1 Realistic Testing of Power Swing Blocking and Out-of
Contents
1. IIL Power Swing Detection Schemes Following is a brief discussion of the methods that have been used for detecting power swings and schemes used for OSB and OST functions Traditional schemes based on me t change of impedance and the newer methods that are implemented in microprocessor based relays are covered A Traditional Rate of change of Impedance Schemes for Detecting Power Swings Traditional methods measure the positive sequence impedance seen by the relay and the rate oF change of impedance During normal operating conditions the measured impedance is the load impedance and that point is far away Irom the distance protection zones When a fault occurs the measured impedance jumps immediately from the load impedance to a point on the impedance plane that represents the fault On the other hand when a power swing occurs the measured impedance moves slowly at some trajectory in the impedance plane and ata rate depending on he sip frequency between the machines This large difference in the speed of movement of impedance in used to differentiate between faults and power swings This method of differentiation is usually implemented using two impedance measuring elements separated by some impedance AZ and the use of a timer When the measured impedance crosses one element i starts the time A fault is declared if the swing crosses and operates the second impedance element before the timer expires On the other hand if t2. Swing Detection Schemes Some relays implement OST in two steps using Jec impedance elements an outer ch rscteristic a middle characteristic and an inner characteristic as shown in Fig 12 and two timers To be declared an unstable condition the power swing locus has to enter the outer hatacteristic to start timer 1 stay between the outer and inner characteristic until timer 1 times out move inside the middle characteristic to Mart a second timer 2 and stays between the middle und inner characteristics until timer 2 expires and then enters the inner characteristic A third mer with a very short delay may also be used which i stated by the inner characteristic and will expire as long as the swing locus stays within the inner characteristic E 2 Triple Concentric Characteristics for snd then will assert an OST condition PSB and OST functions A variation ofthis scheme is shown below in Fig 13 using mho elements tbe setting angle of each zone will affect the shape whether it will be a circles lens or tomato Additional blinders may be used to restrict the resistive reach of the mb elements GP WZ H Fig 13 Triple Characteristics for Power Swing Detection B Non Traditional Methods for Detecting Power Swings a Continuous superimposed A The superimposed current method compares the present values of currents with a buffer that is taken wo cycles eer Figure H states how the method work S
3. be used delay resetting when the swing locus exits the outer lider The same two blinder scheme can be used for out of step tripping except that the timer s sting is usually shorter than when used for PSB When a power swing occurs and the locus crosses the outer blinder the timer start if the timer tos expires before the swing locus roses the inner blinder a power swing situation is detected Ifthe swing continues its irsjectory and crosses the inner blinder an OST condition is declared The OST function can cither be set o trip immediately or wait until the swing locus leaves the inner blinder and er the imer tos expires Tripping immediately when the inner blinde is crossed is referred tos trip on the way in or predictive wipping or early tipping The other case is referred toas ip on the way out or delayed tipping In some applications tripping is allowed only when pole sipping occurs after a set number of times the circuit breaker is not rated for out of phase tripping early tripping when the slip angle is approaching 180 could damage the breaker In this ease wipping on the way out i used and on additional time may be used so that ripping will occur only when the slip angle i approaching zero degrees If tbe uppication calls for very fast tipping before the swing develops into a full ouat sep condition where vollae is further depressed early ripping is used but one must be certain that the circuit breaker will be capable
4. from one side a timcris stared Iit stay between he vo blinders and the timer expires the locus ler crosses the other binder going inthe opposite tection from where it entered an out of step condition is declared Fig 10Single Blinder Scheme Used for Ouolsep Tripping Fanction The single blinder scheme cannot be used for OSB function to block a distance relay for stable power swings because the impedance locus has to cross the two blinders on both sides the line As in the case of the two blinder scheme tripping can be delayed to allow the sip angle be reduced before tripping the circuit breaker Tripping only after certain number of pole slips cun also be implemented c Concentrie Characteristie Schemes The rate of change of impedance method is implemented in many relays using two concentric impedance measuring characteristics separated by some impedance value AZ or AR or AX The operation is based the sume principle a the two blinder scheme Fig 11 shows some of these concentric characteristics for detecting power swings and providing PSB sand OST functions The inner impedance characteristic can be an existing distance protection zone orit may be an additional impedance element specifically fr the purpose of detecting power swings The outer impedance element is usually and additional element specifically used for detecting power swings Fig 11 Concentric Characteristics Used in Power
5. just above the electrical center When Es is smaller than Ey the trajectory sa circle below the electrical center When the frequency fy is larger bat f the direction of the swing is from right to left whereas it will move from letto right is sis smaller than fg perm Fig 6 Impedance Trajectories Seen By a Distance Relay at Bus S Fig 7 shows a family of circles for different ratios of EE and diferent slip angles Fig 7 Impedance Trajectories for Various Ratios of IEE Ina rel power system the frequencies of the machines are not constant and the voltage amplitudes are dictated by other factors as well After a system disturbance the system will have power swings mostly small swings but some will be large enough and others may be unstable power swings Fig S illustrates some of the power swing tajeclories as seen by relay located on bus that could occur in power system Two distance protection zones are also plotted Fig 8 Impedance Trajectories Seen By a Distance Relay at Bus S during Power Swings As shown in Fig 8 the distance zones ean sce some ofthe power swings and can potentially operate und rip the circuit breaker Some operations may be desired but many will be undesired tripping particularly when the swings are stable To prevent tripping power swing blocking PSB schemes also called out of tep blocking OSB schemes are used to in protection schemes to block ripping of selected distance zones
