic, 1) The extreme wind in WEng may be calculated by three alternative methods called the annual maximum (AM), peak-over-threshold (POT) and spectral correction (SC) methods. The theory differ for the three methods, but to generalize we could say that they first fit a statistical model for extreme events and then used that model to estimate the level of the extreme event with a fifty year return period. The uncertainty of the 50-year extreme wind estimate depends on the uncertainty of the fitted model. Read more in the WEng help file section 'WAsP Engineering modelling| Extreme winds', specifically in the subsections called 'Extreme wind estimators' for the AM method, 'Extreme wind POT method' for the POT method and 'Spectral correction' for the SC method. 2) The footnote in IEC 61400-1 Ed.4 expresses the coefficient of variation (COV) by the alpha and beta parameters of the Gumbel distributions. Thus, COV depends on the fitted model and not directly on the uncertainty of the fit. However, the uncertainty is closely related to the Gumbel alpha parameter, so you could say that there is an indirect relation. Unfortunately, WEng no longer reports the fitted alpha and beta parameters, but if you look in the 'WAsP Engineering modelling| Extreme winds| Extreme wind estimators' section of the WEng help file you will find an easy way to calculate them by yearly maxima exactly as in WEng when using the AM method. Alternatively, you can read the 100-year event from the graphs in WEng showing extreme winds as function of return period and then use the formula in the footnote of the IEC standard. With best regards, Morten

Hi Windfrosch, WAsP does not support this directly, but as a work-around method you could try a modified power curve for the turbines you wish to exclude from the windfarm power curve. This modified power curve must have the usual Ct-coefficient curve but a close-to-zero power curve. Ideally that should be exactly zero power for all wind speeds, but then WAsP would fail to import the file, so better use tiny values instead. The easiest way to modify the turbine power curve is to - export to a WTG file, - open the WTG file in the WAsP turbine editor, - export to the old POW format - open the POW file in an ASCII editor - reduce the power scaling parameter (second parameter, third line), say from 1000 to 0.0001 - save the modified POW file You now have a model for an incredibly inefficient turbine, producing almost no power but with the usual wake loss at the other turbines. Now go back to the WAsP project, insert the modified POW file in the subgroup representing turbines you want to exclude from the windfarm power curve and recalculate. As a check you might compare the NetAEP of the turbines of interest to their NetAEP using the original WTG for all turbines. I got small deviations on the last digit, but hope that is OK. Finally, when you have the special-purpose wind farm power curve, you should revert to the original WTG file for all turbines. Cheers, Morten