Performance Exhibit Attachments
Performance Exhibit Attachments
Attachment A
SkySpecs Performance Setup Guide
Contents
1.Introduction…………………………………………………………………… 3
2.ScopeofSkySpecsPerformanceAnalysis…………………………. 3
3.DataTransfer………………………………………………………………… 3
3.1.Option1:SourcePlatform API………………………………….. 4
3.2.Option2:CSVFileTransfer……………………………………… 5
4.InputDataRequirements………………………………………………… 5
5.SupportedDeviceTypes………………………………………………….. 8
6.AnalysisScheduling……………………………………………………… 14
7.ResultsDelivery……………………………………………………………. 14
8.Glossary……………………………………………………………………… 15
9.References………………………………………………………………….. 15
1.Introduction
TheSkySpecsPerformanceanalysistoolsdesignedtoprocesswindturbineoperationaldataand producediagnosticresultsandadvisoryinformation.Thefollowingdocumentdescribesthesteps that must be taken to implement the system and the requirements that must be fulfilled in order for the system to function correctly.
2.ScopeofSkySpecsPerformanceAnalysis
ThescopeoftheSkySpecsPerformanceAnalysis can besummarisedasfollows:
- Performance analysis: detailed checks of turbine power curves including advanced correction and filtering techniques. Combination of methodology from the IEC61400-12 standard, internally developed physics-based methods and machine learning to produce highly accurate power curves. Use of A.I. driven techniques to generate diagnostic information and recommendations.
- Control system checks: analysis of the overall turbine efficiency as well as a detailed assessment of the pitch drive and yaw drive, checking issues such as static and dynamic response, northing and alignment. Identification of outlying behaviour using machine- learningbasedmodelstrainedonextremelylargedatasets,todeliverrecommendationsfor turbines requiring corrective action to produce optimised performance.
- Componenthealthchecks:assessmentofthe currenthealthstatusofmajorcomponents such as main bearing, gearboxes, generator bearings, generator electrical system, transformer electrical system, yaw drives, pitch drives, blades. Furthermore, predictive methods will be used to identify components with elevated risk of future failure based on system modelling and trend analysis. Subject todata availability, a range of signals will be assessed and combined for the health checks.
Differentanalysistoolsrequiredifferentinputdata.Detailsoftheoptionsfordatatransfer,required signals and supported turbine types are provided below.
3.DataTransfer
Inorderforthesystemtofunctioncorrectly,operationaldatamustbetransmittedtotheSkySpecs Performancesystem containing specific Signals and in a format that must be consistent withthe SkySpecs standard. The input data requirements are presented below in Section 4.
UsersmaychoosefromtwopossiblesolutionsforsubmittingdatatotheSkySpecsPerformance system. The first, and preferred approach, is via an API interface provided by the data source (typically a SCADA system, second-level SCADA system or data platform) and allowing the SkySpecs Performance system to request and receive the required data.
The second option is that CSV files (comma separated variable) may be generated either automatically or manually and upload to an SFTP server, from where the files are then automatically imported and processed by SkySpecs. Further details and requirements for these two approaches are provided below.
3.1.Option1:SourcePlatformAPI
Many modern data storage platforms such as first and second level SCADA systems provide a standardised interface, or API through which data can be requested and extracted from the underlying databases. The SkySpecs Performance systemhas beendesigned toconnecteasily tosuchAPIs, andSkySpecs willsupporttheestablishmentof anewinterfacetoany systemthat satisfied the following requirements:
- All information must be provided to allow access to the data storage, including user access credentials,theaddressthroughwhichthesystemcanbeaccessedonline,andanyspecific access key or other security measures used to restrict access.
- Full documentation must be provided describing how a connection should be established to the API and describing in what format requests for data should be submitted.
- Data in the storage must use a consistent naming convention, so that Signals may be correctly identified independent of the specific Site, turbine manufacturer or project to which they belong. For each Signal, a single reference name must be used independently of the source Site or Device.
- Each item of information in the database (e.g. time series data, alarms, events) must be assigned to a single Device and each Device must be identified by a unique name or code.
- EachindividualDevicemustbeassignedtoasingleSite(windpark,windfarm,windpower plant). Multiple Devices may be assigned to a single Site.
- The data storage system and the API must be continuously online, with a total yearly availability of at least 99.5%.
- The response time of any request sent to the API may not exceed 20 seconds for a single Device with a single Signal consisting of records with a resolution of 10-minutes, and for an overall Period of 1 month.
- If an analysis is to be performed for a large number of Devices and / or Signals, then the overallresponsetimeofasinglerequestsenttotheAPImaynotexceed24hours.Therefore, for very large analysis tasks, the API must be capable of parallel processing, such that this overall time limit is not exceeded.
- Any single data transfer task performed via the API must be completed without interruption. In case interruption does occur, the affected analysis task will not be completed and SkySpecswilldeliveranerrormessage.Sucheventsarenottheresponsibilityof SkySpecs.
Assuming the above requirements are satisfied, SkySpecs must be provided with all relevant inputs (scope of expected analysis, API documentation, list of Sites and Devices for which analysis is to be performed), then SkySPecs will implement the required data interface. The implementation will typically be complete within two weeks, with a maximum completion time of onemonth.SkySpecsmaychoosenottoimplementaninterfacetoaspecificplatformincasethe technical complexity is unusually high.
SkySpecswillnotberesponsibleforresolvinganysecurity-relatedissuesrelatingtodataaccess. This includes the creation of user accounts, managing data encryption solutions, changing of firewall settings, opening of access ports, investigation into valid data mapping. All such topics must be managed by the provider of the source platform.
