This report presents the methods and results of the tests conducted at the Power Network Demonstration Centre (PNDC) at the University of Strathclyde (UoS) to evaluate the impact of communication performance on the operation of the Enhanced Frequency Control Capability (EFCC) scheme. This work is part of the EFCC project led by National Grid under Ofgem’s Network Innovation Competition (NIC) funding framework. This report is Part 3 of a set of three reports, where the other two reports focus on the testing of the EFCC scheme’s local and wide-area operational modes respectively. The operation of the EFCC scheme requires two main types of communication networks, i.e. the wide-area communication between the Regional Aggregators (RAs) and Local Controllers (LCs), and regional communication between Phasor Measurement Units (PMUs) and RAs. In this work, both types of communication networks were emulated using a communication emulator and the impact of the communication performance of both networks have been evaluated for the following communication parameters: latency, latency with jitter, the rate of loss of packets and Bit Error Rate (BER). The tests firstly evaluated how these different emulated communication conditions would affect the EFCC controllers in receiving and processing the data (i.e. how the confidence levels in the EFCC controllers were affected under the various emulated conditions). Then the performance of the EFCC scheme during frequency and fault events was tested with the emulated communication conditions. The EFCC controllers handle degraded communication conditions through buffering data over a certain time window (referred as "buffering window") and perform appropriate interpolation to deal with data losses. In the fixed latency tests, it was found that, for 100 ms buffering window, the maximum latency limits at the regional and wide-area networks are around 82 ms and 78 ms respectively. For LCs, even if one of communication links between the RAs and the LCs experience a communication latency larger than the maximum limit, it could lead to the EFCC controllers exhibiting different behaviours. If the LC misses data from two out of three RAs (i.e. the confidence level becomes smaller than the configurable threshold of 50%), this will lead to the LC losing wide area visibility and it will automatically switch to local mode.
|Place of Publication||Glasgow|
|Publisher||University of Strathclyde|
|Commissioning body||National Grid|
|Number of pages||67|
|Publication status||Published - 13 Nov 2018|
- enhanced frequency control capability
- National Grid