Characterization using laser-based technique for failure Si PV module

Transcription

1 SAYURI-PV, Tsukuba, 4th Oct, 2016 Characterization using laser-based technique for failure Si PV module Y. Ishikawa, 1 M. A. Islam, 1 K. Noguchi, 1 H. Iida, 2 Y. Takagi, 2 and H. Nakahama 2 1: NAIST, 2: NISSHINBO Information Device Science Laboratory, Graduate School of Materials Science, Nara Institute of Science and Technology

2 Failure of Si PV modules 2/26 PV power plant Failure by Interconnector Cell Encapsulant No method to check (only in indoor/ one module) Detectable in EL Optical image EL image Nondestructive/portable analysis system EL: Electroluminescence Information Device Science Laboratory

3 Nondestructive analysis for PV modules 3/26 Method Imaging Electrode Electroluminescence (EL) Current Injection 〇 Thermography Current injection 〇 Laser Beam Induced Current (LBIC) Light (laser) Photoluminescence (PL) Light (laser) 〇 EL: Crystal defects Defects in connector Thermo.: hot-spot defects in connector PL: Crystal defects Lifetime mapping LBIC: minority lifetime mapping Check the performance of PV modules No information of Encapsulant Information Device Science Laboratory

4 Information Device Science Laboratory Laser-based nondestructive analysis 4/26 Our method for analysis of failure Si PV modules, Raman spectroscopy for Encapsulant Transient absorption spectroscopy (TAS) and microwave-photoconductivity decay method ( -PCD) for the analysis of minority carrier lifetime of Si solar cells in modules Outline: Raman spectroscopic analysis for Encapsulant deterioration > Highlight of this method TAS and -PCD method for failure Si PV modules (PID) > Effect of PID on the carrier lifetime

5 Encapsulant Properties 5/26 Encapsulant properties that needed for high efficiency and long life PV module High efficiency Long term durability Transparency UV durability Thermal stability Low stiffness Highly adhesive Low WVTR High cross link Acid free High volume resistivity Information Device Science Laboratory

6 Information Device Science Laboratory Encapsulant Properties 6/26 Encapsulant properties that needed for high efficiency and long life PV module High efficiency Long term durability Reduce Jsc Transparency Reduce transparency & eff. UV Affected durabilityby UV Discoloration Thermal stability Low stiffness Delamination Reduce Highly adhesive properties HLow 2 O ingress WVTR Moisture Reduce High cross crosslink Corrosion Acetylation Acid free High Reduce volume volume resistivity resistivity PID

7 UV & T Hydrolysis Failure Modes (Encapsulant Degradation) 7/26 T: thermal energy Glass Si-solar cell Na & K -V Na Migration PID T NaOH OH - Ag 2 O + H 2 O T Reduce EVA adhesion o Formation of acetic acid through deacetylation is responsible for the failure mode related to EVA Back sheet Information Device Science Laboratory T & RH% H 2 O ingress

8 Acetic acid formation Failure Modes (EVA Degradation) 8/26 Failure modes related to EVA encapsulant Photos of typical examples of each failure mode. Discoloration UV exposure at higher operating temperatures Corrosion Moisture ingress through laminated edges Reduce transparency (T) power degradation. Increase in Rs and/or decrease in T decrease in power. Delamination Gradual solder bond failures and/or busbar discoloration FF losses due to solderbond degradation and inadequate use of bypass diodes EVA degradation is also partially responsible for the potential induced degradation Information Device Science Laboratory

9 Acetic acid ( g/g) Increase of carbonyl Encapsulant Degradation in lab test 9/26 Pressure Cooker Test (PCT) Carbonyl: cm -1 Ester: cm -1 Parameters: 130 o C, 2 atmp., 168 hrs. P-type Si SC (a) point of interest for examining the IR and Raman spectroscopy Fluorescence from Polyenestructure ー CH 2 -CH 2 -CH 2 =CHn m (b) IR-spectra Fluorescent Intensity: I 1819 /I 2854 (c) Raman spectra (laser: 532 nm) Methylene wag. Information Device Science Laboratory Fluorescent Intensity (%)

