2
whether the efficiency varies with the weather in the way we expect it to or in some other way.
Thus, the ideal irradiance sensor for monitoring PV performance would also vary with the
weather in a way that mimics how a “good” or “defect-free” PV product is expected to vary. In
other words, if one can accurately measure the total useable incident irradiance under any
environmental conditions (fuel in) one can determine if the PV system is generating electricity
per expectations (energy out).
Pyranometers
Meteorologists around the world use pyranometers to quantify the sunshine. Pyranometers
respond to the change in temperature when the sunlight heats a black surface. The pyranometer
gives a voltage signal that is directly proportional to the irradiance as measured in watts per
square meter. The pyranometer is specially designed to accept light from all angles, to have a
flat response to light from the ultraviolet to the far infrared, and to have a stable output regardless
of sky conditions and changing ambient conditions. A variety of products are on the market
today, typically with the most stable instruments having the highest prices. Typically
pyranometers are recalibrated about once per year, since the output can drift. A photograph is
shown in Fig. 1 of a typical state-of-the-art commercial unit.
Reference cells
PV reference cells can also be used to measure irradiance. However, they work in a very
different way: photons with energy above the band gap of the PV material are converted directly
into positive and negative charges that can be collected and used in an external circuit. The
reference cell generates a current that is dependent on the number and spectral distribution of the
photons. Typically, the current of the reference cell is measured by measuring the voltage across
a small resistor that is included in the reference cell package. This voltage is calibrated under the
reference spectrum [ASTM G173, or IEC 60904-3 spectrum] at 1000 W/m
2
, 25 °C using
standard techniques [IEC 60904-1]. Figure 2 shows an ESTI type reference cell where two half
cells are utilized to measure irradiance and cell temperature. Figure 3 compares the responsivity
of thermal based pyranometers with PV sensors and the reference spectrum. Like a PV module a
reference cell responds to light from all angles, but typically shows an increased reflectance, and,
therefore, a decreased efficiency for light that arrives at a glancing angle. If a solar reference cell
is constructed with typical PV cells, glass, encapsulant, and backsheet the spectral and angular
response will closely match that of the PV modules generating energy in the power plant. The
reference cell does not respond to photons with energy less than the band gap, implying that they
will be insensitive to changes in this part of the spectrum when used for quantifying the
broadband meteorological irradiance. Or, in other words, a reference cell is designed to measure
the irradiance that is available to a PV module for conversion into electricity (fuel in) rather than
being designed to measure the broadband irradiance. The close spectral match of the reference
cell to a PV system minimizes scatter in the data due to variable spectral conditions. The
attributes that introduce uncertainty when reference cells are used to characterize the weather are
the same as the attributes that make them more ideal for characterizing the PV system
performance.