Uncertainty of Measurement Error in Intelligent Electronic Devices

Transcription

1 Uncertanty of urement Error n Intellgent Electronc Devces Po-Chen Chen 1, Student Member, IEEE, Yma Dong 2, Member, IEEE, Vuk Malbasa 1, Member, IEEE, and Mladen Kezunovc 1, Fellow, IEEE 1 Department of Electrcal and Computer Engneerng Texas A&M Unversty College Staton, TX , U.S.A. pchen01@neo.tamu.edu, vmalbasa@tamu.edu, and kezunov@ece.tamu.edu Abstract Ths paper focuses on methodology to quantfy uncertanty n measurements obtaned from Intellgent Electronc Devces (IED). IEDs have emerged n dstrbuton systems as a prevalent source of measurements n montorng and protecton, as well as for dfferent knds of applcatons beyond IED s prmary purposes. These measurement devces are nstalled across a system, from substatons down to the customer locatons, and provde measurements of a wde array of quanttes. We report how IED measurements respond to external dsturbances, whch may lead to possble accuracy mpacts n varous applcatons. The example used to llustrate the approach s hghly accurate fault locaton n dstrbuton systems based on voltage sag measurements. Index Terms Fault locaton, Intellgent Electronc Devce, measurement uncertanty, measurement unts, smart grds. I. ITRODUCTIO The U.S. Department of Energy s ntatve GRID 2030 [1] has demonstrated a vson for the future modernzaton of power systems. Kezunovc et al. [2] have descrbed the nnovaton process n the near future, whch wll requre varatons n the archtecture and components of the current power grd, where Intellgent Electronc Devces (IEDs) are one of the essental components for montorng and protecton. To date a wde varety of IEDs have become avalable and felded. Some IEDs are performng waveform samplng n support of ther prmary applcatons (dgtal relays, dgtal fault recorder) [3]-[5], whle others provde synchronzed measurements n support of other applcatons (synchrophasor IEDs) [6]-[10]. Furthermore, some devces provde energy measurements and power qualty ndcators (smart meters and power qualty meters) [11]-[14]. Fault locaton usng IEDs plays a crtcal role n dstrbuton systems for outage management and servce restoraton. For voltage sag based fault locaton [15]-[20] usng data from varety of IEDs may n some cases have an advantage of detectng accurate fault 2 Electrocon Internatonal Inc. Ann Arbor, MI USA dongyma@gmal.com locaton by matchng the voltage measurements, whle mpedance-based methods may provde multple ambguous estmatons [21]-[24]. As an example, voltage sag based fault locaton methods requre accurate voltage measurements from the correspondng IEDs. Therefore, how the performance of IEDs vares wth external dsturbances becomes crtcal for accurate estmatons [25], [26]. In ths paper, we have characterzed the uncertanty n measurements of a commercal IED product n order to valdate ts use for voltage sag based fault locaton. The devce was tested n the Relay Testng Laboratory of Smart Grd Center, Texas Engneerng Experment Staton (TEES). The tests were amed at determnng the parameters of measurement under gven fault scenaro smulatons. II. PREPARATIO FOR TESTIG A. Requrements of IEDs used for Fault Locaton urements As an mportant factor contrbutng to the accuracy of fault locaton, the characterstcs of measurement was determned by testng the IED product. Two characterstcs of IED data were explored durng the testng: qualty (accuracy), and avalablty (how frequently the data s collected, the length of recordng, etc.). B. Test Sgnal Model The test sgnal model descrbes the test waveforms. The steady-state sgnal model used n tests s x( t) = k= 0 A cos(2π kf t + ϕ ) + A r( ), (1) k 1 k r t where k s the order of harmonc, A k s the ampltude of k th harmonc, ϕ k s the ntal angle, f 1 s the fundamental frequency, A r s the ampltude of random nose, and r(t) s the Ths materal s based upon work supported by the Department of Energy under Award umber(s) DE-OE Fundng for ths effort comes from DOE through project ttled A real-tme montorng, control, and health management system to mprove grd relablty and effcency awarded to ABB, Xcel Energy, and TEES.)