6. of safely ripping on out of phase conditions In some applications both OSB and OST function are required The wo blinder scheme cun be use for both OSB and OST functions in the sume relay Some system disturbances produce power swings tha are slower and stable while other more severe disturbances can cause unstable and out of step conditions with an impedance trajectory that move much faster This difference in speed between these conditions can be used to identify if a power swing is Stable or unstable Two timers are used one imer toss with a longer setting is used for identifying stable swings and another timer tosr is used for identifying unstable swings When the power swing impedance locus crosses he outer blinder both mers Loss and lose are stated If the swing locus crosses he inner binder before both tiers expire a fault is identified and no OSB or OST occur IF only tosr expires and not tos before the inner blider is crossed an unstable power swing condition is declared and OST is asserted I the swing is slow both timers expire before the inner blinder operates nd the swing is deemed to be stable and only OSB is asserted b Single Blinder Scheme of OST Function The singl binder detection scheme is shown in Fig 10 and can be used for detecting unstable power swing ar out f sep conditions heee logic to determine the from where the swing entered and where tef When a power swing impedance locus esses onc binder
7. the locally measured voltage and o is the angle difference between V and the local current as shown in Fig 19 For the purpose of power swing election tis the rate of change of the SCV that provides the main information about the power swing and the small eror in the local estimate of SCV bas itle impact in the detection DF power swings Fucher approximating tat th source vollges ES and ER are equal obe posv sequence Ele ponive sequence swing center voltage SCV canbe spied scn 1 c0s 2 a The value of SCV is zero when te voltage ange is between the vo machines RIR and maximum when te sip anl is et Theta change of vlige at the cecticl center ie the time deve of SCV 2692 ran i ae This equation provides the relationship between the rate o change of SCV and the two machine slip frequency B In Fig 20 SCVI and the rate of change of SCV are ploted assuming a constant slip frequency of 1 rads Note that when the angle between the two machines is zero the voltage is normal and the rate of change of SCV is also zero when the two machines are out ahes with 8 equal to180 the rale of change of SCV is at its Fig 20 Plot of SCVI and its Rate oF change The SCV and its rate of change method has the Following advantages lt The SCV is independent of the system source and line impedances lt The SCV is bounded with a lower limit of zero and an upper limit of one per unit regardless of system i
8. Bg is the initial rotor angle of machine R C isthe frequency of machine S fu is the frequency of machine R The currents and voltages at bus S amd bus R are calculated by the formulas ses en Heen D n d Iein W lt Hlan els a ainan Rela eie zer dii The above formulas are for phase A Power swings are generally three phase phenomena and the phase B and Phase C quantities are 120 away from phase A Fig 5 shows a plot of the vollages and currents at Bus S during an unstable power swing or out of step condition JE KAI U Up x K SI H 4 VR ech bee Ji M HA T MII MAAS j en I UM WILD I eve MeN Fig 5 Voltage and Current Waveforms at Bus S Te positive sequence impedance seen by the relay at bus S in the direction o he line at any point is time is calculated by the equation Z4 Daman a Fig 6 shows the plot of the impedance locus seen by a distance relay t Bus S The trajectory of the swing locus depends on the ratio of the machine voliage amplitudes BE and where the system electrical center is located The electrical center isthe midpoint of the total system impedance Zc When both machine voltages amplitudes are equal the swing locus is a straight line perpendicular to Zy and passing through the electrical center where 8 180 A point on ibis line where 8 90 is also shown When E is larger than Eg the impedance trajectory follows a large circle that passes
9. HARMA YEA T bail Fig 14 Continuous Superimposed Current Method A delta current Al is detected ifthe difference is greater than 56 ofthe nominal current A continuous Al measurement for three cycles indicates a power swing condition and asserts SB provided that o distance protection element has operated A reset delay timer maintains the block to ensure bat it does not reset where the swing current passes through 3 natural minimum and AL detection might reset Another timer can be set t limit the amount of time for blocking to allow the distance protection zones to i The continuous superimposed Al current detection method can detect very fast power swings that are hard to detect with conventional schemes especially for heavy load conditions On the other hand very slow slip rates below He where the AI between two cycles is less iban the threshold of 5 nominal current are hand to detect with the AL method but his situation normally does not develop into swings that will enter the protection zones b Continuous Impedance AZ calculation The continuous DZ calculation method monitors the trajectory of impedances to detect the power swing conditions References 4 7 and 9 provide a detailed description of this method During power swing conditions he impedance trajectory generally moves in an llipical path The method uses an algorithm to calculate the values of R and X and compares them with previous values that are memorized Impe
