Optical Properties of Solar Absorbers – Results on Round Robin and Guidelines

The scope of the standard ISO 22975-3:2014 is the determination of the long term behavior and service life of selective solar absorbers for use in vented flat plate collectors used in typical solar domestic hot water systems or combisystems. With this scope, this standard can be the basis for the certification of solar absorbers in the framework of Solar Keymark European Certification Scheme. Measurements of optical properties are fundamental to support this certification. Six test laboratories participated in a Round Robin test for the measurement of solar absorptance and thermal emittance of different absorber coatings. The samples used were one PVD coating and two selective paint coatings all on aluminum substrates. The results were submitted anonymously and were analyzed according to standard ISO 13528:2015. The results obtained showed good agreement among the measurements made by the participants and allowed the preparation of guidelines to guarantee quality of the measurements of optical properties of solar absorbers. A proposal to include absorber coatings in the Solar Keymark certification scheme was prepared.


Introduction
In the framework of the international standardization committee ISO TC 180, standards for collector components and materials are being developed. The first standard published is ISO 22975-3:2014. The scope of this standard covers the determination of the long-term behavior and service life of selective solar absorbers for use in vented flat plate collectors used in typical solar domestic hot water systems or combisystems. This standard can be the basis for the certification of solar absorber coatings in the framework of the Solar Keymark Certification Scheme. ISO 22975-3:2014 uses a failure criterion based on the optical properties for the evaluation of the durability of the absorbers. The measurement of the optical properties, namely solar absorptance and thermal emittance, is to be performed according to Annex A of ISO 22975-3:2014. In order to evaluate the quality of the measurements performed according to this standard, six laboratories cooperated to perform a Round RobinRound Robin for the measurement of these parameters.

Round Robin Procedure
The main purpose of this Round Robin was to compare the solar absorptance and thermal emittance measurements of absorber coatings for flat plate collectors. The Round Robin procedure was agreed among the six test laboratories and based on Annex A of ISO 22975-3:2014. Six samples of three different selective coatings (one PVD and two paint coatings) were distributed to the participating laboratories and two measurement rounds were performed. Each laboratory measured three samples during the first round, then sent those three samples to the next laboratory and finally measured the next three samples that it had received from another laboratory during the second round.
Results were submitted anonymously to the evaluator. The standard ISO 13528:2015 was used to analyze the results of the Round Robin.

Evaluation of spectral absorptance values and spectral emittance values
According to Annex A of ISO 22975-3:2014, the evaluation of the solar absorptance (α) is based on the measurement of spectral near-normal/hemispherical reflectance of the sample, R s () and measurement of spectral near-normal/hemispherical reflectance of the reference, R r (). The spectral range for this measurement is from 0.30 m to 2.50 m, with at least 40 wavelengths measured. The calculation of near-normal/hemispherical spectral absorptance () is given by: where: is the near-normal/ hemispherical reflectance of the reference.
According to Annex A from ISO 22975-3 (2014), the evaluation of the thermal emittance (T) is also based on the measurement of spectral near-normal/hemispherical reflectance of the sample, R s () and measurement of spectral near-normal/hemispherical reflectance of the reference, R r ().
The spectral range for this measurement is from 2.0 m to 50.0 m. If only a narrower wavelength range is available for measurements (e.g. up to 25 m), the last measured value can be considered as constant up to 50 m The calculation of near-normal/hemispherical spectral emittance (T) is given by:

( e q . 2 )
where: is the near-normal/ hemispherical reflectance of the reference.

Determination of solar absorptance and thermal emittance of absorber surfaces
According to Annex A from ISO 22975-3:2014, the solar absorptance α s is calculated from the following expression: where () is the near-normal/hemispherical spectral absorptance as defined in Equation 1 is the reference solar spectral irradiance (ISO 9050:2003) λ is the selected measuring wavelength ∆λ is the respective wavelength interval The thermal emittance (T) is calculated from the following expression: where is the value of the Planck function for black body radiation at the selected wavelengths and temperature () is the near-normal/hemispherical spectral emittance as defined in Equation 2 λ is the selected measuring wavelengh ∆λ is the respective wavelength interval The Planck function is defined by the following expression: , ( e q .

Analyses of Round Robin results
The analyses of the Round Robin results was performed according to standard ISO 13528: 2015.
For the calculation of the standard deviation for proficiency assessment (Round Robinn),  pt , the normalized interquartile range was considered, as seen in Eq. 6. This is considered as a robust estimator of the standard deviation of the participant results. .

