Home

WindRose User's Guide

1. Temperature MULT Temprature OFFSET C Solar Radiation MULT Time OFFSET sec 600 A sample of the correction coefficients table sheet Input The third table of the Input sheet includes correction coefficients per each file The correction applied has the form a x f and is intended for the processing of measured quantities when l A systematic error occurred during the measurements campaign i e wrong calibration factors additional offset for the zero of the wind vane etc 2 Recorded data have different units temperature in Fahrenheit degrees wind speed in miles h wind direction in radians etc If no values are given then the multipliers and the offsets fare set to 1 0 and 0 0 respectively The following important issues have to be pointed out when applying correction coefficients If the wind speed has to be adjusted by means of the correction coefficients y arx fp then its standard deviation is automatically multiplied by the coefficient a When adjusting wind directions if the resultant direction angle falls out of the 0 360 interval then is automatically adjusted Applying correction coefficients on the time date should be done with caution Although there is no problem using the offset in seconds in order to simulate a wrong day or time setting the multiplier o if z 1 0 can seriously affect the reliability of the measurements as it modifies the sampling rate time step
2. Now come back to the previous screen where the Enable Macros option is not anymore grayed if not close Excel and re open WindRose XLS Then check the box A ways trust macros from this source Click on Enable Macros and start working with WindRose XLS Security Warning H WindRose COM WindRose xls contains macros by Details WindRose Macros may contain viruses It is always safe to disable macros but if the macros are legitimate you might lose some functionality Disable Macros Enable Macros More Info Finally you may check that WindRose 1s among Excel s Trusted Macros Moreover you may re set the Security Level to High Security WindRose v Trust all installed add ins and templates Trust access to Visual Basic Project No virus scanner installed PAGE 54 WINDROSE Wind Data Analysis software REFERENCES l IEC 61400 12 International Electrotechnical Commission Wind Turbine Generator Systems Part 12 Wind turbine Power Performance Testing 1997 Introduction to Wind Energy E H Lysen CWD May 1983 The Netherlands Atmospheric turbulence Models and Methods for Engineering Applications H Panofsky J Dutton John Wiley amp Sons 1984 Structure of Atmospheric Boundary layer Z Sorbjan Practice hall 1989 USA Weibull coefficients estimation D Douvikas Centre for Reneable Energy Sources CRES June 1999 Wind data analysis
3. Calcul of mean turbulence at wind speed m s 10 bin width m s 1 Wind Turbine selected NEG Micon 750 48 v Exponent coefficient a 0 08 Weibull method paper method M low limit m s high limit m s 16 Wind Speed Uncertainty le TOT MET a oT LSY UE O a 0 06 Anemometer s max speed m s 40 Data Logger s precision bits O Site Correlation file name Export processed data file name PAGE 10 The necessary parameters for the wind data analysis sheet Input WINDROSE Wind Data Analysis software The wind turbine for which energy calculations will take place The exponent coefficient of the Power Law which is used to extrapolate the Mast s wind speed at the Hub height of the wind turbine This coefficient is used only when one anemometer is present on the Mast If two or more anemometers are present then this coefficient is calculated analytically per wind speed bin and direction sector more details in SHEAR The given value refers to 5m s Since it is not realistic to consider a constant value for the entire wind speed range a simple model is used to expand it every 5m s this value is divided by two Thus if 0 08 is set then at 10m s is 0 04 at 15m s is 0 02 etc A linear regression is used for the intermediate values and the double of it is considered at Om s 0 16 in the above example The Weibull distribution calculation method In brief 2 methods are proposed for the c
4. 14d Lae 2 9 Dr 00 37 00 12 1 2000 Am ed 146 6 zm ZI Go bn Eo cie 2401010 E 9 13 1 1441 9 2 6 PUR 0057 12 11 0810 i 1245 142 4 2 8 PUN 0107 12 1 2000 o 12 6 140 6 2 5 19 6 OL bey 0 2006 S ip 146 5 e 2159 12700 1201 2000 12 Io i44 il 3 0 Zi 137 00 12012000 Le 2 iL 309 PUN 14297700 12012000 ies 150 2 3 0 20 4 01S 700 12012000 12 58 148 3 2 zu 020700 12012000 WindSpeed Dika Usdv Gust Time Date 14 5 LG 2 9 po Ong 00 12 01 1463 144 6 ZO 241 9 063 7 Q0 O02 1 13 9 146 5 2 9 Zale are 000 0 127 PAGE 8 Intentionally for better understanding the first lines of the date and time columns contain various format types Wind Speed m s Wind Direction Wind Speed SDV Lm o 2 o im i Q E o E Temperature C Temperature C CN c 9 pr amp Uo Z Sent s o Solar Radiation W m2 Fy Gust of Wind Speed m s Solar Radiation W m2 5 Gust of Wind Speed m s v Ke o Q xe Ss Gen e p 7 gt o Lines to skip ro Lines to skip WINDROsE Wind Data Analysis software 14 1 Ala 6 2 22o T 00 37 00 2000 12 1 13 4 146 6 2 DS 00 4730 2000 12 1 13 1 EU 2 6 20 4 0057 00 12 11 12 4 ea E DD 20 4 0107 2000 12 01 12 6 140 6 Ze 19 6 0117 ZIP EP ONE 12 8 146 5 3 2 21 9 1 27 00 20001201 12 4 ial 1 3 0 Pa 1 37 00 20001201 13 2 147 8 3 0 20 4 1 47 00 20001201 13 3 150 3 3 0 20 4 Cts 700 20001201 12 8 148 3 2 5 DE
5. D Foussekis CRES April 1994 Wind Energy Conversion Systems L L Freris Prentice Hall 1990 Atmospheric Science An Introductory survey John M Wallace Peter V Hobbs Academic Press Inc 1997 PAGE 55 Air Density kg m as a function of Elevation Comparison of the 2 methods Case B Method 2 Temperatures C 5 10 1 269 1 247 1 257 1 235 1 245 1 223 1 233 1 211 1 221 1 200 1 210 1 188 1 198 177 187 165 175 154 164 143 152 132 141 121 130 1 110 119 1 099 1 108 1 088 1 097 1 077 1 087 1 067 1 076 1 056 1 065 1 046 1 055 1 035 1 044 1 025 1 034 1 015 1 024 1 005 1 013 0 994 1 003 0 984 Comparison of the first two methods for the air density variation with height sm e e e JE 1 1 ile 1 1 1 1 1 1 ili 1 ile 1 1 1s
6. Imports the input settings from other vvindRose files OK including older versions Choose whatto import lid Iv Input Sheet IV indCarr Sheet PowerCurve Sheet Excel file created with vvindRose Je sold data w indRase 14 lz Browse PAGE 5 Input Note that wildcards and P are accepted to specify multiple files WINDROSE Wind Data Analysis software THE INPUT WORKSHEETS Four tables compose this worksheet which deals with the input data and the necessary parameters required by the program The first table contains the names of the ASCII files to be read together with the column numbers of the requested quantities wind speed direction etc The number of the lines to skip in the beginning of each file containing blank lines or comments is also given here Multiple data files can be specified by using wildcards in their names If the order of the files is not the chronological one from older to newer data then sorting of data 1s performed automatically without any user action However additional memory is needed for that Some of the columns of the ASCII data file can be assigned as logger s battery level the rain the signal quality of Lidars CNR for Windcube points in fit for ZephIR etc The acceptable range of each control signal can be assigned in WindRose s Advanced Obptions menu Column Humber m E CET CE gt E J eS cS 2 z TX D a 8l E File Na
7. in the table of the znput sheet Column Numbers File Names Wind Speed m s Wind Direction Wind Speed SDV Temperature C Solar Radiation W m2 Gust of Wind Speed m Lines to skip c data site1 st010604 000 21 17 13 9 5 33 29 25 22 18 14 10 6 37 20 16 12 8 4 33 Multiple anemometers and vanes Example form of Input sheet s 1 table Case with 5 anemometers and 3 wind vanes PAGE 9 WINDROSE Wind Data Analysis software Measurements Height above Ground level m 10 20 30 40 50 Setting the corresponding anemometers heights Input sheet s 2d table In this example the 21 column contains the wind speed at 10m height The second table of the rnput sheet contains all the necessary parameters characterising the analysis to be performed These parameters are The measuring period start and end dates The time interval of the measurements 10 minutes 1 hour etc The limit for the calm which is the value of the wind speed below which the response of the wind direction measuring device is not reliable recommended value 2m s The number of the direction sectors 16 12 or 8 that the program will use recommended value 16 Title1 Site s name Title2 comment Start from 18 12 1999 End at 29 5 2000 Minutes between data 10 Limit for calms m s 2 Number of Direction Sectors fs Measurements Height above Ground level m 10 above Sea level m 300
8. per day of a month PAGE 33 21 22 736 151 252 61 313 WINDROSE Wind Data Analysis software Variation of the Wind Turbine Energy production per month 4 WT Production kWh QOo Wind Speed m s WT Production kWh Wind Speed m s Month WI s energy production and mean wind speed per month TempT TempG If in the Input worksheet column numbers for the temperature were supplied then in these 2 worksheets similarly to the previous ones one can find for each month the mean hourly temperature per day T April Mean Temperature 16 2 C 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 Variation of the mean temperature per hour during 1 month period Note some missing data between 3 and 5 April SRadT SRadG Again provided that in the Input worksheet column numbers were specified for the solar radiation then these 2 worksheets contain the sum of the hourly solar radiation per day for each month of the year Similarly to the above the tables and the graphs are in the snaar and SRadG Worksheets respectively Some pyranometers solar radiation measuring devices record during the night small negative values instead of zeros due to a limited precision of their A D converter WindRose takes this into account by considering these data as zero so that the correct hourly daily sums are calculated PAGE 34 S Rad Wh m2 TempData WINDROSE Wind Data Analysis software De