10. Realistic Testing of Power Swing Blocking and Out of Step Tripping Functions O Verzosa Jr Jun Doble Engineering Company Abstract Sudden changes inthe power system such as faults nenvork changes due to line trip outs and disconnection ofa large load or ofa generating plant forces the remaining generators o adjust and setle to new stable conditions In some cases the swings are so large that some generators run out of step and lose synchronism With generators having different frequencies these power swings result in voltages and currents in different parts of the networks o swing in amplitude and phase angle The impedance measurements based on these varying voltages and currents will also oscillate If the measured impedance becomes very small and enters the distance relay zones it can lead to an undesired trip of the distance relay Distance relays have power swing blocking features to detect these conditions and prevent a misoperation for stable power swings However during the blocking period a fault cam occur on the line and the distance relay must detect this condition and unblock the distance zones For severe power swings and out of step conditions it may be desirable to trip on certain transmission lines to separate parts of the system into sub grids that will eventually settle at stable states instead ofa total system collapse and also making it easier 1o restore the power system An out of step tripping function can
11. and other similar laboratory tools are extremely expensive I also requires highly specialized personnel to operate them and to do the modeling of various components and run various case scenarios It is laborious and time consuming These sophisticated power system simulation tools EMTP and stability programs as well as RTDS are necessary when deciding if PSB and OST functions are needed in the protection of the power system and for maintaining its stability and deciding if remedial action such as system separation is required The application of PSB and especially OST functions need to be thoroughly studied before they are used and put inservice The settings have to e calculated especially forthe blinder based conventional detection methods and verified that they work for various conditions that they are intended for In any ease itis impossible to study and simulate every possible scenatio as there are too many combinations and permutations of generation load faul types fault resistance faul ocation ak inception angles relay and circuit breaker operating times and network configurations Only the most plausible scenarios are studied and simulated and these can go up to hundreds of cases Eventually ihe protective relays will have to be installed configured and set in the field before putting tem in service h R necessary hat the relay configuration and settings are correct and verified by testing In many cases COMTRADE fil
12. apid advancement of personal computer technology and modern tet sets realistic simulation of power systems that were the domain of laboratories and mainframe computers are now possible to perform with ease in the field The paper described two simulation methods the classical two machine model and another one based on a single source where one specifies impedance points and the rate of change of impedance between points The fied est engineer and technician has now at Ne or her disposal the tools o easily and realistically simulate stable and unstable power swings and faults to perform functional testing o the protection relays together with the PSB and OST functions The technician can use these testing methods for both non conventional and conventional power swing detection methods and schemes These new tools are not only for feld testing but can be used in the laboratory for testing the performance of PSB and OST functions for targeted boundaries of the distance zone and power swing detection characteristics and specific slip conditions and fault combinantions More complex situations and specific power system application of PSB and OST however ean only be simulated using the more elaborate tools ike EMTP and RTDS COMTRADE files from these programs are can be used playback using modem test sts VL References 1 Power Swing and Out of Step Considerations on Transmission Lines IEEE PSRC WG Do A Report tothe Power Systems Relayin
13. be implemented asa standalone relay or added as another function in a distance relay to detect these conditions and perform a controlled trip the transmission line The traditional approaches for swing and outof step detection include parallel single or double blinders on each side ofthe line angle concentric impedance elements which may be of circular shape or quadrilateral shape with a timer to measure the impedance trajectory time from when the outer element detects the power swing until the time that the inner element detects the swing Ifthe time is longer than the set time the power swing blocking function will block distance elements from undesired tripping whereas it will not block for fils More recent methods of detecting these conditions include continuous rate of change of impedance swing center voltage and superimposed current Testing relays that employ these different methods of detection is a challenge Traditional state simulation testing methods o simulate specific points in the impedance plane may work for the traditional detection schemes but will not work for the more recent power swing detection schemes A transient playback of recorded power swings or transien files from EMTP and stability programs provide a realistic test but require highly specialized personnel and iis difficult to target the impedance zones totes An easier way of simulating realistic waveforms has been developed that can test all types of power swing a
14. cheme together with the PSB and or OST functions Visualize on an R X diagram graphical user interface GUD the following the characteristics of the PSB and OST elements the characteristics ofthe protection zones the power swing impedance trajectory Use the GUI to aid in entering simulation points Provide visual plot of the voltage and current waveforms Noneed for additional hardware just the existing test sets and laptops Be able to simulate required system conditions for testing such as stable power swings unstable power swings load and faults Be able wo target specific areas ofthe protection zones and the zones of the PSB and OST functions by specifying the trajectory of the power swing impedance with litle ha Beable to control of source voltages machine frequencies and electrical center The simulation testing methods are described below a Classical two machine model The simple clasical two machine simulation model uses the same model shown in Fig 1 Fi 25 shows the required parameters and data Most of the data are automatically filled out based on other data that are required for testing the main distance protection functions The main parameter that needs to be changed is the frequency of the sources to control the slip frequency The technician can also control he trajectory of the swing locus by specilying the location of the electrical center as a percentage of the line l
