Photovoltaic systems generally do not contain serviceable elements, which is one of the main advances of the PV technology in relation to other technologies. Nevertheless several technical problems can guide to less energy yield and subsequently to a loss of money. That is why PV systems needs frequent monitoring and checks, beginning with the check of the feed-in over alarms generated by inverters up to online monitoring systems. Additionally a visual inspection should be done frequently, which may show surface pollution, damages or for example birds droppings (which may reduce the power of the module significantly!)
Several standards like the EN 62446 specifies necessary measurements for the technical documentation of a newly build or already working PV installation.
At least if a loss of energy yield is detected, the reason for this misbehaviour has to be detected urgently. Different instruments allow the inspection of different attributes of the PV generator: thermal imaging cameras can show regions with higher (hot spot) or lower temperatures, where these temperature differences may be the result of technical problems in the module under test. Electro-luminescence cameras are able to show defects in the structure of the PV cells very clearly.
But none of these both methods allow the determination of the Peak Power (which is the maximum power under Standard Test Conditions) of the modules under test, which is relevant for claims since this value is warranted at the sales date.
That is why it is so important to check the real installed Peak Power directly after installation and each time you are in doubt. Ideal is the check at the installed generator, since every dismounting/mounting of modules cost a lot of time, money and includes the risk of damages to the modules.
Characteristic Uncovers Errors
After the measurement of the IV-curve of the PV generator (module, string or even array) the curve tracers series PVPM returns the following values: short circuit current Isc, open circuit voltage Voc, the complete characteristic of the generator (that means its behaviour under load between short circuit and open circuit), the current maximum power (which is not the peak power!), the measured temperatures and the irradiance.
Based on these values the PVPM device uses our patented methods to calculate the values estimated at STC (1000W/m2, 25°C, AM1.5): currents, voltages, the peak power, the internal series resistance Rs, the internal parallel resistance Rp, and some more.
The comparison from measured and estimated values (data sheet) gives a quick picture of the performance as well as possible problems of the PV generator.
- low peak power, low fill factor
- low currents
- low voltages
- high series resistance (connector/cabling problems), mostly combined with low peak power
- irregular shape of the IV-curve (mismatch, partial shading, pollution)
The PVPM carries out the measurements very fast (between 20ms and 1s), so that a fluently process in the measurement of big PV fields is given. PVPM devices for up to 1000Vdc and up to 100Adc (for special purposes) are available.
While using this technology a check of the PV generator is done quick and with high accuracy on site.
This measurement approach can be found in-depth in the book „Photovoltaik Engineering – Handbuch für Planung, Entwicklung und Anwendung“, Andreas Wagner, Springer Verlag, ISBN 978-3-540-30732-7.
Properly Prepare Measurement
As for a complete IV-characteristic measurement the PV generator also will be shorted, it must be disconnected from any load or inverter before measuring (otherwise the measurement would short-circuit as well the input of the inverter, which will result in danger and serious damage). On the other hand the short-circuiting of a PV generator, even over a longer period of time, is possible without damage for the PV generator.
One of the elements that decide about the accuracy of the results is the irradiance reference sensor. The standard describes a reference cell, that behaves spectral identical to the module under test. That is why we prefer calibrated reference cells, that are build like a small module, with solar glass cover, EVA embedding and Tedlar back surface. The cell has to be the same material as the module (for example a mono-crystalline sensor for a mono-crystalline module). This set-up will behave like the tested module in nearly every detail and allows good precision of the results.
As explained above, the irradiance sensor is essential for the accuracy of the STC calculations.
Unfortunately these kind of sensors are not available for every technology. Additionally such sensors for thin-film modules, would show (more or less) the same dynamic behavior as their “big brothers” – no favorable conditions for a calibrated reference.
Hint: even a regular PV module can be used as a sensor in special cases.
The irradiance sensor is attached before the characteristics measurement directly beside or next to the PV modules. Theoretically it is conceivable to align the sensor at a different location (in the yard) with the same angles (azimuth and elevation) to the sun as the modules. But since the sensor receives not only direct irradiation but as well diffuse light, the result of the irradiation measurement will depend on the ambiance: a green lawn or a white gravel pit will reflect different portions of light…
This potential source of error should be considered!
