NETA Handbook Series I, Protective Relay Vol 1-PDF

The Anatomy of a Shermco Field Service Engineer Reliable and Responsive Part of a large international team available 24/7

Back by the Best From the field, to our full-service rotating apparatus services

NETA-Certified Qualified to perform to the highest standards

Protective Relaying Handbook

Trusted Advisor Offering independent service across all brands

Safety First On record as an industry leader in safe work practices

Volume 1

Cutting-Edge Trained from utility transmission to low-voltage distribution

Our experience in commercial, industrial, generation and specialized sectors, such as oil refineries, pulp and paper, steel and wind power, uniquely qualifies us to handle the largest, most complex assignments. Our people are trained on the latest technologies and safety practices. And, we support them with full-service rotating apparatus services and a large fleet of service vehicles, ready to respond 24/7. Corporate Headquarters | 2425 East Pioneer Drive, Irving, Texas 75061 p. 972.793.5523, 888.SHERMCO | f. 972.793.5542 | www.shermco.com

Why trust your vital power services to anyone less? Count on an industry leader. Visit www.shermco.com and learn more today.

Published by InterNational Electrical Testing Association

Odds are, you won’t find a job anywhere that a Shermco field service engineer can’t handle. As one of the world’s most respected electrical maintenance and testing companies, Shermco delivers A+ service from utility transmission to low-voltage distribution.

Published by InterNational Electrical Testing Association

Protective Relaying Handbook Volume 1

Published by InterNational Electrical Testing Association

Protective Relaying Handbook Volume 1

Table of Contents Dynamic-State Relay Testing .....................................................................................1 A. T. Giuliante

Introduction to Dynamic Testing .............................................................................4 D. L. Tierney

Through-Fault Testing — the Ultimate Test for Protection Schemes Prior to Energizing ........8 Roderic L. Hageman

Automated Test Point Calculations for Electronic Relay Testing and Coordination .......11

Lonnie C. Lindell and Steven R. Potter

Test & Maintenance Tips for Protective Relays ..........................................................14 Scott Cooper

Using 1op Characteristics to Troubleshoot Transformer Differential Relay Misoperation ...........................................................16 Michael Thompson and James R. Closson

Motor Protection Fundamentals ............................................................................27 Bernie Moisey

Meaningful Testing of Numerical Multifunction Protection Schemes ........................30 Jay Gosalia

Using Dynamic Testing Techniques for Commissioning and Routine Testing of Motor Protection Relays ....................................................................................35 Benton Vandiver III, P.E.

Commissioning Numerical Relays — Part One ..........................................................37 James R. Closson and Mike Young

Steady State vs. Dynamic Testing ............................................................................44 Steven Stade

Published by

InterNational Electrical Testing Association 3050 Old Centre Avenue, Suite 102, Portage, Michigan 49024

269.488.6382

www.netaworld.org

Protective Relaying Handbook Volume 1

Table of Contents (continued) Dynamic State and Other Advanced Testing Methods for Protection Relays Address Changing Industry Needs ...........................................46 Kenneth Tang

Acceptance Testing a Synch Circuit ..........................................................................51 Steven C. Reed, P.E.

Partial Differential Relaying ...................................................................................53 Baldwin Bridger, P.E.

Modern Relays and Software Provide Valuable Tools for Analysis ..............................54 Scott Cooper

Understanding and Analyzing Event Report Information ...........................................57 David Costello

NOTICE AND DISCLAIMER NETA Technical Papers and Articles are published by the InterNational Electrical Testing Association. Opinions, views, and conclusions expressed in articles herein are those of the authors and not necessarily those of NETA. Publication herein does not constitute or imply any endorsement of any opinion, product, or service by NETA, its directors, officers, members, employees, or agents (hereinafter “NETA”). All technical data in this publication reflects the experience of individuals using specific tools, products equipment, and components under specific conditions and circumstances which may or may not be fully reported and over which NETA has neither exercised nor reserved control. Such data has not been independently tested or otherwise verified by NETA. NETA makes no endorsement, representation, or warranty as to any opinion, product, or service referenced in this publication. NETA expressly disclaims any and all liability to any consumer, purchaser, or any other person using any product or service referenced herein for any injuries or damages of any kind whatsoever, including, but not limited to, any consequential, special incidental, direct, or indirect damages. NETA further disclaims any and all warranties, express or implied including, but not limited to, any implied warranty or merchantability or any implied warranty of fitness for a particular purpose. Please Note: All biographies of authors and presenters contained herein are reflective of the professional standing of these individuals at the time the articles were originally published. Titles, companies, and other factors may have changed since the original publication date. Copyright © 2009 by InterNational Electrical Testing Association, all rights reserved. No part of this publication may be reproduced in any form or by any means, electronic or mechanical, without permission in writing from the publisher.

1

Protective Relaying Handbook — Volume 1

Dynamic-State Relay Testing NETA World, Winter1999-2000 Issue by A. T. Giuliante ATG Exodus

The traditional method of testing individual relay functions using steady-state calibrations is no longer a viable test method for testing modern multifunction relays. Today, relay designs include innovative numerical techniques that enhance relay performance by combining a number of measuring criteria and by optimizing the relay’s operation for power system conditions. If these relays are tested under the pseudo power system conditions created by steady-state testing, problems in testing and understanding the relay’s operation can occur. In addition, the time for testing individual elements would be excessive because of the time required to reconfigure each individual element tested.

Relay Test Methods A report from IEEE, Relay Per formance Testing, discusses the methods of steady-state, dynamic-state, and transient testing of modern relays. A steady-state test is defined as applying phasors to determine relay settings by slowly varying relay input. Obviously, this test method does not represent power system faults. Dynamic-state test is defined as simultaneously applying fundamental frequency components of voltage and current that represent power system states of prefault, fault, and postfault. Utilizing this technique results in faster relay testing because, in most cases, relay elements do not need to be disabled in order to test a relay function. Transient testing is defined as simultaneously applying fundamental and nonfundamental frequency components of voltage and current that represent power system conditions obtained from digital fault recorders (DFR) or electromagnetic transient programs (EMTP).

Dynamic Relay Testing Dynamic relay testing means testing under true simulated power system conditions. Depending on the level of testing required, test values can be easily calculated with PC-based short circuit or EMTP programs. For dynamic-state testing, a short-circuit program would be used to calculate the

fundamental component of voltage and current values for prefault and fault conditions. For transient simulations, an EMTP program would be used to create waveforms that represent the fault condition. Dynamic-state testing and transient simulations provide a faster and more meaningful way to test relays and relay systems. These techniques provide the user with a far better understanding of how the relay system performs and can aid both relay application and test engineers in evaluating relay operations. Dynamic-state testing is based on a power system model that is used to simulate different events selected according to the application. Events are played back through power system simulators that also monitor scheme performance. Each event is modeled to simulate conditions for the tested relay circuit but only for the time period needed to test.

Why Use Dynamic-State Testing? Modern relay systems are multifunction digital devices that are designed to provide complete protection for a power system component. Some of the newer designs have over 2,000 setting possibilities and require extensive configuration and setting procedures. The traditional method of testing individual steady-state calibrations, one at a time, is no longer a viable method because of the excessive time it would require to reconfigure for each individual element tested. In addition, traditional test methods were designed on the assumption that users did not have test equipment for testing relays under power system conditions. So traditional test procedures were developed using basic test equipment components such as variacs, phase shifters, and load boxes. With today’s modern test equipment, power system conditions can easily be simulated. By making a profile of the operation of the scheme, malfunctions can be found faster because it is easier to identify the changes in areas that do not operate the way they are expected.

2

Protective Relaying Handbook — Volume 1

Advantages of Dynamic-State Testing

What Is Needed for Dynamic-State Testing?

Some of the advantages offered with dynamic-state testing as compared to traditional test methods are:

The test method involves testing the complete scheme with dynamic-state simulations that model the power system the relay scheme will protect. Computational programs such as One Bus, One Liner, CAPE and other mathematical calculation tools such as spreadsheets and Mathcad can be used to model the power system in order to derive the fault voltages and currents for the power system event.

• Complete relay scheme tests For each simulated power system event, the performance of the complete relay scheme is tested. The high power capability of power system simulators allows the user to test the complete relay scheme. This provides a faster way to test relays since relay settings or configuration need not be changed as they would if individual circuits were tested one at a time. The performance of all scheme responses, including unfaulted phase units, can be evaluated since the model and simulators generate three-phase wye voltages and currents. This allows the accurate modeling of power system events. In addition, if the relay scheme includes programmable logic, simulated events which test how the complete system logic operates must be used to assure the relay logic is performing as intended. Contact races and operating and resetting of measuring units may be common problems. Therefore, the complete relay scheme needs to be tested as a whole to insure proper operation and proper nonoperation under simulated power system conditions.

• Realistic relay operating time tests The operating time of many line relay systems depends upon the system impedance ratio (SIR). With dynamicstate testing, different SIRs can be modeled to determine the range of relay operating times. The traditional test method never considers the affect of SIR on relay performance.

• Evaluation of future relay operations The testing provides significant advantages of obtaining more reliable test results which confirm the configuration, settings and correct operation of the protection scheme while significantly reducing test time. Since the test results describe how the relay scheme operates under power system conditions, the test data becomes a useful relay performance database. When the relay system is in service and operates for a power system event, its performance can be compared to the relay performance database to determine if the relay scheme has operated correctly. Many companies have experienced that after a questionable operation has occurred and a request for investigation was made, no findings could be gained from the steady-state test method in most cases since only the set points of individual components were checked. To meaningfully investigate a questionable operation, the actual power system conditions at the time of the incident need to be simulated to be able to observe the reaction of the system as a whole.

Dynamic-State Test Procedure 1. Create a dynamic-state test plan. The test plan for dynamic-state testing depends on the type of protection to be tested and how it is configured. The intent of the test plan is to test the relay scheme’s operation under simulated dynamic-state conditions. 2. Calculate values for simulated fault conditions. A power system model of a two-machine equivalent system can be used to aid in the calculation of voltage and current values for line relay testing. For the application to be tested, line and source values are entered. Faults are simulated on the model with varied fault locations, resistances, and load flows according to the tests defined in the test plan. Each case is a test that will characterize the scheme operation for reach and direction (faults behind and in front) and for the various zones and combinations of zones. For reach tests, the fault locations are defined according to the accuracy of the unit being tested. For a zone one relay with plus-or-minus five percent accuracy, an operation test would be defined at 95 percent of setting (op case). A test for no operation would be defined at 106 percent of setting (non-op case). These two cases confirm the accuracy of the zone one relay. The reach tests are conducted for phase and ground distance relays. For phase distance tests, use a phase-to-phase fault type; for ground distance tests, use a phase-to-ground fault type. 3. Make dynamic-state test cases. Each test case requires three-phase voltage and current values. For reach and direction tests for line relay schemes, three states are usually defined for each test case. The prefault state provides balanced three-phase voltages to the relay long enough to stabilize the relay before a fault is simulated. The prefault time assures that the relay will have the correct memory circuit response. The test time for the fault state must be long enough to operate the tested zone of protection but short enough not to operate the next overreach zone of protection. In this way, the faulted zone can be tested without disabling the adjacent overreaching zones. The postfault time is required to reapply restraint voltage after the test to prevent any spurious operations.

Protective Relaying Handbook — Volume 1 4. Playback with power system simulators. Depending on how many relay functions are configured there may be a number of cases to run. However, it only takes seconds to run a dynamic-state test so that 150 tests will take approximately five minutes. To run a steady-state test with this amount of detail will take significantly longer because of all the communications that are required with the relay to reconfigure its settings. Also, dynamicstate testing gives true relay operating performance for each power system event tested.

Conclusion Dynamic relay testing has allowed users to significantly decrease the amount of time needed for testing while increasing the quality of the test and the documentation of results. Dynamic relay testing has also provided the user with the capability of developing an understanding of the power system and the protection scheme’s function within that power system. Utilities have used dynamic relay testing to find problems that were unexplained with previous test methods. Incident reports can now be meaningfully investigated. A.T. Giuliante is President and Founder of ATG Exodus. Prior to forming his Company in 1995 Tony was Executive Vice President of GEC ALSTHOM T&D Inc.-Protection and Control Division, which he started in 1983. From 1967 to 1983, he was employed by General Electric and ASEA. In 1994, Tony was elected a Fellow of IEEE for “contributions to protective relaying education and their analysis in operational environments.” He has authored over 35 technical papers and is a frequent lecturer on all aspects of protective relaying. Tony is a past Chairman of the IEEE Power System Relaying Committee 19931994 and has degrees of BSEE and MSEE from Drexel University 1967 and 1969.

3

4

Protective Relaying Handbook — Volume 1

Introduction to Dynamic Testing NETA World, Winter1999-2000 Issue by D. L. Tierney Doble Engineering

Steady state testing is used to verify the settings for a relay element. The test quantities are applied to the relay and held steady for a predetermined time equal to or greater than the operating time of the relay. If the relay does not respond, the test quantities are raised or lowered by a small increment less than the resolution of the relay. The test quantity is then reapplied to the relay for the same predetermined time. This procedure is repeated until the relay operates.

Figure 2 — Dynamic State Waveforms

Figure 1 — Steady State Waveforms

The dynamic state test, on the other hand, is used to determine the relay’s response to power system conditions. All applied test quantities are simultaneously switched between states. Each state represents a different steady-state power system condition. One state may represent prefault conditions, while the next represents the fault followed by the postfault condition. More states may be added to represent evolving faults or reclosing. However, the dynamic state test does not include the high frequency and dc components found in many faults. To simulate a more realistic power system disturbance requires a transient simulation test. This test contains the

nonsteady state frequency components, magnitude, phase relationships, and duration the relay will see, unlike the dynamic test which uses stepped sine wave states to simulate the different power system conditions. The transient simulation test uses continuous waveform for each test quantity. The waveform itself contains the prefault and fault power system conditions. These waveforms can come from actual disturbances as recorded by digital fault recorders or the relays themselves. Another source of transient waveforms can be software programs such as Electro-Magnetic Transient Program (EMTP) or MathCAD. In the real world, relays respond to changing or transient conditions. These dynamic conditions are not simulated using stepped sine wave testing. Dynamic state testing and transient simulation testing are effective test methods. Because transient testing requires more complex data sets, dynamic tests are far easier to prepare and produce better results than steady-state testing. This article deals with the dynamic state testing of protection systems.

5

Protective Relaying Handbook — Volume 1

• The relay event log on microprocessor relays or digital fault recorders (DFR) can tell when to apply signals and when to remove them. The DFR can also record the magnitudes of the fault.

Figure 3 — Transient Waveforms

Why use dynamic state testing? Dynamic state testing can be used in every stage of relay testing: • During evaluation testing, dynamic testing can be used to simulate current reversal or power swings to compare the performance of different relays.

• During acceptance testing, dynamic testing can be used to test internal relay elements such as blocking for loss of potential with high loads, or testing the reset time of the keying output when a fault changes from a forward zone 2 to a reverse zone 3. • During commissioning testing, dynamic testing can be used to test the relay in the protection system. For example, fault can be applied to two relays at the same time to test a back blocking scheme or breaker failure scheme. • During troubleshooting, dynamic testing can be used to simulate faults for which relays did not operate as expected. • During routine testing, dynamic tests can be used for rapid go/no-go testing of protection systems.

What equipment will you need to start dynamic state testing? To start dynamic state testing you are going to need two basic pieces of equipment, dynamic state simulation software and high power active sources.

Sources of data for dynamic state testing: Data used in dynamic state testing can come from a number of different sources: • Phasor fault calculations.

• Two terminal line fault simulation software such as GE Fault.

• Multi-bus fault simulation software such as Aspen or CAPE.

• “What If ” simulation. In the absence of DFRs or microprocessor relay event logs, the “What If ” simulation is used. Dynamic state simulations are written to test what if the ground fault current was 200 amperes higher then the fault simulation software said it was. What if the relay did not receive breaker fail initiate until the fault evolved from a single line-to-ground to a double line-to-ground fault?

• When constructing acceptance tests the relay instruction manual may contain part or all of the data needed to construct the dynamic state acceptance test.

Getting started Having the right software and equipment to run a dynamic state test is only the start. How many sources are needed? Are there enough test instruments to run the tests? How can the protection system be tested in parts with a limited number of sources? Can the entire scheme be tested at once? How many states are needed to test a function of the scheme? Which test leads are required and where do they get connected? The following can help answer some of these questions and more: • One-line diagram

• Identify the quantity and type of relays and the number of sources.

• From the relay type information, determine the required current, voltage, and control power of sources. Relays located in different CT and PT strings will require more sources. • Also from the relay type information, determine source burden. If the burden exceeds the capacity of a single source, consider breaking current strings and using additional slaved sources.

• Identify the type and number of dynamic state tests. For example, add two states for each reclose cycle and select appropriate state durations.

• Identify relay settings and then calculate appropriate test quantities to determine expected time delays.

• Three-line diagram

• Determine isolation points for sources to avoid feeding active relaying. • Avoid backfeed potential transformers.

• Determine injection points for current and potential sources.

• Relaying schematic

• Writing the test plans and expected results: • What equipment is expected to operate?

6

Protective Relaying Handbook — Volume 1 • What equipment is not expected to operate? • What targets should the test produce?

• How many cycles should a given state be?

• What points should be isolated to avoid tripping inservice equipment?

• Identification of the type and number of dynamic state tests based on number of trip paths shown in the schematic.

• Identification of the type and number of logic outputs.

• Are the test instruments required to supply signaling normally supplied by other equipment or devices that can not be operated during the dynamic state test?

• Connection Diagrams

• Connection points for current and potentials • Connection points for logic outputs

• In some cases, the physical location of devices

Choosing the correct number of sources In any dynamic state test each state should contain the correct number of quantities for the protection system being tested. Relay burdens and test instrument power ratings must be taken into account when choosing the number of sources required to run the test. Electromechanical relays require more energy to operate than most solid-state and microprocessor relays. In any case, energy requirements climb with each relay added to a test. In some cases, with a high-impedance ground, the neutral must be broken and ground relays must be operated with different sources. The following is a typical list for voltage and current sources when testing electromechanical impedance relays with an electromechanical directional ground overcurrent relay: • A-phase relaying voltage • B-phase relaying voltage

• C-phase relaying voltage

• A-phase relaying current

voltage polarization. In many microprocessor relays the prefault period allows the relay to build voltage memory. In some cases, the prefault is set to zero volts and current to test switch-on-to-fault logic. Typical time duration for prefault is about 60 cycles.

• Fault states vary in number from simulation to simulation. If you are testing single line-to-ground fault with no reclosing relay, a single fault state will do. If you are testing with an evolving fault you will need one state for each stage of the fault. The first state will have a single line-to-ground fault for one or two cycles. The next state will have a double line-to-ground for two, three, or four cycles. A three line fault follows. This data should be as close to real fault levels as practical. In some cases where the fault occurs on a line close to a strong source, the secondary current will approach levels that test equipment can not provide.

• Postfault usually occurs at the end of the dynamic state test. However, postfault states can occur between fault states also. In this case you are simulating the reclose interval. In postfault state the breaker is open so the line currents are zero. The voltages, on the other hand, are either zero or full potential depending on where the relay potential transformers (PT) are located, i.e., on the line or on the bus.

Test lead considerations When pushing high currents, the impedance of test leads becomes a factor. There are several ways to minimize the impedance of the test lead, but it can not be eliminated. • Keep the test leads as short as possible. Shorter test leads have less impedance.

• Do not use the instrument ground as the return path for grounded-wye systems

• Do not coil excess test leads. Coiling the test leads turns them into an inductor. This inductance increases the impedance of the test leads. • Twisted pairs could be used to cancel mutual inductance. This inductance would otherwise increase the impedance of the test leads.

• B-phase relaying current

• Larger gauge test leads. Using a larger gauge test lead will decrease the resistance of the test lead.

• Polarizing voltage

Test lead connection point considerations

• C-phase relaying current • Polarizing current

• 3Io or ground current

Choosing the correct number of states The data should contain voltage and current values for prefault conditions, fault conditions and postfault conditions.

• In most cases prefault is set to normal load conditions to allow the relay to stabilize. In the case of an electromechanical distance relay the prefault state applies the

Where the test leads are connected is one of the most important factors in dynamic testing.

• When connecting test potentials always make sure you will not backfeed potential transformers. Make sure the PTs are isolated by pulling fuses, opening test switches, or by whatever practices are used by your company.

• Potential and current test leads should be connected to test as much of the wiring in the scheme as possible. Relays make up only part of the protection system scheme.

Protective Relaying Handbook — Volume 1 Test switches, cutout switches, meters, transducers, digital fault recorders (DFR), relays from other schemes, and wire make up the rest of the total scheme either as part of the scheme or by sharing currents and potentials. In any case they all can affect how a protection scheme responds.

• When testing breaker-and-a-half scheme one set of current transformers (CT) will have to be disconnected before the test can be conducted. Failure to do so will give the test current multiple paths. As a result, the devices under test will not receive the correct currents.

• When testing protection schemes with a primary and backup system or two primary systems one scheme at a time, care should be taken to not disable or cause an operation of the in-service scheme. Unless each system has its own CTs, it is better, in this case, to jack the currents and potentials into the individual relays.

Logic output considerations If your test requires using the relay’s digital inputs for emulating contact closures, your test instrument will require logic outputs.

• Pay attention to the ratings of the device being driven by the logic output relays. These can be low-power signaling relays. Using them to trip or open high-power devices such as trip/close coils can and will damage the relays. • Study the scheme and connection diagrams well before connecting the logic output contacts. Following are things to avoid:

• Connecting battery positive to battery negative. • Tying different battery banks together.

