IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT, VOL. 56, NO. 5, OCTOBER 2007 1871 On the Measurement of Power-Quality Indexes for Harmonic Distortion in the Presence of Capacitors Nicola Locci, Carlo Muscas, Member, IEEE, and Sara Sulis, Member, IEEE Abstract—Different techniques have been recently proposed in the literature for the identification of the sources of harmonic distortion in electric-distribution systems. Each of them has its own advantages and limitations as far as reliability, accuracy, and practical usability are concerned. In this paper, these techniques are grouped into three main fundamental criteria in order to inves- tigate their performance in the presence of capacitors. A case study is analyzed by means of both experimental tests and computer simulations. The results put in evidence some major problems that may concern the information provided by the considered methodologies, in virtue of the interaction of the capacitors with the other network components. Index Terms—Capacitors, harmonic distortion, localization of the disturbance, measurement on power systems, power quality. I. I NTRODUCTION I N A LIBERALIZED electric market, where the quality of the product provided by a utility can be a critical competitive item and can represent a clause of the contracts for supplying energy to the customers, measurements aimed at both, evaluat- ing the so-called power quality and attributing the responsibility for its degradation, are assuming more and more importance. Harmonic pollution is one of the major concerns of power quality, and many proposals can be found in the literature to share the responsibility for harmonic distortion in an electric- distribution system [1]–[11]. On the other hand, it should be observed that, while the loads that produce harmonics are undoubtedly the nonlinear loads and, therefore, mainly the power-electronic equipment, other network components can play a very important role in the propagation of the harmonics over the distribution system. Capacitors are the most important example. Capacitors are widely employed in power systems to com- pensate the poor power factor of inductive loads. As well known, they allow the root-mean-square (rms) value of the current in the feeder to be reduced by keeping constant the active power provided to the load, and this results in reducing the stress in the generation and distribution system, as well as the line losses. However, this behavior can be optimal only if the network is operated under steady-state sinusoidal conditions. When these nominal conditions are not held, capacitors may limit their beneficial action and may even cause serious prob- lems in the power plant. This is mainly due to their harmonic impedance, which decreases for increasing frequency, thus Manuscript received June 30, 2005; revised March 15, 2007. The authors are with the Department of Electrical and Electronic Engineer- ing, University of Cagliari, 09123 Cagliari, Italy (e-mail: carlo@diee.unica.it). Digital Object Identifier 10.1109/TIM.2007.903597 causing the currents they absorb to be more distorted then the voltages at their terminals and increasing the risk of resonance with the inductive part of the line impedance. In short, capacitors cannot introduce by themselves harmonic distortion, but they can amplify the effects of the distortion existing in the network. Therefore, the question arises of how the above considerations should be taken into account by a power-quality monitoring system aimed at identifying (and, in case, fining) the disturbing equipment in power systems. In this paper, the main theoretical criteria adopted to perform the task of attributing the responsibility for harmonic distortion are shortly recalled and their behavior in the presence of ca- pacitor banks, installed in one or more nodes of the plant, is analyzed. The investigation is performed by means of a simple but chal- lenging case study. Experimental tests are provided to illustrate the real operating conditions of the system, and simulations are performed to rigorously deal with the numerical details of the problem. The attention is focused on some important metrological issues, namely, the correctness of the definition of the measur- and (i.e., the actual capability of the methods based on these approaches of providing reliable information about the respon- sibility for the disturbances) and the possibility of measuring with sufficient accuracy the desired quantities. II. POWER-QUALITY I NDEXES FOR HARMONIC-POLLUTION METERING Several approaches aimed at identifying the main subjects responsible for harmonic distortion in power systems can be found in the literature. The simplest methods are based on mea- surements performed in a single point of the network (usually the point of common coupling, PCC) [1]–[6], whereas the most recent studies suggest the use of distributed systems designed to perform simultaneous measurements in several points of the network [7]–[9]. Despite their different mathematical formulations, the meth- ods can be conceptually grouped in three main sets, accord- ing to the fundamental principle they are based on. In the following, these three main criteria will be briefly recalled. On the other hand, for the sake of brevity, the details of the different implemented techniques will not be described. For these mathematical formulations, the reader is addressed to the referenced papers. The three considered criteria are as follows: 1) Determination of the loads that increment the harmonic distortion [4], [5], [9]: Several approaches are based on this criterion, being different essentially in the electrical 0018-9456/$25.00 © 2007 IEEE