9. we finally obtain the following non linear x function ey He 1 2 Q T 14 x u Resolving g x 0 1s achieved by applying the Newton iterative method starting from an initial value For the v iteration we can PAGE 50 WINDROSE Wind Data Analysis software write v l v 4 0 8 x cd 1 v 1 Its 1 x derivative of g x is 8g x T AT x 2T 4 2x 14 x BE C r 1 x oro here I W Q ax The gamma function s derivative I x is calculated using the relationship x END In x 1 that finally gives roj ZO Using 5 that is the asymptotic approximation of the Gamma function leads to the following form for the B x function B x v2m x e 4G a In J 900 2x Ox Keeping just the first 4 terms we obtain B x J2zx ve A E He in ae as 3 X x X X As a result B x is approximated permitting the calculation of the g x at the v iteration 2 Method Using the data distribution Weibull paper This method also referred as Weibull paper can be applied only if the wind speed distribution is already known The cumulative form of the Weibull distribution 1 e the probability that the wind speed is greater than a given value is written as Olu 2 u AG Taking twice the logarithm we obtain In In Q u k n u k n C Therefore if we plot a graph with the X and Y axes to be n u and PAGE 51 Power Curve correction WINDROSE Wind Data A
10. 6 5 m s WT Annual Energy Production amp cap factor calculated from data distribution 4 157 412 3 kWh 23 71 calculated from Weibull distribution 4 121 376 8 kWh 23 51 Best Sector in Energy contain NNW 41 22 2nd best Sector in Energy contain SSE 20 64 Best Sector in Time distribution NNW 29 95 2nd best Sector in Time distribution SSE 23 59 Main Directions Mean Wind Speed variation N SIN Nw 10 NNE lett NW WNW ENE W E wsw ESE S Time Distribution E 15m s 1 10 15m s 3 5 10m s 2 5m s LLL 56 u N NNW NW WNW Ww WSW SW SSW S SSE SE ESE E ENE NE NNE The main results of the WindRose program worksheet Results PAGE 16 WINDROSE Wind Data Analysis software The mean wind speed Note that if more than 1 year data exist then apart from the mean value of all the wind speed data the annual mean is also provided The calculation method of this value see below assures that non stationary effects such as seasonality are not taken into account The turbulence intensity at the specified wind speed range The maximum wind speed and the date of its occurrence The maximum gust Isec value and the date of its occurrence if a gust column was set in the Input sheet The mean power of the wind per area in units of kW m The total wind energy per area kWh m The coefficients of the Weibull d
11. Each point of the drawn line represents the mean value of all the data within this sector i e the value of the distribution at 0 is the mean value of all the data found within the sector 2 8125 2 8125 It is reminded here that in order to calculate both time and energy distributions per direction sector only data with wind speeds greater than the calm limit set in the Input worksheet are taken into account If the user wants to change this and take into account the total data he has to set the calm limit to Om s and re run the program Finally note that the energy refers to the energy of the wind not the estimated electrical energy of the wind turbine Although in most of the cases the two distributions are identical differences might exist in case of very high wind speeds greater than the cut out wind speed where the wind turbine 1s supposed to shut down 8 The low limit is actually the calm threshold value as set in the Input sheet below of which the response of the wind direction measuring device is not reliable PAGE 20 Shear WINDROSE Wind Data Analysis software This Excel worksheet is created automatically only if multiple anemometers are present The first graph shows the mean wind speed values as measured along with the average vertical wind shear Wind shear calculations are based on the power law u z u z 5 ref where the Ze denotes the height at which the wind speed u z 1s measured
12. NNW sector have an average wind speed of 8 4m s varying from 6 5m s to 9 8m s and average direction 321 varying from 315 to 327 Therefore for the wind speed the slope and the offset of the linear regression are calculated as o 1 022 and B 1 141 and for the wind direction the difference of the means is Ag 321 330 5 9 5 During the prediction phase if in February the wind speed of the Reference site is u 6 2m s and the wind direction is 330 then the predicted values u for the other site are u au tp 1 022 6 2 1 141 7 48m s and q ot Ag 330 9 5 320 5 Similarly the same procedure is followed for the correlation prediction of the turbulence intensities The predicted gust value is taken as Umax predicted 3 USDV predicted Notes on the methodology example e The correlation results depend on the common data period Therefore 1f the Reference time series were 5 PAGE 39 WINDROSE Wind Data Analysis software or 15 years long the correlation results would be the same and based to the common 9 month period e Apart from the partial correlation results per wind speed bin and per direction global correlations are also calculated i e a per wind speed bin no matter the direction b per wind direction for all wind speeds above calm e Assume that SSE wind directions appeared only during January when no data exist for the Target time series Therefore no cor
13. gt S 5 io OL 2 Q o o oO Qa 5 5 O O D gt o qu m ox 5 NUN n Uu 2 O Month Mean wind speed variation per month Two very essential tables are presented in this worksheet The 1 one shows the variation of the turbulence intensity 0 U 100 over the wind direction and the wind speed The 2 one shows the total data distribution over both wind speed and direction The following points should be noted here The binning of the wind speed is done in 25 steps of 1m s starting from the Om s m s with the 26 one gathering all data higher than 25m s The table of the turbulence intensities does not show the first five velocity bins the Excel s Hide cells option 1s enabled as no physical meaning can be extracted from turbulence intensities when the mean wind speed approaches to zero In that case the ratio o U takes very high values dominating the rest ones in the graphical representation of the table see worksheet 3D PAGE 27 3D WINDROSE Wind Data Analysis software Data distribution vs Wind Speed amp Wind Direction ENE E ESE SE SSE sS SSW SW WSW 92 134 140 127 158 220 218 204 122 78 77 71 81 37 34 27 26 12 7 11 8 4 0 4 5 4 8 0 2 4 1 6 4 2 3 3 1 1540 3298 1900 1309 1270 2565 4325 18684 18383 2698 223 Tables worksheet Data distribution per wind speed and direction The number of the wind direction sectors is the one selected in the Inp
14. ncaa tio oem a 33 TempT TempG SRadT SRadG a 34 URSI BEI NM 35 Correlation Prediction of missing data MCP method WindCorr WOFKSDGQt s i oet xo ER Pee eet 36 Methodology asia iuis amd euis ota O nap Raid 38 Appendix Advanced Options aiti te ma COR S SORA ER lameness 45 Air density variation with Height ssssssse 47 Weibull distribution methods seeeeees 49 Power Curve corree Holocaust ee tni t rise ead 52 Installation issues WindRose XLS Security wWarning ccccccccccccceeeeeeeeees 53 References 1 1L RII eme emer rere mer re re essere serrer eis 55 WINDROSE Wind Data Analysis software INTRODUCTION WindRose is a software tool dedicated to the analysis of wind characteristics speed direction turbulence temperature It is not a standalone program but an Add In to the Microsoft Excel 2000 XP 2003 2007 for the Windows 9x ME NT4 2000 XP Vista 7 operating systems The analysis results are stored graphically and numerically into spreadsheets which can be further used as ordinary Excel files The program is designed to provide all the results of the data analysis in a customisable form to meet any particular needs Thus the user can rearrange all the graphs resize them change their colours copy or link them to other sheets or programs 1 e embedded links to Microsoft Word document create new tables using the numerical r
15. of the measurement campaign The fourth table of the Input sheet deals with the calculation method of the air density which seriously affects the wind energy calculations If data contain time series of atmospheric pressure and temperature PAGE 12 Power Curve WINDROSE Wind Data Analysis software then this table is not used since the air density is calculated from real data complying with IEC 61400 12 In the opposite case three are the possible choices Air Density Method 1 p f z Method 1 p 2 Fiz Method 2 p F B z Use the table below dt n m dE ind mi P f r Method 2 anh Selection of the method for the Air Density correction The first method uses an empirical formula relating air density only to the anemometer elevation site height anemometer height The second method calculates air density using both the height and the temperature Two additional values are required the vertical temperature gradient recommended value 6 5 C km and the estimated mean ground temperature for the measurements period If temperature data exist into the data files then the recorded mean value is used instead The third method uses 12 preset values one per each month of the year These values can be retrieved either from a nearby meteorological station or set intuitively The mean value of the air density used during the calculations is given in the Results
16. one per each month of the year These values can be retrieved either from a nearby meteorological station or set intuitively The probability density function f u of the Weibull distribution 1s given by the formula f u pu 1 where k and C are the characteristic parameters of the distribution Two methods exist for the calculation of these parameters 1 Method Using only the Mean and SDV values Solving 1 for k and C requires that the mean value u and the standard deviation o must be expressed as a function of k and C From the mean value definition o0 00 k 1 f u 4 E u uf du Lr e Bj du A y NOINE k Using the transformation t we obtain C 1 k u ct and au du k C H AS elt e Pol 0 Therefore Recalling the gamma function definition r g eds g gt 0 0 we deduce that PAGE 49 WINDROSE Wind Data Analysis software l u cl 1 2 k Now for the standard deviation o 1s valid to write E u py j o EQ p Following the same transformation we have amp e erra k So finally we conclude Dri NA PAS l O eprz r i d 3 The Weibull distribution in its cumulative form P u lt u is written as Or PENNAS 4 The following asymptotic series is used as good approximation of the gamma function T DL ex gu TAX 5 where A x Le ee een l l 139 571 with gom rean 7 ee 12 288 51840 2488320 Dividing 2 and 3 and setting x
