单排竖放平单轴跟踪双面组件发电模型-尉元杰-阿特斯太阳能

Yuanjie.Yu, Jn.Jaubert, Baohua.He System technology centerCanadian Solar Inc. Dec 4th , 2018CSIQ NASDAQ ListedA comparison of bifacial PV system energy yield modeling tools applied to 1P single-axis tracker system 2 Field trial results01 Bifacial system simulation analysis 02 Conclusion 053 3 Bifacial module with leading half-cut cell technologyDouble glass2010 Worldwide leading double glass module manufacturerKu (Half-cell)2010 Worldwide No. 1 Ku module manufacturer Capacity5GW, Shipment3GWBiKu2017 Worldwide NO.1 manufacturer to offer bifacial Ku poly module， shipment300MWCanadian Solar’s BiKu： Based on Eight years’ track record 4 FrontRear• State of the art MCCE (Metal Catalyzed Chemical Etching)• PERC structure using ALD Al2 O3 passivation• 5 busbar design / MBB multi-busbar (9)• Controlled LID/LeTiD• Excellent low light response• Lower temperature coefficient• Enables Bifacial cellsMain CharacteristicsCanadian Solar Poly PERC technology introduction 5 Changshu bifacial field trial overviewChangshu project located in Suzhou, Jiangsu Province, East China, total system size is 28 kWdc.Climate type is typical Humid with hot summer.Latitude:39.7°Longitude:106.8° 6 Item Mounting type Module type Module quantity DC capacity (Kw) InverterProduct type Nominal power MPPT number Max input currentFT1 Single axis tracker(1*10) 3U-MS-FG(single facial) 10 3.71 GROWATT 1000TL3-S 10kW 2 13AFT2 10 3.70FT3 3U-MB-FG(bifacial) 10 3.67 10kW 2 13AFT4 10 3.69 FT2FT1 FT3 FT4Single facial bifacial Changshu bifacial field trial basic information Albedo meters Silicon sensor on back side 7 Field trial data cleaning methodologyData processing item DescriptionData source DC related parameters and climate data with 1min intervalBasic outlier filter ü PArray-DC 5W; ü POA 20W/m2; ü Events and missing data ü Box plot extreme and mild outliers for Power/irradiance Comparison criterion Specific energy yield calculation (kWh/kW) with front side powerBifacial gain definition 8 Bifacial field trial testing results- Energy yield Data source: DC side on field trial platform in Changshu, China. Testing period: Feb-Oct, 2018 Monthly albedo During normal operation 231 days, accumulated bifacial gain for the tracker system is 9.5% under average albedo 26%.Specific Energy yield (kWh/kW) 9 Field trial results01 Bifacial system simulation analysis 02 Conclusion 053 10 ü Canadian Solar s own simulation tool, CASSYSü Freeware and open-source: http://canadiansolar.github.io/CASSYS/ ü Inputs and underlying model are very similar to PVsyst ❖ Some alternate models are available like spectral correction.❖ Can be run with any time step and in batch modeü NREL’s Python code named BifacialVF. (Only used for irradiance simulation)ü Freeware and open-source: https://github.com/NREL/bifacialvfü One of the most popular commercial simulation software tools, PVsyst. Energy modeling methodology-software introduction 11 Simulation VS measurement - Energy yieldBifacial energy yield difference between simulation and measurement can match well within reasonable uncertainty range. （ Sim.-Test)/Test (%)CASSYS PVsyst-1.9% 2.6% （ Sim.-Test)/Test (%)CASSYS PVsyst-0.7% 3.6% 12 Different simulation approach on backside irradiance and energy yieldMeasured DNI, GHI. Measured Front POA, Monthly Albedo Ambient temperatureSimulation A (DNI+GHI) Simulation B(POA) BifacialVF Irradiance and Power simulationIrradiance 13Irradiance VSMeasured DNI, GHI, （ Using Changshu Ambient temperature and Albedo） Measured Front POA， Rear POA, Albedo, Ambient temperature， DC power Irradiance Simulation A (DNI+GHI) comparisonSimulation A Measurement Limitation：The irradiance data in two neighboring places were adopted.Remark: Filter the measured POA irradiance difference more than 50W/m2. 14 Simulation A (DNI+GHI) VS measurement- Front POA irradiance Note: Model transposition uncertainty(2-3%) and measured difference between the pyranometer and Si sensor(1-3%) should be considered, （ Sim.-Meas.)/Meas.(%)NREL CASSYS PVsyst1.8% 3.8% 5.2% Three approaches showed reasonable uncertainty expectation on front irradiance simulation. 15 Simulation A (DNI+GHI) VS measurement-Rear POA irradiance（ Sim.-Meas.)/Meas.