15. cting relays to be used for the application For the more important line applications these studies are required and transient files are available However in some cases they are not readily available in the feld for the test technician to use Or there may be so many test cases that one does not know what to use h is also difficult to target the impedance zones and the PSB and OST characteristic boundaries B Playback disturbance records from DFRS Power system disturbances such as faults and swing recorded by digital fault recorders DFRS and protective relays with transient recording features can be played back to relays using modern protection test sels These files are usually exported by the DFR and relay software in COMTRADE Format These tests provide the most realistic testing for the specific events for which they were recorded but such recordings are limited To test other relays with a different seting and to tet for other power swing conditions there is usually not enough of these records that are applicable C Transient network analyzers and Real time digital simulators Protection relays with power swing blocking and ripping functions can be tested in the laboratory with a very realistic closed loop simulation of the power system and various system fault and switching conditions using transient network simulator and Real Time Digital Simulators RTDS The RTDS can also export transient files in COMTRADE format However RTDS
16. dances are calculated every 14 cycle To detect a power swing it uses several criteria for monotony continuity and Smoothness and it uses thresholds that are caleulated automatically from previous samples Monotony ensures that the direction of movement has not changed by evaluating the derivates of R and X Continuity ensures thatthe impedance trajectory is not stationary and this requires thatthe distance between consecutive R and X values exceed a threshold Smoothness ensures uniform movement o the impedance trajectory with mo abrupt changes and this requires that the ratios of successive differences of R and X are below some threshold xx Bla Fig 17 Smoothness geg for power swing detection The algorithm dynamically adapts o the trajectory by automatically calculating the ibresholds for the next calculations considering previous values After afew about 6 successive calculations where the stated criteria are fulfilled a power swing condition is declared Once a power suing is detected gradually changes such a stable swing changing direction will not result is removal of the power swing declaration unles the criteria are not fulfilled for another few successive calculations and the impedance locus is inside one ofthe protection zones such as the presence of a fault Blocking ofthe protection zones is started only when the swing impedance enters a starting polygon characteristic This characterist
17. ds of Testing PSB and OST Functions Like other protection functions itis essential that the power swing functions are satisfactorily tested prior to placing them in service and testing them together with the other main protection functions as a complete functional protection system is important Testing relays that employ these different methods of power swing detection and verifying performance requirements isa challenge A Playback Transient Data from Electromagnetic Transient Programs EMTP and Transient Stability Studies Transient simulation data from EMTP ATP programs provide very realistic test und a large number of iles can be generated Raw data from stability programs with simulation steps of a quarter cycle or less cun also be transformed into COMTRADE fies for testing However these programs require highly specialized personnel to perform such studies and these tools are generally very expensive It requires a large amount of time to model the power system or at least the relevant parts of the system and a huge number of case studies for various conditions of generation and load mix faults relay operating times line outages etc Playing back transient COMTRADE files from system studies into modern test sets is one of best ways to test the protection application and verify thatthe relays that will be used meet the application requirements and such testing be performed in a laboratory This process ould be a part of sele
18. ength which can be in the range of 100 to 300 of the line This recalculate the source impedances n order to bring the electrical center to he desired location Changing the source voltage amplitudes controls whether the trajectory is below or above the electrical center The actual trajectory is displayed in an R X diagram together with the line impedance and the relay characteristics This allows one to contro whether or not the swing locus should pass through zone selected zones of protection The zoomed out entire swing locus is shown in Fig 26 The relay impedance characteristics ean be entered in a separate user interface used for testing the main protection functions wnceren h8 Y aage aren Fig 25 Classical Two Machine Simulation Model Fig 27 shows a plot of voltage and current waveforms The points where the relay elements operated are also shown AER neng pe 26 Zoom out Showing Circular Impedance Trajectory Timers can be set up to measure the operating time of relay trips and other elements including operation of the outer and inner power swing characteristics and when OSB asserts EH pe 27 Voltage and Curent Waveforms Seen By Relay Tew tee ees Jace rti rs c ve a er BECH Fig 28 Timing of Various Relay Elements 1 Simulating any number of