10 Acetic acid (ug/g) Information Device Science Laboratory Encapsulant Degradation (field aged modules) 10/26 A. Manufactured 1993, in use for 19 years 280ug/g B. Manufactured in 1989, in use for 23 years 40ug/g C. Manufactured 2001, in use for 11 years 510ug/g 190ug/g 640ug/g 44ug/g 23ug/g 130ug/g 84ug/g 55ug/g 33ug/g Raman Fl. Intensity Ratio (%) Raman intensity ratio vs. amount of acetic acid

11 Raman Fl. Intensity Ratio (%) Encapsulant Degradation (Raman/ power loss) 11/26 19modules Year in service Year in service vs. Raman Fl. Intensity for Six different field aged module Raman Fl. intensity ratio/power loss We found that Raman Fl. Intensity ratio has good correlation with the amount of acetic acid, year in service, and the power loss. Islam et al. Investigation of the EVA degradation Information and prediction Device of Science reliability Laboratory by the Raman spectroscopy, 32 EUPVSEC, 2016

12 Information Device Science Laboratory Summary for Raman spectroscopy 12/26 The complete encapsulant degradation mechanism has been proposed. The degradation start by decompose of acetic acid from EVA followed by water ingress. The degradation start from the module edge. The power degradation of a module has a linear relation with the acetic acid decomposition. As like IR-spectroscopy, Raman spectroscopy could be used to determine the acetic acid decomposition. Advantages is that- Raman spectroscopy is non-destructive, can make a portable detection system, can detect the acid condition without removing the module from PV field.

13 Information Device Science Laboratory Laser-based nondestructive analysis 13/26 Our method for analysis of failure Si PV modules, Raman spectroscopy for Encapsulant Transient absorption spectroscopy (TAS) and microwave-photoconductivity decay method ( -PCD) for the analysis of minority carrier lifetime of Si solar cells in modules Outline: Raman spectroscopic analysis for Encapsulant deterioration > Highlight of this method portable system TAS and -PCD method for failure Si PV modules (PID) > Effect of PID on the carrier dynamics (lifetime etc.)

14 Information Device Science Laboratory PID mechanism in p-type Si SC 14/26 Hypothesis of PID mechanism in p-type Si solar cells Na + Na + Na + Na Glass EVA Elec. SiN Emitter (N + ) Base (P) 1.Na migration to elec. 2.Accum. of Na ion at emitter layer 3.Form inversion layer at the surface of emitter layer 4.Reduce shunt resistance (leakage current increase) 5.Disappear of depletion layer/rectifying property no EL emission, hot-spot

15 Information Device Science Laboratory PID mechanism in p-type Si SC 15/26 Hypothesis of PID mechanism in p-type Si solar cells Na + Na + Na + Na Glass EVA Elec. SiN Emitter (N + ) Base (P) 1.Na migration to elec. 2.Accum. of Na ion at emitter layer 3.Form inversion layer at the surface of emitter layer 4.Reduce shunt resistance (leakage current increase) 5.Disappear of depletion layer/rectifying property no EL emission, hot-spot

16 Information Device Science Laboratory PID mechanism in p-type Si SC 16/26 Hypothesis of PID mechanism in p-type Si solar cells Na + Na + Na + Na Glass EVA Elec. SiN Emitter (N + ) Base (P) 1.Na migration to elec. 2.Accum. of Na ion at emitter layer 3.Form inversion layer at the surface of emitter layer 4.Reduce shunt resistance (leakage current increase) 5.Disappear of depletion layer/rectifying property no EL emission, hot-spot Naumann et al., SOLMAT 120 (2014) 383.