3.2.Option2:CSVFileTransfer
AsanalternativesolutiontothedirectAPIconnectiondescribedabove,theuseroftheSkySpecs Performance system may instead prefer to generate files ina CSV formatand upload these to a securefilestoragelocationprovidedbySkySpecs.Inthiscaseitisimportantthatthecontentand structure of the files is consistent with the SkySpecs requirements. Note that for or each Signal, a single reference name must be used independently of the source Site or Device. For further information concerning this approach, see Ref 1.
In addition to uploading time-series data via theCSV File Transfer solution, alarm logs may also optionally be uploaded to the SkySpecs Performance System. For details of the required file format see Ref 2.
4.InputDataRequirements
In order for the SkySpecs Performance system to function properly, specific input Signals must be available. These Signals must be in the form of SCADA 10-Minute logs, consistent with the current standard for operational data logging used in the wind industry.
Inordertoensurethebestpossibleaccuracyofresults,datashouldbeprovidedwhichisofhigh quality.Signalsshouldfreeofgapsoroutliersandshouldaccuratelyrepresentthetrueoperating conditions of the turbine. Although the SkySpecs Performance system includes tools that are capable of automatically identifying quality issues in the data, SkySpecs does not accept any responsibility forthequality ofthedataandalso does notperformanymanipulationorcorrective actioninordertoremoveanydetectederrors.Therefore,itisimportantthattheenduserisaware of any accepts responsibility for providing input data of the highest possible quality.
The input Signals required by the SkySpecs Performance system are shown in Table 1, as well as the scope of analysis for which the Signal is used. Some of the Signals shown are not mandatory but should be made available for best overall system performance. Note that timestamps (TimeStamp, optionally also CMS_TimeStamp) must be provided in precise 10- minute intervals.
i4SEESignalName | Description | AnalysisScope |
TimeStamp | 10-minute timestamp [DD.MM.YYYY HH.MM.SS] | All |
PerformanceAnalysis | ||
WNAC_ExTmp_Avg | Averageambienttemperature [°C] | ComponentHealthChecks |
WNAC_IntnTmp_Avg (optional) | Averagenacelletemperature[°C] | Performance Analysis ComponentHealthChecks |
WTUR_W_Avg[kW] | Average power | Performance Analysis ComponentHealthChecks Control System Checks |
WTUR_W_Min (optional) | Minpower [kW] | Performance Analysis ComponentHealthChecks Control System Checks |
WNAC_Dir_Avg | Average nacelledirection[deg.] | Performance Analysis ComponentHealthChecks Control System Checks LoadsandReliability |
WNAC_WdDir_Avg | Averagewinddirection[deg.] | PerformanceAnalysis ControlSystemChecks Loads and Reliability |
WNAC_PriWdSpd_Avg | Averagewindspeed,primary sensor[m/s] | Performance Analysis ComponentHealthChecks Control System Checks Loads and Reliability |
WNAC_PriWdSpd_Sdv (optional) | StandardDeviationwindspeed,primary sensor [m/s] | PerformanceAnalysis ControlSystemChecks Loads and Reliability |
WNAC_SecWdSpd_Avg (optional) | Averagewindspeed,secondary sensor[m/s] | PerformanceAnalysis |
WROT_RotSpd_Avg | Averagerotorspeed[rpm] | Performance Analysis ComponentHealthChecks Control System Checks LoadsandReliability |
WROT_PtAngBl1_Avg (optional) | Averagepitch angle,blade1[deg.] | ControlSystemChecks LoadsandReliability |
WROT_PtAngBl1_Sdv (optional) | Stdevpitchangle,blade1[deg.] | ControlSystemChecks LoadsandReliability |
WROT_BrgGtmTmp1_Avg | Averagemainbearingtemperature,driveend [°C] | ComponentHealthChecks |
WROT_BrgGtmTmp2_Avg (optional) | Averagemainbearingtemperature,non- drive end [°C] | ComponentHealthChecks |
WGEN_GnBrgTmp1_Avg | Averagegenbearingtemperature,drive-end [°C] | ComponentHealthChecks |
WGEN_GnBrgTmp2_Avg | Averagegenbearingtemperature,non-drive end [°C] | ComponentHealthChecks |
WGEN_GnWiTmpPh1_Avg (optional) | Averagegenwinding temp,phase 1[°C] | ComponentHealthChecks |
WGEN_GnWiTmpPh2_Avg (optional) | Averagegenwinding temp,phase 2[°C] | ComponentHealthChecks |
WGEN_GnWiTmpPh3_Avg (optional) | Averagegenwinding temp,phase 3[°C] | ComponentHealthChecks |
WTRF_WiTmpPh1_Avg (optional) | Averagetransformerwindingtemp,phase1 [°C] | ComponentHealthChecks |
WTRF_WiTmpPh2_Avg (optional) | Averagetransformerwindingtemp,phase2 [°C] | ComponentHealthChecks |
WTRF_WiTmpPh3_Avg (optional) | Averagetransformerwindingtemp,phase3 [°C] | ComponentHealthChecks |
WTRM_GbxOilTmp_Avg | Averagegearboxoiltemperature[°C] | ComponentHealthChecks |
WTRM_GbxBrgTmp_Avg | Averagegearbox bearingtemperature [°C] | ComponentHealthChecks |
WTRM_GbxParticleCount_Tot (optional) | Totalparticlecountin10-minuteinterval[-] | ComponentHealthChecks |
CMS_Tag (optional) | Referencetagidentifyingspecificfrequency range | ComponentHealthChecks |
CMS_TimeStamp (optional) | 10-minutetimestampatwhichaCMS measurement was performed | ComponentHealthChecks |
CMS_RMS (optional) | RMSvalue,foreachspecificCMSTag& TimeStamp | ComponentHealthChecks |
CMS_Power (optional) | AveragepowerforeachCMSTag& TimeStamp | ComponentHealthChecks |
Table1:InputSignalsrequiredbySkySpecsPerformance
5.SupportedDeviceTypes
ForanyanalysisrequestsubmittedtotheSkySpecsPerformancesystem,theDeviceTypemust be identified using the standard SkySpecs naming convention:
[Manufacturer Name][TurbineType Description][TurbineRating, inKW]
The three entries are separated by a space, and each entry may be a combination of numbers, letters or special characters. The Device Types which have to date been successfully analysed to date using SkySpecs Performance are listed in
Table2.
