ASTM Photovoltaic Performance Standards Their Use at the National Renewable Energy Lab
The performance of photovoltaic devices is typically rated in terms of their peak power with respect to a specific spectrum, total irradiance and temperature. The PV Cell and Module Performance Laboratory at the National Renewable Energy Laboratory in Golden, Colo., has been measuring the performance of cells and modules for the U.S. terrestrial PV community since 1980. NREL typically calibrates 200 cells and modules per month. The laboratory follows the procedures described in ASTM International standards for calibrating its primary reference cells (E 1125), spectral responsivity measurements (E 1021), secondary reference cells (E 948), secondary modules (E 1036), concentrator modules (E 2527), and multi-junction cells and modules (E 2236).
A calibration certificate for ISO 17025, General Requirements for the Competence of Testing and Calibration Laboratories (American Association for Laboratory Accreditation number 2236.01) is available for a restricted set of samples. The laboratory also provides independent PV efficiency measurements to help minimize inflated efficiency claims from appearing in the literature or press. The technical basis for many of the PV performance standards were developed over the years by this group.
Primary PV reference cells are used to determine the electrical performance of PV devices. Primary photovoltaic reference cells conforming to the package specifications of E 1040 are calibrated at NREL in natural sunlight using the relative spectral response of the cell (E 1021), the relative spectral distribution of the sunlight, and a tabulated reference spectral irradiance distribution such as provided in G 173, following E 1125. National PV calibration facilities can obtain an ISO 17025 accredited calibration for this method because of its international importance.
The measurement apparatus (Figure 1) uses an absolute cavity radiometer that is traceable to the World Radiometric Reference to measure the total irradiance in a 5° field of view, a spectral radiometer (E 1341), and a zero bias current-to-voltage converter to measure the short-circuit current (Isc) of a PV cell. The field of view of the four cells that are calibrated at the same time and the spectral radiometer are matched to 5° of the cavity radiometer. The linearity of a primary reference cell’s Isc must be verified using E 1143.
Figure 1 NREL primary reference cell calibration apparatus showing, from left to right, four PV cells mounted on a temperature-controlled plate, Licor model LI-1800 spectral radiometer, Eppley model AHF absolute cavity radiometer.
The apparatus shown in Figure 2 was used to develop an alternative method that is in the balloting process. This method relies on the relationship that the Isc produced by light lamp A alone plus the Isc from lamp B alone must be equal to the Isc from lamps A and B together to be linear.1 This difference is expressed as a percentage deviation from linearity and is measured at a variety of light levels over the irradiance range of interest, which is typically 400-1100 Wm-2.
Figure 2 PV reference cell linearity measurement test bed showing two sets of tungsten light sources , shutters and filter wheels holding seven neutral density filters.
The performance of PV cells and modules with respect to standard reporting conditions are defined by a reference temperature (25 °C), total irradiance (1,000 Wm-2), and spectral irradiance distribution (E 490, G 159, and G 173). The terrestrial PV community is currently using G 159, but will switch to G 173 once the balloting is complete for the revised International Electrical Commission standard 60904-3, Photovoltaic Devices: Part 3: Measurement Principles for Terrestrial Photovoltaic (PV) Solar Devices with Reference Spectral Irradiance Data. This will allow an orderly international switch to a new reference spectrum.
This switch will have an impact of less than 1 percent in the peak-watt rating for silicon modules. However, the PV industry is worth over $10 billion per year, producing over 2.5 gigawatts of cells last year and growing over 40 percent per year, so even a one percent change is significant.2 The major impact of switching from ASTM G 159 to G 173 is for the direct reference spectrum, which is used to evaluate concentrator cells.3 Currently there are no consensus standards for evaluating concentrator cells, but national PV calibration labs in Germany, Japan and the United States have informally agreed on reference conditions for concentrator cells and modules.4 For samples that are within a limited scope of packages and sizes, we can issue an ISO 17025 compliant calibration certificate under A2LA accreditation number 2236.01.
For devices that come to NREL for calibration, we first log in the device based on information from a cover letter or request form. We then measure the total area or aperture area, which is crucial for determining its efficiency, and obtain its spectral responsivity. For cells to be tested to ASTM standard E 948, we measure the spectral responsivity with our grating or filter-based system following E 1021 (Figure 3). For modules tested to E 1036, however, spectral responsivity is generally provided to us by the manufacturer or we measure it on a representative cell. The module performance rating standard E 1036 was put into law when it was referenced in that energy policy law in 2005. We can measure the spectral responsivity of a module, but it is time consuming and prone to artifacts.
Figure 3 Spectral responsivity measurement system using 64 filters with 10 nm bandpass to cover a range from 290 to 2000 nm.
Next, we use the information on spectral responsivity (E 1021) to calculate the spectral mismatch (E 973) between the test device and a primary reference cell for the simulator that will be used for the subsequent current vs. voltage (I-V) measurement. This indicates how the simulator should be adjusted so that the test device yields the same current under the simulator’s spectrum as would be expected from the reference spectrum. We require that the simulator for cell or module measurements be class A according to E 927, which sets limits on the simulator’s match to the reference spectrum, spatial nonuniformity of irradiance and temporal instability of irradiance.
We then measure the I-V characteristics of the device under simulated conditions (Figure 4). For modules, we also measure I-V performance under natural sunlight, which enables us to determine module response under “real” conditions. After measurement, the results under natural sunlight are translated to standard conditions. For concentrator cells, we measure the I-V characteristics as a function of light level. We can also adjust the spectrum to a limited extent for multijunction concentrators to make the current match as close as possible to reference conditions.
Figure 4a Cell I-V test bed for measurements following testing to standard E 948.
Figure 4b Module IV test bed following standard E 1036
After all measurements have been made, we carefully review the results for anomalies or procedural errors. Finally, we prepare a report for the client which can be as simple as a presentation of data tables or as involved as a document that contains description, analysis, data and recommendations. //
1 K. Emery, S. Winter, S. Pinegar, and D. Nalley, “ Proc 4th World Linearity Testing of Photovoltaic Cells,” Proc. IEEE 4th World Conference on Photovoltaic Energy Conversion, May 7-12, 2006, Waikoloa, HI.
2 PV News, Vol. 26, March 2007, published by Prometheus Institute.
3 K. Emery, D. Myers, and S. Kurtz, “What is the Appropriate Reference Spectrum for Characterizing Concentrator Cells?,” Proc. 29th IEEE Photovoltaic Specialists Conf., New Orleans, LA, May 20-24, 2002, pp. 840-843, IEEE, New York, 2002.
4 M.A. Green, K. Emery, D.L. King, Y. Hishikawa, and W. Warta “Solar Cell Efficiency Tables (version 29),” Progress in Photovoltaics Research and Applications, vol. 15, pp. 35-40, 2007.