Understanding Power System Frequency
Analytics
274 views ◎247 downloads ⇓
Abstract
Frequency of an electrical signal is defined during steady state conditions, although it can be problematic to measure. However, power system frequency can often fluctuate during power swings and other anomalous events. These events can be characterized as non-periodic signals where frequency is not generally defined [1]. In industry, phasor measurement units (PMU) are used by utility companies to measure voltage and current in the power system. These digital samples obtained from the PMU are included within an algorithm where phasors and frequency of the power system are computed. To report these values with desired accuracy, the frequency of the system must be well known. The PMU’s are often subjected to varying signal types and depending upon the implemented algorithm the computed fundamental frequency component can vary [2, 3]. These signal types are, but not limited to: phase modulated, amplitude modulated, phase angle step, magnitude step, frequency ramp, harmonics, inter-harmonics, etc. These signal types may lead to discrepancies between fundamental frequency values reported by different PMUs processing the same signals measured from the power system. This demonstrates that an advanced understanding of frequency needs to be addressed. The study was conducted to determine the inconsistency between PMUs sold by different manufacturers. To determine if there is an underlying inconsistency between multiple PMU manufacturers, tests were performed to provide insight into the implemented guidelines stated within IEEE std. C37.118.1a-2014. Four compliant PMU devices were subjected to several signal types listed within the IEEE standard. To ensure the implemented frequency method is robust, dynamic events must be considered so that they also measure the fundamental frequency accurately when subjected to varying signal parameters. The maximum reporting difference, via IEEE std. test events, was 72.6 mHz. However, a frequency excursion event captured on the power system was replayed to each of the devices and the reported fundamental frequency values differed up to 748.3 mHz in some instances. This is an order of magnitude greater for the real system event versus the IEEE std. test conditions. In addition to testing the four individual PMU devices and alternative frequency estimation method is proposed.The new proposed frequency method was designed to be fast and accurate for varying signal conditions. The new method allows for frequency reporting at the same rate as the sampled signal. The method was tested for accuracy and compliance using the same test signals defined by the IEEE std. C37.118.1a-2014. For the tested signal conditions, the frequency method resulted in error as little as 0.8 mHz for steady state signals and yielded a maximum error of 49.1 mHz for a phase modulated signals. The test results showed that the new proposed frequency method was compliant with the IEEE std. C37.118.1a-2014 for all tested signal types.