Reference | DeviceType |
ID1 | Acciona AW1323300 |
ID2 | Acciona AW1403000 |
ID3 | AccionaAW771500 |
ID4 | AlstomECO801670 |
ID5 | ArevaM5000-1165000 |
ID6 | Bonus44 600 |
ID7 | Bonus541000 |
ID8 | Bonus541300 |
ID9 | Bonus621300 |
ID10 | Bonus742000 |
ID11 | BonusB44 600 |
ID12 | BonusB822300 |
ID13 | DewindD61250 |
ID14 | EnerconE-1013000 |
ID15 | EnerconE-1013050 |
ID16 | Enercon E-1013075 |
ID17 | Enercon E-1032350 |
ID18 | EnerconE-1152990 |
ID19 | EnerconE-1153000 |
ID20 | EnerconE-1153050 |
ID21 | EnerconE-1153200 |
ID22 | EnerconE-1154200 |
ID23 | EnerconE-1264000 |
ID24 | Enercon E-1264200 |
ID25 | EnerconE-1266200 |
ID26 | EnerconE-1267580 |
ID27 | Enercon E-126b4000 |
ID28 | Enercon E-1384200 |
ID29 | Enercon E-138b3500 |
ID30 | EnerconE-1414200 |
ID31 | EnerconE-40 600 |
ID32 | EnerconE-40b500 |
ID33 | EnerconE-48 800 |
ID34 | EnerconE-53 800 |
ID35 | EnerconE-581000 |
ID36 | Enercon E-58b1000 |
ID37 | EnerconE-661500 |
ID38 | EnerconE-661800 |
ID39 | EnerconE-662000 |
ID40 | EnerconE-701500 |
ID41 | EnerconE-702000 |
ID42 | EnerconE-702040 |
ID43 | EnerconE-702050 |
ID44 | EnerconE-702300 |
ID45 | EnerconE-702310 |
ID46 | EnerconE-822000 |
ID47 | EnerconE-822050 |
ID48 | EnerconE-822300 |
ID49 | EnerconE-822350 |
ID50 | EnerconE-823000 |
ID51 | EnerconE-922000 |
ID52 | EnerconE-922300 |
ID53 | EnerconE-922350 |
ID54 | eno822000 |
ID55 | FuhrlanderFL 2500 |
ID56 | FuhrlanderFL701500 |
ID57 | FuhrlanderFL771500 |
ID58 | GamesaG1142000 |
ID59 | GamesaG1142100 |
ID60 | GamesaG1142625 |
ID61 | GamesaG1262625 |
ID62 | GamesaG1323465 |
ID63 | GamesaG58850 |
ID64 | GamesaG802000 |
ID65 | GamesaG902000 |
ID66 | GamesaG972000 |
ID67 | GE1.5-S1500 |
ID68 | GE1.5-SL1500 |
ID69 | GE1.5SLE1500 |
ID70 | GE1.5XLE1500 |
ID71 | GE2.5-120 2500 |
ID72 | GE2.5XL2500 |
ID73 | GE2.752750 |
ID74 | GE2.85-1032850 |
ID75 | GE5.5-158 5500 |
ID76 | GEGE-1001790 |
ID77 | GEGE-1002850 |
ID78 | GEGE-1032500 |
ID79 | GEGE-1033200 |
ID80 | GEGE-1162300 |
ID81 | GEGE-1202500 |
ID82 | GEGE-1202750 |
ID83 | GEGE-1585300 |
ID84 | GETW600-43 600 |
ID85 | GETW600a-46600 |
ID86 | Goldwind GW1002500 |
ID87 | Goldwind GW1092500 |
ID88 | HSW571000 |
ID89 | LeitwindLTW801650 |
ID90 | MiconM23001000 |
ID91 | MiconM530-26 250 |
ID92 | Multibrid1165000 |
ID93 | NEGMicon600 |
ID94 | NEGMiconM1500-43600 |
ID95 | NEGMiconNM601000 |
ID96 | NEGMiconNM72C1500 |
ID97 | NEGMiconNM72C2000 |
ID98 | NEGMiconNM821500 |
ID99 | NordexN1002500 |
ID100 | NordexN1172400 |
ID101 | NordexN1173000 |
ID102 | NordexN1173600 |
ID103 | NordexN1313000 |
ID104 | NordexN1313600 |
ID105 | NordexN1494500 |
ID106 | NordexN29250 |
ID107 | NordexN541000 |
ID108 | NordexN601300 |
ID109 | NordexN621300 |
ID110 | NordexN802500 |
ID111 | NordexN902300 |
ID112 | NordexN902500 |
ID113 | NordexS701500 |
ID114 | NordexS771500 |
ID115 | NordtankNTK31 300 |
ID116 | NordtankNTK43 600 |
ID117 | NordtankNTK641500 |
ID118 | RepowerMD701500 |
ID119 | RepowerMD771500 |
ID120 | RepowerMM702000 |
ID121 | RepowerMM822000 |
ID122 | Senvion3.0M1223000 |
ID123 | Senvion3.2M1143170 |
ID124 | Senvion3.2M1143200 |
ID125 | Senvion3.4M1043370 |
ID126 | Senvion3.4M1143370 |
ID127 | Senvion4.2M1184200 |
ID128 | Senvion48 600 |
ID129 | Senvion5M1265000 |
ID130 | Senvion6.2M1266200 |
ID131 | Senvion6.2M1526330 |
ID132 | SenvionMD701500 |
ID133 | SenvionMD771500 |
ID134 | SenvionMM702000 |
ID135 | SenvionMM822000 |
ID136 | SenvionMM822050 |
ID137 | SenvionMM922000 |