( e q . 6 )
where: A correction for non-homogeneity of the distributed samples was introduced by considering the estimator of standard deviation given by Eq. 7: where s s is the "between samples" standard deviation as defined in ISO 13528:2015. The comparison of the results was made using the z-score value defined by: ( e q . 8 ) where: x pt is the assigned value and x i the measured value.
The assigned value of the proficiency test (Round Robin) is the median of the values measured by each participant.
The interpretation of z-scores according to ISO 13528:2015 is:  A result that gives |z|  2.0 is considered to be satisfactory;  A result that gives 2.0  |z|  3.0 is considered to be questionable;  A result that gives |z|  3.0 is considered unsatisfactory.

Results and discussion
To check the homogeneity of the test items (samples distributed), a measurement of the optical properties of all samples was made by LNEG before the distribution of samples.
According to ISO 22975-2 the test samples to be used in the durability tests shall have standard deviation in the determined values of solar absorptance of less than 0.01 and for the determined values of thermal emittance of less than 0.04.
The sets of three samples, each with six specimens, distributed and initially measured by LNEG showed the following standard deviations (table 1), which satisfy the criteria of ISO 22975-2. As a criterion for homogeneity, the standard ISO 13528:2015 indicates that the "between specimens" standard deviation, s s , and the standard deviation for proficiency assessment,  pt , shall satisfy Eq. 9: , ( e q . 9 ) The calculation of the "between sample" standard deviation was performed according to section B3, of Annex B of the standard and used, for each sample type, three repeated measurements for each of six samples. The number of specimens recommended by ISO 13528:2015 for the homogeneity check is 10 but the Round Robin only has six partners; for this reason, only six specimens of each sample are considered.
For each sample type and for each parameter measured (solar absorptance and thermal emittance), the values of s s and  pt are listed in Table 2. For calculation of  pt , all measurements were considered (first and second round measurement). In Table 3, the median values x pt of each parameter and sample type are presented as well as the corresponding standard deviation  pt of the proficiency test and the values corresponding to the z-score limits. Table 4 shows the same values considering the homogeneity correction.

Guidelines
Based on the results of the Round Robin, guidelines were prepared for the determination of the parameters, solar absorptance and thermal emittance ,. The guidelines present the necessary equations for calculation of the parameters, which are as indicated in section 2. of this article.
Additional aspects to be considered in the measurement are: a. Samples with a linear structure (typical of rolled metal) should be mounted for be measured according to the schematic representation of Fig. 7. This is to increase the probability of determining an accurate near-normal/hemispherical reflectance value for the spatially inhomogeneous reflection ("streak") in widespread commercial integrating spheres. To evaluate whether this is necessary, photographs with high resolution e.g. using SEM -scanning electron microscope, may be needed, but often the structures are visible to the naked eye. b.
Adequate sampling for representativeness of production line variation but the fulfilling requirements of ISO 22975-3:2014; c.
Correct maintenance of measuring equipment -annual maintenance is a good practice; d.
Traceability of the calibration spectra of reference materials R_r (λ) used in equations 1 and 2.

Proposal for Solar Keymark Scheme Rules
In the framework of the Solar Certification Fund project that supported this work, a proposal for revision of Solar Keymark Scheme rules was prepared to allow the possibility to certify absorber coatings according to ISO 22975-3:2014.
It is considered that the selection of samples for the absorber coating certification shall be done such that it is representative of the whole production (e.g. samples from the left side of the coil, from the middle and from the right side). The test report shall include photos of the samples used in the tests (before and after each test) and the test laboratory shall store samples in good conditions for any necessary future measurements or tests. For instance, the standard conditions for sample conditioning can be considered as 23 ± 2 °C and 50 ± 5 % relative humidity, according to the standard ISO 3270:1984.

Conclusions
It is possible to conclude that the measurement of solar absorptance and thermal emittance can be done with a majority of satisfactory results by the different test laboratories. Guidelines were prepared that can support other laboratories to perform good quality measurements. For future such interlaboratory comparisons intended to demonstrate measurement proficiency, it is strongly recommended that the homogeneity of specimens for a single sample type is ensured before distribution, so that the effects of sample inhomogeneity do not have to be compensated by additional statistical analysis.
based on the results of the Round Robin a document with guidelines for the measurement of the optical properties