• Backfeeding different devices. Make sure that only the device(s) that are intended to operate are energized.

• Is the device being driven by the logic output contacts looking for dry contacts or wet contacts? In other words, is the device supplying the voltage or is the test instrument supplying the voltage? • Is the device being driven by the logic output contacts looking for open-to-close, close-to-open, voltage-to-no voltage, or no voltage-to-voltage transitions?

What am I forgetting? • Are you connected to the correct relay? This is the number one cause of misoperations during scheme testing, misidentification of relays. Do not let this happen to you. Take the time to mark off the adjacent relays so you do not accidentally operate the wrong relay. • When testing protection systems with breaker failure schemes or other similar schemes, isolation points for these relays should be opened. These open isolation points (cutoff switches or test switches) will prevent

7 tripping of in-service equipment in the event that the breaker failure relays or similar relays operate.

• Do not wait until you start testing to find out whether your test leads are connected correctly or whether the phases are rolled in the wiring. Turn on the potentials and currents one at a time or together at different phase angles and/or magnitudes. Then trace the quantities through the PT and CT strings to verify each phase.

• To trip the breaker or not to trip the breaker, that is the question. The actual breaker should be tripped at least once to ensure that the relay contacts can handle the trip current. Reason number two to trip the breaker is to ensure that the voltage drop across the wiring during tripping is not a factor in the operation of the breaker during a fault. Another reason is to test the breaker’s “a” and “b” contacts connected to the protection scheme. However, for all other breaker trips, the use of a breaker simulator is recommended to save wear and tear on the breaker, especially high-voltage breakers. • Many breaker simulators in use today are built around a lockout relay. One problem with the lockout breaker simulator is its speed. The lockout relay operates in less then eight milliseconds. This is faster than most breakers and can give different test results when used instead of operating the breaker. Therefore, time delay circuits may be needed to slow down the tripping and the closing of lockout breaker simulator boxes.

• When connecting the breaker simulator care should be taken not to backfeed signals. • Do you want the station oscillograph or digital fault recorder to operate for each and every test? If not, you may want to temporarily disconnect the triggers to devices.

• When testing schemes with transfer trip, operation of local relays may cause the remote breaker to operate or change state. Care should be taken to isolate these signals if you do not want to operate the remote breaker. In a future issue of NETA World, a specific example with connection details, source selection, and state calculations will be presented. Dennis Tierney has been a Senior Applications Engineer for Relay Protection at Doble Engineering Company for approximately one year. Prior to this position, for eleven years, Dennis worked at the Salt River Project in Phoenix, Arizona, in relay protection, power quality and, SCADA. Before working at the Salt River Project, he worked in HVDC and Communications at the Los Angeles Department of Water and Power. Dennis graduated from Arizona State University in 1982 with a Bachelors of Science Degree in Electrical Engineering.

8

Protective Relaying Handbook — Volume 1

Through-Fault Testing — the Ultimate Test for Protection Schemes Prior to Energizing PowerTest 2000 (NETA Annual Technical Conference) Roderic L. Hageman PRIT Service, Inc.

Concept

Source

The concept of through-fault testing is not new to our industry. Take, for example, primary injection testing of low-voltage circuit breakers. Technicians routinely inject fault level current through these breakers to verify pickup and timing of the associated trip units. Most technicians have learned that single-phase injection can create problems with the ground fault elements, masking the pickup and timing of the phase functions. If there is no way to defeat the ground fault element at the trip unit, an injection in one pole and out another will cause cancellation of the ground fault current. Other through-fault tests are frequently made on substation bus ground fault schemes and bus differential schemes. The fault current for all of these tests is typically provided by a single-phase, high current test set that can deliver thousands of amperes at a very low voltage. The procedures are relatively safe due to the low voltage. Typically, one of the primary hazards is the temperature rise of the test set leads or connections. For the same reasons that the procedures described above are performed, similar tests are desirable for more sophisticated protective relay and metering schemes. In these schemes, phase angles are as important as current magnitude for the correct operation of the scheme. A prime example is that of a transformer differential scheme. The primary current and the secondary current will differ, not only due to the ratio of the protected transformer, but also due to any other phase angle shifts caused by delta-wye configurations. Electromechanical relays typically require that the current transformer connections correct for the delta-wye shifts. Modern microprocessor-based relays can be programmed to account for the shifts internal in the relay. However these corrections are made, it is desirable to perform an overall system test to confirm that the design and installation provide protection without nuisance tripping.

In some cases utilities will actually stage faults on the power system. This amounts to deliberately short circuiting a transmission line or distribution feeder and energizing it at normal voltage. Obviously, this could be damaging to the system, and if the relaying systems do not work correctly, severe damage to the power system can occur. With the prominence of modern computer-operated dynamic test sets, GPS synchronizing, and end-to-end testing, the need for this type of staged fault testing is decreasing dramatically. Although I have not seen it, I have heard of using system generators to provide the desired level of fault current. Because the power system impedance is primarily reactive, fault currents require very little real power. If the generator’s excitation system can be adjusted to produce a relatively low voltage compared to the normal system voltage, fault current can be controlled and kept to a reasonable magnitude. A relatively easy way to provide the fault current and yet control its magnitude is to use a low voltage source and the impedance of a transformer to limit the fault current. This transformer might be, for example, the actual transformer in the part of the distribution system that is being tested. If this is not convenient, a transformer of the appropriate ratio, impedance, and kVA size might be available from a rental agency.

Metering In setting up the through-fault test procedure, it is necessary to take into consideration the available metering. Older electromechanical phase-angle meters might require 0.5 ampere or more to reliably determine phase angle. Modern power meters typically have a sensitivity as low as 50 milliamperes.

9

Protective Relaying Handbook — Volume 1 It is convenient to use a polyphase power meter rather than the “spaghetti jungle” associated with all of the individual meters necessary to monitor the full system. Additionally, modern power meters have functions such as event memory, printing facility, and even on-screen phasor diagrams. In the planning, do not forget to determine where and how to obtain the signal sources that are to be measured. In some cases, relay test switches or test plugs can provide the secondary variables. In other cases, microprocessor-based relays can actually provide the desired information either directly on the relay display or via a computer.

Example The example comes from a 600 MW peaking station. A partial one-line is shown in Figure 1. The test was performed initially to assess problems with the 345 kV line differential scheme. However in the process, several problems with the transformer differential schemes were uncovered. There are several considerations when making the calculations: A. Current magnitude at all system voltages must be high enough to provide adequate current transformer secondary values to reliably register on available metering. B. Current magnitude must be at a level that does not overload system components. C. A source of sufficient kVA capacity and correct voltage level must be available. The two most common three-phase low voltage systems in the United States are 208Y/120V and 480Y/277. This example was calculated knowing that a 1000 kVA, 480Y/277 source was available on site for construction power.

Calculation of Fault Current Calculations are made in per unit and usually most conveniently on the transformer base of the transformer used as the fault limiting impedance.

kVABase *1000 Z Base = ——————— 3 * Base 18*1000 = ————— = 2.818 1.732*3,688 Those with experience in per unit calculations will recognize that ZBase is more easily calculated as: (kVBase )2 Z Base = —————— MVA Base 182 = ——— = 2.817 115 Calculating the actual ohmic impedance of the transformer is as follows: Zactual = ZBase * Zpu From the transformer nameplate we find the %Z and %Z 10.5% Zpm = —— = ——— = 0.105 100 100 ZActual = 2.817 * 0.105 = 0.2958 If we connect a 480Y/277V source to the 18 kV winding and short circuit the 345 kV system, the following current will flow on the 18 kV system: V L-N = ———— Fault Actual

277 V = ———— = 936 A 0.2958

MVA Base: Base Rating of GSU Transformer = 115 MVA = 345 kV VBase: GSU Primary GSU Secondary = 18 kV MVA Base *1000 = ——————— Base 3 *kV Base at 18 kV:

Base

115*1000 = ————— = 3,688 1.732*18

on the 345 kV system current will be: Fault

18 kV = 936* ———— = 48.8 A kV

Now let us check on some of the considerations we listed earlier. First, does our available source have sufficient capacity?

10

Protective Relaying Handbook — Volume 1

Source kVA = 3*

Fault

*kV L-L

= 1.732 * 936 * 0.48 = 778 kVA So our 1000 kVA source was large enough, but it was prudent to remove existing loads; therefore, the tests were planned for the lunch hour. The electrical contractor installed a temporary run of two 500 kcmil cables per phase with necessary barricades and warning tape. Before the contractor installed the cable, calculations were made to insure adequate current would be available in the CT secondaries. The first problem was uncovered here. The CT ratios on the 345 kV system were not the same at each end of the protected line and could not be made equal by tap selection. The coordination engineer was notified and new settings were developed to accommodate the problem.

Calculation of CT Secondary Current At the peaking station the 345 kV CTs were 1200/5 ratio: Relay

48.8 A Fault = ——— = ——— = 0.2033 A CTR 240

At the transmission substation the 345 kV CTs were 2000/5 ratio: Relay

48.8 A = ——— = 0.122 A 400

Both currents were larger than the minimum 50 mA required for reliable phase angle measurement by the meters we were using. The CTs on the 18 kV side at the peaking station were 8000/5 ratio and provided more than adequate current for monitoring the 18 kV winding currents in the transformer differential relays: Relay

936 A = ——— = 0.585 A 1600

After checking that the through-fault current magnitude was less than any of the components, we were finally ready to proceed with the test. The source was turned on, and the transformer differential lockout relay immediately tripped the MOD. Fortunately, the transformer differential relay had event recording, and it was soon apparent that there were significant problems with the 87T circuits. Since the utility engineers were waiting, we elected to disable the 87T and continue the tests on the 87L system. Those tests

went very well. The actual current was almost exactly what was calculated, and phase angle measurements confirmed that the input currents to the line differential relays were as indicated on the drawings. Since the 87L currents appeared to be correct relative to the drawings, we joined with the utility engineers to review the entire scheme. It was determined that the problem was with the design and not with the components. A convenient place to reverse the polarity on one set of relays was located, and that system was finally functional. Once the main objective of the through-fault testing was accomplished, the unexpected transformer differential relay trip became the focus. A number of problems were found with this system. First, the design engineer had reversed the primary and secondary inputs causing an extreme ratio mismatch. Further analysis of the event indicated one of the three CTs on the primary winding was reversed in polarity. This, despite the fact that the CTs had been tested for ratio and polarity, and the secondary circuits had been injected back to the relay. Although this example is somewhat extreme in terms of the number of problems found, typically, through-fault testing will find a problem or problems in the protection circuits. Roderic Hageman is President of PRIT Service, Inc. His firm has provided consulting and testing for electric power distribution systems for more than 25 years. He received his B.S. in Electrical Engineering from Iowa State University and is a registered professional engineer. Mr. Hageman has served two terms as President of the InterNational Electrical Testing Association (NETA) and nine years as a member of NETA’s Board of Directors. He has three times been named NETA’s Man of the Year and continues to be very active in NETA

Our Goal: Partner with clients to enhance system safety and reliability by providing leading edge independent electrical testing and engineering services. CE Power provides: Protective Relay Testing and Calibration Protective Relay Upgrade Services Acceptance Testing Commissioning Services Equipment Repair, Retrofit and Upgrade Preventive Maintenance Power Monitoring Arc Flash Hazard Analysis Engineering Studies/Power System Evaluation CE Power specializes in: 480V—765kV Plant Substation Alternative Energy

CE Power Solutions of Ohio 4500 West Mitchell Avenue Cincinnati, OH 45232 800.434.0415 513.563.6150 phone 513.563.6120 fax [email protected]

24/7 Emergency Service Available Nationwide 800-434-0415 www.cepowersol.com

CE Power Solutions of Wisconsin 3255 West Highview Drive Appleton, WI 51914 800.434.0415 920.968.0281 phone 920.968.0282 fax [email protected]

11

Protective Relaying Handbook — Volume 1

Automated Test Point Calculations for Electronic Relay Testing and Coordination NETA World, Summer 2000 Issue Lonnie C. Lindell and Steven R. Potter SKM Systems Analysis, Inc.

Modern software can be used to automate relay setting selection, documentation and test point specification. Whereas electro-mechanical relays are built to have a specific time-current characteristic, microprocessor-based relays are available with programmable selections of time-current curve shapes and a wide range of possible settings. To automate the generation of time-current curves necessary for relay coordination and testing, most microprocessor-based relays provide equations that can be used to generate the curves. These equations can be used in simple spreadsheet programs to generate time-current curves and to calculate test points with very little effort. The equations can also be used in more sophisticated programs for relay coordination and test point specification. In its simplest form, a spreadsheet can automate calculation of test points. Spreadsheets can also be used to generate complete setting sheets to document a more extensive series of tests. It is important to note that a separate spreadsheet may be required for each type of relay since the equations, equation constants and setting ranges may vary between different relays. Often an existing spreadsheet will require only minor changes to be tailored for a new relay. Using a spreadsheet to generate the test points directly from the relay equation is substantially more efficient than reading points from the relay curves. The spreadsheet is also more consistent and more reliable than reading from the curves. A simple spreadsheet example is shown in Figure 1. In this sample spreadsheet, entering a time dial value automatically displays the calculated test points based on the equation shown. New current multiples can also be selected by simply changing the cells with M=2, M=3 and M=5 for 2, 3 and 5x current multiples.

Test Point Calculation for Relay ABC Equation:

T = AD/(MN-C) + BD + K

Constants:

A B N K C

7.7624 0.02758 2.0938 0.028 1

Time Dial

D

1

Test Point @

M 2

X

=

Time 2.430

Test Point @

3

X

=

0.920

Test Point @

5

X

=

0.332

Figure 1 — Sample spreadsheet for calculating test points

Spreadsheets combined with scientific plotting programs can be used to plot the relay time-current characteristics by entering the relay equations. Mathematics and plotting software combinations such as MathCAD™ can also use the relay equations to display the relay time-current characteristics. While these methods can plot a single curve, they stop short of providing complete relay coordination and system protection functions.

12

Protective Relaying Handbook — Volume 1

System protection software that can incorporate published relay equations to generate time-current coordination curves and specify relay test points is widely available. These programs display damage and performance curves for power system components such as motors, generators, transformers and cables as well as time-current response curves for relays, fuses, circuit breakers and other protective devices. Figure 2 displays a sample coordination drawing that includes low-voltage motor protection with a motor circuit protector and thermal-magnetic breaker; feeder protection with a fuse; and transformer protection with a relay and medium-voltage breaker. Figure 3 displays a sample coordination drawing that includes medium-voltage motor protection, feeder protection, and transformer protection with a combination of relays. The relays include both electromechanical and electronic equation-based relays. Many of the system protection and coordination programs can also generate relay test points. A sample report that includes relay settings and test points is displayed in Figure 3. The report was automatically generated by the system protection and coordination program used to produce the coordination drawing shown in Figure 3. Combining the coordination, reporting, and test point generation in a single application saves time and minimizes errors. The important capabilities of system protection and relay test point specification software include: ●

●

●

● ●

Figure 2 — Sample protective coordination drawing

Representation of device curves by equations and/or tabular data Coordination of devices at different voltage levels on the same drawing Tabular reports that include device settings and test points High quality output of protective coordination drawings Large user-expandable library of protective device characteristics.

Using software to automate relay setting selection, documentation, and test point specification offers several benefits to design and test engineers and technicians: ● ● ●

● ●

●

It improves power system safety.

It improves power system reliability.

The use of standard equations and models reduces the chance for human error.

It promotes consistent design and testing practices.

It provides a consistent presentation format which enhances understanding between multiple engineers and technicians.

The use of standard software improves efficiency and saves time and money. Figure 3 — Sample protective coordination drawing

13

Protective Relaying Handbook — Volume 1 Device Name: Relay A Description: MULTILIN-SR745 Xfmr Relay-5A CT Sec AIC Rating: N/A Current Rating: 300A / 5A Setting: 1) OC Pickup 1.0 2) Mod Inverse 5.0 1.0 3) Inst OC Pickup 7.0

Bus Name: Bus Voltage: Fault Duty: Curve Multiplier: Test Points: @2.0X, 3.787s @4.0X, 1.909s

Line 1, 69kV 69000.0V 200000.0A 1.00000

Device Name: Relay B Description: WESTINGHOUSE-CO-7-50/51 13800.0V AIC Rating: N/A Current Rating: 1500A / 5A Setting: 1) Tap 5.0 2) Time Dials 3.0

Bus Name:

Bus 1, 13.8kV Bus Voltage:

Fault Duty: Curve Multiplier: Test Points: @2.0X, 2.150s @5.0X, 0.970s

12000.0A 1.00000

Bus Name:

Bus 1, 13.8kV Bus Voltage:

Fault Duty: Curve Multiplier: Test Points: @2.0X, 9.200s @5.0X, 1.250s

200000.0A 1.00000

Device Name: Relay C Description: WESTINGHOUSE-CO-11-50/51 13800.0V AIC Rating: N/A Current Rating: 600A / 5A Setting: 1) Tap 5.0 2) Time Dials 5.0 3) INST (High) 35.0

Figure 4 — Sample setting table including automatic relay test point specification

With these substantial benefits and a relatively small investment in time and resources needed to implement a software solution, there is no reason to use traditional time-current curves for selecting relay test points for equation-based electronic relays. From simple spreadsheets to sophisticated protective coordination software, using published relay equation data will substantially automate system protection and relay test point specification. Lonnie C. Lindell is General Manager of SKM Systems Analysis, Inc., an electrical engineering company specializing in power system analysis software development. He received a BS from the Iowa State University School of Engineering and an MBA from the University of Phoenix. He has over 15 years’ experience in the application of engineering computer software, is active in education and engineering presentations, and is a member of the IEEE. Steven R. Potter is a senior support engineer for SKM Systems Analysis, Inc. where he specializes in protective coordination and protection equipment computer modeling. He received his BSEE from San Diego State University. He has over eight years’ experience in the application of engineering computer software, is active in engineering education, and is a member of the IEEE.

14

Protective Relaying Handbook — Volume 1

Test & Maintenance Tips for Protective Relays NETA World, Winter 2000-2001 Issue by Scott Cooper Beckwith Electric

Beckwith Electric protective relays incorporate several self-checking routines that continuously monitor critical functions. When an internal fault is detected the relay safely removes itself from ser vice and closes the diagnostic contact. These self-test functions, however, can not determine the integrity of a status input or trip circuit nor detect small problems in CT or VT circuits. To verify the integrity of these circuits, we recommend routinely checking the relay’s metering during normal operation and performing the diagnostic test procedure during outages. The output trip circuits can be verified by exercising the output relays and checking the external trip circuits for correct operation. This combination of internal self-diagnostics, input verification, and output testing assures that the relay is ready to protect the system. This maintenance should be performed according to each company’s schedule. To prevent a layer of

Figure 1 — Screen from IPSplot® Oscillograph Analysis Software showing a differential trip. The vertical variegated line in center indicates the breaker tripping and subsequent Beckwith relay operation. The suspected cause is a wiring problem in their CT circuit.

insulating silver oxide from fouling the case contacts, we recommend periodically reseating M-0420 and M-0430 relays in the drawout case. One of the most useful and often overlooked diagnostic features of our relays is the oscillographic recorder. With the recorder, up to 170 cycles (96 cycles in the M-0420 and M-0430 relays) of prefault input waveforms can be recorded automatically. The recorder may be triggered manually or by the operation of any output or input combination chosen by the user. Once triggered, this waveform data can be easily transferred from the relay using the IPScom® Communications Software. The waveform may then be analyzed using the available IPSplot® Oscillograph Analysis Software. The resultant data can be a valuable tool in determining the root cause of a relay operation. If periodic functional testing is desired, consider that a single-phase or even a three-phase test set can not duplicate system conditions for a relay which has seven current inputs and four voltage inputs. Consequently, the technician has to disable or alter the setpoints of other functions to prevent interference with the function under test. This could result in the relay being placed back in service with a critical function accidentally disabled. To minimize this possibility, use the IPScom software shipped with the relay to save the relay’s data file before testing. Then write the same file back to the relay af ter testing. This practice can dramatically reduce the possibility of setting errors while also providing a convenient record of “as found” settings. Successful functional testing of these relays involves a few steps. First, study the functional description from the relay instruction book, carefully noting any special features. Second, connect the relay exactly as it will be connected to the system. Third, isolate the function under test with the IPScom software’s configuration screen. Fourth, apply the nominal quantities and check the metering using the IPScom software’s secondary metering screen. Finally, ap-

Protective Relaying Handbook — Volume 1 ply the test quantities and check your results. If the results are not satisfactory, check the secondary metering screen again with the fault quantities applied. If incorrect, check connections and inputs; if correct, check the function logic description and testing instructions. By performing this routine maintenance as required, you are helping to ensure the integrity and reliability of the protective relay. Scott Cooper, Field Service Engineer, joined Beckwith Electric Co. in 1997. His responsibilities include training, commissioning, and troubleshooting protective relays for customers. He is also instrumental in testing new relay products and custom-engineered systems. Scott was previously an electronics technician at Beckwith testing protective relays and conducting failure analysis and individual component evaluations. He is a member of IEEE.

15

16

Protective Relaying Handbook — Volume 1

Using Iop Characteristics to Troubleshoot Transformer Differential Relay Misoperation PowerTest 2001 (NETA Annual Technical Conference) Michael Thompson and James R. Closson Basler Electric

Abstract – When a transformer differential relay operates with no obvious transformer fault, system operators have a serious decision to make. Is there a transformer fault, or did the relay operate incorrectly? Testing the transformer requires significant time, with the associated direct and indirect costs to do so. On the other hand, reenergizing a faulted transformer can lead to catastrophic equipment failure. This scenario of a questionable transformer operate occurs more often than we would like to think, particularly during the equipment commissioning process. Several conditions can cause differential relay false tripping. These conditions can cause false trips from external faults, or simply increased transformer loading. Some indication is needed that the relay is not operating as desired before an incorrect operate happens. A potential problem can be identified by monitoring the operating condition of the differential relay. Indications provided by this monitoring can serve as a warning if the settings or connections are not correct. This paper will explore the issues contributing to transformer differential false trips, and suggest methods to alleviate this issue.