17. peaks greater than m s VERTICAL PROFILE CHECKS Warn if Mean vind Speed ratio W291 is greater than Reject if Mean vind Speed ratio LI2 LI1 is greater than Warn if Gust Vind Speed ratio W291 is greater than Reject if Gust vind Speed ratio W201 is greater than Far 0 UW 5 multiply the above ratios by For 5 LI 10 multiply the above ratios by For 1 0 Ll 15 multiply the above ratios by For 15 U multiply the above ratios by Warn if Mean vvind Direction diff is greater than ded Reject if Mean vind Direction diff is greater than ded MISCALLENEOUS Min monthly completeness when using the 12 monthly distributions 95 Wind Shear Reject values of a power law coe if R2 is less than Reset Default Values ATTENTION the decimal separator for this PC is ie pi 3 14 OK Cancel List of the Advanced Options of the WindRose program PAGE 46 Air density variation with height WINDROSE Wind Data Analysis software Examples of the WindRose log file when an error occurs quA AOS IA oO m SEI SOT Value beyond the error limit for the vertical profile Umax 45m 39 41 Umax 30m 212 55 while Umean 45m 7 51 Umean 30m 6 98 Boe Tele clo enka txt Lines 21742 OIO O00 Z OOM ar EEror Values beyond the error limit for the vertical profile Dirl 230 0 Dir2 329 3 while Umean 45m 10 64 Umean 30m 9 93 Umax 45m 13 76 Umax 30m 12 87 Files 3 cata sicel cxt Lines 5334 WindRos
18. results of the correlations windCorr sheet Note that the contents change according to the direction selection of the drop down button Selected Direction NNE Wind Speed of Target Site m s 0 2 4 6 8 10 12 14 16 18 20 Wind Speed of Ref Site m s Another result of the WindCorr sheet Mean wind speeds correlation per direction sector Output Files Apart from the results presented in the worksheet s tables and graphs three ASCII files are created containing the detailed results The 1 one contains the concurrent pair of data ie in the above example 9 month data which were used for the correlation prediction The 2 one includes time series created by applying the calculated coefficients to the Reference time series in the given example these are the 12 month long data Note that even if data existed 1n the time series there are not included PAGE 42 WINDROSE Wind Data Analysis software The 3 one is again the predicted data but only for the missing data period 1 e in the above example the predicted data for January February and October Finally the 4 table of the windcorr sheet contains the correlation results per wind direction for all wind speeds above 5m s Particular notice should be given to the angular shift of the wind direction vector Correlation of directions for U gt 5m s Dir Difference deg Mean Wind Speed of Ref site m s NNE NE ENE E ESE SE SSE S SSW SW WSW W WNW N
19. sheet More details on the air density calculation methods and on the power curve corrections are given in the Appendix Pairs of columns containing power curves of various wind turbines compose this worksheet Some representative power curves are included by default but the user can add more or delete the existing ones One restriction applies here The user has to respect the form of the existing power curves 1 e put in the Row cell the model name of the wind turbine in the Row2 cell of the next column the tower height choose the power control strategy etc Note e A linear interpolation is performed in order to calculate the exact electrical power of the wind turbine between the given values of the power curve e Whenever the measured wind speed is lower than the minimum value or greater than the maximum value of the power curve s wind speed then the power is considered as zero and the wind turbine as out of operation PAGE 13 WINDROSE Wind Data Analysis software ied jas sa NEG Micon 750 48 diameters 48 tow er 45m control stall control stall n U m s Power kW 3 0 700 4 19 5 5 53 1 6 97 4 600 7 155 3 8 244 6 O Ref Power Curve 9 349 2 5 Correction applied 10 462 2 11 564 5 400 12 640 5 S 13 696 3 e 14 729 8 300 15 745 5 16 750 17 744 6 200 18 734 8 19 723 20 711 9 m 21 701 4 22 694 3 0 nes 0 5 10 U Mis 20 25 30 25 700 6 Left A sample of a power cu
20. that 32000 data per graph an external graph library is used to create them in a form of a PNG picture Consequently these graphs cannot further customized as other Excel graphs in WindRose XLS This Sheet is created only if multiple anemometers are present and sufficient data exist 1e several months Ideally cup anemometers should be calibrated in a wind tunnel after the measurement period in order to assure the accuracy of the measurements during the entire measurement campaign Another inferior possibility is to perform an in situ comparison between the top and the control anemometer usually 1 5m below the top one This procedure is defined in the Annex K of the IEC 61400 12 1 standard Briefly the procedure can be summarized as follows Data are filtered for a narrow direction sector and classified in bins of 1m s Then a linear regression is performed between the two anemometers using data from the initial 1 2 months Finally the regression results are applied to the last 1 2 months of the measurements and compared to the real data In Situ comparison of anemometers at 28 5m and at 30m wind directions 348 75deg 11 25deg Evaluation period regression from 12 1 2008 to 8 3 2008 Application period deviation results from 16 11 2008 to 11 1 2009 Anemometer at 28 5m m s 4 5 6 7 8 9 10 11 12 13 14 15 16 16 y 0 9618x 0 3264 C2 14 13 12 11 2 10 m Squared Sum of Deviations O Co
21. where p 1 225kg m Therefore the air density at the specific site elevation needs to be evaluated using 3 different methods including or not temperature time series etc Finally this Jocal air density and the type of the power control of the wind turbine 1 e stall or pitch provide the corrected power curve according to the IEC 61400 12 recommendation Two charts are shown below the results table A pie chart presenting schematically the most prevailed directions of the site and a polar chart showing the variation of the mean wind speed as a function of the direction Finally the data distribution in time per direction sector 1s given Each bar is composed by 4 parts depending on the percentage of time that PAGE 19 WINDROSE Wind Data Analysis software the wind speed at the specific sector falls in the 4 preset ranges 2m s 5m s 5 10m s 10 15m s gt 15m s WindRose This sheet is composed by the total and the monthly wind roses distribution chart in polar co ordinates They present the wind speed distribution 96 per direction in the time and energy domain May N B 9o energy time NNW 57 NNE N y 2 AA WNW ANS KLAN NN i ISX 10 XN w 0 E Ld ZA A x lt X we bd ci X lt X B d pa M SW IN XK ad S SSW SSE S Samples of a site s wind roses total and monthly Note that for maximum precision the polar plots are divided in 64 sectors each 5 625 360 64 wide
22. 020700 20001201 Site Name SOMEWHERE Site Number 1 Start Time 18 57 03 01 2000 Finish Time 15300 03 21 2000 Total Time 25 day s 20 hour s 3 minute s DATE TIME Anem A Anem A Anem A Anem A Vane A Other An Other An Other An Other An Wind Van Average Avg Dev Minimum Maximum Average dms OE 10 manic 10 minut LO miLmuT LO mantic m s m s m s OLE time m s enka time 2 7 Mar 1 2000 18 57 IORS ea T9 236 18 59 Ll 211 18 59 m a E Mar 1 2000 19300 TED 0 8 9 4 221 19 05 Se 219 19 04 222 0 E Mar 12 2000 19310 eZ Qu 10 6 226 19317 qood 224 19 19 222 2 ONDER oO D Mar D DER L21 oR 10 0 225 19322 13 1 228 19325 226 4 BS 5 7 6 2 Mar 1 2000 19 30 11 2 0 7 9 6 228 19 31 12 4 219 19 30 227 8 EBSEBEEEIRCRE T Mar 1 2000 19340 MORO 0 6 9 1 225 19344 12 0 23393 19 49 226 4 ERES E A Mar 1 2000 19 50 10 0 8 9 1 243 19 54 12 7 245 19 59 236 2 eS rr Mar 1 2000 20 00 RORA 10 ead 277 20304 13 0 250 20 06 250 3 Wa E E 2 Mar 1 2000 20 10 11 4 1 2 8 2 246 20 15 15 8 245 20 18 253 1 EXEXEXNZEDUNDS z Mar 1 2000 20 20 12 2 1 5 o 6 9u 202231 14 3 245 20525 251 7 MERE 8 21n 13 Multiple anemometers vanes within the same mast When multiple anemometers and wind vanes are present WindRose provides some additional results 1 e wind shear Weibull distributions per height etc In that case the column numbers of each device should be separated by a semicolon Furthermore the heights of the anemometers should be set accordingly in ascending order
23. An experienced user will find this screen particularly useful as he can quickly deduce the quality of the data spikes fault operation of a device etc Displaying a smaller data window 1s also possible when the View Mode is set to Zoom X amp Y or Zoom X if zooming 1s done only in the time axis When a smaller data window is displayed the user can smoothly move forward or backward by selecting Pan X amp Y or Pan X just for the time axis and holding down the mouse while dragging it to the desired direction PAGE 15 Results WINDROSE Wind Data Analysis software THE OUTPUT WORKSHEETS This one page sheet presents the summary of the performed analysis The displayed quantities are Measurements Period from 4 2 2002 to 3 2 2003 Mean Annual Wind Speed at 40m height 6 2 m s general mean 6 0 m s Mean Turbulence Intensity at 10m s 9 2 Max 10min Average Wind Speed 27 2 mis 6 1 2003 07 50 Maximum Gust 34 1 m s 6 1 2003 07 40 Uncertainty of Wind Speed measurement 0 2 m s Mean Wind Power 307 8 Watt m Total Wind Energy 2477 1 kWh m Weibull Distribution constants shape factor k 1 62 Ter 3 f Te co am 09 6 7 m s Total number of valid data 48280 Included number of calms lt 2m s 5593 AEL Ke ELE 4280 8 1 Expected W T energy production 3 740 376 3 kWh Gamesa G80 2 0MM W T Capacity Factor 23 2 Pave 1 148kg m Calc Annual Mean Wind Speed at 67m height 6 6 m s general mean
24. Medium You can choose whether or not En run potentially unsafe macros C Low not recommended You are not protected from potentially unsafe macros Use this setting only if vou have virus scanning software installed or vou are sure all documents vau open are safe Ma virus scanner installed Now open WindRose XLS A security warning appears Choose Details Security Warning Hi WwindFiose COM WindRiose xls contains macros by WindRase Macros may contain viruses It is always safe to disable macros but if the macros are legitimate you might lose some Functionality Always trust macros from this source Disable Macros Enable Macros More Info Then select View Certificate and afterwards Install Certificate PAGE 53 WINDROSE Wind Data Analysis software Digital Signature Details Certificate General Advanced General Details Certification Path Digital Signature Information This digital signature is OK Certificate Information aa RUURE This certificate is intended for the following purpose s eEnsures software came from software publisher Name WindRose eProtects software from alteration after publication E mail Not available Signing time Not available View Certificate Issued to WindRose Countersignatures Issued by WindRose Name of signer E mail address Timestamp Valid from 31 12 00 to 31 12 06 Details ssuer Statement
