Parameters System Management for Voltage Sag s

Transcription

1 Parameters System Management for Voltage Sag s F. D. MOYA (a), J.J. PÉREZ (b), L. H. CORREA (c), J.A. TUMIALÁN (d), R. MORENO (e) CALPOSALLE/AVARC research group - La Salle University Codensa S.A. ESP Carrera 2 N 10 70, Bogotá COLOMBIA (a) frmoya@unisalle.edu.co, (b) jjperez@unisalle.edu.co, (c) lcorrea@unisalle.edu.co, (d) jtumialan@unisalle.edu.co, (e) rmoreno@endesacolombia.com.co Abstract: - This paper presents an analysis about classification of three-phase unbalanced dips (sags). This paper describes some detection techniques and basic characterization of voltage sag s. This work also illustrates a methodology developed by the authors for the study of quality power in Bogota about voltage dips in industrial and commercial users. Power quality compliance limits in Colombia Electrical is showed. Finally this work suggests a power quality parameters system management (monitoring, recording, storing and shipping). Key-Words: - Quality power, Voltage sags, Dips, Bollen methodology, Wavelet transformed. 1. Introduction Currently, the distribution systems consist of a large number of interconnections, attend a large number of industrial, commercial and residential customers and are immersed in markets that offer a particular regulation and state intervention, depending on the country and the region they are. The current scene is such that there are traditional and simple techniques that make use of calculating of wave rms value; oriented techniques to the origin or perturbation source, methods based on Fourier transformed and Wavelet transformed and recently hybrid techniques have been developed to combine some of the traditional methods. This paper initially presents a quick overview of the mentioned techniques and then focuses on the detailed description of the detection techniques by determining the rms value and Wavelet transformed. Besides the conceptual part, this article presents the basis and general characteristics of the applications developed for the detection and basic characterization (parameters of duration and magnitude) of real voltage sags, which occurred in industrial circuits of medium voltage of distribution system of Bogota (Colombia). Power Quality" [1], the power quality refers to a wide variety of electromagnetic phenomena that characterize the voltage and current at a given time and at a given location on the power system. This standard classifies electromagnetic phenomena into several groups: Short duration variations Long duration variations Waveform distortion Voltage imbalance Voltage fluctuations Power-frequency variations Transients Regarding Bollen [2] define power quality, as the conjunction of physical voltage-current signal characteristics for a given time and a specific space, according to each country, with the goal of satisfy specific or implicit requirements of the finals users. For the case of short duration variations, the term "sag" is used in the power quality community as a synonym to the IEC term "dip". The standard defines sag as a decrease to between 0.0 and 0.9 pu in rms voltage or current at the power frecuency for durations of 0.5 cycle to 1min. 2. Power quality phenomena According with the IEE Std "Recommended Practice for Monitoring Electric ISBN:

2 together with a characteristic magnitude and phaseangle jump. The characteristic magnitude and phase-angle jump are defined as the absolute value and the argument of the complex phasor representing the voltage in the lowest phase for a type D dip, and the voltage between the two lowest phases for a type C dip. Figure 1 Short duration variation Short duration voltage variations are almost always caused by fault conditions, the energization of large loads that require high starting currents, or intermittent loose connections in power wiring. Voltage sags are usually associated with system faults but can also be caused by switching of heavy loads or starting of large motors. A common fault in power systems is a single line to-ground (SLFG) is adjudicated to atmospheric discharges, winds, and contact with animals or tree, between others. Others faults are phase phase to ground fault (PPFG) and three-phase faults. In this context, the 70% to 80% of transfer line faults are monophasic faults, between 20% and 15% are biphasic faults and the resting 5% corresponds to three-phase faults [3]. All these agreements define the main different kind of sags that could be found in power quality analysis according to the different kind of fault and occurrence level. 3. Classification of three-phase unbalanced dips A classification of three-phase unbalanced dips was proposed in [2] and [4]. The classification considers threephase, single-phase and phase-to-phase faults, star and delta-connected equipment and all types of transformer connection. It was further assumed that positive- and negative-sequence source impedance are equal. This resulted in four types of three-phase unbalanced dip, shown as a phasor diagram in Figure 2. Type A is due to three-phase faults, types B, C and D are due to singlephase and phase-tophase faults. Type B contains a zero-sequence component, which is rarely transferred down to the equipment terminals. Three-phase equipment is normally connected in delta or in star without neutral connection, and single-phase low-voltage equipment is connected between phase and neutral, but the number of dips originating in the lowvoltage system is small. Therefore the vast majority of three-phase unbalanced dips at the equipment terminals are of type C or type D, so that a distinction between type C and D is sufficient, Figure 2 Types of three-phase unbalanced dip [2] An overview of different sags is given in Figure 3. Despite of, sag occurs in different time period per line those events correspond to the same fault phenomena and are considered single sag. This classification has been defined according to IEC [5]. Figure 3 Short duration magnitude variation 4. Techniques of detection of voltage sags One of the first studies focused to detect the origin of voltage sags is due to [6]. The authors proposal is based on the analysis of the changes experienced by the instantaneous active power during the transient that occurs in a sustained decrease of voltage. The Figure 1 illustrates the registrar location (MP), the flow of preset energy, the upstream point of the registrar (A) and the downstream point of the measurer (B). ISBN:

