Identification and analysis of partial shading breakdown sites in CuInxGa(1-x)Se2 modules

Steven Johnston, Harvey Guthrey, Lorelle Mansfield, Timothy Silverman, Mowafak Al-Jassim, Elizabeth Palmiotti, Andreas Gerber, Angus Rockett

Research output: Contribution to journalArticlepeer-review

27 Scopus Citations

Abstract

In this study, CuInxGa(1-x)Se2 (CIGS) mini-modules are stressed under reverse bias, resembling partial shading conditions, to predict and characterize where failures occur. Partial shading can cause permanent damage in the form of ‘wormlike’ defects on thin-film modules due to thermal runaway. This results in module-scale power losses. We have used dark lock-in thermography (DLIT) to spatially observe localized heating when reverse-bias breakdown occurs on various CIGS mini-modules. For better understanding of how and where these defects originated and propagated, we have developed techniques where the current is limited during reverse-bias stressing. This allows for DLIT-based detection and detailed studying of the region where breakdown is initiated before thermal runaway leads to permanent damage. Statistics of breakdown sites using current-limited conditions has allowed for reasonable identification of the as-grown defects where permanent breakdown will likely originate. Scanning electron microscope results and wormlike defect analysis show that breakdown originates in defects such as small pits, craters, or cracks in the CIGS layer, and the wormlike defects propagate near the top CIGS interface.

Original languageAmerican English
Pages (from-to)1-5
Number of pages5
JournalSolar Energy
Volume161
DOIs
StatePublished - 2018

Bibliographical note

Publisher Copyright:
© 2017

NREL Publication Number

  • NREL/JA-5K00-70651

Keywords

  • Accelerated aging
  • Breakdown voltage
  • CIGS
  • Imaging
  • Reliability
  • Thermal analysis

Fingerprint

Dive into the research topics of 'Identification and analysis of partial shading breakdown sites in CuInxGa(1-x)Se2 modules'. Together they form a unique fingerprint.

Cite this