Advanced DVR with Elimination Zero-Sequence Voltage Component for Three-Phase Three-Wire Distribution Systems

Transcription

1 Avance DVR with Elimination Zero-Sequence Voltage Component for Three-Phase Three-Wire Distribution Systems Margo Pujiantara * Heri Suryoatmojo ** Maurihi Heri Purnomo * Mochama Ashari * Takashi Hiyama ** Non - member Stuent - member Non - member Non - member Member Dynamic Voltage Restore (DVR) is a power electronics evice to protect sensitive loa when voltage sag occurs. The magnitue an phase of compensate voltage in DVR epen on grouning system an type of fault. If the system is floating, the zero sequence components o not appear on the loa sie. Meanwhile, in a neutral groune system, voltage sag is extremely affecte by zero sequence components. This paper presents new metho to mitigate unbalance voltage sag cause by zero sequence components. To avoi the impact of zero sequence components, in this simulation the value of zero axis components is etermine by using -q-0 axis metho. From simulation result shows that this metho able to compensate voltage sag with compensation error is 0.99%. Keywors: voltage sag, DVR, zero sequence components. Introuction Accoring to EPRI report (995), the revenue losses ue to poor power quality in U.S. business were $400 billion per year (). Power quality problems are cause by ynamic or non-linear loas an interaction between the loa an network. Two of the main problems in the fiel of power quality are voltage sag an instantaneous power loss. In aition, voltage sag has two main parameters incluing magnitue an time uration (). DVR is one of the best evices mitigating short voltage sag. DVR can provie the most cost effective solution to mitigate voltage sags by establishing the proper voltage quality level that is require by customer. When a fault happens in a istribution network, suen voltage sag will appear on ajacent loas. DVR installe on a sensitive loa restores the line voltage to its nominal value within the response time of a few millisecons. Most of the DVR esign stuies are base on the assumption of the balance three-phase system. An almost all the researches are base on the three-phase threewire systems ()~(4). Uner the foregoing assumption, only the restoration of positive-sequence an the compensation of negative sequence are taken into consieration, while zerosequence components are ignore. But the real power plant an istribution systems are generally using neutral grouning. Therefore the zero-sequence components will appear ue to groun fault in the system. Zero-sequence * Department of Electrical Engineering, Sepuluh Nopember Institute of Technology (ITS) Surabaya, Inonesia 60 ** Grauate School of Science an Technology Kumamoto University Kurokami, Kumamoto city, Japan currents can cause large voltage rops because of the voltages inuce by the coupling inuctances of the lines. Thus the influence of the zero-sequence components must be consiere when a DVR is esigne (5). This paper presente DVR uses -q-0 axis metho which concerns zero sequence voltage compensation to compensate voltage sags in istribution system with neutral grouning. The etaile switching DVR is moele by Matlab an the control scheme to eliminate zero-sequence voltage is propose.. Voltage Sag in Distribution System Accoring to IEEE Stanar, voltage sags can be efine as rms variation with a magnitue between 0 % to 90 % of nominal voltage an uration between 0.5 cycles to one minute. Voltage sags are cause by fault such as short circuits, overloa an starting of large motor. Voltage sags are the most common power isturbance which certainly gives the impact especially to the several types of equipment : ajustable spee rives, process control equipment, an computers are notorious for their sensitivity (6). Moel of istribution system is epicte in Figure. Which is raial an has a power plant with neutral groune system. In this moel, power plant has Grouning Main Bus DVR Fig.. Moel of istribution system Normal Loa Bus Groun Fault Sensitive Loa Bus