3 consists of using wavelet transformed (multiresolution) to obtain the approximation and detail coefficients, then calculate the r.m.s. value by the method of coefficients. Figure 4 Energy flow and registrar location One of the most complete studies about nature, characterization and determination of the origin of voltage sags is found in [7]. The author includes, in previous studies, various techniques or methods to detect its origin and explains its philosophies and ways of operation. The discussed techniques refer to the following methods: distance RELE, slope of the system trajectory, sign of resistance, disturbances in the power and energy, cause - event and a real component of the stream. The most recent work related to the detection of the sag origin is due to [8]. In this work, the author presents a formulation to use the transformation, through symmetrical components of the unbalance sags. The author proposes two new methodologies and focuses on simulations on a distribution network of test, using the computer packages of specialized software matlab, simulink, and simpower systems, connecting different types of loads. The paper cites 194 examples, 174 simulations and 20 field tests. The use of the Fourier transformed is well known and widely used in signal analysis [9], [10]. It is based on carrying a signal of the time domain (t) to the frequency domain (jw). The transformed throws then the magnitudes and the angles of the frequency components of the original signal. The detection method of the numerical matrix is focused on voltages with harmonic content and can detect the beginning and an end of sag, its duration and phase change in tension. The original method is proposed by [11] and is quoted by [10]. The method of evaluation of the peak value, proposed by [12], it is quoted by [10]. The method makes a monitoring to the evolution of the peak value. Obviously if there are steady state conditions, the peak value will remain constant. If there are variations in the voltage the peak value will increase or decrease. In the following sections, the interest will focus on the detailed description of the r.m.s. value techniques and wavelet transformed. The first method consists of obtaining the r.m.s. value (effective value of a voltage wave) and the second 4.1. Detection by comparison of rms value The equation (1) also due to [12], shows the discretized version of the calculation of the effective value of the voltage wave. To calculate a first value r.m.s. on a wave cycle, it takes N samples and these N samples are applied the equation (1). The following r.m.s. value is calculated discarding the samples of the first semi-cycle (refreshing of the information) and considering the values of the next semi-cycle. In this case it is said that occurs a leap to the next value of crossing by cero and that the r.m.s. value that is calculated has a leap of semicycle (see subscript of the left term in the equation (1)). Figure 5 Change in the r.m.s. value of the voltage during a sag [12] (1) Different limits can be used to identify the beginning of sag, although it is a common standard to consider 90% of the rated voltage. The described technique is establishes by the standard IEC [5] to implement in the monitoring equipment, recording and analysis of power quality Calculation of r.m.s. Voltage Using the Analysis Wavelet Coefficients Method The underlying algorithm of DWT (Discrete Wavelet Transform) [13], consists in the decomposition and reconstruction of the signal using wavelet function. The decomposition form ISBN:

4 consists of two filters FIR with responses to the impulse low pass and high pass, respectively, followed by decimation by two. Therefore, if at the input there are signal samples, at the output will have approach coefficients ca (from the response of high pass filter), as illustrated in Figure 6. Figure 6 Operating diagram of the algorithm [15] Yoon and Devaney provide in the reference [13] the theoretical basis and demonstrate the practical application of the measure of r.m.s. value and the power/energy of an electrical signal directly from the coefficients of the wavelet decomposition of that signal. The energy of a signal in the wavelet domain is calculated by the application of the Parseval s theorem, which establishes that if the scale functions and wavelet form an orthonormal system, then the energy of the signal f (t) is: Being j the number of decomposition levels The Vrms voltage is defined by (3) (2) If there are j resolution levels, the rms voltage of a voltage signal v (t) analyzed over a time T (N samples) is: (4) 5. Power Quality Compliance Limits in Colombia The national utility regulator in Colombia (Comisión de Regulación de Energía y Gas, or CREG) is clearly working towards power quality compliance limits. The recently enacted resolutions CREG and CREG require continuous monitoring of several power quality characteristics. Once statistics have been collected the regulator will move to implement compliance targets. The CREG resolutions require reporting of 10-minute values for voltage interruptions (number and duration), voltage deviations greater than 10% from nominal and greater than 60 seconds in duration flicker Pst, voltage unbalance, and voltage dips or swells lasting longer than one-half cycle. 6. Case Analysis Bogota Bogota D.C. is the most important city in Colombia, with a rapid growth of its PIB nearly 6% (2009) and with approximately seven million of inhabitants in 2009, making it a large, industrial, commercial and business center. Codensa E.S.P. is the network operator of the city and one of the main problems are voltage dips in industrial and commercial sectors. The work in [15] presents a methodology to takes into account the extraction of information from network analyzers installed in the feeder circuits in 34.5 kv. The analyzers equipment uses as measuring the voltage UDIN, through interpretation of the standard IEC , and an extensive review of the analyzer Nexus equipment operation 1252, it develops the necessary methodology to extract the information stored in the fields of the database. It methodology was developed to characterize and quantify voltage dips from Codensa E.S.P. The network analyzer Nexus 1252 allows a wide variety of making and recording of information which is based on its previous settings before the connection [14]. According to the preset configuration for the measurement by the Nexus 1252, it creates a database that stores the primary information for monitoring voltage dips and other events that occur in the equipment connection point. In concrete form, thirteen events of real sags of the circuits were randomly selected of three substations of the distribution system of the city that occurred during the years 2009 and 2010 [15]. In Matlab software were developed two fundamental applications of detection and characterization (calculation of rms value by windows and by coefficients of wavelet transformed). The results of the two methods were compared by a threshold value per unit (0,9 in this case). The algorithms let identify if any of rms values of the phase voltage (in per unit) low from 0,9 and mark the number of occurrence sample. Taking as an exact value the ISBN:

5 calculation by the method of rms by the window, we observe a half relative error in duration and magnitude of 1.6% and 1% successively between the two methods which are satisfactory values for the analysis of sags. From the processing of this information and knowing the duration and the magnitude of each one of the events, it can analyze the total table of voltage sags and generate the diagram in 3D number of sags based on duration and voltage rms and it can observe graphically that the most of the occurred event has short durations and a higher percentage in larger voltage drops to 80 90% of voltage, ), as illustrated in Figure 7. Número de huecos de tensión < < < < < < < < <0.9 >=0.9 Tiempo en segundos Total de huecos de tension=13 >80-90 >70-80 >60-70 >50-60 >40-50 > >10-20 >20-30 % Tensión rms Figure 7 Diagram representing the amount of voltage sags versus time and rms voltage [16] 7. Electrical Power Parameters System Management (monitoring, recording, storing and shipping) For verification of power quality regulations compliance (as issued by the CREG), could have a hardware-software integration aimed for two main functions: First, sending reports and event information off limits set by regulation. On the other hand, monitoring and electrical parameters recording information to electrical operator for that it can characterize events identify source and users that may have an impact on its current events. Hardware infrastructure and facilities could include communication servers robust enough for processing and receiving large amounts of information. Distribution substations could have capture equipment and computers for storage, as well as from communications equipment to send information to central control. Software requirements could include man-machine interfaces, management software, databases and programs to set alarms, produce reports, fire information delivery. Some of the functional requirements that could include the application are: The fire alarm configuration and capture information in any medium voltage feeder circuit from the monitoring center. This information would be focused on times and durations of sags, voltage distortion percentages out of limits, flicker, etc. The software must be able to run real-time applications (for comparison of parameters with limits, implementation of statistical routines, and sending information or summary) and runs off-line (to characterize events, identify origin, cause or summary reports, etc.). The report sags events could be sent in summary form (magnitude and duration) to the monitoring center, accepting the existing standards (IEC ) in the sense of identifying a sag from the time any phase falls below 0.9 in pu and until all phases are above 0.9 pu The harmonic event reporting which breach the existing standards (percentages of distortion out of bounds) could be performed after regular running routine statistical procedures for purposes of observing changes in percentages. The event report voltage fluctuations (flicker) may be periodic executions after having the purpose of statistical calculations set forth in the regulation of CREG. 8. Conclusion The results of obtained duration and magnitudes with the techniques described in this paper are very similar. The most significant differences are seen in the shorter duration sags, that is to say, those lasting between two and four cycles. For verification of power quality regulations compliance, electrical operators could have a hardware-software integration to send reports and event information set by regulation. Electrical operators must monitor and recording electrical parameters information to characterize events to identify source and users that may have a big impact. ISBN:

