1864 IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 24, NO. 4, OCTOBER 2009 Investigating the Option of Removing the Antialiasing Filter From Digital Relays Sukumar M. Brahma, Senior Member, IEEE, Phillip L. De Leon, Senior Member, IEEE, and Rajesh G. Kavasseri, Senior Member, IEEE Abstract—Digital relays traditionally employ sampling rates of less than 100 samples/cycle. In order to avoid aliasing due to fault transients, these relays employ an analog antialiasing filter before critical-sampling (Nyquist rate) the input waveforms coming from instrument transformers. In many applications of electrical en- gineering, oversampling (greater than the Nyquist rate) has long been used to simplify the requirements of an antialiasing filter with a sharp cutoff; in some cases, the filter can even be eliminated. This paper investigates this option for a digital relay. The performance of a traditional digital relay is compared with a method that uses oversampling without using an antialiasing filter. By processing a comprehensive array of fault waveforms from Electromagnetic Transients Program simulations, a suitable oversampling rate is suggested. A comparison of phasor estimates using the traditional relay and the proposed method is made for different operating and fault conditions. The results suggest that oversampling can elimi- nate the antialiasing filter traditionally employed in digital relays. Index Terms—Aliasing, analog-to-digital converter (ADC), dig- ital relay, discrete Fourier transform (DFT), power system protec- tion. I. INTRODUCTION D IGITAL relays use sampling rates ranging from 8 sam- ples/cycle to as high as 96 samples/cycle [1]. During the inception of a fault, the voltage and current waveforms are su- perimposed by transients. The amount and duration of transients depend on factors, such as the instant of fault with respect to the voltage waveform, the type of fault, the location of fault on the line, and the damping available in the system. Faults occurring at instants when the voltage waveform is around its peak value are the most severe in terms of transients. Typically, voltage waveforms experience more severe transients than cur- rent waveforms. Digital relays use the discrete Fourier trans- form (DFT) of the sampled signal to estimate the phasor value of the fundamental. To avoid aliasing, especially during a fault, all digital relays employ an analog (low-pass) antialiasing filter before sampling the voltages and currents with an analog-to-dig- ital converter (ADC) [1], [2]. This type of filter introduces a time-delay of 1.5–2 ms in the phasor estimation depending on the sampling rate chosen [2]. This filter can also be relatively expensive. Manuscript received October 16, 2008; revised February 10, 2009. Current version published September 23, 2009. Paper no. TPWRD-00765-2008. S. Brahma and P. De Leon are with the Klipsch School of Electrical and Computer Engineering, New Mexico State University, Las Cruces, NM 88003 USA (e-mail: sbrahma@nmsu.edu; pdeleon@nmsu.edu). R. Kavasseri is with the Department of Electrical and Computer Engi- neering, North Dakota State University, Fargo, ND 58102 USA (e-mail: rajesh.kavasseri@ndsu.edu). Digital Object Identifier 10.1109/TPWRD.2009.2028802 In many applications, oversampling (i.e., ), where and are the sampling and Nyquist frequencies, respec- tively, has long been used to simplify the requirement of an an- tialiasing filter with a sharp cutoff at . If the oversampling rate is selected so that any aliased frequencies are extremely small or below the noise floor, then the antialiasing filter can be made less sharp or, in some cases, even be eliminated, reducing cost and delay [3]. Many commonly available ADCs utilize over- sampling for these reasons. Very inexpensive ADC chips are currently available that use oversampling up to a few hundred kilohertz. The digital music industry today is able to produce excellent sound reproduction by using very simple or no analog prefiltering in their prod- ucts. Since adopting inexpensive oversampling can eliminate a comparatively more expensive analog filter and the associated time delay, it is worthwhile to investigate the possibility of re- moving antialiasing filters from digital relays through oversam- pling. This paper investigates this possibility. A Power System Relaying Committee (PSRC) report on software models used in relays indicates that oversampling is used in the newest generation of relays, but the main purpose of oversampling is oscillography [4]. These relays still use an analog antialiasing filter, and the sampling rate used for phasor estimation is obtained by decimating the data sampled at a higher frequency. In our extensive literature search, including a patent search, there is no published document that investigates the phasor-estimation function of a digital relay without using analog prefiltering. This paper describes the process where a rationally chosen oversampling rate is tried out on a comprehensive array of fault waveforms generated by using the Electromagnetic Transients Program (EMTP). Through these trials, we show that phasor es- timates by using the chosen oversampling rate and without using an antialiasing filter are practically the same as the phasor esti- mates from a conventional digital relay that uses an antialiasing filter. Factors, such as fault type, fault location, fault instant, fault resistance, and prefault conditions are varied while gener- ating the fault waveforms. Based on the similarity of the phasor estimates for these various conditions, conclusions are drawn to emphasize the effectiveness and feasibility of the proposed ap- proach. II. SIMULATION APPROACH In order to obtain a preliminary estimate of the required sam- pling rate to avoid aliasing, the spectral content of the sampled waveform should be measured. In order to perform such spectral analysis, a fault-voltage waveform was obtained by using EMTP simulation. A 240-kV, 225-km, two-terminal transmission line with substantially different source impedances at both ends was 0885-8977/$26.00 © 2009 IEEE Authorized licensed use limited to: New Mexico State University. Downloaded on December 10, 2009 at 14:03 from IEEE Xplore. Restrictions apply.