2 Vb Vabc Vabc to q0 V ref PI V Vq q0 to abc PWM Vc Va Fig.. Phasor three voltage vector iagram for single phase fault Source V T Loa Vabc Vq ref V0 = 0 Fig. 5. Conventional DVR control scheme abc to q0 V ref Vq ref V0 ref V Vq V0 Vq t to polar coorinat Vq V Vq t an V0 fuzzy polar rule t V0 Fig. 6. Propose DVR control scheme abc PWM to q0 rate capacity of 0 MW, 6.6 kv; both of feeer are equal R = 0.04 Ohm, L =.84 mh; normal loa is P = 4 MW, Q = MVAR; Sensitive loa is P = 5 MW, Q= MVAR. Most of the faults in power systems are single phase to groun fault. When single fault occurs in normal loa bus feeer, sensitive loa feeer will suffer voltage sag. Phasor of three voltage vectors iagram for single phase fault is epicte in Figur. Before fault occur, voltage in sensitive loa bus can be escribe using following equation : Va V b Vc Control Storage Energy Storage j j VSI Fig.. Topology of DVR obtaine V sag V ref V inj Fig. 4. Basic DVR phasor iagram () During fault, the voltage mentione in equation can be replace with Va V cos jv sin () V b j V j c Where V is voltage sag magnitue an is phase angle jump. Zero sequence components uring fault can be obtaine by V o ( VaV b Vc) Vo ( V cos jv sin ) () In the equation, the resultant of three vectors is not zero, it means that zero sequence components are generate uring groun fault. Therefore the process of eliminating the zero sequence components is require.. Dynamic Voltage Restorer DVR is power electronics evice which is installe in series with the istribution line system as can be seen in Figure. DVR uses semiconuctor evice to maintain voltage of sensitive loa by injecting voltage whose magnitue an phase can be controlle. The DVR is able to control the voltage across a sensitive loa by injecting an appropriate voltage phasor through an injection transformer. Nominal voltage will be compare with voltage sags in orer to get a injecte voltage by DVR.

3 Switching Line As. Za(k) grae N( ) α 90 P( ) / / Sector B O D(k) 45 Sector A p(k) (k) Zs(k) grae Degree G( D ( k)) 5 80 Fig. 7. Polar Form 0.0 D (k) Magnituo Fig. 8. Fuzzy Polar Membership Function Basically, the ieal DVR injection voltage can be obtaine as following Vinj Vref Vsag (4) Input Signal t Za Zs Rule Fuzzy Polar U Where Vinj is DVR injection voltage, V ref is reference pre V is voltage sags. Basic DVR phasor fault voltage an sag iagram epicte in Figure Conventional Control Scheme The simple iagram of conventional control scheme of DVR is shown in Figure 5. Transformation of a-b-c to -q-0 axis makes it easier for the control system to work. Control using the -q-0 transformation also avois the processing elay inherent in working with root mean square phasor values. Value of parameters V ref = an Vq ref = 0. Almost of all conventional DVR use wye-elta transformer to block zero sequence components. There for control metho concern to an q axis only. Conversion of a-b-c to -q-0 axis is shown in equation 5. v 0 va v sin t sin t sin t. vb (5) vq cost cos t cos t vc On istribution system without zero sequence component (floating or using elta wining), uring groun fault V 0 in - q-0 axis is 0. But in system istribution with neutral grouning, uring groun fault V 0 in -q-0 axis is not 0. Zero sequence components in -q-0 axis uring groun fault can be obtaine by V Fig. 9. Fuzzy polar control iagram ( Va Vb V ) (6) 0 c Because this control metho is concerne with only an q components, this scheme cannot compensate for zero sequence components. 5. Propose Control Scheme Propose control scheme in simplifie block iagram of DVR is shown in Figure 6. Voltage sag restorer metho employe is the comparison between real time voltage an voltage in a-b-c axis converte to -q-0 axis, using -q-0 reference voltage. Reference voltages are taken of V ref, V qref 0 an V 0ref 0. The ifference V ref with measurement is the error signal which shows the value of voltage rop V. It is clear that not only -q voltage is being compensate as many researches i, but also, q, an 0, so that the asymmetrical voltage sag (consisting of zero sequence component) can be restore. This paper using fuzzy polar application metho (7)~(8) (9) to improve the conventional PI compensator. Fuzzy polar consists of basic parameters: erivative multiplier (As), the angle membership function (, an raius membership function (Dr). Operation values of polar coorinate are shown in equations 7 to 9. p(k) [Zs(k) AsZa(k)] (7)

