2652 IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT, VOL. 59, NO. 10, OCTOBER 2010 Accuracy Analysis of Algorithms Adopted in Voltage Dip Measurements Daniele Gallo, Member, IEEE, Carmine Landi, Member, IEEE, and Mario Luiso, Member, IEEE Abstract—This paper analyzes the accuracy of algorithms com- monly adopted in instruments devoted to the detection and charac- terization of voltage dips (which are also called sags). This analysis is particularly interesting, because the results of dip measurements are utilized for calculation of severity levels and site indexes, which are important parameters not only in the assessment of the quality level of power supply but also in the selection of equipment with proper intrinsic immunity. Instruments for dip measurement still have unresolved technical and theoretical issues related to the characterization of their metrological performances, so it can be found that different instruments are significantly in disagreement in some actual measurements. This paper moves a step into the direction of deepening the knowledge about the measurement of voltage dips, pointing out the limits incident to the adoption of the detection algorithms indicated in the standards. It starts with a discussion about parameters that characterize voltage dips, in agreement with the standard. Then, analytical calculations of some systematic deviations in the event characterization, which are introduced by the most diffused dip detection algorithms, in simplified measurement situations, are presented, underlining their remarkable impact. The obtained relations are experimen- tally verified on a commercial power quality instrument, forecast- ing its systematic deviations. Index Terms—Measurement accuracy, power quality instru- ment, sag, voltage dip. I. I NTRODUCTION P HENOMENA that involve root mean square (rms) vari- ation in the supply voltage such as long and short inter- ruption, overvoltage, and voltage dip (in U.K. English) or sag (in American English—the two terms are equivalent) are cur- rently the power quality issues with greater economical impact. In fact, industrial customers, in particular, highly suffer from regular production stoppages due to these phenomena [1]–[3]. Remarkable voltage reductions are caused by a short circuit or an earth fault that is close to a substation that will force the voltage to a very low value in one or more phases. Smaller reductions are caused by timely varying loads. Usually, the reduction ends within a short time due to automatic switching actions, fault reparation, or load stabilization. These phenomena can be classified as voltage dips or in- terruption, depending on the event duration and the minimum voltage magnitude reached during the event. For the purpose of this paper, the authors will refer only to dip events as Manuscript received July 8, 2009; revised November 9, 2009; accepted November 9, 2009. Date of publication April 8, 2010; date of current version September 15, 2010. The Associate Editor coordinating the review process for this paper was Prof. Alessandro Ferrero. The authors are with the Department of Information Engineering, Second University of Naples, 81031 Aversa, Italy (e-mail: daniele.gallo@unina2.it; carmine.landi@unina2.it; mario.luiso@unina2.it). Digital Object Identifier 10.1109/TIM.2010.2045256 their durations are typically less than 0.1 s, thus presenting greater measurement problems. Nevertheless, the conclusions and procedures described in the following can be extended to interruptions or swells with minor changes. Many IEEE groups and task forces are working to develop a recommended practice for converting a suitably sampled volt- age and current data set into specific power quality categories and to describe with specific attributes each category. In partic- ular, the IEEE 1159.2 Working Group focuses on events such as dips and other nonharmonic events that are still without full technical definitions. The translation from sets of digital data to statistically comparable events would be used for purposes of comparing power suppliers, comparing susceptibility qualities of equipment, and evaluating performance versus specification or contract. Therefore, a recognized set of digital definitions will benefit all the stakeholders of the electrical energy market [4]–[7]. However, instruments for dip measurement still have unre- solved technical and theoretical issues related to performance characterization. Because of this, different implementations that fully meet the definitions reported in the standards [6], [7] can still significantly disagree in some actual measurements. It mainly happens because the standards do not include a sound and well-defined procedure for the verification of their performance. This paper moves a step into the direction of deepening the knowledge about the measurement of voltage dips, pointing out the limits that are inherent with the adoption of the detection al- gorithms indicated in the standards. The obtained results could be useful in defining a comprehensive test protocol that is able to assess the performance of measurement instruments. This paper starts with a discussion about measurement parameters that characterize voltage dips in agreement with standards. Then, the dip detection algorithms compliant with the standard are analyzed, presenting the systematic deviations introduced in event characterization. In particular, systematic deviation due to fundamental frequency synchronization and delays in the detection of the beginning and ending of a voltage dip are analytically calculated, in simplified conditions, underlining their remarkable impact. Finally, the obtained relations are used during verification of the measurement accuracy of a commercial power quality instrument to forecast the systematic deviations: some experimental results are shown. II. STANDARD DEFINITIONS FOR VOLTAGE DIPS EN 61000-4-30 [7] provides the first international defin- itions and measurement methods for the characterization of voltage dips. This standard is the fundamental reference for 0018-9456/$26.00 © 2010 IEEE