25. OSE Wind Data Analysis software to the height site elevation anemometer height can be done as follows depending on the user s selection Input sheet 4 Table 1st Method This method mainly used when temperature and pressure measurements are not available employs an empirical formula in which air density 1s a function only of the height m from the sea level p 1 226 g 3108910 gt Height 2nd Method Another method assumes the atmosphere s adiabatic variation therefore dp dz BP where g 9 81m s 1s the gravity acceleration p the air density and z the height Using the ideal gas equation p RpT where R 287 Joule deg kg is the gas constant p the atmospheric pressure and T the temperature in Kelvin K we obtain dp sy p RT Assuming that the temperature T varies with height T To Iz With T 288 K 273 15 and J is the vertical temperature gradient usually taken as 6 5 K km Integrating the above equation Weibull distribution WINDROSE Wind Data Analysis software that relates the air density to the temperature T and the height z of a site Note that if the ASCII data files do not contain temperature columns then a mean temperature value for the measurement period has to be set in the fourth table of the rnput sheet At the end of the document a comparison table is given showing the differences of the first two methods 3rd Method This method uses 12 preset values
26. W NNW N A graphical representation of the Directions Correlation together with the mean wind speed WindCorr sheet When averaging wind direction values due to the discontinuity of the 0 360 point it is checked whether data exist in all the 4 quarters of the circle In that case results appear in red to remind that no reliable conclusions can be made Correlation of directions for wind speeds gt 5m s Direction of From To Number Mean Direction Mean Direction Corr Coeff ad z d nba Ron d Ref Site of data of Ref Site of Unc Site speed speed 11 25 33 75 56 25 78 75 101 25 123 75 146 25 168 75 191 25 213 75 236 25 258 75 281 25 303 75 326 25 348 75 Red color Mean values result from the 4 quarters of the trig circle Wind Direction correlations WindCorr sheet PAGE 43 WINDROSE Wind Data Analysis software Generate Predicted time series using the Tables with the results of the statistical analysis S Locate S penn nf Table common Incomplete time series U t UREA mean amp DIR Ref nean data Ref dt ms NE SE sS SW W NW N H 1 4 6 nik i Pred mE B G RUNE PE lw AN 3E MEAN t me e Inc Shift time series by AT Prediction method Wind Speed Use the regression coefficients of the wind speed bin to obtain the predicted value Ref E Wind Direction Use the differences o
27. alculation of the k C coefficients a using only the mean and the standard deviation of the wind speed and b using the data distribution restricted or not to a specific range 1 e 4 16m s The second method is recommended More details can be found in the Appendix WEIBULL DISTRIBUTION The wind speed uncertainty 1n case of analogue anemometers 1 e the output signal of the anemometer is analogue voltage These parameters are not used in case of anemometers producing pulses revolution The name of the intermediate file that must be created when site correlation will take place details in the CORRELATIONS chapter The name of an ASCII file into which all the row data are exported This is particularly useful when delivering long term data 1 e yearly that are stored originally in many files and or to which correction coefficients were applied The 2 values below the selected Weibull calculation method are taken into account only if the 2nd method is selected 6 Uncertainty calculation for pulsed ouput anemometers will be added in the next program release 7 The created export file includes all the corrections so no corrections should be applied if it is further processed PAGE 11 WINDROSE Wind Data Analysis software Correction Coefficients E o LL LL oO iS 9 m Je xe go X fess lt o o Wind Speed MULT Wind Speed OFFESET m s Wind Direction MULT Wind Direction OFFSET
28. and a is the wind shear exponent coefficient Mean vertical wind profile Height m 0 2 4 6 8 Wind Speed m s Mean wind speed per height measurements amp mean vertical shear sheet Shear The next table shows the measured mean and max wind speeds per direction and per height At the end the calculated values are given for the hub height of the selected wind turbine The variation of the Weibull coefficients per height and per direction is given in another separate table Both tables provide results at hub height based on the extrapolated wind speed time series at that height from the higher anemometer using the wind shear results The last table of this worksheet presents in detail the average value of the exponent coefficient a of the power law per wind speed and per direction For each time step one value of a is calculated Values of a are considered valid only if the goodness of fit R gt 0 95 Note that a this value is user selectable from Advanced Options of the WindRose menu and b it 1s displayed on the top of that table along PAGE 21 WINDROSE Wind Data Analysis software with the data percentage satisfying that condition providing thus an indication of the validity of the Power Law at the specific site Measured wind speed black dots and calculated wind shear Goodness of fit criterion Left R2 0 95 rejected Right R2 gt 0 95 accepted Average and Max 10min wind sp
29. campaigns must contain data from a common time period Otherwise correlations cannot be calculated Based on the common data calculations are performed to estimate correlation coefficients p linear regression coefficients a b of the best line fit and the goodness of fit R All the above are calculated per wind speed range and per direction sector per wind speed range but for all directions per direction but for all the wind speeds above calms for all wind speeds no matter the direction It 1s reminded here that correlation coefficient values close to 1 0 show similar in phase variations while values close to 0 0 show Goodness of fit R 1 with N 2 N ssE y and ssM gt y i l i l where y ax c b a b are the slope and the offset of the best line fit Correlation coefficient Mz xl x2 ib p where xliz n and X212 are the O 0 common data of the two time series x1 kal x2 with mean values y and y and standard deviations o4 and o I m PAGE 36 WINDROSE Wind Data Analysis software irrelevant variations uncorrelated phenomena Furthermore goodness of fit values close to 1 0 show that the best fit line represents very well the data cloud using the linear regression coefficients Input parameters The first column of the windcorr sheet contains the necessary parameters for the program to run The main input is 2 files co
30. cember 2003 Total 54877W m2 Data completeness 99 8 3500 3000 2500 2000 1500 1000 500 0 1 3 5 7 9 11 13 15 17 19 2i 23 25 27 Day Solar radiation daily sum during one month period worksheet SRadG Finally in this worksheet the detailed information which is indispensable for the all graphs appearing to all the above mentioned worksheets is presented In order to improve the readability of the sheet the data are grouped in sets of columns of different colours It 1s reminded here that by double clicking on the every Excel graph the user can see the corresponding data columns when selecting X Values and Names and Values Moreover new graphs can be created using all the listed data which will be updated automatically each time the program runs PAGE 35 29 31 WindCorr WINDROSE Wind Data Analysis software CORRELATION MISSING DATA PREDICTION This worksheet uses the MCP method Measure Correlate Predict and deals with the correlation between the measured data at two different sites Before using it two files one per site have to be created by running WindRose twice setting two different file names in Input sheet s appropriate cell When done the program runs by selecting the Calculate Correlations option in the toolbar of the Excel s main menu Site Correlation file name Creating the necessary file for the Correlations sheet Input Important Note e The two measurement
31. d to a specific site and consequently can be applied to several WindRose xls files EA Microsoft Excel WindRose xls File Edit View Insert Format Toole Data Window WindRose Help Adobe PDF Deh e T amp Bam WE Run EB E Iculate Correlations Arial TH 2 EE Ca Results in English E B 7X Clear Data B Changing the preset parameters of the WindRose program These parameters include among others the acceptable limits of each quantity the way spikes are detected the tests to be performed to check frozen anemometers etc Most of them are straightforward and self explanatory After each run they are all stored into the TempData sheet for traceability reasons Consequently it should be always possible to reproduce the same results even if some of the advanced parameters were modified they appear in red in TempData sheet When multiple anemometers are used a check is performed between the simultaneous wind speed values per height If strange values are noticed then a warning or an error is produced and an entry 1s added into the Log file They are based on the acquired experience and refer to each pair of successive anemometers and or vanes Note that 1f a mast has 4 anemometers then 3 successive tests will be performed for the anemometer pairs 1 2 2 3 and 3 4 The alarm conditions are summarized below e Mean Wind Speed Vertical Profile U2 Ul lt
32. diurnal distribution of only one specific month then the program should be run once per year specifying accordingly the start and end dates in the Input worksheet March April Mean Wind Speed 8 6m s Mean Wind Speed 6 7m s 9 ON FD OO NY Wind speed m s Wind speed m s 0 2 4 6 8 10 12 14 16 18 20 22 0 2 4 6 8 10 12 14 16 18 20 22 Hour Hour Example of the Diurnal distribution per month 12diurnal worksheet The 4 most common representations of the wind distribution are plotted in the present worksheet The first two show the distribution of the wind based on the energy not the expected energy production of the wind turbine as a function of the wind direction 1 one and the wind speed 2 one The last two show the time based distribution of the wind as a function of the wind direction 3 one and the wind speed 4 one Energy Distribution vs Wind Speed Energy kWh m 123 45 6 7 8 9 0 1 12 8 4 15 16 17 18 19 20 21 22 23 24 25 Wind speed m s Wind energy distribution versus wind speed BarCharts worksheet The main figure of this worksheet shows the Weibull distribution that fits the wind data All the wind directions have taken into account when calculating the shape k and the scale C coefficients of the Weibull distribution in this graph If multiple anemometers are present the graph refers to the reference anemometer usually the highest one PAGE 30 WINDROSE Wi