4 t (secon) Fig % sag cause by two phase to groun fault. t (secon) Fig % sag cause by three phase fault D( k) Zs( k) ( As. Za( k)) (8) ( k ) tan ( As.Za( k ) / Zs( k )) (9) Given the input signal Zs, the controller nee the signal erivative to get Za. Point p(k) is represente by input Zs as x axis an Za as y axis. To be use for fuzzy polar controller, input form p(k) in rectangular form shoul be transforme to polar form, D(k) as magnitue an (k) is the angle. The polar form of fuzzy polar is shown in Figure 7. The other factors which is also neee in this control system is maximum control signal Umax. Defuzzification rule as fuzzy polar output (U) for the control system is shown in equation 0 (0). t (secon) Fig.. 50% sag cause by two phase to groun fault restore by DVR. t (secon) Fig.. Zero sequence voltage using conventional control scheme. t (secon) Fig.. Zero sequence voltage using propose control scheme U( k ) G( D( k ))[ N( ( k )) P( ( k ))]. Umax (0) Where Umax is the maximum allowable control signal. G(D(k)) is the membership value of magnitue D(k), while N((k)) an P((k)) are membership value of the angle (k). Membership function of fuzzy polar is shown in Figure 8. Error signal V, V, Vq, Vq, V0, an V0 are then transforme to polar -q-0 form using equations. 7, 8, an 9. The result becomes the input of fuzzy polar control. Figure 9 is the simple moel of fuzzy polar with input an output. In reality, there is only one input, Zs. But it nees erivative signal Za so that it can be converte to polar coorinate by equations 7, 8, 9. Result of error compensation from fuzzy polar control is control signal which shows the value of the voltage will be injecte to the system by the inverter. 6. Result t (secon) Fig % sag cause by two phase to groun fault restore by DVR. Simulation was one using Matlab SimPower System. Simulation uration is 4 cycles. Parameters of fuzzy polar are shown in table. It is simulate that a short circuit fault occurs in normal loa bus (Fig. ) for cycles, which causes voltage sag in sensitive loa bus. DVR installation in sensitive loa bus is intene to restore the voltage which is istorte by sag. To see the propose control scheme performance, sags of 0%, 50%, an 70% in sensitive loa bus will be simulate. Voltage sag cause by two phase to groun fault an the compensate voltage are shown in Fig. 0 an Fig.. Fuzzy polar DVR can restore the voltage up to % as seen in Fig.. Fig. shows the value of zero sequence voltage in sensitive loa bus using conventional DVR control scheme. Fig. shows the value of zero sequence voltage at sensitive loa bus using propose control scheme. From Fig. an Fig. can be seen that the value of zero sequence voltage