Reviewing Differential Relaying Principles When assessing relay system operation, a basic understanding of differential relay operation is necessary. A summary of the concepts follows:

Figure 1 — General Differential Principle

Differential relaying offers the highest selectivity and, therefore, the highest speed and most secure type of system protection. In theory, a differential relay compares the currents into and out of the protected zone. If the sum of the currents is not zero, the relay will operate. This is shown in the phasor diagram, Figure 2. The sum of the currents is identified as the operate (Iop) or unbalance current. The relay does not acknowledge conditions external to the protected zone. Accordingly, coordination delay times are not necessary, and sensitivity can be optimized.

Figure 2 — Phasors of Ideal Non-Fault Condition

17

Protective Relaying Handbook — Volume 1 Differential relaying relies on the quality of the incoming currents from current transformer secondaries. Therefore, CT performance is of particular concern in this application. Although the relay must be desensitized to ensure security for all non-fault conditions, it must remain highly sensitive to faults within the zone of protection. To accomplish this, a fixed minimum pickup setting is commonly used, as well as percentage restraint. Percentage restraint increases the amount of unbalance, or operate, current needed to actuate the relay based on the current flowing through the protected equipment. The restraint setting, or slope, defines the relationship between restraint and operate currents (See Figure 3). Relays vary in the way they define the restraint value in the calculation of Iop/Irestraint percentage ratio. Two common methods are to take the average of the two currents (current entering the zone and current exiting the zone) or to take the maximum of the two currents to use in the percentage ratio.

Figure 3 — Percent Restraint Characteristic

Transformer Differential Specifics Transformer differential relaying does have some complications, which can be the source of errors in connections and set-up. As noted, differential relaying is based on virtually balanced current into and out of the protected zone. However, a transformer is not a balanced current device. The currents into and out of a transformer will differ by the inverse of the transformer’s voltage ratio. Thus, the associated currents need to be adjusted to represent a balance during non-fault conditions. To a great extent, this adjustment can be accomplished with the selection of the system current transformers. The final balancing is accomplished in the relay’s TAP settings. The TAP settings scale the input currents, effectively defining per unit values. The success of this balancing is measured by the mismatch, which is the percentage difference between the ratio of the currents seen by the relay and ratio of the relay taps.

Figure 4 — Transformer Differential Relaying

There are also conditions on the power system that create unbalance currents in a transformer but do not represent transformer faults. When system voltage is applied to a transformer at a time when normal steady-state flux should be at a different value from that existing in the transformer, a current transient occurs, known as magnetizing inrush current. The differential relay must detect energization inrush current and inhibit operation. Otherwise, the relay must be temporarily taken out of service to permit placing the transformer in service. In most instances this is not an option. The harmonics in faults are generally small. In contrast, the second harmonic is a major component of the inrush current. Thus, the second harmonic provides an effective means to distinguish between faults and inrush. Almost every transformer differential relay available inhibits operation based on the 2nd harmonic content of the energization current. A parallel ‘high set’ operate level is included to ensure that larger faults will still be detected during energization. The high set, unrestrained element is also provided to ensure operation for a heavy internal fault such as a high side bushing flashover. This high grade fault may result in CT saturation, which can generate significant harmonics that may restrain the sensitive harmonic restrained element. This is shown in Figure 5. External faults can also cause unbalanced currents in a power transformer, depending on the transformer’s connections. A Wye connected transformer winding can act as a power system ground source, providing ground current to external faults. This unbalanced current must be blocked from the differential circuit to ensure relay security. This blocking is usually achieved by a Delta connection in the associated relay input transformer circuit, which traps the zero sequence (ground) current component. This delta connection can be achieved either with the current transformers, or, if an option, within the transformer differential relay itself.

18

Protective Relaying Handbook — Volume 1

Figure 5 — Simplified Block Diagram

An important issue with transformer differential relaying is the phase shifts inherent in most transformer connections. A delta connection in a power transformer affects a 30° phase shift in the associated currents. Since the differential relay compares the currents on an instantaneous basis, this phase shift will create an unbalance, which must be compensated. This compensation is usually achieved with a corresponding delta connection in the CT secondary circuits and must be coordinated with any zero sequence blocking connections required.

Many transformers are connected with delta windings on the high side and wye windings on the low side. This provides isolation between the power system voltages and a ground source for detecting faults on the low voltage side. The three-line drawing, Figure 6, shows a delta/wye transformer with the associated phase shifts. In this example, the phase shift is accomplished by connecting the CT’s on the wye side in a delta configuration. The required phase shift compensation can also be accomplished within the differential relay. This is desirable for several reasons. Probably the most important of these is that it allows the CT’s to be connected in wye, making them easier to connect and verify during installation.

19

Protective Relaying Handbook — Volume 1

Figure 6 — Phase Shifts in Transformers

The presence of a Load Tap Changer (LTC) in transformers will also affect differential relay operation. Usually, these taps provide the possibility of modifying the voltage ratio 10% for voltage or Var control. This ratio variance, in turn, varies the current ratios. This variation is usually within the security margin provided by the relay’s restraint characteristic. For a given LTC position, the ratio of operate current to restraint current will remain constant, as shown in Figure 7.

Figure 7 — Operate Characteristics with Proper Configuration (10% Mismatch)

Connection Concerns Almost all nuisance trips associated with transformer differential relay applications can be attributed to incorrect relay settings or CT connections or mismatch. During a through-fault condition, the differential operating current due to mismatch can approach the current rating of the transformer. These typical mistakes will be discussed, along with their effects on relay performance. For each case discussed, the TAP settings are presumed to be set to the transformer’s full load current. This defines the 1 per unit value to be equal to full load. This is the easiest setting to calculate, and simplifies analysis. The minimum pickup of the transformer differential relay is taken as 0.35 times TAP for this discussion, or when Iop = 35% of transformer full load, given the defined setting. A restraint slope of 40% of maximum restraint current is assumed. The % of Maximum characteristic is preferred because it uses information from the best performing CT to restrain the relay. A relay using % of Average restraint current would provide different results but the concepts are the same. In modern numerical differential relays, the restraint characteristic may be user-selectable.

20 Single Restraint Input If one set of current transformers is not connected to the differential relay or the current transformers are shorted out, the differential relay acts as an overcurrent relay. Given this scenario, I op = I restraint.

Protective Relaying Handbook — Volume 1 Under this condition, increased loading will cause the relay to operate. This operation will occur when Iop exceeds 35% of transformer full load (based on the setting presumptions). This will be when the load (restraint) current reaches 17.5% of full load (or 17.5% of TAP setting). This condition is plotted on the characteristic graph in Figure 11.

Figure 8 — Transformer Differential Phasors with Missing Input Current

When the single input current exceeds the minimum pick-up the relay will operate. So for this scenario, the transformer will trip at 35% of full load under this condition.

Figure 11 — Operate Characteristic with Reversed Input Current

There are two problems that can occur with phase shift compensation. The engineer performing the work can forget to apply compensation or compensation can be incorrectly applied. When a transformer includes a phase shift, a corresponding adjustment must be made in the relay scheme. This is generally accomplished by connecting the relay input currents in delta, and can be done either at the CT inputs or within the relay’s circuitry. The proper correction is shown in phasor diagram in Figure 12.

Figure 9 — Operate Characteristic with Missing Input Current

Current Transformer Lead Reversal Reversing a current transformer lead, or group of leads, is the simplest mistake made when wiring a new panel or upgrading a protection system. Since the differential relay compares the transformer currents, CT polarity is extremely important. When a CT lead is reversed, the resulting unbalance current is double the normalized load current. That is Iop = 2 * I load. Assuming balanced currents (proper TAP settings), Iop = 2 * I restraint. This is shown in the phasor diagram, Figure 10.

Figure 12 — Transformer Differential Phasors with Proper Phase Shift Adjustment

If phase shift compensation is not performed when the application requires it, there will be a resulting Iop in the relay. As load increases, the relay will begin to see an unbalance. The differential relay will interpret this unbalance as a fault and operate. Phasor analysis, Figure 13, shows that an uncompensated 30° phase shift will cause an unbalance current that is approximately half the normalized load current. That is Iop = 0.5 * I load.

Figure 10 — Transformer Differential Phasors with Reversed Input Current Figure 13 — Phasor Diagram with Missing Phase Shift

21

Protective Relaying Handbook — Volume 1 If this condition exists, the relay will operate with increases in load, unless the restraint slope setting is larger than 50%. The relay will operate when Iop exceeds 35% of transformer full load (based on the previous setting presumptions). This will occur when the load (restraint) current reaches 68% of full load (or 68% of TAP setting). Figure 14 shows this situation.

Figure 16 — Phasor Diagram with Wrong Phase Shift

The relay will operate when the load (restraint) current reaches 35% of full load (or 35% of TAP setting) as shown in Figure 17. This is a similar level of load to the scenario where one side of the differential zone is completely missing as shown in Figure 9. Figure 14 — Relay Operate Characteristic with Missing Phase Shift

Another error can occur by incorrectly applying a phase shift. For example, shifting the relay input on the delta side of a delta/wye transformer. While the required phase angle adjustment is achieved, the necessary zero sequence blocking is not provided. In this case, the differential relay will operate for external ground faults on the wye side of the transformer. This condition is not detectable by taking readings under balanced loading conditions. The other incorrect shift is a phase shift in the wrong direction. As shown in Figure 15, there are two ways to apply a delta connection. Each affects a 30° phase shift, but in different directions. If the wrong connection is applied, it will result in a 60° difference rather than proper phase compensation. Again, this will cause a non-fault, or false, Iop, and the relay will operate with increasing load. Phasor analysis, Figure 16, shows that a 60° difference in the relay currents will cause an unbalance current equal to the normalized load current. That is I op = 1 * I load.

Figure 17 — Operate Characteristic with Wrong Phase Shift

Transposed Tap Settings

Incorrect TAP settings can occur when the TAP settings for the relay are transposed. That is, the high side TAP setting is applied to the low side input, and vice versa. The resulting relay performance will depend on how closely matched the current signals into the relay are. If the currents into the relay are very close, the TAP settings will also be similar, and relay security may not be affected. However, if the inputs are substantially different, the resulting unbalance will likely cause the Figure 15 Two Delta Applications relay to operate and cause a nuisance trip.

22

Protective Relaying Handbook — Volume 1

For example, presume a condition where the currents to the relay are 3.8 amps on the high side and 4.2 amps on the low side. The proper relay TAP settings would be 3.8 for the high side input and 4.2 for the low side input. If the settings are transposed, the current magnitudes will be incorrectly scaled. This results in a mismatch of 22%, as shown below. Mismatch = (current ratio) - (TAP ratio) smaller of above

Figure 18 — Characteristic with Bad Tap Settings

with proper settings:

Mismatch = with transposed settings: Mismatch =

(3.8/4.2) - (3.8/4.2) = 0% (3.8/4.2) (3.8/4.2) - (4.2/3.8) = 22% (3.8/4.2)

In this example, the security of the relay will depend on the setting of the restraint slope. At a slope setting of 15%, the relay will operate on increasing load, when the I restraint exceeds about 1.6 multiples of TAP or at 160 % of transformer full loading. At a slope setting of 40%, it would not operate on load. However, the security margin would be reduced by this mismatch. Figure 18 shows this example.

3 Factor Neglected In Tap Settings

Another TAP setting problem that can occur is to overlook the magnitude increase associated with a delta connection in the current circuit. This is a by-product of the phase shift adjustment, and must be taken into account. The magnitude shift is the square root of 3, or 1.73. This magnitude compensation must be included if the delta compensation is achieved with CT connections. It may or may not be required if the delta compensation is achieved internal to the relay. Care must be taken to review the operating characteristics of the relay in question when calculating tap factors. This problem is mitigated in some numerical relays that are capable of automatically calculating their own tap adjust factors. Using the previous example of 3.8 and 4.2 as the currents into the relay, assume that the 4.2 amps current requires a phase shift. The delta compensated 4.2 amps is now effectively 4.2*1.73=7.3 amps for the differential element. Thus, for the delta side of the transformer, 3.8 amps = 1PU and, for the wye side of the transformer 7.3 amps = 1PU. The proper current ratio is now (3.8/7.3) rather than (3.8/4.2). If the protection engineer overlooks this, the resulting mismatch will be: Mismatch = (3.8/7.3) - (3.8/4.2) = 73% (3.8/7.3) This will clearly cause a problem. The relay will operate at 48% of transformer full load current in this case. The effect of this setting error is shown in Figure 19.

Figure 19 — Relay Operate Characteristic with Missing √3 Factor in Taps

Checking and Troubleshooting Differential Circuits Field personnel can apply the lessons noted in this paper in order to troubleshoot CT connections and rectify problems. For example, a quick simple check of measuring the current in the operate coil of the differential relay may be sufficient to detect the gross problems described such as reversed polarity or one CT completely missing. However, many of the problems identified result in relatively small mismatches. This check also does not acknowledge the fact that the relay can adjust for magnitude mismatch by its tap settings. For example, a properly designed differential relay circuit with one tap set at 5 amps and the other set at 10 amps would result in 5 amps of operate current under full load balanced conditions. On one side of the zone 5 amps = 1PU, while on the other side of the zone 10 amps = 1 PU. In electromechanical relays, Ioperate is the sum of the currents, which would be 10 – 5 = 5 amps for this example. A better approach is to measure and record both the magnitude and angle of the restraint currents at each terminal of the relay. For example, the criteria should be: • The ratio of the magnitudes of the restraint current on each phase should be equal to the ratio of the magnitudes of the tap settings.

Protective Relaying Handbook — Volume 1 High _ Side _ Current Low _ Side _ Current

High _ Side _ Tap Low _ Side _ Tap

• The currents on each phase relay should be nearly exactly 180° out of phase.

Differential Current Monitoring as a Diagnostic Tool Modern relays with internal phase compensation do not allow the field engineer to do it the old way with phase angle and magnitude readings. It is necessary to see the values seen by the differential element after they have been manipulated inside of the relay, and this cannot be done by direct measurement. Other methods must be employed. As this paper has noted, there are many connection or setting problems that can cause incorrect operations in transformer differential relays. The task is to detect these problems before an incorrect relay operation. Differential current monitoring is a diagnostic function designed to aid in the installation and commissioning of differential relays, especially on transformer banks. This function attempts to identify and prevent false trips due to incorrect polarity, incorrect angle compensation, or mismatch. During transformer commissioning, it would be particularly useful to analyze the system installation and create a record of the settings and measured currents. The differential current monitoring function can create a differential check record like the sample shown in Figure 20. These records are also useful when comparing the present system characteristics to the characteristics at commissioning during troubleshooting to determine if something has changed. The differential check record shown in Figure 20 is an example of a differential current check record developed by a numerical differential relay. This particular example is from an actual installation. The names and dates on the record have been changed. Upon putting load on the transformer bank after installing the upgraded protection, the differential relay alarmed, triggering the diagnostic routine to generate this report, and tripped. The relay’s trip outputs were not connected at the time. The first grouping of information in the record is the date and time the record was captured and the basic relay identification. The second grouping is a record of the CT and transformer connection settings and the 87 (differential) settings that were entered by the user. The third grouping is a report of the tap and angle compensation factors that the relay is using for each of the three phase CT input circuits. It is important to note that the angle compensation cannot be entered manually. The angle compensation is calculated by the relay based on the CT and transformer connections. Additionally, the tap compensation setting may be entered manually or automatically calculated.

23 As mentioned earlier in the paper, a transformer delta winding can be configured in two ways: Delta IA-IB or Delta IA-IC. The type of delta and the normal phase sequence of the system determines whether the phase shift will be +30 degrees or –30 degrees. From the information in the report, it can be noted that the user has described the transformer winding connected to CT circuit 1 of the relay as a delta with DAB (Delta IA-IB) connections; and the transformer winding connected to CT circuit 2 of the relay is described as a wye configuration. This would be a pretty safe assumption based on the fact that an ANSI standard delta high-side/wye low-side transformer uses this configuration so that the low side lags the high side by 30 degrees when system phase sequence is ABC. The fourth grouping of information in the record attempts to identify polarity and angle compensation errors by looking at the phase angle differences of compared phases. The differential alarm is set whenever the minimum pickup or the slope ratio exceeds the differential alarm, percent of trip setting. If the differential alarm is set and neither the polarity alarm nor the angle compensation alarm is set, a mismatch error is identified indicating that the most likely cause of the alarm is incorrect tap settings. In this example, the record clearly identifies that the problem appears to be with the angle compensation. The fifth grouping of information (MEASUREMENTS) displays the measured and calculated currents at the time of the differential record trigger. The relay measures secondary current and develops the tap and phase compensated currents for use by the differential element. Primary current (MEASURED I PRI) is calculated simply as the secondary current multiplied by the CT turns ratio. Secondary current (MEASURED I SEC) is the current actually measured by the relay. Angle compensated current (ANGLE COMPENSATED I) is the measured secondary current with phase compensation applied. Tap compensated current (TAP COMP I) is the tap and phase compensated current actually used by the differential function. From this information, it is easy to see how the relay goes about compensating for magnitude and angle differences between the two sides of the zone of protection. The final two lines of the report give the most critical information. IOP is the operating current. SLOPE RATIO is the ratio of IOP to the restraint current (in this case it is the maximum of the two TAP COMP I currents). These values should be compared to the settings shown earlier in the report to determine if the relay is in a trip or alarm condition. Figure 21 shows the A phase currents before and after compensation plotted on a polar graph. From the information in Figures 20 and 21, it is easy to see that the internal phase compensation is the opposite of what it should be and that the currents were shifted 30 degrees the wrong way. In this installation the transformer being protected was actually a delta IA-IC/wye configuration and that the low side leads the high side by 30 degrees. Changing the transformer connection parameters in the relay’s settings, corrected the problem.

24

Protective Relaying Handbook — Volume 1

This facility of modern relays can also be used to simplify commissioning and documentation. To verify correct CT circuit connections, internal phase, zero sequence and tap compensation settings for the differential functions, load should be placed on the protected zone and a differential check record triggered, recorded, and examined. The check record can then become a permanent relay commissioning record.

Summary Differential protection is simple in concept. Measure the current that goes in versus what goes out. If there is a difference, there must be a short circuit within the protected zone and a trip should occur. When the protected zone includes a transformer, the situation is not so simple and

special considerations must be made. One of the greatest challenges is compensation for phase angle and magnitude differences. The paper describes the effects of many of the possible errors that can be made in installing and checking out a transformer differential circuit. Proper installation checks and final in-service readings can detect these problems and ensure reliable and secure operation. The paper describes these traditional final in-service checks. However, with modern solid state and numerical differential relays, traditional checkout procedures may not be capable of detecting all possible errors. For this type of relay, diagnostic routines and reporting functions can make up for this. It is important for the relay technicians and engineers to make use of these advanced features to ensure proper operation of the protection system.

Annoted Differential Check Record

Figure 20 — Annotated Differential Check Record

Protective Relaying Handbook — Volume 1

Figure 21 — In-Service Current Circuit Verification Form

25

26 Bibliography

1. Blackburn, J. Lewis, Protective Relaying Principles

and Applications, Second Edition, Marcel Dekker, Inc., New York, 1998

2. ANSI/IEEE C37.91-1985, IEEE Guide for Protective Relay Applications to Power Transformers

3. Criss, John, and Larry Lawhead, “Using Transformer Differential Iop Characteristics to Measure Near-Trip Conditions”, Protective Relay Conference at Georgia Institute of Technology, April 1997. Jim Closson received his BS from Southern Illinois University at Carbondale, and an MBA from the University of Laverne. Prior to rejoining Basler Electric as a Protection and Control Product Manager, he served as a Regional Application Engineer for Basler Electric. He has also held managerial and sales positions with Electro-Test, Inc. and ABB. He has taught courses on Electrical Power Systems Safety, Ground Fault Applications and Testing, and Power System Maintenance. Mr. Closson is a Senior Member of the IEEE and serves on the Power Distribution Subcommittee for the Pulp and Paper Industry Committee of the IAS and on the Transportation Subcommittee for the Petrochemical Industry Committee of the IAS. Michael Thompson served nearly 15 years at Central Illinois Public Service Co. where he worked in distribution and substation field operations before taking over responsibility for system protection engineering. He received a BS, Magna Cum Laude from Bradley University in 1981 and an MBA from Eastern Illinois University in 1991. During his years at Bradley University, Mike was involved in the cooperative education program and worked in electrical engineering and maintenance at a large steel and wire products mill. Mike is Senior Product and Market Manager for the Protection and Control Product Line at Basler Electric. Mr. Thompson is a member of the IEEE.

Protective Relaying Handbook — Volume 1

The Anatomy of a Shermco Field Service Engineer Reliable and Responsive Part of a large international team available 24/7

Back by the Best From the field, to our full-service rotating apparatus services

NETA-Certified Qualified to perform to the highest standards

Protective Relaying Handbook

Trusted Advisor Offering independent service across all brands

Safety First On record as an industry leader in safe work practices

Volume 1

Cutting-Edge Trained from utility transmission to low-voltage distribution

Our experience in commercial, industrial, generation and specialized sectors, such as oil refineries, pulp and paper, steel and wind power, uniquely qualifies us to handle the largest, most complex assignments. Our people are trained on the latest technologies and safety practices. And, we support them with full-service rotating apparatus services and a large fleet of service vehicles, ready to respond 24/7. Corporate Headquarters | 2425 East Pioneer Drive, Irving, Texas 75061 p. 972.793.5523, 888.SHERMCO | f. 972.793.5542 | www.shermco.com

Why trust your vital power services to anyone less? Count on an industry leader. Visit www.shermco.com and learn more today.

