Module Reliability Assessment Using IR and EL Imaging Techniques

Transcription

1 9/23/14, IEA PVPS Task 13 WS, 29th EU PVSEC Amsterdam Module Reliability Assessment Using IR and EL Imaging Techniques M. Köntges Institute for Solar Energy Research Hamelin Extract of TASK13 report and new work Review on Failures of Photovoltaic Modules Primary authors: M. Köntges, S. Kurtz, C. Packard, U. Jahn, K. A. Berger, K. Kato, Th. Friesen, Haitao Liu, M. Van Iseghem

2 Outline Introduction of Thermography (TG) and Electroluminescence (EL) technique What physical quantity do we see? Interpretation of the images EL/TG How much do the images allow to assess the reliability?

3 Steady state thermography (TG) Wavelength 8 μm - 14 μm T b T r Min. 700 W/m² irradiation at the module array Measure T in: ε(t 4 T b4 )+ ρ(t r 4 T b4 )= 0 [2] ε T ρ Typical emissivity values are for the glass 0.95 for the polymer back sheet Angle of view relative to surface normal 0 to 40 glass 0 to 45 polymer back sheet Out of angle module appear too cold [1] [1] C. Buerhop, H. Scheuerpflug, R. Weißmann: The Role of Infrared Emissivity Of Glass on IR- Imaging of PV-Plants, Proc. 26th EUPVSEC (WIP, Hamburg, Germany, 2011), pp [2] C. Buerhop, D. Schlegel, C. Vodermayer, M. Nieß: Quality control of PV-modules in the field using infrared-thermography, 26th EUPVSEC (WIP, Hamburg, Germany, 2011), pp

4 Camera position Chose camera height h c so that view angle γ < 40 to normal of glass surface A too high position cause self reflection Be aware of other reflections: clouds, houses, trees Often high position necessary Long stick Lifting ramp/ladder Octocopter β h c d α γ=40 Image courtesy of DuPont. h m d= 1.5 m, h m =0.5 m l 1 Red framed dot: self reflection [1] B. Weinreich, Feldstudie zur Modul- und Generatorqualität auf Basis thermografischer Messungen über 100 MW, Proc. 28th Symposium Photovoltaische Solarenergie (OTTI, Bad Staffelstein, Germany, 2013), ISBN

5 Steady state thermography and their interpretation Pattern Description Possible failure reason One module warmer than others One row (substring) is warmer than other rows in the module Single cells are warmer, not any pattern (patchwork pattern) is recognized Module is open circuited - not connected to the system Short circuited (SC) sub-string - Bypass diode SC, or - Internal SC Whole module is short circuited - All bypass diodes SC or - Wrong connection Electrical measurements Module normally fully functional Sub-strings power lost, reduction of V oc Module power drastically reduced, (almost zero) strong reduction of V oc Remarks, Chapter Safety Power Check wiring A System failure Probably burned spot at the module One diode shunted Check wiring all diodes shunted B(f) A when ext. SC, B(f) when Diodes SC TG outdoor const. or E const. or E Single cells are warmer, lower parts and close to frame hotter than upper and middle parts. Massive shunts caused by potential induced degradation (PID) and/or polarization Module power and FF reduced. Low light performance more affected than at STC - Change array grounding conditions - recovery by reverse voltage (PID) A C (v,h,t)

6 recognized IEA INTERNATIONAL ENERGY AGENCY - Wrong connection reduction of V oc Diodes SC Single cells are warmer, lower parts and close to potential induced Steady state thermography and ced. their Low light conditions interpretation One cell module clearly warmer than the others Part of a cell is One row (substring) is warmer warmer than other rows in the module Pointed heating Single cells are warmer, not any pattern (patchwork pattern) is recognized Massive shunts caused by frame hotter than degradation (PID) upper and middle and/or polarization Possible failure Pattern parts. Description reason - Module Shadowing is open circuited effects - not - connected Defect cell to the - system Delaminated cell - Broken cell Short circuited - Disconnected (SC) sub-string string - cell Bypass interconnect diode interconnect SC, ribbon or - Internal SC - Artifact Whole - Partly module is short shadowed, circuited e.g. - bird All bypass dropping, lightning diodes SC or - protection Wrong rod connection Module power and FF redu- performance more Electrical affected than measurements at STC Module Power decrease normally fully not necessarily functional permanent, e.g. shadowing leaf or lichen Drastic power Sub-strings reduction, FF power lost, reduction reduction of V oc Power reduction, dependent Module power on form drastically and size of the reduced, cracked (almost part zero) strong reduction of V oc - Change array grounding - recovery by reverse Remarks, voltage Safety Power Chapter (PID) (cell cracks) Probably burned (burn marks) spot at the module (interconnects) One diode shunted Crack detection after Check detailed wiring visual inspection of the all cell diodes possible shunted (cell cracks) A B(f) B(f) B(f) A when ext. SC, B(f) when Diodes SC C (v,h,t) outdoor Visual Check inspection wiring A A, System needed, cleaning failure (cell mismatch) or B(f) B, shunted cell or (delam.) C(m, tc, h) C(m, const. or tc) E C(m, tc) const. or E Sub-string part remarkably Single cells hotter are than warmer, others lower when equally parts and shaded close to frame hotter than upper and middle parts. Sub-string with missing Massive or shunts open circuit caused bypass diode potential induced degradation (PID) and/or polarization Massive Isc and power Module reduction power when and FF part redu- of this sub-string ced. Low light is shaded performance more affected than at STC May cause severe fire - Change hazard array when hot grounding spot is in this sub-string conditions - recovery by reverse voltage (PID) A, A B(f) TG A, C C (v,h,t)

