Search

We perform a global horizontal (GHI) and direct normal (DNI) surface solar irradiance comparison between the National Solar Radiation Database (NSRDB V3) and four Canadian Solar Spectral Irradiance Meter Network (CanSIM) stations for 2018-2019. On average, under all-sky conditions, the GHI from the NSRDB had the mean absolute error (MAE) and the root mean square error (RMSE) of 14% and 18%, respectively, while for the DNI those metrics were 15% and 24%, respectively. By using the sky condition classification obtained from a spectral pyranometer, we estimated that 7% to 17% (depending on location studied) of NSRDB cloudy sky data is misattributed as such due to satellite parallax and snow cover errors. The lower bound on parallax error varied from 5% and 12%, while the upper bound on the cloud-snow discrimination error ranged from 2% to 5%.

This manuscript describes a decomposition algorithm for deriving the broadband direct normal (DNI) and diffuse horizontal (DHI) irradiances from one-minute spectral global horizontal irradiance measurements performed by the SolarSIM-G. The algorithm was calibrated and validated at four stations across Canada (Ottawa, Varennes, Egbert, and Devon) and at one station in China (Xianghe), which cumulatively represent a seven-year data set. For the DNI estimation, the root mean square error (RMSE) ranged from 26 W/m2 to 48 W/m2, while the largest mean bias error (MBE) was 4 W/m2. For the DHI estimation, the RMSE ranged from 14 W/m2 to 27 W/m2, while the largest MBE was 3 W/m2. In addition, the integrated DNI and DHI errors for each station were less than 1% and 2%, respectively. The described method is an alternative to other measurement techniques for obtaining all three irradiance components.

A comprehensive uncertainty analysis validates a Solar Spectral Irradiance Meter (SolarSIM) for accurately resolving the spectral and broadband direct normal irradiances (DNI), spectral aerosol optical depth (AOD), precipitable water vapour and atmospheric total column ozone amounts. The derivation of these parameters from four SolarSIMs were compared to reference instrumentation at the Physikalisch-Meteorologisches Observatorium Davos and World Radiation Center (PMOD/WRC) in Davos, Switzerland in September 2015. The SolarSIMs are the first instruments to ever simultaneously participate in the 12th WMO International Pyrheliometer Comparison, Fourth Filter Radiometer Comparison, and First Spectroradiometer Comparison. The SolarSIMs’ DNI data were compared to the World Standard Group’s PMO2 absolute cavity radiometer, with World Radiometric Reference factors ranging from 0.999674 to 0.994610 for the best and the worst performing devices, respectively. In addition, the SolarSIMs’ spectral DNI data was compared against PMOD’s Precision Spectral Radiometer. The mean difference of the spectral DNI was found to be less than 5% for wavelengths above 400 nm. The SolarSIMs’ measurements of AOD data were compared against PMOD’s Precision Filter Radiometer triad. The median AOD differences and their standard deviations were found to be 0.0046 ± 0.0044, 0.0016 ± 0.0034, 0.0018 ± 0.0026, and 0.0041 ± 0.0022 for 368 nm, 412 nm, 500 nm, and 865 nm, respectively. The SolarSIMs’ measurements of precipitable water vapour were compared against PMOD’s Cimel CE318 sun photometer. The median difference and the corresponding standard deviation averaged 1 ± 0.2 mm for all SolarSIMs. Furthermore, the SolarSIMs’ measurements of total column atmospheric ozone were compared against PMOD’s Brewer MkIII spectrophotometer. The median difference and the corresponding standard deviation averaged 6 ± 7 DU for all SolarSIMs.

In this work, we demonstrate the capabilities of the SolarSIM-D2 in deriving the spectrally corrected DNI for a hypothetical concentrating photovoltaic (CPV) mono-module consisting of Azur Space’s 3C42 triple-junction solar cell with a silicone on glass Fresnel lens. The annual spectrally corrected and measured DNI sums in Ottawa, Canada were 1114.7 kWh/m2 and 1145.6 kWh/m2, respectively, representing a difference of 2.8% - which is significant in the context of developing a CPV powerplant.

A solar spectral irradiance meter (SSIM), designed for measuring the direct normal irradiance (DNI) in six wavelength bands, has been combined with models to determine key atmospheric transmittances and the resulting spectral irradiance distribution of DNI under all sky conditions. The design principles of the SSIM, implementation of a parameterized transmittance model, and field performance comparisons of modeled solar spectra with reference radiometer measurements are presented. Two SSIMs were tested and calibrated at the National Renewable Energy Laboratory (NREL) against four spectroradiometers and an absolute cavity radiometer. The SSIMs’ DNI was on average within 1% of the DNI values reported by one of NREL’s primary absolute cavity radiometers. An additional SSIM was installed at the SUNLAB Outdoor Test Facility in September 2014, with ongoing collection of environmental and spectral data. The SSIM’s performance in Ottawa was compared against a commercial pyrheliometer and a spectroradiometer over an eight month study. The difference in integrated daily spectral irradiance between the SSIM and the ASD spectroradiometer was found to be less than 1%. The cumulative energy density collected by the SSIM over this duration agreed with that measured by an Eppley model NIP pyrheliometer to within 0.5%. No degradation was observed.