19. es are not always available or those that are available do not verity the settings There will also be times when testing will be required for maintenance and for trouble shooting The test technician in the eld will not be able to rely on the power system engineer to provide transient iles Field personnel are often pressed for time to complete numerous tests ofthe protection system and ave Wenner to Finish the tests and put the relay inservice Easier and simpler ways of simulating faults and power swing conditions that ae readily available tothe est engineer or technician inthe Field are needed and these will be discussed below D State Simulation Test Methods Traditionally state simulation methods of simulating impedance points by injecting several static states of voltage and currents have been employed for testing distance relays and power swing blocking and out of sep tripping relays This type of test works successfully hen testing conventional power swing detection schemes that are based on stepped rate ol change methods using blinders und concentric characteristic that are separated by some AZ and fixed timers Even this method of testing may require complex calculations especially When one has to target impedance points at other angles away forthe R axis of the impedance planc A gr phical user interface using a point and click with the mouse on the impedance plane with the distance relay tipping zones and the PSB a
20. g Committee ofthe IEEE Power Engineering Society July 2005 2 W A Eimore Editor Protective Relaying Theory and Applications Marcel Decker New York 3 GER 3180 Applicaton of Ou oF tep Blocking and Tripping Relays by John Berdy HI Dr Jurgen Holbach Siemens AG Raleigh NC USA New Out OF Step Blocking Algorithm for Detecting Fast Power Swing Frequencies NT Annual Western Protective Relay Conference Washington State University Spokane Washington October 21 23 2003 5 Demetrios Tziouvaras and Daging Hou Outof Step Protection Fundamentals und Advancements 30th Annual Wester Protective Relay Conference October 21 28 2003 Spokane Washington 16 G Benmousal D Hou and D Triouvaras Zerosetng Power Swing Blocking Protection 31st Annual Western Protective Relay Conference October 19 21 2004 Spokane Washington I7 Proper detection and treatment of power swing to reduce the risk of Blackouts Blumschein Y Yelgin M Kereit Siemens AG Energy Sector 8 MICOMbo P443 EN M CA2 Technical manual ALSTOM 2011 wwwalstom com 9 Siemens AG Energy Sector User Manual Distance Protection 7SA522 V4 61 Ordering Nr C53000 G1176 C155 5 Online Available hip wwv siprotec com 10 GEK 113248 D90PIus Instruction Manual revision1 8x Copyright 2010 GE Multilin 215 Anderson Avenue Markham Ontario Canada L6E 1183 11 SEL 421 4 5 Relay Instruction Manual Date Code 20101221 2010 by Schweitzer Engineeri
21. he timer expires before the impedance locus crosses he second impedance element then the impedance movement is declared a power swing as Two Blinder Scheme for PSB and OST Functions The rate of change of impedance method is implemented in many relays using a pair of blinders separated by some impedance value AZ and a imer loss The Iwo blinder scheme is shown in Fig 9 Two parallel blinders are placed to the right of the line impedance und another pair on the ef side The timer toa is started when the swing crosses the outer blind Fig 9 Two Blinder Power Swing Detection Scheme pa fault condition the impedance locus moves quickly tothe fall impedance point and the inner binder operates immediately well before the time expires Incase of a power swing witha slower movement of the impedance locus tos Will expire before the swing locus crosses the inner blinder in which case power swing condition is declared IF the swing locus enters the region between the two blinders and stays there before the timer hes expires a power swing condition is declared This scheme ean be used for a power swing blocking function to and block the distance relay protection zones from tripping in ease the swing locus proceeds further into zone operating characteristic A reset timer started when the PSB asserts is normally used to force it to deassert when the timer expires even if the impedance locus slays inside the inner blinders Another timer may also
22. howing Step Changes The main drawback of this testing method stems from the act tha it does not represent physical reality in an clectrical system because of step changes or sudden jumps in the voltage and current waveforms Real power swings are slow moving and smooth as opposed 10 faults that jump instantly Power swing detection methods tha are based on continuous measurements of small delta currents or delta impedances behave unpredictably because of unrealistic tst quantities Step changes are seen by some detection methods as faults or switching conditions and not power swings Even when using a large number of tates which is Very laborious and error prone in the hope of smoothing out the discontinuities such tests still cannot sucessfully test these non conventional power swing detection methods because measurements are made every cycle or more often E Need for Simpler Easy to Use Testing Methods that Approximate Reality The protection est engineer or technician needs to have simpler easy to use tools in the feld or testing PSB and OST functions and such tools should consider the following Simple and easy to use by technicians with litle power system background equiting minimum input from the user Approximate physical reality reasonably well to allow proper testing o all conventional and non conventional methods and schemes of PSB and OST functions i should allow complete functional testing of ine protection s
23. ic is automatically calculated to encompass all the protection zones to be blocked This method of power swing detection handles slip frequencies more than 7H and requires no calculations Tor PSB applications The dynamie calculation of thresholds allows the method to handle more complex multi machine oscillations c Swing Center Voltage V cos and Its Rate of Change In a two machine equivalent system the electrical center is the electrical midpoint of the total impedance of the line and two sources When a power swing occurs due toa slip in frequency between the tuo machines the voltage at the electrical center goes to zero when the angles between the wo machines are 180 apart The voltage at the electrical center of the system is referred to as the swing center voltage SCV References SEL manual PSRC ABB manual provide more details of the method Figure 18 illustrates the voltage phasor diagram ofa general two machine system with the SCV shown as the phasor from the origin to ihe point v g 18 Voltage Phasor Diagram of a Two Machine System Ina homogenous high voltage power system the impedance angle 0 is close lo 90 and the diagram can be redrawn as shown in Fig 19 Fig 19 Veose is a Projection of Local Voltage VS Onto Local Current I At the location ofthe relay using the ocally available voltages and currents the voltage at the swing center can be approximated SCV lt I mee an Where IN is the amplitude of
