Home

D5.1-Part 3_Tool Box for modeling ETN components

1. Date 19 08 2011 Page 45 DEL WPS 1 Part 3 Toolbox for model ing ETN components V3 doc 5s PEGASE COOPERATION u DC EUROSTAG u DC UDE Figure 6 5 Test 2 results for Model 0 difference cannot be identified 2 When alternating chopper switching on and off occurs as in Test 4 the difference in DC Date 19 08 201 1 voltage may be observed p REC EUROSTAG p REC UDE p SEC EUROSTAG p SEC UDE Page 46 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc iz PecAse 99 COOPERATION q REC EUROSTAG q REC UDE q SEC EUROSTAG q SEC UDE u_DC EUROSTAG u DC UDE Figure 6 6 Test 4 results for Model 0 difference in DC voltage doesn t influence on active and reactive power This can be explained by the difference in the way EUROSTAG and UDE software handles discontinuities In any case this difference doesn t influence on active and reactive power calculated 6 4 3 Conclusion on test results Tests made prove that EUROSTAG implementation of VDC HVDC model in general follows UDE prototype Some differences may be observed but they are not significant and do not produce any notable error in sense of modeling the whole power system 6 5 EUROSTAG limitations The following limita
2. ccccccceceeeeeseeeeeeeeeeeseeeesseeeesseeeesaeeeenas 39 6 1 Versions of IMPIOMENtATtION cccceescecceeseeeceeeseecceuseeesseueeecceuseeeseaeeessaeeeeseeeesssaeeeessaneeesseass 39 6 2 Implementation of specific parts of the model cccccseeececeeeeeeeceeeeeeeeeeeeeeeeseeesaeeeeesseeeeeseees 39 6 2 1 FLO doleo Nm E E 39 6 22 Current conversion cccccccecccceceeeeaeeeeseeeeececeeeeeeeeeeeeeaaaaaausaseeeeeeeeeeeeeeseesaaaaaaaagsseesess 40 6 2 3 PUY STCHSSIS IOC M MERE 40 6 2 4 Losses modeling sssssssssesssseeeee eene nennen nnn nnnnna nnn rss nna nnn risen annia nns 41 6 2 5 Implementation of converter reactance esssesssssesssseeeeeene nennen nnns 41 6 3 Model parameters GESCIrIPtiOn ccccccssseeceeeeaeeeeeeeeeeeeeeeeeeeeeeseeeesseeeeeeeeseaeeeeesaeeeeeesssageeeeeeaas 42 64 MOC SITS S WING MR 42 6 4 1 Sample power grid cccccccccseseseeceeeecceeeeseeeeeeeeeeaueeseeeeeeeeessaeeeeeeeeesssaeaaeeeeeeesssaaaaeess 42 pda ESU U eee eee ere ne oe e as EEUU RUNE MMC PUT ENSE UU MUN ee 44 6 4 8 Conclusion on test results sss nnns 47 65 OG Gini a NOMS sisii m 47 7 Digital loss of synchronism protection ccceeccseeeceeeeeeeeneeeseeeeeueeneeeseeeeeueeaeeeseeesaes 49 Mets Model descriptos E M 49 2 INOLES ON IMPICMENTATION wcascaveas svicraanaranianra
3. e eo se ee e e 9 PEGASE COOPERATION i g L s uo e C v Ale UE B Rs Figure 16 2 EUROSTAG model of under voltage protection Date 19 08 2011 Page 104 DEL WPS5 1 Part 3 Toolbox for model ing ETN components V3 doc 5 PEGASE EA VOLTAGE POSITIVE 1 pB u ra Lr r1 ir Figure 16 4 Modelling of time delay and preset length of output signal Date 19 08 201 1 Page 105 COOPERATION DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc 5s PecAse COOPERATION Block diagram of under voltage protection in MATLAB SIMULINK environment is presented in Figure 16 5 If LI Pick up Wa Pick up setting pu Threshold 12 0 Voltage Subtract 1 lf U gt Pickup input m Analogue Hysteresis ESI Sum Switch on at 1 5 Drop out setting pu Picdeup if 2 Switch off at 1 5 Drap aut if 2 Threshold 22 0 Do nothing if 0 Subtract 2 f U gt Drop out Data Type Conversion 1 Resettable Small Integrator Transport reset is on Delav falling edge 0 004 sec B to bre ak dE algebraic Minimum Relational Data Type uis M 1 Trip Operator Conversion Z Command t e Duration Time elationa sec Resettable Delay Operator Integrator Setting sec I EIS vs falling edge Hysteresis Binary Release Binary Data Type Subtract Switch an at a 5 Conversi
4. T Loon d aon On enn Cen WO en Xeon dns n Lan i pp tal S gin zs pin NN a qnap ey pisi WO ay pipun Genel Bie i Mw adj ELE i ES 4 OO EFI sianop JHF 7 a Can zs Cain l 295 4 uanginig z wis uas s qnop oy B Cen tea gn en Loin n eon Cp n EFI zaunimAago IUjEI2U addy eeg lIEuongejes HMU aianop Ny eun zen zal a qnap lt i tapa Burg gun 2 pen EPI Wo S pase 101818 31u jusuifiac lal t Ln dt Ln 8m C zn pero zn Xen dn to rn ON aT n i Bun i ajqeyasay power system mina for two ter lon General model of differential protect 10 Figure 13 element ui g E e wm Y EE e Z uh es ut E ac a wv 9 EM z dz is As Page 89 Date 19 08 2011 PEGASE COOPERATION For the development of model of differential protection in EUROSTAG software an Above Line block could be used see Figure 13 11 that describes a general line equation A x B y C and returns binary 1 or 0 depending on the coefficient B and whether the given point lies above or below the defined line ABOVE LIME Figure 13 11 The Above Line block in EUROSTAG s model editor The part of a model that outputs binary 1 if a point in tripping characteristic lies in the tripping area see Figure 13 12 is presented in Figure 13 13 ylixl Tripping area vaca Ems xl x2 x3 Figure 13 12 Tripping characteristic that is mod
5. Zero sequence over current protection is used to protect the network against earth faults The protection is activated if the residual current 310 I1 I2 I3 rises above the setting threshold The residual current corresponds to the current flowing through earth 3 4 10 16 Negative sequence over current protection is used to protect a rotating machine motor or generator against current unbalances This unbalance may come from e the power source transformer or generator which does not supply a symmetrical three phase voltage e other consumers who do not constitute a symmetrical load causing the power supply network to have a voltage unbalance e from a two phase power supply following a blown fuse or a phase being cut e from an inversion of phases due to a connection error Both protections operate in a similar way to the phase or positive sequence over current protection Therefore the same models can be applied only with different input signals that can be easily obtained using EUROSTAG measurement blocks LURRENT LURRENT PHASE 1 1 ka Figure 12 1 EUROSTAG measurement blocks for different current signals Date 19 08 2011 Page 81 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE COOPERATION 13 Differential protection 87 ANSI IEEE C37 2 standard device number 87T transformer 87G generator 87B busbar IEC 61850 logical node name PDIF Input signals Current phase and
6. eeeeeeeseseeeeeeseeeeeeeneeeeennn nennen 22 4 5 2 Voltage set point modification ssssesssssessssssesesseeeeneeeennnnne nnns 24 Date 19 08 2011 Page 3 DEL WPS5 1 Part 3 Toolbox for model ing ETN components V3 doc s E PeGASe ee COOPERATION 4 6 EUROSTAG limitations ssssseseeeeeee eene nnne nenne nennen nnn nnn nnn niin nnn nn nnns 25 4 6 1 FU cU crc 26 4 6 2 Limited integrator block with variable liMItS ccccccsseeeeeeeeeeeeeeeeeseeeeeeeseeaeeeeeeseseaees 27 5 High Voltage Direct Current HVDC Transmission System 28 Date MOCC IO mem 28 SN NMNBCR ise T UU mU UU LM 29 S E eene ot TOT m 29 SO OVENI E 29 5 3 2 Simplified HVDC VSC Dynamic Model ssseeeeeeessseseeeennennnnneennnnnnnnnns 30 5 9 9 DC LINK MOdel EM 33 5 3 4 Sending end converter SEC model cccccccseeeeeeceeeeeesseeseeeeeeeeessueeeeeeeeeeesssaaaeees 33 5 4 Demonstration Examples ccccccccsssccccessececeessceccesececsaueeecseuseeeseaseeessageeessaeeesseaseeessneesesaeees 34 5 4 1 Test network and controller parameters eeeeeeseeeeeeeeeenerne nnn 34 5 4 2 Simulation results eeeeessssssssssssssessseeeeeeee nennen nennen nnn nnn nnne nnns nnns 35 6 Implementation of the HVDC VSC model
7. means that pickup drops out when the pickup value of approx 95 is undershot For a new pickup the time counter starts at zero Other curve shapes IEEE GE Type IAC I2t and other standard curve shapes might be also considered in addition to IEC Date 19 08 2011 Page 73 DEL WPS5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE 11 COOPERATION Directional over current protection ANSI IEEE C37 2 standard device number IEC 61850 logical node name PTOC Input signals Current phase or positive sequence Output signals Binary 1 or 0 Date 19 08 2011 Page 74 Directional over current relays are constructed using a normal over current unit plus a unit that can determine the direction of the power flow in the associated distribution system element In addition to the relay current this second unit usually requires a reference signal to measure the angle of the fault and thus determine whether or not the relay should operate Generally the reference or polarization signal is a voltage but this can also be a current input 3 The ability to differentiate between a fault in one direction or another is obtained by comparing the phase angle of the operating current phasor which varies directly with the direction of the fault and some other system parameter that is not dependent on the fault location This constant parameter is referred to as the polarizing quantity For directional relays applied to operate
8. ANSI IEEE C37 2 standard device number 81U under frequency 810 over frequency IEC 61850 logical node name PTUF under frequency ELEM Variations in the power supply frequency may be due to 3 e Overloads when the network is fed by a limited power source internal generation plant operating while cut off from the utility network e Faulty operation of a generator frequency regulator e Cutting off of a generation plant from an interconnected network e Power supply being cut off in an installation fitted with large motors The motors then feed the installation the time it takes for the flux to be extinguished with a decreasing frequency The protection in its simplest form compares the network frequency to a minimum or maximum threshold frequency Generally it includes a definite time delay Structurally models for over under frequency and voltage protection are the same therefore models for frequency protection are given in Figure 17 1 and Figure 17 2 without further explanation Resettable Small Integrator Transport resets on Delay falling edge 0 004 sec to break algebraic Minimum Relational Data Type i l Trip Operator Conversion 1 loop Frequen a Relay Command Pick up setting Curation Analogue f i i Drop outsetting po stable us iue sec Integrator Setting sec resets on AND Trip falling edge Hysteresis Binani Release Binan Data Type Subtract Switch an at 0 5 C
9. COOPERATION ins PeGASe A COOPERATION d Example 4 A step change of REC AC voltage reference from 1 0 p u to 0 7 p u Response of REC and SEC active power to Auge o7 0 3 pu Response of REC and SEC reactive power to AUpec o 0 3 pu 0 0 2 0 4 0 6 0 8 Us 1 Figure 5 13 Response of a voltage change at the receiving end Date 19 08 2011 Page 38 DEL WPS 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE COOPERATION 6 Implementation of the HVDC VSC model The HVDC VSC model was then implemented on EUROSTAG The results were then compared with those obtained using another power system simulation package known as PSD an in house simulation software used by UDE 6 1 Versions of implementation Base VSC HVDC model includes complex DC link model with capacitor charging and discharging modelling It was implemented in two versions referred as Version 1 and Version 2 In Version 1 converter reactance is modelled inside VSC HVDC model whereas in Version 2 it must be accounted in power flow file Since complex DC link model slows down overall computational process two simplified models were proposed Model A and Model B that are based on Version 2 model which is also referred henceforth as Model 0 In Model A simplified model for DC link is used In Model B in addition to this model of current controller is also simplified For Model 0 Model A Model B two alternatives ar
10. 128 No 4 pp 396 401 Jiuping Pan Reynaldo Nuqui Le Tang and Per Holmberg VSC HVDC Control and Application in Meshed AC Networks IEEE PES General meeting Pittsburgh Pennsylvania July 20 24 2008 Feltes C Erlich I amp Koch F Fault Ride Through of DFIG based Wind Farms connected to the Grid through VSC based HVDC Link 16th Power Systems Computation Conference July 2008 Glasgow Scotland Lidong Zhang and Hans Peter Nee Multivariable Feedback Design of VSC HVDC Connected to Weak AC Systems Member IEEE and Senior Member PowerTech 2009 28 June 2 July 2009 Bucharest Romania Rich Hunt Michael L Giordano Thermal Overload Protection of Power Transformers Operating Theory and Practical Experience 59th Annual Protective Relaying Conference Georgia Tech Atlanta Georgia April 27th 29th 2005 Christophe Preve Protection of Electrical Networks ISTE Ltd 2006 ISBN 10 1 905209 06 1 IEC 60255 8 Electrical Relays Thermal Electrical Relays Second Edition 1990 SIPROTEC Miulti Functional Protective Relay with Local Control 7SJ62 64 V4 7 Manual Siemens T60 Transformer Management Relay UR Series Instruction Manual Revision 5 2x GE Multilin 2007 MiCOM P139 Feeder Management and Bay Control Software Version 605 611 612 Technical Manual Areva Transformer Protection IED RET670 Technical Reference Manual ABB 2006 Blackburn J L Domin T J Protective relayin
11. 61850 logical node name PTOC Input signals Current phase or positive sequence Output signals Binary O or 1 From the modeling point of view the model of inverse time over current protection is essentially the same as for instantaneous definite time over current protection The only difference concerns the timer element 3 4 8 13 The time delay depends on the ratio between the current measured and the operating threshold The higher the current means the shorter the time delay see Figure 10 1 Operating zone Figure 10 1 Inverse time protection Inverse time protection operation is defined for instance by IEC 60255 3 This standard defines several types of inverse time protection that are distinguished by the gradient of their curves standard inverse very inverse or extremely inverse time protection 4 10 12 Trip time curves according to IEC Normal Inverse Type A Very Inverse Type B Date 19 08 2011 Page 68 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE COOPERATION Extremely Inverse Type C Long Inverse Type B For all characteristics t trip time in seconds Tp setting value of the time multiplier I fault current Ip setting value of the pick up current Pickup threshold Approximately 1 10 Ip The tripping times for I Ip gt 20 are identical with those for I Ip 20 When the current element picks up the time delay of the trip signal is calcul
12. MACHINE WF REACTIVE POWER Unit Mvar sim MACHINE WF ACTIVE POWER Unit MW sim MACHINE WEF REACTIVE POWER Unit Mvar Figure 4 4 Active reactive power following a fault Date 19 08 2011 Page 22 DEL WPS 1 Part 3 Toolbox for model ing ETN components V3 doc s PecAse Using simplified model sim VOLTAGE AT NODE NWF Unit kV M EUROS AG Using detailed model sim VOLTAGE AT NODE NWF Unit kV EE URC AC Figure 4 5 Bus voltage after a fault Using simplified model sim VOLTAGE ANGLE AT NODE NWF Unit deg H URC A z Date 19 08 2011 Page 23 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc COOPERATION M Using detailed model M EUROS AG Figure 4 6 Voltage angle after a fault This test case shows proper model operation in case of voltage drop Both active and reactive power dropped when fault begins due to converter blocking After short circuit is eliminated fast recovery of reactive power output and slow recovery of active power can be seen The same curves for detailed model simulation are shown below 4 5 2 Voltage set point modification The purpose of this test case is to verify correct modeling of current limiter Set point modification was modeled at time t 10 s As a result of this action reactive power output of the WF shall increase As current limitation is modeled the same way in detailed and simplified model only simplified model was tested w
13. Page 42 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE COOPERATION Table 9 Bus data Node name Base Generated power Generated power voltage NREC 220 NRECSYS 220 220 NRECSYSI 220 NRECSYS2 154 0 0 220 20 220 00 00 00 EE NE Lo Table 10 Branch data Sending and Base Resistance Reactance Semi shunt Semi shunt Ratio receiving node power conductance susceptance names NONE ee mes Pp ppp NRECSYS Mec P qp tq m oj mo m NRECSYSI mes Ceo ee NRECSYS2 Mus om on m m Jp S n NSECSYS scssi o o NSECSYSI NSECSYS 100 NSECSYS2 Results of steady state computation are shown in the tables below Table 11 Power flow results bus voltages Nodassne ace Voltage Voltage voltage magnitude angle Neecsys 20 1 o NRECSYSI 20 1 o NRECSYS2 20 07 9 NSECSYS 20 1 o INSECSYSI 20 1 9 NSECSYSZ 220 o7 9 Date 19 08 2011 Page 43 DEL WPS5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE COOPERATION Table 12 Power flow result line flows names 6 4 2 Test results The models was tested according to test program prepared by UDE The results was compared to those acquired from UDE modeling environment that was considered as reference The program includes for test cases change of Psgc ref from 1 0 to 0 9 p u change of Psgc ref from 1 0 to 0 5 p u change
14. Part 3 Toolbox for model ing ETN components V3 doc PEGASE COOPERATION e The current phase in relation to the polarizing voltage is in a range referred to as the tripping zone Unfortunately in pan European simulator only the positive sequence phasors are available therefore it is not possible to reproduce any of the existing direction determination approaches Let us develop a simple directional element using only positive sequence quantities The task is simplified by the fact that only three phase faults should be taken into account To begin with let us calculate the phase displacement between the positive sequence quantities as shown in equation 14 3 17 18 p Z ZU 14 In this case the forward zone of the protection see Figure 11 2 can be shown as a half plane where the active power detected by the protection is positive IA 5 IA gt cos o 20 15 Nia Nila Forward Forward o 90 Jut dF EL UE Oa ua LLLI Reverse Figure 11 2 A plane that represents forward and reverse direction of current active power One voltage and five current phasors are displayed The reverse zone of the protection on its turn is a half plane where the active power detected by the protection is negative lt o lt gt cos g lt 0 16 tla In purely resistive circuits voltage and forward current waveforms are in phase and the phase displacement is 0 In purely inductive and capacitive c
15. Region of Kurskaya NPP COOPERATION p Zheleznogorsk Russia Yuzhnaya S Kursk E I Ietalurgicheskaya S Anti emergency system of Kurskaya NPP includes two parts out of step protection at each generator and automatic unloading system The latter system is only enabled when Russian and Ukrainian power systems are not operated in parallel and therefore the grid is weakened It automatically reduces NPP total generation by dropping a generator when one of predetermined events short circuit or grid element disconnection occurs The logic that determines which generator and in what conditions it should be dropped is presented in Figure 20 3 this logic can be implemented in EUROSTAG using Macroblocks to define the conditions of generator dropping and Macroautomatons to enable the dropping via events Eurskaya HPF Vz250 kV Eurskaya MPP Wed kV for more than 4 25 3 Disconnection of 750 330 kV transformer at Eurskaya HPF Total generation of units 1 2 l p more than SUL MV 4 p more than 1100 MY etarting factor Generation of unit 2 5 p more than 7 UD MY Drop generator at unit 3 or 4 Drop generator at unit 3 or 4 D rop generator at unit 2 Total generation of the HPP 3 p more than 2500 MY i p more than 2650 IW TI0 330 KV trans former load p more than 500 MV to 330 kV bus 7 p more than 300 MV ta 750 EV bus Control action Figure 20 3 Part of a statio
16. and angle Output signals Binary 1 or 0 The loss of synchronism between power systems or a generator and the power system affects transmission line relays and systems in various ways Some relay systems such as segregated line differential relay systems will not respond to an OOS out of step condition and other relays such as over current directional over current and distance relays may respond to the variations of voltage and currents and their phase angle relationship In fact some of the above relays may even operate for stable power swings for which the system should recover and remain stable 14 25 26 30 31 Instantaneous phase over current relays will operate during OOS conditions if the line current during the swing exceeds the minimum pickup setting of the relay Likewise directional instantaneous over current relays may operate if the swing current exceeds the minimum pickup setting of the relay and the polarizing and operating signals have the proper phase relationship during the swing Time over current relays probably will not operate however it most likely will depend on the swing current magnitude and the time delay settings of the relay Phase distance relays model measure the positive sequence impedance for three phase faults The positive sequence impedance measured at a line terminal during an OOS condition varies as a function of the phase angle separation 6 between the two equivalent system source voltages Distance r