There is no “Standard Watt-Peak”
Speaking of accuracy: the accuracy of performance measurements is quite limited in photovoltaic. The “Physikalisch-Technische Bundesanstalt PTB” in Braunschweig, the highest instance for this kind of technology, is able to calibrate for example crystalline single cells with an accuracy of slightly better than 1%.
Every authority depending on these results will have unavoidable less accuracy. In the laboratory of TÜV Rheinland in Cologne an accuracy of 3% for crystalline modules is reached with a still high technical overhead. On-site measurements under natural sunlight can reach an accuracy of about 5%.
A high-accuracy “Standard Watt Peak”, comparable to the “Standard Meter” in Paris for the length, does unfortunately not exist for PV technology.
The Process of Measuring in Detail
The IV-characteristic measurements require sunny, stable weather with a sufficient irradiance in the plane of the modules to be measured. The standard responsible for these measurements (IEC 60904) requests an irradiation of at least 800W/m2. According to our own studies irradiances of 500-600W/m2 are already suitable to carry out a performance check.
To force the PV generator into the various areas of its IV-characteristic, a controlled transistor is used for small power sources (cell or single module). The use of capacitors has prevailed at higher power (strings or arrays): the capacitor is discharged before the measurement, then the PV generator will charge the capacitor. The maximum current of the generator, the short-circuit current, will flow now briefly into the empty capacitor, the voltage is near 0V. While charging the capacitor, the voltage rises and the current decreases until, with full charged capacitor, no more current will flow – the open-circuit voltage is reached.
During this process, many current-voltage points are measured by the meter. This can be represented immediately as diagram on the graphical display. Since this process of charging is very smooth you can measure as well module types with high capacity.
The speed of measurement plays an important role for the accuracy. On the one hand the IV-curve should be recorded in shortest possible time, so that fluctuations in ambient conditions have only a limited impact on the result, on the other side, the characteristic may be distorted if the measurement runs too quickly: many (mostly thin-film) modules have a more or less high capacity, and exactly this capacity “bends” the IV-curve at too short measuring times. Inspection time should be significantly more than 10 ms for a single module for this reason to obtain a precise and reliable characteristic. Unfortunately, this requires a large load capacitor, which inevitably affects the dimensions of the instrument.
What Can the Measurement Fulfill?
It has already been mentioned, that the appropriate generator power, converted to STC conditions (peak power) can be compared directly with the manufacturer’s data. It is clear to read, whether the PV generator reaches the required performance. A second result of the IV-curve measurement (with e.g. a device PVPM1000CX) is the internal series resistance. This value indicates whether the resistances of wiring, plugs, etc. existing in the PV generator are within the range of tolerable values.
If this error appears in the results, it can be eliminated and the correctness subsequently confirmed by a quick new measurement of the characteristic curve – additional power losses are avoided. A measurement of the series internal resistance is possible only with instruments of the type PVPM (patented method).
Shadings are Frequent
At PV generators, a frequent problem is the performance loss due to (partially) shading of modules/strings. Circumstances and shading effects are sometimes harde to notice with the naked eye and are therefore easily overlooked. But even small shaded areas can lead to significant performance and loss of earnings. Shadow effects are very easy to identify in a measured IV-characteristic. The “bump” in the IV-curve immediately suggests this problem. Of course a technical problem of the module can produce the same shape of the IV-curve.
If only one module is affected, and if this module is not found immediately, a little trick can help: cover one module completely with something like a cardboard box and make an IV-curve measurement. If the right module was covered the error will disappear from the IV-curve (because the covered module does not produce power and thus will not affect the IV-curve).
The measurement of I V characteristic is a meaningful instrument to assess the state of the PV generator and allows, compared to the thermal imaging, the determination of power losses, as well as a quantitative evaluation of errors in the generator.
The summary for the practitioner: Fast error detection is the most important requirement for a quick troubleshooting. And an early detection of generator defects facilitates warranty processing.
For further information please contact: Four-C-Tron, Phone: 080-25252506, Email : firstname.lastname@example.org
About the Author