Two global solar spectral irradiance meters (SolarSIM-Gs) were deployed and tested at the National Renewable Energy Laboratory in Golden, USA from 18 September to 6 October 2017. The instruments were installed in the global horizontal orientation at the Solar Radiation Research Laboratory and their measurements were compared against colocated secondary reference spectroradiometers and secondary standard pyranometers. The SolarSIM-Gs’ spectral global horizontal irradiance (GHI) accuracy was shown to be $\lt 5$% on average per wavelength in the 290-1650 nm range for over 1,900 analyzed spectra. The broadband GHI measurement accuracy for two instruments was shown to have mean bias and standard deviation of 0.97, 1.23 W/m2 and 5.27, 5.65 W/m2, respectively, for over 10,000 analyzed data points.

This study presents the results of the 4th Filter Radiometer Comparison that was held in Davos, Switzerland, between September 28 and October 16, 2015. Thirty filter radiometers and spectroradiometers from 12 countries participated including reference instruments from global aerosol networks. The absolute differences of all instruments compared to the reference have been based on the WMO criterion defined as 95 % of the measured data has to be within 0.005 ± 0.001/m (where m is the air mass). At least 24 out of 29 instruments achieved this goal at both 500 and 865 nm, while 12 out of 17 and 13 out of 21 achieved this at 368 and 412 nm, respectively. While searching for sources of differences among different instruments, it was found that all individual differences linked to Rayleigh, NO2, ozone, water vapor calculations and related optical depths and air mass calculations were smaller than 0.01 in AOD at 500 and 865 nm. Different cloud detecting algorithms used have been compared. Ångström exponent calculations showed relatively large differences among different instruments partly because of the sensitivity of this parameter at low AOD conditions. The overall low deviations of these AOD results and the high accuracy of reference aerosol network instruments demonstrated a promising framework to achieve homogeneity, compatibility and harmonization among the different spectral AOD networks in the near-future.

Three of seven stations have been deployed as part of the Canadian Solar Spectral Irradiance (CanSIM) network situated in Ottawa, Varennes and Egbert to measure long term spectral variation of the direct normal (DNI) and global horizontal irradiances (GHI) across the country. Every station is equipped with a solar tracker, SolarSIM-D2+, SolarSIM-G+, and SR20 pyranometer, reporting the spectral DNI, GHI, diffuse horizontal irradiance (DHI) and aerosol optical depth in the 280–4000 nm range, broadband DNI, GHI, and DHI, atmospheric total column ozone and water vapour amounts. The spectral GHI as measured by the SolarSIM-G+ was within 5% as compared to EKO MS-700 spectroradiometer in 350–1050 nm range on 17 March 2017. The difference in the GHI as reported by SolarSIM-G+ and SR20 pyranometer from all stations was within 2% on 14 April 2017. Furthermore, on this day, the daily GHI sum for the Ottawa, Varennes, and Egbert stations was 7.01, 6.95, and 7.11 kWh/m2, respectively, while the daily DNI sum was 10.65...

A next generation solar spectral irradiance meter (SSIM) was installed at the University of Ottawa solar test site in September 2014, for ongoing collection of environmental and spectral data. The instrument's performance is compared against a commercial pyrheliometer and a spectroradiometer during an eight month study where ambient temperature fluctuates from −30°C to 30°C. The cumulative solar energy measured by the SSIM over the duration of the experiment agreed to within 0.5% as compared to the Eppley pyrheliometer. A good spectral agreement between the SSIM and the ASD FieldSpec 3 spectroradiometer is observed, with the integral of the spectral irradiance agreeing to within 1% for both instruments. No degradation is observed at any point during this investigation.

The SSIM prototype was designed at the University of Ottawa as a cost-effective alternative to a field spectrora-diometer. The instrument was installed at the University of Ottawa’s CPV testing facility in September, 2013, collecting the environmental and spectral data from October, 2013 to March, 2014. The SSIM’s performance was compared against an ASD field spectroradiometer and an Eppley pyrheliometer during a six month study. It was observed that the SSIM can accurately reproduce the solar spectrum and the direct normal irradiance. The mean difference between the SSIM and the Eppley pyrheliometer was within ±1.5% for cloudless periods in October, 2013. However, bandpass filter degradation and moisture ingress limited the long term performance of the device.

One of the challenges associated with forecasting and evaluating concentrating photovoltaic system (CPV) performance in diverse locations is the lack of high-quality spectral solar resource data. Various local atmospheric conditions such as air mass, aerosols, and atmospheric gases affect daily CPV module operation. A multi-channel filter radiometer (MFCR) can be used to quantify these effects at relatively low cost. The proposed method of selectively sampling the solar spectrum at specific wavelength channels to spectrally reconstruct incident irradiance is described and extensively analyzed. Field spectroradiometer (FSR) measurements at the University of Ottawa's CPV testing facility (45.42°N, 75.68°W) are fed into our model to mimic the outputs from the MCFR. The analysis is performed over a two year period (2011-2012), using 46,564 spectra. A recommendation is made to use four aerosols channels at 420, 500, 780, and 1050 nm, one ozone channel at 610 nm and one water vapour channel at 940 nm, all of which can be measured with ubiquitous Si photodiodes. A simulation of this MFCR channel configuration produces an RMS error under 1.5% over 96% of the 350-1830 nm range, when compared with the FSR, for the 2012 data set in Ottawa.

We discuss intensity and temperature effects on solar cell performance. We simulate and measure these effects in III-V triple-junction concentrator cells incorporating quantum dots under AM1.5D illumination. Reducing series resistance improved simulated efficiency by 4.3%, and improving sub-cell design reduced power degradation with increasing temperature.