17 Information Device Science Laboratory PID mechanism in p-type Si SC 17/26 Hypothesis of PID mechanism in p-type Si solar cells Na + Na + Na + Na Glass EVA Elec. SiN Emitter (N + ) Base (P) 1.Na migration to elec. 2.Accum. of Na ion at emitter layer 3.Form inversion layer at the surface of emitter layer 4.Reduce shunt resistance (leakage current increase) 5.Disappear of depletion layer/rectifying property no EL emission, hot-spot

18 PID mechanism in p-type Si SC 18/26 Hypothesis of PID mechanism in p-type Si solar cells Na + Na + Na + Na Glass EVA Elec. SiN Emitter (N + ) Base (P) 1.Na migration to elec. 2.Accum. of Na ion at emitter layer 3.Form inversion layer at the surface of emitter layer 4.Reduce shunt resistance (leakage current increase) Increase surface recomb. decrease carrier lifetime? 5.Disappear of depletion layer/rectifying property no EL emission, hot-spot EL image Thermo image Information Device Science Laboratory

19 Effect of PID on carrier dynamics 19/26 TAS and -PCD method for failure Si PV modules (PID) > Effect of PID on the carrier dynamics (lifetime etc.) EL imaging in electron behavior needs electrodes and specific power supplier. EL detection: possible after complete deterioration Need removing of modules from the system to make detail measurement TAS detects free carrier s dynamics at arbitrary point PL also detects carrier s behavior through the dynamics with photo-luminescence. Information Device Science Laboratory

20 Information Device Science Laboratory Transient Absorption Spectroscopy 20/26 Wavelength conversion Fig. Experimental set up TAS parameters Pump: 532 nm Probe: 400nm-800 nm Power: 4.1 mj

21 Transient Absorption Spectroscopy 21/26 *Shukla et al. ACS Nano, (2016),*Dugar et al., RSC Adv. (2015) Fitting equation: A = A d1 exp(-t/t 1 ) +A d2 exp(-t/t 2 ) + A d3 exp(- t/t 3 ) E F Schematics of the proposed carrier transition mechanism in TAS. τ 1 is fast decay directly associated with carrier trapping to the mid gap defect state (around Fermi level) in the vicinity of the conduction band, τ 2 is slow decay that is attributed to carrier trapping to the deep-level state from the conduction band and τ 3 is slowest decay attributed to carrier relaxation to the valence band edge. Information Device Science Laboratory Time ( s)

22 Reflectivity Information Device Science Laboratory Carrier lifetime measurement by -PCD 22/26 Microwave reflectivity: high Pulsed laser Microwave reflectivity: low wafer electron hole Just after pulsed laser irradiation After relaxation Microwave reflectivity is proportional to carrier conc. 1 1/e 0 Time

23 Normalized carrier lifetime (µ-pcd) Carrier lifetime (µ-pcd) for PID tested module 23/26 PID condition: 1000 hrs., 85 o C /85% RH, V Mono-(p) Si Module Linear relation between the normalized EL intensity and normalized carrier lifetime measured by µ-pcd for PID tested module Information Device Science Laboratory Normalized EL intensity

24 τ 1 (μs) τ 2 (μs) Information Device Science Laboratory τ 3 (μs) Carrier lifetime (TAS and µ-pcd) 24/ Fresh PID 300h τ μpcd (μs) Fresh PID 300h τ μpcd (μs) Fresh PID 300h τ μpcd (μs) µ-pcd can demonstrate the deterioration of carrier lifetime happened by PID like TAS meas. Possible in portable way in future

25 Information Device Science Laboratory Summary for TAS / -PCD 25/26 A comparative study of carrier dynamics through TAS has been conducted. Low carrier lifetime has been observed at the PID affected area. Very high decay rate was found in the 1000 hrs. PID tested module. EL intensity may have a linear relationship with the carrier lifetime. -PCD can demonstrate the carrier lifetime like TAS.

26 Information Device Science Laboratory Acknowledgement 26/26 This work is supported by NEDO, Japan. The measurement of TAS was conducted at the Nano-Processing Facility, supported by IBEC Innovation Platform, AIST. Thank you for your kind attention.