ID138 | SenvionMM922050 |
ID139 | SiemensGamesa1323300 |
ID140 | SiemensGamesa1323400 |
ID141 | SiemensGamesaSG-5-1706200 |
ID142 | SiemensGamesaSG-5-170 6300 |
ID143 | SiemensGamesaSG-5-1706400 |
ID144 | SiemensGamesaSG-6.6-170 6600 |
ID145 | SiemensSWT-1.3-621300 |
ID146 | SiemensSWT130-3.63600 |
ID147 | SiemensSWT142-4.14100 |
ID148 | SiemensSWT-2.3-1012300 |
ID149 | SiemensSWT-2.3-1082300 |
ID150 | SiemensSWT-2.3-822300 |
ID151 | SiemensSWT-2.3-932300 |
ID152 | SiemensSWT-3.03000 |
ID153 | SiemensSWT-3.0-1133000 |
ID154 | SiemensSWT-3.2-1133200 |
ID155 | SiemensSWT-3.6-1073600 |
ID156 | SiemensSWT-3.6-1203600 |
ID157 | SiemensSWT-3.6-DD3600 |
ID158 | SiemensSWT-6.0-1546300 |
ID159 | SiemensSWT-DD-130-4.34300 |
ID160 | SuedwindS771500 |
ID161 | SuzlonS881800 |
ID162 | TackeTW1.5s-701500 |
ID163 | TackeTW600-43600 |
ID164 | Vensys1002500 |
ID165 | Vensys70-771500 |
ID166 | VestasV1001450 |
ID167 | VestasV1001800 |
ID168 | VestasV1001950 |
ID169 | VestasV1002000 |
ID170 | VestasV1002200 |
ID171 | VestasV1053600 |
ID172 | VestasV1102000 |
ID173 | VestasV1102200 |
ID174 | VestasV1123000 |
ID175 | VestasV1123075 |
ID176 | VestasV1123300 |
ID177 | VestasV1123450 |
ID178 | VestasV1173300 |
ID179 | VestasV1173450 |
ID180 | VestasV1173600 |
ID181 | VestasV117-23450 |
ID182 | VestasV1263300 |
ID183 | VestasV1263450 |
ID184 | VestasV126-23450 |
ID185 | VestasV1263800 |
ID186 | VestasV126-23450 |
ID187 | VestasV1363450 |
ID188 | VestasV1363600 |
ID189 | VestasV1363800 |
ID190 | VestasV1364200 |
ID191 | VestasV1504200 |
ID192 | VestasV1504300 |
ID193 | VestasV1505600 |
ID194 | VestasV1625600 |
ID195 | VestasV1626000 |
ID196 | VestasV1626200 |
ID197 | VestasV1649500 |
ID198 | VestasV42600 |
ID199 | VestasV42650 |
ID200 | VestasV47660 |
ID201 | VestasV47705 |
ID202 | VestasV52850 |
ID203 | VestasV802000 |
ID204 | VestasV901800 |
ID205 | VestasV902000 |
ID206 | VestasV903000 |
ID207 | VestasV39500 |
ID208 | WindworldWW-52750 |
Table2:Device Typenaming,accordingtotheSkySpecsconvention
Analysis may also be performed for Device Types not listed in the table, however the machine learningalgorithmsusedbySkySpecsPerformancewillfirstrequiretraining,forwhichaminimum volume of data is required. Furthermore, it is important to note that for specific Device Types it may not be possible to apply one or more of the SkySpecs Performance analysis tools, for example due to unsuitable hardware configuration of the turbine or specific type dependent behavioural characteristics which differ from the Device Types already assessed.
6.AnalysisScheduling
Once a connection to the SkySpecs Performance analytics system has been established as describedabove,anddataisreadytobeautomaticallytransferredintherequiredformat,analysis tasks can be triggered. SkySpecs will implement a scheduling service to automatically trigger defined analysis tasks, according to the agreed scope of supply. The scheduling service will ensure that results are generated at the start of each month and made available for delivery to the customer.
7.ResultsDelivery
Once an analysis task has been completed, results will be discussed with the customer in a monthlyfeedbackmeeting.Furthermore,anautomateddeploymentsystemwillbeusedtodeliver reports to the customer via a range of optional channels, including:
1)FullreportstransferredtoasecurefileserversuchasMSOneDrive,wheretheycanbe accessed by specific stakeholders within the customer organisation.
