Solar simulator

Werner Herrmann TÜV Rheinland Energy GmbH, 51101 Cologne, Germany

Introduction

Solar simulators are specific test devices that are used to evaluate the characteristics of solar cells and PV modules (PV devices). They use a light source with a spectral distribution similar to natural sunlight. Measurements with solar simulators are often referred to simply as indoor measurements. The advantages of solar simulators are obvious:

• Measurements are independent on the weather conditions.
• A high reproducibility is achieved because the test conditions can be adjusted to the desired ranges of module temperature and irradiance.

For testing and qualification of PV devices solar simulators are mainly used for two purposes:

a) Determination of the performance characteristics of PV devices

In this area, so-called pulsed solar simulators with a pulse length typically less than one second are used. The maximum output power determination (power measurement) of PV devices, which is based on the measurement of the current-voltage characteristic (I-V curve), is the best-known representative.

The most important application case of solar simulators is the power rating of PV devices. A performance comparison is made using the nominal power, which is commonly referenced to the Standard Test Conditions (STC). These are defined in the standard IEC 61836:

• Effective irradiance: 1000 W/m²
• Spectral irradiance distribution: Air Mass (AM) 1.5 global acc. to IEC 60904-3
• PV device temperature: 25°C

Measuring techniques for solar simulators are aiming to measure as close as possible to these conditions.

b) Long term irradiance exposure of PV devices

Solar cells and PV modules should provide long-term stability of electrical parameters and used materials such as polymeric components.

In this area, so-called steady-state solar simulators are used, which have irradiation periods of several days or weeks.

Application range of solar simulators in PV

In the field of PV, the application range of solar simulators includes the following areas:

a) Performance measurement of PV devices (solar cell or PV module) at STC level: IEC 60904-1,IEC 60904-3, IEC 60904-8, IEC 60904-10

In the PV laboratory this comprises the calibration measurement of PV devices (performed at STC), whereas in PV production lines it means the power class sorting of cells or PV modules at the end of the production process.

b) Determination of PV module parameters for temperature and irradiance correction: IEC 60891

The knowledge of the I-V correction parameters (such as temperature coefficients) is relevant when the measurement conditions of I-V measurement differ from the STC. The greater the deviation of the measurement conditions from the target conditions (i.e. STC), the greater the uncertainty related to I-V correction.

c) Irradiance – temperature (G-T) matrix of PV modules (Energy rating matrix): IEC 61853-1

This standard defines the measurement of the so-called performance matrix of a PV module, which covers the electrical performance under variable device temperature and irradiance. For that purpose, the PV module is to be installed in a test chamber, in which the temperature can be set in the range of 15°C to 75°C. Irradiance variation in the range of 100 W/m² to 1100 W/m² is typically done with attenuator masks or by changing the lamp power. Both methods may have an influence on the spectral irradiance and the spatial non-uniformity of irradiance in the test area. Consequently, re-calibration of the solar simulator of the may become necessary for different irradiance levels (see section 4).

d) PV module energy rating - Angular responsivity: IEC 61853-2

Transmittance losses of the cover glass of PV modules become relevant for angles of incidence larger than 50°. This IEC standard defines the measurement of the angular transmittance of a PV module. The PV module is rotated in the light field of the solar simulator. For this application, a slow divergence of the light beam is required to keep spatial non-uniformity effects low.

e) PV module spectral responsivity: IEC 60904-8

By putting successively narrow band filter glasses into an appropriate position of the beam, the spectral responsivity of a solar module can be determined. If an LED-based solar simulator is used, the spectral responsivity can be measured by the variation of different LED colours.

f) Electrical stabilization of PV modules: IEC 61215-1-X

This series of IEC standards was developed for the range of thin-film PV modules, which have the property that they are not electrically stable, but dependent on the exposure to temperature and irradiance. The stabilization measurement requires a continuous irradiation and therefore the use of a steady-state solar simulator, with irradiation periods of several days.

g) Quantification of output power degradation: IEC 61215-1

This standard defines test procedures for qualification testing of PV modules. In the testing sequences solar simulators are used to monitor changes in the electrical performance of PV modules that may occur during stress tests such as damp heat, thermal cycling or mechanical load. Due to the high reproducibility of power measurements, changes in the electrical output power of less than +/-0.5% can be detected.

