Chap21-370-397 20/06/02 16:03 Page 37021Relay Testingand CommissioningIntroduction 21.1Electrical type tests 21.2Electromagnetic compatibility tests 21.3Product safety type tests 21.4Environmental type tests 21.5Software type tests 21.6Dynamic validation type testing 21.7Production testing 21.8Commissioning tests 21.9Secondary injection test equipment 21.10Secondary injection testing 21.11Primary injection testing 21.12Testing of protection scheme logic 21.13Tripping and alarm annunciation tests 21.14Periodic maintenance tests 21.15Protection scheme design for maintenance 21.16References 21.17

Chap21-370-397 20/06/02 16:04 Page 37121Relay Testingand Commissioning21.1 INTRODUCTIONThe testing of protection equipment schemes presents anumber of problems. This is because the main functionof protection equipment is solely concerned withoperation under system fault conditions, and cannotreadily be tested under normal system operatingconditions. This situation is aggravated by theincreasing complexity of protection schemes and use ofrelays containing software.The testing of protection equipment may be divided intofour stages:i.type testsii.routine factory production testsiii.commissioning testsiv.periodic maintenance tests21.1.1 Type TestsType tests are required to prove that a relay meets thepublished specification and complies with all relevantstandards. Since the principal function of a protectionrelay is to operate correctly under abnormal powerconditions, it is essential that the performance beassessed under such conditions. Comprehensive typetests simulating the operational conditions are thereforeconducted at the manufacturer's works during thedevelopment and certification of the equipment.The standards that cover most aspects of relayperformance are IEC 60255 and ANSI C37.90. Howevercompliance may also involve consideration of therequirements of IEC 61000, 60068 and 60529, whileproducts intended for use in the EEC also have to complywith the requirements of Directives 89/336/EEC and73/23/EEC. Since type testing of a digital or numericalrelay involves testing of software as well as hardware,the type testing process is very complicated and moreinvolved than a static or electromechanical relay.Network Protection & Automation Guide 371 Chap21-370-397 20/06/02 16:04 Page 37221.1.2 Routine Factory Production Testswould take 4 years to write the functional type-testspecifications, 30 years to perform the tests and severalThese are conducted to prove that relays are free fromyears to write the test reports that result. Automateddefects during manufacture. Testing will take place attechniques/ equipment are clearly required, and areseveral stages during manufacture, to ensure problemscovered in Section 21.7.2.are discovered at the earliest possible time and henceminimise remedial work. The extent of testing will beElement Range Step Sizedetermined by the complexity of the relay and pastI>10.08 - 4.00In 0.01Inmanufacturing experience.I>20.08 - 32In 0.01InDirectionality Forward/Reverse/Non-directionalRCA -95 to +95 121.1.3 Commissioning TestsCharacteristic DT/IDMTThese tests are designed to prove that a particularDefinite Time Delay 0 - 100s 0.01sIEC Standard Inverseprotection scheme has been installed correctly prior toIEC Very Inversesetting to work. All aspects of the scheme areIEC IDMT Time DelayIEC Extremely Inversethoroughly checked, from installation of the correctUK Long Time Inverseequipment through wiring checks and operation checksTime Multiplier Setting (TMS) 0.025 - 1.2 0.025of the individual items of equipment, finishing withIEEE Moderately Inversetesting of the complete scheme.IEEE Very InverseIEEE IDMT Time Delay IEEE Extremely InverseUS-CO8 Inverse21.1.4 Periodic Maintenance ChecksUS-CO2 Short Time InverseTime Dial (TD) 0.5 - 15 0.1These are required to identify equipment failures andIEC Reset Time (DT only) 0 - 100s 0.01sdegradation in service, so that corrective action can beIEEE Reset Time IDMT/DTtaken. Because a protection scheme only operates underIEEE DT Reset Time 0 - 100s 0.01sfault conditions, defects may not be revealed for aIEEE Moderately Inversesignificant period of time, until a fault occurs. RegularIEEE Very InverseIEEE IDMT Reset Time IEEE Extremely Inversetesting assists in detecting faults that would otherwiseUS-CO8 Inverseremain undetected until a fault occurs.US-CO2 Short Time InverseTable 21.1: Overcurrent relay element specification21.2 ELECTRICAL TYPE TESTSVarious electrical type tests must be performed, asfollows:Three phase non-directional pick up and drop off accuracyTest 1over complete current setting range for both stagesThree phase directional pick up and drop off accuracy21.2.1 Functional TestsTest 2over complete RCA setting range in the forward direction,current angle sweepThe functional tests consist of applying the appropriate21Three phase directional pick up and drop off accuracyinputs to the relay under test and measuring theTest 3over complete RCA setting range in the reverse direction,performance to determine if it meets the specification.current angle sweepThey are usually carried out under controlledThree phase directional pick up and drop off accuracyTest 4over complete RCA setting range in the forward direction,environmental conditions. The testing may be extensive,voltage angle sweepeven where only a simple relay function is being tested.,Three phase directional pick up and drop off accuracyas can be realised by considering the simple overcurrentTest 5over complete RCA setting range in the reverse direction,voltage angle sweeprelay element of Table 21.1.Test 6Three phase polarising voltage threshold testTo determine compliance with the specification, the testsAccuracy of DT timer over complete setting rangeTest 7listed in Table 21.2 are required to be carried out. This isAccuracy of IDMT curves over claimed accuracy rangeTest 8a time consuming task, involving many engineers andTest 9Accuracy of IDMT TMS/TDtechnicians. Hence it is expensive.Test 10Effect of changing fault current on IDMT operating timesMinimum Pick-Up of Starts and Trips for IDMT curvesTest 11When a modern numerical relay with many functions isAccuracy of reset timersTest 12considered, each of which has to be type-tested, theEffect of any blocking signals, opto inputs, VTS, AutorecloseTest 13functional type-testing involved is a major issue. In theVoltage polarisation memoryTest 14case of a recent relay development project, it wasTable 21.2: Overcurrent relay element functional type testscalculated that if one person had to do all the work, it 372 Network Protection & Automation Guide Chap21-370-397 20/06/02 16:04 Page 37321.2.2 Rating Testsseconds. This is carried out between all circuits and caseearth, between all independent circuits and acrossRating type tests are conducted to ensure thatnormally open contacts. The acceptance criterion for acomponents are used within their specified ratings andproduct in new condition is a minimum of 100M. Afterthat there are no fire or electric shock hazards under aa damp heat test the pass criterion is a minimum ofnormal load or fault condition of the power system. This10M.is in addition to checking that the product complies withits technical specification. The following are amongstthe rating type tests conducted on protection relays, the21.2.7 Auxiliary Suppliesspecified parameters are normally to IEC 60255-6.Digital and numerical protection relays normally requirean auxiliary supply to provide power to the on-board21.2.3 Thermal Withstandmicroprocessor circuitry and the interfacing opto-isolated input circuits and output relays. The auxiliaryThe thermal withstand of VTs, CTs and output contactsupply can be either a.c. or d.c., supplied from a numbercircuits is determined to ensure compliance with theof sources or safe supplies - i.e. batteries, UPS,specified continuous and short-term overload conditions.generators, etc., all of which may be subject to voltageIn addition to functional verification, the pass criterion isdips, short interruptions and voltage variations. Relaysthat there is no detrimental effect on the relay assembly,are designed to ensure that operation is maintained andor circuit components, when the product is subjected tono damage occurs during a disturbance of the auxiliaryoverload conditions that may be expected in service.supply.Thermal withstand is assessed over a time period of 1sfor CTs and 10s for VTs.Tests are carried out for both a.c. and d.c. auxiliarysupplies and include mains variation both above andbelow the nominal rating, supply interruptions derived by21.2.4 Relay Burdenopen circuit and short circuit, supply dips as apercentage of the nominal supply, repetitive starts. TheThe burdens of the auxiliary supply, optically isolatedduration of the interruptions and supply dips range frominputs, VTs and CTs are measured to check that the2ms to 60s intervals. A short supply interruption or dipproduct complies with its specification. The burden ofup to 20ms, possibly longer, should not cause anyproducts with a high number of input/output circuits ismalfunction of the relay. Malfunctions include theapplication specific i.e. it increases according to theoperation of output relays and watchdog contacts, thenumber of optically isolated input and output contactreset of microprocessors, alarm or trip indication,ports which are energised under normal power systemacceptance of corrupted data over the communicationload conditions. It is usually envisaged that not morelink and the corruption of stored data or settings. For athan 50% of such ports will be energised in anylonger supply interruption, or dip in excess of 20ms, theapplication.relay self recovers without the loss of any function, data,settings or corruption of data. No operator interventionis required to restore operation after an interruption or21.2.5 Relay Inputsdip in the supply. Many relays have a specification thatRelay inputs are tested over the specified ranges. Inputs21exceeds this requirement, tolerating dips of up to 50msinclude those for auxiliary voltage, VT, CT, frequency,without operation being affected.optically isolated digital inputs and communicationIn addition to the above, the relay is subjected to a numbercircuits.of repetitive starts or a sequence of supply interruptions.Again the relay is tested to ensure that no damage or data21.2.6 Relay Output Contactscorruption has occurred during the repetitive tests.Protection relay output contacts are type tested toSpecific tests carried out on d.c. auxiliary suppliesensure that they comply with the product specification.include reverse polarity, a.c. waveform superimposed onParticular withstand and endurance type tests have to bethe d.c. supply and the effect of a rising and decayingcarried out using d.c., since the normal supply is via aauxiliary voltage. All tests are carried out at variousstation battery.levels of loading of the relay auxiliary supply.21.2.7 Insulation Resistance21.3 ELECTROMAGNETIC COMPATIBILITY TESTSThe insulation resistance test is carried out according toThere are numerous tests that are carried out toIEC 60255-5, i.e. 500V d.c. 10%, for a minimum of 5determine the ability of relays to withstand the electricalNetwork Protection & Automation Guide 373 Chap21-370-397 20/06/02 16:04 Page 374environment in which they are installed. The substationthat the relay can withstand an interruption in theauxiliary supply without de-energising, e.g. switchingenvironment is a very severe environment in terms of theelectrical and electromagnetic interference that canoff, and that when this time is exceeded and it doestransiently switch off, that no maloperation occurs.arise. There are many sources of interference within asubstation, some originating internally, others beingIt simulates the effect of a loose fuse in the batteryconducted along the overhead lines or cables into thecircuit, or a short circuit in the common d.c. supply,substation from external disturbances. The mostinterrupted by a fuse. Another source of d.c. interruptioncommon sources are:is if there is a power system fault and the battery issupplying both the relay and the circuit breaker trip coils.a.switching operationsWhen the battery energises the coils to initiate theb.system faultscircuit breaker trip, the voltage may fall below thec.lightning strikesrequired level for operation of the relay and hence a d.c.interrupt occurs. The test is specified in IEC 60255-11d.conductor flashoverand comprises a interruptions of 2, 5, 10, 20, 50, 100 ande.telecommunication operations e.g. mobile phones200ms. For interruptions lasting up to and including20ms, the relay must not de-energise of maloperate,A whole suite of tests are performed to simulate thesewhile for longer interruptions it must not maloperate.types of interference, and they fall under the broadumbrella of what is known as EMC, or ElectromagneticThe relay is powered from a battery supply, and bothCompatibility tests.short circuit and open circuit interruptions are carriedout. Each interruption is applied 10 times, and forBroadly speaking, EMC can be defined as:auxiliary power supplies with a large operating range,The ability of equipment to co-exist in the samethe tests are performed at minimum, maximum, andelectromagnetic environmentother voltages across this range, to ensure complianceover the complete range.It is not a new subject and has been tested for by themilitary ever since the advent of electronic equipment. EMCcan cause real and serious problems, and does need to be21.3.2 A.C. Ripple on D.C. Supplytaken into account when designing electronic equipment.This test (IEC 60255-11) determines that the relay is ableEMC tests determine the impact on the relay under testto operate correctly with a superimposed a.c. voltage onof high-frequency electrical disturbances of variousthe d.c. supply. This is caused by the station battery beingkinds. Relays manufactured or intended for use in thecharged by the battery charger, and the relevant waveformEEC have to comply with EEC Directive 89/336/EEC inis shown in Figure 21.1. It consists of a 12% peak-to-peakthis respect. To achieve this, in addition to designing forripple superimposed on the d.c. supply voltage.statutory compliance to this Directive, the followingrange of tests are carried out:60.00a.d.c. interrupt test50.00b.a.c. ripple on d.c. supply test40.0021c.d.c. ramp test30.00d.high frequency disturbance test20.00e.fast transient test10.00f.surge immunity test0.00g.power frequency interference testTime (ms)h.electrostatic discharge testFigure 21.1: A.C. ripple superimposed on d.c.supply testi.conducted and radiated emissions testsFor auxiliary power supplies with a large operating range,j.conducted and radiated immunity teststhe tests are performed at minimum, maximum, andk.power frequency magnetic field testsother voltages across this range, to ensure compliancefor the complete range. The interference is applied usinga full wave rectifier network, connected in parallel with21.3.1 D.C Interrupt Testthe battery supply. The relay must continue to operateThis is a test to determine the maximum length of timewithout malfunction during the test. 374 Network Protection & Automation Guide Chap21-370-397 20/06/02 16:04 Page 37521.3.3 D.C. Ramp Down/Ramp UpBurst period, 300 msThis test simulates a failed station battery charger, whichVBurst duration (1/15 ms)would result in the auxiliary voltage to the relay slowlyramping down. The ramp up part simulates the batterybeing recharged after discharging. The relay must powertup cleanly when the voltage is applied and notmaloperate.5 ns rise time, 50 ns pulse widthVThere is no international standard for this test, so individualmanufacturers can decide if they wish to conduct such atest and what the test specification shall be.Repetition period tFigure 21.3: Fast Transient Test waveform21.3.4 High Frequency Disturbance TestThe product is energised in both normal (quiescent) andThe High Frequency Disturbance Test simulates hightripped modes for this test. It must not maloperate whenvoltage transients that result from power system faultsthe interference is applied in common mode via theand plant switching operations. It consists of a 1MHzintegral coupling network to each circuit in turn, for 60decaying sinusoidal waveform, as shown in Figure 21.2.seconds. Interference is coupled onto communicationsThe interference is applied across each independentcircuits, if required, using an external capacitive couplingcircuit (differential mode) and between eachclamp.independent circuit and earth (common mode) via anexternal coupling and switching network. The product isenergised in both normal (quiescent) and tripped modes21.3.6 Surge Immunity Testfor this test, and must not maloperate when theThe Surge Immunity Test simulates interference causedinterference is applied for a 2 second duration.by major power system disturbances such as capacitorbank switching and lightning strikes on overhead lineswithin 5km of the substation. The test waveform has anopen circuit voltage of 4kV for common mode surges and2kV for differential mode surges. The test waveshapeconsists on open circuit of a 1.2/50ms rise/fall time anda short circuit current of 8/20ms rise/fall time. Thegenerator is capable of providing a short circuit test0Timecurrent of up to 2kA, making this test potentiallydestructive. The surges are applied sequentially undersoftware control via dedicated coupling networks in bothdifferential and common modes with the productenergised in its normal (quiescent) state. The product21shall not maloperate during the test, shall still operateFigure 21.2: High Frequency Disturbancewithin specification after the test sequence and shall notTest waveformincur any permanent damage.21.3.5 Fast Transient Test21.3.7 Power Frequency InterferenceThis test simulates the type of interference that is causedThe Fast Transient Test simulates the HV interferencewhen there is a power system fault and very high levelscaused by disconnector operations in GIS substations orof fault current flow in the primary conductors or thebreakdown of the SF6 insulation between conductorsearth grid. This causes 50 or 60Hz interference to beand the earthed enclosure. This interference can eitherinduced onto control and communications circuits.be inductively coupled onto relay circuits or can bedirectly introduced via the CT or VT inputs. It consists ofThere is no international standard for this test, but onea series of 15ms duration bursts at 300ms intervals, eachused by some Utilities is:burst consisting of a train of 50ns wide pulses with verya.500V r.m.s., common modefast (5ns typical) rise times (Figure 21.3), with a peakvoltage magnitude of 4kV.b.250V r.m.s., differential modeNetwork Protection & Automation Guide 375 Chap21-370-397 20/06/02 16:04 Page 376applied to circuits for which power system inputs are not1.current and voltage applied at 90% of setting,connected.(relay not tripped)2.current and voltage applied at 110% of setting,Tests are carried out on each circuit, with the relay in the(relay tripped)following modes of operation:3.main protection and communications functions1.current and voltage applied at 90% of setting,are tested to determine the effect of the discharge(relay not tripped)To pass, the relay shall not maloperate, and shall still2.current and voltage applied at 110% of setting,perform its main functions within the claimed tolerance.(relay tripped)3.main protection and communications functionsare tested to determine the effect of the21.3.9 Conducted and Radiated Emissions TestsinterferenceThese tests arise primarily from the essential protectionThe relay shall not maloperate during the test, and shall stillrequirements of the European Community (EU) directiveperform its main functions within the claimed tolerance.on EMC. These require manufacturers to ensure that anyequipment to be sold in the countries comprising the21.3.8 Electrostatic Discharge TestEuropean Union must not interfere with otherequipment. To achieve this it is necessary to measure theThis test simulates the type of high voltage interferenceemissions from the equipment and ensure that they arethat occurs when an operator touches the relays frontbelow the specified limits.panel after being charged to a high potential. This is exactlythe same phenomenon as getting an electric shock whenConducted emissions are measured only from thestepping out of a car or after walking on a synthetic fibreequipments power supply ports and are to ensure that whencarpet.connected to a mains network, the equipment does not injectinterference back into the network which could adverselyIn this case the discharge is only ever applied to the frontaffect the other equipment connected to the network.panel of the relay, with the cover both on and off. Twotypes of discharges are applied, air discharge and contactRadiated emissions measurements are to ensure that thedischarge. Air discharges are used on surfaces that areinterference radiated from the equipment is not at anormally insulators, and contact discharges are used onlevel that could cause interference to other equipment.surfaces that are normally conducting. IEC 60255-22-2This test is normally carried out on an Open Area Testis the relevant standard this test, for which the testSite (OATS) where there are no reflecting structures orparameters are:sources of radiation, and therefore the measurementsa.cover on: Class 4, 8kV contact discharge, 15kV airobtained are a true indication of the emission spectrumdischargeof the relay. An example of a plot obtained duringb.cover off: Class 3, 6kV contact discharge, 8kV airconducted emissions tests is shown in Figure 21.5.dischargeThe test arrangements for the conducted and radiatedIn both cases above, all the lower test levels are alsoemissions tests are shown in Figure 21.6.tested.When performing these two tests, the relay is in aThe discharge current waveform is shown in Figure 21.4.quiescent condition, that is not tripped, with currents21and voltages applied at 90% of the setting values. This10090is because for the majority of its life, the relay will be inRise Time = 0.7 to 1.0 ns.80the quiescent state and the emission of electromagneticCurrent specified for 30 ns and 60 ns7060interference when the relay is tripped is considered to be50of no significance. Tests are conducted in accordance4030with IEC 60255-25 and EN 50081-2, and are detailed in20Table 21.3.1000 10 20 30 40 50 60 70 80 90Frequency Range Specified Limits Test LimitsTime, ns30 - 230MHz 30dB(V/m) 40dB(V/m)Figure 21.4: ESD Current Waveformat 30m at 10mRadiated230 - 1000MHz 37dB(V/m) 47dB(V/m)at 30m at 10m79dB(V) 79dB(V)The test is performed with single discharges repeated on0.15 - 0.5MHzquasi-peak quasi-peak66dB(V) average 66dB(V) averageeach test point 10 times with positive polarity and 10Conducted73dB(V) 73dB(V)times with negative polarity at each test level. The time0.5 - 30MHzquasi-peak quasi-peak60dB(V) average 60dB(V) averageinterval between successive discharges is greater than 1Table 21.3: Test criteria for Conducted andsecond. Tests are carried out at each level, with the relayRadiated Emissions testsin the following modes of operation: 376 Network Protection & Automation Guide Chap21-370-397 20/06/02 16:04 Page 37710090Quasi-peak limits80Average limits70Typical trace60504030201000.1 1 10 100Frequency, MHzFigure 21.5: Conducted Emissions Test PlotScreened roomAnte-chamberAccess panelE.U.T.Impedance networkSupport/analysisequipment(a) Conducted EMC emissions test arrangement10m21AntennaE.U.T.TurntableEarth Plane(b) Radiated Emissions test arrangement on an OATSE.U.T. - Equipment under testFigure 21.6: EMC test arrangementsNetwork Protection & Automation Guide 377 Chap21-370-397 20/06/02 16:04 Page 37821.3.10 Conducted and Radiated Immunity Testsoperate their radios/mobile phones without fear of relaymaloperation.These tests are designed to ensure that the equipment isimmune to levels of interference that it may be subjectedIEC 60255-22-3 specifies the radiated immunity tests toto. The two tests, conducted and radiated, arise from thebe conducted (ANSI/IEEE C37.90.2 is used for equipmentfact that for a conductor to be an efficient antenna, itbuilt to US standards), with signal levels of:must have a length of at least 1/4 of the wavelength of1.IEC: Class III, 10V/m, 80MHz -1000MHzthe electromagnetic wave it is required to conduct.2.ANSI/IEEE: 35V/m 25MHz - 1000MHz with noIf a relay were to be subjected to radiated interference atmodulation, and again with 100% pulse150kHz, then a conductor length of at leastmodulation=300x106/(150x103x4)IEC 60255-22-6 is used for the conducted immunity test,=500 mwith a test level of:would be needed to conduct the interference. Even withClass III, 10V r.m.s., 150kHz - 80MHz.all the cabling attached and with the longest PCB tracklength taken into account, it would be highly unlikely21.3.11 Power Frequency Magnetic Field Teststhat the relay would be able to conduct radiation of thisfrequency, and the test therefore, would have no effect.These tests are designed to ensure that the equipment isThe interference has to be physically introduced byimmune to magnetic interference. The three tests,conduction, hence the conducted immunity test.steady state, pulsed and damped oscillatory magneticHowever, at the radiated immunity lower frequency limitfield, arise from the fact that for different site conditionsof 80MHz, a conductor length of approximately 1.0m isthe level and waveshape is altered.required. At this frequency, radiated immunity tests can23.3.11.1 Steady state magnetic field testsbe performed with the confidence that the relay willconduct this interference, through a combination of theThese tests simulate the magnetic field that would beattached cabling and the PCB tracks.experienced by a device located within close proximity ofthe power system. Testing is carried out by subjectingAlthough the test standards state that all 6 faces of thethe relay to a magnetic field generated by two inductionequipment should be subjected to the interference, incoils. The relay is rotated such that in each axis it ispractice this is not carried out. Applying interference tosubjected to the full magnetic field strength. IEC 61000-the sides and top and bottom of the relay would have4-6 is the relevant standard, using a signal level of:little effect as the circuitry inside is effectively screenedby the earthed metal case. However, the front and rearLevel 5: 300A/m continuous and 1000A/m short durationof the relay are not completely enclosed by metal and areThe test arrangement is shown in Figure 21.7.therefore not at all well screened, and can be regarded asan EMC hole. Electromagnetic interference whendirected at the front and back of the relay can enterfreely onto the PCBs inside.When performing these two tests, the relay is in aInduction coil21quiescent condition, that is not tripped, with currentsand voltages applied at 90% of the setting values. Thisis because for the majority of its life, the relay will be inE.U.T.Induction coilthe quiescent state and the coincidence of anelectromagnetic disturbance and a fault is considered tobe unlikely.However, spot checks are performed at selectedfrequencies when the main protection and controlGround planefunctions of the relay are exercised, to ensure that it willoperate as expected, should it be required to do so.The frequencies for the spot checks are in generalselected to coincide with the radio frequency broadcastE.U.T. - Equipment under testbands, and in particular, the frequencies of mobilecommunications equipment used by personnel workingFigure 21.7: Power frequency magneticin the substation. This is to ensure that when working infield set-upthe vicinity of a relay, the personnel should be able to 378 Network Protection & Automation Guide Chap21-370-397 20/06/02 16:04 Page 379To pass the steady-state test, the relay shall notopen contacts intended for connection to trippingmaloperate, and shall still perform its main functionscircuits, in accordance with ANSI/IEEE C37.90within the claimed tolerance. During the application of3.1.0kV r.m.s., 50/60Hz for 1 minute across thethe short duration test, the main protection functionnormally open contacts of watchdog orshall be exercised and verified that the operatingchangeover output relays, in accordance with IECcharacteristics of the relay are unaffected.60255-521.3.11.2 Pulsed magnetic fieldThe routine dielectric voltage withstand test time may beThese tests simulate the magnetic field that would beshorter than for the 1 minute type test time, to allow aexperienced by a device located within close proximity ofreasonable production throughput, e.g. for a minimum ofthe power system during a transient fault condition.1 second at 110% of the voltage specified for 1 minute.According to IEC 61000-4-9, the generator for theinduction coils shall produce a 6.4/16s waveshape withtest level 5, 100A/m with the equipment configured as21.4.2 Insulation Withstand for Overvoltagesfor the steady state magnetic field test. The relay shallThe purpose of the High Voltage Impulse Withstand typenot maloperate, and shall still perform its main functionstest is to ensure that circuits and their components willwithin the claimed tolerance during the test.withstand overvoltages on the power system caused by21.3.11.3 Damped oscillatory magnetic fieldlightning. Three positive and three negative high voltageimpulses, 5kV peak, are applied between all circuits andThese tests simulate the magnetic field that would bethe case earth and also between the terminals ofexperienced by a device located within close proximity ofindependent circuits (but not across normally openthe power system during a transient fault condition. IECcontacts). As before, different requirements apply in the61000-4-10 specifies that the generator for the coil shallcase of circuits using D-type connectors.produce an oscillatory waveshape with a frequency of0.1MHz and 1MHz, to give a signal level in accordanceThe test generator characteristics are as specified in IECwith Level 5 of 100A/m, and the equipment shall be60255-5 and are shown in Figure 21.8. No disruptiveconfigured as in Figure 21.7.discharge (i.e. flashover or puncture) is allowed.If it is necessary to repeat either the Dielectric Voltage orHigh Voltage Impulse Withstand tests these should be21.4 PRODUCT SAFETY TYPE TESTScarried out at 75% of the specified level, in accordanceA number of tests are carried out to demonstrate thatwith IEC 60255-5, to avoid overstressing insulation andthe product is safe when used for its intendedcomponents.application. The essential requirements are that therelay is safe and will not cause an electric shock or firehazard under normal conditions and in the presence of asingle fault. A number of specific tests to prove this maybe carried out, as follows.2121.4.1 Dielectric Voltage WithstandDielectric Voltage Withstand testing is carried out as aroutine test i.e. on every unit prior to despatch. Thepurpose of this test is to ensure that the product build isas intended by design. This is done by verifying theclearance in air, thus ensuring that the product is safe tooperate under normal use conditions. The following testsare conducted unless otherwise specified in the productTime5kV peakRise time (10 % to 90 %) = 1.2 sdocumentation:Duration (50 %) = 50 s1.2.0kV r.m.s., 50/60Hz for 1 minute between allFigure 21.8: Test generator characteristicsterminals and case earth and also betweenfor insulation withstand testindependent circuits, in accordance with IEC21.4.3 Single Fault Condition Assessment60255-5. Some communication circuits areexcluded from this test, or have modified testAn assessment is made of whether a single faultrequirements e.g. those using D-type connectorscondition such as an overload, or an open or short circuit,2.1.5kV r.m.s., 50/60Hz for 1 minute across normallyapplied to the product may cause an electric shock or fireNetwork Protection & Automation Guide 379 Chap21-370-397 20/06/02 16:08 Page 380hazard. In the case of doubt, type testing is carried out21.5.2 Humidity Testto ensure that the product is safe.The humidity test is performed to ensure that theproduct will withstand and operate correctly whensubjected to 93% relative humidity at a constant21.4.4 Earth Bonding Impedancetemperature of 40C for 56 days. Tests are performed toClass 1 products that rely on a protective earthensure that the product functions correctly withinspecification after 21 and 56 days. After the test, visualconnection for safety are subjected to an earth bondinginspections are made for any signs of unacceptableimpedance (EBI) type test. This ensures that the earthcorrosion and mould growth.path between the protective earth connection and anyaccessible earthed part is sufficiently low to avoiddamage in the event of a single fault occurring. The test21.5.3 Cyclic Temperature/Humidity Testis conducted using a test voltage of 12V maximum and atest current of twice the recommended maximumThis is a short-term test that stresses the relay bysubjecting it to temperature cycling in conjunction withprotective fuse rating. After 1 minute with the currentflowing in the circuit under test, the EBI shall not exceedhigh humidity.0.1.The test does not replace the 56 day humidity test, but isused for testing extension to ranges or minormodifications to prove that the design is unaffected.21.4.5 CE MarkingThe applicable standard is IEC 60068-2-30 and testA CE mark on the product, or its packaging, shows thatconditions of:compliance is claimed against relevant European+25C 3C and 95% relative humidity/+55C 2C andCommunity directives e.g. Low Voltage Directive95% relative humidity73/23/EEC and Electromagnetic Compatibility (EMC)are used, over the 24 hour cycle shown in Figure 21.9.Directive 89/336/EEC.10096%9095%95%90%8021.5 ENVIRONMENTAL TYPE TESTS80%7015minVarious tests have to be conducted to prove that a relaytemperatureEnd of temperaturecan withstand the effects of the environment in which itriseTimeis expected to work. They consist of: the following tests:+55C1.temperature2.humidity3.enclosure protection4.mechanicalThese tests are described in the following sections.0.5h+28C21+25C3hTime+22C3h12h0.5h 6h21.5.1 Temperature Test24hTemperature tests are performed to ensure that aFigure 21.9: Cyclic temperature/humidityproduct can withstand extremes in temperatures, bothtest profilehot and cold, during transit, storage and operatingFor these tests the relay is placed in a humidity cabinet,conditions. Storage and transit conditions are defined asand energised with normal in-service quantities for thea temperature range of 25C to +70C and operating ascomplete duration of the tests. In practical terms this25C to +55C.usually means energising the relay with currents andvoltages such that it is 10% from the threshold forDry heat withstand tests are performed at 70C for 96operation. Throughout the duration of the test the relay ishours with the relay de-energised. Cold withstand testsmonitored to ensure that no unwanted operations occur.are performed at 40C for 96 hours with the relay de-energised. Operating range tests are carried out with theOnce the relay is removed from the humidity cabinet, itsproduct energised, checking all main functions operateinsulation resistance is measured to ensure that it haswithin tolerance over the specified working temperaturenot deteriorated to below the claimed level. The relay isrange 25C to +55C.then functionally tested again, and finally dismantled to 380 Network Protection & Automation Guide

Chap21-370-397 20/06/02 16:08 Page 381check for signs of component corrosion and growth.The acceptance criterion is that no unwanted operationsshall occur including transient operation of indicatingdevices. After the test the relays insulation resistanceshould not have significantly reduced, and it shouldperform all of its main protection and communicationsfunctions within the claimed tolerance. The relay shouldalso suffer no significant corrosion or growth, andphotographs are usually taken of each PCB and the caseas a record of this.21.5.4 Enclosure Protection TestEnclosure protection tests prove that the casing systemFigure 21.10: Relay undergoing seismic testand connectors on the product protect against the ingressof dust, moisture, water droplets (striking the case at pre-1.2Adefined angles) and other pollutants. An acceptable levelA0.8Aof dust or water may penetrate the case during testing,but must not impair normal product operation, safety orPulse shape (half sine)cause tracking across insulated parts of connectors.+0.2A0-0.2A0.4DD D21.5.5 Mechanical Tests2.5D 2.5DMechanical tests simulate a number of different2.4D =T1mechanical conditions that the product may have to6D =Tendure during its lifetime. These fall into two categories2D - duration of nominal pulsea.response to disturbances while energisedA - peak acceleration of nominal pulseT- minimum time for monitoring of pulse when conventional1b. response to disturbances during transportationshock/bump machine is usedT- asTwhen a vibration generator is used(de-energised state)21Tests in the first category are concerned with theFigure 21.11: Shock/Bump Impulse waveformresponse to vibration, shock and seismic disturbance.The test levels for shock and bump tests are:The tests are designed to simulate normal in-serviceShock response (energised):conditions for the product, for example earthquakes.These tests are performed in all three axes, with the3 pulses, each 10g, 11ms durationproduct energised in its normal (quiescent) state. DuringShock withstand (de-energised):the test, all output contacts are continually monitored21for change using contact follower circuits. Vibration3 pulses, 15g, 11ms durationlevels of 1gn, over a 10Hz-150Hz frequency sweep areBump (de-energised):used. Seismic tests use excitation in a single axis, using1000 pulses, 10g, 16ms durationa test frequency of 35Hz and peak displacements of7.5mm and 3.5mm in the x and y axes respectively belowthe crossover frequency and peak accelerations of 2.0gn21.6 SOFTWARE TYPE TESTSand 1.0gn in these axes above the crossover frequency.Digital and numerical relays contain software toThe second category consists of vibration endurance,implement the protection and measurement functions ofshock withstand and bump tests. They are designed toa relay. This software must be thoroughly tested, tosimulate the longer-term affects of shock and vibrationensure that the relay complies with all specifications andthat could occur during transportation. These tests arethat disturbances of various kinds do not result inperformed with the product de-energised. After theseunexpected results. Software is tested in various stages:tests, the product must still operate within itsspecification and show no signs of permanenta.unit testingmechanical damage. Equipment undergoing a seismicb.integration testingtype test is shown in Figure 21.10, while the waveformfor the shock/bump test is shown in Figure 21.11c.functional qualification testingNetwork Protection & Automation Guide 381 Chap21-370-397 20/06/02 16:08 Page 382The purpose of unit testing is to determine if an21.6.3 Unit Testing Environmentindividual function or procedure implemented usingBoth Dynamic and Static Unit Testing are performed insoftware, or small group of closely related functions, isthe host environment rather than the targetfree of data, logic, or standards errors. It is much easierenvironment. Dynamic Unit Testing uses a test harnessto detect these types of errors in individual units or smallto execute the unit(s) concerned. The test harness isgroups of units than it is in an integrated softwaredesigned such that it simulates the interfaces of thearchitecture and/or system. Unit testing is typicallyunit(s) being tested - both software-software interfacesperformed against the software detailed design and byand software-hardware interfaces - using what arethe developer of the unit(s).known as stubs. The test harness provides the test datato those units being tested and outputs the test resultsIntegration testing typically focuses on these interfacesin a form understandable to a developer. There are manyand also issues such as performance, timings andcommercially available testing tools to automate testsynchronisation that are not applicable in unit testing.harness production and the execution of tests.Integration testing also focuses on stressing thesoftware and related interfaces.21.6.4 Software/Software Integration TestingIntegration testing is black box in nature, i.e. it does nottake into account the structure of individual units. It isSoftware/Software Integration Testing is performed intypically performed against the software architecturalthe host environment. It uses a test harness to simulateand detailed design. The specified software requirementsinputs and outputs, hardware calls and system calls (e.g.would typically also be used as a source for some of thethe target environment operating system).test cases.21.6.5 Software/Hardware Integration Testing21.6.1 Static Unit TestingSoftware/Hardware Integration Testing is performed inStatic Unit Testing (or static analysis as it is often called)the target environment, i.e. it uses the actual targetanalyses the unit(s) source code for complexity, precisionhardware, operating system, drivers etc. It is usuallyperformed after Software/Software Integration Testing.tracking, initialisation checking, value tracking, strongTesting the interfaces to the hardware is an importanttype checking, macro analysis etc. While Static Unitfeature of Software/Hardware Integration Testing.Testing can be performed manually, it is a laborious andTest cases for Integration Testing are typically based onerror prone process and is best performed using athose defined for Validation Testing. However theproprietary automated static unit analysis tool. It isemphasis should be on finding errors and problems.important to ensure that any such tool is configuredPerforming a dry run of the validation testing oftencorrectly and used consistently during development.completes Integration Testing.21.6.2 Dynamic Testing21.6.6 Validation TestingDynamic Testing is concerned with the runtime21The purpose of Validation Testing (also known asbehaviour of the unit(s) being tested and so therefore,Software Acceptance Testing) is to verify that thethe unit(s) must be executed. Dynamic unit testing cansoftware meets its specified functional requirements.be sub-divided into black boxtesting and white boxValidation Testing is performed against the softwaretesting. Black boxtesting verifies the implementationrequirements specification, using the targetof the requirement(s) allocated to the unit(s). It takes noenvironment. In ideal circumstances, someoneaccount of the internal structure of the unit(s) beingindependent of the software development performs thetested. It is only concerned with providing known inputstests. Validation Testing isblack boxin nature, i.e. itand determining if the outputs from the unit(s) aredoes not take into account the internal structure of thecorrect for those inputs. White boxtesting is concernedsoftware. For relays, the non-protection functionswith testing the internal structure of the unit(s) andincluded in the software are considered to be asmeasuring the test coverage, i.e. how much of the codeimportant as the protection functions, and hence testedwithin the unit(s) has been executed during the tests.in the same manner.The objective of the unit testing may, for example, be toEach validation test should have predefined evaluationachieve 100% statement coverage, in which every line ofcriteria, to be used to decide if the test has passed orthe code is executed at least once, or to execute everyfailed. The evaluation criteria should be explicit with nopossible path through the unit(s) at least once.room for interpretation or ambiguity. 382 Network Protection & Automation Guide Chap21-370-397 20/06/02 16:08 Page 38321.6.7 Traceability of Validation TestsPower system simulators can be divided into two types:Traceability of validation tests to software requirementsa.those which use analogue models of a poweris vital. Each software requirement documented in thesystemsoftware requirements specification should have at leastb.those which model the power systemone validation test, and it is important to be able tomathematically using digital simulation techniquesprove this.21.7.1 Use of Power System Analogue Models21.6.8 Software Modifications - Regression TestingFor many years, relays have been tested on analogue modelsRegression Testing is not a type test in its own right. Itof power systems such as artificial transmission lines, or testis the overall name given to the testing performed whenplant capable of supplying significant amounts of currentan existing software product is changed. The purpose of[21.1]. However, these approaches have significantRegression Testing is to show that unintended changeslimitations in the current and voltage waveforms that canto the functionality (i.e. errors and defects) have notbe generated, and are not suitable for automated,been introduced.unattended, testing programmes. While still used on aEach change to an existing software product must belimited basis for testing electromechanical and static relays,considered in its own right. It is impossible to specify aa radically different approach is required for dynamicstandard set of regression tests that can be applied as atesting of numerical relays.catch-all for introduced errors and defects. Eachchange to the software must be analysed to determine21.7.2 Use of Microprocessor Based Simulationwhat risk there might be of unintentional changes to theEquipmentfunctionality being introduced. Those areas of highestrisk will need to be regression tested. The ultimateThe complexity of numerical relays, reliant on softwareregression test is to perform the complete Validationfor implementation of the functions included, dictatesTesting programme again, updated to take account ofsome kind of automated test equipment. The functionsthe changes made.of even a simple numerical overcurrent relay (includingall auxiliary functions) can take several months ofRegression Testing is extremely important. If it is notautomated, 24 hours/day testing to test completely. Ifperformed, there is a high risk of errors being found insuch test equipment was able to apply realistic currentthe field. Performing it will not reduce to zero theand voltage waveforms that closely match those foundchance of an error or defect remaining in the software,on power systems during fault conditions, the equipmentbut it will reduce it. Determining the Regression Testingcan be used either for type testing of individual relaythat is required is made much easier if there isdesigns or of a complete protection scheme designed fortraceability from properly documented softwarea specific application. In recognition of this, a newrequirements through design (again properlygeneration of power system simulators has beendocumented and up to date), coding and testing.developed, which is capable of providing a far moreaccurate simulation of power system conditions than has21.7 DYNAMIC VALIDATION TYPE TESTINGbeen possible in the past. The simulator enables relays21to be tested under a wide range of system conditions,There are two possible methods of dynamically provingrepresenting the equivalent of many years of sitethe satisfactory performance of protection relays orexperience.schemes; the first method is by actually applying faultson the power system and the second is to carry out21.7.2.1 Simulation hardwarecomprehensive testing on a power system simulator.Equipment is now available to provide high-speed, highlyThe former method is extremely unlikely to be used accurate modelling of a section of a power system. Thelead times are lengthy and the risk of damage occurringequipment is based on distributed microprocessor-basedmakes the tests very expensive. It is therefore only usedhardware containing software models of the variouson a very limited basis and the faults applied areelements of a power system, and is shown in Figure 21.12.restricted in number and type. Because of this, a provingThe modules have outputs linked to current and voltageperiod for new protection equipment under servicesources that have a similar transient capability and haveconditions has usually been required. As faults maysuitable output levels for direct connection to the inputsoccur on the power system at infrequent intervals, it canof relays i.e. 110V for voltage and 1A/5A for current.take a number of years before any possible shortcomingsInputs are also provided to monitor the response of relaysare discovered, during which time further installationsunder test (contact closures for tripping, etc.) and thesemay have occurred.inputs can be used as part of the model of the powerNetwork Protection & Automation Guide 383

Chap21-370-397 20/06/02 16:08 Page 384Figure 21.12: Digital power system simulatorfor relay/protection scheme testingtest results being available on completionsystem. The software is also capable of modelling thedynamic response of CTs and VTs accurately.A block schematic of the equipment is shown in FigureWhere it is desired to check the response of a relay or21.13, is based around a computer which calculates andprotection scheme to an actual power system transient,stores the digital data representing the system voltagesthe transient can be simulated using sophisticated powerand currents. The computer controls conversion of thesystems analysis software and the results transferreddigital data into analogue signals, and it monitors anddigitally to the simulator, or the event recorder recordingcontrols the relays being tested.of the transient can be used, in either digital or analogue21.7.2.2 Simulation softwareform as inputs to the simulator model. Output signalconversion involves circuits to eliminate the quantisationUnlike most traditional software used for power systemssteps normally found in conventional D/A conversion.analysis, the software used is suitable for the modellingAnalogue models of the system transducerthe fast transients that occur in the first fewcharacteristics can be interposed between the signalmilliseconds after fault inception. Two very accurateprocessors and the output amplifiers when required.simulation programs are used, one based on time domainand the other on frequency domain techniques. In both21This equipment shows many advantages over traditionalprograms, single and double circuit transmission lines aretest equipment:represented by fully distributed parameter models. Thea.the power system model is capable of reproducingline parameters are calculated from the physicalhigh frequency transients such as travelling wavesconstruction of the line (symmetrical, asymmetrical,b.tests involving very long time constants can betransposed or non-transposed), taking into account thecarried outeffect of conductor geometry, conductor internalimpedance and the earth return path. It also includes,c.it is not affected by the harmonic content, noisewhere appropriate, the frequency dependence of the lineand frequency variations in the a.c. supplyparameters in the frequency domain program. Thed.it is capable of representing the variation in thefrequency dependent variable effects are calculatedcurrent associated with generator faults and powerusing Fast Fourier Transforms and the results areswingsconverted to the time domain. Conventional currenttransformers and capacitor voltage transformers can bee.saturation effects in CTs and VTs can be modelledsimulated.f.a set of test routines can be specified in software andThe fault can be applied at any one point in the system andthen left to run unattended (or with only occasionalmonitoring) to completion, with a detailed record ofcan be any combination of phase to phase or phase 384 Network Protection & Automation Guide Chap21-370-397 20/06/02 16:08 Page 385IALinearID/ACTCurrentBinterpolationconversionmodelamplifiercircuitsICVDUVAEquipmentundertestI/OComputerLinearVKeyboardSub-D/A VoltageCVTBinterpolationsystemconversionmodelamplifiercircuitsVDUVCContactstatusmonitorKeyboardStorageKey :CT- Current transformerSignallingCVT- Capacitor voltage transformerChannelVDU - Visual display unitSimulationCommunications(When required)To second RTDSlink to secondRTDSFigure 21.13: Block diagram of microprocessor-based automated relay test systemto earth, resistive, or non-linear phase to earth arcing faults.power frequencyFor series compensated lines, flashover across a seriesh.the use of direct coupled current amplifiers allowscapacitor following a short circuit fault can be simulated.time constants of any lengthThe frequency domain model is not suitable fori.capable of simulating slow system changesdeveloping faults and switching sequences, therefore thej.reproduces fault currents whose peak amplitudewidely used Electromagnetic Transient Program (EMTP),varies with timeworking in the time domain, is employed in such cases.k.transducer models can be includedIn addition to these two programs, a simulation programbased on lumped resistance and inductance parametersl.automatic testing removes the likelihood ofis used. This simulation is used to represent systems withmeasurement and setting errorslong time constants and slow system changes due, form.two such equipments can be linked together toexample, to power swings.simulate a system model with two relaying points21.7.2.3 Simulator applicationsThe simulator is also used for the production testing of21The simulator is used for checking the accuracy ofrelays, in which most of the advantages listed abovecalibration and performing type tests on a wide range ofapply. As the tests and measurements are madeprotection relays during their development. It has theautomatically, the quality of testing is also greatlyfollowing advantages over existing test methods:enhanced. Further, in cases of suspected malfunction ofa relay in the field under known fault conditions, thea.'state of the art' power system modelling data cansimulator can be used to replicate the power system andbe used to test relaysfault conditions, and conduct a detailed investigationb.freedom from frequency variations and noise orinto the performance of the relay. Finally, complexharmonic content of the a.c. supplyprotection schemes can be modelled, using both therelays intended for use and software models of them asc.the relay under test does not burden the powerappropriate, to check the suitability of the proposedsystem simulationscheme under a wide variety of conditions. To illustrated.all tests are accurately repeatablethis, Figure 21.14(a) shows a section of a particular powersystem modelled. The waveforms of Figure 21.14(b) showe.wide bandwidth signals can be producedthe three phase voltages and currents at the primaries off.a wide range of frequencies can be reproducedVT1 and CT1 for the fault condition indicated in Figureg.selected harmonics may be superimposed on the21.14(a).Network Protection & Automation Guide 385

Chap21-370-397 20/06/02 16:08 Page 386N3GLInfinite bus4G CB3 CT3FF FFCT4 CB434Line 28G9GLR3LR4CT2CB1 CT1CB2 11GFFF12load 1Line 1load 2VT1VT2load 3LR1LR2Relay 1 Relay 2(a) Example power systemVVaVVbVVcIIaIIbFigure 21.14: Example of application studyIIc0 0.080.16 0.24 0.32 0.4 0.560.48(b) Voltages and currents at VT1/CT121.8 PRODUCTION TESTINGProduction testing of protection relays is becoming farmore demanding as the accuracy and complexity of the21products increase. Electronic power amplifiers are usedto supply accurate voltages and currents of high stabilityto the relay under test. The inclusion of a computer in thetest system allows more complex testing to be performedat an economical cost, with the advantage of speed andrepeatability of tests from one relay to another.Figure 21.15 shows a modern computer-controlled testbench. The hardware is mounted in a special rack. Eachunit of the test system is connected to the computer viaFigure 21.15: Modern computer-controlledtest benchan interface bus. Individual test programs for each typeof relay are required, but the interface used is standardBecause software is extensively tested at the type-for all relay types. Control of input waveforms andtesting stage, there is normally no need to check theanalogue measurements, the monitoring of outputcorrect functioning of the software. Checks are limitedsignals and the analysis of test data are performed by thecomputer. A printout of the test results can also beto determining that the analogue and digital I/O isproduced if required.functioning correctly. This is achieved for inputs by 386 Network Protection & Automation Guide Chap21-370-397 20/06/02 16:08 Page 387applying known voltage and current inputs to the relayb.general inspection of the equipment, checking allconnections, wires on relays terminals, labels onunder test and checking that the software has capturedthe correct values. Similarly, digital outputs areterminal boards, etc.exercised by using test software to actuate each outputc.insulation resistance measurement of all circuitsand checking that the correct output is energised.d.perform relay self-test procedure and externalProvided that appropriate procedures are in place tocommunications checks on digital/numerical relaysensure that only type-tested software is downloaded,there is no need to check the correct functioning of thee.test main current transformerssoftware in the relay. The final step is to download thef.test main voltage transformerssoftware appropriate to the relay and store it in theEPROM fitted in the relay.g.check that protection relay alarm/trip settingshave been entered correctlyh.tripping and alarm circuit checks to prove correct21.9 COMMISSIONING TESTSfunctioningInstallation of a protection scheme at site creates aIn addition, the following checks may be carried out,number of possibilities for errors in the implementationdepending on the factors noted earlier.of the scheme to occur. Even if the scheme has beeni.secondary injection test on each relay to provethoroughly tested in the factory, wiring to the CTs andoperation at one or more setting valuesVTs on site may be incorrectly carried out, or theCTs/VTs may have been incorrectly installed. The impactj.primary injection tests on each relay to proveof such errors may range from simply being a nuisancestability for external faults and to determine the(tripping occurs repeatedly on energisation, requiringeffective current setting for internal faults (essentialinvestigation to locate and correct the error(s)) throughfor some types of electromechanical relays)to failure to trip under fault conditions, leading to majork.testing of protection scheme logicequipment damage, disruption to supplies and potentialhazards to personnel. The strategies available to removeThis section details the tests required to cover itemsthese risks are many, but all involve some kind of testing(a)(g) above. Secondary injection test equipment isat site.covered in Section 21.10 and Section 21.11 details thesecondary injection that may be carried out. SectionCommissioning tests at site are therefore invariably21.12 covers primary injection testing, and Section 21.13performed before protection equipment is set to work.details the checks required on any logic involved in theThe aims of commissioning tests are:protection scheme. Finally, Section 21.14 details the tests1.to ensure that the equipment has not beenrequired on alarm/tripping circuits tripping/alarmdamaged during transit or installationcircuits.2.to ensure that the installation work has beencarried out correctly21.9.1 Insulation Tests3.to prove the correct functioning of the protectionAll the deliberate earth connections on the wiring to be21scheme as a wholetested should first be removed, for example earthingThe tests carried out will normally vary according to thelinks on current transformers, voltage transformers andprotection scheme involved, the relay technology used,d.c. supplies. Some insulation testers generate impulsesand the policy of the client. In many cases, the testswith peak voltages exceeding 5kV. In these instancesactually conducted are determined at the time ofany electronic equipment should be disconnected whilecommissioning by mutual agreement between thethe external wiring insulation is checked.clients representative and the commissioning team.The insulation resistance should be measured to earthHence, it is not possible to provide a definitive list ofand between electrically separate circuits. The readingstests that are required during commissioning. Thisare recorded and compared with subsequent routinesection therefore describes the tests commonly carriedtests to check for any deterioration of the insulation.out during commissioning.The insulation resistance measured depends on theThe following tests are invariably carried out, since theamount of wiring involved, its grade, and the siteprotection scheme will not function correctly if faults exist.humidity. Generally, if the test is restricted to onea.wiring diagram check, using circuit diagramscubicle, a reading of several hundred megohms should beshowing all the reference numbers of theobtained. If long lengths of site wiring are involved, theinterconnecting wiringreading could be only a few megohms.Network Protection & Automation Guide 387 Chap21-370-397 20/06/02 16:08 Page 38821.9.2 Relay Self-Test Procedurerobust moving coil, permanent magnet, centre-zero type.A low voltage battery is used, via a single-pole push-Digital and numerical relays will have a self-testbutton switch, to energise the primary winding. Onprocedure that is detailed in the appropriate relayclosing the push-button, the d.c. ammeter, A, should givemanual. These tests should be followed to determine ifa positive flick and on opening, a negative flick.the relay is operating correctly. This will normally involvechecking of the relay watchdog circuit, exercising all21.9.3.2 Magnetisation Curvedigital inputs and outputs and checking that the relaySeveral points should be checked on each currentanalogue inputs are within calibration by applying a testtransformer magnetisation curve. This can be done bycurrent or voltage. For these tests, the relay outputs areenergising the secondary winding from the local mainsnormally disconnected from the remainder of thesupply through a variable auto-transformer while theprotection scheme, as it is a test carried out to proveprimary circuit remains open; see Figure 21.17. Thecorrect relay, rather than scheme, operation.characteristic is measured at suitable intervals of appliedUnit protection schemes involve relays that need tovoltage, until the magnetising current is seen to rise verycommunicate with each other. This leads to additionalrapidly for a small increase in voltage. This indicates thetesting requirements. The communications pathapproximate knee-point or saturation flux level of thebetween the relays is tested using suitable equipment tocurrent transformer. The magnetising current shouldensure that the path is complete and that the receivedthen be recorded at similar voltage intervals as it issignal strength is within specification. Numerical relaysreduced to zero.may be fitted with loopback test facilities that enableCare must be taken that the test equipment is suitablyeither part of or the entire communications link to berated. The short-time current rating must be in excess oftested from one end.the CT secondary current rating, to allow for theAfter completion of these tests, it is usual to enter themeasurement of the saturation current. This will be inrelay settings required. This can be done manually viaexcess of the CT secondary current rating. As thethe relay front panel controls, or using a portable PC andmagnetising current will not be sinusoidal, a moving ironsuitable software. Whichever method is used, a check byor dynamometer type ammeter should be used.a second person that the correct settings have been usedIt is often found that current transformers withis desirable, and the settings recorded. Programmablesecondary ratings of 1A or less have a knee-point voltagescheme logic that is required is also entered at this stage.higher than the local mains supply. In these cases, astep-up interposing transformer must be used to obtainthe necessary voltage to check the magnetisation curve.21.9.3 Current Transformer TestsThe following tests are normally carried out prior toenergisation of the main circuits.Test plug isolatingVariable transformercurrent transformers250V 8Afrom relay coils21.9.3.1 Polarity checkABCAPPPP21250V_ +To relay21SSa.c. supplyV21PcoilsS11Step-up transformerPS22if requiredMain circuitbreaker open_ +AFigure 21.17: Testing current transformermagnetising curve21.9.4 Voltage Transformer TestsVoltage transformers require testing for polarity andFigure 21.16: Current transformerpolarity checkphasing.21.9.4.1 Polarity checkEach current transformer should be individually tested toverify that the primary and secondary polarity markingsThe voltage transformer polarity can be checked usingare correct; see Figure 21.16. The ammeter connected tothe method for CT polarity tests. Care must be taken tothe secondary of the current transformer should be aconnect the battery supply to the primary winding, with 388 Network Protection & Automation Guide Chap21-370-397 20/06/02 16:08 Page 389the polarity ammeter connected to the secondaryCorrect phasing should be further substantiated whenwinding. If the voltage transformer is of the capacitorcarrying out on load tests on any phase-angle sensitivetype, then the polarity of the transformer at the bottomrelays, at the relay terminals. Load current in a knownof the capacitor stack should be checked.phase CT secondary should be compared with theassociated phase to neutral VT secondary voltage. The21.9.4.2 Ratio checkphase angle between them should be measured, andThis check can be carried out when the main circuit isshould relate to the power factor of the system load.first made live. The voltage transformer secondaryIf the three-phase voltage transformer has a broken-voltage is compared with the secondary voltage showndelta tertiary winding, then a check should be made ofon the nameplate.the voltage across the two connections from the broken21.9.4.3 Phasing checkdeltaVandV, as shown in Figure 21.18. With theNLrated balanced three-phase supply voltage applied to theThe secondary connections for a three-phase voltagevoltage transformer primary windings, the broken-deltatransformer or a bank of three single-phase voltagevoltage should be below 5V with the rated burdentransformers must be carefully checked for phasing.connected.With the main circuit alive, the phase rotation is checkedusing a phase rotation meter connected across the threephases, as shown in Figure 21.18. Provided an existing21.9.5 Protection Relay Setting Checksproven VT is available on the same primary system, andthat secondary earthing is employed, all that is nowAt some point during commissioning, the alarm and tripnecessary to prove correct phasing is a voltage checksettings of the relay elements involved will require to bebetween, say, both A phase secondary outputs. Thereentered and/or checked. Where the complete scheme isshould be nominally little or no voltage if the phasing isengineered and supplied by a single contractor, thecorrect. However, this test does not detect if the phasesettings may already have been entered prior to despatchsequence is correct, but the phases are displaced by 120from the factory, and hence this need not be repeated.from their correct position, i.e. phase A occupies theThe method of entering settings varies according to theposition of phaseCor phaseBin Figure 21.18. This canrelay technology used. For electromechanical and staticbe checked by removing the fuses from phasesBandCrelays, manual entry of the settings for each relay(say) and measuring the phase-earth voltages on theelement is required. This method can also be used forsecondary of the VT. If the phasing is correct, only phasedigital/numerical relays. However, the amount of data tobe entered is much greater, and therefore it is usual toAshould be healthy, phasesBandCshould have only asmall residual voltage.use appropriate software, normally supplied by themanufacturer, for this purpose. The software also makesthe essential task of making a record of the data enteredAmuch easier.BCOnce the data has been entered, it should be checked forAcompliance with the recommended settings ascalculated from the protection setting study. Whereappropriate software is used for data entry, the checksV121C Bcan be considered complete if the data is checked priorto download of the settings to the relay. Otherwise, aV2check may required subsequent to data entry byinspection and recording of the relay settings, or it mayVbe considered adequate to do this at the time of dataNentry. The recorded settings form an essential part of theVcommissioning documentation provided to the client.VLV221.10 SECONDARY INJECTION TEST EQUIPMENTSecondary injection tests are always done prior toVprimary injection tests. The purpose of secondary1injection testing is to prove the correct operation of theprotection scheme that is downstream from the inputs toAB CPhase rotationthe protection relay(s). Secondary injection tests aremeteralways done prior to primary injection tests. This isFigure 21.18: Voltage transformerphasing checkNetwork Protection & Automation Guide 389

Chap21-370-397 20/06/02 16:11 Page 390because the risks during initial testing to the LV side ofthe equipment under test are minimised. The primary(HV) side of the equipment is disconnected, so that nodamage can occur. These tests and the equipmentnecessary to perform them are generally described in themanufacturer's manuals for the relays, but brief detailsare given below for the main types of protection relays.21.10.1 Test Blocks/Plugs for SecondaryInjection EquipmentIt is common practice to provide test blocks or testFigure 21.19: Modern test block/plugssockets in the relay circuits so that connections canreadily be made to the test equipment without21.10.2 Secondary Injection Test Setsdisturbing wiring. Test plugs of either multi-finger orsingle-finger design (for monitoring the current in oneThe type of the relay to be tested determines the type ofCT secondary circuit) are used to connect test equipmentequipment used to provide the secondary injectionto the relay under test.currents and voltages. Many electromechanical relayshave a non-linear current coil impedance when the relayThe top and bottom contact of each test plug finger isoperates and this can cause the test current waveform toseparated by an insulating strip, so that the relay circuitsbe distorted if the injection supply voltage is fed directlycan be completely isolated from the switchgear wiringto the coil. The presence of harmonics in the currentwhen the test plug is inserted. To avoid open-circuitingwaveform may affect the torque of electromechanicalCT secondary terminals, it is therefore essential that CTrelays and give unreliable test results, so some injectionshorting jumper links are fitted across all appropriatetest sets use an adjustable series reactance to control thelive side terminals of the test plug BEFORE it is inserted.current. This keeps the power dissipation small and theWith the test plug inserted in position, all the testequipment light and compact.circuitry can now be connected to the isolated relayside test plug terminals. Some modern test blocksMany test sets are portable and include precisionincorporate the live-side jumper links within the blockammeters and voltmeters and timing equipment. Testand these can be set to the closed or open position assets may have both voltage and current outputs. Theappropriate, either manually prior to removing the coverformer are high-voltage, low current outputs for useand inserting the test plug, or automatically uponwith relay elements that require signal inputs from a VTremoval of the cover. Removal of the cover also exposesas well as a CT. The current outputs are high-current,the colour-coded face-plate of the block, clearlylow voltage to connect to relay CT inputs. It isindicating that the protection scheme is not in service,important, however, to ensure that the test set currentand may also disconnect any d.c. auxiliary supplies usedoutputs are true current sources, and hence are notfor powering relay tripping outputs.affected by the load impedance of a relay elementcurrent coil. Use of a test set with a current output that21Withdrawing the test plug immediately restores theis essentially a voltage source can give rise to seriousconnections to the main current transformers andproblems when testing electromechanical relays. Anyvoltage transformers and removes the test connections.significant impedance mismatch between the output ofReplacement of the test block cover then removes thethe test set and the relay current coil during relayshort circuits that had been applied to the main CToperation will give rise to a variation in current from thatsecondary circuits. Where several relays are used in adesired and possible error in the test results. The relayprotection scheme, one or more test blocks may be fittedoperation time may be greater than expected (never lesson the relay panel enabling the whole scheme to bethan expected) or relay chatter may occur. It is quitetested, rather than just one relay at a time.common for such errors to only be found much later,Test blocks usually offer facilities for the monitoring andafter a fault has caused major damage to equipmentsecondary injection testing of any power systemthrough failure of the primary protection to operate.protection scheme. The test block may be used eitherFailure investigation then shows that the reason for thewith a multi-fingered test plug to allow isolation andprimary protection to operate is an incorrectly set relay,monitoring of all the selected conductor paths, or with adue in turn to use of a test set with a current outputsingle finger test plug that allows the currents onconsisting of a voltage-source when the relay was lastindividual conductors to be monitored. A modern testtested. Figure 21.20 shows typical waveforms resultingblock and test plugs are illustrated in Figure 21.19.from use of test set current output that is a voltage 390 Network Protection & Automation Guide Chap21-370-397 20/06/02 16:11 Page 391V relay/sourceSaturation level ofmagnetic circuit (current)limited only by D.C.resistance ofTimerelay coilRelay with saturationof CDG magnetic circuit(phase shift from CDGinductive load shown).a) Relay current coil waveform distorted due to use of voltage sourceSinusoidal CURRENT whenchanging impedance of relayis swamped out by highsource impedanceTimeTypical VOLTAGE waveformappearing across relaycurrent coils with sinusoidal Iabove the relay setting (10 x shown).21b) Undistorted relay current coil current distorted due to use of current sourceFigure 21.20: Relay current coil waveforms3-phase output set. Much greater precision in thesource the distorted relay coil current waveform givessetting of the magnitudes and phase angles is possible,rise to an extended operation time compared to thecompared to traditional test sets. Digital signals toexpected value.exercise the internal logic elements of the relays mayModern test sets are computer based. They comprise aalso be provided. The alarm and trip outputs of the relayPC (usually a standard laptop PC with suitable software)are connected to digital inputs on the PC so that correctand a power amplifier that takes the low-level outputsoperation of the relay, including accuracy of the relayfrom the PC and amplifies them into voltage and currenttripping characteristic can be monitored and displayedsignals suitable for application to the VT and CT inputs ofon-screen, saved for inclusion in reports generated later,the relay. The phase angle between voltage and currentor printed for an immediate record to present to theoutputs will be adjustable, as also will the phase anglesclient. Optional features may include GPS timebetween the individual voltages or currents making up asynchronising equipment and remote-located amplifiersNetwork Protection & Automation Guide 391

Chap21-370-397 20/06/02 16:11 Page 392to facilitate testing of unit protection schemes, anddigital I/O for exercising the programmable scheme logicof modern relays.The software for modern test sets is capable of testingthe functionality of a wide variety of relays, andconducting a set of tests automatically. Such sets easethe task of the commissioning engineer. The softwarewill normally offer options for testing, ranging from atest carried out at a particular point on the characteristicto complete determination of the tripping characteristicautomatically. This feature can be helpful if there is anyreason to doubt that the relay is operating correctly withthe tripping characteristic specified. Figure 21.21illustrates a modern PC-based test set.Traditional test sets use an arrangement of adjustabletransformers and reactors to provide control of currentand voltage without incurring high power dissipation.Some relays require adjustment of the phase betweenthe injected voltages and currents, and so phase shiftingtransformers may be used. Figure 21.22 shows thecircuit diagram of a traditional test set suitable forovercurrent relay resting, while Figure 21.23 shows thecircuit diagram for a test set for directional/distanceFigure 21.21: Modern PC-based secondaryinjection test setrelays. Timers are included so that the response time ofthe relay can be measured.testing will be largely determined by the client21.11 SECONDARY INJECTION TESTINGspecification and relay technology used, and may rangefrom a simple check of the relay characteristic at a singleThe purpose of secondary injection testing is to checkpoint to a complete verification of the trippingthat the protection scheme from the relay inputcharacteristics of the scheme, including the response toterminals onwards is functioning correctly with thetransient waveforms and harmonics and checking ofsettings specified. This is achieved by applying suitablerelay bias characteristics. This may be important wheninputs from a test set to the inputs of the relays andthe protection scheme includes transformers and/orchecking if the appropriate alarm/trip signals occur atgenerators.the relay/control room/CB locations. The extent of21ACoarseRangecontroladjusting CTreactorKKI21Fine controlI>Relay250Vvariablecoila.c. supplytransformerStartStoptimertimerInjectionRelay short-circuitingBackingMediumtransformerswitchtransformercontrol10% controlreactorRelay current,I= Ammeter reading (A)KxK12Figure 21.22: Circuit diagram of traditional test set for overcurrent relays 392 Network Protection & Automation Guide

Chap21-370-397 20/06/02 16:11 Page 393Fault A-NSupply switchX( )AB440V22.5C3 phasepN4 wire supply20.0ChokeVar ia ble17.5tr ansf orme rRelay15.0c ont rol12.5adjusting CT10.0APAA7.5P AA>Vvoltage element440/110V5.0phasepVariable transformer for current controlshiftingg2.5transformerV VoltmeterTo other voltage0.0A Ammeterelements-2.5PA Phase angleof relayyunder testmeter-5.0(if required)-7.5Figure 21.23: Circuit diagram for traditional-10.0test set for directional/distance relays-15.0-10.0-5.0 0.05.0 10.0 15.0 R( )Figure 21.24: Distance relay zone checkingusing search technique and tolerance bandsThe testing should include any scheme logic. If the logicis implemented using the programmable scheme logicfacilities available with most digital or numerical relays,Xappropriate digital inputs may need to be applied andoutputs monitored (see Section 21.13). It is clear that aZmodern test set can facilitate such tests, leading to anreduced time required for testing.*21.11.1 Schemes using Digital or NumericalRRelay TechnologyPSB-ZoneThe policy for secondary injection testing varies widely.In some cases, manufacturers recommend, and clientsaccept, that if a digital or numerical relay passes its self-Figure 21.25: Testing of power swingtest, it can be relied upon to operate at the settings usedblocking element discrete pointsand that testing can therefore be confined to those partsof the scheme external to the relay. In such cases,secondary injection testing is not required at all. Moreoften, it is required that one element of each relay(usually the simplest) is exercised, using a secondaryinjection test set, to check that relay operation occurs at21the conditions expected, based on the setting of therelay element concerned. Another alternative is for thecomplete functionality of each relay to be exercised. Thisis rarely required with a digital or numerical relay,probably only being carried out in the event of asuspected relay malfunction.To illustrate the results that can be obtained, FigureFigure 21.26: Simulated power swing waveform21.24 shows the results obtained by a modern test set21.11.2 Schemes usingwhen determining the reach settings of a distance relayElectromechanical/Static Relay Technologyusing a search technique. Another example is the testingof the Power Swing blocking element of a distance relay.Schemes using single function electromechanical orFigure 21.25 illustrates such a test, based on usingstatic relays will usually require each relay to bediscrete impedance points. This kind of test may not beexercised. Thus a scheme with distance and back-upadequate in all cases, and test equipment may have theovercurrent elements will require a test on each of theseability to generate the waveforms simulating a powerfunctions, thereby taking up more time than if a digitalswing and apply them to the relay (Figure 21.26).or numerical relay is used. Similarly, it may be importantNetwork Protection & Automation Guide 393 Chap21-370-397 20/06/02 16:11 Page 394to check the relay characteristic over a range of inputof VTs/CTs may not then be discovered until eitherspurious tripping occurs in service, or more seriously,currents to confirm parameters for an overcurrent relayfailure to trip on a fault. This hazard is much reducedsuch as:where digital/numerical relays are used, since the currenti.the minimum current that gives operation at eachand voltage measurement/display facilities that exist incurrent settingsuch relays enable checking of relay input values againstii.the maximum current at which resetting takesthose from other proven sources. Many connection/wiringplaceerrors can be found in this way, and by isolatingtemporarily the relay trip outputs, unwanted trips can beiii.the operating time at suitable values of currentavoided.iv.the time/current curve at two or three points withPrimary injection testing is, however, the only way tothe time multiplier setting TMS at 1prove correct installation and operation of the whole ofv.the resetting time at zero current with the TMS at 1a protection scheme. As noted in the previous section,primary injection tests are always carried out afterSimilar considerations apply to distance and unitsecondary injection tests, to ensure that problems areprotection relays of these technologies.limited to the VTs and CTs involved, plus associatedwiring, all other equipment in the protection schemehaving been proven satisfactory from the secondary21.11.3 Test Circuits for Secondary Injection Testinginjection tests.The test circuits used will depend on the type of relayand test set being used. Unless the test circuits aresimple and obvious, the relay commissioning manual will21.12.1 Test Facilitiesgive details of the circuits to be used. Commonly usedAn alternator is the most useful source of power fortest circuits can also be found in Chapter 23 of referenceproviding the heavy current necessary for primary[21.1]. When using the circuits in this reference, suitableinjection. Unfortunately, it is rarely available, since itsimplifications can easily be made if digital or numericalrequires not only a spare alternator, but also sparerelays are being tested, to allow for their built-inbusbars capable of being connected to the alternator andmeasurement capabilities external ammeters andcircuit under test. Therefore, primary injection is usuallyvoltmeters may not be required.carried out by means of a portable injection transformerAll results should be carefully noted and filed for record(Figure 21.27), arranged to operate from the local mainspurposes. Departures from the expected results must besupply and having several low voltage, heavy currentthoroughly investigated and the cause determined. Afterwindings. These can be connected in series or parallelrectification of errors, all tests whose results may haveaccording to the current required and the resistance ofbeen affected (even those that may have given correctthe primary circuit. Outputs of 10V and 1000A can beobtained. Alternatively, modern PC-controlled test setsresults) should be repeated to ensure that the protectionscheme has been implemented according tohave power amplifiers capable of injecting currents up tospecification.about 200A for a single unit, with higher current ratingsbeing possible by using multiple units in parallel.2121.12 PRIMARY INJECTION TESTSThis type of test involves the entire circuit; currenttransformer primary and secondary windings, relay coils,trip and alarm circuits, and all intervening wiring arechecked. There is no need to disturb wiring, whichobviates the hazard of open-circuiting currentAtransformers, and there is generally no need for any250V a.c.switching in the current transformer or relay circuits.supplyThe drawback of such tests is that they are timeconsuming and expensive to organise. Increasingly,reliance is placed on all wiring and installation diagramsbeing correct and the installation being carried out asInjection transformerVariable transformerper drawings, and secondary injection testing being250/10 + 10 + 10 + 10V40A10kVAcompleted satisfactorily. Under these circumstances, theprimary injection tests may be omitted. However, wiringFigure 21.27: Traditional primaryinjection test seterrors between VTs/CTs and relays, or incorrect polarity 394 Network Protection & Automation Guide Chap21-370-397 20/06/02 16:11 Page 395If the main current transformers are fitted with testthe residual circuit, or relay display, will give a reading ofwindings, these can be used for primary injection insteada few milliamperes with rated current injected if theof the primary winding. The current required for primarycurrent transformers are of correct polarity. A readinginjection is then greatly reduced and can usually beproportional to twice the primary current will beobtained using secondary injection test equipment.obtained if they are of wrong polarity. Because of this, aUnfortunately, test windings are not often provided,high-range ammeter should be used initially, for examplebecause of space limitations in the main currentone giving full-scale deflection for twice the ratedtransformer housings or the cost of the windings.secondary current. If an electromechanical earth-faultrelay with a low setting is also connected in the residualcircuit, it is advisable to temporarily short-circuit its21.12.2 CT Ratio Checkoperating coil during the test, to prevent possibleoverheating. The single-phase injection should beCurrent is passed through the primary conductors andcarried out for each pair of phases.measured on the test set ammeter, A1 in Figure 21.28.The secondary current is measured on the ammeter A2 orTemporarythree-phaserelay display, and the ratio of the value on A1 to that onshort circuitA2 should closely approximate to the ratio marked onAthe current transformer nameplate.Primary250V a.c.injectionsupplytest setBRelayABCTest plugCinsulationuATemporaryshort circuitFigure 21.29: Polarity check on main currenttransformersRelayPP121.12.4 Primary Injection Testing of Relay ElementsS1As with secondary injection testing, the tests to becarried out will be those specified by the client, and/orthose detailed in the relay commissioning manual.S2Digital and numerical relays usually require far fewerPP2tests to prove correct operation, and these may beRelay or test blockcontact fingersrestricted to observations of current and voltage on theAArelay display under normal load conditions.1Primary injection21.13 TESTING OF PROTECTION SCHEME LOGIC21test setProtection schemes often involve the use of logic to250Va.c. supplydetermine the conditions under which designated circuitbreakers should be tripped. Simple examples of suchFigure 21.28: Current transformer ratio checklogic can be found in Chapters 9-14. Traditionally, thislogic was implemented by means of discrete relays,separate from the relays used for protection. Such21.12.3 CT Polarity Checkimplementations would occur where electromechanicalIf the equipment includes directional, differential oror static relay technology is used. However, digital andearth fault relays, the polarity of