2)Individual“cards”createdforeachpriorityissueandautomaticallyinsertedintocommercially available planning tools including Trello, Atlassian JIRA, MS Planer.
SkySpecsPerformanceistypicallycapableofprovidingacustomsolutionforresultsdelivery,to suit individual customer preferences.
8.Glossary
TheterminologycommonlyusedwithintheSkySpecsPerformanceanalyticsandsupporting documentation is defined in Table 3 below:
Term | Definition |
Device | Anindividualwind turbine. |
Event | Asingleobservation,suchasawarningoralarm,occurringataspecificpointin time and associated with a single Device. |
Execution | A single activation of a SkySpecs Performance analysis, processing data for a definedPeriod(consistingofoneorseveralIntervals)andforoneormoreSite(s) and Device(s). |
Interval | Aunitoftimewithinwhichasingleanalysisisperformed.TheIntervalmay(but must not) be shorter than the Period. |
Metric | Aspecificmeansofmeasuringthebehaviourofasignal,typicallyresultingina Result. |
Period | Aunit oftimewithinwhichabatchof analysistasksareperformed. ThePeriodis definedwithastarttimeandanendtimeandmay(butmustnot)includemultiple Intervals to be analysed in a single Execution. |
Result | Asingle,scalarvalueresultingfromaspecificcalculationperformedbyananalysis task. |
Signal | Aseriesofmeasurementdata,usuallyrepresentedbyavectorconsistingof multiple entries recorded with a fixed resolution. |
Site | Asinglegeographicallocationuponwhichawindpowerplantislocated,consisting of one or several Device(s). |
Table3:SkySpecsPerformancecommonlyused terminology
Attachment B
SkySpecs Performance
CSV File Upload Guide
Contents
1.Introduction…………………………………………………………………… 2
2.FileStructure…………………………………………………………………. 2
3.SupportedSignal Names…………………………………………………. 5
4.SupportedDeviceTypes…………………………………………………. 7
5.FileUploadProcedure…………………………………………………….. 7
6.Glossary……………………………………………………………………….. 9
7.References……………………………………………………………………. 9
1.Introduction
TheSkySpecsPerformanceanalysistoolsaredesignedtoprocesswindturbineoperationaldata and produce diagnostic results and advisory information. In order for the system to function correctly,datamustbeprovidedinaformatthatmustbeconsistentwiththeSkySpecsstandard. Users may choose from two possible solutions for submitting data to the system. The first, and preferred approach, is via an API interface, connecting the data source (typically a SCADA system)totheSkySpecsPerformancesystem.Forfurtherinformationconcerningthisapproach, see Reference [1]. The second option is to generate CSV (comma separated variable) files and uploadthesetoanFTPserver,fromwherethefilesareautomaticallyimportedandprocessedby SkySpecs.
ThisdocumentprovidesdetailsoftheCSVFileUploadsolution,includingtherequiredfilecontent and structure, as well as the procedure for uploading the files to the FTP server.
2.FileStructure
The CSV upload solution requires files to be prepared in a defined format and structure. It is anticipated that such file generation should be performed by an automated service, which regularly extracts data from the source (typically a SCADA database) and uploads to an FTP server (refer to Section 5 below).
An example of a file that has been prepared with the correct format is shown in Table 1, and generalrequirementsfortheformatandstructureoffilesaresummarisedinTable2.Thespecific requirements for the content of each column in the file are detailed in Table 3. All of these guidelines must be satisfied for error-free upload of CSV files to the SkySpecs Performance system.
![]() |
Table1:AnexampleofaCSVfilewiththecorrectformat
Item | Requirement |
Header | Thefirstrowofthefileisusedtodefinethecolumnheaders.Thefollowing information must be included, in the order shown in the example of Table 1: –SiteId –DeviceId –DeviceType –DataResolution –TimeStamp –Signals(oneormore) Thenaming must bepreciselyasshownhereandiscasesensitive. Any header information entered in the second or subsequent rows will cause errors. |
Mainbody | ThemainbodyofthefilecontainstheoperationalSCADAdata,organised with one Signal per column, and one time step per row. |
Columnseparators | Columnsareseparatedbyasemicolon“;”.Commaseparation‘,’isnot accepted and will cause errors. |
Decimalpoints | Decimalpointsarerepresentedbyacomma“,”.Useofapoint“.”as decimal will cause errors. |
Thousandseparators | Thousandseparatorsarenotneeded,butpoints“.”maybeused.Commas “,” will cause errors. |
FileSize | Themaximumsize ofa singlefilemaynot exceed30MB. Attemptsto upload larger files will cause errors. |
Timeperiod | ThetotaltimeperiodcoveredbyasingleCSVfilemayvaryfromafewhours to several years (although note that the file size limitation may not be exceeded). It is recommended that in a single CSV file, data should be included for a single reporting period, typically one month. Therefore, longer periods should be divided into multiple files, each containing data for a single time period. |
SitesandDevices | A single file may contain data for multiple Sites. For each Site, multiple Devices may be included. Since each CSV file will result in a single results Output,forlargerfleetsitisrecommendedthatasinglefileshouldbeused for each individual Site, and all Devices at that Site should be included in this file. |
Table2:GeneralrequirementsfortheCSVfilestructure
Item | Requirement |
SiteId | SiteId is usedto define the geographicallocationof thewindpower plant. Letters, numbers or special characters may be used, the complete entry will be treated as a character string. |
DeviceId | DeviceId is used to define an individual wind turbine. Letters, numbers or special characters may be used, the complete entry will be treated as a character string. |
DeviceType | DeviceTypeisusedtodefinethemanufacturerandratingofaspecificwind turbine.Letters,numbersorspecialcharactersmaybeused,thecomplete entry will be treated as a character string. Note that the recommended i4SEE syntax should be used, as defined in Section 4. |
DataResolution | DataResolutiondefinesthetime,inseconds,betweenanytwosubsequent entries.Thedataresolutionmustbeconstantthroughoutthefile.Atypical value is 600s for SCADA 10-minute data. |
TimeStamp | TheTimeStampmustbedisplayedinoneofthethreesupportedformats, which are as follows: dd.MM.yyyy H:mm, dd.MM.yyyyH:mm:ss dd.MM.yyyyHH:mm |
Signals | Anynumberofsignalsmaybeincludedinthesubsequentcolumnsofthe file. Signal naming shouldbe accordingto the i4SEE standard. Refer to Section3. |
Signal Values | Signalvaluesaretreatedas“doubles”(double-precision64-bit).Upto15 decimal places are accepted. |
Table3:Specificrequirementsforthecontentofeachcolumn
3.SupportedSignalNames
ThesignalnamescurrentlysupportedbytheSkySpecsPerformancesystemarelistedinTable
4.OnlythesespecificnameswillbeacceptedintheCSVfile.