24. ied Simulation Points for Smooth dei Between Points The method calculates the voltages and currents based on a simple model with a source behind bus S where the relay s located and the impedance of the point from the relay location s shown in Fig 22 For ground faults the residual factor is considered The complex impedance Zt presented to he relay Varies with ime and is calculated foreach time step Lms o the simulation according to the specified points and rate of change of impedance The current li and the bus voltage Va for each simulation step are calculated in the usual vay This test method allows one to test simple scenario where an end fault occurs and is cleared and the impedance locur moves back to the load area t then a stable power swing ensues and enters into zone ta the tine protection and then oes back to the load area IF SB works coretly and blocks no tipping should occur Such atest scenario is shown in the Fig 34 pe M Test Scenario With Fault and Power Swing The table of points in Fig 35 shows more details about each point including the speed of movement between points dz4dt the duration of each point as well as the source impedance ZsMag and ZsAng behind the relay which can be change by the user if desired The initial point is a load where the impedance is 36 8650 at 9 7 fora duration of 05s with a source impedance of 7 8 at 847 A fault point 2 occurs i
25. ied out for unbalanced faults such as an A phase to ground fault and a phase to phase fault outside the inner set of blinders but within zone 1 these resulted in faster tipping within 3 cycles Ths simulation testing again confirmed thatthe negative sequence directional element resets PSB immediately as described in the relay instruction manual This method of simulating faults and power swings by means of user specified points and a specified smooth rate of change of impedance makes it easier for the test engineer and lechnician to conduct more complex testing scenarios in the field as well as a laboratory V Summary and Conclusions When power systems are subjected to faults and large power system disturbances power swings both stable and out o step conditions ean cause transmission protective relays to undesirably ip and can cause even more severe disturbances Power swing blocking functions detect these conditions and prevent undesired operation of relays In other severe eases of system instability separating parts of he system may be required to save it from total collapse Outof step tripping functions detect unstable conditions at specific points in the power system where separation may be desired The paper discussed the conventional and some non conventional methods of power swing detection for PSB and OST functions PSB and OST functions must be fully tested before putting them in service Complete functional testing of the main
26. ig 30 Directed Power Swing Trajectory The first case for early wipping or trip on the way in resulted in tipping at 51222 cycles from the start ofthe test before the swing slip angle reaches 180 as shown in Fig 3 The second case for tip on the way out resulted in wipping at 54 7 cycles from the start of he test well afer the swing leaves 180 going left as shown inFig 32 Fig 32 Tore ren Both cases shown above were tested fr a slip frequency of 5 Hz Additional tests down to 3Hz and up to 6Hz showed similar relays For another test at 2H only PSB operated At THz OST stopped working The classical two machine method is very easy to use for unstable power swings A more complet testing is also required to test for stable power swings where the swing locus does mot cross the line impedance User Specified Points and Smooth Rate of Change of Impedance Between Points This method is illustrated in Fig 33 It involves specifying several impedance points using mouse clicks on the R X diagram specifying the rate of change by which the impedance moves smoothly between points and how long a point may stay stationary ata certain location to simulate a fault or load condition For swings the rate of change of impedance dads slow and for Faults the movement can be very Fast or instantaneous r TT ms nine Fig 33 User specif
27. mpedance parameters Hucan detect slow swings and very fast power swings In practice relays are usually designed for security and dependability and some minimum and maximum threshold of dSCV to guarantee PSB detection from 0 1Hz to 7H Abee decrease the sensitivity ofthe power swing detector the positive sequence impedance should be within a starter zone which is automatically determined based on the characteristics ofthe protection zones tobe blocked Application of the SCV method for PSB functions requires no settings and no stability studies The V cosy method is also used in out of step or pole slip protection and can detect sip rates of 0 2Hz to SH with additional eriter for thresholds for security 4 Some Limitations The three non conventional methods can easly detect very fast swings but may need to be complemented by conventional rate of change of impedance method for extremely slow power swings with slip frequencies of He or less In general this situation would not develop into a power swing condition that enters the distance protection zones Under very slow power swings conventional schemes can be set with much smaller AZ between the inner and outer blinders since there will be a lot of time for the impedance to travel between the two blinders One can also set the complementary conventional inner blinder as close a possible to the distance zones to be clocked C Other Power Swing Detection Methods There ate other powe
28. n zone 2 for 0 15 and is then cleared jumping back to the load area point 3 After 0 05s a power swing then ensues where the swing locus moves from point 3 to point 4 ata rate of 12862 crossing the outer and inner characteristics which should result in PSB operation It then moves ata slower rae of 750s to point 5 which is in zone 1 but tripping is expected to be blocked Then it moves out of zone 1 and back to the load area through points 6 and 7 ana T POETAE Zei Del ei oe Fig 36 shows the voltage and current waveforms for the above simulation points and how the relay responded h shows the Outer characteritic operate first then PSB asserting before the Inner characteristic could operate Fig 36 Plot of Waveforms and Relay Operation No tripping ofthe distance elements occurred when the the swing locus entered zone 1 fora few cycles This confirmed that the relay responded correctly for this simulation scenario Another testing scenario may involve the occurrence of a power swing where the PSB function blocks zone 1 2 and 4 While the block is asserted a 3 phase fault occurs on middle of the protected line as shown in Fig 37 This will est how the relay responds and how quickly itean remove the block and clear the fault Depending on the algorithm employed by a relay it could allow fast tipping in about 2 cycles or more or it may have to wait for
