Pyranometer Versus Reference Cell Measurements

Synchronous measurement of irradiance on the solar generator plane using a pyranometer, together with a reference cell calibrated by the module vendor (see Figure 2.45), yield revealing results. Most of the HG readings obtained using reference cells are several percentage points lower than those obtained with a highly accurate CM 11 or MC 21 pyranometer.

Figures 2.46-2.49 show this type of comparison of readings obtained using Siemens M1R reference cells (with 2% accuracy) versus the readings obtained with pyranometers at a PV installation in Liestal

image180

Figure 2.45 Synchronous measurement of irradiance on the module plane using a Siemens M1R mono-c-SI reference cell and a heated and ventilated Kipp&Zonen CM 11 pyranometer

Подпись:
Irradiance GPyr and GRef on a Fine Summer Day

Time (CET, Central European Time)

Irradiance GPyr and GRef on an Overcast Winter Day

image18260

50

CM

E

40

c

О

30

о

с

Ъ

20

ь-

10

0

5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

Time (Central European Time CET)

Figure 2.47 Synchronous measurement of irradiance on the module plane using a Siemens M1R mono-c-SI reference cell and a heated and ventilated Kipp&Zonen CM 11 pyranometer on an overcast winter day at a PV system with a 30° inclination angle in Liestal (elevation 40 m). Interestingly, the pyranometer readings are considerably lower

0 2 3 4 5 6 7 8 9 10 11 12

Month

Figure 2.48 Synchronous measurement of the monthly mean of total daily irradiation HG on the module plane using a Siemens M1R mono-c-SI reference cell and a heated and ventilated Kipp&Zonen CM 11 pyranometer on a sunny summer day at a PV system with a 30° inclination angle in Liestal (elevation 340 m) from 1997 to 2000. Interestingly, the multi-year mean pyranometer readings are 4.2% higher (inclination angle 30°, elevation 340 m) and one on the Jungfraujoch (inclination angle 90°, elevation 3454 m). A more precise analysis reveals that on sunny days in particular (see Figure 2.46) the reference cell readings are generally considerably lower than the pyranometer readings, whereas on days with little direct beam radiation the crystalline reference cells often register somewhat more radiation (see Figure 2.47).

Figure 2.49 Synchronous measurement of the monthly mean of total daily irradiation Hg on the module plane using a Siemens M1R mono-c-SI reference cell and a heated and ventilated Kipp&Zonen CM 21 pyranometer on a sunny summer day at a PV system with a 90° inclination angle on the Jungfraujoch (elevation 3454 m) from 1997 to 2000. Interestingly, the multi-year mean pyranometer readings are 4.2% higher

Inasmuch as there are clear days and overcast days during any given year, these effects on the monthly and annual mean readings cancel each other out to some extent. That said, as a rule the multi-year mean irradiation readings from reference cells were several percentage points lower than the pyranometer readings at the same fixed PV system (see Figures 2.48 and 2.49). Similar and in some cases larger discrepancies were observed at other installations and by other researchers [2.10].

The spectrum of the light source used by vendors to calibrate their reference cells (G — 1 kW/m2) probably contains somewhat more usable light for the solar cells than is the case with natural light. However, in view of the fact that (a) vendors use the same devices to calibrate their modules and reference cells, and (b) vendor module power outputs are specified accordingly, such discrepancies should be taken into account in order to obtain more precise calculations of PV system energy yield.

Such discrepancies could perhaps be reduced for measurement purposes by using the calibration procedure proposed by [Ima92]. However, inasmuch as the spectrum of sunlight is determined not only by the air mass (AM) number but also by current atmospheric water vapour content, calibrations of reference cells in natural sunlight against pyranometers will inevitably be prone to random errors due to the composition of the solar spectrum at any given moment. Hence pyranometer and reference cell G and H readings on various days with similar weather conditions may well vary – in the present case, as noted, by up to several per cent. It would be ideal if reference cells that are highly weather resistant were available from an internationally recognized test lab for each cell technology and if such reference cells were calibrated extremely accurately relative to stable primary standards of the same type.

This spectral mismatch between natural sunlight and the light sources used for calibration is the source of some of the discrepancies in the solar generator correction factor kG and the anticipated value for the ideal case, namely 1 (see Section 2.6 and further details in Chapters 7 and 8). Consequently, calculation simulation software for PV system energy yield, which (as is normally the case) meteorologists use in tandem with pyranometer radiation readings and that do not consider spectral mismatches, mostly provide energy yield figures that are too high by several percentage points, even if other key factors such as reduced radiation attributable to low angles of incidence, horizons, shading, power loss and so on are duly taken into account and the requisite additional work is performed.

Yields as secure as the gold in Fort Knox.

Since 1936, the United States government has safely stored its gold at Fort Knox. Gold has been a dependable investment for centuries. Wise investors today put their money in PV. So it’s only natural that Fort Knox is equipped with a PV plant. No wonder the security experts have chosen inverters by КАСО new energy. But watch out: Some would even steal for maximum PV yields. We say why bother when you can simply buy а КАСО inverter. They are the safest investment around. Ask a dealer today!

new energy.

Updated: August 3, 2015 — 2:31 pm