Published by InterNational Electrical Testing Association

Odds are, you won’t find a job anywhere that a Shermco field service engineer can’t handle. As one of the world’s most respected electrical maintenance and testing companies, Shermco delivers A+ service from utility transmission to low-voltage distribution.

Published by InterNational Electrical Testing Association

27

Protective Relaying Handbook — Volume 1

Motor Protection Fundamentals PowerTest 2001 (NETA Annual Technical Conference) Bernie Moisey Northern Alberta Institute of Technology

Summary

Voltage

To enter set points into modern management type motor protection relays for a specified motor the end user must be familiar with all of the motor characteristics and is able to interpret the technical data supplied by the manufacturer. Without this knowledge the protection scheme could result in one that over or under protects the motor.

The voltage on the nameplate of a motor may differ from the system nominal voltage, i.e. 4000 volts on the nameplate connected to a 4160-volt system. In most cases when the motor is started, the voltage at the motor terminals will “sag”. To ensure that sufficient voltage is present to accelerate the load the starting voltage must be calculated and then limits set with an under voltage relay. If the voltage “sags” on start the locked rotor/starting current will decrease. Because the motor is driven into saturation at rated starting voltage, the starting current is not directly proportional to voltage. This must be considered when entering set points for locked rotor protection. Setting alarm set points for current unbalance requires that one must be able to determine an acceptable current unbalance by converting the normal system voltage unbalance to current unbalance. Set points are also required for over voltage and reclosing when the residual voltage is present.

Motor Specifications The following motor data could be considered minimum requirements for a protection scheme on a large motor: Horse power; voltage rating; full load speed; type of motor; frequency; full load torque; breakdown torque; locked rotor torque; service factor; NEMA design; insulation class; symmetrical locked rotor amps at rated voltage; type of enclosure; maximum temperature rise at specified load; ambient temperature; kVA code; current at 100%, 75%, 50% and no-load; power factor at full load; no-load and locked rotor current; efficiency; cold and hot safe stall time; power factor correction data; load inertia; rotor inertia; load torque during the acceleration period; time-current and hot and cold thermal limit curves; motor starting and accelerating curves; speed curves at different voltages; performance curves; permissible starting sequence; minimum time between starts; number of starts per hour and residual voltage data.

Symmetrical Components Most microprocessor based motor protection relays use symmetrical components in thermal and unbalance algorithms. Some relays estimate positive, negative and zero sequence quantities while others use the actual sequence equations. A good understanding of these fundamentals is required to select appropriate set points and to design test circuits to verify relay operation.

Grounding Electrical systems may be ungrounded, direct or solidly grounded, low impedance grounded and high impedance grounded. In all cases the magnitude of the charging current or the line to ground fault current must be known. The ground element of the relay must be connected to detect this abnormal condition and disconnect the motor as quickly as possible. The two most common methods of connecting ground relays to the system are using a zero sequence current transformer and the residual connection. In each connection it is possible for the ground element to receive a false signal, which could result in the motor being taken off line. Compensation must be considered when determining set points to minimize nuisance trips. The residual connection uses three current transformers. False ground fault signals can occur due to unbalanced phase burdens, asymmetrical starting current and the normal mismatch of the three cur-

28 rent transformers. Compensation for these false signals can be achieved by increasing the pick up or by increasing the time delay. False signals can enter the ground relay through the zero sequence connection. If any triplen harmonic is present in the primary circuit this will pass through the zero sequence current transformer and appear as a ground fault. When two motors are connected to the same bus, the running motor can trip out when the other motor is started. A “sagging” bus voltage combined with the residual voltage and noise that is generated during the starting sequence can result in a trip. Compensation for the zero sequence connection is achieved by a short time delay set point, not instantaneous. In high impedance grounded systems, the neutral limiting resistor limits the fault current to a magnitude of 1 to 10 amps. High impedance faults may be difficult to detect and low set points may result in false trips. When this is the case, the use of a low pick up directional relay with an angle of maximum torque, current leading voltage should be considered.

Thermal Limit Curves Large motor manufacturers include thermal limit curves as part of the specifications. One is called the cold thermal limit curve and the other is referred to as the hot thermal limit. The cold thermal limit curve is the limit of the motor when the motor temperature is equal to or less than the specified ambient, usually 400 C. The hot thermal limit curve is the thermal limit of the motor when it is operated in the maximum ambient temperature, at specified rise and specified load. All thermal limit curves consist of the following three curves: locked rotor; failure to accelerate and running overload. The locked rotor and failure to accelerate are voltage dependent. These limit curves usually are plotted on semilog paper and the slope of the hot curve can be different from the slope of the cold curve. When the limit curve is given for a motor that can be started at two different voltages, 100% voltage and 80% or 90% voltage, the locked rotor thermal curve appears as a straight line and the failure to accelerate thermal limit curve is for the lower starting voltage. When the starting voltage is determined for a specific motor, the limit curves must be altered to reflect this condition. The time between the cold safe stall time and the hot safe stall time can be of short duration, long duration or, in the case of a motor that is “ring” limited, the hot safe stall time can be equal to the cold safe stall time. Also the acceleration time can be greater than the safe stall time.

Thermal Protection Thermal protection includes protecting the motor during starting, acceleration and running. Manufactures of microprocessor based motor protection relays will supply I2t protection curves that are available in a database and the end user selects one that “fits” the motor characteristics. Other manufacturers allow the end user to generate a custom I2t protection curve. This is accomplished by entering

Protective Relaying Handbook — Volume 1 time – current values in a look-up table. The custom curve provides flexibility and results in a more reliable protection scheme. When designing the thermal protection scheme the engineer or technologist must determine the degree of protection. Is the motor to be over protected or allowed to operate at the maximum thermal limit? To accomplish this, an understanding of the relay’s thermal algorithm is required. Microprocessor relays allow the I2t protection curve to “move downwards” when the motor temperature increases. This is accomplished by multiplying all values in the time – current look-up table by a constant. The complete protection curve moves “up or down” by the same proportion. The thermal algorithm can be biased by stator RTD inputs or if RTDs are not used the biasing is accomplished with a full load thermal capacity reduction set point. Hot and cold thermal limit curves can be parallel or have different slopes and may have acceleration time that is greater than the safe stall time. In each case care must be taken to insure that the motor is protected in all three modes of operation, i.e. the I2t protection curve must never intersect the thermal limit curve. Where motors have a variable starting voltage and a long acceleration period, one may consider selecting a motor protection relay that has a voltage dependent I2t protection curve. When using this type of relay it is necessary to manipulate the thermal limit curve supplied by the motor manufacturer or at the time of ordering the motor, request limit curves for minimum and maximum starting and accelerating voltages. Motors with acceleration times greater than the safe stall time may fail to restart after a normal shutdown. If this situation arises then it is necessary to adjust the thermal algorithm so that the hot motor inrush current does not intersect the I2t protection curve.

Protection – Phase Current Set points are required for over current conditions that result from three phase and phase-to-phase faults that may occur on the load side of the current transformers. Mechanical jam or rapid trip set points may be required to prevent the motor from stalling when maximum or breakdown torque is exceeded. Under current protection may be used as secondary protection to protect the mechanical load from damage, i.e. a pump that uses the product as lubrication. A phase sequence set point may also be required. By entering the proper sequence the relay now has the ability to select the proper symmetrical component equations and prevent operation in the reverse direction. When using an instantaneous element to clear faults insure that the disconnect has the required interrupting capacity. For some contactor applications it may be necessary to disable the instantaneous device. Also the asymmetrical starting current must be allowed for. If the sensitivity is too great, a small difference between the starting current and the maximum three-phase fault current, consider using differential protection. Differential protection requires that all six leads from the motor be accessible.

29

Protective Relaying Handbook — Volume 1 Unbalanced Protection Most microprocessor type motor protection relays have two types of unbalanced protection. One type uses alarm and trip set points. The other type of unbalanced protection involves biasing the thermal algorithm. When the motor draws an unbalanced current, the relay will calculate an equivalent balanced current that will produce the same motor heating. This equivalent current, not the actual motor current is used to determine the trip time. The equivalent current must be greater than the pick up for the algorithm to be enabled. Depending upon the manufacturer of the relay the K factor, a ratio of negative sequence rotor resistance to positive sequence rotor resistance may be pre-determined or entered by the users. The ambient temperature plus the rise under ideal conditions plus the rise due to the unbalanced current drawn determines the temperature of a motor. The temperature rise due to unbalance depends upon the amount of unbalance and the amount of load on the motor. Care should be taken to prevent the motor from being disconnected when it is not stressed. Most algorithms have flexibility that allows the end user to determine at what percent unbalance and percent overload the algorithm is enabled. Motors with a service factor of 1.0 have the I2t protection curve enabled at 1.15 times the full load current, while a 1.15 service factor motor allows the protection curve to be enabled at 1.25 times the full load current. By entering the appropriate service factor as a set point the relay then determines at what unbalance the algorithm is enabled. Other relays allow a set point to be entered as to when the protection curve can be enabled. Typical values for enabling the protection curve are 1.01 to 1.25 times the full load current of the motor. Knowing the unbalanced algorithm equation allows the protection engineer to calculate the percent unbalance that is required to enable the algorithm. A typical equation is,

I2

Ie

2

K

I1

2

I2 is per unit negative sequence current.

Ie is the equivalent current calculated by the relay when the motor draws unbalanced current.

I 1 is per unit positive sequence current, or load component. K is the ratio of negative sequence rotor resistance to positive sequence rotor resistance.

IEEE states that for every 3% voltage unbalance the temperature rise of the motor will increase 25%. Motor protection relays do not use voltage unbalance in algorithms, they use current unbalance. An approximation of converting voltage unbalance to current unbalance is, for every 1% voltage unbalance at the terminals of the motor, the percent current

unbalance will be approximately equal to the per unit starting current expressed as a percent.

Control The emergency start feature allows the operator to restart a hot motor by resetting the thermal algorithm to zero percent thermal capacity used. In some cases the relay will not allow an emergency restart if the motor temperature exceeds the stator RTD trip set point. Start inhibit, when enabled, prevents a restart until sufficient thermal capacity is available. Thermal capacity required to start the motor is a “learned “ feature. Acceleration timer, when enabled, will lock out the motor if it does not come up to speed in the specified time.

Backspin timer prevents a restart when the direction of motor rotation is opposite to the norm, i.e. a down hole pump. Time between starts, is a set point that controls the minimum time between when the motor is first started and when another start is allowed.

Anti jogging, when enabled, prevents a series of rapid startstop operations. It can be used for a lock out condition that prevents a restart when residual voltage is present.

`Phase reversal prevents the motor from starting in the “wrong” direction.

Some relays have auxiliary contacts and logic that allow the motor to start on a reduced voltage, i.e. wye-delta, autotransformer, etc.

Starter failure is a signal from the starter to the relay that implies that the contacts have changed state. This is the same as the 52b contact signal of an electrically operated breaker.

Conclusion To properly protect a motor the end user must be familiar with the motor characteristics and load requirements. An understanding of the microprocessor based motor protection relay algorithms allows for flexibility. Do not disable algorithms because you do not understand them.

Reference “Concepts of Motor Protection” written by B.H. Moisey Bernie Moisey has been an instructor at the Northern Alberta Institute of Technology for 33 years and is currently teaching in the power systems and protective relaying section. Bernie has presented motor protection seminars in Canada, United States, South America, and Australia. He acts as a consultant for major manufacturers of protective relays designing and upgrading protection algorithms. He is actively involved in application engineering.

30

Protective Relaying Handbook — Volume 1

Meaningful Testing of Numerical Multifunction Protection Schemes PowerTest 2001 (NETA Annual Technical Conference) Jay Gosalia Doble Engineering Company

Introduction: Generally, an electromechanical relay is a single function module. Multiple electromechanical relays wired together create a complete protection scheme. Electromechanical relay design is not very complex and is relatively easy to test. Traditionally, complete protection scheme testing was not done. Test engineers tested each individual component of the scheme and assumed that if the individual component worked correctly then the complete scheme worked correctly. This is true to some extent as the scheme is relatively simple; hence, the assumptions stated above provided satisfactory results. Advances in Digital Signal Processing technology and microprocessor design allow the user to do more with protection today than older technology provided. These advances now allow the complete protection scheme to be designed by combining different protection functions in software logic. With modern, high-power computers, the protection designer accurately simulates power system conditions using tools like EMTP, ATP, etc. to develop new protection methodology. This new methodology provides efficient protection where protection functions are implemented using a numerical algorithm. Protections are now designed to work effectively under power system disturbance. Efficient protection is not the only benefit of this advanced technology. Additionally, these advances allow the protection designer to put a host of different features in a box resulting in a reduction of the protection packaging size. Numerical multifunction protection provides a myriad of different options to cover a wide range of applications. Changing the software logic functions to customize the protection scheme for the individual application is easy. The degree of customization available to the user depends upon the design of the protection scheme. This flexibility improves the functionality of the protection multi-fold, and at the same time makes the protection scheme complex. This added complexity increases the possibility that the

engineers will commit errors in the protection configuration. The relatively complex protection design poses a number of challenges for the relay test instruments. The traditional test methods will not be sufficient to evaluate the performance of multifunction numerical protection. A Protection test instrument should provide the functionality to enable the user to evaluate the capability of the multifunction digital protection scheme. This paper describes how to test multifunction numeric protection effectively. It also discusses test instrument requirements for testing and evaluation of multifunction protection schemes. Use of today’s technological advances in the personal computers, in testing and in the evaluation of a protection scheme under actual power system condition using COMTRADE standard and simulation programs like EMTP and ATP are discussed in this paper.

Numerical Protection Functions and Settings: Numerical protection provides a host of protection schemes along with control and metering functions in one package. The user configures the protection to suit the application by enabling certain functions and disabling unnecessary functions. The connection between the various functions is done through software links. Software links are equivalent to a wiring connection between the protection elements in an electromechanical protection scheme. The difficult part to the user is that unlike electromechanical schemes, in numerical protection it is not apparent which elements are active or how they are interconnected. In numerical protection, these details are only available via Human Machine Interface (HMI) if provided or by connecting a PC, typically via serial port, to the protection. Some of the line protection scheme can have setting parameters as high as 250. Understanding how to calculate settings and how to implement the various function links are not trivial tasks.

31

Protective Relaying Handbook — Volume 1 To apply the protection for the application, perform the following tasks: • Enable the required protection elements

• Disable the protection elements which are not required

• Calculate the settings for all required protection elements

• Set up the links between various protection elements typically using the software function links

• Assign output contacts for trip, close, control, and annunciation functions • Assign logic inputs for various functions

• Configure the protection for metering display and set up communication option to SCADA or sub-station automation system if required These tasks require careful evaluation of the application and a good understanding of the capability of the numerical protection. Once this task is completed, the user needs to create an application verification plan to confirm that the protection meets the intended need.

Dynamic-State Testing: As described above, numerical protection provides many protection and control functions in one box. It is not an easy tack to verify the proper setting and functionality of each individual function. Traditional Steady-state Testing of an individual function requires the protection element under test to be isolated and at the same time to deactivate the other protection element, so that it does not interfere with the protection function under test. Reconfiguring the protection typically allows user to test protection using Steady-State techniques. Some of the numerical protection can be put in a diagnostic mode or test mode, which facilitates testing of the protection element for Steady-State Test. This is a time consuming task and requires intimate knowledge of the protection scheme. To verify the application, Dynamic-State Testing of the protection is the logical choice. To simplify the testing of numerical protection DynamicState Testing can be used. Dynamic-State Testing means testing under simulated power system conditions. A report from IEEE Power System Relaying Committee entitled Relay Performance Testing discusses how Dynamic-State Testing and transient simulations provide a far better understanding of how the protection system performs. By making a profile of the operation of the scheme, malfunctions can be found faster because it is easier to identify the changes in areas that do not operate the way they are expected. DynamicState Testing allows fundamental frequency components to switch synchronously and thereby represent power system events. The synchronous switching between the pre-fault, fault, and post-fault conditions allows users to simulate a power system event easily and quickly. Dynamic-State Testing does not simulate transient components of the power system event. In most applications, this may be acceptable, as what is being verified is the configuration of the scheme,

protection algorithm, and interaction between the various elements of the numerical protection. Some modern-line protection schemes use a superimposed component to detect the direction of the fault. For such type of protection, transition between two states of the Dynamic-State simulation needs to be examined to confirm that transition from one state to other state does not create problems with the protection. Figure 1 shows the functional block diagram for Dynamic-State Testing of a protection scheme. Protection Under Test

Source Inputs

Protection Outputs

Logic Inputs

Power System Simulator V

I

State change Definition

Power System State Simulation Controls Sources and outputs

Figure 1 — Functional Block Diagram for Dynamic Testing of a Protection Scheme

Dynamic-State Testing, as mentioned above, controls the voltage and current applied to the protection under test. By simulating voltage and current phasors for various states of the power system, protection response can be analyzed. The transition from pre-fault to fault to post-fault state can be programmed by the user by monitoring the response from the output contacts of the protection. During every state, logic outputs can be controlled to simulate events like carrier signal, breaker position, etc. Dynamic-State Testing, thus by controlling current/voltage phasors, logic outputs and monitoring protection output contact response in a real time can simulate power system events accurately which allows user to test protection response easily.

Distribution Numerical Protection Scheme: Consider the typical distribution numerical protection scheme, which includes the following protection functions in a package. • Device 51 phase and ground time over current functions (Directional or non-directional) • Device 50 high set over current protection typically protection provides 2 high set over current protection • Device 79 Multi shot auto-reclose function • Device 25 Sync check function

• Device 87 Bus differential protection

• Device 50 BF Breaker fail protection scheme

• Device 81 under and over frequency protection

32

Protective Relaying Handbook — Volume 1

Some other protection functions may include load-shedding scheme, VT/CT supervision functions, and Cold start supervision. Discussion of the complete protection scheme and its application is beyond the scope of this paper.

Protection Testing: How Much And When? Before applying the protection scheme, testing needs to be performed to confirm the suitability of the scheme for the application. When testing a protection scheme, it is important to test the configuration, which is intended to be the configuration in service. It is impossible to test all possible configurations for the scheme. For evaluation of the protection scheme, it is necessary to check the configuration that exercises the maximum elements of the scheme. Another question that arises here is when and how much testing should be done when the firmware of the protection changes? The manufacturer of the digital protection scheme upgrades the protection firmware time to time. The reason for these changes can be a problem fix or improvement in the protection performance. This practice is going to continue, as such changes most of the time do not require changes in the hardware or scheme wiring. It is mostly the software changes. When protection is in the service and firmware changes are required, then the protection scheme should be at least tested to ascertain the proper operation of the protection. Figure 2 describes how some functions of a distribution protection scheme can be tested using Dynamic-State Testing. 52a I

Current Detector

& Breaker Fail Timer

Protection Algorithm

BFR Trip

Trip

Figure 2 — Typical Breaker Fail

Breaker Fail Protection: Breaker fail protection provides the protection when a breaker fails to trip. Protection senses the fault and issues a trip command. If the breaker fails to clear the fault, the breaker fail timer times out, and if the protection continues to sense current through the breaker, then it provides breaker fail output. The breaker fail output contact generally picks up the lock-out relays to trip the associated breaker to clear the fault. Figure 2 shows a typical block diagram of breaker fail logic. To test breaker fail protection properly, the following power system conditions should be simulated: • Breaker fail condition

• Normal trip condition

• Delayed breaker trip condition

Dynamic-State Testing can simulate all of the above conditions. Dynamic-State Testing allows the user to simulate a power system event by creating different states of the power system.

• First state of the test simulates the normal load condition, which is Pre-Fault State. This condition should be simulated for time sufficient for protection to come to a quiescent state.

• Fault State should follow the pre-fault condition by simulating fault current and fault voltages. Test instrument in this state should monitor the response of the output contacts from the protection scheme. The protection test instrument should monitor trip output and breaker fail output contacts along with other protection contact outputs. • Program the test instrument to simulate Post-Fault State on sensing the operation of breaker fail output contact. It is important that the test instrument should monitor all or as many output contacts as possible of the protection scheme. This allows the user to analyze the protection performance by checking response of the output contacts. The user can compare the protection response with the expected response to analyze the operation. By programming the transition from Fault State to PostFault State on trip contact, normal trip condition can be simulated. The test instrument should be able to simulate breaker operation by simulating breaker trip and close operation with applicable delays like three-cycle trip time and five-cycle close time. Programming breaker trip time to a desired value can simulate delayed breaker trip. The effect of the 52a contact on the logic can be analyzed by simulating 52a contact from the test instrument. Similarly, if the logic is effected by any other logic input the test instrument could simulate the same. By recording the transition of protection contacts, protection trip time, and breaker fail time delay, the resetting time of the current detectors and the breaker fail margin time can be measured. These parameters are very important for the proper operation of the breaker fail protection. As mentioned above by monitoring as many output contacts as possible the proper operation of the protection scheme can be ascertained.

Bus Protection Logic: One common feature in distribution numerical protection is simple bus protection logic. Microprocessor based protection allows simple logic bus protection to be included for distribution protection. Figure 3 describes the simple bus protection logic.

33

Protective Relaying Handbook — Volume 1 Incomer Block Trip

F1

Numerical Protection Reclosing Logic:

F2 A

of logic outputs should be available. When all of the logic outputs are independent of each other, they provide ample flexibility for simulation of any power system event the user conceives.

