Experimental studies on the rtEthernet-based centralized fault management system for smart grids

Abstract

The distribution smart grids in future are likely to have looped structures with a variety of new electronic devices to integrate renewable energy. However, tremendous challenges are brought to the fault management system (FMS), particularly the ultra-fast speed of fault localization with data acquisition at many different locations over a large geographical area. In this work, we propose a centralized fault management system (CFMS) with operation resilience based on real-time Ethernet (rtEthernet) protocols for the protection of smart grids. First, we propose a model to estimate the communication cycle time and the mis-synchronization error based on rtEthernet protocols. The model can be used to design CFMS over a large geographical area with a large number of power system devices. Second, we propose to use the instantaneous values of sample points in the percentage differential protection algorithm, instead of the value of root mean square (RMS), so that ultra-fast speed for data sampling and fault localization can be achieved. Third, we have established a comprehensive test-bed for the CFMS by using our large-scale real-time digital system (RTDS) to simulate the power system. Our extensive experimental studies have demonstrated that the proposed CFMS achieves ultra-fast speed within a sub-cycle and protects the smart grids successfully.