NotethatthesenamesareconsistentwiththestandardIEC61400-25,butitmaybenecessaryto map the names provided by the SCADA system of the turbine manufacturer to satisfy this standard.
SkySpecsSignalName | Description |
TimeStamp | Timestamp[DD.MM.YYYYHH.MM.SS] |
WNAC_ExTmp_Avg | Averageambienttemperature[°C] |
WNAC_IntnTmp_Avg | Average nacelletemperature[°C] |
WGEN_GnAPh1_Avg | Averagegenwindingcurrent,phase1[A] |
WGEN_GnAPh2_Avg | Averagegenwindingcurrent,phase2[A] |
WGEN_GnAPh3_Avg | Averagegenwindingcurrent,phase3[A] |
WTUR_TotEnergyProd_Totmax | Totalenergyproductioncounter,accumulated value[kwh] |
WTUR_W_Avg | Averagepower[kW] |
WTUR_W_Max | Maxpower [kW] |
WTUR_W_Min | Minpower [kW] |
WTUR_W_Sdv | St.dev power [kW] |
WNAC_Dir_Avg | Average nacelledirection[deg.] |
WTUR_cosPhi_Avg | Averagepowerphase angle[deg.] |
WGEN_GnVPh1_Avg | Averagegenwinding voltage,phase1[V] |
WGEN_GnVPh2_Avg | Averagegenwinding voltage,phase2[V] |
WGEN_GnVPh3_Avg | Averagegenwinding voltage,phase3[V] |
WNAC_WdDir_Avg | Averagewinddirection[deg.] |
WNAC_PriWdSpd_Avg | Averagewindspeed, primary sensor[m/s] |
WNAC_PriWdSpd_Sdv | St devwindspeed,primary sensor[m/s] |
WNAC_SecWdSpd_Avg | Averagewindspeed, secondarysensor [m/s] |
WNAC_SecWdSpd_Sdv | Stdevwindspeed, secondary sensor[m/s] |
WROT_RotSpd_Avg | Averagerotorspeed[rpm] |
WROT_PtAngBl1_Avg | Averagepitch angle,blade1[deg.] |
WROT_PtAngBl1_Sdv | Stdevpitchangle,blade1[deg.] |
WROT_PtAngBl2_Avg | Averagepitch angle,blade2[deg.] |
WROT_PtAngBl2_Sdv | Stdevpitchangle,blade2[deg.] |
WROT_PtAngBl3_Avg | Averagepitch angle,blade3[deg.] |
WROT_PtAngBl3_Sdv | Stdevpitchangle,blade3[deg.] |
WTRM_GbxOilTmp_Avg | Averagegearboxoiltemperature[°C] |
WTRM_GbxBrgTmp_Avg | Averagegboxbearingtemperature[°C] |
WROT_BrgGtmTmp1_Avg | Averagemainbearing temperature,driveend[°C] |
WROT_BrgGtmTmp2_Avg | Averagemainbearing temperature,non-driveend [°C] |
WGEN_GnBrgTmp1_Avg | Averagegen bearing temperature,driveend[°C] |
WGEN_GnBrgTmp2_Avg | Averagegen bearingtemperature, non-driveend[°C] |
WGEN_GnWiTmpPh1_Avg | Averagegenwinding temp,phase 1[°C] |
WGEN_GnWiTmpPh2_Avg | Averagegenwinding temp,phase 2[°C] |
WGEN_GnWiTmpPh3_Avg | Averagegenwindingtemp, phase3[°C] |
WTRF_WiTmpPh1_Avg | Averagetransformerwindingtemp,phase1[°C] |
WTRF_WiTmpPh2_Avg | Averagetransformerwindingtemp,phase2[°C] |
WTRF_WiTmpPh3_Avg | Averagetransformerwindingtemp,phase3[°C] |
WROT_PtMotTmpBl1_Avg | Averagepitchmotor temperature, blade1[°C] |
WROT_PtMotTmpBl2_Avg | Averagepitchmotortemperature,blade2[°C] |
WROT_PtMotTmpBl3_Avg | Averagepitchmotortemperature,blade3[°C] |
WTRM_GbxParticleCount_Tot | Totalparticlecountin10-minuteinterval[-] |
Table4:SignalnamessupportedbySkySpecs
4.SupportedDeviceTypes
As showninTable2, eachentryintheCSV filemust beassignedtoaDeviceType. This informs theSkySpecsPerformancesystemspecificallywhichtypeofturbineproducedthedata.Although theuseofanysyntaxwillbeacceptedbythesystem,forthebestresultsqualityitisrecommended to adhere to the SkySpecs standard syntax for device naming. This is defined as follows:
[ManufacturerName] [TurbineTypeDescription] [TurbineRating, in KW]
The three entries are separated by a space, and each entry may be a combination of numbers, letters or special characters. Some examples of popular device types are shown in Table 5:
Example | DeviceType |
1 | EnerconE-1013075 |
2 | EnerconE-852350 |
3 | GE1033200 |
4 | NordexN1173000 |
5 | SenvionMM822000 |
6 | Siemens SWT1012300 |
Table5:ExamplesofDeviceTypenaming,accordingtotheSkySpecsconvention
5.FileUploadProcedure
TheprocedureforuploadingdataintheformofCSVfilesisdescribedbelow:
- DatatobesubmittedforprocessingbytheSkySpecsPerformancesystemmaybeuploaded via the SkySpecs CSV File Uploader.
