Inverter and PV System Technology
Industry Guide 2013

Many inverters record the most important operational data, evaluate the data automatically and, in the event of a fault, send the operator notifications via email, text message or internet. This is sufficient for basic plant monitoring. However, it only allows obvious faults, such as fault currents or total failure, to be recorded.

In order to determine whether a PV plant is producing optimal yields, the plant data needs to be measured continually, and preferably compared with the actual radiation values present. This is due to the fact that currents and voltages, and consequently feed-in capacities, constantly change depending on meteorological conditions. The operator can only determine whether or not the PV plant’s operational data indicate optimal functioning by directly comparing them with insolation data.

Measuring insolation and output

Solar radiation is measured using either pyranometers or PV sensors (reference cells). A third – more indirect – possibility is to compare a plant’s data with meteorological information and yields from PV plants in the vicinity.

Pyranometers measure insolation with great accuracy. They essentially consist of one or two hemispherical glass domes, a black platelet that acts as an absorbing surface, the thermal elements positioned below this and a metal casing. Solar radiation heats the absorbing surface, the warming of which is directly dependent on the insolation. Insolation can thus be ascertained from the temperature difference between the absorbing surface and the white metal casing. Pyranometers are installed horizontally when meteorological data is needed and in the module plane when PV output requires monitoring.The advantage of high measuring accuracy is, nevertheless, opposed by a serious disadvantage: Due to their thermal functionality, pyranometers are relatively sluggish, which means that they are incapable of accurately detecting rapid insolation fluctuations caused, for example, by scattered clouds.

PV sensors, which are also installed in the module plane so that they are exposed to the same insolation conditions as the modules, provide a cost-effective alternative to the accurate, but slow and expensive, pyranometers. A PV sensor consists of a solar cell which supplies power in proportion to insolation. This power is, however, also dependent on the operating temperature of the solar cell, which means that a temperature sensor is necessary in order to offset thermal effects and determine the exact insolation. However, owing to its limited spectral response, the solar cell cannot detect certain portions of the insolation, and reflection losses may also occur. PV sensors are therefore much less accurate in their measurements of insolation than pyranometers. Despite this, they are often used to monitor PV plants. This is because a PV sensor can be selected to correspond to a plant’s modules. For example, a PV plant consisting of CI/GS/Se thin-film modules is monitored by a PV sensor with a CI/GS/Se solar cell. This simplifies the comparison of instantaneous values, which means that operational faults and defects can be recognized quickly.

Like silicon-based PV sensors, pyranometers measure absolute insolation in watts per square meter (W/m² ). Thanks to their technological differences, however, they are capable of recording slightly different solar spectrums (while silicon cells’ restricted spectral response means they are only able to perceive the “silicon spectrum”, pyranometers are able to register the entire solar spectrum). Consequently, when measuring global irradiation, pyranometers invariably record higher levels of insolation than silicon sensors, and, conversely, lower performance ratio (PR) values, as a result of global insolation being used to calculate the denominator of the PR formula. When exposed to the same level of insolation, modules produce a greater output on a cooler day than on a warm day, meaning that it may be necessary to measure the operating temperature of the modules in order to determine the exact output.

Comparisons with regional meteorological data mean that pyranometers and PV sensors are no longer required. Yield simulations are calculated using data supplied by neighboring meteorological offices and compared with the actual yield. Operators can also check their own performance data by examining the yield of nearby PV plants. Both methods have the disadvantage that faults often go unrecognized for hours or even days. Furthermore, the validity of this comparison is limited by regional differences such as cloud cover or vegetation. A rule of thumb is that a PV plant with a generator output of 100 kWp or more should be fitted with an on-site insolation measuring system. Nevertheless, it is also worth measuring on-site insolation in plants with lower capacities.

Insolation data obtained from satellite pictures may also be consulted in order to determine whether the PV plant is running efficiently. The yields are recorded hourly and sent to a server via the internet once a day. There, the data are compared to the yields expected. This method achieves an average accuracy – although not very quickly – comparable to plant monitoring with PV sensors. If a fault is identified, it often cannot be rectified immediately because the target value and actual value of the yield are only compared once a day.

Another method of monitoring a plant is the continuous comparison of output supplied by the individual module strings (string monitoring). If all the strings have been installed with the same orientation, then their output should always be the same. If it is possible that partial shading could occur, this is known in advance. Therefore, if a string unexpectedly falls behind the others this means that there must be a fault. String monitoring is a quick, simple and effective method of identifying yield losses.

If operational data are saved on the internet, a service provider (or “technical plant manager” in the case of large-scale installations) can assume the task of monitoring the plant and then inform the operators of any faults which occur, or even take independent measures to rectify them.

Causes of faults resulting in yield reduction

Yield losses can generally be attributed to three causes of faults. Component faults, installation faults and faults caused by external influences.

Component faults are more frequently found in inverters than modules. These can be due to production faults, aging or thermal overload of the inverters. Such faults often lead to the complete failure of either the PV plant or the part of the generator connected to the defective inverters. An increasing number of inverter manufacturers are, therefore, now providing long-term guarantees and service contracts. PV modules are not as badly affected by thermal overload as inverters, but rather by external influences, although this happens over relatively long periods of time. As shown by several experiments running for extended periods of time during the 1980s, crystalline solar modules are able to supply power for 20 years without showing significant signs of aging. Provided that the manufacturer has put a sound quality management system in place, production faults are often identified in the factory, meaning that broken cells or incomplete lamination only rarely appear in a PV plant as component faults.

Installation faults rarely result in complete plant failure but only in partial yield reduction. Sometimes, installation faults only start to take effect after a certain time, which means that they are recognized far too late. If, for example, modules are installed so close to one another that there is no longer an expansion gap, the glazing may crack due to the effects of temperature and wind. Individual modules or even whole strings will continue to fail as a result of electrical connections not being installed carefully enough. Insulation can also be adversely affected by installation faults. For this reason, it is wise to use an automatic insulation monitor, which is integrated into some inverters.

External influences primarily affect PV modules. Over the decades, UV radiation from the sun will lead to light aging. The darkening of the plastic film (browning) can lead to a reduction in module output (degradation). Weather-induced aging is only observed relatively rarely in the plastics, in which the solar cells are embedded. Cell damage occurs more often, which is caused by shading and subsequent excessive heating (hot spot). Bypass or string diodes may be damaged by thermal overload or overvoltages. Inverters are not normally directly exposed to meteorological conditions although they are adversely affected by circuit feedback, for example.