17. angle phase or positive sequence from all sides of the protected object Output signals Binary 1 or 0 There are protection functions that are simple from the point of view of general operational principles but at the same time are very complicated from the point of view of development of simplified models Let us take as an example the numerical busbar differential protection Very stringent demands regarding security availability and selectivity are imposed on the busbar protection as false operation of it can lead to very serious consequences 3 5 16 19 24 1 Security The protection may not under any circumstances trip in the event of an external fault as the loss of a busbar in general causes shut down of a large section of the system Tripping is therefore always dependent on a number of criteria 2 Availability The failure to trip or the delayed tripping of a busbar fault creates the risk of system instability and may under critical circumstances cause a system collapse due to the voltage collapse on the busbar Furthermore severe damage could be caused at the fault location due to the very high short circuit current level 3 Selectivity In multiple busbar substations only the busbar section which is affected by the fault must be cleared so that the power supply via the other busbars is ensured For this purpose a separate measuring unit for each bus section is required along with a complex image of the isolator posi
18. cb5 R gt a3 R b3 R lt a4 R b4 Rectangular A IX Xp tilted R lt Rp X tg Q R Rg X tg Date 19 08 2011 Page 92 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc 5s PecAse e O Re Ke Re COOPERATION aids JANG lt Rpl E Measured impedance Z for phase to phase e g for loop consisting of phase A and B and phase to ground e g for phase A elements are as follows 25 29 UA Upg dA Y p ud A B U Z m 35 ph gnd DE ee Where Uape and IAgc are phase voltages and currents phasors Ip is zero sequence current phasor and kg is a compensation factor calculated from positive and zero sequence line parameters by equation 36 Date 19 08 2011 Page 93 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE Date 19 08 2011 Page 94 COOPERATION Zn Z kg 36 KE UA 36 If distance protection model is based upon calculated positive sequence impedance where U1 is positive sequence voltage and I1 is positive sequence current then the measured impedance is U 37 I 37 Zi Using for instance the equation 37 and equations defining the tripping characteristic as well as some fragments from previous models we can easily create a model for distance protection see Figure 14 1 In the considered model two important features are missing 25 29 1 Power swing detection blocking element Distance protections are usua
19. gt H T Data T ue d 3 Minimum Relational Data Type SU ECL Tri Operator Conversion z to daubley1 Pu EE Relational c j d to double esettable Operator Z omman Integrator 1 Duration reset is on sec FE i A E Tri falling edge AND P Trip 5 1 Hysteresis binary Release Manuel Dueh Data Type Subtract Switch on at 0 5 Conversion 1 i E Time Signal use Switch aff at 0 5 Delay binan Setting sec EN 4 Block Manual Switch 2 Signal binary Logical Operator R Rset x U4 magnitude I1 u1 I4 wolt magnitude AND amperes degrees n T Iul ur Re mi L4 fu Sim m phase B Cplx R1 X1 degrees N phase degrees or Ltd aje Fi degrees Figure 14 1 Block diagram of distance protection When developing an EUROSTAG model of distance protection a dedicated block Inside Polygon can be applied see Figure 14 2 With this single block a complex polygon characteristic can be easily modeled INSIDE POLYGON Figure 14 2 The Inside Polygon block in EUROSTAG s model editor What remains to be done is to calculate impedance Z split it into real and imaginary parts and connect R and X signals to the Inside Polygon block Date 19 08 2011 Page 95 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE COOPERATION 15 Power swing detection ANSI IEEE C37 2 standard device number IEC 61850 logical node name RPSB Input signals Current and voltage phase
20. linked to controlled power transmission line In addition user automatons shall be defined to control outputs of the device and trigger line tripping Common framework may be observed in the example of implementation 7 3 Model parameters description Description of model parameters adjustable by user and their default values are given in the table below Table 13 Model parameters description Name Default Unit Description value DELTA Frame size for catching phase angle near 180 S i i EXTBLOCK 0 External blocking N2ST Setpoint number of cycles for the second stage i i Z i pa PHI degrees Line impedance angle SS p __ Setpoint number of cycles for the third stage NI pu 7 4 Model testing 7 4 1 Sample power grid The model was tested in a simple scheme shown at the figure NHV1 NHV2 NGEN SYS GEN amp e a pot SYS lt NHV1 The device is supposed to be installed at NHV 1 and control NHV1 NHV2 2 line Date 19 08 2011 Page 50 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc 33 PeGASe e e COOPERATION Table 14 Input node data Node name Base Soo Pone voltage Non x 9 067 0 0 NHVI 30 0 0 9 0 nav s 0 9 9 0 Ys 380 49i 1i 0 0 Table 15 Input branch data Sending and Base Resistance Reactance Semi shunt Semi shunt Ratio receiving node power conductance susceptance names Results of steady
21. of Vsgc from 1 0 to 0 7 p u change of Vggc from 1 0 to 0 7 p u For each test active power of SEC and REC reactive power of SEC and REC and DC voltage for Model O version were measured and compared to the reference Simulation output from EUROSTAG is shown together to that obtained from UDE on the same diagrams All diagrams are in per units over time in seconds Tests show that the model developed produce results close to the reference in most cases with some exceptions described below 1 In Test 1 small difference in REC reactive power and DC voltage may be observed q REC EUROSTAG q REC UDE q SEC EUROSTAG q SEC UDE Date 19 08 2011 Page 44 DEL WPS 1 Part 3 Toolbox for model ing ETN components V3 doc iz pecase Gee COOPERATION _ u DC EUROSTAG u DC UDE 0 40 0 60 0 80 Figure 6 4 Test 1 results for Model 0 there is small difference in results This is probably caused by lag blocks that were added to EUROSTAG model to avoid algebraic loops The difference is insignificant in absolute value and may be notice only when parameter measured changes very slightly As seen from the figures below when chnageo of reactive power or DC voltage is higher the difference is almost unnoticable q REC EUROSTAG q REC UDE q SEC EUROSTAG q SEC UDE
22. or line during off line contingency studies In this case the fault can be simulated by simply switching off the faulted power system element This approach will be far more reliable than reliance on heavily simplified model Figure 13 3 Imitation of operation of line differential protection Date 19 08 2011 Page 84 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE COOPERATION Figure 13 4 Imitation of operation of generator differential protection Figure 13 5 Imitation of operation of transformer differential protection For off line trainings a very simple and general differential protection element can turn out useful for demonstrative purposes In this case the main point of model would be to show the differential principle and typical pick up characteristic Let us develop a simple and general model for differential protection To begin with the tripping characteristic should be chosen for the model 0 2 4 6 3 Figure 13 6 Typical three section tripping characteristic with differential current on y axis and restraint or bias current on x axis Tripping area is above the characteristic DEL WPS 1 Part 3 Toolbox for model ing ETN components V3 doc Date 19 08 2011 Page 85 PEGASE Percent Differential New Site 1 New Device 1 Setti EG SETTING PARAMETER Operating Characteristc Graph View Enabled COOPERATION Operation Characteristic Graph New Site 1 New Device 1 Setti
23. power system Stable Power Swing a power swing is considered stable if the generators do not slip poles and the system reaches a new state of equilibrium i e an acceptable operating condition DEL WPS5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE Date 19 08 2011 Page 99 COOPERATION Unstable Power Swing a power swing that will result in a generator or group of generators experiencing pole slipping for which some corrective action must be taken Out of Step Condition Same as an unstable power swing Electrical System Center or Voltage Zero it is the point or points in the system where the voltage becomes zero during an unstable power swing In this chapter we will develop a two blinder model for power swing and out of step mode detection The two blinder scheme shown in Figure 15 4 is based on the same principle of measuring the time needed for an impedance vector to travel certain delta impedance The time measurement starts when the impedance vector crosses the outer blinder RRO and stops when the inner blinder RRI is crossed If the measured time is above the setting for delta time a power swing situation is detected If the blinders are set in parallel to the line impedance then they are optimized for the delta impedance measurement because the power swing impedance vectors will normally enter the protection zones at an angle of nearly 90 degrees to the line angle Depending on certain network condition
24. resistance of the chopper if and when it is activated and a block indicating that the corresponding block becomes active only when the DC voltage increases beyond a pre specified threshold REC active power REC SEC Losses SEC active power Figure 5 8 DC link model 5 3 4 Sending end converter SEC model The SEC is responsible for transmitting the active power injected in to the sending end of the transmission system by maintaining the AC voltage set point The sending end current controller and connection to the grid is similar to the REC model Active current reference is calculated from the desired power active power to be transmitted through the HVDC line The reactive current control loop can be used to control the SEC terminal voltage on the AC side or to guarantee a constant power factor injection to the AC grid Again this is similar to the REC side controller Date 19 08 2011 Page 33 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc 33 PeGASe 9 ha COOPERATION Active Current Reference Calculation P SEC ref Vgc PQ Priority QsEC O A Power Factor Controller SEC l l poe l Magnitude SEC ref gk en quee NM P Y Limitation jQ SEC rf 7 1 mem Sere n gt G 1 r d gt ISEC max SEC ref aya sT l ge e RLI II GN a Below 0 9 pu TSEC max 9 voltage reactive QsEC SEC current priority VSEC 0 REC Terminal Voltage Controller VSEC ref Gv y Figure 5 9 Send
25. seconds Although pickup of the relay elements is based mostly only on the fundamental harmonic component of the measured currents false device pickup due to inrush is still a potential problem since depending on the transformer size and design the inrush current also comprises a large fundamental harmonic component For that reason relays often feature an integrated inrush restraint function It prevents the pickup of the over current elements if inrush conditions are present Inrush current contains a relatively large second harmonic component twice the nominal frequency which is nearly absent during a fault current The inrush restraint therefore is based on the evaluation of the 2nd harmonic present in the inrush current Digital filters are used to conduct a Fourier analysis of all three phase currents to detect the magnitude of 2nd harmonic Cross blocking Since the harmonic restraint operates individually per phase the protection is fully operative even when the transformer is switched onto a single phase fault whereby inrush currents may possibly be present in one of the healthy phases It is however possible to set the protection in a way that when the permissible harmonic content in the current of only one single phase is exceeded not only the phase with the inrush current but also the remaining phases and over current elements in them are blocked DEL WPS 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE Da
26. starting current sometimes is acquired and the resulting increase of temperature rise is suppressed This means that the calculated temperature rise is kept constant as long as the starting current is detected Emergency start up of electrical machines When machines must be started for emergency reasons operating temperatures above the maximum permissible operating temperatures sometimes are allowed emergency start Then exclusively the tripping signal can be blocked via a binary input Since the thermal profile may have exceeded the tripping temperature after startup and dropout of the binary input has taken place the protection function features a programmable run on time interval which is started when the binary input drops out and continues suppressing a trip signal Tripping by the overload protection is defeated until this time interval elapses This binary input affects only the tripping signal It has no effect on the fault condition logging nor does it reset the thermal profile Influence of the negative sequence current To determine the heat rise of rotating machines the relay can take into account the negative sequence component Indeed the rotating field corresponding to the negative sequence component induces a double frequency rotor current which causes considerable losses Behavior in the case of input signal failure The situation with a loss of input signal actual current temperature etc is considered here Usually dependi
27. state computation are shown in the tables below Table 16 Node voltages Node Base Voltage Voltage some voltage magnitude angle NAVI sys 30 1 09 Table 17 Power flow names NGEN NHVI 600 0 130 7 Test results Below test results are shown Each stage was tested separately with other stages disabled by setting corresponding automaton Threshold 1 value to 10 Test procedure consists of modeling three phase fault at NHV1 node that lasts from 10 0 to 10 7 s This causes generator GENI loss of synchronism For the time of short circuit condition the device is blocked by setting corresponding setpoint to 1 Then the respective device stage causes line tripping during first second or fifth cycle of asynchronous operation Date 19 08 2011 Page 51 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc SZ PeGASe e f d t B Test for stage 1 sim Line phase angle ui FU R L4 y fA 3 Figure 7 1 Test for stage 1 sim MACRO AUT ALARMB VARIABLE ALARI sim MACRO AUT ALARMB VARIABLE SLIPDN T sim MACRO AUT ALARMB VARIABLE SLIPUP IN E a ROS IZ G Figure 7 2 Test for stage 2 sim Line phase angle lll FLU R g S y fA G sim MACRO AUT ALARMB VARIABLE ALAR sim MACRO AUT ALARMB VARIABLE SLIPDN r sim MACRO AUT ALARMB VARIABLE SLIPUP IE IROS Z G sim Line phase angle i d FU R OS y fA 3 Date 19 08 2011 Page 52 DEL WP5 1 Part 3 Toolbox for mod
28. time systems ABS VALUE BIGSOM f COSBLK f Hal Lookup Tables j a Event handling Laplace j Mathematical Operations i Matrix aru P 1e Iu gt Electrical al j Integer eret dr F Boeg E 2 a Port amp Subsystem EXPBLK m GAIMBLK INVBLK E j a Zero crossing detection zz vise Signal Routing es j in Signal Processing a p 4 Implicit loc FP sans TERN aa Annotation 5 bers BIET MATMAGPHI 4 Sinks LOGBLK f MATZREIM j I Sources s l Thermo Hydraulics M Demonstrations Blocks 1 User Defined Functions a a MIN Max FF 1 MAX MIN ra MAXMIN l MAX f MIN f E 8 I a A 4 u a gt M X P T 7 F E Figure 20 1 Library of standard elements in SCILAB XCOS and EUROSTAG software Let us take as an example the SPS of Kurskaya nuclear power plant in Russia Kurskaya NPP is situated in the southern part of Center UPS near Ukrainian border Its total capacity is equal to 4000 MW that is allocated in 4 generating units Kurskaya NPP is connected to the power system by three 750 kV overhead lines and several 330 KV lines A simplified plot of the surrounding power grid is shown in Figure 20 2 Date 19 08 2011 Page 114 DEL WPS 1 Part 3 Toolbox for model ing ETN components V3 doc mas PeGASe MH ovobryanskaya r5 IY y Kurskaya MPP Ukraine ol Powerplant 750 kV line 3 Substation 330 kV line Figure 20 2
29. 0225 0 023 0 0235 0 024 0 0245 0 025 0 0255 Response of DC link voltage to Ap ret 0 1 pu 1 005 pu 1 0 995 Uu DC 0 99 0 985 0 98 0 975 0 97 0 965 0 0 2 0 4 0 6 08 t s 1 Figure 5 10 Response to a SEC power step change Date 19 08 201 1 Page 35 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE COOPERATION b Example 2 A step change of SEC reference power from 1 0 p u to 0 5 p u Response of REC and SEC active power to Ap 0 5 pu Response of REC and SEC reactive power to Ap 0 5 pu 0 0 0 2 0 4 0 6 og ts pu i 0 01 1 M MM 0 015 40 41 0 025 0 03 Response of DC link voltage to Ap re 0 5 pu 1 04 E XUu DC 0 96 0 92 0 88 0 84 0 8 Figure 5 11 A step change involving 50 SEC power Date 19 08 2011 Page 36 DEL WPS 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE 1 5 pu 0 5 0 5 1 5 Response of REC and SEC active power to Au o 0 3 pu p REC p_SEC 0 2 0 4 Response of DC link voltage to Ap 0 1 pu Example 3 A step change of SEC AC voltage reference from 1 0 p u to 0 7 p u Response of REC and SEC reactive power to Au o 0 3 pu Figure 5 12 Voltage step change at the sending end Date 19 08 201 1 Page 37 DEL WPS 1 Part 3 Toolbox for model ing ETN components V3 doc
30. 7 Graphical explanation of different reset times of voltage and current vectors is presented in Figure 11 6 DEL_WP5 1 Part 3 Toolbox for model Date 19 08 2011 Page 77 ing ETN components V3 doc COOPERATION Different reset times 160 170 Figure 11 6 Graphical explanation of different reset times of voltage and current vectors S00 a00 ae tA1 ang A1y 100 a J 4 Figure 11 7 Possible problem with angle difference calculation due to different reset times of current and voltage To avoid this the angle difference can be calculated as shown in equation 20 p ZI Z Uj if Z4 ZUi lt T p ZH Zu 2 n if 41 ZU1 gt 20 p Z1 24U 2 1 if 21 ZU lt 2 The block diagram of suggested directional determination unit is presented in Figure 11 8 It is modeled using equations 18 19 and 20 Date 19 08 2011 Page 78 DEL_WP5 1_Part_3_Toolbox_for_model ing_ETN_components_V3 doc PEGASE COOPERATION u1 Phase degrees 11 Phase degrees Relational Switch Operator Threshold 0 5 t Criteria u Z Threshold SHIFT degrees AND Forward Fund Logical Operator F onward Detection Ground for unused inputs Reverse Rew Logical Operator 180 Reverse Detection Figure 11 8 Block diagram of direction determination unit As with inverse time over current protection it seems it is better to develop a dedicated block for direction determina
31. Deliverable D5 1 Modeling requirements for the ETN Part 3 Toolbox for modeling ETIN components Proprietary Rights Statement This document contains information which is proprietary to the PEGASE Consortium Neither this document nor the information contained herein shall be used duplicated or communicated by any means to any third party in whole or in parts except with prior written consent of the PEGASE Consortium Grant Agreement Number 211407 implemented as Large scale Integrating Project Coordinator Tractebel Engineering S A Project Website http www fp7 pegase eu e o t e e9 Se o 26 s PEGASO eee COOPERATION o Document Information Document Name D5 1 Part 3 Toolbox for the modeling the ETN components ID DEL WPS 1 Part 3 Toolbox for modeling ETN components V3 doc WP 5 Task 5 3 subtasks 5 3 1 5 3 2 5 3 3 5 3 4 Revision 1 Revision Date 30 06 2011 Author UDE RTU amp ESP Diffusion list STB SSB and Coordinator Approvals Final version to be submitted to the European Commission SS compa Date visa O O RTU Wee Antans Sauhats Author Vladimir Chuvychin RTU Yee Vadims Strelkovs T RTU red Fekadu Shewarega UDE Dedi Istvan Erlich UDE Auo Alexander Rubtsov ESP Task Leader RTU WP Leader UDE Documents history 1 01 05 2011 First draft version RTU UDE TE amp ESP Contribution to the report Contribution to the report Final
