Energy Convers. Manag.: Evaluating the Global Applicability of Spectral Indices for Photovoltaic Performance Analysis
The research article was published by Christian A. Gueymard from Solar Consulting Services.
- Spectral indices are commonly used to assess the spectral impact on different photovoltaic cells.
- A study was conducted on three indices (APE, BF, and UF) across six photovoltaic-centric wavebands.
- Thousands of direct and global spectra were generated and analyzed using the SMARTS model.
- These spectra represent a broad range of atmospheric conditions, suitable for diverse geographical applications.
- APE demonstrated the best performance; however, no index is truly injective due to compensations between blue and red shifting effects.
Photovoltaic (PV) cells are sensitive to the incident spectrum, which impacts power output. Output measurements are typically taken under standard testing conditions (STC), including temperature, irradiance, and spectrum distribution. STC defines the reference spectral distribution, often at Air Mass (AM) 1.5. Commonly used reference spectra are AM1.5G and AM1.5D. Spectrum irradiance and atmospheric conditions significantly affect broadband irradiance and spectrum distribution. Recent studies evaluated spectral effects on different PV cell technologies through simulations. To simplify analysis, three spectral indices (APE, BF, and UF) were introduced for PV cell studies. However, there is disagreement on whether these indices accurately represent specific spectra, necessitating further research. The goal of the article is to determine the existence of reliable spectral indices for characterizing any specific spectrum.
Average Photon Energy (APE): A spectral index representing the photon energy distribution of the incident spectrum. It is calculated as the ratio of the photon energy flux to the photon count flux of the incident light, indicating the average energy of the incident photons. A larger APE signifies a blue shift in the incident spectrum, while a smaller APE indicates a red shift.
Useful Fraction (UF): A spectral index describing the effectiveness of the incident spectrum on photovoltaic cells.
Blue Fraction (BF): BF is a spectral index used to describe the blue light component in the incident spectrum. In this study, the boundary is set at 650 nm, and the formula is as follows:
Recent research has focused on the variation of spectral indices (APE, BF, and UF) of PV cells under different atmospheric conditions and air masses (AM). Under identical atmospheric conditions, the air mass (AM) ranges from 1 to 26. Taking AM≤7 as an example, the direct normal irradiance (DNI) in the WB#6 scenario demonstrates that APE and BF decrease as AM increases, while UF exhibits unexpected stability. Similar observations are noted for global horizontal irradiance (GHI), indicating the unreliability of UF, an issue not reported in the literature to date. On the other hand, altering atmospheric conditions (AM1.5) involves atmospheric mass, aerosol optical thickness, Ångström exponent, and precipitable water vapor, leading to diverse effects on spectral indices. Different atmospheric conditions have a complex impact on spectral indices. An increase in air mass leads to redshift, further redshift occurs with increased aerosol levels, and blueshift is caused by water vapor.
Overall, APE exhibits stronger discriminative capability and is beneficial for practical measurements, while UF tends to overly manifest neutral effects with smaller differences. The variations in APE are more pronounced than in BF, suggesting APE’s advantage in distinguishing different scenarios, which is crucial given that experimental uncertainties often mask subtle differences between conditions. Furthermore, different atmospheric conditions and air masses may offset the differences in spectral indices. The author compared analogous spectra with the same APE but under different atmospheric conditions. The results showed inconsistent variations in water vapor absorption in the NIR region compared to the UV/visible spectrum, indicating limitations in the assumption of spectral index uniqueness.Hence, when analyzing and comparing various photovoltaic technologies, it is essential to carefully consider the intricate interplay among atmospheric conditions, air masses, and spectral indices.
- Used SMARTS model to generate thousands of direct and global solar spectra covering diverse atmospheric conditions.
- Calculated APE, BF and UF spectral indices from modeled spectra across 6 PV-relevant wavebands.
- Analyzed impact of inputs like air mass, aerosol, water vapor on indices.
- Evaluated index performance compared to reference spectra.
- Assessed bijectivity between indices and spectral distributions.
In the discussion, it was observed that the spectral indices (APE, BF, UF) play a crucial role in representing solar irradiance for PV cells. UF exhibited inconsistent behavior, indicating limitations. APE and BF were reliable, with APE showing higher responsiveness to atmospheric changes, making it preferable. Notably, the study highlighted a key finding: none of the indices were bijective due to compensations between red-shifting and blue-shifting atmospheric processes, leading to identical index values for different spectral distributions. This insight underscores the complexity of accurately characterizing solar spectral distributions.