Real-time power system oscillation monitoring using synchrophasor measurements

Abstract

Advanced Measurement Infrastructure (AMI) constitutes one aspect necessary to realise a Smart Grid. Advanced monitoring is necessary for reliable operation and control of a modern power system. Enabling measurement instrumentation to perform multiple measurement functions has benefits in creating versatile monitoring systems that is available to all levels of network management operations. Power system instrumentation that can record network coherent data with sufficient certainty in time-stamping to be within the IEEE C37.118.1 requirement for voltage and current synchrophasor has become affordable and accessible in general. Methods for Low Frequency Oscillation (LFO) monitoring in power systems are evaluated from first principles. These monitoring functions are traditionally exclusively available through specialised hardware and software solutions in Wide Area Monitoring Systems (WAMS). Similar features can be made available cost-effectively by setting proper requirements to the signal analysis required to extract the LFO characteristics. Minimal sophistication should be required in the software needed for this purpose. This is demonstrated to be possible. Two LFO monitoring methods, with vastly different mathematical underpinnings, were chosen according to their potential for real-time monitoring of LFOs. The modified Yule- Walker method fitting an ARMA model, and a wavelet method using orthogonal wavelet bases were used. The opportunity for real-time operational application of the LFO results from the two methods are comprehensively studied. The two methods were first tested on simulated power system data. Results confirmed the ability of both methods to characterise LFOs in recorded data with sufficient accuracy and within reasonable processing time. Advantages and disadvantages are substantiated from the results obtained. Field data were emulated and recorded by the synchrophasor recorders used in this research. Results obtained by the LFO analysis applied to this data, indicated that the measurement instruments used does not introduce additional inaccuracies in the two methods. A case study of a Southern African transmission line validated the potential value of both methods in LFO monitoring of real power systems. Practical synchrophasor data were recorded over a 12-month period for this purpose. It is shown that typical WAMS functions do not need to be the exclusive domain of data produced by PMUs at transmission level. Power system instrumentation that record network data coherently can also produce synchrophasor data and minimal sophistication is needed to extract LFO information. A cost-effective solution to obtain additional information on small signal stability performance of power systems is possible and can be deployed at all levels of power system operation.