2 functon generatng random numbers from a normal dstrbuton wth μ=0 and σ =1. C. Test Scenaros The test scenaros shown n Table I nclude those for ampltude change, phase angle change, frequency change, and harmonc nterference. The objectve s to valdate n phasor ampltude and angle. Each scenaro descrbed n Table I has been replcated 10 tmes. The reference condtons of Table I are V, k = 1 I, k = 1 Ak _ voltage =, Ak _ current =, 0, k 1 0, k 1 [ ϕ1a, ϕ1b, ϕ1c ] = [ 0, 2π /3,2π /3], A r = 0, V = 100 V, and I = 1 A under 100% rated secondary value, constant phase, and nomnal frequency. D. Valdatng Quanttes In the steady state test scenaros the mean and varance of measurement was calculated and we demonstrate the process usng voltage ampltude. The n test scenaro s defned as Vmeas. Vref. =, (2) V where V meas. s the measurement value from IED to secondary voltages, V ref. s the reference voltage, and V s the secondary-based voltage. Then the mean Mean e _ Vamp and the varance Var _ of can be computed as TABLE I. Varyng Quantty ame Lne Voltage Ampltude Phase Current Ampltude Phase-to-Phase Voltage Angle Dfference Lne Voltage Ampltude Phase-to-Phase Current Angle dfference e Vamp = 1 e Vamp = Mean _, (3) SEARIOS FOR STEADY-STATE TEST Varyng Quantty Varable A k _ voltage and Varyng Range Step- Length % 10% A k _ current 0 500% 50% ϕ1a ϕ1b, 1B ϕ1c π / 6 2π / 3 π / 6 ϕ1c ϕ1a 3 A k _ voltage 0 120% 10% 1A ϕ1b 1B ϕ1c ϕ1c ϕ1a and π / 6 2π / 3 π / 6 Frequency f Hz 0.5 Hz Harmoncs Ampltude A k, k = 3, 5, 9 5% and 10% 5% Var = 1 e _ Vamp = ( Mean 2 e _ ), (4) where s the number of measurement samples. Errors n current ampltude measurements were calculated usng the same formulatons (2)-(4) by changng the voltage quanttes to current quanttes. Ths also apples to voltage and current angle measurements. III. IMPLEMETATIO OF LAB TEST A. Laboratory Setup The hardware setup procedure s shown n Fg. 1. Test waveforms were generated and sent to analog sgnal amplfer va Input/Output (I/O) box and Dgtal/Analog (D/A) converter by Relay Assstant [27]; the Analog Sgnal Amplfer generates analog waveforms based on ts nput and sends them to the IED; ampltude measurements were read drectly from IED s front panel screen; angle measurements were recorded by manually trggerng the dsturbance recordng functon, and retreved from the dsturbance reports. The Techron 7780 Standard 3-phase Voltage/Current Mult-Amplfer System [28] from AE Techron was used as a sgnal amplfer set. After calbraton, s n output ampltudes ere wthn 0.1% rms, and s n phase angle dfferences between outputs were wthn 0.5 degrees. Whle performng laboratory experments the ampltudes and reference quanttes were obtaned usng the Fluke 8010A multmeter. When testng the angle, waveforms were captured by a Tektronx TDS 320 osclloscope and the angle dfference between two phases was calculated from tme delays between zero-crossng ponts on two waveforms by Angleph1_, (4) 360 ) f = ( Tzero cros sn g, Tzero crossn g, ph1 where Angle ph1 _ s the angle dfference n degrees between phase 1 and phase 2 (1 and 2 beng A, B or C), Tzero cros sn g, Tzero cros sn g, ph1 s the tme delay between Fgure 1: Hardware setup. 0