33. e Warning 12 Error g 5 e Max Wind Speed Vertical Profile U2 UI e Warning 1 8 Error 2 0 The final value of depends on the wind speed If 0 lt u lt 5 then g 2 5 If 5 lt u lt 10 then e e 2 0 PAGE 45 WINDROSE Wind Data Analysis software If 10 lt u lt 15 then g 1 5 If 15 lt u lt 65 then e 1 0 e Mean Wind Direction Vertical Profile Ag for U gt 5m s Warning Ag 25deg Error Ag 45deg e Spike detection for consecutive time steps Umean 10m s Umax 13m s e SDV constant or invariant during consecutive time steps Advanced Options WindRose control parameters GENERAL CHECKS Reject Vind Speed data if less than mis Reject Wind Speed data if greater than mis Reject ind Direction data if less than deg Reject ind Direction data if greater than ded Reject identical Wind Dir data for 3 successive time steps 0 no 1 yes Reject Gust Vind Speed data if less than m s Reject Gust Wind Speed data if greater than m s Check if Gustis greater than Average Wind Speed 0 no 1 yes Reject Usdy data if Turb Intensity is less than to Reject Usdy data if Turb Intensity is greater than 93 Reject successive data if Usdv 0 and U Ucalm na 1 ves Reject Temperature data if lezs than deg Reject Temperature data if greater than degt SPIKE DETECTION Successive Mean Vind Speed peaks greater than m s Successive Gust Vind Speed
34. e considered as one column instead of 3 Additionally right after the column name we put the character n 1 e for the 10 column we write 10n Some models of the CAMPBELL data loggers use another particular date format An integer number is given indicating the number of days elapsed from the beginning of the year Thus for the 26 September 1998 it 1s written 1998 269 This case is treated by adding the letter c right after the date column number ie for the 10 column we write 10c However the day column i has to be next to the year column i e 11 in the previous example ii should not be the file s last one and iii is normally counted when numbering the columns of the ASCII file Below some examples are given showing how the znput sheet should be filled WindSpeed Dir Usdv Gust Time Date 14 3 SHG 2 9 NT 0 7 GRIP 19 00 14 8 144 6 2 8 21 9 OelTeoo L L 2700 ee 13 9 146 5 ZEE O27 200 RETI 12 0 ome 14 1 144 0 2 9 22 0 00 37 00 1 12 2000 mee 13 4 146 6 2m zu COAG e v le EE o 2XCYOIG ME 121 UC 2 6 20 4 0057 w 018 o Oo 2d 1492 3 ZO BO d 0107 1 12 2000 3 12 6 140 6 2 5 19 6 0117 1 12 2000 z 12 8 146 5 3 2 2159 Qt mao Ole 22201910 126 Ta S30 Al AL Oil SH T0 e OE 2 000 19 7 jm 350 20 i 27 00 03021201010 ies 150 2 EX 2d OL 700 O11 22000 12 8 148 3 2 5 Zu 020700 01122000 WindSpeed Disa Usdv Gust Time Date 1452 137 6 2 9 pr 0 7 1248 4010 14 3 144 6 DS PU Occ 72 VO 127 17 00 n 13 9 146 5 2 Dalal Qro 00127 70 oe
35. e expects power curves given in the PowerCurve sheet to be normalized as recommended by IEC sea level 15 C 1013 25mbar At higher altitudes and different temperatures air density varies considerably and its calculation has to be done accurately since it influences significantly the energy calculations Ideally for accurate air density measurements atmospheric pressure temperature and relative humidity have to be measured simultaneously with the wind speed In the opposite case 3 empirical methods are used to deduce the air density Analytically Case A using atm pressure and temperature data Usually atmospheric pressure 1s measured at a low height 1e data logger level assumed value 2m thus Bmeas data are extrapolated to hub height Hn according to the following formula B B 1013 25 1 2 25577 10 H 1 2 25577 10 Ba Then the air density can be calculated using the following equation corr RT R T Ba 4 0378 P where R 287 Joule deg kg T is the measured temperature in K g the relative humidity kot P the vapor pressure in hPa The last depends on the temperature and 1s calculated as follows P 2 05 10 7 0081867 Finally if relative humidity 9 data are not available then a constant value of 0 5 50 is assumed Case B Atm pressure time series not available When atm pressure is not recorded then air density estimation due PAGE 47 WINDR
36. e two series see above e The time shift for which the max Correlation coefficient occurred Finally in the last section are written e The correlation uncertainty of the wind speed e The global coefficients of the best line fit all wind speeds no matter the direction e The goodness of fit R The calculation of the correlation uncertainty 1s performed analytically per each wind speed bin and direction sector of the Reference site as follows VN where o is the standard deviation of the concurrent wind speeds of the Target site and N the number of data per each interval The displayed value is an average weighted value based on the data wind speed and direction distribution correlation uncertainty The 3 table of the windcorr sheet presents the detailed correlation results per direction and wind speed bin Note that this is not a static table and its contents can change depending on the selected direction in the drop down button PAGE 41 WINDROSE Wind Data Analysis software Correlations Table Select direction of the Reference Site MEN LUE Wind Speeds Directions Turb Intensities Unc site Unc Wind speed bin Number of bur Mean Correlation Sloni A site B vir aie Otel Ref site m s data value Coefficient P Mean b value m s value value value m s value 0 21052 0 15078 0 77246 0 42402 0 54443 0 08434 0 01391 0 62122 0 00230 0 10204 0 04968 The table with the detailed
37. ean hourly wind speed per day February 2000 Mean Wind Speed 8 6m s Data completeness 99 9 21 18 15 12 O UA Variation of the mean wind speed per hour during the month of January 24 31 744 points plotted Similarly to the previous 2 worksheets DIRhourT and DIRhourG show the variation of the wind direction per day mean value per hour for each month PAGE 32 WTprodT WTprodG Dir deg El Energy kWh WINDROSE Wind Data Analysis software February 2000 360 315 270 225 x U lt 5m s o 5 U 10 180 135 90 45 e U gt 10m s m s 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 Day Monthly Variation of the mean wind direction per hour These 12 tables wrprodr and 12 graphs wrprodT represent the sum of the hourly wind turbine energy production per day per month of the year Note that the energy production results depend on the data completeness February 2000 Total 267252kWh Mean Wind Speed 8 6m s Data completeness 99 9 20000 16 15000 12 C 10000 8 E 2 5000 4 1 3 5 7 9 11i 13 15 17 19 21 23 25 27 29 31 Day mmm E o U WIprodc Variation of the WI s daily produced energy together with the mean wind speed February 2000 Total exp production by Nordex N50 800 267252kWh Mean Wind Speed 8 6m s 3 5 Hour 1 2 6 8 9 10 11 12 13 14 15 16 17 18 19 20 WrprodT Table showing the hourly sum of the wind turbine energy production kWh
38. eed per direction Height SSW WSW W WNW NNW 3 88 measured 10 88 4 27 1927 4 49 14 37 4 62 14 43 4 72 14 83 calculated Mean wind speed per height and per direction sheet Shear Weibull parameters per Height Height Const NNE ssw WSW W NNW x C k C k C k C k C k Co e E Variation of the Weibull distribution parameters per height and per direction sheet Shear PAGE 22 WINDROSE Wind Data Analysis software Wind Shear exp coef a Data percentage with R 0 95 90 25 NNE ENE SSW WSW W WNW NNW 0 139 0 227 0 371 0 181 0 100 0 170 0 391 0 391 0 124 Wind shear exponent results per wind speed and direction sheet Shear Here it is underlined that internally the above table is calculated also significant variability of the exp coefficient a with these two parameters Thus when calculating the wind speed at a higher height ie hub height if such values of a exist per hour per month per wind speed and per direction then they are preferably used rather than the summarized ones of the above table Diurnal variation of wind shear All directions fms U 10m s 5223 10min averaged data coef a ponent B Power Law Ex a A typical graph showing the diurnal variation of the exponent coefficient a with the hour of the day Result from Lavrio site worksheet Shear At the end of the shear sheet a series of polar graphs are created showing the influence of the mast towe
39. es KANE a CRES Centre for Renewable Energy Sources 19 km Marathon Avenue GR 190 09 Pikermi Greece Tel 30 210 66 033 00 e mail dfousek cres gr Building a competitiva Greece MIMNISTAY CF DEVELOPMENT http www cres gr windrose A wind data analysis tool WindRose User s Guide Document version 8 1 Excel COM Add In version 3 88 Analysis DLL version 5 89 Table of Contents IntFOQUC TION ie oes dedii ies InStallatiOT e Fae cae tad Mac ada daeaseataneadendees 3 Running the program Running WindRose for the 1 time sese 4 Registration Demo version ccceccccccccecceceeeeeeeeeeeeeeeeeeees 2 The Input worksheets Ino cosas toilet siete a t Acad erent tcp tos 6 Multiple anemometers vanes within the same mast 9 POWGICULVE pie ROI ETE MUI LIES Esa Men sp M estes 13 Quick Check of data sseseeeeeeeeeeeenererrnennnnnn 15 The Output worksheets ROS serait koe scenes adi Mri tos cee 16 AV IBBTOSQ s operta obe a dto ta s irse topo 20 eic RU A OR On M RORIS 2l Ae X IK cioe neto te o ra QE vae ESTE mio eU tes cuota 25 Zoe ITO Sereda inia imo ch cena vas capio usan tun dies fas inei cin 26 Hlc 21 ON T EANA E CONN EAE ENT E T RS 28 2pie AZT lera e Cot o R 29 BarCharts Weib ll 2 detto dct bius ead eremo diva diis 30 UhourT UhourG DIRhourT DIRhourG 32 WLprodT WV prod Giantess cele
40. esults etc Data analysis complies with the requirements imposed by the IEC and MEASNET standards Main features e It performs complete statistical analysis of the wind data including Weibull distribution constants per direction sector and global turbulence intensity evaluation and polar plots wind roses of the time and energy distribution of the wind e It correlates data from two sites calculating correlation coefficients globally and per ranges of wind speed and direction As a result it provides the predicted time series for the missing data of a site based on the other site s complete set of data MCP method e Where multiple anemometers and or wind vanes are present the vertical wind shear is calculated per wind speed and per direction sector Thus the extrapolation of the wind speed at higher heights 1e hub height of a wind turbine is performed more accurately e A dozen of power curves from a variety of wind turbines are included providing a good estimate of the expected energy production PAGE 1 WINDROSE Wind Data Analysis software e It has a user configurable time step 10 minutes 1 hour etc and it is able to analyse huge data sets limited only by the available computer memory e In case of measurements with systematic errors a linear correction can be applied for all the measured quantities and separately for each file e It includes 3 methods for the air density correction due to
