Application of advanced digital signal processing tools for analysis of voltage events in power systems

International Journal of Electrical Engineering Education, Jul 2009 by P�rez, Enrique, Barros, Julio

Abstract This paper presents a digital signal processing laboratory developed for analysis of voltage events in power systems. The designed laboratory permits the analysis of simulated voltage events of different magnitude, duration and point-on-the-wave of beginning as well as the analysis of real voltage events from records taken by power quality monitoring equipment, using different signal processing tools such as r.m.s magnitude, Fourier analysis, Kalman filtering and wavelet analysis. This laboratory has been used as a benchmark in the development of new signal processing methods for the complete characterization of voltage events in power systems as well as a powerful teaching tool in electrical power quality courses.

Keywords digital signal processing; power quality; power systems; voltage events

(ProQuest: ... denotes formulae omitted.)

Voltage dips, short interruptions and overvoltages are abnormal and sudden changes in the magnitude of supply voltage and at present they represent one of the most important power quality disturbances because of the effects they can produce on equipment connected to the power system. These voltage events are both unpredictible and unavoidable and their frequency of occurrence depends greatly on the type of power system and on the point of observation.

There is no uniform definition of these power quality disturbances. European Standard EN 50 160 uses the following definitions:1 A voltage dip is a sudden reduction of voltage supply to a value between 90% and 1% of the nominal voltage followed by a voltage recovery, with duration between 10 ms and 1 minute. A short interruption in voltage supply is defined as a condition in which the voltage supply is lower than 1% of the nominal voltage with a duration less than 3 min; otherwise the interruption is classified as a long interruption, and finally, a temporary overvoltage is an overvoltage (>110% of the nominal voltage) of relatively long duration.

On the other hand, IEEE Standard 11592 defines a voltage sag (dip) as a shortduration voltage decrease at the power-system frequency to a value between 90% and 10% of nominal voltage. An interruption is defined when the supply voltage decreases to less than 10% of the nominal voltage for a period not exceeding 1 min, and finally a swell is defined as an increase in the r.m.s. voltage for durations from 0.5 cycles to 1 min, with typical magnitudes between 110% and 180%.

Voltage events are characterised by a pair of data, the magnitude and the duration. The magnitude of the voltage event is the lowest magnitude of supply voltage in the case of voltage dips and interruptions (the highest magnitude in the case of a voltage swell) measured during the event, and the duration of a voltage event is the time difference between its beginning and its end.

There are different methods for the estimation of the magnitude of supply voltage and its time evolution during a voltage event. The computation of r.m.s. voltage is the most simple processing tool and is the method proposed in power quality measurement standards3 but, as is shown in Ref. 4, it has limited performance especially in the case of short duration voltage events. Other signal processing tools have been proposed for power quality applications in Refs 5 and 6. Among the most commonly used alternative methods for analysis of voltage events are Fourier analysis, Kalman filtering and wavelet analysis. A comparative study of the performance of these methods can be seen in Refs 7 and 8. The application of new digital signal processing techniques to exactly characterise the magnitude and duration of voltage events is an important area of research for real-time monitoring and control of power systems. The purpose of this paper is to present the development and assessment of a laboratory designed to test different digital signal processing tools in the detection and analysis of voltage events in power systems, using the definitions of European Standard EN50160.

Digital signal processing tools for detection and analysis of voltage events

The digital signal processing tools considered in this paper for detection and analysis of voltage events are the r.m.s. method, Fourier analysis, Kalman filtering and wavelet analysis. The next subsections present a short review of the main characteristics of these processing methods.

R.m.s. method

The r.m.s. magnitude of a voltage supply is computed in a digital system using the following equation:

...

where vi refers to the voltage samples and N is the number of samples taken in a window. The window size can be selected from a half-cycle to any multiples of half-cycles of the power system frequency. In the case of a voltage event it is necessary that the new value of the supply voltage after the event is entirely within the sampling window to obtain the correct r.m.s. value. Thus, depending on the window length and on the time interval for updating the values, the magnitude and the duration of a voltage event can be very different as is reported in Refs 4, 9 and 10.

The r.m.s. method is simple and easy to implement, but shows a limited performance in the detection and in the estimation of the magnitude and duration of voltage events, mainly for short duration voltage events. Another limitation of this method is that it does not provide information about the phase angle or the point-on-wave where the event starts.

The Urms(1/2) magnitude is used in IEC standard 61000-4-30 for detection and evaluation of voltage dips, overvoltages and interruptions. The Urms(1/2) value is defined as the r.m.s. voltage measured over 1 cycle, commencing at a fundamental zero crossing, and refreshed each half-cycle. A voltage dip or an interruption begins when the Urms(1/2) magnitude is below the dip threshold and ends when the Urms(1/2) voltage is equal to or above the dip threshold plus the hysteresis voltage. On the other hand, a voltage swell begins when the Urms(1/2) voltage is above the swell threshold and ends when this magnitude is equal to or above the swell threshold minus the hysteresis voltage.