neta handbook series ii - protective vol 3-pdf

PRO

TECT

IVE

RELA

Y

HANDBOOK

Published By

VOLUME 3 SERIES II

IntellirentSponsored by

PROTEC

TIVE RELA

Y HA

ND

BOO

K VOLU

ME 3 SERIES II

Published by NETA

- The InterNational Electrical Testing A

ssociation

PRealy V3 Cover.indd 1 2/6/14 11:07 AM

the test equipment answer

T 972.317.0479 | www.intellirentco.com | contact us at [email protected]

Test

Equipment

Rental

PROTECTIVE RELAY HANDBOOK

VOLUME 3

Published by InterNational Electrical Testing Association

■■ Relay and Protection Systems

■■ Switchgear and Breakers (Low, Medium and High Voltage)

■■ Cable and Bus

■■ Transformers

■■ Batteries

■■ Motors and Rotating Apparatus

■■ Watthour Metering

■■ Power Quality and Consumption Analysis


Test Equipment Rental

888.902.6111972.317.0479info@intellirentco.comwww.intellirentco.com

We support testing of




■■ Transformers

■■ Batteries








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 © 2013 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.


3050 Old Centre Avenue, Suite 102, Portage, Michigan 49024

269.488.6382

www.netaworld.org

PROTECTIVE RELAYHANDBOOK

VOLUME 3

TABLE OF CONTENTS

Better, Faster, and More Efficient Relay Testing...We Have the Technology!........................................................................................5Chris Werstiuk, Manta Test Systems

Using Hall-Effect Sensors to Add Digital Recording Capability to Electromechanical Relays.......................................................................................24Amir Makki, Softstuf, Inc., Sanjay Bose, The Consolidation Edison Company of New York, Tony Giuliante, ATG Consulting and John Walsh, Sean Breatnech Technical Services (SBTS) LLC

An Improved Constant Source Impedance Testing Method for MHO Distance Relays............................................................................................29Jason Buneo and Rene Aguilar, Megger

Instantaneous Ground Fault Relays (50GS) and Zero-Sequence CTSPowel Technical Brief #68......................................................................................35Baldwin Bridger, Powell Electrical Manufacturer Co.

Modern Protective Relay Techniques:Using a 94-Year-Old Concept to Protect Electrical Equipment......................................36Suparat Pavavicharn, Basler Electric Company



269.488.6382www.netaworld.org

TABLE OF CONTENTS CONTINUED

Multi-Function Numerical Protection Relays Using Symmetrical Components for More Reliable and Secure Protection............................................................................41Steve Turner, Beckwith Electric Co., Inc.

Newer Solid-State Relays Offer Enhanced Flexibility for SCADA Solutions in the Municipal Environment.............................................................46Lynn Hamrick and Owen Wyatt, Shermco Industries

Relay Maintenance Has Changed, Hasn’t It?..............................................................49Kerry Heid, Magna Electric Corp.

Modern Protective Relay Techniques: Using a 94-year Old Concept to Protect..................51Suparat Pavavicharn, Basler Electric Company

Industry Advisory – NERC Relay Safe Work Practices....................................................55Scott A. Blizard, American Electrical Testing Co., Inc.

Solving Relay Misoperations with Line Parameter Measurements.....................................58Will Knapek, OMICRON

An Introduction to End-to-End Testing...........................................................................60Chris Werstiuk, Manta Test Systems

GPS Testing, The Future of Testing................................................................................61William Knapek, OMICRON electronics

NETA Accredited Companies ....................................................................................65



269.488.6382www.netaworld.org

1. INTRODUCTIONModern Microprocessor relays are much more powerful than

their predecessors and testing one of these relays can be a daunting task for the average relay tester. All of this new power increases the relay’s complexity exponentially which can make installation mistakes easier to create and harder to find. The modern relay tes-ter needs to adapt to new technologies to find the most effective test plan possible to make sure the relay has been installed correct-ly, and will operate when required after the testing is completed. The purpose of this paper is to discuss the possible test techniques available to help the reader determine which techniques will be most effective for his/her skill level and available technology.

2. A BRIEF HISTORY OF PROTECTIVE RELAYSWe will start with a brief history of protective relays to compare

the different generations and understand their basic operation.