5 ecreases from 0. pu to 0.04 pu. DVR propose metho is able to reuce the zero sequence components. Symmetrical voltage sag cause by three phase fault an the compensate voltage are shown in Fig. 4 an Fig. 5. Fuzzy polar DVR can restore the voltage up to 99.0 % as seen in Fig. 5. Several types of fault have been simulate an the result can be seen in table. The table shows that voltage sags cause by both symmetrical an asymmetrical fault can be restore well. Simulation results show that DVR using this metho can restore both symmetrical an asymmetrical voltage sags very well. Table, Fuzzy Polar Parameters No Direct input Quarature input q Zero input 0. As Dr. α 90º 90º 90º 4. Umax Table. Voltage sag an harmonics restoration. Voltage Sag Restoration (%) Error (%) 0 % GF % GF % GF % F % F % F % FG % FG % FG % F % F % F Conclusion DVR with the technique of elimination zero sequence at istribution system phase wire use neutral grouning was moele by Matlab SimPower System an fault cause by zero sequence in istribution system was simulate an analyze. Also, the effects of zero-sequence components were simulate an iscusse. This metho can reuce zerosequence components very well. Simulation results show that DVR using this metho can restore both symmetrical an asymmetrical voltage sags very well. The average error of DVR voltage sag compensation is 0.99%. This metho can be aapte to the conventional DVR moel. Refferences [] John S. Hsu, "Instantaneous Phasor Metho for Obtaining Instantaneous Balance Funamental Components for Power Quality Control an Continuous Diagnostics," IEEE Transactions on Power Delivery, Vol., No.4, pp , October 998. [] Zhang, Liong, Math H.J.Bollen, Characteristic of Voltage Dips (Sags) in Power System, IEEE Transaction on Inustrial Electronic, February 004 [] K.Chan, an A. Kara, "Voltage sags mitigation with an Integrate Gate Commutate Thyristor base Dynamic Voltage Restorer," Harmonics an Quality of Power Proceeings, Proceeings. 8 th International Conference On Volume:, pp , 998. [4] Ming Fang, A. I. Gariner, A. MacDougall an G.A. Mathieson, "A Novel Series Dynamic Voltage Restorer for Distribution Systems," Power System Technology, Proceeings. POWERCON International Conference on Volume:, pp. 8-4, 998. [5] Ding Hongfa, Gao Jun an Duan Xianzhong, "New Concepts of Dynamic Voltage Restoration for Three-Phase Distribution Systems," Proceeings of IEEE Power Engineering Society Summer Meeting Conference, vol, pp.47 4, July 000. [6] Math H. J. Bollen, Unerstaning Power Quality Problems:Voltage Sags an Interruptions, New York, IEEE Press, 999. [7] Margo P, M Hery-Purnomo, M Ashari, T Hiyama : Balance Voltage Sag Correction using Dynamic Voltage Restorer Base on Fuzzy Polar Controller, ICICIC 007 Conference Proceeings, Kumamoto Japan, September 007. [8] Margo P, M Hery-Purnomo, M Ashari, Zaenal PA, T Hiyama : Compensation of Balance an Unbalance Voltage Sags using Dynamic Voltage Restorer Base on Fuzzy Polar Controller, International Journal of Applie Engineering Research (IJAER) - RESEARCH INDIA PUBLICATIONS, IJAER 455 Vol. No.7, Delhi Inia, 008. [9] Fransisco Jurao, manuel Valvere, Voltage Correction By Dynamic Voltage Restorer Base on Fuzzy Logic Controller IEEE Transaction on Inutrial Electronics, may 00. [0] Thomas H. Ortmeyer an T. Hiyama, Frequency Response Characteristics of The Fuzzy Polar Power System Stabilizer, IEEE Transactions on Energy Conversion, Vol. 0, No., June 995. Appenix GF : groun fault. F : phase-phase fault FG : phase-phase-groun fault F : -phase-fault Bibliography MARGO PUJIANTARA, born in 965, got his bachelor egree from Institut Teknologi Sepuluh Nopember (ITS), Surabaya in 989 an got his master egree from Institut Teknologi Banung (ITB) in 995. He is currently a lecturer in Electrical Engineering Dept. in ITS an is actively involve in researches in the fiel of inustrial electronic. He is currently pursuing his octoral egree in Electrical Engineering Dept. ITS. HERI SURYOATMOJO (Stuent Member) was born in Magetan, East-Java, Inonesia on June 980. He receive his B.E an M.E. egrees in Electrical Engineering from Sepuluh Nopember Institute of Technology (ITS) in 004 an 006 respectively. He is currently a Ph.D. stuent at Kumamoto University. His research interest concerns on strategy an optimal configuration of renewable energy.

6 MAURIDHI HERY PURNOMO got his bachelor egree from Institut Teknologi Sepuluh Nopember (ITS) in 985. He got an M.S. an Ph.D from Osaka City University, Osaka, in 995 an in 997. He joine ITS as lecturer in 985 an became Professor in 004. His research fiel is intelligent system applications on electric power systems operation, control an management. Currently he is a IEEE Member. MOCHAMAD ASHARI got his bachelor egree from Institut Teknologi Sepuluh Nopember (ITS) in 989 an joine ITS as lecturer in 990. In 998 an 00, he got his Master an PhD egrees from Curtin University, Australia. His research fiel is inustrial electronics an applications, incluing harmonics filter esign, solar home systems, etc. He has receive many research grants from ADB, JICA, an Inonesian Government. Currently he is the hea of Electrical Engineering Dept in ITS. TAKASHI HIYAMA (Member) was born on March 4, 947. He receive his B.E., M.S., an Ph.D. egrees all in Electrical Engineering from Kyoto University in 969, 97, an 980, respectively. Since 989, he has been a professor at Department of the Electrical an Computer Engineering, Kumamoto University. His current interests inclue the application of intelligent system to power systems operation, management, an control. He is a senior member of IEEE, a member of Japan Solar Energy Society, an member of IEEJ.