This Master's thesis presents a solution to maintaining the voltage in distribution networks within the prescribed ranges. The regulation problem emerged because of the ever-increasing number of renewable energy sources implemented into distribution networks.
We created an algorithm capable of regulating both the HV/MV regulatory transformers and the MV/LV regulatory transformers. As our simulation tool, we used the open-coded programme OpenDSS, and connected it to the programme pack Matlab so that we could more easily analyse, process and graphically present the data.
Firstly, we presented the influence of distributed generation on the voltage profile, the distributional transformer and its evolutionary trends. Furthermore, we discussed the way to define the network in the simulation tool, which we used to simulate the power flow during the whole day. In the first phase, we analysed the primary state module, where there is no renewable energy sources present and the regulation network is in the classic mode. In the next phase, we introduced the renewable energy sources in the network and observed how the network reacted (the voltage flow, power, losses, drops in the voltage in the presence of a particular element etc.). We upgraded the classic voltage regulation method by integrating the regulatory distributional MV/LV transformers. We also created an algorithm to automatically regulate the voltage when loaded. In contrast to the classic method where the adjustable off load tap change is used, the regulatory distributional transformers use the automatic on load tap change. To function, the regulatory algorithm requires data about the referential point of measure. We explored both the classic and the central operational modes of the regulation algorithm, where we use different points of measure.
In the last phase, we evaluated the maximal number of integrated renewable energy sources in the LV network considering various regulatory modes. Because the influence of the integration of renewable energy sources in the distribution networks heavily depends on the location of integration and the network load, we implemented the simulations iteratively. We took the average of the total results given by the iteration method based on the 90% of successful results.
|