Main components of solar simulators and range of technical design

Various types of solar simulators are used as light source to determine the current-voltage (I-V) characteristics of PV devices as defined in IEC 60904-1. These can work as single lamp systems or multiple lamp systems, where a single lamp just irradiates a part of the test area. The following chapters only consider flash-type (pulsed) solar simulators, which are typically used for output power measurement of PV modules. Nonetheless many of the considerations are equally relevant to steady-state simulators. Pulsed solar simulators consist of the light unit, a testing chamber and the I-V data acquisition system. Table 1 gives some technical information about the main components and Figure 1 shows typical designs of solar simulators.

Light unit	Test room	I-V data acquisition unit
Most widely used are Xenon lamps that require a specific pulse generator, which is capable to deliver a constant discharge current. The generator needs to be recharged after each flash. Xenon lamps are high-pressure gas discharge lamps. They use the radiation emission of an arc, have high luminance levels and have a continuous spectrum. Optical filtering methods are required to match the AM1.5 spectrum. In the recent years the LED simulator technology has rapidly developed. As a single LED only emits light in a limited spectral range, a range of colors must be used to cover the wavelength range of interest and deliver a continuous spectrum. Both types of light source are capable to deliver a pulse length in the range of 100 ms.	This can be a dark room (i.e. tunnel) with reflective or non-reflective walls or baffles in the beam of light to suppress internal reflections. Usually the design implies several meters distance between the lamp and the test area, which requires a lot of space. This type is typically used for PV module calibration measurement in testing laboratories. Also compact solutions for solar simulators are available. Here a number of lamp(s) are operated in a casing with integrated optical features for light homogenization in the test area (such as reflectors or diffusors). These table-top configurations typically show less than one meter distance between the light source(s) and the PV module to be measured. Both solar simulator types are illustrated in Figure 2.	Accurate I-V measurement requires a data acquisition system, which allows time synchronous measurement of irradiance in the test area, module voltage and module current. The irradiance measurement serves not only to adjust the irradiance to the target value, but also to monitor irradiance fluctuations during the pulse (I-V data acquisition). The PV module is connected to an electronic load, which drives through the total I-V curve in the course of a flash. The resulting I-V curve is typically composed of several hundred I-V data points. Active electronic loads are capable to measure the I-V curve in 4 quadrants. They also offer the possibility to measure the dark I-V curve (carrier injection by electrical means instead of light), which may be required for analysis purposes.

Table 1: Main components of solar simulators used for output power measurement of PV modules

Figur 1 und Figur 2

Quality indicators for solar simulators

Solar simulators are not perfect light sources, and the quality of emitted light can strongly influence the result of the PV power measurement. In particular, the parameters explained in the following sub-chapters must be considered:

Pulse length

The pulse length of a flash determines the I-V data acquisition time of power measurements. For commercially available solar simulators it lies in the range of 10 ms to several 100 ms. It must be noted that a long pulse length may be required for c-Si PV modules with high efficiency cells, such as heterojunction or PERC technologies. Transient effects associated with the internal capacitance may cause artefacts that show up in I-V curve segment around the maximum power point (MPP), if the voltage sweep rate is too high.

Whether a PV module is sensitive to transient effects or not can be determined by comparing the I-V curves resulting from the forward and backward voltage sweep directions. To solve the problem for short pulse length solar simulators, multi-flash measurement techniques that measure and assemble segments of the I-V curve are available. Furthermore, various software solutions that control the distribution of the I-V data points on the curve or imply a specific correction procedure have been developed.

Spectral irradiance distribution of the lamp

The responsivity of solar cells is strongly dependent on the wavelength. Spectral deviations of the solar simulator spectrum to the AM1.5 reference spectrum become a relevant error source when reference device and test device are not spectrally matched. In this case, the measured irradiance with the reference device is to be corrected with a spectral mismatch factor. The procedure for spectral mismatch correction is defined in IEC 60904-7.

The situation becomes even more complex for multi-junction PV devices. Here spectral differences will cause a current mismatch between junctions. In such cases specific methods of light filtering or tuning the simulator light spectrum must be applied to reduce measurement errors.

Uniformity of irradiance in the test area

Non-uniform illumination of a PV module is a major uncertainty source for output power characterization of PV modules. It will cause a scatter of short circuit currents (I_SC) of serially connected cells, which overlaps the normal scatter that is caused by cell manufacturing tolerances.