B

C

Figure 3 — Simple Bus Protection Logic

Protection at the incomer provides backup for A, B, and C feeder protections. Close up fault on any of the feeders A, B, or C (Fault F2 in the above figure), both feeder and back up protections high set element will pick up. If the fault is close up then it is difficult to be distinguished by back up protection as bus or feeder fault. The typical logic employed checks that if high set element of incomer protection picks up and receives no block operation from the feeder protection then it is a bus fault. (Refer to the Fault at F1 in Figure 3.) Incomer protection should trip the incomer breaker after a delay of 50-mSec (or any time set by the user). This assumes that this is a radial system. Testing of this logic at the incomer protection can be accomplished by Dynamic-State Testing. The first step is to define the states of the power system to simulate the feeder fault. • First state of the test simulates the normal load condition, which is Pre-Fault State. • Fault State should follow the pre-fault condition by simulating fault current and fault voltages. Test instrument in this state should simulate the block-input signal using logic output.

• Program the test to simulate Post-Fault State on sensing the operation of trip output contact. During this state normal voltage and no current should be provided which simulates breaker open condition. By not exercising the block signal, a bus fault can be simulated. Refer to the Fault at F1 in Figure 3. During this condition, monitor all the output contacts to analyze the response of the incomer protection for the feeder fault. Testing of such type of scheme is relatively simple when all the elements of the scheme are available for testing. However, it may not be feasible all the time to have a complete scheme available, typically when evaluating the protection for application in the lab. Therefore, if a complete scheme is not available, Dynamic-State Testing can simulate the operation of the protections not available during evaluation. Accurate control of logic outputs for Dynamic-State Testing allows user to simulate various fault conditions for protection scheme. For this purpose, an adequate number

Testing and evaluating the Reclosing logic in the multifunction protection requires control of multiple inputs to the protection and monitoring various output signals from the protection on time basis. The test instrument can be programmed to simulate different power system events such as transient fault, permanent fault, three-phase fault, multiphase fault, single-phase fault or combination of any of the above. During the fault simulation, the test instrument can be programmed to simulate breaker closing/tripping operation, operation of other associated control and protection element by controlling logic outputs of test instrument. During the simulation, test instrument can be programmed to record the protection response, which can be evaluated by the user.

Numerical Line Protection: Numerical line protection provides many enhanced functions for the effective protection of the transmission line. Over and above the 3-zone 3-phase, phase and ground impedance protection, numerical protection can provide the following enhanced functionality: • Switch on to the fault

• Power swing blocking • VT/CT supervision • Long memory

• Reclosing and Synchrocheck functions • Time over current back up • Breaker fail protection

• Distance to fault function Dynamic-State Testing tests the above functions easily by simulating power system phasors, controlling logic output for simulation of power system events and recording protection response. Dynamic-State Testing allows users to create test plans independent of the protection manufacturer. Dynamic-State Testing simply simulates power system conditions and allows users to analyze the response of the protection.

Transient Testing: To test state-of-the-art protection such as a superimposed directional comparison protection, some special capabilities are expected from the test instrument. This type of protection operates very fast and works on the transient component during the fault. For such a protection type, Dynamic-State Testing may not be adequate. Transient Testing provides the accurate simulation of power system

34 events. It is an important tool the user has, to perform a thorough evaluation of protection scheme operation. Transient Testing can be performed using the data created by an EMTP program or data recorded by DFRs. To check such types of protection, Transient Simulation should be used. Actual DFR record or an EMTP/ATP generated record can be used to test such type of protection. COMTRADE data files help the user to perform Transient Testing. Modern DFRs and digital protection can record power system events in a COMTRADE data format. It is also possible to create COMTRADE data files from simulation programs such as EMTP and ATP. The COMTRADE format is a collection of data around an event point, recorded at regular time intervals, to define the characteristics of voltage, current and the status of digital channels before, during and after an event. High playback rate of the transient event using COMTRDADE data is necessary to simulate a power system event accurately. Test instrument should be able to play the transient data at high playback rate. To simulate dc-offset, dc-coupled amplifiers are required for Transient Testing. EMTP/ATP can generate transient data at 50-microsecond intervals to simulate a transient event accurately. This requires that the test instrument play back the data at rate of 20 kHz. Test instruments, which can play transient data at this rate, can easily have a bandwidth of six kHz, which is adequate for simulation of transient conditions. Along with the control of the voltage and current values on a sample-by-sample basis, an instrument should be able to simulate logic output with 0.1-mSec accuracy and should be able record response of the protection with 0.1-mSec accuracy. The ability of test instrument to record multiple inputs and timers along with the ability to control logic outputs with 0.1-mSec accuracy is very important in testing modern multifunction numerical protection.

Test Instrument Requirements: A test instrument should provide a sufficient number of logic inputs, in order to allow the user to monitor multiple contact output from the protection under test. The numerical protections’ output contacts are typically under software control. The output contacts can be allocated to perform different functions depending on the application. To evaluate the protection performance, it is important to simulate power system conditions, record, and analyze the status of all output contacts during test interval. It is important to ensure that the contacts, that operated were supposed to operate, and that the contacts that did not operate were not supposed to operate. At the same time it is necessary to ensure that pick up and drop out timings of the contacts are as expected. The instrument should also be able to start multiple timers to analyze the timing sequence of the various events. To simplify the analysis of the event, the timer should be able to be started and stopped by the user by defining the trigger where trigger definition includes logical combination of inputs.

Protective Relaying Handbook — Volume 1 Summary: Advances in the microprocessor and digital signal processing technology allow the user to employ multifunction numerical protection for the protection of power system apparatus. Modern state-of-the-art protections are designed to work under power system conditions. Software logic provides enormous flexibility to the user to design protection logic for the application. To properly test and evaluate the protection for the application, Dynamic-State Testing or Transient Testing should be used. Dynamic-State Testing simulates power frequency phasors, which can be useful in evaluating the protection algorithm. Transient Testing simulates power system events accurately by the simulation of transient conditions. The simulation of power system conditions, includes simulation of voltage and current values along with the simulation of the events like breaker operation, carrier signals, etc. Numerical protection can only be tested effectively by playing back such events. The modern test instrument employs similar technology as used in the protection, which allows simulation of the power system events easily and effectively. A modern test instrument provides dc-coupled amplifiers with adequate power along with 8 to 16 isolated logic inputs and outputs. Such a high number of inputs and outputs are necessary to test numerical protections effectively. Response time of such inputs and outputs should be on the order of 100 microseconds. To simplify the analysis of protection response, test instrument provides timing functionality where time interval can be recorded for user specified events. The start and stop functionality of the timer based upon the logical combination of inputs further simplifies the protection response evaluation.

References: 1. ANSI/IEEE C37.111.1991 Standard Common Format for Transient Data Exchange (COMTRADE) for Power Systems

2. IEEE Special Publication # 96TP115-0 Relay Performance Testing, Power System Relaying Committee, Report of Working Group I 13. 3. Jodice, J.A. and Giuliante, A.T., “A New Philosophy for Protection Diagnostics,” Proceedings of the Sixty-Third Annual International Conference of Doble Clients, 1996, Section 6-7. Jay Gosalia is presently working at Doble Engineering Company as Product Manager: Diagnostic Instruments. He has over 22 years of experience in the power engineering field, 17 of which have been dedicated to the development and marketing of protective relays. Before joining Doble Engineering, he was the US Sales and Marketing Manager at GEC Alsthom T&D, Protection and Control Division for 13 years. Prior to GEC, he worked at ABB in the Circuit Breaker Division as a design and development engineer. Mr. Gosalia, an active member of the IEEE Power System Relaying Committee, has authored several technical papers on protective relays. He has a BS in electrical engineering and MS in computer science.

35

Protective Relaying Handbook — Volume 1

Using Dynamic Testing Techniques for Commissioning and Routine Testing of Motor Protection Relays NETA World, Summer 2001 Issue by Benton Vandiver III, P.E. Omicron electronics Corp. USA

Traditional methods of testing modern motor protection relays applied to large motors can be a daunting task due to the specific motor data required to determine the numerous relay settings for proper protection. The use of a modern three-phase test system and recording/playback of dynamic test files can provide many timesaving and analytical advantages resulting in reduced costs when commissioning motor protection and later routine testing.

Traditional approach and inherent problems Applying a motor protection relay typically requires specific knowledge of approximately 40 parameters of a large motor as a minimum. Calculations are required to determine the actual settings to be used in the modern motor protection relay (comprising 50-150 settings) based on the motor data and its actual application. It is not uncommon to apply a motor to a power system where the nominal system voltage differs from the motor nominal rating. This alone affects several critical settings necessary to adequately protect the motor. Improper settings will result in over or under protection resulting in various problems. These complexities make testing very difficult with conventional single-phase test sets. Even traditionally defined tests using a three-phase test set can not prove the settings are correct for actual operating conditions of the motor in its application. Ramped output of voltage(s) and current(s) or even a sequence of static output states simulating start, load acceleration, and full load scenarios do not represent the actual power system conditions. The numerous variables and system parameters cause simple test calculations to fall short in verifying correct motor operation and protection for the specified application. Creating realistic values for testing voltage sag, zero-sequence voltage, negative-sequence current, asymmetrical starting current, or locked-rotor current

settings, just to mention a few, is nearly always an educated guess. A more precise and predictable test method would certainly be preferred.

The dynamic testing approach Utilities have always understood the advantages of dynamic fault testing to verify system protection schemes. Due to the technical advancements in modern protection relays and their needs for dynamic waveforms to prove their appropriate protection operation, the dynamic testing method is becoming the only testing choice. Large, expensive power system simulators have been used for these tests in the past. In recent years, advanced DSP technology and power electronics now provides this capability in cost-effective, portable, modern three-phase test systems. This test method utilizes the IEEE Comtrade Transient Data Exchange standard as the source of the dynamic test files. These Comtrade files can be produced using PC-based mathematical power system modeling programs or directly captured transient events from the actual power system using several available recording devices (fault recorders, protective relays, power quality monitors, digital meters, etc.). Evaluating a system misoperation or testing new protection ideas is easily performed using these dynamic files. By replaying the recorded files with a modern three-phase test set to the protection relay, the dynamic response can be analyzed based on the expected response. It is this method of recording the power system waveforms, their analysis, and replaying the recorded dynamic Comtrade files which can simplify the challenge of testing a modern motor protection system. The calculation of critical motor protection settings for each application becomes uniquely deterministic, for instance, based on direct analysis of the recorded waveforms during a motor start sequence. This test procedure is surprisingly simple, and future routine tests become virtually automatic.

36

Protective Relaying Handbook — Volume 1

Getting the dynamic data As noted, sources to obtain the dynamic waveform test files may be more readily available than we think for the motor protection testing challenge. Modern motor protection relays are capable of capturing the measured voltage and current waveforms in oscillography files. These can easily be converted or directly exported in the Comtrade file format for use by most modern test sets. Other system devices like power quality meters are typically applied in these applications and also can be used to capture the waveform data required. Other sources may include digital fault recorders, multifunctional digital meters, and power quality analyzers. The newest three-phase test systems, however, are also capable of monitoring and capturing these power system waveforms, essentially becoming a digital fault recorder. This then provides an all-in-one testing solution for the modern motor protection scenario.

A new motor protection testing procedure Based on available technology and information, it is obvious that quality data eliminates the guesswork of calculating the critical settings for a modern motor protection application. One such approach would be: 1) 2)

3)

4) 5) 6)

7) 8)

Use base motor data and known power system data to enable the minimum protection in the motor protection relay and still allow safe motor starts.

Capture the motor’s voltage and current waveforms, (preferably in Comtrade format) during the typical commissioning procedures. The more start and run scenarios the better. Save each scenario as a separate file.

9)

Save all Comtrade test files and compile a total test plan for this motor protection. When the next routine test is scheduled, recall the test plan and replay them to verify the relay’s correct response.

10) Concurrently, during the routine maintenance, record any new motor starts for use in motor performance assessment.

Eliminate the trial and error and reduce testing time By utilizing the recorded dynamic files of the motor, the proper settings are directly determined by analysis, and the motor is spared the needless exercise of unnecessary motor trips, starts, and lockouts. Proper motor protection settings for thermal damage is easily determined from the actual negative-sequence current and zero-sequence voltage values determined from the same analysis of the test files. Commissioning of new motor installations is dramatically reduced using this new procedure. Simultaneously, benchmark performance data for both the dynamic file capture procedure and comparing them to the original commissioning dynamic test files, measuring key parameters. This also provides the inherit ability to perform a routine test automatically using the saved commissioning test plan. The advantage of utilizing the identical test files and expecting the same relay test results is obvious. This also establishes a repeatable historical trend of the relay’s performance, which is only possible by consistently using the identical test files.

Conclusion This procedure has been used in actual commissioning of several large motor protection schemes that resulted in:

Use analysis software to view the captured motor starts. Determine critical settings by measuring actual voltage sag, maximum starting current, acceleration time, current unbalance, CT mismatch, negative-sequence current, zero-sequence voltage, full-load current, and other actual performance data.

• The advantage of utilizing actual motor/system performance data to determine the protection settings that eliminated over/under protection issues.

Replay the Comtrade test files to the motor protection relay and assess it for correct response and operation.

• Availability of dynamic commissioning data, where future comparative analysis has the potential to identify trends before problems result.

Use the actual performance measurements to then make the calculations required for the remaining settings of the motor protection relay.

Utilize the editing features of the analysis software and modify copies of the recorded Comtrade files to decrease/increase voltage and current values to fault levels, increase full-load current, increase acceleration times, etc. and save as additional test files.

Playback these fault files to the motor protection relay and ensure it protects the motor as expected. Place the motor in service with confidence the protection is optimized.

• Significantly reducing the total commissioning test time and time to in-service.

• Grounding and unbalance problems were identified and resolved before the motors were placed in service, again saving potential downtime.

Benton Vandiver III, P.E. is currently Technical Director for OMICRON electronics Corp. USA in Houston, TX. A 1979 BSEE graduate of the University of Houston, his primary responsibilities include market-focused product development, strategic sales, test application development, and product training for Omicron products in North and South America. An IEEE member, he has authored or co-authored many technical papers for various conferences and journals in North America over his 23 years in the power systems industry.

37

Protective Relaying Handbook — Volume 1

Commissioning Numerical Relays Part One (See Protective Relaying Handbook – Volume 2 for Part Two)

NETA World, Summer 2001 Issue by James R. Closson and Mike Young Basler Electric Co.

Modern numerical relays have many new features that were not available in electromechanical or analog designs. These new features include setting groups, programmable logic, and adaptive schemes. Although these features make numerical relays very powerful, they also create a need for reviewing commissioning methods. Although there are several references regarding commissioning of electromechanical relays, there are no written standards that address the testing and commissioning of multifunction, numerical protective relays. Therefore, most methods employed today are based on experience. Although there are many methods that give good results, this article suggests one approach for changes to commissioning tests and revised documentation of relay settings. Commissioning protective relays requires three primary tasks relay personnel should perform: • Calibration of the relays • Functional tests • In-service readings Relay calibration confirms that the relay will respond according to design and set point when voltage and/or currents are applied at the relay terminals. Functional testing confirms that the proper breakers trip or close according to the design when the relay contacts close. The functionality of all ac and dc schemes should also be checked. Finally, in-service readings are taken as soon as the equipment is placed in service and has load current flowing. In-service readings confirm that, with a given load present, the proper voltages and currents appear at the relay terminals.

Calibration and functional testing of relays must be done before the equipment is placed in service. In-service readings must be taken immediately after load is on the equipment. The equipment is not released to the dispatcher or plant operator until the in-service readings are correct. There are other commissioning tasks important to protective relays such as testing instrument transformers, meggering control cables, confirming transformer taps, and so on. However, this paper will focus on calibration, functional testing and in-service readings as areas directly affecting the commissioning of numerical relays. For the purposes of this paper, electromechanical and solid-state relays will be referred to as “traditional” devices while multifunction, microprocessor-based designs will be referred to as “numerical.” Although there are significant differences in electromechanical and solid-state devices, the methods used for testing and commissioning are similar, whereas numerical relays must be approached differently.

Commissioning Traditional Relays Relay calibration is performed using the manufacturer’s instruction manual and the relay setting sheets. In the simple overcurrent example provided in Figure 1, the setting sheet would include identifying information about the station name, the feeder number, and relay model numbers. The actual settings are: • CT ratio • 50 element tap • 51 element tap • 51 element inverse time curve selection • 51 element time dial.

38

Protective Relaying Handbook — Volume 1

Figure 1— Overcurrent circuit ac schematic

This information will readily fit on one sheet of paper. Setting records, in either hard copy or electronic form, are easily stored. Once the settings are determined, they are applied to the relays by selecting taps, adjusting dials, or setting switches. Secondary current values are then injected into the relay using a test set. Pickup of the 50 and 51 elements is checked against the settings, and adjustments are made to bring the relay within calibration limits. Timing tests of the 51 elements are also made to ensure the time dial and curve settings are correct. The output contacts are monitored during these tests to verify relay operation. Functional testing, also called circuit tests or trip checking, is another keystone activity of commissioning. This is not the place to cut corners. The intent is to confirm that the protection and controls work as intended and also that they have no unintended consequences. Making checks to see that the design works right is called a positive test. Making checks to see that the design does not work incorrectly is called a negative test.

Positive Tests When the phase overcurrent relays in Figure 2 operate, the circuit breaker will trip directly. In order for the ground relay to trip the breaker, the cutout switch must be closed. Each overcurrent relay element is operated one at a time to confirm that each works. The relay contact should be forced to close with the test set instead of applying a jumper across the contact. Targets are confirmed after each trip. Open the ground cutout switch, then attempt to trip with the ground relay; the breaker should not trip. With the ground relay trip contact still closed, turn the ground cutout switch back on and confirm that the breaker trips. This proves the cutout switch prevented tripping. These are all positive tests. The negative tests are more difficult to define. The relay engineer must look for ways the circuit could operate in an unintended manner. For example, when the circuit in Figure 2 is taken out of service, the fuses are pulled or the switch is opened.

Figure 2— Simple overcurrent circuit dc schematic

If the fuses are not correctly labeled, the wrong circuit could be de-energized. Personnel working on the circuit could be harmed or inadvertently cause the breaker to trip because the fuses were labeled incorrectly and control power was never removed from the circuit under test. This is an unintended consequence that would not be discovered by confirming only that the circuit operates correctly under normal conditions.

Negative Tests Figure 3 is an example of a circuit similar to the one in Figure 2 with the drawing changed to reflect a circuit wiring error. Rather than the ground cutout switch being present only in the ground relay circuit, it is now in the tripping circuit for all relays. If the same positive testing procedure were followed as before, this wiring error would not be found. All relay test results would be positive, including cutout switch actuation while testing the ground relay. If testing were terminated at this point, this wiring error would go undetected. The ground cutout switch would open all tripping circuits when operated. A negative test can be performed by confirming that the phase relays trip the circuit breaker while the ground cutout switch is in the open position. Complete negative testing involves checking every possible combination and permutation in the circuit. Although not extremely difficult in a simple circuit, it can become very complex when there are 20 to 30 circuit elements. Because it may not be practical to check every possible combination, test engineers should concentrate on the ones that are most common, such as incorrect wiring or identification.

39

Protective Relaying Handbook — Volume 1

Figure 4 — In-service readings Figure 3 — Miswired overcurrent circuit dc schematic

Commissioning Numerical Relays In-Service Readings The purpose of in-service readings is to verify that the correct quantities of current and/or voltage appear in the relay. In-service tests are performed with a phase angle meter, voltmeter and ammeter. The in-service check will prove the ac wiring, CT and VT ratio, the direction the relay is looking, phasing, direction and magnitude of switchboard metering, and direction and magnitude of SCADA metering. Several weeks of manpower are invested to get a line terminal ready for commissioning. Therefore, the value of an hour expended to perform a complete in-service test is inexpensive insurance. Prior to releasing equipment for service, power flow in the primary bushings must be determined in order to compare it to secondary values. Using Figure 4 as an example, current readings can be taken in the A-phase primary bushing with a tong ammeter on its extension stick. Secondary current and phase angle readings can then be taken in the current coil of the relay with an ammeter and phase angle meter. By taking the CT ratio into account and comparing the two current readings, CT ratio and appropriate current to the relay coil can be verified. The phase angle reading compares A-phase current to a reference voltage to prove the relay is looking at the polarity of A-phase current and not some other phase. When these readings are confirmed to be correct, the equipment can be released for service.

Although the same basic steps apply to commissioning numerical relays, the nomenclature changes somewhat. Where springs and drag magnets or extender boards provided the ability to calibrate traditional relays, the modern numerical relay is calibrated at the factory before shipment. Therefore, the numerical relay can only be checked or tested to ensure that it operates within the parameters specified by the manufacturer. Because numerical relays contain a variety of functions (86, 27, 59, 50/51, etc.), functional testing takes on a new meaning. Providing a stimulus and observing the response of each function within the numerical relay becomes a function test. With this in mind, our commissioning steps are: • Relay tests/checks • Functional tests • In-service readings Today’s relay engineer must make adjustments to testing techniques to accommodate the enhanced capabilities of numerical relays. These adjustments to technique must consider multiple setting groups, custom internal logic schemes, built in logic switches, dynamic setting capability, internal phase compensation, diagnostic screens, communications, security, oscillography, alarms, and more.

Relay Calibration As noted before, calibration of numerical relays is usually not required since there are no adjustments to be made. There are no trim pots, switches, or selectors with which to make settings and adjustments. If the relay does not operate within tolerance there is no way to adjust it, so calibration, as we know it, can not be made. However, each relay should be checked to make sure it is operating correctly. Secondary injection is still used to make the test, and the output contacts still should be monitored.