41. ews of the results are presented in the next worksheets Mean Daily Wind speed variation Wind speed m s 0 20 2 93 12 3 93 1 4 93 21 4 93 11 5 93 31 5 93 20 6 93 10 7 93 Date Wind speed variation per day mean amp max values e The 1 one shows the evolution of the daily mean wind speed as well as the maximum wind speed occurred per day for the given time step i e 1f 10minutes series are recorded the maximum value is of course not the instantaneous Isec gust but the maximum 10minutes wind speed occurred during this day e The 2 figure presents the Diurnal distribution Actually this is the distribution of the wind speed as a function of the hour of the day All the data are taken into account here independently of the wind PAGE 26 Tables WINDROSE Wind Data Analysis software direction The Diurnal distribution per month is presented in the 12diurnal worksheet Diurnal Distribution Wind speed m s BMEBN MUNI LITI O O 12 3 74 15 6 7 amp 9 10 1112 13 14 15 16 lf 18 19 20 21 22 23 Hour of the Day Distribution of the wind speed per hour of the day e Finally the 3 figure shows the mean wind speed per month If more than a year period is examined each monthly mean wind speed 1s calculated from all the corresponding months Mean Monthly Wind Speed Variation over a year Wind Speed m s O Q E
42. f the projection of the 2 distributions data and Weibull to 1 year Thus the notion of missing data is not applicable to them e If the given time period defined by the Start Stop dates is one year then the 3 values should converge In fact if there is no missing data then the 1 and the 2 values should be practically the same If the 3 value is close enough to the other two it means that the given Weibull distribution coefficients k C fit PAGE 18 WINDROSE Wind Data Analysis software well the data and can be used safely to characterize the site e For the same reasons there are also 3 wind turbine capacitor factors How the W T Energy production is calculated e The wind turbine energy production is based on the wind speed estimation at hub height If measurements are not taken at the WT s hub height then the wind speed is extrapolated to that height For that the power law is used in order to calculate the wind shear e When the wind speed is measured at different heights then the detailed wind shear per direction and per wind speed is calculated If the wind speed is measured at only one height then a uniform shear is assumed based on the exponent coefficient a of the power law e Having calculated the wind speed time series at hub height the energy production is evaluated using the corrected power curve e Usually the power curves given by the WT manufacturers refer to the sea level
43. f the mean bin angles as a displacement angle nc Thus Locate the V u 9 n k bin the predicted u 9 is f ee S u i A n k u as Dinky coef ro p p t 4g AT dt AT time Shift where Ag s P am Mean O n k Mean The n k and l m denote the Reference and Incomplete time series bins respectively Apply the time shift dt Table U Ref __ amp DIR Ref mean mean E SE S SW W NW N n U bin Correlate the synchronized data k DIR bin 1 Isolate the Ref data x x xy that belong to the n k mean wind speed amp direction bin 2 Calculate their mean value X pean 3 Calculate the mean value y of the simultaneous data JY c wan Ms 4 Locate the l m wind speed amp direction bin that belongs to 5 Calculate for each bin the following Mer linear regression coefficients a b Y pean aX meant 0 li goodness of fit R correlation coefficient p amp DIR nc mean 1 U bin m DIR bin Methodology of the correlation prediction procedure PAGE 44 Advanced Options Spike detection Warnings errors WINDROSE Wind Data Analysis software APPENDIX The Advanced Options menu of the WindRose includes various parameters which affect the program s behaviour but most of the times are not relate
44. hat time series from data loggers with unsynchronized clocks will be cotrectly processed PAGE 38 WINDROSE Wind Data Analysis software zero show uncorrelated phenomena and should alarm the user about the exploitation of the results Let s now assume for simplicity reasons that the Reference site holds l year long data and the Target site holds 9 month data missing months January February and October Hence the concurrent pairs of data are 9 month long The next step is to select the data of the Reference time series that belong to a specific wind speed bin and direction sector In the same time we investigate their concurrent pairs from the Target time series For both data sets all the statistical quantities 1 e mean values linear regression coefficients etc are calculated and stored Now if WindRose 1s asked to predict missing data then for the wind speed it uses the regression coefficients and for the wind direction the difference of the mean direction values Arithmetic example Assume 16 direction sectors and 2m s wide wind speed bins Suppose that 100 Reference time series data have directions from the NNW sector 326 25 to 348 75 and their wind speed falls into the 6m s 8m s bin The mean value of their directions is 330 5 and their average wind speed is 7 2m s The corresponding concurrent simultaneous 100 data from the Uncompleted time series not necessarily within the 6m s 8m s bin and the
45. istribution that fits all the data The total number of valid data within the specified time period by the Start Stop dates given at the Input sheet The percentage of the missing data The number of calms wind speed below a threshold The expected electrical energy production kWh of the selected wind turbine during the given period of measurements for all the valid data The capacity factor of the wind turbine percentage of the nominal power of the wind turbine at which the machine should operate continuously to produce the expected electrical energy The estimated mean wind speed at the nacelle height of the wind turbine Similarly with the measured mean wind speed if more than 1 year data exist then the given value is the average of 12 monthly values Each monthly average value is weighted with the completeness of the given month for all the years The expected Annual Energy Production AEP in kWh and the corresponding capacity factor Both are calculated by two methods a from the data distribution and b from the Weibull distribution If 12 month data exist then the AEP is calculated from the 12 distributions In the opposite case less than 1 year data the AEP is calculated from the general average distribution The two best direction sectors in wind energy contain The two best direction sectors 1n terms of time PAGE 17 S gt WINDROSE Wind Data Analysis software Note that the annual mean wind speed ca
46. lculation takes into account the various data completeness of each month during all the years of measurements Averaging the 12 months assures that the result a represents the 1 year reference period and b is not biased by data that fail to cover exactly 12 24 36 etc months Below an arithmetic example is given representing the described procedure STEP 1 STEP 2 Month Mean Data Month Mean Data Wind Speed Complet Wind Speed Completeness Jan 2003 MES 100 Jan 9 86 77 Feb 2003 720 100 Feb 7 64 99 Mar 2003 Cee 100 5 Mar 6 6 100 Apres 2003 Gee 60 Apr Uc 80 May 2003 oe 100 May ome 100 Jan 2003 6 8 100 Jun 6 45 Sas Jul 2003 Teal 99 m J l Tel 99 amp Aug 2003 Sq 90 Aug gum TOO Sep 2003 Cr 100 Sep 6 0 100 Oct 2003 6 4 100 Oct 6 4 100 Nov 2003 ST 100 S Nov S1 100 Dec 2003 Teg TS Dec Jus TS amp Jan 2004 ORS 54 Feb 2004 gue JS z Mar 2004 623 100 Apr 2004 820 100 Mean Annual Wind Speed 7 4 May 2004 Wise 100 9 86 644 6 6 36 6 3 6 45 Jun 2004 5 9 82 Plo7 85 art Sd 8 1 7 9 12 9 86 pcc Desde Why 3 values for the expected W T Energy production e The 1 value refers to the energy production of the wind turbine for the given time period while the other 2 refer to the annual production e The 1 value depends on the missing data as it is calculated only by the valid data within the given time period The other 2 values are the outcome o
47. ll as the total data distribution again per wind speed and wind direction PAGE 28 WINDROSE Wind Data Analysis software Turbulence Intensity vs Wind Speed amp Wind Direction SSE Turbulence intensity per wind speed and direction 3D worksheet 12pie In this 2 page worksheet 12 pie charts are plotted one per month of the year showing the two dominant wind directions At the end of the 2 page a table is also given with the numerical representation of the figures In case that a prevailed sector referring to time 1s not among the best two sectors referring to energy then a blue colour is used or a green colour in the opposite case As in the windRose sheet the calms are not included here too January February Main wind directions blue based on the time green based on the energy content 12diumal This one page sheet is composed by 12 bar plots presenting the Diurnal distribution of the data per month of the year all years included In fact it is the hourly variation of the wind speed per month all the directions taken into account Note that if the total diurnal distribution is needed it can be found in the Timechart sheet PAGE 29 Bar Charts Weibull WINDROSE Wind Data Analysis software Important Note e If data exceed l year period the monthly graphs cumulate data from different years When it is necessary to see the
48. mes isi nh tie m p EZ TaT Y 5 T B5 i GET S585 3 a D z a i m mou uu E i den ce os us eee un a an n c e ao Sse ppp fe 2 n 5 es f e UD yp M Hoe dp Upon me Bee E Ee gt gt T T wc T aes ae eee Sa T cclsiteT st0o sx 5913 1721 ese 25 4012 29 2 pde c siteliprocessed i txt TES 25 670 8 10 11 12 13 14 15 File names and column numbers sheet Input Delimiting characters within the ASCII data files It is expected that ASCII files contain some comment lines in the beginning followed by at least 5 columns of data The accepted characters delimiting separating the columns are the following e The space The comma The semicolon The 2 parentheses The tab The double quotation marks PAGE 6 WINDROSE Wind Data Analysis software Time format WindRose accepts as time format when reading ASCII files the following forms lala a Tm SS co aate mini ES 5 Utah S 5 lala Sica nos lola agi o d OE GEI Note that if none of the delimiting characters are present and 5 6 digits are found then the last two represent the seconds the middle two the minutes and the first one s the hour If 3 4 digits are found then the last two represent the minutes and the first one s the hour of the day Finally if just 1 2 digits are found then depending on the time step they are considered either as minutes if the time step lt Ih 1 e 5 is taken as 00 00 05 or as the hour of the day if the time s