32. N_components DEL WP5 1 Part 3 Toolbox for model ing So 1E wo yous d Burgess puana si z n plousem se zin ewe E suiquioa 01 dj Burgess iagram o GOOLE un gal JPUBIS 3n dui Juan 51 Ln d x LL pjoaussiu L AL 1ajdijnpy eui g ug ui sIs818 5AH a aie 382 ual E 2 E 3 Aja au p 211m pe du l Euang ag i FH n OE sBpa uE xy d Buas un sjasal jusunz F 10 Ee1 a u uojeaua 19 EI2U a qeyeasey P adiy eje IEuonEIaH buas jno deig ldonl d x p H Buas dnapig ajgiqaB E siqnop feyay indu HE ag op 2az T pc A 222 zon apa Gule uanginig dr AE ap niis o Jasal puru uodsugn 1ogg1B 4u du JEWS sa aad LU n uu ry MATLAB SIMULINK block d delay characteristic Figure 10 2 Although the equations for inverse time characteristics are relatively simple the resulting model is not if itis developed using standard mathematical blocks and transfer functions Page 70 Date 19 08 2011 PeGASO a COOPERATION In Figure 10 3 EUROSTAG and Figure 10 4 SIMULINK a sample model is presented that includes one equation for inverse time characteristic without logical protection against division by zero and negative or extremely large calculated tripping times Ls 0 14 ES X Lou 0 05 a Figure 10 3 Equation for inverse time characteristic using EUROSTAG ai n Constant 1 Dot Product Time Multiplier Setting Tp Curent Input 7 Divide A M ath F Current Functia
33. RG op ass ZoD aBpa Buil pid doses 101E18 dy D PLUMA Ied oT EUS ajqepasay s qnop op SUCHE Ho uou zZ Woe uo SOPE Uo yap adA EjE Me uia Sisal aJ5 H HF Gas WOE ajqnop op puewwog uo iaaun s ZIoOpEIad du ad Eje lIEuongejeg nun doo 4ESIq op Oa 200 afipa Guryyey AE Bp un S pase 101815 31u s qggass podsugmn E uS Burgess WE f sBEIS Wee Jo UONEA SE Sp 10 WOME AE E Yyopinng JEMUE n QUE utrq e qon EuBis JE TUE py ng e jug uq gonna EuBIS E rniuE p BEEaS sH H or J epour ay opui pu 104g12d j di rn nr ua 4 juEgsuoo FUCIE 2H i i IBISUE aul uanoun 4 1E pl t F juanna g qissiunuasd uinumzeg py 10 Guan PEO png 314 x Guess Juang snBoa EuEg ndu yuanung apg E lon f thermal overload protect iagram o MATLAB SIMLINK block di Figure 8 5 left out of the model in that were order to fit it better into the framework of PEGASE project are discussed below 4 5 6 7 lon features of thermal overload protect 1ca Some typ V3 doc ETN_components DEL_WP5 1 Part 3 Toolbox for model Page 59 Date 19 08 201 1 ing PEGASE Date 19 08 2011 Page 60 COOPERATION Ambient temperature consideration Some manufacturers modify the equation 10 to account the effect of variations in ambient temperature That requires additional input signal for the thermal replica model Also the rated temper
34. UROSTAG software VOLTAGE POSITIVE 1 kV EV I I Lr r1 2H 4 r1 SL rm ag Figure 19 2 EUROSTAG model of over P Q S protection Date 19 08 2011 Page 111 DEL WPS 1 Part 3 Toolbox for model ing ETN components V3 doc a e e eo se ee e e 9 PEGASE COOPERATION _ i g s uo e Ale UE B Rs Figure 19 3 EUROSTAG model of under P Q S protection Date 19 08 2011 Page 112 DEL WPS 1 Part 3 Toolbox for model ing ETN components V3 doc 5 P GASe e COOPERATION These general models can be used also as a part of larger protection or automation systems taking advantage of the existence of blocking and releasing binary inputs Date 19 08 2011 Page 113 DEL WPS5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE COOPERATION 20 Special Protection Schemes SPS Due to great variety of special protection schemes and wide area protection and automation systems authors consider that it is impossible to create ready to use models It is better to have a library of standard elements see Figure 20 1 as well as standard protection and automation models so that end user can create any complex system he wants x Palettes LE Palette browser cos E Palettes E Commonly Used Blocks eel gt E i Continuous time systems E LE Discontinuities i vee Discrete
35. acteristic takes about half a period a change in 6 of 180 the impedance change occurs in about 0 5 seconds When approaches 180 during an OOS the measured impedance falls into the operating characteristic of a distance relay for a particular transmission line The impedance measurement by itself cannot be used to distinguish an OOS condition from a phase fault The fundamental method for discriminating between faults and power swings is to track the rate of change of measured apparent impedance DEL WPS 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE Date 19 08 2011 Page 98 COOPERATION Fault impedance Power Swing PD dR k n in lt Fault Inception E a Nx dX k n MX e pe e n Figure 15 2 Impedance vector during power swing and fault 29 The difference in the rate of change of the impedance has been traditionally used to detect an OOS condition and then block the operation of distance protection elements before the impedance enters the protective relay operating characteristics Actual implementation of measuring the impedance rate of change is normally performed though the use of two impedance measurement elements together with a timing device If the measured impedance stays between the two impedance measurement elements for a predetermined time then an OOS is declared and an out of step blocking signal is issued to block the distance relay element operation Impedance me
36. acturers and technologies While the manufacturer specific models together with reliable parameters enable the simulation of individual units as well as wind farms the effort required to put together such a large data set makes this approach impractical For preliminary system studies or estimating grid code compliance therefore the current trend is directed at the use of open source generic models with parameters that can be calibrated to conform to any given technology or topology The generic model is derived with the quasi stationary model as the basis by making the following simplifications The term rotor flux feedback through Aq is neglected The block containing the slip s a factor is not considered The rotor current in per unit is assumed to be the stator current magnetization current is neglected Theline side converter is not considered explicitly simplification of the quasi stationary model in accordance with the above list results in a generic model shown in As stated above the LSC model is not included Also the air gap torque of the machine is assumed to remain constant and thus the equation of motion is not considered As a result of the assumptions and simplifications direct calculation of the model parameters is not possible One possible way to determine the parameters is comparing the simulation results based on the simplified model with measurement results and employing a heuristic optimization to determine paramet
37. age and shutdown of major portions of the power system Uncontrolled tripping of circuit breakers during an OOS condition could cause equipment damage and pose a safety concern for utility personnel Therefore a controlled tripping of certain power system elements is necessary in order to prevent equipment damage and widespread power outages and minimize the effects of the disturbance Out of step detection methods and types of schemes Out of step protection functions detect stable power swings and out of step conditions by using the fact that the voltage current variation during a power swing is gradual while it is virtually a step change during a fault Both faults and power swings may cause the measured apparent positive sequence impedance to enter into the operating characteristic of a distance relay element A short circuit is an electromagnetic transient process with a short time constant The apparent impedance moves from the pre fault value to a fault value in a very short time a few milliseconds On the other hand a power swing is an electromechanical transient process with a time constant much longer than that of a fault The rate of change of the positive sequence impedance is much slower during a power swing or OOS condition than during a fault and it depends on the slip frequency of the OOS For example if the frequency of the electromechanical oscillation is about 1 Hz and the impedance excursion required to penetrate the relay char
38. asurement elements with different shapes have been used over the time These shapes include double blinders concentric polygons and concentric circles as shown in Figure 15 3 31 Outer Z Element Outer Z Element Distance Element pR Outer Z Element Inner Z Element Inner Z Element Inner Z Element Load Region a b c Figure 15 3 Different double blinder OOS characteristics To guarantee that there is enough time to carry out blocking of the distance elements after an OOS is detected the inner impedance measurement element of the OOS detection logic must be placed outside the largest distance protection region that is to be blocked The outer impedance measurement element for the OOS detection has to be placed away from the load region to prevent inadvertent OSB logic operation caused by heavy loads These relationships among the impedance measurement elements are shown in Figure 15 3 b using concentric polygons as OOS detection elements Below some definitions are given for clarity Power Swing a variation in three phase power flow which occurs when the generator rotor angles are advancing or retarding relative to each other in response to changes in load magnitude and direction line switching loss of generation faults and other system disturbances Pole Slip a condition whereby a generator or group of generators terminal voltage angles or phases go past 180 degrees with respect to the rest of the connected
39. ated using an integrated measurement process The calculated time delay is dependent on the actual fault current flowing and the selected tripping characteristics Once the time delay elapses a trip signal is issued Let us take as an example IEC Normal Inverse Type A characteristic The respective model in MATLAB SIMULINK environment is presented in Figure 10 2 The attention should be drawn to the fact that the pick up threshold in most relays is 1 1 times the current setting Ip in order to avoid division by zero that can take place when I Ip see the equations for inverse time characteristics as well as very large calculated tripping times when I is just slightly larger than Ip In the current model tripping time calculation is switched on see switch element in the block diagram as soon as relay picks up I 2 1 1 Ip To avoid negative calculated time values when the relay has not yet picked up or division by zero time delay calculation element is switched off if I Ip Date 19 08 2011 Page 69 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc COOPERATION 1o01E128d j Fa1B a7 CUE UIQ zZ qois EuBis JENUE py xag 204 089 lt i Jug ur l Jg EuBis Z ugang JENUE p SsEa 2 adA ELEC alay pays aq UED a suangnbsa JURY i d Jad iyo awg zi Cen n EuBis 120007 zn jguBis appl py Lan EuBis addy Gopan Foul s EuBiIS ime t ion with inverse f O C protect V3 doc ET
40. ational environment The changes include the following 1 Model inputs are supplied directly with voltage active and reactive power in p u 2 Model outputs provide active and reactive current instead of power This modification was made to improve model performance in EUROSTAG 3 Simple lag with time constant 7 0 001 s is applied to model outputs to avoid algebraic loops in EUROSTAG 4 2 Implementation of specific parts of the model 4 2 1 Simple lag with variable time constant This block used in converter blocking model was represented as simple lag with feedback loop with variable gain K Figure 4 1 Figure 4 1 Simple lag with variable time constant A Current conversion Current conversion from terminal voltage reference frame Ip Ig to synchronous reference frame Le lim is performed according to the formulas below Iye lg cos p t Io sing lim Ip sing Ig cos qp where terminal voltage angle 4 2 3 PI block with limitation The PI block shown in the detailed model as the figure below is modeled using the scheme given in Figure 4 2 Date 19 08 2011 Page 18 DEL WPS5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE 4 3 Date 19 08 2011 Page 19 Figure 4 2 PI block with limits Model parameters description Description of model parameters adjustable by user and their default values are given in the table below Table 1 Acronyms and default values of model parameters Name Default Unit D
41. ature rise needs to be known Different thermal time constants Other possible adjustment for equation 10 involves the use of two different thermal time constants It can be essential for transformers to account for different cooling methods The change of the transformer thermal time constants can be activated by the cooling equipment via a binary input It is also important for self ventilated electrical machines In a self ventilated machine the thermal time constant at standstill can differ considerably from the time constant of a continually running machine since then the ventilation provides for cooling whereas at standstill only natural convection takes place Machine standstill is detected and the thermal time constants are swapped when the current undershoots some preset threshold value Current limitation In order that overload protection on occurrence of high short circuit currents does not cause short trip times possibly affecting time grading of different protection functions in some relays it is possible to implement current limiting for the overload protection Currents exceeding the specified value are limited to this value For this reason they do not further reduce trip time in the thermal memory Start up of electrical machines On startup of electrical machines the over temperature calculated by the thermal replica may exceed the alarm over temperature or even the trip over temperature In order to avoid an alarm or trip the
42. cAse COOPERATION 3 4 The quasi stationary QS model of the DFIG The quasi stationary mode underlies the assumption that in the synchronously rotating reference frame the transformer voltage in the stator winding can be neglected against the much greater speed voltage 1 e ZK ad jop dt 8 Using a reference frame rotating at angular speed corresponding to the grid frequency o the quasi stationary machine model can then be given in Figure 3 2 and also in block diagram form in lower part Stator voltage equation 3rd order DFIG model dy r j Rd R yy e 2 co Vs Krhri Gee dt I dy r oy a 03 V ra Ty karsls Tn dt ly do dt E I v Rd Ls Y Ral sd KK Pitch control power conversion model Figure 3 2 The quasi stationary mathematical model of the DFIM 3 5 Rotor side converter RSC model The RSC control in a DFIG is responsible for the control of active and reactive power The active power reference is derived from the tracking characteristic of the turbine with the objective of adjusting the generator speed for maximum power generation corresponding to a given wind speed The outer power control loop of the RSC provides the set values for the inner rotor current loop After setting the derivative term in the voltage equation to zero and some re arrangement we have umrin tj g BP i 9 JXs Date 19 08 2011 Page 11 DEL WPS 1 Part 3 Toolbox for
43. dels to represent these systems and devices Another set of activities involves the development of models for complex digital and distributed control systems and special protection schemes that have the capability to affect the operation of a wide area of the interconnected system Because these devices enable a faster control of the system an accurate representation of their behaviour on different time scales is necessary The intermediate reports submitted so far deal with aspects of the modelling of components of the ETN for different purposes such as state estimation security assessment dispatcher training etc In addition to power electronic based modern equipment and systems the reports included discussions of protection and automation systems The reports also dealt with the definition of accuracy requirements for mathematical models of power system equipment on the one hand and methodologies for the validation of those models to ascertain whether these requirements are met on the other In addition to facilitate the sharing of models of the various plant and apparatus used in different countries and control areas of the ETN a common platform for the exchange and the development and improvement of those models is one of the tasks included in this work package This report building on the results of the previous reports and outputs of other work packages introduces the general framework for deriving simplified models for wind turbines volta
44. ditional power system components Included in this consideration are also complex digital and distributed control systems and system protection schemes that have the capability to affect the operation of a wide area of the interconnected system Because these devices enable a faster control of the system an accurate representation of their behaviour on different time scales is necessary for reliable system operation Previous reports in this work package dealt with modelling and validation of power system components and systems in general with particular emphasis on power electronic based electronic interfaces FACTS devices and digital controllers Many of the models that are currently in use have been evaluated and the existing modelling gaps elaborated Furthermore the problems associated with lack of access to manufacturer specific models and their control algorithms for many modern equipment such as wind turbines HVDC VSC etc due to product protection reasons have been highlighted This report building on the results of the previous reports and outputs of other work packages introduces the general framework for deriving simplified models which do not necessarily rely on manufacturer or technology specific data for wind turbines voltage source converter based HVDC transmission systems and several protection and automation systems Additionally the models have been implemented on different computation platforms and results of sample comp
45. during a fault it is important to remember two facts 3 4 10 18 1 The system voltage will collapse at the point of the short circuit Therefore the polarizing voltage must not include the faulted phase 2 The fault power factor is low i e the current lags the voltage by nearly 90 The choice of connections to obtain correct directional discrimination for unbalanced faults usually is restricted to those shown in Figure 11 1 It is possible to determine the best connection of the three for any given set of system and fault conditions by analyzing the phasors within the relay for the most probable conditions of load angles faults and the effect of arc resistance If any of these conditions change the preferred connection will also change Since it is not reasonable to account for all of the possible variations it is necessary to select the best compromise for the widest range of possibilities Of the three shown the 90 connection is usually preferred Figure 11 1 Conventional connections for directional phase relays It is said that the protection detects the direction of the current in reality it detects the sign of the active power Thus the phase displacement between the voltage and the short circuit current must be known Directional phase over current protection is activated if the following two conditions apply to a time equal to the time delay chosen e The current is higher than the setting threshold DEL WPS 1
46. e available in Alternative 1 current controller model is computed in grid synchronous frame and in Alternative 2 it is computed in terminal voltage frame In sense of implementation in EUROSTAG these versions may differ in number of additional time constant blocks added to avoid algebraic loops The table below shows summary of VSC HVDC implementations that are available for EUROSTAG Table 7 Alternative simulations 6 2 Implementation of specific parts of the model 6 2 1 PQ priority block PQ priority blocks are used in both sending SEC and receiving REC end converter models The implementation of this block with selection of either active or reactive current priority depending on voltage level VSEC is shown in Figure 6 1 Date 19 08 2011 Page 39 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE 6 Ip ref ISECMAX ISECMAX COOPERATION VPO Iq GIOSECRE Figure 6 1 Implementation of PQ priority blocks 6 2 2 Current conversion Current conversion from terminal voltage reference frame Ip Ig to synchronous reference frame Le lim is performed according to the formulas below Ipe lp coso t I5sing lim Ip sing Ig cos o where terminal voltage angle 6 2 3 Hysteresis block DC link model requires a hysteresis block to model chopper resistance switching Unfortunately built in Schmidt trigger block that may be found in EUROSTAG causes abnormal