40 Most numerical relays allow a combination of entering the data from the front panel (human machine interface) or through a serial port using a personal computer (PC). To realize the full potential and capabilities of numerical relays, the user must have computer skills in order to interface the relay with a PC. In some cases, such as programming logic schemes, utilization of a PC is a must. Because there is a single algorithm instead of individual measuring elements, there is no need to repeat testing on every phase or every zone. It should be sufficient to test A-phase zone 1, B-phase zone 2, C-phase zone 3, etc. However, automated relay testing can speed up the process so there is no significant time penalty for testing all phase combinations. Impedance relays respond to dynamic conditions differently than steady state conditions. These relays should be tested accordingly, using dynamic testing. Because the numerical relay has extended capabilities, there are more settings to apply. Most numerical relays are multifunction devices that have several relaying functions built into one box, thus adding to the number of settings for each relay. Relay engineers and technicians should think of these devices as systems rather than individual relays because they often include switches, metering, control, and wiring (in the form of logic schemes). It will become apparent that documentation of the settings will become an important factor in correct commissioning of the numerical relay.

Dynamic Testing For complex schemes such as distance, secondary injection with steady state values does not provide the information needed prior to commissioning. The response of these systems can only be measured with tests that simulate the power system: prefault load, fault condition with transients, and postfault conditions. These test cases can be simulated with software or oscillography files recorded during actual fault conditions and replayed to the relay. Testing the logic one function at a time in these complex schemes would be extremely time consuming and still may not prove the scheme works. There are many timing and coordination issues that can be proven only by testing the scheme exactly as it will be when it is in service. If the test cases are played with the help of an automated test set and the entire line protection panel is connected to the test set, then the entire battery of tests can be made without reconnection. By applying a series of faults that change incrementally, balance points can be confirmed for every fault type and every phase combination. Since the testing is dynamic, it is not necessary to disable elements for testing as with steady state. It is always preferable to do the testing exactly as the scheme will be in service. The performance of the distance elements changes with the source impedance ratio (SIR). By running additional tests with a variety of SIR’s, the performance of the tripping elements can be measured and compared against factory performance expectations. This set of data also gives

Protective Relaying Handbook — Volume 1 baseline information that can be used in future tests to see if the scheme is still performing the same as when first installed. Software is available to generate the more than 100 files needed to run all the cases. These files can then be combined with automated test equipment to run the cases in succession and record the results. This is, by far, the most cost effective method with meaningful results. The emphasis of automated testing is on relay characteristics. Control functions, such as external cutout switches, autoreclosing, and SCADA, must still be checked as part of commissioning. If the user has standard schemes and has the ability to program the automated testing, these control features could be added to the automated test schemes with prompts to turn on or turn off controls as part of the automated testing and reporting. Communications-aided schemes can be very difficult to test under actual system conditions because the real communication signals have channel delay and attenuation characteristics that are difficult to simulate under test. The true proof is to allow the test equipment to communicate via satellite so the fault signal is given to both ends simultaneously. Any other method except for staged fault testing will leave questions about how the communications channels will coordinate with each other.

Disabling Elements For Testing Testing multifunction relays may require that certain elements be disabled to accommodate testing. For example, if a simple 50/51 function has both time and instantaneous elements programmed to the same output contact, it may be necessary to disable the 51 element to get an accurate pickup value on the 50 element. Although not a difficult chore in most relays, it does require changing the relay from the inservice setting to perform the test. The preferred method of testing any circuit or relay is to test it exactly as it will be when it is in service. Making changes to the in-service settings after they are loaded into the relay requires that the setting be changed back. This may be risky because there may be dozens of settings that need to be changed. One alternative is to begin by loading a copy of the inservice settings in the relay and disable elements for testing as the need arises. Rather than trying to reverse all the changes when testing is complete, load the original copy of the in-service settings back in the relay. This will ensure that the relay is returned to its as-found settings. The relay engineer should be ever mindful that disabling elements within the numerical relay is akin to removing conductors connecting traditional relays in order to test one of the relays in the switchboard. Forgetting to replace the conductors once testing is completed would have severe consequences. Failing to remember to return the numerical relay logic to its original setting can have the same consequences.

41

Protective Relaying Handbook — Volume 1 In applications where the same testing scheme will be used repeatedly, it may be more convenient to create a setting group dedicated for testing. The relay set points within this setting group can be the same as the in-service group, with elements programmed to individual output contacts where needed for testing.

Testing Setting Group Change The ability to have several groups of relay settings that can be dynamically activated manually or automatically to meet the needs of the system may be one of the most powerful features of numerical relays. When system conditions change, the relay is alerted and the settings are changed instantly. There is no need to compromise a setting to fit two different system conditions In many applications there may only be a need for one or two setting groups, leaving the remaining groups without settings. If the relay should inadvertently be programmed to switch to an unused setting group, the relay would essentially be out of service. This is another instance where making the “negative” test is very important. Even though settings and schematics show no setting group change, negative tests should be performed to ensure there are no unintended consequences of switching to an unused group. Perhaps the easiest method is to identify any automatic or dynamic functions during the setting and commissioning process. They can then be set and tested to operate when needed and not operate when not needed. It is the lack of a negative test that can lead to trouble on an automatic feature. For example, if a setting group were accidentally programmed to change groups five minutes after the 51 element reaches 70 percent of pickup, it might not be noticed during testing. After installation, once the load reaches this threshold and five minutes has expired, the relay will switch to another setting group with no programmed settings. In some applications there may be no need to utilize setting group changes. However, the setting sheet still should list the setting group change and whatever set points or commands are needed to program it for “no setting group change.” Although the setting group change function is not used, it is important to document it so the setting engineer and the commissioning engineer both have identified the function for inclusion in the testing and commissioning. Leaving it off the setting sheet because the function is not needed may mean it will not be checked at all. In this case a setting of zero is important. If any setting groups are not used, copy the in-service group settings to all other unused groups. If the relay inadvertently switches to one of those groups, it will still be in service with operational settings. When more than one setting group is used, copy the normal setting to all of the groups that are not used.

Dynamic or Adaptive Characteristics Dynamic or adaptive characteristics should be handled in a similar manner as setting group changes. Any preprogrammed feature of the relay that can make dynamic changes to the relay while it is in service should be called out on the setting sheet so it can be confirmed during commissioning even though it may not be used. These features may be used to cut out instantaneous or ground elements for relay coordination or change the set point of a protection element. If they are identified on the setting sheet, they can be tested to confirm that they operate as intended and do not operate when not wanted.

Testing Programmable Logic Multifunction relays have, in one box, the equivalent of several single function relays that would be found on the traditional relay panel. The functional schematic of the traditional relay is determined by the wiring from one device to the next. In the numerical relay, the programmable logic takes the place of the wiring. Therefore, programmable logic should be treated the same as switchboard wiring. Logic diagrams should be drawn out and documented on blueprints and included in the construction package or settings file. When functional testing is performed as part of commissioning, testing of the programmable logic should be taken as seriously as functional testing traditional schemes. Programmable logic can be saved and transmitted to the relay electronically, sometimes in the same file as the settings. Saving programmable logic to a file in advance of commissioning is a time saver. However, it should not supplant the need for a hard copy of the logic diagram. The logic diagram is an important document used during commissioning and as a permanent record in the substation drawings for troubleshooting. Figure 5 shows a typical programmable logic scheme for basic overcurrent protection. This is the level of detail required to perform functional tests. Based on this information, the commissioning engineer can begin with the inputs and confirm that input 1 changes as the breaker 52b contact changes state and that the relay correctly identifies the status of the breaker. Input 3 cuts out the ground and negative sequence relays, and so on. Notice that there are enough output contacts to program the 51 element to output 1 and the 50 element to output 2. In this example, the need to disable elements for testing as previously discussed is eliminated. The logic should be tested just as functional testing would be for traditional relays. That means confirming that all inputs, outputs, relay function blocks, logic gates, controls, alarms, and switches perform as intended and do not operate with unintended consequences. This means identifying and performing all positive and negative tests. The sequence of events feature of numerical relays can be used to help sort out the results of logic testing to confirm that the correct elements are asserted, logic has functioned correctly, and timing is correct.

42

Protective Relaying Handbook — Volume 1

tripping duty and reporting via a mechanical target. The seal-in unit will stay picked up as long as trip current is flowing to the trip coil. The output contacts of a numerical relay are usually individually sealed relays rated for 30 ampere tripping duty. However, they will break less than one ampere and will be damaged if opened while trip current is flowing. The output contacts are initiated by the internal trip logic of the relay and, therefore, are independent of trip current. To avoid damaging output contacts used for trip and close duty, a hold-up circuit is typically provided that will allow output contacts to remain Figure 5 — Typical programmable logic scheme closed for 10 to12 cycles regardless of what the logic is doing. Once a trip or close Once a numerical logic scheme has been completely has been initiated, the contact should remain closed long checked out, it is not necessary to repeat the internal logic enough to complete the breaker operation. This type of settesting every time that scheme is used. The scheme is reting is easily overlooked and may not be discovered until the peated verbatim electronically. However, inputs, outputs, and relay is in service. This is another item that should be added operational settings must be checked every time. to the setting sheet document so that it can be properly programmed and checked during commissioning. Testing External Inputs In some cases, the targets of numerical relays have proMost numerical relays use optical isolators (optos) to grammable features such as report last target, report all condition the input circuits as shown in Figure 5. These targets, report initial fault targets, ignore certain targets, optos have some dc voltage that defines their threshold of etc. Electromechanical targets are cumulative. That is, if operation. Typically, this threshold is somewhat higher than multiple system faults occur since the relay targets were last half the battery voltage but below the minimum expected reset, there is no way to determine which targets went with dc bus voltage. which fault. Numerical relays, on the other hand, normally If the plant or substation battery system is ungrounded report only the targets for the most recent fault. Previwith typical battery ground monitoring systems applied, ous target data can be retrieved from event data. Because tests should be performed to confirm that the opto would there may be settings or logic associated with targets, this not operate with a full positive or negative battery ground information should also become a part of the setting and between the opto and an external field contact. On a 130 commissioning procedure. volt dc system, full battery ground between the external field contact and the opto would result in half battery voltage of Making Changes to Existing Settings 65 volts dc to the opto. After the input is tested for proper If a setting change is implemented after the relay is placed operation at normal battery voltage, the test should be rein service, how much testing should be done? First, the field peated at half battery voltage to confirm the opto will not engineer should be armed with the existing settings and operate. It should be noted that some relays have internal the new settings in both electronic format and hard copy jumpers used to set the opto threshold. This, too, should be printout. Th is is useful in the event questions arise about the taken into consideration. as-found settings and also provides a means of returning to the old settings if problems are encountered installing Testing Targets And Output Contacts the new ones. Traditional electromechanical relays commonly use trip and seal-in units in conjunction with the main relay contacts. The main contacts are normally not rated for tripping duty, so the combination trip and seal-in provide

Protective Relaying Handbook — Volume 1 After downloading the existing settings from the relay and comparing them to the original commissioning as-left settings, any discrepancies identified should be documented. A misoperation caused by an initial wrong setting may well be the purpose for implementing the setting changes. If the two existing setting files match, the new settings can be applied to the relay. Although settings are typically applied electronically through a Com. Port, some relay manufacturers offer provisions for settings application via a front panel keypad. Regardless of which method is used, all items that have been changed should be tested. If the change is a relay set point, then secondary injection testing is required. If the change is in the relay’s programmable logic, then a functional test should be performed. If the instrument transformer inputs have been disturbed, then in-service tests should be done. In Volume 2 of the Protective Relaying Handbook series, part two will cover in-service testing, documentation and using the numerical relay as a commissioning tool. Jim Closson received his BS from Southern Illinois University at Carbondale, IL, and an MBA from the University of Laverne, Laverne, CA. Prior to joining Basler Electric as a protection and control product manager, he served as the company’s regional application engineer. He has also held managerial and sales positions with Electro-Test, Inc. and ABB. He has taught courses on electrical power systems safety, ground fault applications and testing, and power system maintenance. Mr. Closson is a senior member of the IEEE and serves on the Power Distribution Subcommittee for the Pulp and Paper Industry Committee and on the Transportation Subcommittee for the Petrochemical Industry Committee. Mike Young received an MBA from Rollins College and a BSET from Purdue University. He worked for Wisconsin Electric Power Company as a relay engineer and for Florida Power Corporation as a field relay supervisor for 21 years. He has authored and presented numerous papers on protective relaying at technical conferences across the United States. He is currently Principal Application Engineer for Basler Electric, a member of the IEEE, and is involved in several working groups of the IAS.

43

44

Protective Relaying Handbook — Volume 1

Steady State vs. Dynamic Testing NETA World, Summer 2001 Issue by Steven Stade Universal Power Products, LLC

The introduction of microprocessor-based relays has created the need to examine existing test practices that have been adopted over the years for electromechanical relays. Although protection philosophies have not changed, the achievement of those philosophies is accomplished in different methods between the two types of relays. When building a protection scheme utilizing electromechanical type relays, a physical layout can be designed and tested on how the process will eventually occur. For example, coils are energized to create an electrical force to close a mechanical contact. This will operate other devices in a similar manner that will eventually accomplish the desired task. With this type of arrangement, each step has to be proven to give confidence of correct operation. The major device to be proven is the relay itself. Electromechanical relays, by nature, are devices that use electrical and mechanical energy to operate. If any part of that operation fails, the unit will not function correctly. To prove the correct operation of the device as intended, currents and/or voltages must be applied to the relay to simulate actual load conditions. Instructions from relay manufacturers and past experiences generally have established testing practices utilizing a steady-state testing environment. The use of microprocessor-based relays requires a protection scheme to leave the physical layout environment that has typically been used and enter a digital world. When viewing a schematic of the design, all that may be seen is a trip contact that will operate a downstream device. Viewing the schematic will not give any details of what creates the trip contact to operate. More information is necessary. This idea carries over into the testing environment. Traditional testing methods may not give all the information needed to prove all the details are in place to produce the desired result. That is because a great deal of the process is being handled with a microprocessor chip, based on settings typically contrived by an engineer and entered by field personnel.

To prove the correct operation of the microprocessorbased relay, three basic items need to be proven: accuracy of the CT and/or PT inputs of the relay, operation of the output contacts, and the correct entry and utilization of settings stored in memory by the microprocessor chip. In short, it is now required to work in a digital world to determine if input received from a CT and/or PT can produce a desired response to an output contact. To accomplish this task, the use of a dynamic-state environment may take precedence over the standard steady-state environment. It is no longer a requirement to raise operating values in a manner to verify mechanical relay operations. It is more important to initiate faults that represent true system conditions to verify the relay will respond as desired. Inputs, outputs, and programming can be proven with correct operation of the relay. Settings are engineered to produce an output for a particular condition. In the microprocessor-based relay, many settings entered in the relay must be coordinated to work with each other in order to create the desired output. To test the pickup of a particular element may prove the correct operation of that element but in many cases will not prove it will function as expected in a system fault. A fault needs to be placed on the relay to prove the element actually will produce a desired output. This can be performed quickly and efficiently by the use of a dynamic state test. It applies the fault condition and waits on relay response. Steady-state testing can be attempted to perform this function, but it will not always give the response expected. The relay may recognize the conditions placed on the relay as abnormal and actually cause the relay to not operate. Dynamic-state testing can be designed to prove the desired accuracy of the relay at the same time the relay is tested to operate for particular fault conditions. Assume that a five percent tolerance is allowed for an element. Apply a fault at a value five percent under the expected pickup. Follow this

Protective Relaying Handbook — Volume 1 with a fault at five percent above expected pickup. If the element does not operate during the first test and operates during the second test, it is proven that the element does operate as expected within the tolerance allowed. One major issue involved in the testing of microprocessor-type relays is the amount of elements that will create a trip condition to the relay output. It becomes an issue to determine the best method of how to isolate and prove particular elements. Should elements be temporarily disabled and enabled later, or should elements be directed to an isolated output contact? In making such a judgment, how do you insure the integrity of the test? These questions can bring about the consideration of a dynamic-state test. With some understanding of the protection environment, dynamic tests can be assembled to isolate elements based on the fault conditions applied to the relay. Zones of protection, zone timers, ground elements, etc. can all be isolated for testing, and the integrity of the testing procedure is protected. Isolating contacts by use of a fault condition must hold true. If a window of operation is not available for an element to operate, what is the use of that element? Dynamic-state testing is a process that applies currents and/or voltages to a relay in states to simulate actual fault conditions. The most basic dynamic test is a two-stage dynamic test with use of a prefault and fault condition. It begins by applying currents and/or voltages in a state that simulates the system in a normal operating state. After a predetermined time, all channels will be revised to apply conditions to the relay that simulates a fault on the system. Outputs from the relay will be monitored or timed to verify expected operation. A three-stage dynamic test utilizes prefault, fault, and postfault conditions. It functions in the same manner as a two-stage dynamic test, only after the prefault and postfault condition, the voltage and/or current outputs are revised to a third state after the output condition occurs. The third state should represent the condition of the system after operation of the output contact. What is monitored is the reaction of the relay. Multistage dynamic tests can be used to simulate reclosing situations or to represent special situations that may be need to be applied to the relay. Dynamic-state testing can be performed effectively with a combination of today’s testing equipment and available software. A variety of test set manufacturers assemble testing equipment utilizing three-phase voltages and currents. Many of these test sets are also internally designed to perform dynamic state test conditions. Use of three-phase currents and voltages becomes important as many microprocessor relays require such inputs to recognize the test condition as an actual fault. Although test sets are built to emulate three-phase power systems and perform dynamic test conditions, it can be a time consuming process to set up conditions manually. To increase efficiency and productivity, the utilization of software becomes an important issue. Software can be used to store fault data, apply the conditions to the test set, and store results in a database. Depending on the software itself,

45 the application may have measures available to automatically calculate fault conditions based on the settings of the relay. Test set manufacturers generally build software that is compatible to their equipment. Available to the market is also a company that specializes in software that communicates with a variety of test sets available from various manufacturers. Programs have become very involved with companies that have adopted dynamic-state testing processes. The process involves the engineering department and field department working together as a team to produce effective results. Many engineering departments have incorporated a fault simulation program based on their electrical system. These programs are used to help configure relay settings based on actual system conditions that should be monitored by the relay. With the ability of the fault simulation program to calculate fault conditions, the conversion to secondary values can configure any dynamic-state test desired. While applying the dynamic state test conditions, incorrect operations have located misapplication of the relay settings. As a result, the company has been able to create more solidity in relay setting practices. Both steady-state testing and dynamic-state testing have a place in the testing environment. As the world of protective relaying is evolving, testing practices have to evolve with it. As with any other new process that is introduced, there is a learning curve involved. But adopting new processes can be beneficial and cost effective as an end result. Steven Stade is a protection maintenance specialist for Universal Power Products, Tulsa, Oklahoma. He has over 15 years experience in relay testing and maintenance. Formerly with Central South West Services now American Electric Power, he managed enterprise-wide testing practices and implemented various software applications to the changing maintenance environment.

46

Protective Relaying Handbook — Volume 1

Dynamic State and Other Advanced Testing Methods for Protection Relays Address Changing Industry Needs NETA World, Summer 2001 Issue by Kenneth Tang Manta Test Systems

Protective relay testing has become increasingly complex recently, reflecting continued changes in the industry and protective relay technology. Many technicians and engineers have been confused not only about how to test and how often to test but what to test. In particular, testing needs for traditional electromechanical devices versus new intelligent and adaptive processor-based protective relays are sometimes similar and sometimes different depending upon whom you ask. There is general agreement that the newer processor-based or digital relays require dynamic testing to verify some functions/elements. However, in actual practice the type of testing performed even on the same model of relay varies significantly from company to company. In order to try to clear up this confusion, let’s first look at the types of dynamic testing available.

uses sinusoidal waveforms with instantaneous changes in phase, amplitude, or frequency (or combinations of all three) possibly combined with linear ramps of these parameters. This has been an accepted method for testing distance relays for quite some time, but it has become necessary even for distribution relays with such features as cold-load pickup blocking and block on expiration of voltage memory. Some elements such as power swing blocking and frequency decay respond only to changes in the input quantities not to their steady state values.

Types of Dynamic Testing Dynamic on/off This is the most elementary type of dynamic testing, started by wiring a switch in series with the test voltage and/or current sources. This type allowed one to test the relay response to an off-to-on or on-to-off step change in voltage/current. This has been done since long ago to measure, for example, the operating time of an overcurrent or overvoltage relay. This can also be used to check the response of an impedance relay to a zero voltage forward or reverse fault, or a switch onto fault condition.

Dynamic state This type of testing recognizes that power system voltages and currents undergo sudden and gradual changes in phase, amplitude, and frequency. Therefore, dynamic-state testing

Figure 1

Dynamic with controlled FIA and dc offset This type of testing recognizes that real-world fault currents have an exponentially decaying dc offset. The rate of decay depends on the system L/R ratio. The magnitude of the offset depends on both the prefault and fault current magnitude and angle and the fault incidence angle. The fault incidence (or inception) angle (FIA) is the electrical angle of the waveform at which the fault occurs. (See Figure 3).

47

Protective Relaying Handbook — Volume 1

Figure 2

Figure 4

Figure 3

The response of modern relays with digital filtering can vary depending on these waveform characteristics. Some distance relays operate in one cycle or less and can only be tested with controlled FIA and properly simulated dc offset. There have been cases of design deficiencies being discovered in digital relays by applying this type of testing.