- Each user of the SkySpecs Performance system is provided with access to a customer- specific FTP upload server. Any uploaded data iskept completely secure, no customer may seeanyfilesfromanothercustomer.Oncefileshavebeenuploadedsuccessfully,norecord of theoriginalfileisstoredwithinthe SkySpecs system,thereforetheuseris responsiblefor keeping an archive should the data be required again at a later date.
- In order to use the file upload service, the user is required to create one or multiple CSV file(s)withaformatthat isfullyconsistentwiththeguidelinesprovidedabove.Theuserthen uploads this file to the “Incoming” folder on the designated ftp server.
- SkySpecs will then automatically import this file for processing. The system will respond to the presence of newly uploaded files within a maximum of 120 seconds.
- Any uploaded files of the wrong type will automatically be moved to the folder named “NotCsv”.
- Filesthatareuploadedwitherrorsintheformatwillautomaticallybemovedtothefolder named “Error”.
- Onceafilehassuccessfullybeenprocessed,itwillberemoved,andaresultsfilewillbe created in the folder named “Processed”.
6.Glossary
TheterminologycommonlyusedwithintheSkySpecsPerformanceanalyticsandsupporting documentation is defined in Table 6 below:
Term | Definition |
Device | Anindividualwind turbine. |
Event | Asingleobservation,suchasawarningoralarm,occurringataspecificpointin time and associated with a single Device. |
Execution | A single activation of a SkySpecs Performance analysis, processing data for a definedPeriod(consistingofoneorseveralIntervals)andforoneormoreSite(s) and Device(s). |
Interval | Aunitoftimewithinwhichasingleanalysisisperformed.TheIntervalmay(but must not) be shorter than the Period. |
Metric | Aspecificmeansofmeasuringthebehaviourofasignal,typicallyresultingina Result. |
Period | Aunit oftimewithinwhichabatchof analysistasksareperformed. ThePeriodis definedwithastarttimeandanendtimeandmay(butmustnot)includemultiple Intervals to be analysed in a single Execution. |
Result | Asingle,scalarvalueresultingfromaspecificcalculationperformedbyananalysis task. |
Signal | Aseriesofmeasurementdata,usuallyrepresentedbyavectorconsistingof multiple entries recorded with a fixed resolution. |
Site | Asinglegeographicallocationuponwhichawindpowerplantislocated,consisting of one or several Device(s). |
Table6:SkySpecsPerformancecommonlyused terminology
Attachment C
SkySpecs Performance
CSV File Upload Guideline Alarms
Contents………………………………………………………………………………………… 27
1.Introduction……………………………………………………………………………………. 27
2.FileStructure……………………………………………………………………………………. 27
3.FileUploadProcedure……………………………………………………………………………………. 29
4.Glossary……………………………………………………………………………………. 31
5.References……………………………………………………………………………… Error!Bookmarknotdefined.
The SkySpecs Performance analysis tools are designed to process wind turbine operational data and produce actionableresultsandadvisoryinformation.Inorderforthesystemtofunctioncorrectly,datamustbeprovided in a format that must be consistent with the SykSpecs standard. Users may choose from two possible solutions for submitting data to the system. The first, and preferred approach, is via an API interface, connecting the data source (typically a SCADA system) to the SkySpecs Performance system. For further information concerning this approach, see Reference [1]. The second option is to generate CSV (comma separated variable) files and upload these to an FTP server, from where the files are automatically imported and processed by SkySpecs.
This document covers specifically Event logs as generated by the turbine controller. Event logs are used by the SkySpecsPerformancesystemtosupportdetailedroot-causediagnosis,especiallyforcaseswhereananomalyis detected through analysis of time-series data. Refer to [2] for details of the file format required for uploading SCADA time-series data.
NotethatsometurbinemanufacturersusedifferentterminologytorefertoEvents,e.g.“Alarms”,“States”or “Warnings”.AlsonotethatSkySpecsmayrequestthattheCustomeralsoprovidesadocumentexplainingthe definition of Event codes for a specific Device Type (reference code versus description).
ThisdocumentprovidesdetailsoftheCSVFileUploadsolution,includingtherequiredfilecontentandstructure, as well as the procedure for uploading the files to the FTP server.