During the I-V measurement of a PV module its current is varied from zero to a value that exceeds the maximum I_SC of the cells. For each impressed value current the module voltage will result from the voltage contributions of individual cells, which result from the working points on their I-V curves. Two cases must now be distinguished: a) Voltage contributions from cells are positive when the I_SC value of the cell is higher than the impressed module current. b) But voltage contributions can be also negative when the I_SC value of the cell is lower than the impressed module current. In the second case the cell will get reverse biased and will operate on its reverse I-V curve. Because reverse cell I-V curves reach up to high negative voltage, the resulting negative voltage contributions can be a multiple of the positive contributions. This will result in a deformation of the I-V curve, which is mainly obvious in the I-V curve segment between I_SC and P_MAX (horizontal leg).

If the cell strings in a PV module are not evenly irradiated, kinks in the horizontal leg of the I-V curve will be additionally observed. These are associated with the function of the bypass diodes.

Because the P_MAX value generally shows a weak dependency on spatial non-uniformity in the range <5%, this parameter shall be preferred for setting the solar simulator.

A specific uncertainty aspect for test laboratories is associated with the experimental set-up. A small sized reference cell is commonly mounted at a fixed position in the test area and measuring the local irradiance at this point. On the other hand, the PV module to be measured spans a large area, which is subject to spatial irradiance non-uniformity. Then, the effective irradiance for the module is given by the average irradiance in the module area, which may differ from the irradiance measured by the reference cell. Therefore, a correction must be applied to the irradiance measured with the reference cell. This will depend on the size of the module under test.

Temporal instability of irradiance

During I-V data acquisition the irradiance is not completely stable but subject to light fluctuations. As the photocurrent generation of cells follows these fluctuations, an irradiance correction for each I-V data point to the target irradiance level is required.

I-V translation of a measured module I-V curve (raw data) to the target irradiance is performed in accordance with correction methods that are defined in the standard IEC 60891. It must be noted that translation errors may occur, which depend on the accuracy of the module I-V correction parameters (such as internal series resistance) and the extent of the irradiance correction.

Today commercially available solar simulators typically show an irradiance fluctuation less than 1% so that the impact of temporal instability plays a minor role in PV power measurement.

Effective irradiance

For power rating measurements of PV modules the lamp power of the solar simulator must be adjustable to give 1 000 W/m² effective irradiance. The effective irradiance reaching the cells results from the irradiance value measured with the reference device multiplied by correction factors associated with spectral mismatch (see section 4.2) and spatial non-uniformity mismatch (see section 6.1).

Performance evaluation of solar simulators

The standard IEC 60904-9 defines various methodologies for determining the classification of solar simulators for three quality indicators (Table 2). Suppliers of solar simulators specify the respective class for each indicator with a specific letter grade (e.g. AAA):

• Spectral match to AM1.5 global reference spectrum as defined in IEC 60904-3;
• Spatial non-uniformity of irradiance in the designated test area;
• Temporal instability of irradiance during I-V data acquisition.

links|mini|628x628px|Table 2: Methodologies for classification of solar simulators according to IEC 60904-9It must be noted that the standard IEC 60904-9 is currently undergoing a revision process. The publication of the third edition of the standard is expected in 2020. As major change the spectral classification will be extended to the wavelength range of 300 nm to 1200 nm to address the needs of PV industry for accurate measurement of high efficiency solar cells such as PERC. Furthermore a new A+ class (twice as good as class A) will be introduced that addresses the advances in solar simulator technology and allows reducing the uncertainty of reference module calibration in test laboratories.

The use of a solar simulator of a particular class does not eliminate the need of quantifying the influences of the simulator properties on the measurement. For this purpose, the influences of spectral mismatch, irradiance non-uniformity mismatch and temporal stability of irradiance shall be analysed. The solar simulator classification (letter grade) should not be used in isolation to imply any level of measurement confidence or measurement uncertainty. Specifically the spectral classification must be carefully considered because it may not be well correlated with the spectral mismatch. For example, a class C spectrum may deliver a lower spectral mismatch for a certain PV technology than a higher class spectrum. Experience has shown that the classification of a solar simulator is not constant but subject to various factors:

• Ageing of lamp (s) with operation time
• Exchange of lamp(s)
• Lamp power settings
• Use of inserts in the beam of light such as optical filters or light attenuator masks
• Ageing or soiling of inserts
• Reflections from the surroundings such as properties of test room walls

Accordingly, the classification only refers to the actual operating conditions. Ideally, the classification as stated in the product specification or test report shall therefore cover the range of operating conditions during the practical use. It is recommended to periodically review the classification of a solar simulator against the requirements of IEC 60904-9.