49. mparison of anemometers Squared sum of deviations m s Anemometer at 30m m s 4 5 5 6 6 7 7 8 8 9 9 10 10 11 11 12 12 13 13 14 14 15 15 16 Wind speed bin m s The main result of the Annex K sheet PAGE 25 Time Charts WINDROSE Wind Data Analysis software The main result of the Annexk sheet is a complex graph with 4 axes In the blue axes x top y right the linear regression between the two anemometers is presented In the green axes x bottom y left the square sum of the systematic and statistical deviations per wind speed bin are plotted The in situ comparison 1s considered successful and a recalibration of the anemometers is not necessary if each bar remains lower than 0 1m s The examined condition can be written as follows 2 U lt 0 1m s bA where AU Uop a Uconrort p with a P being the regression coefficients for the initial period and N the number of data per each wind speed bin N gt 3 10min data The default and recommended parameters for this method are 1 select data from 6m s to 12m s 11 use intervals of 8 weeks for both the first and last part of the measurements and 111 apply the method to the main wind direction sector However these parameters can be modified within the Advanced Options of the WindRose menu Three important figures are presented in this worksheet displaying the results of the wind data analysis in a global form More detailed vi
50. nalysis software In In Q u respectively then the Weibull distribution becomes a straight line with slope k and intercept point to Y axis the quantity k n C from which the C parameter can be calculated Note that in the znput sheet it 1s given the possibility to restrict the wind speed range in which the fit 15 performed so that extreme values accentuated by the logarithms will not affect the results It is suggested to fit the data within the range 4 16m s Following the IEC 61400 12 recommendations the power curve is corrected depending the power control stall pitch control of the wind turbine The formulas used are the following m norm A T stall control Par norm 3 A T pitch control U U 2 Pais PAGE 52 WINDROSE Wind Data Analysis software INSTALLATION ISSUES WindRoseXLS Under certain configurations 1 e Office XP and or Windows XP Security Warning professional you may not be able to open the WindRose XLS file with the included macro enabled This will prevent WindRose to function properly The following steps need to be taken to overcome this situation The purpose is to end up with the WindRose macro to be among the Microsoft Excel s trusted sources Therefore open Excel without WindRose XLS go to Tools Macro Security Set it to Medium Security High Only signed macros From trusted sources will be allowed to run Unsigned macros are automatically disabled
51. nd Data Analysis software 7 Data Distribution Weibull Distribution 1 90 9 0 12 3 4 5 6 7 8 9 0 11 12 13 4 15 16 17 18 19 20 21 22 23 24 25 Wind speed m s Data distribution together with the calculated Weibull distribution that fits the data Some important information is included in the table below the graph Here the coefficients of the Weibull distribution are displayed per wind direction sector together with the corresponding percentage of the total data found in it the mean wind speed the turbulence intensity and the starting and ending values of the direction sector Some third party programs dealing with optimum wind turbine arrangement within a wind farm i e WindFarm by ReSoft require as an input this table s data as well as the site s topography z 5 Direction Angles deg prs ps Data Distrib PON E 11 25 33 75 33 75 56 25 56 25 78 75 78 75 101 25 101 25 123 75 123 75 146 25 146 25 168 75 168 75 191 25 191 25 213 75 213 75 236 25 236 25 258 75 258 75 281 25 281 25 303 75 303 75 326 25 326 25 348 75 348 75 11 25 Weibull distribution coefficients data percentage mean wind speed and turbulence intensity per direction sector In this version of WindRose the calculated Weibull coefficients per month are given only as values at TempData M109 0122 however future versions will incl
52. ntaining the data from the Reference and the Target site As pointed out earlier these files are generated by running twice WindRose before this step After running the Calculate Correlations command the windcorr sheet holds all the results in form of tables and graphs Optionally two files can be also created containing 1 the concurrent pair of data and 11 the predicted time series see S below Output Files The remaining input parameters are e The initial time shift of the 2 time series used to calculate the best possible correlation coefficient see below Methodology e A Boolean value Yes No indicating whether or not an hourly averaging will occur before processing the data Often the hourly averaging provides better correlations The number of the wind speed bins The width of each wind speed bin e The minimum number of common hours that must exist per wind speed bin and direction sector in order to perform reliable calculations recommended 2hours 1 e if the time step is 10min 12 data e The number of the direction sectors 8 12 or 16 Recommendation When correlation results are not satisfactory try less direction sectors wider wind speed bins and hourly averaging 10 Por nearby sites 2 hours is sufficient provided that the clocks are synchronized 11 Note that this number has nothing to do with the one set during the WindRose analysis PAGE 37 WINDROSE Wind Data Analysis software Method
53. o run the program for the first time see below RUNNING THE PROGRAM EA Microsoft Excel WindRnse xls File Edit wiew Insert Format Tools Data Window osha Seay x3 WindRase Adobe PDF Help Run HE Calculate Correlations Results in English Clear Data Import 5 Advanced Options Enter registration code Running WindRose within Excel Microsoft Excel has security levels that allow users to run macros based on whether or not they are digitally signed by a trusted macro developer The WindRose XLS file includes some macros and has a digital signature named WindRose The first time that the WindRose XLS file is opened the user is asked whether or not the macros should be enabled Answer yes so that WindRose runs properly and never asked in the future the same question If you are not allowed to open the WindRose xLS file with the included macros enabled might happen if the Security Level of Excel is set to High then consult the Appendix NSTALLATION ISSUES PAGE 3 Running the program for the first time WINDROSE Wind Data Analysis software RUNNING THE PROGRAM Three sample data files ST FLASH TXT NOMAD TXT and GenASClI TXT containing wind data are included in the package to facilitate the first run of the program Moreover the Input worksheet contains already all the required information to run the wind data analysis for the specific data files All the user ha
54. ology Consider correlating measured data from a Reference site long term data and a Target site short term or incomplete data Sometimes nearby sites log wind data with a time lag depending on their distance and their data loggers clock settings This time lag can be determined by examining the correlation coefficient for various time shifts i e 2h Ih 50min 0 1h 50min 2h of the Target time series in respect to the Reference ones The time shift at which the maximum value of the correlation coefficient 1s found is then inserted into the Target time series by adding it to its time date array and displayed at the Table 2 of the windcorr sheet Input File with the Reference time series c data ref bin Input File with the Target incomplete time series c data target bin Output File for the concurrent pairs of data optional c data sync txt Output File for of the predicted time series optional c data pred txt Locate best correlation by shifting Target series up to hours Transform time series to 1 hour data v Number of wind speed bins Width of wind speed bin m s Min hours of data per bin Number of Dir sectors Correlation Method bcm v The input parameters of the WindCorr Sheet Y Time shift values close to zero show that wind phenomena occur simultaneously to both sites Global correlation coefficients close to 12 This step assures t
55. r shadow on the lower anemometers assuming that the reference anemometer is installed on the top of the mast where the wind flow is undisturbed no matter the wind direction This is achieved by plotting the wind speed ratios above a threshold 4m s PAGE 23 Wind speed at 76m WINDROSE Wind Data Analysis software Tower Influence Anemometer at 40m us Anemometer at 50m Influence of the mast tower shadow on the lower anemometers Polar plot of the wind speed ratio of a lower anemometer versus the top one worksheet Shear In the above example note that at 300 the anemometers ratio falls around 0 5 providing valuable information about a the exact orientation of the boom and b the boom length sufficiency according to the standards A less pronounced drop of the ratio is also observed as expected diametrically at 120 Anemometer at 76m vs Anemometer at 100m Wind vane 1 vs Ref wind vane Main direction only N and U 2m s 8001 points U 4mis 19214 points EENE MEN A MEAN Wind direction de 90 135 180 225 270 315 360 Wind direction deg 5 10 15 20 25 30 Wind speed at100m Comparison of anemometers and vanes mounted on the same Mast worksheet Shear These graphs are automatically created for all the low anemometers in respect to the reference one without any user interaction However PAGE 24 AnnexK WINDROSE Wind Data Analysis software since Excel cannot handle more