3 zero-crossng ponts on two waveforms, and f 0 s the fundamental frequency 60 Hz. To analyze current waveforms the analog current sgnals were scaled down and converted to a voltage sgnal usng a Jamb CT201T accurate current transformer, before connectng to the osclloscope. To analyze voltage waveforms, 10:1 probes were appled to scale the voltage sgnals down before the osclloscope sampled the sgnals. B. Acquston of Phasor urements from IED Steady state sgnals lastng three seconds were sent to the IED and the reference measurement devces. Values were acqured n the mddle of the sgnal generatng process, when readngs from IED and reference measurement devces were stable. Values were read from the IED front panel screen. Angle measurements were retreved from dsturbance reports (example shown n Fg. 2). Recordng of dsturbances was trggered manually. In the meantme, the osclloscope was trggered at the begnnng of the test to produce reference measurement of angle dfference between two phases. C. Pre-processng of Data All of the recorded data was converted to secondary quanttes and compared to the readngs from the reference measurement devces. For nstance, when analyzng results from voltage ampltude tests, values from IED were frst converted to secondary voltages usng V meas _ sec V V =, (5) meas _ prm _ sec V _ prm where the quanttes wth subscrpt of sec are secondary values, the quanttes wth subscrpt prm are prmary values, and V means rated voltage on the prmary or secondary sde of the measurement transformer. ote that converson of current measurements s smlar to converson of voltage measurements. Snce lne voltages (phase-to-phase voltage) were recorded, when calculatng usng (2), 3 V _ ph was used as the rated lne voltage. When processng the results from angle tests, the angle dfferences between two phases were calculated from phase angles usng the dsturbance reports. IV. IED SETUP AD TEST PROCEDURE The IED supports two measurement modes: RMS and DFT. The DFT algorthm flters out harmoncs whle the RMS method does not suppress harmoncs. IED characterzaton ncludes harmoncs and frequency tests whch are affected by the measurement mode. The DFT mode was chosen n ths study. Snce the voltage sag based fault locaton algorthm requres a change n voltage ampltudes we used the DFT n order to flter out the harmoncs. A general procedure for performng the laboratory tests can be summarzed as follows. 1. Calbrate amplfers; 2. Generate test sgnals accordng to Table 1-3 usng Relay Assstant; 3. Calbrate sgnals n Relay Assstant and send the calbrated sgnals to sgnal amplfer; 4. Record measurement values from IED and from reference measurement devces; 5. Move to next scenaro and repeat the process. V. RESULTS AD AALYSIS Results from IED that characterzed the mean quanttes of the sgnal were shown n Fg. 3. Table II provdes the data of V ab n Fg. 3(a) as an example of test results (due to the space lmts we cannot show all of them). Below s a summary of the observatons. 1. Both voltage and current ampltude measurements produced very stable outputs durng repeated tests, wth very low varance. Offsets n both voltage and current ampltude measurements show a negatve correlaton wth the ampltude of test sgnals. The s show a lnear pattern, suggestng that wth proper settngs of the ampltude correcton factors, most of the can be elmnated. 2. o obvous mpacts of frequency and harmoncs have been observed durng tests of ampltude measurements. 3. Errors n angle measurements were not as stable as those n ampltude measurements, but the mean n phase angle dfferences never exceeded ±5 ; the largest absolute detected was o clear pattern n the angle dfference s, frequency or harmoncs n the test waveforms was observed. The s n angle dfferences may be related to the calculated phase angles, or the ncepton angles of the phasor calculaton TABLE II. DATA OF VAB I FIG. 3(A): ERRORS I LIE VOLTAGE AMPLITUDE MEASUREMETS TO THE AMPLITUDE OF TEST SIGAL Fgure 2: Phasor nformaton from dsturbance report. Ampltude (per unt) Mean [%] Maxmum [%] Varance 0 1.9E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E-07