29. nd OST characteristic drawn makes it easy for the test technician or engineer to perform tests that target A specific points For testing a PSB Al function a minimum of three points is needed the fist one in the load area fora pre swing duration one XKR Md RX point between the inner and outer Fig 21 Graphical User Interface or Entering Sa characteristics and a third point Simulation hmpedanee Pointe inside the tripping zones voltages and current fr eae Ee a Ee S Label state are calculated based onasingle er aan TT r3 source model with constance source Er Nimm e impedance providing a dynamic test Em nim ua wi me n c zT ENTE E EX Fig 23 State Table of Voltages and Currents Tor Points Entered in Fig 21 Fig 22 Single Source Model I the state duration ofthe second point i a litle longer than the PSB timer setting it wil block tripping ofthe tipping zone when the impedance moves to point 3 otherwise i the state duration i less than the timer setting PSB will not assert and the tipping zone will operate when the impedance moves to point 3 The procedure is similar when testing OST functions that employ tripping on the way in or an early trip I the OST function is set fr tripping on the way outa fourth impedance point is required and placed outside the outer zone opposite the initial toad impedance Fig 24 Plot of Vaveforms for States s
30. nd out of step detection schemes This paper covers the basics of power swing and outof step phenomena and the various schemes of detecting these conditions and ways of testing the power swing blocking and out of step tripping functions 1 Basic Power System Stability and Power Swing Phenomena Power systems under normal system conditions are operated very close to their nominal frequency of 60 Hz or S0 Hz with very small deviations inthe order of 0 02 He for larger systems to 0 05 Hz for much smaller systems Under these steady state conditions power generation and load are balanced E HS Z Qe 7 ba d peL Two Machine System Model Considering e classical tvo machine system model shown in Fig 1 the power transmited ean be represented by the equation pe a Where Bs isthe voltage of machine S Ey ehe voltage of machine R 7 cis the angle by which E leads Ey EH Ys cis he rolor angle of machine S duo cis the rolor angle of machine R 27 cis he total impedance between the tuo machines consisting of Zs Ze and Z ECH Zs islhe impedance of source S 2 is the impedance of source R 2 ee impedance ofthe transmission Hine The power angle curve shown in Fig 2 graphically describes the power relationship between the power transmitted and the angle between the two ends It shows tht the power transfer increases with increasing angle 8 from 0 to 90 and then decreases beyond 907 Systems are normally operated well bel
31. ng Laboratories Ine 12 IMRKS06312 UEN_A_en Technical 2011 ABB 13 E W Kimbark Power System Stability vol 2 John Wiley and Sons Inc New York 1950 14 C37 233 2009 IEEE Guide for Power System Protection Testing 15 FGTesT software version 2 ference_manual_REL670_1 2 Copyright Doble Engineering Company VIL Biography Quintin Verzosa Jr Jun received his BSEE degree from Mapua Institute of Technology Manila Philippines n 1976 He joined National Power Corporation Philippines in 1978 as a relay engineer and was later promoted to principal engineer and then manager of Power System Analysis amp Protection He became manager of Protection Control amp Communications Engineering Design In 1993 Jun joined GEC ALSTHOM in Hawihome New York where he worked on protection and control applications and testing and later became Protection Systems Engineering Design manager Jun joined Doble Engineering Company in 1998 as a senior protection engineer and s currently manager of Protection amp Automation Engineering Services He is involved in protection test systems application support services and training research and development of protection models and testing algorithms He isa member of IEEE and is actively involved in several Power Systems Relaying Committee PSRC working groups Jun i also involved Cigre working group WG BS45 on testing techniques of protection and automation systems
32. ow the maximum power transfer t 90 the maximum power transfer angle at some power Py corresponding to an angle 6 The maximum power astra e Fig 2 He Power Angle Curve Teel 1a large disturbance occurs such as a Dal onthe ine the power transmitted is suddenly reduced the electrical output during the fault represented by the lower curve in Fig 3 and the electrical output of machine S decreases to Py However the input power Py equivalent to the mechanical torque applied to the generator cannot instantly decrease and this imbalance e the machine rotor to accelerate and he angle inercases This analysis neglects the operation of governors that change the mechanical input and voltage regulators that control the excitation and change the machine voltage amplitudes the altis cleared after some time P When the angle is c where the electrical power is now greater than the mechanical power the machine will start to decelerate However due to inertia the machine rotor angle will continue to increase to Se Ms point Arca representing the accelerating energy will equ to rea 2 epesntng ie deccraing energy Tis i ferred io Sa ECG amp Ka oeh BETEN Figure boss at al clearance anda alle vag the hal h cleared quickly the Area 2 can be equal to Area 1 before the angle can reach the limiting angle 5 and the system will eventually recover after some oscillation and sete at be initial operating condition a