6 Acknowledgements The authors acknowledge the received support through the financing contract RC No 706 concluded between COLCIENCIAS, CODENSA S.A. E.S.P., EMGESA S.A. E.S.P. and La Salle University. The authors wish to thank the engineers Gustavo Arciniegas from La Salle University, Edgar Muela, Bernardo Uribe and Diego Cely. References: [1] IEEE Std , Recommended Practice for Monitoring Electric Power Quality, Institute of Electrical and Electronics Engineers. 70 Pág. New York [2] Bollen, M., Characterization of three- phase unbalance dips ( as easy as one-two-three?), Power Engineering Society Summer Meeting, IEEE Volume 2, July Page (s): vol [3] Grainger, J. Stevenson, W. Power Systems Analysis. McGraw-Hill. México [4] Bollen, M. Understanding power quality problems: voltage sags and interruptions. IEEE Press New York. [5] IEC Std , Electromagnetic compatibility (EMC). 89 Pág. CEI/IEC :2003 [6] E. Belenguer, C. Reineri, N. Aparicio, R. Mampel, Voltage sag source location in distribution networks, International Conference on Renewable Energies and Power Quality (ICREPQ'03), 2003 [7] R. Makaliki, Voltage sag source location in power systems, MsC Thesis, Chalmers Tekniska Hogskola, 2006 [8] B. Polajžer, G. Štumberger, D. Dolinar, Evaluation of different methods for voltage sag source detection based on positive-sequence components, International Conference on Renewable Energies and Power Quality (ICREPQ'09), [9] Alzate, A. López, H. Ramírez, Análisis comparativo de algunas teorías en el dominio de la frecuencia para la detección de distorsiones en sistemas eléctricos de potencia, Scientia et Tehnica, Vol. 14 No. 39, pp , 2008 [10] M. Alonso, J. Sanz, J. Sallán, J. Villa, Comparison of different voltage dip detection techniques, International Conference on Renewable Energies and Power Quality (ICREPQ'09), 2009 [11] C. Fitzer, M. Barnes, P. Green, Voltage sag detection techniques for a dynamic voltage restores, IEEE Transactions on industry applications, Vol. 40 No. 1, 2004 [12] M. Bollen, D. Sabin, R. Thallam, Voltagesag indices - recent developments in IEEE P1564 Task Force, in CIGRE-PES 2003, 2003 [13] W.K. Yoon and M.J. Devaney, Power Measurement Using the Wavelet Transform, IEEE Transactions on Instrumentation and Measurement, Vol. 47, No. 5, October 1998, pp [14] Electro Industries, Manual de Referencia Nexus 1252 Communicator EXT 3.0 Chapter 3. Configuring the Nexus 1250/1252 Meter [15] J. Pérez, L. Correa, F. Moya, J. Tumialan, R. Moreno, Development and implementation of a methodology for the study of voltage dips in Bogotá D.C., 21st International Conference on Electricity Distribution (CIRED- 2011), Frankfurt, 6-9 June, [16] L. H. Correa, J.A. Tumialán, J.J. Pérez, R. Moreno, Comparison of RMS Value Techniques and Wavelet Transformed for Detection and Basic Characterization of Voltage Sag s, VI Simposio Internacional Sobre Calidad De La Energía Eléctrica (Sicel2011), Asunción-Paraguay, Nov Biographies F. D. Moya is Professor from La Salle University, Bogotá, Colombia, Electrical Engineering Program.. J. J. Pérez is with La Salle University, Bogotá, Colombia, Electrical Engineering Program. L. H. Correa is with La Salle University, Bogotá, Colombia, Electrical Engineering Program. J. A. Tumialán is with La Salle University, Bogotá, Colombia, Engineering Program in Automation. R. Moreno is with CODENSA S.A. ESP, Bogotá, Colombia. ISBN: