An inter-laboratory comparison between nine metrology institutes found that there was generally good agreement in the calibration of solar cells on the world photovoltaic scale, but measurable differences in short-circuit current, voltage and power were still revealed due to different methods and conditions.
An international research team compared solar cell measurement methods from nine metrology institutes around the world and found a ‘relatively good’ degree of agreement between them. However, some discrepancies were still observed in important parameters such as short-circuit current, voltage and maximum power, highlighting the differences in measurement procedures and experimental conditions.
The interlaboratory comparison involved nine participating institutions: the German Physikalisch-Technische Bundesanstalt (PTB), designated as the reference laboratory; Finland’s Aalto University; The German Fraunhofer Institute for Solar Energy Systems (Fraunhofer ISE), the Swiss University of Applied Sciences (SUPSI), the Taiwan Industrial Technology Research Institute (ITRI), the German Institute for Solar Energy Research Hamelin (ISFH), the European Commission’s Joint Research Center (JRC), the Institute of Metrology of Bosnia and Herzegovina (IMBiH) and the Austrian Institute of Technology (AIT). No institutions from China took part in this comparison.
The scientists conducted their tests on solar cell designs that conformed to the World Photovoltaic Scale (WPVS), an internationally established calibration system designed in 1995 to ensure traceable and consistent measurements of photovoltaic devices using standardized reference solar cells.
The interlaboratory comparison used four filtered silicon solar cells with WPVS design, consisting of two low-pass (LP1, LP2) and two high-pass (HP1, HP2) filtered devices. Low-pass filtered cells transmit shorter wavelengths while blocking higher wavelengths, while high-pass filtered cells do the opposite by transmitting longer wavelengths and blocking shorter wavelengths, allowing tailored spectral responses for different photovoltaic measurements.
Each cell type included a PT-100 temperature sensor, a 2 cm x 2 cm active area, and Schott glass filters to precisely control which parts of the light spectrum reach the silicon cell. All measurements were performed under standard lighting conditions, with strict temperature and spectral control requirements. The reported values for each cell include extensive uncertainties, with contributions mainly from calibration, spectral mismatch correction and temperature control.
Across all laboratories, uncertainty contributions consistently include reference equipment calibration, spectral mismatch correction, temperature control, optical alignment, and electronic measurement noise. The final extended uncertainties for short-circuit current were found to vary, but generally remained around 0.9-1.6%, reflecting differences in instrumentation and methodology. Spectral mismatch correction emerged as a dominant source of uncertainty in almost all setups due to differences between simulator spectra and illumination reference conditions.
“The results agreed quite well,” the academics explained. “For short-circuit current measured under standard lighting conditions, 37 of the 44 reported values showed agreement within their extended uncertainties. The agreement of the short-circuit current values of participating laboratories was between -2.2% and 3.5%, with normalized errors (EN) ranging between -0.72 to 3.3 for low-pass filtered cells, and -2.1 to 2.8 for high-pass filtered cells. The measurements carried out under natural sunlight have shown the importance of measuring the spectral radiation when using the sun as a source.”
The analysis also showed that high-pass filtered reference cell HP2 showed the largest discrepancies, mainly due to temperature-related shifts in the Schott RG780 filter and possible instability of the spectroradiometer. The results for the maximum power (Pmax) also showed larger deviations than the open-circuit voltage, especially for high-pass filtered cells, with some values exceeding the uncertainty limits. Additional variation arose from differences in temperature control, spectral mismatch correction, and data processing between laboratories.
“Based on the findings, efforts are underway to further reduce the discrepancies,” the scientists concluded. “When measuring under natural sunlight, the uncertainties can be greater than expected if the irradiance spectra are not measured.”
Their findings are available in the study “Results of an interlaboratory comparison of modified reference solar cells between nine metrological institutes”, which was recently published in Results in technology.
In February, an interlaboratory comparison between Fraunhofer ISE and PTB’s CalLab PV modules showed that the two organizations differ by less than 0.15% when measuring the performance of photovoltaic modules. The Fraunhofer team said this was one of the key indications of whether their analysis is on the right track.
Earlier in January 2025, China’s Fujian Metrology Institute (FMI) and the National Photovoltaic Industry Measurement and Testing Center (NPVM) announced the establishment of a metrology traceability system for both silicon and perovskite solar cells. The calibration system consists of a monochromatic light system, a bias light system, a 3D motion measurement platform with temperature control and an electrical measurement system.
This content is copyrighted and may not be reused. If you would like to collaborate with us and reuse some of our content, please contact: editors@pv-magazine.com.