7 Estimate temperature for active & inactive module parts Use linear model for PV module temperature I irr v wind T mod = c + w 1 T amb + w 2 I irr + w 3 v wind Use typical temperature coefficient for module efficiency = (1 - (T mod - 25ºC) c T ) 25ºC T amb Passive cell parts can not convey part of the irradiation. Equivalent to higher irradiance I eff I eff = I irr / (1- ) Passive cell is hotter. Simulated by higher effective irrad. Sym. Unit Value c C 5.1 w w 2 Cm²/W 0.03 w 3 Cs/m c T 1/K ºC T pas = c + w 1 T amb + w 2 I eff + w 3 v wind Coefficients from: G. Tamizh Mani et al., PHOTOVOLTAIC MODULE THERMAL/WIND PERFORMANCE: Long -Term Monitoring and Model Development For Energy Rating, NCPV and Solar Program Review Meeting 2003, p

8 Module temp. [ C] IEA INTERNATIONAL ENERGY AGENCY Steady state thermography (TG) 2.5 to 10 K temperature increase for: shunted by pass diode open circuit PV-module PID-cells Temperature difference increase with irradiation and module efficiency passive cell Local defects give large range of temperatures No rejection criterion defined yet (85ºC - 150ºC?) 45 cell convey energy away Irradiance[W/m²] Steady state thermography identifies relevant defects, because done under working conditions Measurement conditions are limiting the technique Cheap imaging technique High view position necessary (increase costs)

9 Log-in elektroluminescence (EL) Wavelength 1150 nm 40 V 8 A Electronic relay Si CCD camera Current source to stimulate electroluminescence Optional on/off subtraction and nm long pass filter to eliminate extraneous Isc 10% Isc inactive cell parts, PID Logarithmic EL intensity quotient of two areas is proportional to voltage difference V V T ln 1 2

10 Features from the wafer Crystal defects from multicrystalline wafer (multi) Cell efficiency is higher with smaller defect area Edge Wafer (multi) Slightly reduced cell efficiency Striation rings (mono) Slightly reduced cell efficiency No follow up failure

11 Metal paste Contact formation feature A Temperature distribution during firing of metal paste is inhomogeneous, chain pattern of conveyor belt Contact formation - feature B Temperature gradient during firing of metal past from middle to edge Humidity corrosion Corrosion of front finger contacts, probably due to acidic acid of laminate, rapidly reduce FF New end of live effect found in hot humid environment! 1 1 S. Sakamoto et al., 29 th EUPVSEC 2014, Amsterdam, Netherland, 5DV.3.14

12 EL Further failure Local shunt on solar cell (validate with TG) Edge isolation fault Emitter locally defect Slighly reduce V oc and FF, possibly hot spot Shunt by cell interconnect ribbon Reduction of V oc Broken cell interconnect ribbon Smal reduction of FF und I mpp, may generate arcing, 1/3 power loss when last ribbon breaks

13 Three classes of finger failure Failure by screen printing Repeating finger interruption, typically not at cell interconnect ribbon Failure by stringing Irregular finger interruption at cell interconnect ribbon (chipping) Failure by cell crack Finger interruption at cell crack, cell part isolation in progress Follow up failure possible

14 EL Cracks in solar cells Inactive cell parts Need to measure 10% I sc decrease of FF and I mpp Cross crack line along multiple neighbor cells E.g. scratching a module corner on back sheet of a PV module May result in isolation fault, visual inspection needed Cell cracks in solar cells How much is the power loss? Köntges et al., ep Photovoltaik aktuell 7/8 2008, S. 36

15 Lock-in electroluminescence (EL) C I sc A B With cooled 12 bit CCD camera detectable voltage difference: < 150 mv 3500 V VT ln 150mV 10 Voltage drop at break resistance is less at lower current C 1/10 I sc A Therefore low current image is more sensitive to isolated parts B

16 60 cell module simulation Aim: Power loss simulation Module equivalent network and voltage drop across crack LT Spice for electrical simulation Two diode model for each cell Power loss due to defective cell is influenced by breakdown voltage Breakdown model of Alonso-Garcia 1 Break resistance R b connect broken cell area A inactive A active New symbol for broken cell R b A inactive 1 M.C. Alonso-Garcia et al., Solar Energy Materials & Solar Cells 90 (2006)

17 Spice simulation 1 of voltage drop over break resistance R b 60 cells 228 W PV module A active R b A inactive I = 8 A Below noise level At I=8 A one can not detect if the break resistance is high enough to generate power loss [1] M. Köntges, I. Kunze, S. Kajari-Schröder, X. Breitenmoser, B. Bjørneklett, Solar Energy Materials & Solar Cells 95 (2011), pp

18 Spice simulation 1 of voltage drop over break resistance R b 60 cells 228 W PV module A active R b A inactive C Typ B A At I=0.8 A one can detect if the break resistance is high enough to generate power loss I = 0.8 A [1] M. Köntges, I. Kunze, S. Kajari-Schröder, X. Breitenmoser, B. Bjørneklett, Solar Energy Materials & Solar Cells 95 (2011), pp

19 Electroluminescence (EL) High resolution Cost efficient Si CCD camera for dark environment Very expensive camera for daylight imaging Fast image recoding Often difficult to differentiate between feature and failure Automation of failure detection only for production possible Failure cells are detectable by deviation from mean intensity, but a human has to classify further For much more detectable failure and more explanations read TASK13 report

20 Summary: Steady state TG is a cheap technique for failure detection in the field, but is strongly restricted by weather conditions and view angle EL can detect most defects, often difficult to differentiate between failures and effects, inexpensive technique EL for outdoor same advantages like indoor, but quite expensive technique New standards in preparation for EL: IEC (number is applied) More details in TASK13 report Review on Failures of Photovoltaic Modules - available here and on Thanks for financial support: State of Lower Saxony and BMWi under contract number FKZ C.