47. ed The last two sections of the report are devoted to the modelling of protection and automation devices Due to great variety of protection systems in general and special protection schemes and wide area protection and automation systems in particular it was found to be impossible to create ready to use models As a result a library of standard elements as well as standard protection and automation models have been developed which can be put together or adjusted by the end user to suit any practical application DEL WPS 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE is 26 COOPERATION 3 Simplified wind farm wind turbine model 3 1 Introduction As the level of installed wind power generation grows its impact on power system dynamic behaviour is likely to become increasingly significant For performing transient stability studies and quantifying the effect of wind farms on the overall performance indices of the system suitable models of wind turbines and their aggregate models are required So far about 25 vendor specific models have been developed 1 However the use of vendor specific models that contain proprietary information usually requires prior manufacturer authorization and therefore such models are not openly available One possible solution is the development of publicly available generic wind turbine models that can be parametrically adjusted to any practical implementation This would also improve o
48. ed voltages and currents Furthermore the relative magnitude of the protected line and the equivalent system source impedances is another important factor in the performance of distance relays during power swings If the line positive sequence impedance is large when compared with the system impedances the distance relay elements may not only operate during unstable swings but may also operate during swings from which the power system may recover and remain stable 31 Out of step protection philosophy The performance of protective relays that monitor power flows voltages and currents may respond to variations in system voltages and currents and cause tripping of additional equipment thereby weakening the system and possibly leading to cascading outages and the shutdown of major portions of the power system Protective relays prone to respond to stable or unstable power swings and cause unwanted tripping of transmission lines or other power system elements include over current directional over current under voltage distance and directional comparison systems The philosophy of out of step relaying is simple and straightforward avoid tripping of any power system element during stable swings Protect the power system during unstable or out of step conditions When two areas of a power system or two interconnected systems lose synchronism the areas must be separated from each other quickly and automatically in order to avoid equipment dam
49. eeeeeeeeeeeeeeeeas 108 18 Under frequency load shedding c ccccccecccsseecseeeceeeecsueesseecsueecsueesseeessesesseeeeas 110 19 General DFotecllon Mode lesene EEEE 111 20 Special Protection Schemes SPS cccccceecccescecseeeceeeeceececeueeseueeseueesseeesaneeseees 114 MEM c MNNETOTTTu 116 Date 19 08 2011 Page 5 DEL WPs 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE COOPERATION L Executive Summary Modelling usually involves a trade off between the level of detail in the modelling process on the one hand and the capability of the model in reproducing the physical reality and thus the accuracy of the simulation results on the other To strike an optimal balance between complexity of the model and the efficiency of the simulation tools therefore the mathematical model of the system and its components must remain as simple as possible yet detailed enough to capture the underlying physical phenomena As the number of power electronic converters connected to the power system has increased rapidly over the recent past it was deemed necessary within the context of PEGASE project to assess the accuracy and robustness of the models commonly used to represent these devices This was prompted by serious concerns expressed in the professional literature to this effect especially when one considers their overload capability is typically much more limited than the tra
50. el ing ETN components V3 doc PEGASE ic a8 COOPERATION 0 I I I S 0 3 10 15 20 23 30 sim MACRO AUT ALARMB VARIABLE ALARI sim MACRO AUT ALARMB VARIABLE ALAR2 sim MACRO AUT ALARMB VARIABLE ALAR3 sim MACRO AUT ALARMB VARIABLE SLIPDN sim MACROAUT ALARMB VARIABLE SLIPUP M EUROS TAG Figure 7 3 Test for stage 3 7 4 3 Conclusion on test results Test results show that all parts of the model works as desired Stage 1 output signal is set during first cycle of asynchronous operation stage 2 and stage 3 signals are set after 2 and 5 cycles respectively The model is reset to initial stage 2 second after last detected cycle As seen from diagrams system is automatically split into two parts after corresponding signal becomes active This shows correct operation of the complete system 8 Thermal overload protection Overloading transmission lines cables transformers generators motors etc beyond the nameplate rating can cause a rise of temperature of these devices above permissible level As a result the insulation will deteriorate resulting in accelerated loss of life of the equipment The thermal overload protection is used to avoid it 1 2 ANSI 49 IEC PTTR thermal overload function is available in modern numerical protection relays This function depending on the protected object and specific implementation in the relay uses some c
51. el T 1 ms for close G 1 0 Ti 0 005s T 0 03s T 0 5 3 s for open eer tan l sT G 1 0 Ti 0 005s T 0 03s t g terminal voltage angle Figure 3 8 Simplified model including converter blocking option 3 10 Accuracy of the generic model in comparison with the QSS model The purpose of a generic model is to keep the model of the wind turbine as simple as possible by keeping the accompanying loss in accuracy for preliminary system studies or estimating grid code compliance within acceptable limits The generic model will inevitably involve simplification compared to the detailed model The only thing that needs to be ascertained is that these simplifications should not stunt the model to the extent that the underlying physical phenomena are no longer visible Date 19 08 2011 Page 16 DEL WPS5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE COOPERATION Response of Active Current Voltage drop 5076 detailed O55 model GE 3 1 model o 0 1 0 2 0 3 oa tfs os Figure 3 9 Effect of different levels of simplification on the simulation result Although a general statement with regard to loss of accuracy as a result of the use of the generic model is not possible Figure 3 9 exemplifies the nature of the error one commits in using the generic model The blue curve describes the behavior of active current following a grid fault leading to a voltage dip to 50 of the pre fault value when the switch in F
52. elay elements will operate during a power swing stable or unstable if the swing locus enters the distance relay operating characteristic Zone 1 distance relay elements with no intentional time delay will be the most prone to operate during a power swing Also very prone to operate during swings are Zone 2 distance relay elements used in pilot relaying systems for example blocking or permissive type relay systems Backup zone step distance relay elements will not typically operate during a swing depending on their time delay setting and the time it takes for the swing impedance locus to traverse through the relay characteristic Figure 15 1 shows the operation of a Zone 1 distance relay when the swing locus goes through its operating characteristic 31 jx Date 19 08 2011 Page 96 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE Date 19 08 2011 Page 97 COOPERATION Figure 15 1 Zone 1 distance relay characteristic It is important to recognize that the relationship between the distance relay polarizing memory and the measured voltages and currents plays the most critical role in whether a distance relay will operate during a power swing Another important factor in modern microprocessor type distance relays is whether the distance relay has a frequency tracking algorithm to track system frequency Relays without frequency tracking will experience voltage polarization memory rotation with respect to the measur
53. elled in Figure 13 13 by the help of three Above Line blocks Differential and bias currents in the following model are calculated according to the equation 22 taking into account that the calculation has to be done using complex values of currents from both sides of the protected object Date 19 08 2011 Page 90 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE COOPERATION Three lmes 1 CURRENT POSITIVE Lu Lr oc L POSITIVE CURRENT ANGLE Figure 13 13 The part of a model that outputs binary 1 if a point in tripping characteristic lies in the tripping area Date 19 08 2011 Page 91 DEL WPS 1 Part 3 Toolbox for model ing ETN components V3 doc st PecAse 9 ha COOPERATION 14 Distance protection ANSI IEEE C37 2 standard device number IEC 61850 logical node name PDIS Input signals Current and voltage phase and angle Output signals Binary 1 or 0 Distance relay usually has 5 zones of phase fault and earth fault protection Often it is possible to choose between quadrilateral polygon characteristics or mho circles Rectangular characteristics with 2 zones are common for generator distance protection The faulted phase impedance must appear within a tripping zone in order to issue a tripping signal Some possible characteristics of distance elements are presented in the following table 25 29 Quadrilateral Universal X aipR cb polygon X a5 R
54. er protection systems 1s therefore not required allowing for tripping without additional delay DEL WPS 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE COOPERATION Therefore on those relatively rare occasions when there is a genuine fault inside the protection zone one can be almost sure that this fault will be quickly and selectively isolated In case of pan European simulator it means that it is not necessary to develop burdensome model at all Any fault inside the busbar differential protection zone can be treated with immediate opening of respective circuit breakers see Figure 13 2 L Nr 320 TNr 1 D D lt Sd M 301 O xX KS 1 330 NN KS 2 330 M E D L Nr 319 TNr 2 Figure 13 2 Selective isolation of busbar fault Detailed primary schemes of stations and substations will not be applied in the pan European simulator except maybe for the most important nodes and for dispatcher off line training simulations However it must be noted that the arrangement and connection of incoming and outgoing feeders in stations and substations and the number of busbars have an important influence on the supply reliability of the power system Knowledge of the exact primary schemes can be very useful for transmission system operators The same approach can be applied for other differential protections as well Let us suppose that pan European simulator operator wants to simulate a fault at generator transformer
55. ers that minimize the error AG gg c SE ES Z i S 50 ml Figure 3 4 Generic model of variable speed machine 3 7 Model of the full size converter machine and its control system The machines used in conjunction with the full size converter are separately or permanent magnet excited generator SG and asynchronous generator ASG But due to the presence the dc link capacitor the machine including the machine side converter MSC do not directly influence the grid behaviour The generator and the MSC can therefore be modelled as just a controllable current source A typical configuration for the full size converter machine is shown in Figure 3 5 Date 19 08 2011 Page 13 DEL WPS5 1 Part 3 Toolbox for model COOPERATION ing ETN components V3 doc 5 PEGASE See COOPERATION Gearbox Generator A L Line Chopper R U Ube Q Uc Pitch and power MSC and LSC Control Speed control Figure 3 5 Full size converter based WT system configuration including the control system Assuming that z the impedance of the converter choke zz the transformer impedance the current source can be determined as kc withv voltage injected by the converter and z Z z lx Es The corresponding Norton equivalent circuit is given in Figure 3 6 l VZ ig Figure 3 6 T
56. escription value Active and reactive power control KU 2 pu Gainin voltage channel IMAX l pu eis o pac o eme IPMIN IL pu How a paco eim ee HOMI ft 1 pu wa a TC 0001 s Time constant for close shut down of converter TO 3 s Time constant for open recover of converter IIMMAX 5 pu Limi IMMIN 5 pu IREMIN 5 pu d pu Gan 0 0 0 0 00 0 KG TI 0005 s Lag parameters TD 003 s DEL WPS 1 Part 3 Toolbox for model ing ETN components V3 doc COOPERATION 5 PecAse Oc COOPERATION 4 4 Model testing 4 4 Sample power grid The simple scheme shown Figure 4 3 has been used for testing purposes NHV1 NHV2 NGEN NLOAD GEN amp a L NLOAD NHV3 Figure 4 3 Test network Synchronous generator is connected to NGEN node It is equipped with turbine governor and AVR NGEN is considered as a slack bus in steady state computation Load is connected to NLOAD node It is modeled with constant active and reactive power QSS Wind Farm model is used to control the injector which is connected to NWF node NHW node is modeled as PV node in steady state computation Network data are shown in the tables below Table 2 Network power flow data Node Base name Date 19 08 2011 Page 20 DEL WPS5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE COOPERATION Table 3 Test network branc
57. for example is chosen to be oy the voltage current and flux linkage space phasors correspond to the ordinary complex phasors Voltage equations ZK ul i ZS jo 1 dt ZK up Tele oco 2 Flux equations V hb h 3 Wr Shh tyi 4 With Io OBL 5 and L l l 6 Equation of motion do 1 my ul Vis 7 dt T My Sd Sq Sq Where ly Main field inductance l Inductance my Torque at the turbine shaft r Resistance Tu Inertia constant u Voltage phasor V Complex flux linkages Oy Angular velocity of the reference frame 09 Synchronous angular velocity Superscripts subscripts S R Stator rotor d q Direct quadrature axis component Oo leakage inductance ZK In arbitrary reference frame speed x The voltage equations resolved into real and imaginary parts together with the equation of motion 7 constitute the detailed full order model of the doubly fed induction machine The terminal voltage u s forms the link to the rest of the network In the following sections this detailed EMTP type model will be simplified for use in simulations based on phasor approximation in the following steps Simplified model Detailed model gt with the orange box in Figure 3 8 Simplified model Quasi steady state QS model without the orange box in Figure 3 8 Date 19 08 2011 Page 10 DEL_WP5 1_Part_3_Toolbox_for_model ing ETN components V3 doc 5s Pe
58. g Principles and Applications 3rd Edition CRC Press 2007 ISBN 1 57444 716 5 Hewitson L G Brown M Balakrishnan R Practical Power System Protection Elsevier 2004 ISBN 0 7506 6397 9 Page 116 DEL_WP5 1_Part_3_Toolbox_for_model ing_ETN_components_V3 doc PEGASE 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 Date 19 08 2011 COOPERATION MiCOM P125 P126 P127 Directional Non directional Relay Version 6D Technical Guide Areva MiCOM P141 P142 P143 Feeder Management Relays Technical Guide Areva F650 Digital Bay Controller User manual Firmware version 3 7X GE Multilin 2007 Elmore W A Protective Relaying Theory and Applications 2nd Edition Marcel Dekker Inc 2003 ISBN 0 82470 972 1 Reimert D Protective Relaying for Power Generation Systems Taylor amp Francis 2006 ISBN 0 8247 0700 1 Horowitz S H Phadke A G Power System Relaying 3rd Edition John Wiley amp Sons 2008 ISBN 978 0 470 05712 4 Network Protection and Automation Guide Alstom 2002 ISBN 2 9518589 0 6 Jeff Roberts and Armando Guzm n Directional Element Design and Evaluation Schweitzer Engineering Laboratories Inc 2006 John Horak Directional Overcurrent Relaying 67 Concepts Basler Electric 2006 MiCOM P6231 P632 P633 P634 Transformer Differential Protection Software Ver
59. ge source converter based HVDC transmission systems and several protection and automation systems Additionally the models have been implemented on different computation platforms and results of sample computations provided The first two chapters following this introduction deal with the modelling of wind turbines and the implementation of this model on EUROSTAG The report summarises the necessary mathematical steps leading to a generic wind turbine model applicable for Type 3 doubly fed induction generator and Type 4 full converter interface wind turbines The models have been implemented and the error in comparison with a more detailed model quantified The subsequent two chapters focus on simplified modelling of HVDC VSC As is well known the active and reactive power in the HVDC VSC transmission system can be controlled independently from one another and even in situations of zero active power the system can continue to operate to transmit reactive power in support of system voltage during disturbances As this technology gives total flexibility regarding its location in the AC system it is quite probable that the number of HVDC VSC installations will increase significantly in the future As with the wind turbine a simplified HVDC VSC has been derived and implemented It is shown that without the need for exact representation and the use of complex mathematical relationships a fairly acceptable error margin in the simulations can be achiev
60. h data Sending and Base Resistance Reactance Semi shunt Semi shunt Ratio receiving node power conductance susceptance names NHVI NHV2 1 1000 0 002 0 023 0 000 0 279 D NHV1 NHV2 2 1000 0 002 0 023 0 000 0 279 EN NWVZNHVSO Too foor foon foo O14 f NGENNAVE nor foo wem o OO NIVINLOAD 100 n zr fooro i 0 i8 NNsNwE 36 ux ownen a Results of steady state computation are shown in the tables below Table 4 Power flow result bus data Node Bees Voltage Voltage Nano voltage magnitude angle 1 043 0 154 1 084 0 085 NWF 24 000 1 036 0 076 Table 5 Line power flow Sending and Power flow receiving node Active Reactive names Mw NWF NHV3 150 0 4 5 Test results Model was tested in several test cases to verify it operates correctly Date 19 08 2011 Page 21 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc 5s PEGASE COOPERATION M A 4 5 1 Direct metallic short circuit This test case is intended to show the model behavior during fast transient both in detailed and simplified versions as well as to verify that converter blocking is implemented correctly Metallic short circuit on NHV3 bus is modeled at time t 60 s Wind farm power output its terminal voltage magnitude and angle for simplified model simulation are shown at the figure below Using simplified model sim MACHINE WF ACTIVE POWER Unit MW sim
61. hat the optimization by equation 18 can lead to wrong direction determination as the phasor can go into the reverse plane even in case of a forward fault Using for instance the plane showed in Figure 11 3 it can happen in case of a highly capacitive forward fault To avoid wrong direction determination the angle should be limited within the ranges showed in Figure 11 4 and equations 19 Some margin for instance 2 in both directions can be set to separate safely forward and reverse directions 3 4 16 18 Date 19 08 2011 Page 76 DEL_WP5 1_Part_3_Toolbox_for_model ing ETN components V3 doc us PecAse M 9 0 Forward Forward 0 88 0 88 e ll e 9o e 90 77 2 92 180 92 180 Reverse Reverse 180 Figure 11 4 Limits of angle for reliable direction determination Forward if 88 lt lt 88 o Reverse if 92 lt lt 180 or 92 lt lt 180 19 Since ot is continually growing quantity the angle of current and voltage phasors is not constant see Figure 11 5 200 180 100 arg IA1 TT j 180 arg VAL 10 200 0 2 4 Figure 11 5 The continually growing angles of current and voltage If the two waveforms seen in Figure 11 5 do not reset at the same time due to angle difference between current and voltage the angle difference calculation can lead to substantial inaccuracies see Figure 11
62. he current source Norton equivalent of the generator and the MSC As stated above the machine itself can be considered a controlled current source and the converter as a first order delay The resulting block diagram of the full converter machine including the control system is given in Figure 3 7 Current controller converter Current source Figure 3 7 Block diagram of current controller converter and current source Date 19 08 2011 Page 14 DEL WPS 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE 3 8 3 9 Date 19 08 2011 Page 15 COOPERATION This representation after some simplification and re arrangement leads to the same generic model developed for the DFIG Figure 3 4 Current limitation equations The output current limitation may be necessitated by physical control or dynamic limits Short term thermal current limitations define physical limits for the maximum current the system can feed into the grid For the DFIG and full converter machine these current limits are determined by the converter which has a much shorter time constant than the generator Normally active current priority is applicable during steady state operation while reactive current priority may be used during very low or high voltage conditions Depending on the type of priority setting applicable the maximum allowable active or reactive current can be defined as follows H2 2 H2 _ 32 I5 max 7 Lik ly Or D max I I ax l
63. her the AC network voltage is symmetrical or not The use of PWM makes it possible that fast control of the active and reactive power is achieved This capability is beneficial when support of the AC network particularly during disturbances is needed since the control can be optimized to obtain a fast and stable performance during AC system fault recovery The DC voltage control provides DC voltage control using the current as a control variable The AC voltage dependent current limiter acts to keep the AC bus voltage within its upper and lower limits uS PA simplified HVDC VSC Dynamic Model The two AC networks linked to one another by the HVDC line can be represented by their Th venin or Norton equivalent circuits with the control actions adjusting the voltage or the current source as shown in Figure 5 3 and Figure 5 4 REC AC Grid SEC AC Grid User defined block model lcpc JX sec control Link control D Figure 5 3 The Thevenin equivalent circuit of HVDC VSC Date 19 08 2011 Page 30 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc nas PeGASe COOPERATION REC AC Grid SEC AC Grid User defined block model REC DC SEC control Link control VREC iREC irec 7 JXREC Figure 5 4 The Norton equivalent circuit of HVDC VSC The acronyms used in Figure 5 4 have the following meaning REC Receiving end converter SEC Sending end co
64. igure 3 8 is ON The red curve represents the same situation when the generic model is incorporated i e the switch in OFF position 3 11 Sign convention and initialization procedure 3 11 1 Sign convention Generated active power is negative or alternatively consumed active power is positive Inductive reactive power under excited generation mode is positive or capacitive reactive power overexcited generation mode is negative Ut KV and u t pu represents the magnitude of the terminal voltage 3 112 Initialization After incorporating the Detailed Model Assume Detailed Model is switched on at toy then initialize as follows set first order delay block and integral outputs i w t n upper branch Bye Coo AT set first order delay block and integral outputs i 529 lower branch After initialization simulation is continued by incorporating the Detailed Model After switching off the Detailed Model Assume Detailed Model is switched off at torr then proceed with the initialization as follows Setall block outputs and state variables zero Date 19 08 2011 Page 17 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE 4 COOPERATION Implementation of the wind farm model 4 1 Modifications made in EUROSTAG model compared to base model description Base model was slightly modified to make it more coherent to EUROSTAG comput
65. imple representation of oil or winding temperatures due to load but is not truly the temperature of interest oil top oil winding or hot spot temperature Functionally therefore this is essentially an over current function with an asymptotic time delay Thermal replica based protection elements typically include several threshold settings to alarm and trip on increasing temperature conditions The thermal replica model is widely used it is incorporated in many protection devices by different manufacturers and it can be applied to different power system elements Protection tripping time can be easily calculated assuming the protected object is loaded with constant current I 2 3 4 I 2 init Oe ooo min 11 The calculated value will show how much time has left until the actual heat content H will exceed the set overload level Hein Htrip 100 In a similar way as with the protection tripping time the alarm time can be estimated 2 f Hinit talarm T n min 12 I I B Hatarm max Date 19 08 2011 Page 54 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE COOPERATION The calculated value will show how much time has left until the actual heat content H will exceed the set alarm level Hajarm Halarm 80 46 for instance assuming that the load current I will remain constant The previously mentioned operating principles are presented in the Figure 8 1 as a block diag
66. ing end converter model 5 4 Demonstration Examples 5 4 1 Test network and controller parameters To illustrate the usage of the models and the plausibility of the results typical scenarios involving voltage and power step changes have been simulated The configuratiion and parameters of the test system are summarised below REC AC Grid SEC AC Grid VREC V SEC VRECO 1 T j0 pu V RECO 1 jo pu Toy 1 0 pu on vj 1 12 pu off vy 1 02 pu Table 6 Controller parameters To 26 ms REC and SEC current controller parameters of REC and SEC are the same G 0 4pu T 0 005s T 0 028 Xeno Xc 0 13 pu Additional REC model parameters and settings Gae 35 pu Tyo O01s Gye 5 0 pu ieee mm 1 25 pu Date 19 08 2011 Page 34 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc z PecAse COOPERATION Additional SEC model parameters and settings hsc max 7125 pu Proses 4 Pracl Psucl 10 01 1926 of the power passing through each converter 5 4 2 Simulation results a Example 1 SEC reference power changed from 1 0 p u to 0 9 p u For a step changing involving SEC reference power the behavior of power at both at the sending and receiving ends as well as the DC voltage are shown in Response of REC and SEC active power to Ap 0 1 pu Response of REC and SEC reactive power to Ap ret 0 1 pu 0 0215 0 2 0 4 0 6 og ts 4 pu 0 022 0
67. ircuits it is respectively 90 and 90 It means that the forward flowing current phasor will always be located in the forward plane of Figure 11 2 Date 19 08 2011 Page 75 DEL WPS5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE COOPERATION The problem with the approach given in equation 14 is that the fault power factor usually is low i e the current lags the voltage by nearly 90 Thus for typical X to R ratios between 3 and 20 the phase displacement angle is 73 to 87 3 3 lt 5 lt 20 ES tgp gt 71 6 lt lt 87 1 17 With angle values close to 90 direction determination can becomes unreliable To resolve the issue that fault current is typically highly lagging the directional unit can be optimized to detect lagging current conditions rather than 1 0 power factor conditions One approach shown in Figure 11 3 is to phase shift the voltage phasor so that the shifted voltage phasor is in phase with current when current lags the 1 0 power factor condition by some setting SHIFT p Zh ZU SHIFT 18 45 Ui p 28 Q 42 p 45 Forward Forward JU LL WI AL A u M oa uf F Heverse Heverse 135 Figure 11 3 A plane that represents forward and reverse direction with voltage phasor shifted by 45 Now during a fault the angle values are safely in the forward plane for reliable direction determination It should be noted t
68. istics and settings as shown in Figure 10 6 CURRENT Selection of characteristic peu Pick up drop nut and time multiplier settings a CST Block Ed AKIO KS Lol 1 25 Release EST LEEREN Ea CST Figure 10 6 Pseudo EUROSTAG block diagram of O C protection with inverse time delay characteristic Date 19 08 2011 Page 72 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE COOPERATION Some typical features of non directional over current protection that were left out of the model in order to fit it better into the framework of PEGASE project are discussed below 3 4 8 13 User defined curves It should be noted that some manufacturers offer the possibility to utilize user defined characteristics These characteristics are then defined point by point Thus for instance in SIEMENS 7SJ6 relays up to 20 pairs of values current time may be entered The device then approximates the characteristic using linear interpolation The dropout curve may be user defined as well Sq A Fickup Curve l lp T Tp A1T230 1000 000 400 000 100 000 0000 x L1 mq E I Figure 10 7 An example of user defined curve in SIEMENS 7SJ6 relay Drop out behaviour Very often it is possible to select whether the dropout of an element is to occur instantaneously after the threshold has been undershot or whether dropout is to be performed according to some characteristic Instantaneously
69. ith this test case Active and reactive power current output of the WF is shown at the figure below DEL_WP5 1_Part_3_Toolbox_for_model ing ETN components V3 doc Date 19 08 2011 Page 24 sim ACTIVE POWER Dh NHV3 1 Unit MW sim REACT POWER LINENWF NHV3 1 Unit Mvar sim CURRENT ON LINE NWF NHV3 1 Unit kA Figure 4 7 Change in voltage set point As seen from the figure reactive power output rises after set point modification Consequently total current output rises and hits the current limit Then active power output is decreased to keep the current in the range This shows correct operation of current limitation block in the model 4 6 EUROSTAG limitations The following EUROSTAG limitations were discovered during WF model implementation all this limitations prevents simple implementation of switching from detailed to simplified model of WE at runtime 1 EUROSTAG has no features for disabling or enabling parts of the model at runtime Therefore proposed switching from detailed to simplified model of WF at runtime may be implemented only using some workaround 2 EUROSTAG has no ability to change memory block such as integrator simple lag etc state at runtime 3 There is no limited integrator block with variable limits Date 19 08 2011 Page 25 DEL WPS 1 Part 3 Toolbox for model ing ETN components V3 doc s PecAse 9 o COOPERATION Some of these limitations may be overridden by use of
70. ively As a result of this modification controller variables change with grid frequency REC terminal voltage l Grid synchronous coordinates oriented coordinates PQ Active reactive current reference Prec arg es Injected current source Figure 5 6 Current controller in synchronous reference frame Alternative 2 Terminal voltage oriented coordinates If the controller operates in RSC terminal voltage oriented coordinates the controller variables remain constant in steady state conditions in spite of changes in grid frequency In this case the RSC terminal voltage is the reference 1 e it lies along the real axis and the necessary variations in terms of input and output variables is shown in Error Reference source not found Page 32 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE COOPERATION REC terminal voltage oriented coordinates Grid synchronous coordinates PQ PO LREC ref L lL REC Active reactive current reference Prec arg Wine Injected current source Figure 5 7 Current controller in terminal voltage reference frame The current control structure makes use of standard PI controllers as shown in Figure 5 6 and Figure 5 7 The output in both cases is the current source shown in Figure 5 4 5 3 3 DC Link Model The model of the DC link capacitance and the DC chopper are shown in Figure 5 8 The model also accounts for the power dissipated by the
71. lation Experience shows that an increase of 6 K means half the life time DEL WPS 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE COOPERATION Loss of life As the key goal of implementing thermal overload protection functions is to limit the loss of life of the transformer winding insulation the loss of life protection element sometimes is used as well With constant rated load the relative ageing rate L is equal to 1 For values greater than 1 accelerated ageing applies e g if L 2 only half of the life time is expected compared to the life time under nominal load conditions Date 19 08 2011 Page 61 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE COOPERATION 9 Non directional instantaneous definite time over current protection Output signals Binary O or 1 When a fault occurs in a power system the fault current is almost always greater than the pre fault load current in any power system element A very simple and effective relaying principle is the one using current magnitude as an indicator of a fault Over current relays can be applied to protect practically any power system element i e transmission lines transformers generators or motors The operating principle of a relay can be defined as follows 3 4 8 13 gt I gt Tri set p 13 I Ij Do not trip The quantity Iet is known as the pickup setting of the relay The equation 13 describes an ideal relay o
72. lly equipped with power swing blocking to prevent the switching of protected lines in case of the stable swings Blocking is carried out by the measuring of elapsed time between the passing of apparent impedance vector through outer and inner zone of blocking elements If elapsed time is greater than defined value usually 40 50 ms distance elements are blocked inhibited Blocking may affect either phase to phase elements or both phase to phase and phase to ground elements Sometimes the smallest zone the Ist stage is not blocked 2 Load encroachment detection blocking element To exclude the risk of unwanted fault detection by the distance protection during heavy load flow the load region can be cut out from the characteristics The cut out shape is defined by the minimum load resistance and the load area angle The first of these two features will be analyzed and modeled in the next chapter and if necessary can be added to distance protection model to its blocking or releasing input The other concerns mainly the last reserve or back up zones of the distance protection which are not reasonable to take into account in the framework of PEGASE project DEL WPS 1 Part 3 Toolbox for model ing ETN components V3 doc 5s PecAse COOPERATION 9 pex CT TH Accleration Manual Switch 3 Resettable Transport Signal Integrator Delay binary reset is on O 02 sec falling edge to break Igebraic double a Eae
73. model ing ETN components V3 doc PEGASE COOPERATION The voltage drop over the rotor resistance is used as an auxiliary signal to be controlled by a PI controller With the feed forward terms accounted for the PI rotor current controller only needs to put into effect the transition of the rotor current to the set value and compensate for the stator and rotor resistances The corresponding current contro loop is shown in Figure 3 3 upper part Figure 3 3 Block diagram of quasi stationary model of DFIG including rotor current control The line side converter LSC primarily controls the DC link voltage which in simplified simulations can be assumed to remain constant around the rated value and thus not considered However the magnitude of the current set value for the LSC is limited in accordance with the converter rating with a priority for the active current Although the model of the LSC is not incorporated explicitly the current limitation will be considered in the generic model as will be shown later Date 19 08 2011 Page 12 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE 3 6 Simplified model of DFIG and its control system the generic model A comprehensive system study for a large system such as the UCTE involves the simulation of a large interconnected system spanning national boundaries operated by different transmission system operators TSO and incorporating wind turbines of multiple manuf
74. model behavior during simulation An attempt was made to implement hysteresis block utilizing SET RESET block of EUROSTAG but it produced the same picture Eventually hysteresis block was implemented with use of time constant block as shown in Figure 6 2 Date 19 08 2011 Page 40 DEL WPS5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE COOPERATION QVDC VCHON 1 F AND CHST EE Cv lt Figure 6 2 Implementation of hysteresis block Here VDC is an input and CHST is an output of hysteresis block This implementation gives decent quality of modeling but invokes a warning message at initialization phase of time domain simulation This message shall be ignored 6 2 4 Losses modeling Losses in the converters are modeled approximately as a proportion of active power transferred through SEC Parameter KLOS determines the proportion KLOS is set directly in parameters for Version 1 and Version 2 models For Model A and Model B it is calculated by equation 1 FLOSS 1 FLOSS KLOS where FLOSS determines losses 1n each converter In the load flow the following equation must be accounted IPsec F nec KLOS IPsec 6 2 0 Implementation of converter reactance In model Version converter reactance is accounted inside model MACROBLOCK For other implementations Model 0 Model A Model B the reactance must be included in EUROSTAG network file To do this it is sufficient t
75. modification in target MACROBLOCK 4 6 2 Limited integrator block with variable limits Limited integrator block may be modeled by manually controlling limit crossing An example of implementation of this approach is shown at the figure below Date 19 08 2011 Page 27 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE COOPERATION T2 SET POINT Pa 1 Y 6 5 9 10 C ST LIMIT L n OUTUL OUT 0 gt i 0 Fi 0 0 A 11 A SET In POINT 1 ra 1 Y 14 AND ou 13 Y 18 J h OR LIMIT gt ais 7 0 16 A AND 17 A 15 K i l on IPE oy os Figure 4 11 Implementation of limited integrator block with variable limits in EUROSTAG B High Voltage Direct Current HVDC Transmission System 5 1 Introduction Modern HVDC transmission systems use voltage source converters VSC based on self commutated switching devices The superiority of HVDC VSC over the classical HVDC stems from the fact that the outputs of the VSC are determined solely by the control system and not by the AC network s ability to keep the voltage and frequency constant This give
76. n Divide Setting Ip A g Ip CA Bs Constant z Constant 3 Figure 10 4 Equation for inverse time characteristic using SIMULINK Even if the equation for inverse time characteristic can be written directly into some elementary function block it is necessary to provide the possibility to switch between different characteristics In Figure 10 5 a part of a model is presented in which it is possible to choose the necessary characteristic by providing appropriate control signal to the switching elements Date 19 08 2011 Page 71 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc PeGASO COOPERATION Current Input E D 14 uY vue 02 1 Current Setting Ip A Switch 4 Time Delay Calclatian 1 Threshold 0 5 Time Multiplier Setting Tp Selection of Mux characteristic ta combine 3 i signals inta a vector i Switch 2 Time Delay Calclatian 2 Threshold 0 5 SO urs cuc vuc2 32 1 Switch 3 Time Delay Caleclatian 3 Threshold 0 5 a 120 ut WiCurt vuczzi1 13 Switch 4 Time Delay Calclatian 4 Threshold 0 5 round Upper signal ult Middle signal u Lower signal uz Figure 10 5 Part of a model for selection of necessary characteristic The more characteristics are to be included the larger the model will be Taking into account the scale of pan European simulator it seems that it would be better to develop programmatically a dedicated element that would include all the necessary character
77. n s unloading logic unit Date 19 08 2011 Page 115 DEL WPS 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE COOPERATION 21 References 1 2 3 4 5 6 7 3 9 10 11 12 13 14 15 16 17 Date 19 08 201 1 M Behnke et al Development and Validation of WECC Variable Speed Wind Turbine Dynamic Models for Grid Integration Studies AWEA s 2007 WindPower Conference Los Angeles California June 4 7 2007 Fortmann J Engelhardt S Kretchmann J Feltes C Janssen M Neumann T amp Erlich I Generic Simulation Model for DFIG and Full Size Converter Based Wind Turbines 9th International Workshop on Large Scale Integration of Wind Power into Power Systems as well as on Transmission Networks for Offshore Wind Power Plants Oct 2010 Quebec City Quebec Canada Feltes C Wrede H Koch F W amp Erlich I Enhanced Fault Ride Through Method for Wind Farms Connected to the Grid Through VSC Based HVDC Transmission IEEE Transactions on Power Systems 2009 Vol 24 No 3 pp 1537 1546 Shewarega F Erlich amp Rueda Jos L Impact of Large Offshore Wind Farms on Power System Transient Stability IEEE PES Power Systems Conference amp Exposition March 2009 Seattle WA USA Feltes C Wrede H amp Erlich L Dynamic Behaviour of DFIG Based Wind Turbines During Grid Faults IEEJ Transactions on Industry Applications 2008 Vol
78. ng on the relay setting the value of the thermal replica is either reset to zero or maintained in the memory In the latter case when the normal situation is restored the thermal replica uses the stored value for calculation and matches it to the operating conditions Hot spot temperature Calculation of the hot spot temperature and determination of the ageing rate according to IEC 60354 can also be considered these quantities are considered in IEEE C57 91 1995 standard as well however there are terminological differences between IEC and IEEE standards when defining certain parameters Hot spot temperature is the hottest temperature spot in the transformer winding The location of the winding hottest spot is dependent on the physical design of the transformer The protection device calculates the hot spot temperature from the input temperature data and the settings of the relay When a settable threshold of the hot spot temperature element is exceeded an annunciation or a trip is generated Hot spot calculation is done with different equations depending on the cooling method Aging factor The aging factor element detects transformer aging in per unit normal insulation aging The element can be set for alarm or trip whenever the computed aging factor is greater than the user defined pickup setting for the specified time delay The life time of cellulose insulation refers to a temperature of 98 C or 208 4 F in the direct environment of the insu
79. ngs Gro Bl w EY Save 29 Restore amp Default z Reset VIEW A amp Save p Restore e Default 29 Reset VIEW ALL 7 Operation Characteristic Graph New Site 1 New Device 1 Settings Grouped Elements Group 1 Transformer 0 500 pu 15 2 000 pu 10 000 pu 80 96 Adapt 2nd Per phase Inrush Inhibit Level 20 0 96 fo Overexcitation Inhibit Function Disabled Overexcitation Inhibit Level 10 0 96 fo OFF Self reset Differential pu Disabled Mea a E dg de ww gb dB e eo 3 Restraint pu New Device 1 Settings Grouped Elements Group 1 Transformer A Figure 13 7 Atypical tripping characteristic and settings of differential protection for General Electric T60 relay using bended lines 5 Tripping characteristic see Figure 13 6 usually consists of two or three segments characterized by pairs of points and slopes of segments In rare occasions that will not be taken into account the characteristic is more complicated as shown in Figure 13 7 for General Electric T60 relay In general there are several different approaches in calculation of differential and restraining currents Some of them are listed below Currents in 21 24 of course should be amplitude and vector matched 5 6 19 24 For two winding transformer protection lair L l In 7 h b 21 For two winding
80. nverter DC Direct current AC Alternating current The impedances Xggc and Xgrgc represent the equivalent Thevenin impedances with resistances neglected of both external AC circuits a Receiving end converter REC model The function of REC is to transfer the active power fed by the SEC into the HVDC line to the AC grid by maintaining the DC voltage level Active current is used to regulate the DC link voltage The reactive current control loop can be used to control the REC terminal voltage on the AC side or to guarantee a constant power factor at the AC grid terminal and also to support the grid voltage during faults set point s P DC Link Controller B l y IREC max BS REC max v DC ref 1 H ak O ST Voc EM I Sons REC max PQ Priority dREC O Power Factor Controller TO P j RECQ IREC ref i pae Magnitude REC ref SS L Y Limitation Q P wn j PREC ref Pe he P Gin d MS 4 I icona REC ref b ST uo Q PQ 222 2 2 l1 l2 1 Ai REC ref Always active PM current priority QREC 0 7 REC Terminal Voltage Controller pum REC ref e Gy V REC j Figure 5 5 REC control Date 19 08 2011 Page 31 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE Date 19 08 201 1 COOPERATION The REC control structure is shown in Figure 5 5 which implements the following three core functions The PI controller maintaining the DC voltage with the active converter cur
81. o fill in Shunt capacitor field for REC and SEC nodes in EUROSTAG Network Editor with the following value Q SBASE UO X VBASE where X is XREC or XSEC in p u as set in model parameters file UO is voltage at the node in steady state SBASE and VBASE are base power and voltage of the node Please note the sign of the reactive power Date 19 08 201 1 Page 41 DEL WPS 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE oo COOPERATION 6 3 Model parameters description Description of model parameters adjustable by user and their default values are given in the table below Table 8 Model parameters value GI FLOSS RCH TDCL 1 0 0 13 G 04 5 0 1 25 pu Convertercurrentlimit TI 002 TV 0005 0 13 GI u Gain in voltage controller u TI TV 6 4 Model testing 6 4 1 Sample power grid The model was tested in a simple scheme shownFigure 6 3 NRECSYS2 NRECSYS2 NSECSYS2 int NSEC o NSECSYS2 D NSECSYS1 NRECSYS1 NRECSYS REG VSC SEC VSC NeHeS 1S TENA NSECSYS1 Figure 6 3 Test system This network represents test case on which VSC HVDC model was tested on UDE software It is separated into two synchronous zones each containing INJECTOR representing VSC REC or SEC and two infinite bus nodes Number of infinite buses was doubled to enable switching of their voltage that is required by some testing procedures Date 19 08 2011
82. of under frequency protection it is possible to create a typical UFLS model If necessary rate of change of frequency can be used as input signal using frequency measurement and derivative lag elements see Figure 18 1 Figure 18 1 Frequency and rate of change of frequency signals Date 19 08 2011 Page 110 DEL WPS5 1 Part 3 Toolbox for model ing ETN components V3 doc s PecAse 9 ha COOPERATION 19 General protection model It can be seen from the previous chapters that all the models that operate according to some over or under criteria current voltage frequency are actually identical For the framework of PEGASE project it means that it is enough to have two models to cover a wide range of protections functions that operate according to over and under criteria In addition to the already inspected protection functions the general model can be applied also for instance as active P reactive Q apparent power S protection or generator speed protection using the available measurement elements in EUROSTAG software see Figure 19 1 The only difference concerns the input signal For convenience the over under voltage models are repeated in this chapter see Figure 19 2 Figure 19 3 that can be used to monitor different power system parameters as well ACTIVE VOLTAGE ELECTRICAL POWER ANGLE TORQUE REACTIVE POWER POSITIVE POSITIVE POSITIVE 1 1 0 MW Figure 19 1 Measurement blocks in E
83. ombination of ambient temperature oil temperature and measured current to detect the presence of an over temperature condition The scale of the PEGASE project restricts the size and complexity of the model therefore it is suggested to apply the following simplified differential equation for thermal overload protection omitting ambient temperature input signal and some other typical but not substantial features 1 2 3 4 Date 19 08 2011 Page 53 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE COOPERATION 2 Saas L Du Description pu Heat rise For example if H 120 the protected object overheats by 20 in relation to its rated operation conditions min Thermal time constant of winding or oil Imax A Maximum permissible current of the protected object Equation 10 is in accordance with IEC 60255 8 3 and is used to track a first order thermal image replica based on the measured current The thermal replica model calculates a maximum temperature rise based on the measured current the thermal time constant and the maximum permissible current of the protected object The biggest advantage of using a thermal replica for temperature protection is the ease of implementation The function involves only settings in the relay with no need to physically install and connect temperature sensors However this method does not account for ambient temperature and provides only a s
84. on 3 Switch off at 0 5 NOT Block Binary Logical Operator Logical Operator Figure 16 5 MATLAB SIMULINK model of under voltage protection In case of over voltage protection the model is simpler as the hysteresis Schmidt trigger element can be applied EUROSTAG and MATLAB SIMULINK models of over voltage protections are presented in Figure 16 6 and Figure 16 7 respectively Date 19 08 2011 Page 106 DEL_WP5 1_Part_3_Toolbox_for_model ing_ETN_components_V3 doc PEGASE COOPERATION a VOLTAGE uc 1 pem 7 EH CST i 1 1 pem a 37 EET 2 AND Ea CST EE CST ES CST Figure 16 6 EUROSTAG model of over voltage protection Resettable Small Integrator Transport resets on Delay falling edge 0 004 ec to bre ak algebraic Minimum Relational Data Type lanpi er 1 R Trip Operator Conversion 1 input Relay s ias Command Pickup setting Duration Analogue i peer Reschable pum pain eo F ntegrator a Sett resets on ese AND Trip falling edge Hysteresis Binary Release Binan Data Type Subtract Switch on at 5 Conversion 2 Switch off at 0 5 HOT Black Binan Logical Operatori Logical Operator Figure 16 7 MATLAB SIMULINK model of over voltage protection Date 19 08 2011 Page 107 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc 5 PEGASE 9 ha COOPERATION 17 Under frequency and over frequency protection Properties Description
85. on of voltage regulators To load shed the non priority consumer network when an overload occurs To monitor the voltage before sources to carry out a power supply transfer The protection in its simplest form is activated when one or all three of the phase to phase or phase to ground voltages are lower higher than the setting threshold Positive sequence voltage can also be used Generally it includes a definite time delay Suggested block diagram of under voltage protection for PEGASE project is presented in Figure 16 2 In Figure 16 3 and Figure 16 4 the same model is split into two main parts When modeling under voltage protection it should be noted that the reset setting of the protection has larger value than the trip setting therefore standard hysteresis Schmidt trigger element cannot be applied In order to model the pick up drop out feature the logic depicted in Figure 16 1 can be used if the operating point is below pick up setting then 2 is set at the output of the block No 15 in Figure 16 2 and the relay picks up If operating point is above the drop out setting then 2 is set and the relay drops out In case of 0 relay waits for further voltage changes Voltage w Drop out Drop out f 9 C1 setting C C Pick up S i SIS Time Figure 16 1 Pick up and drop out moments during transient process Date 19 08 2011 Page 103 DEL WPS5 1 Part 3 Toolbox for model ing ETN components V3 doc a e
86. ondition is maintained for a time period equal or greater than the time delay setting 0 50 s in this example that is set by the help of Summer element No 21 the binary 1 value appears at the output of the Relay element No 23 relay trips The resettable integrator is a main part of a logic that estimates for how long some criteria have been fulfilled To obtain this functionality in EUROSTAG and Scilab the following equivalent scheme is used see Figure 9 2 and Figure 9 3 The output signal is being reset by providing a large negative input to the integrator when necessary gt Switch Resettable Threshold 0 5 Integrator Integrator reset is on falling edge Canstant Figure 9 2 Modelling of resettable integrator E 3 Resettable integrator CURRENT NT oe i F EA BEL ES CIE oie E Input signal Pick up and 0 5 otherwise j drop out settings zl 1 Time setting 0 1 ass Time setting a COT Block EH AMD E W CST ES Eelease CST Ea CST Figure 9 3 EUROSTAG block diagram of non directional over current protection Date 19 08 2011 Page 63 DEL WPS 1 Part 3 Toolbox for model ing ETN components V3 doc 5s PecAse COOPERATION The same model is presented below in MATLAB SIMULINK as well as SCILAB XCOS environment as these programs are also used in PEGASE project Figure 9 4 shows the full block diagram of the non directional instantaneous definite time o
87. onversion 2 Switch off at 0 5 HOT Block Binan Logical Operator Logical Operator Figure 17 1 Model of over frequency protection Date 19 08 2011 Page 108 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE COOPERATION t If f Pick up Pick up setting pui Frequency input Subtract 1 iff gt Pickup P Analogue Hysteresis Sum Switch on at 1 5 Pickup if 2 Switch off at 1 5 4 Drap aut if 2 Threshold 2 0 Do nothing if 0 Subtract 2 ff gt Drap out Data Type Conversion 1 Drop out setting pu Resettable Small Integrator Transport reset is on Delav falling edge 0 001 sec ta break algebraic biinimu m Relational Data Type liop NN i Trip Operator Conversion Z B da Command E BP Duration Time elationa iu Resettable Delay Operator sec Integrator t Setting sec resets on AND Trip falling edge Hysteresis Binary Release Binary Data Type Subtract3 Switch an atO5 Conversion 3 Switch off at 0 5 NOT Block Binan Logical Operator Logical Operator Figure 17 2 Model of under frequency protection Date 19 08 2011 Page 109 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE COOPERATION 18 Under frequency load shedding During emergency situation in the power system caused by generating power deficiency frequency decline takes place Dynamics of under frequency during the deficienc
88. out setting PON Upper Limit 100 Lower Limit 0 0 10 Time Delay Setting Release NOT Block Logical op AND fm Switch On 0 5 i Binary Switch Off 0 5 Output Trip Figure 9 7 SCILAB XCOS block diagram of non directional over current protection Some additional features that can be modelled by the given model using the blocking and releasing inputs 3 4 8 13 Date 19 08 2011 Page 65 DEL WPS 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE Date 19 08 2011 Page 66 COOPERATION Dynamic cold load pickup function Sometimes it may be necessary to dynamically increase the pickup values if during starting certain elements of the system show an increased power consumption after a long period of zero voltage e g air conditioning systems heating installations motors Thus a general raise of pickup thresholds can be avoided taking such starting conditions into consideration A number of manufacturers have implemented such feature in their relays The dynamic cold load pickup feature can be simulated by blocking one over current protection model and releasing the other when necessary Switch on to fault SOTF logic The instantaneous high current switch onto fault protection function is usually provided to disconnect immediately and without delay feeders that are switched onto a high current fault It is primarily used as fast protection in the event of energizing the feeder while the earth s
89. perating characteristic as shown in Figure 9 1 The relay does not operate operating time is infinite as long as current magnitude is less than I If current magnitude exceeds I gt relay operates taking a definite time T to close its contacts Operating zone gt gt gt I Figure 9 1 Tripping characteristic of two stage over current protection function The previously mentioned operating principles are presented in the Figure 3 2 as a block diagram created in EUROSTAG environment In this model positive sequence current see Current element No 22 in Figure 9 3 is continuously compared with the pick up value see Schmidt Trigger element No 8 in Figure 9 3 that can be set in amperes or in per unit values whichever is more suitable from the point of view of software developers Phase current as an input quantity can be used as well Date 19 08 2011 Page 62 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE COOPERATION As soon as positive sequence current exceeds the pick up value 0 7 kA in this example the binary 1 value appears at the output of the Schmidt Trigger element relay picks up On the other hand relay drops out as soon as current decreases below the drop out value 0 6 kA in this example The binary 1 is converted to 0 5 by the help of Summer element No 36 to form a convenient threshold signal for a Relay element No 11 If the pick up c
90. r replica check zone All three processors must reach a trip decision independently before the trip command is released Independent characteristics for check zone and bus selective zone modification of stabilization current so that it decays at the rate of t 64 ms special restraint to avoid over stabilization by separate addition of the magnitudes of positive and negative currents and selection of the smaller sum different algorithms working in parallel and many more factors influence the operation of 7SS52 24 As a result it is very hard to simplify the logic involved in 78852 or any other similar device from ABB GE Areva etc so that the model would relatively accurately imitate the real protection device Fortunately there are two specific features of busbar differential protection that allow taking very simple approach regarding modeling These features are 1 Rarity of occurrence of busbar faults Busbar faults are on occasion caused by mechanical or insulator defects but are frequently caused by incorrect switching operations Statistically one can count one busbar fault every 10 years 0 63 to 2 faults per feeder in 100 years according to CIGRE survey 21 2 100 selectivity of busbar protection Busbar protection is 10096 selective and therefore only responds to faults within its protected zone The boundary of the protected zone is uniquely defined by the location of the current transformers Time grading with oth
91. r station The control functions of HVDC VSC converter include AC and DC voltage control active and reactive power control and the inner current control together with current output limitation and internal converter voltage limitations For a two terminal HVDC VSC system one of the converters controls the DC voltage and the other the active power Additionally each of the converters can optionally be set in either AC voltage control or reactive power control mode The basic control functions are summarized in Figure 5 2 Date 19 08 2011 Page 29 DEL_WP5 1_Part_3_Toolbox_for_model ing ETN components V3 doc mas PeGASe COOPERATION Or l Upc AC voltage control Reactive power C control bad Active power current voltage e control limit U Q control Figure 5 2 Overview of HVDC VSC control functions ref DC voltage Inner current control control ref The active power control function aims to fulfil system wide objectives such as power flow control and congestion management As a result during normal operation the power and voltage settings are determined by the system operator 6 The lower level control functions are attuned to these prescriptions and the reference values are determined accordingly The current control operates in a coordinate system that is phase synchronous with the fundamental frequency in the network and provides a symmetrical three phase current to the converter regardless of whet
92. ram created in EUROSTAG environment DEL WPS 1 Part 3 Toolbox for model ing ETN components V3 doc Date 19 08 2011 Page 55 PEGASE A E l en ed 5 RE Hz ag H gt b 3 qp I I Lo stun Log Log AL A AL Block signal Release signal Alarm setting CURRENT H G H g Imax setting Figure 8 1 EUROSTAG block diagram of thermal overload protection Date 19 08 2011 Page 56 ing ETN components V3 doc trip signal Lri Lg duration COOPERATION DEL WPS 1 Part 3 Toolbox for model Date 19 08 2011 Page 57 PEGASE COOPERATION The thermal overload protection model is continuously fed with positive sequence current i e this protection function 1s active throughout the simulation process The calculation of the heat content H is based on the fact that the temperature in the windings or oil is proportional to the square of the current and that the temperature increases and decreases exponentially with a certain time constant 1 2 3 4 The part of the model that represents the equation 10 is given in Figure 8 2 CURRENT POSITIVE 1 ES A ix H pu ooo E s 1 x min I max 4 or k4 Figure 8 2 Part of the model that represents the equation 10 Inclusion of two additional inputs see Figure 8 3 significantly extends the application capabilities of model as very often protection functions are blocked or relea
93. rent as a control variable The current magnitude limitation block with active current priority during normal operation The AC voltage control block or alternatively the power factor control dashed lines The voltage controller is a fast acting proportional controller which can include a dead band so that the control action is only then activated when the voltage error is larger than a pre specified value for example 0 1 p u This block is also responsible for grid voltage support during faults In steady state operation the DC voltage control and by implication the d axis component of the REC current has priority In case of grid fault however the priority is switched to reactive current to provide fast voltage support b Calculation of active reactive current reference and inner current control loop The converter control is based on a vector control approach with its rotating reference frame aligned with the grid voltage reference or with the terminal voltage chosen as the reference As a result active power can be controlled through d axis and reactive power through q axis component of the converter current both independently of one another Alternative 1 Grid synchronous coordinates If the controller operates in the grid reference frame then the current phase position should be modified as shown in Figure 5 6 so that the real and imaginary parts of the current correspond to the active and reactive parts of the current respect
94. requency and active and reactive power flow according to the needs of the network Additionally HVDC VSC has the capability to rapidly control active and reactive power independent of one another Sending end converter SEC Receiving end converter A S Hz unn LL ara Ls Figure 5 1 Circuit diagram of a VSC based HVDC HVDC light Active power transfer can be quickly reversed without any change of the control mode and without any filter switching or converter blocking The power reversal is obtained by changing the direction of the DC current without the need for changing the DC voltage polarity as opposed to conventional HVDC HVDC VSC enables a decoupled control of active and reactive power and allows the connection of weak or even passive networks Additionally the high switching frequencies of approximately 1 2 kHz reduce the filter size and the IGBT valves themselves have a smaller size compared to thyristor valves in classical HVDC systems This leads to a smaller footprint of the converter stations and thus makes it more suitable for applications where space requirement is a critical issue 5 3 Control of HVDC VSC 5 3 1 Overview HVDC VSC consists of two voltage source converters one operating as a rectifier and the other as an inverter Both ends of the HVDC line are connected to an AC circuit and the HVDC line linking the two transmits the power from the rectifier to the inverte
95. s this may not be always correct but it can be assumed for simplification One advantage of the blinder scheme for power swing detection applications is that it can be used independent of the distance zone characteristics In addition while the impedance vector is in this delta impedance area the protective relay element can be blocked from tripping as this may be either heavy load or a stable power swing If an unstable swing is detected the mho element can be allowed to trip immediately not recommended or tripping can be delayed until the swing passes through thereby minimizing over voltages across the opened breaker To find the correct settings for the blinders is not always simple and requires a sophisticated grid analysis p Operate i Y Restrain y RO l Unstable Stable Power Swing Power Swing k amp o 4 j ee Figure 15 4 Two blinder scheme 31 Figure 15 5 illustrates the power swing detection model with two blinder scheme DEL_WP5 1 Part 3 Toolbox for model ing ETN components V3 doc COOPERATION pueuiq Bun 18096 4 s qgpur pueuiq Bum lamog 101E18 drj IE21B77 7 yopmg GO 1E go uou GO E uo uoumz PEINS sI5818 5AH b ypg JHF gt uku H 2az sull Guo 1o E12d JEU n HE 98 H sBpa Burg ua si 13281 101E1B21u B qEgaseu doo JDIEIqa JE HE OF as ZOO Eja g pod uElj fug ur EuBig JENUE py WH e a a MN QUE
96. s V3 doc PEGASE Date 19 08 2011 Page 88 COOPERATION The second way to define triple slope biased restraint characteristics is to determine it by the following four pair of points see equations 30 33 and Figure 13 9 This approach is more flexible and allows easier to define the necessary characteristic lai laa las laa 30 loj Ibo Ips Ipa 0 2 4 8 10 Figure 13 9 Tripping characteristic defined by four pairs of points Segment 1 Ip lt Ip lt Ip lh Ip 31 ME Ia2 1a lai b2 bl Segment 2 Ip Ip 1p3 ees ES 32 DE Ia3 7192 1a2 b3 b2 Segment 3 Ip3 S Ip S Tp4 33 no Ip 153 OW Ig4 7 Ig3 Iga b4 Ip3 Figure 13 10 illustrates a general model of differential protection for two terminal power system element The right part of the model is the same as in previous chapters In the left part equations 22 30 33 are modeled For conversion to per unit values a base current of 1000 A is used in this example DEL WPS 1 Part 3 Toolbox for model ing ETN components V3 doc COOPERATION Gg 1E Uo quauis Gig uo young PENS zizalmjz H UE MIQ duj idon aigiqaBE Eag OF 935 ZOO AE B podsuely saalBiap aseyd apie zl yzg nd a aa dug uas anuo apniiuEBE ui epis sasuBap aseyd apis Hz mee nd o sam dug uoiianuas spnjubciEui apis 1n E1a drj JEa1B o7 Cue Ug Pubs z dome ENUE py yO Eg jusuiBas
97. s total flexibility regarding the location of the converters in the AC system since the requirement on the AC network in terms of short circuit capacity SCR is no longer a limiting factor In other words HVDC SVC can feed power even into a passive network Classic HVDC terminals can provide limited control of reactive power by means of switching of filters and shunt banks and to some degree through firing angle control But the HVDC VSC control makes it possible to create any voltage phase angle or amplitude which can be accomplished almost instantly Unlike conventional HVDC converters that normally have a 596 minimum current the HVDC SVC converter can operate at very low power or even at zero power As the active and reactive power are controlled independently at zero active power the full rating can be utilized to transmit reactive power If and when the need arises the same converter can even be used as an SVC Date 19 08 2011 Page 28 DEL WPS5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE COOPERATION 5 2 HVDC VSC basics The converters use a set of six valves two for each phase which are equipped with high power Insulated Gate Bipolar Transistors IGBT The valves are controlled by a control system using pulse width modulation PWM Since IGBTs can be switched on or off as desired output voltages and currents on the AC side can be controlled precisely and the control system thus enables the control of voltage f
98. sed according to some external criterion Trip setting pu or 58 Drop out setting pu or 0 Block signal Figure 8 3 Additional blocking and releasing binary inputs If the releasing or blocking input is not used then the binary 1 should be connected to the logical AND operator Blocking input could be supplemented with logical NOT operator so that binary 0 instead of a 1 would result in no blocking condition DEL WPS 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE COOPERATION Pickup of the thermal overload protection function is stabilized by setting the dropout value see Figure 8 3 the pickup condition is maintained until the current falls below the drop out value thus securing that the function does not drop out too fast Drop out ratio is mostly around 0 95 for new digital protection devices For older electromechanical devices it 1s lower for instance 0 8 The model is further extended to avoid an inconsistent length of tripping impulses that in some cases can be very short See Figure 8 4 where the preset length of output signal is enforced 0 25 s is a typical minimum trip command duration setting and it can be hard coded into a model command duration Figure 8 4 Modelling of preset length of output signal As soon as calculated heat rise exceeds the trip setting i e the maximum permissible temperature that corresponds to the maximum permissible current the trip outpu
99. sion 610 Technical Manual Areva SIPROTEC Differential Protection 7UT613 63x V4 60 Manual Siemens Ziegler G Numerical Differential Protection Publicis Corporate Publishing 2005 ISBN 3 89578 234 3 Busbar Differential Protection IED REB670 Technical Reference Manual ABB 2006 SIPROTEC Multifunctional Machine Protection 7UM62 V4 6 Manual Siemens 2005 SIPROTEC Distributed Busbar Breaker Failure Protection 788522 V4 6 788523 V3 3 788525 V3 3 Manual Siemens 2007 MiCOMho P443 and P445 Distance Relay Software Version 33 Technical Manual Areva Ziegler G Numerical Distance Protection 2nd Edition Publicis Corporate Publishing 2006 ISBN 3 89578 266 1 SIPROTEC Line Differential Protection with Distance Protection 78D52 53 V4 60 Manual Siemens Karel Maslo Martin Kanok Distance Protection Model for Network Simulators 8th International Conference CONTROL OF POWER SYSTEMS 08 June 11 13 2008 Strbsk Pleso High Tatras Slovak Republic SIPROTEC Distance Protection 7SA6 V4 61 Manual Siemens P Kundur Power System Stability and Control McGraw Hill Inc 1993 Power Swing and Out of Step Considerations of Transmission Lines IEEE PSRC WG D6 2005 Page 117 DEL_WP5 1_Part_3_Toolbox_for_model ing ETN components V3 doc
100. sions and the generic model in its final form seeseseseesseeeessss 15 3 10 Accuracy of the generic model in comparison with the QSS model ss 16 3 11 Sign convention and initialization procedure csececccseeeeecceeeeeeceeeeeecsaeeeeeeeueeeeesaueeesaaeeeeesaeess 17 Sdk MEE cies i o m 17 cupri me H 17 4 Implementation of the wind farm model eeesseesseeeeneeenrenenn 18 4 1 Modifications made in EUROSTAG model compared to base model description 18 4 2 Implementation of specific parts of the model cccsececccseeeeecceeeeeeceeeeeeceeeeeeseaseeesaeeessaaeees 18 4 2 1 Simple lag with variable time constant ccceecccceceseeeeeeeeaeeeeeeeeeseeeceeeseaeeeeeeeeaaeeeess 18 4 2 2 Current conversion eeeeeessssssssssssseeeeeeeeee nennen nennen nennen nnne nnns snne nnn nnne Errena 18 4 2 3 PIO Cr WIL MIMO n mtem 18 4 3 Model parameters GeSCTIPtiON cccccccccccceeeessseeeeeeeeeeaeeseeeeeeeeseaaaceeeeeeeeseaaasseeeeeeessssaagseeeeness 19 zr ModeltestiNng ME m E EE EE Ea 20 4 4 1 Sample power grid lssssssesssessessseeseenenen nennen nnne nnns nn nar n nnn n nsns nain ns 20 no TOSTO SU ee 21 4 5 1 Direct metallic short circuit
101. special techniques Below we describe some of them 4 6 1 Reset feature Integrator and simple lag block with reset feature may be modelled with the following schemes Figure 4 8 Modeling of simple lag block with reset feature Figure 4 9 Modeling of integrator block with reset feature To reset block at specific moment t one must change at that moment the value BIAS to y t The problem with this approach is that SINGLE SETPOINT MODIFICATION EVENT from EUROSTAG seq file has no access to model outputs 1 e it can t read the value of y t The solution is to use USER AUTOMATON because SINGLE SETPOINT MODIFICATION EVENT from USER AUTOMATON EVENTS collection can modify setpoint with value read from model The approach is illustrated at the figure below Date 19 08 2011 Page 26 DEL WPS5 1 Part 3 Toolbox for model ing ETN components V3 doc 33 PeGASe Oc COOPERATION Equipment machine node line etc MACROBLOCK through Equipment tab 4 BIAS le ee ee e l MACROAUTOMATON MACROBLOCK SINGLE SETPOINT MODIFICATION MACROAUTOMATON USER AUTOMATON USER AUTOMATON EVENT Figure 4 10 Implementation of simple lag block with reset in EUROSTAG The scheme works as follows the reset of simple lag block is initiated by modification of setpoint block in MACROAUTOMATON MACROBLOCK that may be done by event in seq file This must trigger USER AUTOMATON which in its turn initiates setpoint
102. t signal appears and lasts at least 0 25 s Trip setting usually is set in percent in this example it is 100 The alarm stage is modeled in the same way as trip stage albeit with the lower operation setting To consider that in EUROSTAG the best algorithmic performances are achieved when the output variable is the output of a differential block or in case of an algebraic equation when the variable is varying smoothly a Simple Lag element with short time constant is used as a last or output element of the model see Figure 8 4 It should be noted that all the necessary features of the model are included in the block diagram though it would had been possible to transfer some of its features to the so called User automatons For informative purposes the same model in MATLAB SIMULINK environment is presented as well Figure 8 5 It has some differences with EUROSTAG model due to the differences of available elementary blocks and modeling approach in both programs In this and in the following chapters blocks of the Matlab models in yellow color represent input and output signals but blocks in blue color the settings Date 19 08 2011 Page 58 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc COOPERATION Cayqnop o G Orde go uoums p UGS aU piojEladp SO Te us qoum adj eed EUDE H VEU sI5818 54AH 5 aas Woe a qnop o PUB WLU c UO aay ce Io E12ad UE x ada eeg IEuongjas WMA daa E ELS LM HE
103. tallations will increase significantly As with the wind turbine a simplified HVDC VSC model has been derived and implemented It is shown that without the need for exact representation and the use of complex mathematical relationships which in any case would not be of much help without reliable parameters a fairly acceptable error margin in the simulations can be achieved Protection and automation devices are handled in the subsequent section of the report Due to great variety of protection systems in general and special protection schemes and wide area Date 19 08 2011 Page 6 DEL WPS5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE COOPERATION protection and automation systems in particular it is impossible to create ready to use system models As a result a library of standard elements as well as standard protection and automation models have been developed which can be put together or adjusted by the end user to create any complex system needed Date 19 08 2011 Page 7 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE Date 19 08 2011 Page 8 COOPERATION Introduction The number of power electronic converter based equipment and system connected to the European Transmission Network ETN has increased rapidly over the recent past This work package as part of the overall activities in the PEGASE projected is devoted to the analysis of the accuracy and robustness of the commonly used mo
104. tch on atO 5 Conversion 2 Switch off at 0 5 HOT Block Binary Logical Operator 1 Logical Dperator z Figure 9 5 MATLAB SIMULINK block diagram of non directional over current protection using equivalent scheme instead of resettable integrator Date 19 08 2011 Page 64 DEL_WPS5 1_Part_3_Toolbox_for_model ing ETN components V3 doc PeGASO COOPERATION In case the short circuit duration is close to the relay time delay setting the tripping impulse length can be very short If software developers decide that they can deal with the variable length of output signal simulation time step is decision making criterion here then the model can be greatly simplified see Figure 9 6 Current input Analogue bx ud Time rap out setting Resettable Delay Operator Integrator Setti resets on engia falling edge Release Binary Block Binani Logical Operator 1 Relational Trip Binary Operator 2 Figure 9 6 Simplified MATLAB SIMULINK block diagram of non directional over current protection In Figure 9 7 the full model of non directional over current protection is shown in Scilab format as well Resettable Integrator Continuous fix delay SWITCH Saturation Yes Threshold a 0 5 Delay 0 001 Upper Limit 100 Pass if gt a 9999 Resettable Integrator gt op amp Lower Limit 0 Bick oF Saturation Yes Current Input ick up setting Threshold a 0 5 Drop
105. te 19 08 2011 Page 67 COOPERATION Voltage restraint Voltage control For backup over current protection for instance of generators ordinary over current relays are sometimes difficult to apply due to the decaying characteristic of the fault current The value of the fault current will be progressively reduced due to armature reaction to a value less than full load current Therefore normal over current relays set above the load current or maximum permissible overload cannot be applied to provide time delayed backup protection as they will not operate for these fault conditions For successful application of generator backup protection the operating characteristics of the relay are required to be a function of voltage and current There are two types of relay that are customarily used for these applications namely voltage restrained and voltage controlled over current relays With voltage restrained over current relays when the voltage falls below a set value the operating time of the over current characteristic is continuously reduced with declining voltage In voltage controlled over current relays the operating time characteristic is changed from the load characteristic to the fault characteristic when the voltage falls below a set level DEL WPS 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE 10 COOPERATION Non directional inverse time over current protection ANSI IEEE C37 2 standard device number IEC
106. tem fault and not an out of step trip condition 25 The timer of 25ms is a fixed timer in the logic and not user accessible The same time delay is used in model see Figure 15 5 Instead of a minimum tripping time duration as in previous models blocking time duration is set in this model The characteristic presented in Figure 15 6 can be easily modeled in EUROSTAG by the help of four Above Line elements In Figure 15 7 a part of a model that distinguishes a stable and unstable power swing is presented Timers can be added to this model using the resettable integrator as shown for instance in Figure 9 3 Date 19 08 2011 Page 101 DEL_WP5 1_Part_3_Toolbox_for_model ing ETN components V3 doc 33 PeGASe ee COOPERATION ABOVE LIME Crossing of og two blinders ANDO Crossing of all HN four blinders FE r nm z oO m m Figure 15 7 Part of a model that detects the crossing of blinders Date 19 08 2011 Page 102 DEL_WP5 1_Part_3_Toolbox_for_model ing ETN components V3 doc PEGASe e COOPERATION 16 Under voltage and over voltage protection Properties Description ANSUI IEEE C37 2 standard device number 277 under voltage 59 over voltage IEC 61850 logical node name PTUV under voltage PTOV over voltage Output signals Binary 1 or 0 These protections are used to protect equipment against an abnormally low or high voltage They can also be used 3 To monitor the operati
