PMU Estimate Error Due to Low-Pass Filter Transfer Function

A new tool is being designed to perform Monte Carlo analysis on applications which use Phasor Measurement Unit (PMU) data as input. The purpose of the tool is to help determine the effects of PMU estimate error on applications' output. This paper describes the effect of the fixed filter transfer function on the PMU reports and describes a method to simulate the error from ideal reports. PMUs use a discrete Fourier transform (DFT) to estimate the magnitude and phase of the power system signal. The DFT output is a signal sampled at the internal sampling rate of the PMU with frequency components at both the difference and the sum of the power system signal frequency and the system's theoretical nominal frequency. The difference frequency components are the synchrophasor estimates and the sum frequency component must be removed by filtering. However PMU filters are not ideal, and a portion of the sum signal component is unattenuated by the filter. The unattenuated sum component becomes aliased into the synchrophasors transmitted at the PMU reporting rate. This unattenuated, aliased sum component is a significant error in the PMU reports. Many PMUs use a fixed filter with a known transfer function.