View
2
Download
0
Category
Preview:
Citation preview
NREL is a national laboratory of the U.S. Department of Energy, Office of Energy Efficiency and Renewable Energy, operated by the Alliance for Sustainable Energy, LLC. NREL is a national laboratory of the U.S. Department of Energy, Office of Energy Efficiency and Renewable Energy, operated by the Alliance for Sustainable Energy, LLC.
Transient Behavior in Thin-Film Modules Michael Deceglie
Timothy Silverman Bill Marion
Sarah Kurtz
2 NREL M. Deceglie 2015 PVMRW
Metastable performance in thin-films
Light exposure (1000 W m-2, 55°C) Storage at room temp.
Thin-film modules change performance upon exposure to light and storage in the dark
3 NREL M. Deceglie 2015 PVMRW
Metastabilities vs. permanent changes
• Initial Norm. Pmp: 0.88 • After 165 hours at 55°C: 0.91
Light exposure (1000 W m-2, 55°C) Storage at room temp.
Changes are at often at least partially reversible, especially upon exposure to elevated temperatures
4 NREL M. Deceglie 2015 PVMRW
Metastabilities vs. permanent changes
• Initial Norm. Pmp: 0.88 • After 165 hours at 55°C: 0.91
Light exposure (1000 W m-2, 55°C) Storage at room temp.
Changes are at often at least partially reversible, especially upon exposure to elevated temperatures
Metastabilities: reversible changes Stabilization: Repeatedly achieving metastable state
5 NREL M. Deceglie 2015 PVMRW
Differentiating degradation
No stabilization Degradation or metastable change? P
ower
Unstabilized measurement
Damp heat
6 NREL M. Deceglie 2015 PVMRW
Differentiating degradation
Pow
er
With stabilization Degradation ruled out
Unstabilized measurement
Stabilized measurement
Damp heat
Time
Repeatable stabilization procedure needed before and after reliability test
7 NREL M. Deceglie 2015 PVMRW
“The light state”
Pow
er
“Dark state”
Time
“Light state”
8 NREL M. Deceglie 2015 PVMRW
“The light state”
Pow
er
“Dark state”
“Light state”
Time
?
There is not a singular, well-defined light state
9 NREL M. Deceglie 2015 PVMRW
Time scales of change
Transient changes in performance of two different types of CIGS modules
Changes can occur on different timescales for different modules
10 NREL M. Deceglie 2015 PVMRW
Must define relevant timescale Transient change in CdTe module exposed at 1000 W m-2, constant temperature
We use: <1% per 20 kW h m-2 IEC 61646: <2% in two intervals of 43 kW h m-2
0.4% / 20 kW h m-2 Fit to last 20 kW h m-2
11 NREL M. Deceglie 2015 PVMRW
Temperature dependent states
CdTe module exposed to light at different temperatures
Silverman et al., J. PV 5:1 p. 344, 2015
12 NREL M. Deceglie 2015 PVMRW
Quantifying the temperature effect
• Light soak in chamber until stable o <1% / 20 kWh m-2 change based on in situ IV curves
• Measure IV at room temperature at 30 & 60 minutes after end of light soak
• Repeat light-soak at different back-of-module temperature
• Example: o Light soak at 50°C o Measure RT IVs o Light soak at 70°C o Measure RT IVs o Light soak at 50°C o Measure RT IVs
Discriminates permanent vs. reversible changes
13 NREL M. Deceglie 2015 PVMRW
Temperature-dependent “light states”
Variations in light-soak temperature affect the final
metastable state.
For a 20°C discrepancy: 2.8% change in CIGS 5.8% change in CdTe (Effects not universal)
Different shades indicate different samples. Both 30 min and 60 min measurements shown.
14 NREL M. Deceglie 2015 PVMRW
Temperature-dependent parameters
Voc 3.6 % FF 3.1%
Voc 0.3% FF 2.3%
Differences
Hot light soak Cool light soak
15 NREL M. Deceglie 2015 PVMRW
Light-free stabilization
• Can be quicker and cheaper • Must produce relevant state
Pow
er
Unstabilized measurement
Light-free
Damp heat
Time
Light-based
16 NREL M. Deceglie 2015 PVMRW
Bias at elevated temperature (BET)
Method based on proposal in draft IEC 61215 update: • Collect light I-V curve • Heat to 85°C • Apply forward bias (0.6 to 0.7) × Voc • Cool to room temperature • Collect light I-V curve • Repeat until power is stable
I-V measurement I-V measurement
17 NREL M. Deceglie 2015 PVMRW
BET results
BET Light exposure
Two types of CdTe module were stabilized with the bias at elevated temperature (BET), then placed in light-soak chamber
For some modules, BET does not produce a light-stable state
Silverman et al., J. PV 5:1 p. 344, 2015
18 NREL M. Deceglie 2015 PVMRW
BET results
BET Light exposure
Validation is critical
Silverman et al., J. PV 5:1 p. 344, 2015
19 NREL M. Deceglie 2015 PVMRW
Reproducible measurements are possible
• Round robin o 5 labs o 4 CIGS module types
• Uninterrupted 5 hour light soak
• Module forward biased while cooled
• Prompt IV measurement (within 1 minute)
Largest metastability in study
Deceglie et al., J. PV 5:2 p. 607, 2015
20 NREL M. Deceglie 2015 PVMRW
Conclusion
• “Light state” and “Dark state” are ill-defined o Timescale o Temperature o Stimulus (alternate methods)
• Best practices o Define the timescale of interest o Tightly control the temperature o Validate light-free methods o Control and minimize the delay between light
exposure and measurement
21 NREL M. Deceglie 2015 PVMRW
Acknowledgments
Thank you: • Steve Rummel, Allan Anderberg, Kent Terwilliger, Greg Perrin,
and Keith Emery (NREL) • Daniela Dirnberger and Alexandra Schmid (Fraunhofer ISE) • Stephen Barkaszi (FSEC) • Nicholas Riedel and Larry Pratt (CFV) • Samantha Doshi and Govindasamy Tamizhmani (TUV
Rheinland PTL) This work was supported by the U.S. Department of Energy under Contract No. DE-AC36-08GO28308 with the National Renewable Energy Laboratory.
Recommended