33. protection functions together with the enable PSB and OST functions is important to verify that the relay is correctly configured and set and all elements of the protective relay interact properly In the past the test technician had to rely on highly specialized protection and system engineers to perform extensive transient stability studies and RTDS simulations to come up with COMTRADE files that could be used inthe field to test PSB and OST protections functions Disturbance fles from DFRs and relays with fult recording capabilities also provide very realistic power system quantities for testing However these iles or the right specific files are not always available in the field for the specific application being tested Conventional blinder hased and concentric characteristic power swing detection schemes could be tested with conventional state simulation methods Even state simulation methods require a highly trained test technician to caleulate the test currents and voltages to successfully carry out the tests A graphical user interface in the impedance domain allows the technician to visualize the relay characteristics and set up dynamic tests with simple mouse clicks on the R X diagram Newer power swing detection methods and schemes however cannot be tested with state simulation methods because the abrupt step changes and discontinuities are contrary to physical reality and make the relays behave unpredictably Thanks to the r
34. r swing detection methods that have been used are currently being investigated or being implemented but are not covered in this paper These include the R Dot scheme described in reference 1 the use of Synchrophasors and wavelets D Some PSB and OST Application Considerations and Line Protection Requirements during Power Swing Conditions In order to properly apply PSB and OST functions detailed system studies are necessary o determine where PSB and or OST functions are required Reference 1 discusses these considerations and requirements in detail The usual protection performance requirements such as speed selectivity reliability and sensitivity should be considered also during power swing conditions Some general requirements are simply listed below During a power swing where PSB has asserted the occurrence of a fault balanced and unbalanced should be detected and remove blocking in the shortest possible time To allow protection to clear the Ful The protection should maintain selectivity when PSB is removed due to an unbalanced fault and while the slip angle between machines is close to 180 Protection elements should be secure to external faults during out of step conditions While it may difficult for older ine protection relays to satisfy these requirements newer relays have additional logic to perform much beter These should be considered when applying PSB and OST functions and also during testing IV Metho
35. the PSB duration timer to expire In this particular case the relay PSB function asserted first Then when the three phase fault on the line occurred the PSB reset in 7 eycles and the relay tripped at the same time as shown inthe Oscillograph of Fig 38 Seater er Fig 37 Power Swing and Faultin Zone ao Imer Set of ne 38 Tc far Fault Ine ane Blinder Resets PSB mare This particular relay tested has a feature to reset PSB faster if the impedance remined inside an additional se f inner blinders shown in the Fig 37 next to the line for a certain amount of time The timer automatically adjusts depending on how fast the power swing entered the PSB detection concentric characteristics Simulating a faster swing locus across the wo concentric characteristics makes resetting of PSB faster while simulating a slower swing eme in slower reseting of PSB Additional tests confirmed the theory of operation as described in the relay manual Fig 39 Power Swing and Fault in Zone See 1 eee are Outside the Incr Set of Binders Fig 4 Oseillograph Sowing PSB Assertion and Blocking of Zone 1 Another simulation was made as shown in Fig 39 moving the 3 phase fault outside this inner binder but still within zone 1 resulted in no tipping even for a longer duration of 30 cycles as shown in Fig 40 since the PSB remained asserted It tripped only after increasing the fault duration beyond 60 eycles More simulation tests were carr
36. tv and the system is stable The system oscillation that that occurs is considered a stable power swing On the other hand if he fault clearance is slow angle Gc advances too far that Area 1 becomes large enough such that Area 2 cannot become equal to Area before the angle reaches R m shown in Fig 4 Beyond this p point the electrical power again becomes less than the mechanical power the otor will now accelerate again amd the rotor angle will L2 continue to increase beyond 180 Pole Slipping will occur and the machines will X continue to rotate at different speeds This condition is considered an unstable power Swing or an out f step condition Fig 4 Unstable System due to Slow Clearing Time IL Relay Quantities during Power Swings Referring to Fig the currents and voltages seen by a protective relay at Bus S can be calculated by the Following equations e DEEN D The impedance seen by a distance protective relay at bus is caute as D Zen 2 1 Where 1 s the mmer from Bus to Bus h V e isthe votage at bus s During power swings the frequencies of the two machines fs and fg become unequal and the instantaneous values of the voltages and currents have to be calculated considering the time varying rotor phase angles of both machines The rotor angle of machines S and R at any point in ime ar Pest pg 2a fet D e D Beg Det o Where 8 is the initial rotor angle of machine S
37. ving urs allows one to test OST functions where the relay is st to rp only a set numberof pose Slips Fig 29 shows the waveform where the umber of swing turns or pole slips is Fig 29 Simulation of Maliple Pole Stips The classical two machine model is implemented by calculating the rotor phase angle of each machine for every step d starting from the initial angles and their specified frequencies e is the initial rotor angle of machine S Hg is the intial rotor angle of machine kK PaO Os Sulo as Qu D Be T at us Other quantities ae then calculated in the usual way first the current and then the voltages are each bus using equations iste sent eimi 1 nas re neon D SC an Wild lt Wielt Weien des a Wale a The instantaneous values for the voltage and current waveforms can then be easily calculated Tor phase A with equations it 9 VZ CO sinta OH 20 s 7 Y2 COL sinGev 0 en vd E ICI singu O en Since power swings are three phase phenomena the quantities for phases B and C are just 1207 apart and can easily be calculated with the same adjusted by A 120 Because itis easy lo direct where the swing locus will pass through and to specily the frequencies itis easy to test for OST functions The relay was tested for two different settings tip on the way in as well as tip on the way out Fig 30 shows the points where tipping is expected to occur F
Download Pdf Manuals
Related Search