Calibration procedure for solar simulators

This section aims to provide practical recommendations for the calibration of a solar simulator used for PV power measurements. Calibration of a solar simulator means that the light intensity in the area of the PV module is adjusted in a way that the effective irradiance at the cells matches the target irradiance. For power rating measurements the STC are commonly used as target conditions as these are the basis for power labelling (nominal output power) of a PV module.

There are different calibration requirements for PV laboratories and for PV production lines. Consequently, the calibration procedures differ in some respects.

PV module calibration in the test lab

Calibration measurements of PV modules in test laboratories are aiming to achieve a low measurement uncertainty. For this reason primary calibrated reference cells are used as reference device. For these ±0.5 % calibration uncertainty can be achieved, if performed by specialized national metrology institutes. The c-Si reference cell with dimensions of 2 cm x 2 cm is encapsulated in a specific standardised housing of so-called World-PV-Scale (WPVS) design (refer to IEC 60904-2 for further information).

The reference cell is normally mounted at a fixed position in the test area outside the expectable dimensions of PV modules to be measured. As previously indicated the measured irradiance value must be corrected for spectral and spatial non-uniformity mismatch.

Figure 3 illustrates the correction procedure for non-uniformity mismatch. The coloured irradiance pattern was measured with 8 cm step width in both directions. The pattern is superimposed by the position of a 60 cell PV module and the position of the reference cell. The spatial non-uniformity in the PV module area is 0.87 %.

The correction for non-uniformity mismatch between the positions of the reference cell and the PV module requires the knowledge of two parameters:

• Average irradiance in the PV module area: 1009.9 W/m² in this example
• Irradiance at the location of the reference cell: 1015.7 W/m² in this example

This means that the effective irradiance seen by the PV module is lower compared to the measured irradiance by the reference cell. The irradiance setting of the solar simulator must delivers an average irradiance in the PV module area of 1000 W/m² when the reference cell measures the following irradiance: [math]\displaystyle{ \frac{1015.7}{1009.9}x1000\frac{W}{m^2}=1005.9\ W/m^2 }[/math]

FIGURE 3

As previously explained, the spectral mismatch (SMM) calculation according to IEC 60904-7 delivers a second correction to the measured irradiance, when the spectral responsivities of the reference device and the PV module are different. The effective irradiance for the PV module results from the measured irradiance multiplied with the spectral mismatch factor. Accordingly the output power measurement must be irradiance corrected or the irradiance adjusted by the factor 1/SMM.

PV module in production lines

For a certain PV module type the adjustment of the solar simulator light intensity is performed with a calibrated reference module, which is placed at a position in the test area, where the production line PV modules shall be measured.

In order to keep the uncertainty of production line measurements low, the reference module shall fulfil the following requirements:

• Size adjusted: Both reference module and production line modules should have the same size and electrical circuitry to avoid measurement errors associated with effects of non-uniformity irradiation.
• Spectrally matched: The reference module should be processed with the same cell type (than the production line modules) to reduce uncertainties related to spectral mismatch.
• Optically matched: The optical properties of materials should reflect the bill-of-materials (BOM) of the production line modules to guarantee comparable transmittance effects and internal reflections.

Because of impacts associated with non-uniform irradiation of the PV module (see chapter 4.3), it will not be possible to exactly reproduce both the calibrated short circuit current (I_SC) and maximum output power (P_MAX) at the same time. This offers two possibilities for adjustment of the light intensity:

• The reference module delivers the calibrated I_SC value
• The reference module delivers calibrated P_MAX value.

Table 3 summarises the pros and cons of both methods. In both cases the second parameter is free variable. Therefore, an additional quality criterion should be defined saying that the maximum deviation from the calibrated values should not exceed a certain limit (i.e. ±0.5%). Any larger discrepancy could indicate the following problems:

• Discrepancy on V_OC [math]\displaystyle{ \Rrightarrow }[/math] Module temperature measurement
• Discrepancy on I_SC [math]\displaystyle{ \Rrightarrow }[/math] High spatial non-uniformity of irradiance, damage of the reference module (cell cracks), high ISC production spread of cells
• Discrepancy on P_MAX [math]\displaystyle{ \Rrightarrow }[/math] Contacting technique of PV module (high contact resistance), sweep rate effects due to short data acquisition time, module temperature measurement

