41 corbellini analysis_and_modelling_of_bifacial_pv_modules

41 corbellini analysis_and_modelling_of_bifacial_pv_modules

• 1.
  14th PV PerformanceModeling Collaborative Workshop Modelling of bifacial PV modules Gianluca Corbellini Vasco Medici
• 2.
  24th PV PerformanceModeling Collaborative Workshop • Located in Lugano (Ticino) • 4500 students • 30y+ experience in PV • Outdoor test stand • Swiss PV Module Test Center • Swiss BIPV Competence Centre
• 3.
  34th PV PerformanceModeling Collaborative Workshop Agenda • Outdoor measurement • Modelling • Indoor measurement • Fit of parameters • Results of modelling • Conclusions and next steps
• 4.
  44th PV PerformanceModeling Collaborative Workshop Outdoor measurement Testing of outdoor performance for 12 PV modules between June 2014 and June 2015 Orientation: 45° tilt at 3°E azimuth Location: Lugano (south Switzerland) Sensors: GHI, DHI, front GPOA , back GPOA , TBOM monofacial All modules operating at MPP, I-V curve measurement every 5 minutes Technology ɳ @STC γ (%/°C) ɳ @200W/m2 HJT BIFAC 1 18.6% -0.33% 3.1% HJT BIFAC 2 17.7% -0.26% 3.1% HJT BIFAC 3 17.6% -0.28% 2.3% C-Si 1 16% -0.41% 5.1% C-Si 2 15.9% -0.41% 5.1% HJT MONO 1 19.4% -0.26% 4.9% HJT MONO 2 19.4% -0.26% 4.9% CIS 1 13.8% -0.34% 5.4% CIS 2 13.7% -0.34% 5.4% C-Si 3 13.4% -0.43% 2.4% CdTe 1 10.9% -0.25% 0.5% CdTe 2 10.6% -0.25% 0.5%
• 5.
  54th PV PerformanceModeling Collaborative Workshop Outdoor bifacial vs monofacial All day types Overcast PR ΔkWh/kWp PR ΔkWh/kWp HJT BIFAC 3 101% 13.7% 106% 15.4% HJT BIFAC 2 100% 12.5% 105% 13.9% HJT BIFAC 1 99% 11.8% 104% 13.7% CIS 1 97% 9.0% 95% 3.8% CIS 2 96% 7.7% 97% 5.4% CdTe 2 95% 7.0% 98% 6.5% CdTe 1 94% 5.6% 94% 1.9% c-Si 3 92% 3.4% 98% 6.4% HJT MONO1 92% 2.9% 92% 0.2% HJT MONO2 91% 2.2% 91% 0.1% c-Si 1 90% 0.7% 93% 0.9% c-Si 2 89% Ref. 92% Ref. Bifacial modules show and additional 10 to 13% energy yield with respect to monofacial HJT Strong dependence on sky conditions (diffuse/direct ratio) and position of the sun HJT Bifacial 2 vs HJT Mono 1 in clear sky days
• 6.
  64th PV PerformanceModeling Collaborative Workshop Modelling of Irradiation SKY DIFFUSE RADIATION DIRECT RADIATION GROUND REFLECTED GROUND REFLECTED COMPARISON OF 5 MODELS
• 7.
  74th PV PerformanceModeling Collaborative Workshop Modelling of Performance LOW IRRADIANCE TERMCORRECTED FLASH TEST TEMPERATURE TERM Tbom estimated from VOC and ISC
• 8.
  84th PV PerformanceModeling Collaborative Workshop Indoor Measurement Some parameters have been estimated with indoor testing: - Low irradiance coefficients - Temperature coefficients - Power @STC front - Power @STC back Are these measures correct for bifacials? How do we estimate the bifaciality factor? PB PF
• 9.
  94th PV PerformanceModeling Collaborative Workshop Indoor Measurement Comparison of front reference cell and a cell on the backside at different position 1.09 0.44 0.48 1.12 1.1 0.58 0.48 0.82 1.11 0.56 0.52 0.51 0.93 5% φ ≈ 0.0044 PF = PF + φPB PB = PB + φPF PF = 291.74 W PF = 𝟐𝟗𝟎. 𝟓𝟕 W PB = 261.61 W PB = 𝟐𝟔𝟎. 𝟑𝟎 W Measure of reflections on the backside in the dark room [%]: Bifaciality factor = 0.896 Corrected for JB = 0.948
• 10.
  104th PV PerformanceModeling Collaborative Workshop Modelling of TCELL Cell temperature estimated from a one-diode model tuned using indoor data Validation on 1 year outdoor data of a standard module (framed poly-Si 260W) IPH IS RS RSH ISH V + - ISCVOC TCELL α β Rsh (G) TMODEL = 0.982 TMEAS + 0.748 RMSE = 0.4535
• 11.
  114th PV PerformanceModeling Collaborative Workshop Modelling of TCELL Validation on a standard module (framed poly-Si 260W) TCELL modelled RMSE = 0.4535 IPH IS RS RSH ISH V + - ISCVOC TCELL α β Rsh (G)
• 12.
  134th PV PerformanceModeling Collaborative Workshop Modelling of Performance 4 Variables • GHI • TBOM (VOC ,ISC) • DHI • (Geff) 6 Parameters • PSTC • δ • ai • Albedo • K0 • K1 Accounting as portion of effective ground reflected irradiance Accounting as portion from diffuse irradianceMeasured but can be modelled Estimated at 0.23 (gravel)
• 13.
  144th PV PerformanceModeling Collaborative Workshop Results of modelling RMSE = 8.34W bias = -7.519W std = 3.607W k1 = 0.193 [0.191 0.194] RMSE = 4.163W bias = -1.745W std = 3.780W k0 = 0.437 [0.434 0.441] RMSE = 4.197W bias = -1.392W std = 3.959W k0 = 0.115 [0.113 0.116] k1 = 0.131 [0.129 0.132] RMSE = 2.933W, bias = -0.445W, std = 2.899W k0 = 0.207 [0.205 0.209] RMSE = 4.836W bias = -1.812W std = 4.484W
• 14.
  154th PV PerformanceModeling Collaborative Workshop Conclusions Two parameters model covers very well the backside contribution to module’s power, both terms are significant Reflections in dark room need to be taken in account TCELL(VOC, ISC) is very accurate on monofacials and promising for bifacials
• 15.
  164th PV PerformanceModeling Collaborative Workshop Next Step Improve TCELL(GBF, TAMB) modelling, only from environmental variables Testing of different tilt/azimuth – optimization for climates New standard procedure for nameplate power definition and indoor testing Modelling of LCOE as the key factor for bifacial success
• 16.
  174th PV PerformanceModeling Collaborative Workshop Thank you for your kind attention Thank you for your kind attention Gianluca Corbellini - SUPSI [email protected]