The CSV upload solution requires files to be prepared in a defined format and structure. It is anticipated thatsuchfilegenerationshouldbeperformedbyanautomatedservice,whichregularlyextractsdatafromthesource (typically a SCADA database) and uploads to an FTP server (refer to Section Error! Reference source not found. below).
An example of a file that has been prepared with the correct format is shown in Error! Reference source not found. and general requirements for the format and structure of files are summarised in Table 2. The specific requirementsforthecontentofeachcolumninthefilearedetailedinTable3.Alloftheseguidelinesmustbe satisfied for error-free upload of CSV files to the SkySpecs Performance system.
Table1:AnexampleofaCSVfilewiththecorrect format
Item | Requirement |
Header | Thefirstrowofthefileisusedtodefinethecolumnheaders.The following information must be included, in the order shown in the example of Error! Reference source not found.. –DeviceId –Code –PeriodStart –PeriodEnd –Message Thenamingmustbepreciselyasshownhereandiscase sensitive. Anyheaderinformationenteredinthesecondorsubsequent rows will cause errors. |
Mainbody | ThemainbodyofthefilecontainsalistofEventlogs,organised with one entry per row. |
Columnseparators | Columnsareseparatedbyasemicolon“;”.Commaseparation‘,’ is not accepted and will cause errors. |
FileSize | Themaximumsizeofasinglefilemaynotexceed30MB. Attempts to upload larger files will cause errors. |
Timeperiod | The totaltime period coveredby asingleCSVfile may vary from a few hours to several years (although note that the file size limitation may not be exceeded). It is recommended that in a single CSV file, data should be included for a single reporting period,typicallyonemonth.Therefore,longerperiodsshouldbe dividedintomultiplefiles,eachcontainingdataforasingletime period. |
SitesandDevices | A single file may contain data for multiple Sites. For each Site, multipleDevicesmaybeincluded.SinceeachCSVfilewillresult in a single results Output, for larger fleets it is recommended that a single file should be used for each individual Site, and all Devices at that Site should be included in this file. |
Table2:GeneralrequirementsfortheCSVfilestructure
Item | Requirement |
DeviceId | DeviceId is used to define an individual wind turbine. Letters, numbersorspecialcharactersmaybeused,thecompleteentry will be treated as a character string. |
Code | TheuniquecodeusedtoidentifytheEvent.Thisistypicallya numerical integer. |
PeriodStart | TheTimeStampatwhichtheEventstartedmustbedisplayedin one of the three supported formats, which are as follows: dd.MM.yyyy H:mm, dd.MM.yyyyH:mm:ss dd.MM.yyyy HH:mm |
PeriodEnd | TheTimeStampatwhichtheEventendedmustbedisplayedin one of the three supported formats, which are as follows: dd.MM.yyyy H:mm, dd.MM.yyyyH:mm:ss dd.MM.yyyy HH:mm |
Message | The message can be any free text description of the Event. Spacesmaybeincluded,aswellasupper-caseandlower-case characters. Special characters should be avoided. |
Table3:Specificrequirementsforthecontentofeachcolumn
TheprocedureforuploadingdataintheformofCSVfilesisdescribedbelow:
- Data to be submitted for processing by the SkySpecs Performance system may be uploaded via the SkySpecs CSV File Uploader.
- Each user of the SkySpecs Performance system is provided with access to a customer-specific FTP upload server. Any uploaded data is kept completely secure, no customer may see any files from another customer. Once files have been uploaded successfully, no record of the original file is stored within the SkySpecs system, therefore the user is responsible for keeping an archive should the data be required again at a later date.
- In order to use the file upload service, the user is required to create one or multiple CSV file(s) with a format that is fully consistent with the guidelines provided above. The user then uploads this file to the “Incoming” folder on the designated ftp server.
- SkySpecs will then automatically import this file for processing. The system will respond to the presence of newly uploaded files within a maximum of 120 seconds.
- Anyuploadedfilesofthewrongtypewillautomaticallybemovedtothefoldernamed“NotCsv”.
- Filesthatareuploadedwitherrorsintheformatwillautomaticallybemovedtothefoldernamed “Error”.
- Once a file has successfully been processed, it will be removed, and a results file will be created in thefolder named “Processed”.
TheterminologycommonlyusedwithintheSkySpecsPerformanceanalyticsandsupportingdocumentationis defined in Table 4 below:
Term | Definition |
Device | Anindividualwindturbine. |
Event | Asingleobservation,suchasawarningoralarm,occurringataspecificpointintimeand associated with a single Device. |
Execution | A single activation of a SkySpecs Performance analysis, processing data for a defined Period(consistingofoneorseveralIntervals)andforoneormoreSite(s)andDevice(s). |
Interval | Aunitoftimewithinwhichasingleanalysisisperformed.TheIntervalmay(butmustnot) be shorter than the Period. |
Metric | Aspecificmeansofmeasuringthebehaviourofasignal,typicallyresultinginaResult. |
Period | Aunitoftimewithinwhichabatchofanalysistasksareperformed.ThePeriodis defined withastarttimeandanendtimeandmay(butmustnot)includemultipleIntervalstobe analysed in a single Execution. |
Result | Asingle,scalarvalueresultingfromaspecificcalculationperformedbyananalysistask. |
Signal | Aseriesofmeasurementdata,usuallyrepresentedbyavectorconsistingofmultiple entries recorded with a fixed resolution. |
Site | Asinglegeographicallocationuponwhichawindpowerplantislocated,consistingofone or several Device(s). |
_files/image004.gif)