Analysis of Power System Mode Meters Under Various Oscillatory Conditions

This paper documents the performance of the Least Squares Autoregressive Moving Average plus Sinusoids (LS-ARMA+S) mode meter under non-sinusoidal oscillatory conditions. Originally developed as a way to eliminate the bias experienced by LS-ARMA mode meters in the presence of forced oscillations, the LS-ARMA+S mode meter was shown to be a promising upgrade under both ambient and forced oscillation conditions. In order to further validate it for real-world application, the LS-ARMA+S mode meter has been analyzed under two of the more common system events; ringdown-type transient responses, and hydro turbine rough zone oscillations. It is shown through extensive Monte Carlo simulations that the LS-ARMA mode meter is biased by each of these types of system events, and that the LS-ARMA+S mode meter can detect and compensate for them, thus reducing or even eliminating the bias. These results further suggest that the LS-ARMA+S mode meter extends the LS-ARMA mode meter beyond ambient conditions to include commonly observed oscillatory events.