4 process when dsturbance reports were generated. Incdentally, the tests also reveal the constrant on usng the measurements from the gven IED n scenaros where the fault s cleared very quckly [29], [30]. The reportng mechansm of the devce s desgned such that a new measurement s reported when the accumulaton of change over tme exceeds a threshold. Therefore, the shortest fault duraton for a durng-fault measurement to be reported s 0.4 seconds under the most favorable settngs and the most severe fault condton (voltage at measurement pont drops to zero). Faults lastng less than 0.4 seconds would be nterpreted as transents by the IED and the durng-fault phasor measurements would be reported. In cases where faults are cleared by protectve devces n less than a cycle (sub-cycle fault), employng fast and accurate phasor extracton technques s necessary [20]. VI. COCLUSIOS Ths paper makes several contrbutons: A commercal IED product has been tested extensvely n the lab to determne the parameters of measurement s for fault scenaro smulatons. The mpact of dsturbances on the ampltude and angle of voltage and current measurements of IED was presented. The results demonstrate the observatons of two dfferent dmensons (as mentoned n Secton II-A). The qualty was suffcently precse. The avalablty depends on how bg the changes n ampltudes are and how long the durngfault perod s. ACKOWLEDGMETS The authors gratefully acknowledge Dr. J. Stoups, Dr. M. Mousav, Dr.. Kang, and Dr. X. Feng from the ABB Group for ther feedback and valuable advce. REFERECES: [1] GRID 2030, A natonal vson for electrcty s second 100 years. [2] M. Kezunovc, V. Vttal, S. Melopoulos, and T. Mount, The Bg Pcture: Smart Research for Large-Scale Integrated Smart Grd Solutons, IEEE Power and Energy Magazne, vol. 10, no. 4, pp , Jul [3] S. M. Brahma, P. L. De Leon, and R. G. Kavasser, "Investgatng the Opton of Removng the Antalasng Flter From Dgtal Relays," IEEE Trans. Power Del., vol. 24, no. 4, pp , Oct [4] P. M. Anderson, Power System Protecton, Wley-IEEE Press, [5] Dstrbuted Dgtal Fault Recorder, Dgtal Energy, General Electrc. [6] M. Kezunovc and B. Peruncc, Automated Transmsson Lne Fault Analyss Usng Synchronzed Samplng at Two Ends, IEEE Trans. Power Syst., Vol. 11, o. 1, Feb [7] W. Premerlan, B. Kasztenny, and M. Adamak, "Development and Implementaton of a Synchrophasor Estmator Capable of urements Under Dynamc Condtons," IEEE Trans. Power Del., vol. 23, no. 1, pp , Jan [8] J. De La Ree, V. Centeno, J. S. Thorp, and A. G. Phadke, "Synchronzed Phasor urement Applcatons n Power Systems," IEEE Trans. Smart Grd, vol. 1, no. 1, pp , Jun [9] K. E. Martn, D. Hama, M. G. Adamak, S. Anderson; M. Begovc, G. Benmouyal, G. Brunello, J. Burger, J. Y. Ca, B. Dckerson, V. Gharpure, B. Kennedy, D. Karlsson, A. G. Phadke, J. Salj, V. Skendzc, J. Sperr, Y. Song, C. Huntley, B. Kasztenny, and E. Prce, "Explorng the IEEE Standard C Synchrophasors for Power Systems," IEEE Trans. Power Del., vol. 23, no. 4, pp , Oct [10] C. Zheng, V. Malbasa, and M. Kezunovc, Regresson tree for stablty margn predcton usng synchrophasor measurements, IEEE Trans. Power Syst., vol. 28, no. 2, pp , May [11] H. S. Cho, T. Yamazak, and Mnsoo H, Determnng locaton of applances from mult-hop tree structures of power strp type smart meters, IEEE Trans. Consum. Electron., vol. 15, no. 4, pp , Oct [12] K. Kozy, B. Gou, and J. Aslakson, A Low-Cost Power-Qualty Meter Wth Seres Arc-Fault Detecton Capablty for Smart Grd, IEEE Trans. Power Del., vol. 28, no. 3, pp , Apr [13] J. Momoh, Smart Grd: Fundamentals of Desgn and Analyss, John Wley & Sons, Mar [14] L. Crstald, A. Ferrero, and S. Salcone, "A dstrbuted