	Referencing calibrated ISC of the reference module	Referencing calibrated PMAX of the reference module
Advantage	Almost independent from module temperature and connection technique	Better compensation of non-uniformity effects
Disadvantage	Non-uniform illumination of a module will mainly affect its ISC. Increase of non-uniformity will cause lower ISC. Thus a higher irradiance setting is required to deliver the calibrated ISC, which results in overestimation of PMAX.	Requires a careful module temperature measurement and module connection technique to the I-V load. Faulty contacts will cause a higher irradiance setting to deliver calibrated PMAX. This causes overestimation of module ISC.

References

International test standards

IEC 60891	Photovoltaic devides – Procedures for temperature and irradiance correction to measured I-V curves
IEC 60904-1	Photovoltaic devices – Part 1: Measurement of photovoltaic current-voltage characteristics
IEC 60904-2	Photovoltaic devices – Part 2: Requirements for reference solar devices
IEC 60904-3	Photovoltaic devices – Part 3: Measurement principles for terrestrial photovoltaic (PV) solar devices with reference spectral irradiance data
IEC 60904-7	Photovoltaic devices – Part 7: Computation of the spectral mismatch correction for measurements of photovoltaic devices
IEC 60904-8	Photovoltaic devices – Part 8: Measurement of spectral responsivity of a photovoltaic (PV) device
IEC 60904-9	Photovoltaic devices – Part 9: Solar simulator performance requirements
IEC 60904-10	Photovoltaic devices – Part 10: Methods of linearity measurement
IEC 61215-1	Terrestrial photovoltaic (PV) modules - Design qualification and type approval – Part 1: Test requirements
IEC 61215-1-1	Terrestrial photovoltaic (PV) modules - Design qualification and type approval – Part 1-1: Special requirements for testing of crystalline silicon photovoltaic (PV) modules
IEC 61215-1-2	Terrestrial photovoltaic (PV) modules - Design qualification and type approval – Part 1-2: Special requirements for testing of thin-film Cadmium Telluride (CdTe) based photovoltaic (PV) modules
IEC 61215-1-3	Terrestrial photovoltaic (PV) modules - Design qualification and type approval – Part 1-3: Special requirements for testing of thin-film amorphous silicon based photovoltaic (PV) modules
IEC 61215-1-4	Terrestrial photovoltaic (PV) modules - Design qualification and type approval – Part 1-3: Special requirements for testing of thin-film Cu(In,GA)(S,Se)2 based photovoltaic (PV) modules
IEC 61853-1	Photovoltaic (PV) module performance testing and energy rating – Part 1: Irradiance and temperature performance measurements and power rating
IEC 61853-2	Photovoltaic (PV) module performance testing and energy rating – Part 2: Spectral responsivity, incidence angle and module operating temperature measurements

Publications on non-uniformity impact

W. Herrmann et al., “Modelling of PV Modules – The Effects of Non-Uniform Irradiance on Performance Measurements with Solar Simulators,” 16th EU PVSEC, 2000

C. Monokroussos et al., “Impact of Calibration Methodology Into the Power Rating of c-Si PV Modules Under Industrial Conditions,” in Proc. 28th EU PVSEC, 2013

H. Wilterdink et al.: “Practical Assessment of Power Rating Uncertainties for Industrial Silicon Modules”, 35th EUPVSEC, 2018

Publications on spectral classification

W. Herrmann et al., “Uncertainty of solar simulator spectral irradiance data and problems with spectral match classification”, 27th EU PVSEC, 2012

R. Galleano et al., “Traceable spectral irradiance measurements in photovoltaics: Results of the PTB and JRC spectroradiometer comparison using different light sources”, Measurement, Volume 124, August 2018, pp. 549-559

Publications on measurement techniques for high capacitance PV devices

A. Virtuani et al.: Fast and accurate methods for the performance testing of highly-efficient c-Si photovoltaic modules using a 10 ms single pulse solar simulator and customized voltage profiles, Meas. Sci. Technol. 23 (2012)

C. Monokroussos et al.: Accurate power measurements of high capacitance PV modules with short pulse simulators in a single flash, 28th EUPVSEC, 2012

K. Ramspeck et al.: Accurate efficiency measurements on very high efficiency silicon solar cells using pulsed light sources, 29th EU PVSEC, 2014