modelwinds演唱会

WINDSIM–FLOWSIMULATIONSINCOMPLEXTERRAINASSESSMENTSOFWINDRESOURCESALONGTHENORWEGIANCOASTDr.
A.
R.
Gravdahl1)andDr.
K.
Harstveit2)1)VECTORAS,Industrigaten9,N-3124-Tnsbeg,Norway2)NorwegianMeteorologicalInstitute,ClimatologyDivisionP.
O.
Box32,Blindern,N-0313Oslo,NorwaySummaryWindSimisatailor-madesimulatorforpredictionoflocalwindfields.
WindSimisbasedona3DReynoldsAveragedNavierStokessolver.
Solvingthenon-lineartransportequationsformass,momentumandenergymakesWindSimasuitabletoolforsimulationsincomplexterrain,andinsituationswithcomplexlocalclimatology.
WindSimisusedintheassessmentofwindresourcesalongtheNorwegiancoast.
1.
WindEnergyinNorwayAlthoughthewindclimateinNorwayseemsfavourable,theexploitationofwindenergyinNorwayhassofarbeenverymodest.
However,duringthelastyearsthegeneralinterestforwindenergyhasbecomemorepronounced.
OnthisbackgroundtheNorwegianWaterResourcesandEnergyDirectoratelaunchedaprojectinspring1998,ontheassessmentofthewindresourcesalongtheNorwegiancoast.
2.
MesoscalemodelsBasedondigitalterrainmodelswithheightandroughnessdata,morethan120modelshavebeenestablishedcoveringtheNorwegiancoastfromthesoutherntipatLindesnesuptotheRussianborderinthenorth.
Fig.
1MesoscalemodelsalongtheNorwegiancoast.
Theapproximateextensionofeachmodelis1000km2.
ThewindfieldshavebeencalculatedwithWindSim,a3DReynoldsAveragedNavier-Stokessolver.
Toeachmodeltherehasbeenassociatedwindroses,extractedfromlong-termstatisticalweatherdatafrommeteorologicalstationsalongthecoast.
Intotalmorethan30meteorologicalstationshaveprovidedweatherdata.
Resultsfromatypicalmodelarepresentedinfigure2.
ThemodelistakenfromthesoutherntipofNorwayatLindesnes.
CurrentlythelargestwindmillparkinNorwayissituatedinthisregionatFjeldskr.
Fig.
2Annualmeanwindspeedat50metersheightinVest-Agdercounty,atypicalmodelalongtheNorwegiancoast.
3.
MicroscalemodelsSomeofthemeteorologicalstationsareinfluencedbyspeed-ups.
Particularmicromodelshavebeenestablishedforthesestationsinanattempttofilteroutmicroscaleeffectsfromthelong-termstatisticalweatherdata.
Amicroscalemodeltypicallycoversanareaintheorderof1to10km2.
4.
VerificationVerificationofthenumericalsimulationshasbeenanimportantissueintheproject.
Atvariouslocationswhereexperimentaldatafromseveralmeasurementmastshavebeenavailable,themeasuredspeed-upshavebeencomparedagainstthesimulations.
Thespeed-upisdefinedasthespeedatonelocationdividedagainstthespeedattheotherlocation.
4.
1MesoscaleInthesouthernpartofNorwayinRogalandcounty,measurementsfromthemeteorologicalstationatObrestadsituatedalongthecoastlinewascomparedagainstmeasurementsfromamastintheinlandplateau,"hyJren",atapproximately200metersheight.
Thedistancebetweenthetwomastswas12km.
Speed-upsarepresentedinfigure3.
Themodelcapturesthesignificantmaximumandminimuminthespeed-upforthewinddirectionswest(270deg)andeast(90deg),whichistowardsandfromthecoastline.
0.
090.
0180.
0270.
0360.
0Winddirection(deg)0.
000.
250.
500.
751.
001.
251.
50Speedup()Measurements,meanMeasurements,90%percentileSimulationsFig3.
Speed-upbetweentwomeasurementmastsinRogalandcounty,versusincomingwinddirectionat30metersheight.
4.
2MicroscaleSimilarcomparisonshavealsobeenmadeformicroscalemodels.
ThemeteorologicalstationatFruholmeninFinnmarkcounty,wellknownforitsseverspeed-up,wasmovedtoalessexposedlocationin1989.
Thetwolong-termmeasurementseriespriorandpostto1989constituteagoodverificationcaseonthemicroscale.
Theterrain,withatypicalwindfieldfromsouth,ispresentedinfigure4.
Thepinsmarkthetwopositionsforthemeteorologicalstation.
Thedistancebetweenthepositionsisapproximately200meters,andthedifferentinheightelevationis11meters.
Fig4.
MicromodelforFruholmenwithtypicalwindfieldfromsouth.
Thetwopinsmarkthemeteorologicalmastposition1and2,withposition1atsouthwest.
Thenumericalmodelreproducesthelocalwindconditionsascanbeseeninfigure5.
Thegridresolutioninthemicroscalemodelis10x10meter.
Figure6and7givesthespeed-upatthetwolocationswithreferencetospeedaboveopensea,thefourgraphsrepresentvaluesfromneighbourpointsinthecomputationalgrid.
Evenwiththisrefinedmodelagridindependentsolutionhasnotbeenestablished.
Thecomparisoninfigure5isbasedonthemeanvalueofthefourneighbourpoints.
0.
090.
0180.
0270.
0360.
0Winddirection(deg)0.
800.
901.
001.
101.
201.
301.
401.
501.
60Speedup()MeasurementsSimulationsFig5.
Speed-upbetweentwomeasurementmastsatFruholmenversusincomingwinddirectionat10metersheight.
0.
090.
0180.
0270.
0360.
0Winddirection(deg)0.
800.
901.
001.
10Speedup(U10/U10_sea)MeanNeighbourpointsFig6.
Griddependency,Speed-upwithreferencetoopenseaforneighbourpointsatmastposition1.
0.
090.
0180.
0270.
0360.
0Winddirection(deg)0.
901.
001.
101.
201.
30Speedup(U10/U10_sea)MeanNeighbourpointsFig7.
Griddependency,Speed-upwithreferencetoopenseaforneighbourpointsatmastposition2.
5.
BasisWindSimisbasedona3DReynoldsAveragedNavierStokessolver.
Solvingthenon-lineartransportequationsformass,momentumandenergymakesWindSimasuitabletoolforsimulationsincomplexterrain,andinsituationswithcomplexlocalclimatology.
Differentturbulentclosuresareavailable,thetwo-equationk-εmodelhavebeenusedinthemodelpresentedherein.
Thegridgenerationusessocalledbodyfittedco-ordinates,followingtheterrain,withrefinementtowardstheground.
Formoreindepthinformationaboutthemethodology,see[1].
6.
PublicationResultsfromtheNorwegianwindresourceassessment,micromodellingandverificationsarepublishedontheInternet,http://windsim.
com7.
References[1]GravdahlA.
R.
,"MesoScaleModellingwithaReynoldsAveragedNavier-StokesSolver–AssessmentofwindresourcesalongtheNorwegiancoast",InternationalEnergyAgencyAnnexXI,RisDenmark,October1998.