Performance Evaluation of Large PV SystemsCopyright: © OhWeh, CC-BY-SA license
In the design process of a large PV system on the island of Sardinia, a dynamic simulation of the plant is carried out to support the performance optimization and fault analysis of the plant including the grid interconnection. Several different layouts of the plant architecture are investigated and compared. The design process takes advantage from performance data of existing PV plants. For the assessment of the final realization of the plant, additional measurement devices will be tracking the behavior of the PV installation to optimize the performance of the plant operation.
The simulation includes the PV modules themselves and the environmental factors, such as irradiation, temperature and shading as well as on modeling of the electrical grid with its components. As this simulation scenario involves various physical domains, two simulation tools with their own strengths have to be used. The PV modules and the environment are modeled in the tool PVsyst with a time step of one hour. PowerFactory instead is used to model the grid components both of the plant and the utility grid dynamically in the range of μs. An interface integrates the two simulation tools and exchanges simulation results.
Sensitivity analysis is used to optimize the complexity of the model and decrease the simulation burden. As far as reliable measurements are available, the simulation results both for the individual components and for the entire PV plant are validated against existing data from real PV plants. The modeling of the electrical grid includes the control of the converters as well. Therefore, the behavior of the PV plant for normal and fault conditions can be investigated in the simulation environment.
Alternative architectures of the PV plant are investigated in simulation. In particular, the influence of different DC collector systems is analyzed. As a result, the main factors that affect the efficiency of the PV installation can be determined and steps for an overall energy optimization can be derived.