Not only the operate characteristic but the operate time depends on the source impedance and actual fault conditions. Figure 5 shows that the operate time of a typical distance relay is not constant but increases sharply near the end of its reach and will shift up or down depending on the source-to-line impedance ratio (SIR). Testing the operate time of a distance relay using an arbitrary choice of fault voltage and fault current gives results which can not be related to the expected performance.

Dynamic system model-based testing This type of dynamic testing is one step closer to simulating real-world conditions. It recognizes that relays are applied in specific power system configurations, all of which have their own particular characteristics. It has been shown that modern protection relays (line protection, in particular) have both static and dynamic characteristics. The dynamic characteristic is the one that actually performs under real fault conditions. All properly designed distance relays today (electromechanical or digital) use some form of cross polarization and memory polarization in order to improve security for zero voltage faults and reverse bus faults and to improve fault resistance coverage. These relays have a dynamic characteristic dependent on source impedance, load flow, fault resistance, relay design, and other factors. A typical characteristic is shown in Figure 4. In order to measure this, fault voltage and current waveforms must be calculated using a system model that takes into account the sequence impedances of the source and line and the source voltage(s). In addition, the fault quantities must be stepped from their prefault values to their fault values in order to test dynamic characteristics which are memory based.

Figure 5

It is not practical to do this testing manually; therefore, automatic testing programs are available from some test equipment manufacturers for this purpose. Figure 6 shows an example. Effects of load flow, mutual coupling, and fault resistance on the relay operation can also be evaluated with these programs. Directional relays can also have complex characteristics depending on the type of polarization they employ, the system configuration, and the type of fault conditions. These characteristics can also be effectively tested using these same

48

Protective Relaying Handbook — Volume 1

Figure 6

automatic test programs. This method of testing can be used as a unified approach across many different types of relays because it presents relays with fault scenarios regardless of their design or operating principle.

Fault playback Fault playback testing recognizes that real world fault waveforms contain dc offsets, harmonics, transients, noise, etc. Fault playback testing replays actual fault waveform recordings (or computer-simulated waveforms) into a relay and measures its response. Certain elements, such as high-impedance ground fault detection elements, are best tested in this way because these waveforms are not normal sinusoids. The device is presented with a very close approximation to the real world waveforms, limited by the accuracy and resolution of the original recording and its sample rate. Some people may not consider this as dynamic testing but classify this as transient testing, apart from dynamic testing. However, the point is that it is one step closer to simulating the real world.

End-to-end testing End-to-end testing is the testing of transmission line protection systems using GPS time synchronized simultaneous injection of test waveforms at two (or more) remote terminals that are connected together by a communications

channel. The waveforms may be sinusoidal waveforms as described above in dynamic state testing or may be transient waveforms described above in fault playback testing. In end-to-end testing, we can perform testing which was not possible by testing a single relay or terminal, such as simulating race conditions such as those that occur in current reversals. The primary benefit of this type of testing has been the ability to simultaneously test protection settings, logic, auxiliary relays, communication equipment, and the communications channel for a complete line protection.

Closed loop testing In this type of testing, the stimulus waveforms change dynamically based on the response of the device under test. A simple case of this may be testing a multishot autoreclosing relay. A more complex case may be testing the clearing sequence for a multiterminal line where the fault voltages and currents change after each breaker opens in succession. Not only the waveform shape but also its duration are regarded as important in simulating real-world fault scenarios. This type of testing is typically done with large simulators but some types of portable test equipment have some of these capabilities built in.

49

Protective Relaying Handbook — Volume 1

Figure 7

What Type of Dynamic Testing Should Be Done? In deciding the type of dynamic testing that should be done, the most important consideration is: What is the test trying to prove? a) Does the specific device meet its published specifications?

b) Is a specific installation of the device set properly and functioning as expected? c) Does the design meet the general needs of the application?

d) Has the device performance changed or degraded? e) Does the device work properly as part of an entire system? f ) Does the design meet the specific needs of an application?

Traditionally, users have been focused on purposes a, b, and d in the above, but now the trend is toward focusing on purposes b, e, and f. Why? Manufacturers of relays are performing intensive type-testing to verify their designs. They are providing built-in self-test features. Digital technology has led to higher overall relay reliability. As a result, many companies are moving toward simplifying commissioning and periodic maintenance tests. Once the basic protection software algorithms have been verified, they should not need to be reverified unless, of course, there is a software upgrade. The characteristics should not change or degrade with time. Tests should still be done to ensure that the relay settings and logic are correct and that the relay is wired correctly

in the system. Some basic functional tests with the relay installed in the system, monitoring the actual trip (or other) output and with the in-service settings should suffice. We have seen people test with “test settings” meant for the convenience of some test program which have no relationship to the in-service settings. This may prove a or c in the list but does nothing for the other objectives. On the other hand, the complexity of newer devices means that there are more things to go wrong and more combinations of conditions that may cause different or unexpected behavior. Digital relays can embody extreme complexity by employing digital memory, counters, timers, and software algorithms. These can base their operation on multiple inputs, quantities calculated from these inputs, and a memory of the past values of these inputs and calculated quantities. An example is a motor protection relay that employs thermal memory and adjusts the overload protection curve based on the measured heat accumulated in the motor. This has led to more thorough and complicated relay testing by users to evaluate devices for their particular application. There is a general trend toward the dynamic system model based testing, end-to-end testing, and fault playback test methods. This is for a number of reasons: • These methods check the performance of the protection system as a whole, not individual elements. • Many functions can be tested simultaneously and faster.

• These methods present the relay system with conditions that it will likely encounter in its application.

• These tests have a wider coverage, including, in some cases, objectives a, b, d, e, and f in the list and generally do not repeat tests performed by the manufacturer. • Certain key aspects can only be tested using these methods

This is not to say that other types of testing will diminish; they will continue to have a useful place. An understanding of the device and the system being tested is important to help one choose the applicable method. Consideration should also be given to what test hardware and software is required, how much time it will take, and whether one has sufficient training and knowledge to not only set up the test cases but to evaluate the correctness of the results.

50 Summary Many dynamic test methods are at our disposal today. Choosing the right method requires an understanding of the methods themselves, an understanding of the design of the devices being tested and their specific application, and, most importantly, clarity about the purpose of the testing to be performed.

References MTS-1700 Universal Protective Relay Test System Operation Manual, Manta Test Systems, 2000 Application Note: Practical Mho Distance Relay Testing with the MTS-1710, Manta Test Systems, 1995

Testing Impedance Characteristics of Transmission Line Relays, Elmo Price, Georgia Tech Protective Relay Conference, May 1999. Testing Modern Protective Relays, R.J. Martilla, Canadian Guide to Protection and Control, 1999

Relay Performance Testing, IEEE Power System Relaying Committee, 1996 Kenneth Tang is the Technical Support Manager at Manta Test Systems, a supplier of advanced electronic test products and services for electrical power system protection, control, and measurement.

Protective Relaying Handbook — Volume 1

51

Protective Relaying Handbook — Volume 1

Acceptance Testing a Synch Circuit NETA World, Summer 2001 Issue by Steven C. Reed, P.E. Electric Power Systems

On any new electrical installation startup, acceptance testing is an important step to ensure the correct operation of protective equipment and the safety of personnel. Synchronization circuits are some of the most critical parts of any electrical distribution system, especially the first time they are put into use. If for any reason there is an error in the engineering, application, or wiring there is a great concern for the safety of personnel and possible equipment damage. There can be no short cuts in the startup of a synchronization circuit.

Basic Description A synchronization relay (25) is used to verify that the voltages on either side of a breaker are within appropriate voltage magnitude and phase relationship prior to initiating a breaker close. The voltage comparison is made between the bus and the line. See Figure 1 for a basic example. The relay close circuit will operate after an enable signal is received from each of the following circuits: voltage difference, phase difference, voltage monitor, and time delay. The voltage difference (delta voltage) setting compares the magnitude of the bus voltage to the line voltage. If the delta voltage is less than the set limit, it enables a close signal for the voltage difference. If the magnitude of the voltage difference is exceeded between the line and the bus, the breaker close signal will be blocked. The phase difference setting measures the phase angle between the line and bus voltage. If the measured angle is less than the window setting, the phase difference setting will be enabled. If the phase-angle reading between the bus and line voltage is greater than plus or minus the window setting, the close circuit will be blocked. The voltage monitor setting allows programmable options that may best suit the system operating conditions, such as allowing breaker closure for live line, dead line, live bus, and dead bus.

Live line (LL) – enabled when the line voltage is greater than the setting. Dead line (DL) – enabled when the line voltage is less than the setting.

Live bus (LB) – enabled when the bus voltage is greater than the setting. Dead bus (DB) – enabled when the bus voltage is less than the setting.

The time window setting is an adjustable time delay that allows the relay close circuit to be enabled after all previous conditions have been met.

Pre-Energization Tests Before any work is performed the field technician must review the manufacturer’s literature to gain a full understanding of the design and capabilities of the relay. The engineering drawings shall be compared to the manufacturer’s literature as a way of confirming the correct use of the relay. The coordination study settings shall also be reviewed to verify the correct overall use and design (for example, LL, DL, LB and DB). After the relay operational review has been completed, each component of the synchronization circuit should be tested on the line side and bus side. 1. Insulation and ratio test of any PT or CCVT.

2. Insulation test and verification of secondary wiring.

3. Verify proper secondary grounding at the PT or CCVT.

4. Verify proper primary and secondary phasing of the line and bus PTs.

5. Calibrate synchronization relay in accordance with settings.

52

Protective Relaying Handbook — Volume 1

Energization Tests Although any problems should have been detected in the pre-energization tests, the synch system is not ready to be utilized. Assuming a generator is on the line side of the synchronization system, it is not worth risking damage to the generator should a mistake be made during the verification of the synchronization scheme. A back-up plan should be in place for verification that the synchronization scheme is correct. As an example of the various methods for verification, we will assume the following: a generator is on the line side of the synch system, the synch system is set up across a generator breaker, and a step-up transformer feeds a high-voltage transmission breaker. The synchronization relay PTs are connected on the line and bus side of the generator breaker. There are three acceptable procedures allowing for correct verification in this case: isolation from utility grid while the generator runs, a backfeed from utility while the generator is isolated, or the use of two sets of phasing sticks (least desirable). The first method is the isolation from the utility grid. The generator may be started, and the generator breaker closed onto dead bus only! The high-voltage transmission breaker will have already been opened, locked, and tagged out. The open transmission breaker will isolate the generator from the utility system. Most likely three PTs will be located on the line side and load side of the generator breaker. Verify voltage through phase angle measurements at the synch relay. No voltage should exist between similar phases (Line A-Bus A, Line B-Bus B, and Line C-Bus C). Obtain proper voltage readings from line-to-ground and phase-to-phase on the line and bus feeds. Verify that the synch system is operating in accordance with all set parameters. The second method is to backfeed the utility system by isolating the generator. A portion of the generator bus will need to be removed to allow a backfeed from the utility system. Proper clearances and safety standards need to be met prior to the initiation of this test. The high-voltage breaker will be closed, backfeeding the step-up transformer. The generator breaker needs to be closed (temporary adjustment to the breaker-closing scheme), to backfeed the generator bus. This allows both the line and bus PTs to be energized. Most likely three PTs will be located on the line side and load side of the generator breaker. Verify voltage through phase angle measurements at the synch relay. No voltage should exist between similar phases (Line A-Bus A, Line B-Bus B, and Line C-Bus C). Obtain proper voltage readings from line-toground and phase-to-phase on the line and bus feeds. Verify that the synch system is operating in accordance with all set parameters. The third method may need to be utilized if either of the other two energization procedures can not be met. This method is the least desirable due to the number of personnel required and risks of electrical hazards. Two sets of phasing sticks should be used to monitor the line and bus voltage. The line side is fed from the generator. The bus side is fed from the utility. One phase set will monitor voltage on the line side and the bus side of A-phase. The second phase set will monitor the line side and bus side of C-phase.

Figure 1 — Connections for a Typical Application

All necessary industry, site, and customer safety standards should be followed, such as the use of blast suits, since this is considered hot work. Location of the monitoring will be critical for personnel safety but must be determined based on each job and rating of equipment. This procedure will take at least four technicians to complete: two technicians (minimum) with phasing sticks, one (with radio) monitoring both phasing sticks, and one (with radio) monitoring synchronization process. It should be determined that the synch relay will allow closure of the breaker just before the voltage at both phasing sticks is zero. If both phases are in synch, the third phase must also be in synch. The synch relay should never allow closure unless both phasing sticks show zero voltage. During this entire process the generator breaker close scheme should be disabled as a protective measure. After the pre-energization and energization tests have been completed, a new synchronization scheme can be operated with full confidence that it is within proper operating conditions. Steven C. Reed has a BS in electrical engineering from Villanova University, a MBA from the Olin School of Business at Washington University in St. Louis, and has professional engineering licenses in multiple states. Steve has worked at Electric Power Systems for 12 years and served as a field engineer, system protection engineer, and now serves as regional manager. He is a NETA Certified Technician Level III.

53

Protective Relaying Handbook — Volume 1

Partial Differential Relaying NETA World, Summer 2001 Issue by Baldwin Bridger, P.E. Powell Electrical Manufacturing Co.

“Partial differential” relaying is a form of overcurrent relaying frequently used to detect main bus overcurrent faults and to back up feeder overcurrent relaying. The basic circuit is shown in the one-line diagram. Note that this is a double-ended substation with two main breakers and a tie breaker. The partial differential relaying concept can not be used on a straight radial distribution system. True bus differential relaying compares all currents entering and leaving a switchgear bus. Within the limits of the accuracy of the CTs and the relays, true bus differential relaying will detect all faults on the protected bus. Since all currents are taken into account, the relays can be very fast. Bus differential relaying, however, provides no backup to the feeder overcurrent relaying, so additional overcurrent relays are required on main and tie breakers to provide this backup function. Also, high speed bus differential relaying can be quite expensive, and many switchgear users do not feel that it is economically justified. Partial differential relaying sums the currents entering or leaving a switchgear bus through main and tie breakers. If a fault exists on the protected bus, the currents will add in the relays, but if fault current is flowing through the bus to a fault on another bus, the currents will subtract and the relays will not respond. If the fault is on a feeder, the partial differential relays will act as backup to the feeder overcurrent relays. Similar protection can be obtained by using separate overcurrent relays on each main and tie circuit breaker. However, proper coordination of the overcurrent protection requires that the tie breaker relays coordinate with the feeder relays and that the main breaker relays coordinate with the tie breaker relays for a total of three steps of relaying at this bus. Using the partial differential circuit, however, elimi-

nates one step of coordination, since the same relays serve both the main and the tie breakers without compromising coordination. This reduces the time delay required for the main breaker relays and improves the chances of getting good coordination with upstream relays which are often on the utility system serving the substation. This improved coordination is the principal benefit of partial differential relaying. Reprinted with permission of Powell Electrical Manufacturing Co.

Baldwin Bridger, PE, is recently retired Technical Director of Powell Electrical Manufacturing Co., Houston, Texas. He has worked as an engineer and engineer manager in the design of low- and medium-voltage switchgear since 1950, first at GE and since 1973 at Powell. He is a Fellow of IEEE and a past president of the IEEE Industry Applications Society

54

Protective Relaying Handbook — Volume 1

Modern Relays and Software Provide Valuable Tools for Analysis NETA World, Fall 2001 Issue by Scott Cooper Beckwith Electric Co.

In the past, the only indications of a trip were an alarm, a target flag, and a tripped breaker or lock-out relay. There was no data to aid in determining what happened, how it happened, or the extent of the damage. This often resulted in days of unnecessary testing and inspections or, worse, placing faulted equipment back on line, which can be a safety hazard or can cause further equipment damage. This article discusses present-day fault recording and failure analysis using modern digital relay technology. Unlike their electromechanical and static predecessors, digital relays provide a number of valuable tools to aid in determining exactly what happened and the extent of damage. To provide appropriate indication for the operators, the relay front panel prominently displays some basic information: the relay’s operational status, current trip state, and the function tripped during the most recent event. More detailed target event information may be retrieved via the keypad interface as well as remotely via PC or PLC systems. Unlike traditional alarm panels which are wired directly to relays, these systems may use RS232, RS485, or modem communication connections and a variety of communication protocols to interrogate the relay. Currently, detailed trip data may be accessed, processed, and appropriately displayed to operators, technicians, engineers, and management in different locations. The troubleshooting process normally starts with the sequence of events record. Each target is stored in order and is identified by a time stamp which corresponds to the relay clock time of the first trip. In installations where multiple devices are present, all time stamps may be automatically synchronized using an IRIG-B network. To prevent data loss, event data is stored in nonvolatile memory in case of input power interruption to the relay. Typically, relays can automatically store 24 or more separate trip events, depending upon the application. Each trip event may contain a number of individual targets that are identified as being either picked up or tripped. At the inception of each event, digital relays also log I/O status and metering quantities.

Target History:

Screen from Beckwith Electric’s IPScom® Communication Software showing a relay’s target history during testing. The left panel shows the events available and the right panel displays the I/O status, line side currents, and targets in the selected event.

In addition to the target data, most digital relays also incorporate some form of waveform recorder or oscillograph record. The oscillograph record can be used to identify the sequence of events, aids in verifying the validity of the relay’s operation, and speeds troubleshooting by helping to identify the faulted phase or component. Oscillographs can also provide necessary data for engineers adjusting a relay set point to overcome a normal transient. In the event of an actual fault, oscillographs provide insight on how far a parameter was out of specification and for how long. This is often crucial data for determining what testing or inspections are needed after an incident. Of course, if the worst should happen, the data may be used as evidence in court. During normal relay operation, the relay continuously saves waveform and I/O data to RAM. After an oscillograph event is triggered, that portion of memory is reserved. The

55

Protective Relaying Handbook — Volume 1 resultant waveform data may then be downloaded to a PC and saved as a file for analysis. In the event that redundant or overlapping relays are used, both oscillograph recorders should be set to trigger if either relay trips. Inevitably, in the event of an actual fault, one relay will trip first and open the breaker, while the other relay may not have the opportunity to time out even if the same settings are used on both. For generator protection, this is especially true with slow developing faults or abnormal operating conditions such as loss of field (40) or volts per hertz (24). A second oscillograph can also help confirm a suspected incorrect trip or relay failure.

59N:

Screen from IPSplot® Oscillograph Analysis program showing an actual generator stator ground fault. After the trip, meggering the stator could not verify the problem. However, from the trace, the operators could see arcing and breakdown, so more testing was ordered. The stator subsequently failed high potential test. Having an accurate oscillograph record probably prevented this trip from being dismissed as a relay problem, and serious machine damage was averted.

Inadvertent Energization:

Screen from Beckwith Electric’s IPSplot® Oscillograph Analysis program showing an actual generator inadvertent energization. From the trace, we verified a secondary current equal to 24A for a period of 4 cycles before the relay and breaker opened the circuit. Input 1 shows the indicated breaker position, output 2 shows the Beckwith relay tripping. The cause of this event was operator error.

To analyze the event waveforms, one must obtain the relay settings, target information, and review oscillographs from the event. Next, study the target information to determine what function tripped first. Finally, use the oscillograph record to calculate the associated values at the time of the trip. The waveform can also be used to verify that the relay is operating correctly. Relay or input circuit problems can be indicated by waveforms that are inconsistent with actual generator operation. For example, was a corresponding neutral voltage or current recorded during a 51V inverse time overcurrent operation? Once the validity of an operation is established, the waveforms can be used for system troubleshooting. For example, identification of a miswired CT circuit is easy with an oscillograph of the differential trip.

Since the oscillograph records both pre- and posttrigger data, the record can also be used to verify proper trip circuit operation. This is easily accomplished by comparing the relay trip time mark, the breaker open indication, and the cessation of currents. This process verifies that the breaker indicated open and did, in fact, open within the anticipated operating time. Oscillograhs can also be used to check for breaker problems like unequal pole operation. After analysis, resultant oscillograph files may be converted to the Comtrade format (see Transformer Inrush figure) for playback using modern test equipment. Comtrade is a common waveform file format used by most test equipment manufacturers. This capability provides a new and exciting means to evaluate equipment based on actual fault conditions. In the past, the scope of normal relay testing did not include the harmonics, dc offsets, and CT saturation that may be present in actual operation. This playback capability can be used to verify that a relay operates correctly for an actual system fault. It can also be used to verify a relay does not improperly trip during plant transients (see Relay Comparison graph).

56

Protective Relaying Handbook — Volume 1 Modern digital relay technology provides valuable tools for determining the cause and extent of trip events and faults. By using fault recording and failure analysis functions available in digital protective relays, users can save time and money by eliminating unnecessary testing and inspections and avoiding equipment damage. Scott Cooper, Field Service Engineer, joined Beckwith Electric Co. in 1997. His responsibilities include training, commissioning, and troubleshooting protective relays for customers. He is also instrumental in testing new relay products and custom-engineered systems. Scott was previously an electronics technician at Beckwith for two years testing protective relays and conducting failure analysis and individual component evaluations. He is a member of IEEE and served in the US Navy for six years in the nuclear reactor controls division. He has served as a senior technician with Seapower Engineering.

Transformer Inrush:

Screen from Omicron’s Transview Analysis program showing a “black start.” The left panel shows the distorted waveforms are the result of energizing a bank of load transformers. The right panel shows the harmonics present. The 100 percent dc offset and harmonics were causing the installed generator relay’s differential element to trip.

Relay Comparison:

Excel Spreadsheet comparing the tripping performance of two digital relays from different relay manufacturers. The above transient was captured and played through the two relays using an Omicron 256-6 test set. From the graph, the differential element of relay B can be set to more sensitive and is much more tolerant to the black start transient than the currently installed relay A. Accurately identifying the problem and proving the solution quickly solved this customer’s problem.