56. relation results exist and consequently the prediction will be based on the global correlation results per wind speed bin no matter the direction e Again assume that only during October another missing data period extremely high wind speeds gt 25m s occurred from the NNW direction However correlation results exist for the NNW direction but only up to 18m s Therefore the predicted values of these high speeds will be based on the global NNW correlation results 1 e all the wind speeds above calms from that sector Mean Number of pen oletion General Data start at Data end at Wind Speed concurrent Max Wind at amet Bacertalnty regression eia shift min for the speeds m s data Speed Tarcet ite coefficients Correlation g Number of data coefficient 4464 1 2 2000 2 3 2000 23 50 6 91 m s slope offset m s R2 0 926 0 11 0 876 0 162 0 856 4176 1 2 2000 0 00 29 2 2000 23 50 6 21 General results of the Correlations windCorr sheet Output Tables The 2 table of the windcorr sheet contains some statistics of the two input files and the global correlations results In the beginning are given e The total number of the data in file e The dates of the first and last data points e The mean wind speed Then in the next section are written e The number of the common data PAGE 40 WINDROSE Wind Data Analysis software e The maximum correlation coefficient after the time shifting of th
57. rve table sheet PowerCurve Right Its graphical representation including the correction due to the site elevation height sheet Input PAGE 14 WINDROSE Wind Data Analysis software QUICK CHECK OF THE DATA Once the input sheets are filled correctly the user can launch the WindRose by selecting Run in English or Run in Greek depending on the language in which the results will be displayed Execution time depends on the amount of data normally a Pentium 4 PC will take 7 minute to run l year data When the execution is finished the figure below will appear displaying all the processed valid data This is an important feature of the program since Excel graphs cannot contain unlimited amount of data and the user interaction zooming scrolling is limited Select part of the Ev data for detailed view e Wind Speed m s c c 5 071 n 1 1 2000 3 12 17 5 1 2000 10 15 56 9 1 2000 17 19 35 14 1 2000 0 23 15 18 1 2000 7 26 5 Time View Mode ZoomX amp Y v Wind Direction C ZoomX Click the m ouse REPLOT Close SDV Wind Speed C PanX amp Y button on the graph Hold it down and git v Wind Speed Change of the View Mode Gust Wind Speed C PanX Detailed view of all the processed data Note that a zoom operation is possible by selecting a region with the mouse b scrolling panning is achieved by holding the mouse down and dragging it to the desired direction
58. s to do is to select from the Excel Main Menu WindRose and then Run The status bar of the Excel should then display the different steps of the program s execution At the end the results of the analysis are stored in forms of graphs and or tables into the different worksheets Before calculating the correlations predictions between 2 sites WindRose has to be run twice once per each site The WindRose Excel workbook is composed by the following 24 worksheets dee Input Zr SUhOUEI I7 DIRhouErG 2 PowerCurve 10 TimeCharts Log DIBRBOULI 3 Results Tibe p 19 TempG 4 WindCorr T2 12000 mg 20 TempT Ju Tables 13 12pie 21 SradG 6 Weibull 14 BarCharts Z2 XOT Te Upolar 15 WtprodG 23 TempData e UhourG 16 WtprodT 24 Air Density Two worksheets Input and PowerCurve are used for defining all the necessary parameters to run the program The remaining worksheets are used for presenting graphically and numerically the results of the analysis The content of all the worksheets 1s explained in detail in the following paragraphs Note e When working with the original WindRose XLS keep it unchanged for future use by saving it at the end of the analysis using the Save As option with another name 2 f the default program path c Program Files NWindRose was changed during the installation then modify accordingly the paths of the 3 provided sample files in the Input sheet 3 During the program exec
59. tep gt Ih 1 e 5 1s taken as 05 00 00 Other custom formats 1 e the 12 hour form using am and pm e 1 50pm and the French style n i e 13h50 can be supported on demand Finally there are cases where time comes in the form of incremental steps from the beginning of the day Thus the time column is filled with integers increasing from to 144 time step 10min or from 1 to 1440 time step 1min etc In this case we add the character s right after the column number For example if the time column was the 9t We put 9s Date format WindRose defaults to the European date format but also accepts other date formats If the date has one of the following European formats then it is directly read ddmmyy ddmmyyyy d m yy dd mm yyyy d m yy OIC el VA Ote ny c elect vy The acceptable delimiting characters for the date format are If the American format is used mm dd yy mmddyyyy etc then right after the column number of the date we put the character u 1 e for the 10 column we write 10u If the international date format is used yymmdd yyyy mm dd etc then after the column number of the date we put the character i 1e for the 10 column we write 101 Measurements data coming from the NOMAD data logger use another date format I e for the 26 september 1998 it 1s written PAGE WINDROSE Wind Data Analysis software Sep 26 1998 When reading such files the above sequence should b
60. the site elevation height and to the temperature which is critical for the correct calculation of the wind energy e It produces monthly charts and tables with the per hour variation of wind speed wind direction expected wind turbine s power temperature and solar radiation How it works WindRose reads ASCII files containing columns of data an output format supported by the majority of data loggers Five columns of data are necessary for the program to run wind speed wind direction standard deviation of the wind speed time and date Several formats are supported when dealing with date and time Whenever temperature atm pressure humidity solar radiation flow inclination are recorded the appropriate analysis is performed If the wind speed s standard deviation is not recorded the program can still run using the following trick provide the same column numbers for both U wind speed and Usdv and then ignore the Turbulence Intensity results which will be constant to 100 PAGE 2 Digital Signature WINDROSE Wind Data Analysis software INSTALLATION Assuming that Microsoft Excel is properly installed WindRose can be installed using the SETUP EXE program Administration privileges are required in Windows NT4 2000 XP At the end of the installation the WindRose option is added in the main menu of the Excel Figure 1 By opening the WindRose XLS file default location c Program Files WindRose you are ready t
61. ude a graphical representation The 2 graph of the worksheet is the wind data accumulated probability based on the data distribution not on the Weibull distribution It is a useful tool to estimate the percentage of the time that the wind speed exceeds a specific value 1 e how much of the time the wind speed exceeds the cut in or the cut out speed of the wind turbine PAGE 31 UhourT UhourG DIRhourT DiIRhourG Wind Speed m s WINDROSE Wind Data Analysis software Cumulative Probability 100 90 4 80 4 70 4 60 50 40 30 4 20 Cumulative Probability 96 10 4 0 0123456789 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Wind speed m s Cumulative data probability i e 33 of the time the wind speed is higher than 10m The 3 graph shows the variation of the turbulence intensity over the wind direction The calculation is done using only wind speeds within the specified range in the Input worksheet usually 10m s 1m s Obviously if a direction sector has no data in the specified wind speed bin a zero 1s displayed The discussion held in the description of the Tables worksheet about averaging turbulence intensities applies here too Safe conclusions can be obtained only when significant amount of data are found in each direction sector These 2 worksheets contain monthly tables UhourT and graphs UhourG representing the evolution of the m
62. ut worksheet 8 12 or 16 As mentioned before each direction refers to the mean angle of the corresponding sector For example suppose that 16 sectors are selected the default and recommended value then the width of each sector is 22 5 360 16 and the North winds are all whose angle falls into the 348 75 11 25 sector The mean turbulence intensity per wind speed and direction bin 1s calculated using the classical averaging methods no correction takes place Although not precise this simple method provides a satisfactory approximation of the true turbulence intensity which could be calculated only if the detailed wind speed time series were available The number of successive rotations of the wind vane clockwise and counter clockwise is given in the 2 table These revolution numbers indicating the number of times that a wind turbine would have been rotated around itself provide a good estimate of the number of times that the electrical power cables are twisted For this calculation all the data including calms have been taken into account considering that even during low wind speeds the wind turbine usually rotates for an optimum orientation Obviously the accuracy of the two numbers depends on the data completeness low number of missing data This worksheet contains the 3D graphical representation of the Tables worksheet showing the turbulence intensity distribution per wind speed and direction as we
63. ution and under certain circumstances the PC may not respond fot a few seconds PAGE 4 Demo version Registration e Import OLD WindRose XLS files WINDROSE Wind Data Analysis software The demo version of the WindRose for unregistered users incorporates all the features of the full version but is limited to analysing measured data that do not exceed 31 days Registered users are given a code which is entered using the WindRose menu within Excel EA Microsoft Excel WindRose2 xls File Edit View Insert Format Tools Data Window Deis g vd GA Paste Special Arial 10 H7 U E zm Ss WindRose Help WE Run Calculate Correlations Adobe PDF Results in English File Names Wind Speed m s Wind Direction Wind Speed So Temperature C 3 AProgram FilesW indRasesl flashtxt 5 8 amp 4 2 1 4 Program Files WVindRose GenAscliltxt 1 2 3 4 5 b 5 Program FileswWyvindRasesMomad txt ded B 21n If the PC s OS is Windows Vista then before entering the code turn off the User Account Control Control Panel Security Center and turn it back on afterwards Excel files created with older versions of WindRose can be imported using the Import option of the WindRose menu see figure below Regarding older files this version of WiNDROSE xLs has some new input parameters and graphs therefore some graphs and tables may not display correctly im Import E X

WindRose User's Guide

Contents

Download Pdf Manuals

Related Search

Related Contents