system for electrc power qualty measurement," IEEE Trans. Instrum.., vol. 51, no. 4, pp , Aug [15] R. A. F. Perera, L. G. W. Slva, M. Kezunovc, and J. R. S. Mantovan, Improved fault locaton on dstrbuton feeders based on matchng durng-fault voltage sags, IEEE Trans. Power Del., vol. 24, no. 2, pp , Apr [16] S. Lotffard, M. Kezunovc, and M. J. Mousav, "A Systematc Approach for Rankng Dstrbuton Systems Fault Locaton Algorthms and Elmnatng False Estmates," IEEE Trans. Power Del., vol. 28, no. 1, pp , Jan [17] Y. Dong, C. Zheng, and M. Kezunovc, Enhancng Accuracy Whle Reducng Computaton Complexty for Voltage sag based Dstrbuton Fault Locaton, IEEE Trans. Power Del., vol. 28, no. 2, pp , Apr [18] P.-C. Chen, V. Malbasa, M. Kezunovc, and Y. Dong, Senstvty of Voltage Sag Based Fault Locaton n Dstrbuton etwork to Sub- Cycle Faults, submtted to IEEE/PES General Meetng, [19] P.-C. Chen, V. Malbasa, and M. Kezunovc, Senstvty Analyss of Voltage Sag Based Fault Locaton Algorthm, submtted to Power Systems Computaton Conference (PSCC), Wroclaw, Poland, [20] P.-C. Chen, V. Malbasa, and M. Kezunovc, Locatng Sub-Cycle Faults n Dstrbuton etwork Applyng Half-Cycle DFT Method, accepted for publcaton n IEEE/PES Transmsson and Dstrbuton Conference and Exposton (T&D), [21] A. A. Grgs, C. M. Fallon, and D. L. Lubkeman, A fault locaton technque for rural dstrbuton feeders, IEEE Trans. Ind. Appl., vol. 29, no. 6, pp , Dec [22] L. Yuan, Generalzed fault-locaton methods for overhead electrc dstrbuton systems, IEEE Trans. Power Del., vol. 26, no. 1, pp , Jan [23] S. Das,. Karnk, and S. Santoso, Dstrbuton fault-locatng algorthms usng current only, IEEE Trans. Power Del., vol. 27, no. 3, pp , Jul [24] R. H. Salm, M. Resener, A. D. Flomena, K. R. Cano de Olvera, and A. S. Bretas, Extended fault-locaton formulaton for power dstrbuton systems, IEEE Trans. Power Del., vol. 24, no. 2, pp , Apr [25] M. Kezunovc, Smart Fault Locaton for Smart Grds, IEEE Trans. on Smart Grd, vol. 2, o. 1, pp 11-22, Mar [26] M. Kezunovc, L. Xe, and S. Grjalva, "The role of bg data n mprovng power system operaton and protecton," Bulk Power System Dynamcs and Control - IX Optmzaton, Securty and Control of the Emergng Power Grd (IREP), pp. 1-9, Aug [27] Relay Assstant for Transent Relay Testng, Megger and Test Laboratores Internatonal (TLI) product datasheet. [Onlne]. Avalable: [28] 7700 Seres Power Supply Amplfers Techncal Manual, techncal manual, AE Techron, January [29] C. J. Km. and T. O. Balek, "Sub-cycle ground fault locaton Formulaton and prelmnary results," IEEE/PES Power Systems Conference and Exposton (PSCE), pp. 1-8, Mar [30] R. Moghe, M. J. Mousav, J. Stoups, and J. McGowan, "Feld nvestgaton and analyss of ncpent faults leadng to a catastrophc falure n an underground dstrbuton feeder," IEEE/PES Power Systems Conference and Exposton (PSCE), pp. 1-6, Mar

5 (a) (b) (c) (d) (e) (f) (g) (h) () (j) (k) (l) Fgure 3: All the quanttes are referred to the test sgnal, and the quanttes of y-axs are referred to mean s n the sub-fgure descrptons; (a) lne voltage ampltude vs. lne voltage ampltude; (b) lne voltage ampltude vs. frequency; (c) lne voltage ampltude vs. harmonc level; (d) phase voltage angle dfference vs. phase voltage angle dfference; (e) phase voltage angle dfferences vs. frequency; (f) phase voltage angle dfference vs. harmonc level; (g) phase current ampltude vs. phase current ampltude; (h) phase current ampltude vs. frequency; () phase current ampltude vs. harmonc level; (j) phase current angle dfference vs. phase current angle dfference; (k) phase current angle dfference vs. frequency; (l) phase current angle dfference vs. harmonc level.