57

Protective Relaying Handbook — Volume 1

Understanding and Analyzing Event Report Information NETA World, Fall 2001 Issue David Costello Schweitzer Engineering Laboratories, Inc.

When faults or other system events occur, protective relays record sampled analog currents and voltages, the status of optoisolated inputs and output contacts, the state of all relay elements and programmable logic, and the relay settings. The result is an event report, a stored record of what the relay found, and how and why it responded. Readily available information, product instruction manuals, and assistance from analytic software equips the user with the necessary tools to determine if the response of the relay and the protection system was correct for the given system conditions. Each time the power system faults and relays capture data, the results are ready-made test reports. By analyzing the actual relay and system performance, utilities are saving money by extending or eliminating traditional routine tests. Regulatory agencies require the installation of disturbance monitoring equipment and postfault event analysis. Relays with event reporting help meet these requirements. Information recorded in relay event reports are valuable for testing, measuring performance, analyzing problems, and identifying deficiencies prior to causing a misoperation. The ability to quickly and accurately analyze event data is useful.

This article supports efforts through a real-world example which demonstrates the process of changing raw data into useful information.

How to Analyze an Event Report 1. Understand the expected or desired operation 2. Collect event reports and other information

3. Look for possible exceptions and/or unexpected elements 4. Compare actual operation to expectations 5. Utilize manufacturers’ data and software 6. Develop and test solutions

This example is a step-by-step tutorial on analyzing an event report, valuable lessons, and problem resolution. Before analyzing event report details, begin with a basic understanding of what took place, or what should have. This process generally involves reviewing the relay settings and logic, obtaining the relay history report, and gathering any additional information that may be helpful (for example, known fault location, targets from other 1st Event Report: relays, breaker operations, SCADA and CARNALL CCT.# 2522 SN# 96143025 Date: 8/25/99 Time: 11:54:43.479 personnel records). The event report is Event : AB T Location: 0.12 Shot: 0 Targets: INSTABQ used to verify that the actual operation Currents (A pri), ABCQN: 3766 3551 239 6124 20 matches the expected operation. Historical information was down2nd Event Report: loaded from a distribution relay that had CARNALL CCT.# 2522 SN# 96143025 Date: 8/25/99 Time: 11:54:44.083 to be closed by SCADA after tripping Event : CG T Location: 0.21 Shot: 0 Targets: INSTCQN to lockout. The relay controls a recloser Currents (A pri), ABCQN: 0 3 3932 3930 3931 which is mounted on a steel stand within the substation and powered from the 3rd Event Report: substation dc battery. Daily routine CARNALL CCT.# 2522 SN# 96143025 Date: 8/25/99 Time: 12:09:41.758 requires utility employees to investigate Event : ABC Location: 5.67 Shot: 2 Targets : all out of the ordinary events, including Currents (A pri), ABCQN: 443 654 403 478 105 failures to automatically reclose and lockout events.

58

Protective Relaying Handbook — Volume 1 The pickup of the 50H element is (30 amperes secondary) • (CTR=120:1), or 3600 ampere primary. We should, therefore, expect the initial INST A B trip target for a 3766 ampere fault. The next expected sequence for this relay is to open the recloser, time on =N the first reclosing open interval, then automatically reclose. The first reclose =Y attempt should be after an open delay =0 of 900 cycles, or 15 seconds (79OI1 set=0 ting). However, the second event is an instantaneous C-to-G trip only 0.604 second after the first event. What would cause a fault to occur during a recloser open period during timing our first reclose attempt? The analytic software plots of the first (Figure 1) and second (Figure 3) event reports confirm our suspicion of a recloser failure and flashover inside the recloser tank. In Figure 1, the initial A- to B-phase fault is evident. The first digital element to assert is the 51P time-overcurrent pickup, the most sensitively set element. This triggers the event report as expected by the ER = 51P setting. To determine which element caused the trip, identify the point in time where the trip asserts (OUT T) and look for any other =RE element transitions at the same point. The pickup of the instantaneous phase overcurrent element, 50HP, asserts at the same instant the trip output asserts, while the 51P element is shown picked up but still timing to trip. The 79 reclosing element prepares to time to a reclose by changing from the reset state to the cycle state when the relay trips. IN6, programmed to monitor a 52a auxiliary contact, comes open two cycles after the trip indicating the recloser has opened. After adjusting the scaling on the C-phase current channel in the analytic assistant software (Figure 1), we can see that the C-phase interrupter did not open fully as current continues to flow. The trip coil monitor, IN3 = TCM, is an optoisolated input wired as a voltage divider to monitor the health of the trip coil (refer to Figure 2). When the recloser is closed and the trip output contact is not asserted, the TCM input allows a few milliamperes of current to flow through the trip coil. The voltage drop is across the relay TCM input because the input has a much higher impedance than the trip coil (roughly 1000 times greater). In the first five cycles of Figure 1, the TCM is asserted, indicating the trip circuit was intact. At the time of trip, the TCM input deasserts, initially because of the closed trip contact and then because of the open 52a auxiliary in the trip circuit.

Partial Display of As Set Settings for CARNALL CCT.# 2522 SN# 96143025 CTR =120.00 79OI1 =900 79OI2 =2700 79RST =600 M79SH =11011 50C =99.99 50NL =99.00 51NP =12.00 51NTD =15.00 51NC =3 51NRS 50L =99.99 50H =30.00 51P =5.01 51TD =2.50 51C =4 51RS 52APU =0 52ADO =0 TSPU =0 TSDO TKPU =0 TKDO =0 TZPU =0 TZDO S(123) = A(12) = B(12) = E(34) = F(34) = K(1234) = L(1234) = A1(1234) = A2(1234) = V(56) = W(56) = X(56) = A3(1346) = A4(2346) =TCMA TR(1246) =50H+51T RC(1246) =TF ER(1246) =51P TDUR =5 TFT =30 IN1 =DC IN2 =DT IN3 =TCM IN4 IN5 = IN6 =52A

Figure 1— C-Phase Interrupter Fails to Open

In order to understand normal relay operation, examine the output contact logic and determine what elements in the relay are actually used in this application. In this relay, we notice that only two elements are programmed to cause a trip (TR equation), the nondirectional phase instantaneous 50H element and phase time-overcurrent 51T element.

59

Protective Relaying Handbook — Volume 1

Figure 2 — Trip Coil Monitor

In the second event, the failed interrupter flashes over to the recloser tank 0.604 second after the first trip occurred. In Figure 3, you can see the 79 reclosing element immediately goes to lockout. The relay is designed to drive its reclosing element to lockout if a trip occurs before reclosing has been attempted. This prevents reclosing after a flashover across an open pole or internal tank failures such as this. Therefore, the operation of the relay was correct, and the cause of the failure to reclose was a recloser failure condition. The information gathered in the first two events indicates that C-phase carried current for at least 0.721 second (the difference between the trigger times of each report, 0.604 second, plus 7 additional cycles of fault data in event two). The fault current seen for the majority of this time was only around 50 amperes primary. Could a recloser failure element have been used to clear this fault before it developed a more severe 4000 ampere fault?

component magnitudes are calculated. At the end of the first event, the C-phase current is only 0.42 ampere secondary (3IO = Ia + Ib + Ic = 0.412 A, as well). As set, the overcurrent elements used for tripping and those not used for tripping are set much too high to see the 0.412 ampere phase and residual current flowing through the failed interrupter, so the trip failure logic, as set, is ineffective. In this relay, the elements which unlatch the trip output and trip failure timing are the same elements that prevent the reclosing relay from resetting after an automatic reclose. Set a residual overcurrent element 50NL to 0.25 ampere secondary to provide sensitive breaker failure supervision for unbalanced faults. The event reports in the history of the relay are reviewed to insure normal load unbalance is not greater than (0.25 ampere secondary) • (CTR=120:1), or 30 amperes primary. With this setting, our trip failure logic would have detected the unbalanced condition as a result of the stuck C-phase interrupter. Programming an output to close when a trip failure (TF) is detected could provide a trip to a back-up protective device (the transformer differential relay), assert an alarm to the SCADA system to initiate maintenance, and prevent a more intense fault. Supervising the trip failure element with a phase overcurrent relay is more challenging in this relay but can still be done. The maximum prefault load current in the relay’s history of events was 130 amperes primary, or 1.08 amperes secondary. If we set any element other than 50C in this relay below load, our reclosing relay will be prevented from resetting (the trip and trip failure unlatch elements are the same elements used to allow the recloser to reset). The logic and wiring in Figure 4 allow a sensitive 0.5 ampere secondary setting for 50C to be used for phase current supervision of the trip failure logic while not interfering with the reclosing reset logic.

Figure 4 — Recloser Trip Failure Logic for Phase Faults Below Load

Figure 3 — Failed C-Phase Interrupter Flashes to Ground

The recloser failure element as set in this relay is only intended to cancel reclosing. The TF or trip failure bit asserts if none of the overcurrent elements in the relay (with the exception of the 50C element) have dropped out TFT cycles after a relay trip is initiated. If the overcurrent elements drop out, the trip failure element stops timing. Using the analytic assistant software, phase current and symmetrical

We assume that the C-phase interrupter eventually opened because no backup protective device operated and the beginning of the third event (see Figure 5) indicates the C-phase current is zero. The dispatcher instructed a local switchman to report to the substation because the SCADA system indicated the recloser was open and in lockout. Approximately 15 minutes after the initial trip, the third event captures the SCADA close operation. By noticing that IN4 (reclosing enable) is deasserted, we verify that the switchman manually turned automatic reclosing off. By noticing that IN1 (direct close) is asserted, we verify that SCADA was used to remotely close the recloser once. The recloser closed without incident.

60

Protective Relaying Handbook — Volume 1 Figure 5 indicates the cold load inrush and the brief pickup of the 51P time-overcurrent element. A cold load pickup scheme can be enabled through settings such that it is automatically put in service when the recloser is open and locked out for a long period of time. After a successful close, the scheme automatically adjusts to the original settings. When the scheme is active, the relay modifies the pickup of the phase time-overcurrent element to a higher value while keeping the same curve and time dial settings to maintain coordination with upstream devices.

Figure 5 — SCADA Close of a Failed Distribution Recloser

The third event emphasizes the importance of using a manual close delay. In newer recloser controls and substation relays, front panel operator controls are built in so that traditional control switches can be eliminated. For safety, the user may add a settable time delay to the operation of the front panel operator controls. This delay allows an operator to initiate a manual close by pushing the close button and then walking away to a safe distance before the close signal is actually sent by the relay to the recloser or breaker. The associated red close LED flashes as the timer counts down. This safety improvement can be made in older relays such as the one in this example by wiring the manual close switch contact to a programmable relay input and time-delaying the close output with programmable logic as follows. See Figure 6. S(123) TSPU K(1234) TKPU L(1234) V(56) A4(2346)

Figure 7 — Cold Load Pick-Up Scheme Improves Security and Maintains Coordination

To enable the cold load pickup scheme, an element called 52BT that follows the recloser status is used. 52BT is the inverse of 52AT (see Figure 8). With the settings shown, the modified pickup will be in service for 52APU time, a settable value. After the time expires, the pickup is forced back to the original value. If 52ADO exceeds all reclosing relay open interval time settings (79OI1, 79OI2, etc.), the cold load pickup scheme will be disabled through the reclose cycle. The 52AT element drops out, and the 52BT element asserts when 52ADO expires after the recloser goes to lockout. The 52ADO is effectively the loss-of-diversity time delay. The 52APU is effectively the time limit before minimum pick-up is restored following a close operation.

= IN5 IN5 is energized by a momentary manual close switch =0 TSDO = 300 = IN5 =0 TKDO = 315 = ST = KT*!L =V A4 is a time-delayed close output from the relay to the close coil

Figure 6 — Delayed Manual Close Setting Option Improves Safety

61

Protective Relaying Handbook — Volume 1 50L 51P 52APU B(12) C(12) F(34) G(34) X(56) Y(56) TR(1246)

=7.50 =5.00 =30 =51T =50L =52BT =52AT =B*C*F =B*G =X+Y

51TD =2.50 52ADO =3600

51C

=4

This was a B-to-G trip. The reclose operation was successful for the fault. Had a reclose failure 51RS =Y not occurred, an investigation of the first event, since it appeared at first glance to be a normal trip and reclose event, also may not have happened. Further investigation proves that the C-phase interrupter experienced a problem during the initial trip as well. During that event, however, the reclose occurred before the fault evolved into a larger problem. Utilizing the analytic software to calculate phase current and symmetrical component magnitudes indicates there was sufficient current (1.3 A C-phase and 3I0) to assert the revised recloser failure logic.

Conclusion In summary, analysis of this series of event reports: • Reviewed the anticipated protection system behavior for a given fault. • Used simple analysis techniques and analytic software to unravel complex event details. Figure 8 — Effect of 52APU and 52ADO Settings on Relay Word Bits 52AT and 52BT

The relay generated ten event reports in just over 20 minutes according to its history. In addition to the three events reviewed here, there were six event reports triggered by brief downstream B-to-G faults. The oldest event in the history buffer was time stamped 11:47:33.395 (shown in Figure 9).

• Verified correct operation of the relay.

• Revealed a recloser failure and the need for maintenance on the C-phase interrupter.

• Showed two occurrences of the breaker problem, indicating that event analysis can expose problems such as these before they become more extreme.

• Identified a weakness in the as-set trip failure settings and provided data to develop an improved set of settings and logic that would have identified this problem the first time, notified SCADA, and locked out the problem equipment.

• Highlighted the need for safety improvements through breaker failure logic, local and remote indication, manual close operation safety delays, and failed recloser lockout. • Indicated multiple faults in the same vicinity, suggesting a problem at a specific line location requiring further investigation (trees invading line). • Demonstrated the need for cold load pickup logic to prevent misoperation on inrush.

Figure 9 — Another Failure of the Interrupter Recorded by Event Reports

• Illustrated the power of multifunction relays and programmable logic in developing solutions to each problem identified in the event reports (i.e., the power to solve all problems exists in the relay already installed).

62 The analytic assistant software used to create the oscillographic images generated COMTRADE files used to reproduce the fault sequence through test equipment into relays in the laboratory. By doing this, the improved logic solutions shown proved to function correctly, solving problems for the actual system fault using the existing relay. The event report files are stored as documentation for regulatory agencies and as proof of relay operations test. For more information, please download the technical paper “Understanding and Analyzing Event Report Information” at http://www.selinc.com/techpprs.htm. David Costello has a BS in electrical engineering from Texas A&M University. From 1991 to 1996 he was employed with Central Power and Light and Central and Southwest Services, Inc., where he worked as a system protection engineer. In 1996, he joined Schweitzer Engineering Laboratories, Inc. (SEL) as a field application engineer. He currently holds the position of Regional Service Manager and is responsible for SEL customer support in the Southcentral and Southwest United States.

Protective Relaying Handbook — Volume 1

NETA Accredited Companies The following is a listing of all NETA Accredited Companies as of August 2011. Please visit the NETA website at www.netaworld.org for the most current list. A&F Electrical Testing., Inc...................................................................................Kevin Chilton Advanced Testing Systems ............................................................................Patrick MacCarthy American Electrical Testing Co., Inc. ......................................................................Scott Blizard Apparatus Testing and Engineering ....................................................................... James Lawler Applied Engineering Concepts .................................................................... Michel Castonguay Burlington Electrical Testing Company, Inc. ........................................................... Walter Cleary C.E. Testing, Inc. ........................................................................................... Mark Chapman CE Power Solutions of Wisconsin, LLC............................................................. James VanHandel DYMAX Holdings, Inc. ....................................................................................... Gene Philipp Eastern High Voltage ....................................................................................... Joseph Wilson ELECT, P.C. .................................................................................................Barry W. Tyndall Electric Power Systems, Inc. .................................................................................. Steve Reed Electrical and Electronic Controls ..................................................................... Michael Hughes Electrical Energy Experts, Inc............................................................................... William Styer Electrical Equipment Upgrading, Inc. .......................................................................Kevin Miller Electrical Maintenance & Testing, Inc........................................................................ Brian Borst Electrical Reliability Services ..................................................................................Lee Bigham Electrical Testing, Inc. ................................................................................. Steve C. Dodd Sr. Elemco Services, Inc. ...................................................................................... Robert J. White Hampton Tedder Technical Services ....................................................................... Matt Tedder Harford Electrical Testing Co., Inc. ................................................................... Vincent Biondino High Energy Electrical Testing, Inc..................................................................... James P. Ratshin High Voltage Maintenance Corp. ........................................................................... Eric Nation HMT, Inc. .........................................................................................................John Pertgen Industrial Electric Testing, Inc. ........................................................................ Gary Benzenberg Industrial Electronics Group ................................................................................. Butch E. Teal Industrial Tests, Inc. .............................................................................................. Greg Poole Infra-Red Building and Power Service ............................................................ Thomas McDonald M&L Power Systems, Inc. .................................................................................. Darshan Arora Magna Electric Corporation ................................................................................... Kerry Heid Magna IV Engineering – Edmonton ...................................................................Jereme Wentzell Magna IV Engineering (BC), Ltd. ........................................................................ Cameron Hite

Setting the Standard

MET Electrical Testing, LLC .......................................................................... William McKenzie National Field Services...................................................................................... Eric Beckman Nationwide Electrical Testing, Inc. ...............................................................Shashikant B. Bagle North Central Electric, Inc. ...............................................................................Robert Messina Northern Electrical Testing, Inc. .......................................................................... Lyle Detterman Orbis Engineering Field Service, Ltd. ....................................................................... Lorne Gara Pacific Power Testing, Inc. ...................................................................................Steve Emmert Phasor Engineering ........................................................................................... Rafael Castro Potomac Testing, Inc. ........................................................................................... Ken Bassett Power & Generation Testing, Inc.......................................................................... Mose Ramieh Power Engineering Services, Inc. ..................................................................... Miles R. Engelke POWER PLUS Engineering, Inc. ...................................................................Salvatore Mancuso Power Products & Solutions, Inc. ........................................................................ Ralph Patterson Power Services, LLC ........................................................................................ Gerald Bydash Power Solutions Group, Ltd ...........................................................................Barry Willoughby Power Systems Testing Co. ............................................................................... David Huffman Power Test, Inc. ..............................................................................................Richard Walker POWER Testing and Energization, Inc. ............................................................... Chris Zavadlov Powertech Services, Inc. ................................................................................... Jean A. Brown Precision Testing Group .................................................................................... Glenn Stuckey PRIT Service, Inc. ........................................................................................ Roderic Hageman Reuter & Hanney, Inc....................................................................................... Michael Reuter REV Engineering, LTD ................................................................................ Roland Davidson IV Scott Testing, Inc................................................................................................Russ Sorbello Shermco Industries ............................................................................................... Ron Widup Sigma Six Solutions, Inc. ....................................................................................... John White Southern New England Electrical Testing, LLC ................................................. David Asplund, Sr. Southwest Energy Systems, LLC .......................................................................Robert Sheppard Taurus Power & Controls, Inc. ............................................................................... Rob Bulfinch Three-C Electrical Co., Inc.................................................................................James Cialdea Tidal Power Services, LLC ....................................................................................Monty Janak Tony Demaria Electric, Inc. ............................................................................ Anthony Demaria Trace Electrical Services & Testing, LLC ...................................................................Joseph Vasta Utilities Instrumentation Service, Inc. ........................................................................Gary Walls Utility Service Corporation.................................................................................. Alan Peterson Western Electrical Services ......................................................................................Dan Hook

Setting the Standard

About NETA NETA (InterNational Electrical Testing Association) is an association of leading electrical testing companies; visionaries, committed to advancing the industry’s standards for power system installation and maintenance to ensure the highest level of reliability and safety. NETA is an accredited standards developer for the American National Standards Institute (ANSI) and defines the standards by which electrical equipment is deemed safe and reliable. NETA is also the leading source of specifications, procedures, testing, and requirements, not only for commissioning new equipment but for testing the reliability and performance of existing equipment.

QUALIFICATIONS OF THE TESTING ORGANIZATION An independent overview is the only method of determining the long-term usage of electrical apparatus and its suitability for the intended purpose. NETA Accredited Companies best support the interest of the owner, as the objectivity and competency of the testing firm is as important as the competency of the individual technician. NETA Accredited Companies are part of an independent, third-party electrical testing association dedicated to setting world standards in electrical maintenance and acceptance testing.

Hiring a NETA Accredited Company assures the customer that: • The NETA Technician has broad-based knowledge — this person is trained to inspect, test, maintain, and calibrate all types of electrical equipment in all types of industries. • NETA Technicians meet stringent educational and experience requirements in accordance with ANSI/NETA Standard for Certification of Electrical Testing Technicians, (ANSI/NETA ETT). • A registered Professional Engineer will review all engineering reports. • All tests will be performed objectively, according to NETA specifications, using calibrated instruments traceable to the National Institute of Science and Technology (NIST). • The firm is a well-established, full-service electrical testing business.

CERTIFICATION NETA Certified Technicians conduct the tests that ensure that electrical power equipment meets the ANSI/NETA standards’ stringent specifications. Certification of competency is particularly important in the electrical testing industry. Inherent in the determination of the equipment’s serviceability is the prerequisite that individuals performing the tests be capable of conducting the tests in a safe manner and with complete knowledge of the hazards involved. They must also evaluate the test data and make an informed judgment on the continued serviceability, deterioration, or nonserviceability of the specific equipment. NETA, a nationally-recognized certification agency, provides recognition of four levels of competency within the electrical testing industry in accordance with ANSI/NETA Standard for Certification of Electrical Testing Technicians, (ANSI/NETA ETT).

Setting the Standard