107. tion as in Figure 11 9 possibly according to the previously presented algorithm 26 CURRENT ex k Eu x LST 1 1 e5 CST 999 RET 5 Em AND EE CST 2s CST 1 Protection release block according to the EJ direction criteria CST Figure 11 9 Pseudo model with dedicated direction determination element Date 19 08 2011 Page 79 DEL_WPS5 1_Part_3_Toolbox_for_model ing ETN components V3 doc zu PecAse M This approach would be especially beneficial in case of directional inverse time over current protection see Figure 11 10 CURRENT Selection of characteristic pom Pick up drop out and time multiplier settings 2 CHT e ia AMO LCST l E Release tT 1 Protection release E according to the EH ection criteria eT Figure 11 10 Pseudo model with dedicated elements for inverse time characteristic and direction determination Date 19 08 2011 Page 80 DEL WPS 1 Part 3 Toolbox for model ing ETN components V3 doc s PecAse 9 ha COOPERATION 12 Zero residual and negative sequence over current protection ANSI IEEE C37 2 standard device number 46 negative sequence over current 50N instantaneous zero current 51N inverse time zero current 67N directional zero current IEC 61850 logical node name PIOC instantaneous PTOC definite time inverse time Input signals Current zero or negative sequence Output signals Binary 1 or 0
108. tion on the usage of the Schmidt block to model hysteresis has been found Date 19 08 2011 Page 47 DEL WPS 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE COOPERATION The figure below shows block diagram that uses Schmidt trigger block This model is derived from VSC HVDC REC model and simplified to leave only the parts that are essential for demonstration Figure 6 7 Block diagram that demostrates abnormal behaivor of EUROSTAG The model has been correctly compiled and is used to control current injector in conditions of fourth test case see 6 4 2 During simulation the output produced by the integrator block 29 was inconsistent with its input The figure below shows the input and the output of the block DEL WPS 1 Part 3 Toolbox for model ing ETN components V3 doc Date 19 08 2011 Page 48 cE pecase Mh Figure 6 8 The input bottom and the output top of the integrator block 29 The figure shows that while the input is positive at given time period the output doesn t grow steadily as it should but experience random variations Nevertheless this modeling issue has been solved using the scheme described in 6 2 3with acceptable results 7 Digital loss of synchronism protection 7 1 Model description Loss of synchronism automation ALAR is a device designed to detect asynchronous mode of operation in power system and to initiate separation of asynchronously operating parts to prevent f
109. tions for the allocation of the feeder currents and distribution of the trip commands The stringent demands that are put on busbar differential protection result in very complicated internal logic In Figure 13 1 the structure of Siemens numerical busbar protection 7SS52 is presented as an example This protection is fully numerical and consists of a central unit and de centralized bay units 24 Date 19 08 2011 Page 82 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE Date 19 08 2011 Page 83 COOPERATION Figure 13 1 Structure of the numerical busbar protection 78852 Due to the simple current comparison the principle of differential protection is very straight forward If the current entering into the protected zone is not equal to the current leaving this zone the current differences at the ends of the protected zone indicates the fault In real life application such as 7SS52 the fault recognition depends on the busbar configuration current transformer placement and position of all isolators and circuit breakers This information along with measured currents is transferred from the bay units to the central unit for further processing Trip decision is based on 3 independent measurements 2 measurements are based on busbar configuration two independent processors execute the protection algorithm on alternate data samples and the third measurement considers all busbar sections independent of the isolato
110. transformers or other two terminal power system elements such as generators lair h L 22 Ig h b Calculation of differential and restraining currents for three or four winding protection lait lj I5 I3 L4 In 1 IIl is uf 23 For multi terminal protected objects such as busbars Date 19 08 201 1 Page 86 DEL WPS 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE COOPERATION lar h I I5 c Ii Ig 24 Ig I ls E A little bit more common is approach given in equations 22 and 24 for two terminal and multi terminal power system elements respectively Typical three segment characteristic can be defined in two ways The first one is to use input data given in 25 and presented in Figure 13 8 la Ibi 152 Ip Ki K2 k3 25 k3 k2 I kl Idl 0 2 4 6 8 Ibl Th2 Th3 Figure 13 8 Possible input data for differential protection tripping characteristic The tripping criterion for the considered input data is formulated as Segment 1 a if the restraint bias current is in the following region 0XI lt Ig 26 b and differential current exceeds the one calculated by 27 I3 5 Ela 27 c then the trip signal can be issued Segment 2 Ip S Ip S 1p 28 lg gt k gt Ip Iy ky Ip Tli Segment 3 Ip S Ip lt Ip 29 Ig gt K5 Ip 152 K2 152 Ii Ki Toi Ha Date 19 08 2011 Page 87 DEL WP5 1 Part 3 Toolbox for model ing ETN component
111. ts final form Fault Ride Through FRT is now a general requirement on wind turbines Accordingly wind turbines must remain connected to the grid during faults by introducing new technologies if need be FRT of wind turbines is necessary at least for two reasons To be able to continue with active power in feed immediately after the fault clearance To provide voltage support during and after the fault period to reduce the size of the voltage dip area within the grid The model is extended to account for this operational requirement by incorporating a block which increases or reduces the reactive power in feed whenever the voltage exits a dead band of 10 above or below the rated value Figure 3 8 represents the model in its final form The model as a whole both with and without the orange box constitutes the simplified model The only difference between the two options including or excluding the orange box is the level of simplification which one chooses to adopt for a particular simulation DEL WPS 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE 2 COOPERATION signal Wind Farm QSS Model 3 Teia a sae Model lo 01 pu lo max 10 min Ip a deadband K 20 10 h kV i T i 7 X u t pu y kV amp 9 t Mvar 5 0 O u 3 2 z ol f E p max l MVA I Mvar d 10pu 69 P r MW pul mE u t pu amp Fuad ir 00 i ap PU y e 40 Detailed Mod
112. urther fault propagation ALAR is a local automation device i e it detects asynchronous mode on specific line and initiates the line tripping Principles of asynchronous mode detection may vary They may base on measurements of active power current or impedance Modern type devices detect loss of synchronism by comparing phase angle between voltage phasors on the ends of power transmission line Voltage phasor on one end is measured while phasor on the opposite end is calculated The model of power transmission line used for calculation is shown at the figure U U Z Date 19 08 2011 E m 49 PEL WPS 1 Part 3 Toolbox for model ing ETN_components_V3 doc PEGASE COOPERATION Phasors of U and I are measured and U phasor is computed according to formula The device has three stages of operation First stage is activated when phase angle between line ends becomes greater than 180 degrees Second and third stages are activated after certain cycles of asynchronous operation Cycle counter is reset if time between cycles exceeds certain value set by TRESET parameter This brings the device to initial state The model has five output variables one for each stage and two outputs which show whether the part of the system to which the model is connected goes faster or slower than the other part 41 25 Notes on implementation The logic of the device is implemented in macroautomaton macroblock ALAR Corresponding macroautomaton shall be
113. utations provided In the process particular emphasis was placed on wind turbine modelling as the level of installed wind power generation continues to increase steadily and its impact on power system dynamic behaviour is becoming increasingly significant The report describes the necessary steps for deriving a simplified model for Type 3 doubly fed induction generator and Type 4 full converter interface wind turbines The models have been implemented and the error in comparison with a more detailed model quantified The report subsequently focuses on simplified modelling of HVDC VSC As is well known voltage source based HVDC is able to create any voltage phase angle or amplitude almost instantly Unlike conventional HVDC converters that normally have a 596 minimum current the HVDC SVC converter can operate at very low power or even at zero power As the active and reactive power are controlled independently in situations of zero active power the full rating can be utilized to transmit reactive power and to support system voltage especially during disturbances The clear superiority of HVDC VSC over the classical HVDC also stems from the fact that the outputs of the VSC are determined solely by the control system and not by the AC network s ability to keep the voltage and frequency constant As this technology gives total flexibility regarding the location of the converters in the AC system it is quite probable that the number of HVDC VSC ins
114. utrq Eu amp is JE MWe py azEB asH s 4 s A MN Bua lamog Gor JE uo usus SOPE ue uou 215313 5 H HF Gas uag AE 8 1 aL saBpa 181a uon si asa 101E1B21u B qEgase a qnop a Passos siapulq 1E Bum 12008 4 aqe eun tapua pn pn tn LANIE LAN pTM ANAE Can e Japulig KELAK NABANI png mag pens Con e Japullg one cien tani CL LINE LINE panne un Lapua snis ene Qon Gea nos CL png naar pens Cn zjndui po Jaquinu ppo ue H andl 40 ES Lx Bl i g G t sasalBiap asEud l L GssiBspl _ r aseyd kn zi yzg T H sa1a dug Ex Ho xdg spnyuBEu gayon NE ble En kl apnyu ew al a QU i B Inl i i x ion model detect ing f power sw iagram o Block d Figure 15 5 D E e E wm Y e 23 Hig p uh es in 3 a E wv 9 E z zz z ef A E O O rT to qv a be eel rT CN NS co O Y qv C PEGASE COOPERATION Four lines in two blinder scheme are defined in the model using eight pair of points and the following line equation where r1 x1 r2 x2 are two pairs of points defining one line X X L r 2n 38 eI Al r x The modeled two blinder characteristic is presented in Figure 15 6 20 10 20 Figure 15 6 Example of two blinder characteristic In Areva P443 relay if the disturbance takes less than 25ms to cross two blinders the relay will consider this to be a power sys
115. vacuinadecvnandesaenuesiapianieus E E EE 50 7 3 Model parameters GESCIIDtION ccccccsseeececcceeseeceeeceeeeeeeeeeeeescecesseaseceeessaaseceesesaaseeeesssageeeeseaas 50 Ts so b iom 50 VAs Sample power gid TTE 50 FAs FOSO US ororen pet e A E E E E E E EOE ENEA 51 7 4 3 Conclusion on test results sisse nennen nnne nnns 53 Date 19 08 2011 Page 4 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE M 9 MEM Ov Toad PlOlOGUON a cccscrdcccecaduedercedvenes dad dett du dod e hd etae de oti dd i dedidit du dress 53 9 Non directional instantaneous definite time over current protection 62 10 Non directional inverse time over current protection eeseeeseeeeeeee eee 68 11 Directional over current protection ccceccceecceeeeseeeceeeceeeeaeeeseeeeeueeeeeeseeessueeneeenaees 74 12 Zero residual and negative sequence over current protection 81 19 JAerenial DOECO ca creterctdennes didit adii o dad did dad diede dti qa indie adidieeidadiaes 82 tA DISCE P OG CUOI mm T m UU TERT 92 15 Power swing detection eessessssssessseeseee eene nennen nennen nnne nnne nnne nnns 96 16 Under voltage and over voltage protection ccececeee cece eect eeeeeeeeeeeeeeeaeeeeeaeeeeeas 103 17 Under frequency and over frequency protection ccccccccsececeeeeeeeee
116. ver current protection that includes drop out feature blocking and releasing inputs and provides a preset length of output signal Small Minimum Trip badare Transport Integrator Del Command iesean elay Duration falli d 0 004 sec Avi sec Sg edge ta break algebraic loop Relational Data Type Dperatorz Conversion 1 LP Current input En Anal ick up setting Analogue Drop out setting pouT Time Relational ESAHAOIE Delay Operator 1 Integrator t Setting sec resets on AND Trip falling edge Hysteresis Binary Release Binary Data Type Subtract Switch on atO 5 Conversion 2 Switch off at 0 5 HOT Block Binary Logical Operator 1 Logical Operator Figure 9 4 MATLAB SIMULINK block diagram of non directional over current protection using resettable integrator In Figure 9 5 the same model as in Figure 9 4 is presented with the exception of the resettable integrator element that is replaced with an equivalent scheme Minimum Trip Integrator Command Upper Saturation 100 Curation Lower Saturation 0 sec Constant P double Integrator Upper Saturation 100 Relational Data Type Lower Saturation 0 Operator 2 Conversion 1 Constant ul X1 Current ES ERI input Elay UE incus Pideup setting Threshold 0 5 Time Relational Drop out setting Delay Operator 1 agaad Setting sec we pee s eee Hysteresis binary aA Data Type Subtract Swi
117. ver time the quality of the models and their portability across simulation platforms For purposes of developing generic models wind turbines are classified as follows Type 1 conventional directly connected induction generator Type 2 wound rotor induction generator with variable rotor resistance Type 3 doubly fed induction generator Type 4 full converter interface In this report the necessary steps for deriving a simplified model for Type 3 and Type 4 wind turbines is described 3 DFIG based WT system configuration The following figure summarizes the major components of a DFIG based wind turbine system and the necessary control tasks The block denoted as operating control is responsible for coordinating the pitch angle and converter controls in addition to performing the necessary supervisory control measures for safe and automatic operation which will not be considered further Measurements Crowbar MSC PWM LSC PWM HN Operating Control mM Converter Control Figure 3 1 DFIG based WT system configuration including the control system Date 19 08 2011 Page 9 DEL WPS5 1 Part 3 Toolbox for model ing ETN components V3 doc PEGASE COOPERATION 3 3 Detailed full order model of the DFIG The complete set of mathematical relationships that describe the dynamic behaviour of the machine is given below The superscript ZK denotes an arbitrary reference frame rotating at the speed x If x
118. version Date 19 08 2011 Page 2 DEL WP5 1 Part 3 Toolbox for model ing ETN components V3 doc PeGASe M Table of content Deliverable D3 m asara EENE Era aE aE EEEE ENE 1 Modelng requirements Tor the ETN uscarea aE 1 Part 3 Toolbox for modeling ETN components ccccseeceseeeeseeeeeeeeeseeeeeeeeseeeesaeeeeaeeeeas 1 1 Executive Summary ssessssessssssssseseeee eene nnne nn snni sn sna sr sea sns na sanas a rara nnns 6 2 Introduction MM 8 3 Simplified wind farm wind turbine model seeeseeeeseeeeee 9 Seda AUCO Me T 9 3 2 DFIG based WT system configuration cccccceeccccecseeeeeeeecaeseeceeeceeeseceeesaaaseeeeeseeaeeeeesseaeeeesesaaess 9 3 3 Detailed full order model of the DFIG eslseeeesssseeseeneeeenen nennen 10 3 4 The quasi stationary QS model of the DFIG ccc ccccccccssseeeeeseeeeeeeeeeeeeeeeessessaeaeeeeeeeeeeees 11 3 5 Rotor side converter RSC model sseesssssssesssssssseeeee eene nennen nnne nnn nnn nnn 11 3 6 Simplified model of DFIG and its control system the generic model 12 3 7 Model of the full size converter machine and its control system seseeeeessssssee 13 3 8 Current limitation equations sssssesesssesseeeseennnn nnne nnn nnne nnn arr n nnn nsns n na nnns nnn nns 14 3 9 Further exten
119. witch is closed but can also be used every time the feeder is energized in other words also following automatic re closure If necessary the SOTF function can be easily modeled by shortly releasing the advanced over current protection model during feeder energization Voltage start Voltage start logic can be implemented by releasing the protection if voltage decreases below a preset value Blocking during power swings or re closure cycle Protection can be blocked in case of power swings or during automatic re closure cycle to avoid unwanted trip during these conditions Circuit Breaker Failure Protection The circuit breaker failure function ensures fast back up tripping of surrounding breakers A current check with extremely short reset time is used as check criteria to achieve a high security against unnecessary operation A current check can be performed using the non directional over current protection model Some typical features of non directional over current protection that were left out of the model in order to fit it better into the framework of PEGASE project 3 4 8 13 Inrush restraint When the numerical relay is installed for instance to protect a power transformer large magnetizing inrush currents can flow when the transformer is energized These inrush currents may be several times the nominal transformer current and depending on the transformer size and design may last from several milliseconds to several
120. y Where imax maximum allowable current ip_ max maximum allowable active current ig max maximum allowable reactive current ip actual active current i actual reactive current Active and reactive current output may further be limited as a function of the voltage Some of the reasons that may necessitate the reduction of the active current may be reactive current priority during grid faults active current restriction to fulfil specific grid code requirements active current restriction as the result of the voltage going outside the standard operating range and grid stability limits especially at weak connection points Similarly reactive current limitation may be needed as a result of active current priority during steady state operation Dynamic control limits are needed to allow for an improved representation of active power restoration process following a grid fault Common reasons that give rise to ramp rate limitation after a grid fault are the need to limit shaft and gear box loads and due to limits on rate of change of the dc link voltage Control delays in the converter the settling time needed by the PLL to re synchronize after deep voltage dips and phase angle jumps following fault clearance may act to slow the active power increase Manufacturers implement the delay as first order lag or as a linear ramp rate limitation The model preferably should allow for both implementations Further extensions and the generic model in i
121. y of generation in the power system can have very different character It depends on the value of disturbance response of emergency automation and governor system For gradual increase in load or for sudden but mild overloads electric power system generating units governors will sense speed change and increase power input to the generator Here so called primary and secondary frequency control is activated Severe system disturbances can result in fast frequency drop which makes impossible fast governor and boiler response To restore frequency load shedding or spinning reserve involvement may be activated If governor action cannot activate spinning reserve quickly enough to restore the system to its normal operating frequency frequency actuated automatic load shedding UFLS serves as a last resort tool to prevent the system from collapse Up to date automatic load shedding system practically in the most power systems foresees disconnection without time delay or with small delay part of the load on under frequency The number of load shedding steps and value of load to be shed vary for the different power systems Some power systems use rate of change of frequency as additional factor to shed a load As under frequency load shedding is tracking frequency and operates when the frequency decreases below some preset value the model for one stage of UFLS is essentially the same as model for under frequency protection By combining several models

D5.1-Part 3_Tool Box for modeling ETN components

Contents

Download Pdf Manuals

Related Search

Related Contents