Power distribution systems and protection devices have traditionally been designed and built assuming a centralized power supply with high performance generation units. Due to the appearance and penetration of distributed generation (DG) the originally one-way power flow is not true anymore in all cases: depending on the capacity of DG units the direction of power flow may turn. This bidirectional power flow can disrupt the operation of protection devices. In a faulty situation, it can possibly cause the failure of necessary operation.
A further issue is that islanded operation can occur when the distributed generation system with some loads is disconnected from the utility power system. There is no internationally accepted practice yet that describes how these DG units should work in islanded operation. It is forbidden in the most countries and the distributed generators must automatically disconnect themselves from the network under these circumstances. The protection system of small-scale generators must be able to detect the changed operating conditions and must give a tripping command.
The examination of the effects of the grid-connected distributed generation units are essential because of the reasons mentioned above. One possible way of this examination is the modelling of the electric network and the relay systems, and running computer simulations. In this thesis the EMTP-ATP simulation environment was used for this purpose.
In order to give a short review of the problem I present the solutions of different manufacturers to detect islanded operation and I present the program package EMTP-ATP. After that I build a medium voltage compensated network that is suitable for the investigation of the effects of the distributed generators on the protection systems. Then I build the models of the different islanding detection schemes in ATPDraw and I confirm their proper operation using the medium voltage network model mentioned above. Finally, I briefly describe the possibilities of testing real protection devices with the Omicron CMC356 testing device.