Environmental Engineering Reference
In-Depth Information
which is used in this work, provide a cheap and reliable possibility to detect faults
or power losses in photovoltaic plants (Luther and Schumacher, 1991; Eicker et al .,
2005).
Today, simulation tools are mostly used during the design phase to predict the yield
of the PV plant. During plant operation, monitoring of the energy production is often
considered as sufficient for supervision of the plants. Mostly, only the plant AC power
is monitored, as the data is easily available from the inverter. Some attempts have been
made to develop low-cost I - V curve sequencers to measure the module performance
in more detail (van Dyk et al ., 2005), but such systems are not standard practice.
Advancedmonitoring systems use data loggers, which regularly transmit themeasured
values via a modem or Internet connection to a server, where the data is visualized
and significant faults can be detected by analysing the measured performance (Beyer
et al. , 2004).
Simulation has been used in the analysis of monitoring results to determine the loss
mechanisms, but mostly not as an online tool (Reinders et al ., 1999). Only within
the PVSat project with its satellite-based meteorological data is a full simulation
model used, also based on the INSEL environment to calculate the theoretical yield
of the plant once per month (Stettler et al. , 2005). As the error of satellite-based
hourly irradiance measurements is over 20%, deviations of system yield from the
potentially possible yield can only be detected above this error range (Lorenz et al .,
2005).
In this work, the role of simulation tools for the commissioning of PV plants and
their subsequent operating period is analysed. During the commissioning period, it is
important to determine accurately the real installed power and to check the correct
functioning of all components. As themeasurements are taken under varying irradiance
and temperature conditions, models are needed to transfer the measured power to
standard test conditions. Different methods of PV generator parameter extraction for
the simulation models are analysed.
If the PV power plant has a longer lasting follow-up, it is important to implement
automatic routines for fault detection. Here a new methodology is demonstrated for
online remote simulation and fault detection. The required communication technology
for the data exchange is explained and online simulation of a grid-connected PV
generator is carried out.
6.3.1 Photovoltaic System Simulation
An accurate physical model for most PV modules is the two-diode equation for the
current density j , which requires six fitted parameters to describe the two exponential
functions for the Shockley and recombination diodes (factors c s and c r and series
resistance r s ), the photocurrent with factors c 0 and c T and the current loss through the
parallel resistance r p . The diode factors are 1 for the Shockley diode term and 2 for
Search WWH ::




Custom Search