(%)NREL CASSYS PVsyst10.6% 32% 24%Three approaches showed aggressive results on rear irradiance simulation? Additional factors that influence the results need to be studied.MBE: Sim.-Meas. 16 sensor Module sensorsensorTube1.5 m 1.3 m 1.4 mSimulation limitation– Backside irradiance sensor height Sensor height difference is one of simulation deviation sources. 17 Average of three sensor irradiance is about 35.6% higher than middle irradiance sensor.sensor Module sensorsensorTube1.5 m 1.3 m 1.4 mRemark: ü Module installation height 1.5m used as simulation input.ü Measured rear side irradiance heights are 1.4m and 1.3m for edges and middle positons respectively. （ Sim.-Meas.)/Meas.(%)Approach NREL CASSYS PVsystBefore correction 10.6% 32% 24%After correction -25% -3.6% -11.6%After correction, the simulation results become conservative. It is not very easy to match very well.Simulation A (DNI+GHI) VS measurement- Corrected Rear POA irradiance 18 Simulation A (DNI+GHI) VS measurement-Rear POA irradiance 19 Measured Front POA， Albedo, Ambient temperature， DC power Measured Front POA, Monthly Albedo Ambient temperatureVS MeasurementSimulation BSimulation B (POA) approach comparisonIrradiance and Power Irradiance and Power 20 Simulation B (POA) approach VS measurement- Rear POA irradiance For rear side irradiance: CASSYS and PVsyst are still aggressive?（ Sim.-Meas.)/Meas.(%) Approach CASSYS PVsystBefore corrected 24% 14%After corrected* -11.6% -21.6%Remark： Average of three sensor irradiance is about 35.6% higher than middle irradiance sensor. Height brings additional uncertainty. 21 Simulation B (POA) VS measurement - Energy yield 22Irradiance and Power One minutemeasured minutely albedo VSOne hourmeasured hourly albedo One hourmeasured monthly albedoMeasured Front POA, Albedo Ambient temperatureDifferent simulation time intervals comparison Measured Front POA， Rear POA, Albedo, Ambient temperature， DC powerIrradiance and PowerMeasurementSimulationCASSYS can provide alternative time step from 1min to hourly for key parameter input. 23 Different time intervals - Rear POAMinutely simulation results is a little better than hourly and monthly on back irradiance level.（ Sim.-Meas.)/Meas.(%) Time step Minutely albedo Hourly albedo Monthly albedoMBE/AV 22.1% 22.7% 23.6% 24 Different time intervals – Energy yieldSimulation results with different time steps show the same level in energy yield side. Current monthly albedo should be enough for bifacial energy yield simulation. （ Sim.-Meas.)/Meas.(%)Time step Minutely albedo Hourly albedo Monthly albedoMBE/AV -1.5 % -0.8% -1.1% 25 Field trial results01 Bifacial system simulation analysis 02 Conclusion 053 26 ConclusionØ Three simulation approaches (NREL,PVsyst,CASSYS) seem relatively conservative considering the limitation of sensor type, height, position, etc.Ø Energy yield simulation and measurement in Changshu field match well within reasonable uncertainty range. Rear side simulation under low albedo is not sensitive to energy yield impact.Ø Different time intervals for albedo input are not sensitive under low albedo for bifacial energy yield simulation. 27 Next stepIn order to simulate more accurately next step, we willü Set up more accurate irradiance sensor testing system with more reasonable positionü Simulate through stronger irradiance simulation software( like Rhinoceros +DIVA )ü Albedo testing method and calculation needs to align the same.ü Special modules can be made to monitor and verify irradiance non-uniformity on back side.ü ……Thanks for continuous contribution to bifacial energy modeling from PVsyst, NREL,SANDIA,CASSYS,. Canadian Solar will share learnings about simulation experience based on more field test data! 28 New Field trial testing in WuhaiLocation Mountingtype Module type Module Qty String Qty Invert Remark1-17Red zone Landscape3*6 CS3U-355PB-FG 216 12 SungrowSG80KTL-M The same BOM for bifacial and monofacial moduleCS3U-355P-FG 216 12Total 24 strings Key parameter ValueTilt 35°Height above the ground 0.5m(fixed）pitch 8m(fixed)Ground albedo 0.3Mounting structure Triangle steel structureFoundation Screw steel tubeCable connection Up and down string connectionBack side irradiance monitoring Ground Albedo monitoringSingle facial module bifacial module 29