Adaptive protection for microgrids based on communication (PROOF)

The aim oft his research was to contribute with decentralization of energy technologies by investigating technical aspects such as performance of new adaptive protection schemes, anomaly detection in microgrids and cyber-attacks. The research focused on simulation and test in a real-time environment a microgrid setup and run different topologies scenarios in which overcurrent protection will adapt according to the current state of the microgrid like open switches, high Distributed Energy Resources (DER) generation, low DER generation, etc. To do this, the switches and protective devices must communicate to each other.

During the research we explored 3 use cases connected to each other with the ultimate purpose of increasing security levels in microgrid, in the cyber-physical layers of energy systems. In the first, the Council on Large Electric Systems (CIGRE) low voltage benchmark microgrid was modeled on the RTDS and data was collected using a script generated for automatic fault reproduction and data collection into a .cvs file. The fault data was made by changed parameters such as fault resistance, fault point location and microgrid scenario. The main purpose of this use case wastousethecollected data in further research for hard fault detection of fault in microgrids in presence of high fault resistance and low contribution of current from the DERs.

 The second case purpose was to test a adaptive setting topology in the low voltage CIGRE benchmark. We had chosen 3 scenarios for the microgrid and collected the measurements needed to implement optimization in order to choose the adequate settings for a pair of relays that will maintain coordination and minimum operating constraints. By the use of a central management controller a logic implemented based on the state of the switches that determines the topology then it will send communication signals to the pair of relays to set accordingly to the state of the grid and therefore maintain microgrid protected from faults.

 The third case was related to cyber security in microgrids. We divided this testing into two parts, in the first part simple diffential and overcurrent relay protection was tested. The test consisted on injecting GOOSE messages from a third party PC to the relays via Local Area Network (LAN) with the objective to overwrite the messages that the controller is sending. With this we managed to send fake settings to the relays making ineffective the ones being send by the controller. In the second part of this test, we placed a rasberry pi and by controlling remotely the sent again the fake settings to the low voltage CIGRE bechmark. We show how changing the setting under some scenarios, can trigger the protection and therefore open the microgrid switches under normal operation conditions.