A) Electro-Mechanical RelaysElectro-mechanical relays are considered the simplest form of

protective relays. Although these relays have very limited operat-ing parameters, functions, and output schemes, they are the foun-dation for all relays to follow and can have very complicated me-chanical operating systems. The creators of these relays were true geniuses as they were able to apply their knowledge of electrical systems and protection to create protective relays using magne-tism, polarizing elements, and other mechanical devices that mim-icked the characteristics they desired.

The simplest electro-mechanical relay is constructed with an in-put coil and a clapper contact. When the input signal (current or voltage) creates a magnetic field greater than the mechanical force holding the clapper open, the clapper closes to activate the appro-priate control function. The relay’s pick-up is adjusted by chang-ing the coil taps and/or varying the core material via an adjusting screw. The relay has no intentional time delay but has an inherent delay due to mechanical operating times. (see figure 1)

Figure 1: Example of Clapper Style Relay

Figure 2: Typical Electro-Mechanical Relay with Timing Disk

The next level of Electro-mechanical relay incorporates an inter-nal time delay using a rotating disk suspended between two poles of a magnet. When the input coil’s magnetic strength is greater than the mechanical force holding the disk in the reset position, the disk will begin to turn toward the trip position. As the input signal (voltage or current) increases, the magnetic force increases, and the disk turns faster. The relay’s pick-up is adjusted by chang-ing input coil taps and/or adjusting the holding spring tension. The time delay is altered by moving the starting position and varying the magnet strength around the disk. Time characteristics are pre-set by model. (see figure 2)

The next level of complexity included polarizing elements to determine the direction of current flow. This element is necessary for the following protective functions to operate correctly:

• Distance (21)

• Reverse Power (32)

• Loss of Field (40)

• Directional Overcurrent (67)

These relays used the components previously described but will not operate until the polarizing element detects that the current

BETTER, FASTER, AND MORE EFFICIENT RELAY TESTING...WE HAVE THE TECHNOLOGY!

PowerTest 2012by Chris Werstiuk, Manta Test Systems

5Protective Relay Handbook

is flowing in the correct direction. Polarizing elements use resis-tors, capacitors, and comparator circuits to monitor current flow and operate a clapper style contact to shunt or block the protective functions accordingly. (see figures 3 and 4)

Figure 3: Example of an Electro-Mechanical Relay Polarizing Element

Figure 4: Typical Polarizing Element Electrical Schematic

As electro-mechanical relays are largely dependent on the in-teraction between magnets and mechanical parts, their primary problems are shared with all mechanical devices. Dirt, dust, corro-sion, temperature, moisture, and nearby magnetic fields can affect relay operation. The magnetic relationship between devices can

also deteriorate over time and cause the pick-up and timing char-acteristics of the relays to change or drift without regular testing and maintenance.

These relays usually only have one or two output contacts and auxiliary devices are often required for more complex protection schemes. Figure 5 depicts a simple overcurrent protection scheme using electromechanical relays for one feeder. Notice that four re-lays are used to provide optimum protection and any single-phase relay can be removed for testing or maintenance without compro-mising the protection scheme.

Figure 5: Typical Electromechanical Overcurrent Trip SchematicB) Solid State Relays

As technology progressed and electronic components shrunk in size, solid-state relays began to appear. The smaller, lighter, and cost-effective solid-state relays were designed to be direct replacements for the electro-mechanical relays. However, this generation of relays introduced new, unfore-seen problems including; power supply failures and electronic component failures that prevented relays from operating, and sensitivity to harmonics that caused nuisance trips. Protective relays are the last line of defense during an electrical fault, and they must operate reliably. Unfortunately, early solid-state relays were often unreliable, and you will probably find many more electro-mechanical relays than solid state relays in older installations.

Solid-state relays used electronic components to convert the analog inputs into very small voltages that were monitored by electronic components. Pick-up and timing settings were ad-justed via dip switches and/or dials. If a pick-up was detected, a timer was initiated which caused an output relay to operate. Although the new electronics made the relays smaller, many models were made so they could be inserted directly into exist-ing relay cases allowing upgrades without expensive retrofitting expenses. Early models were direct replacements with no addi-tional benefits other than new technology, but later models were multi-phase or multi-function.Page 6/ Werstiuk

These relays used the components previously described but will not operate until the polarizing element detects that the current is flowing in the correct direction. Polarizing elements use resistors, capacitors, and comparator circuits to monitor current flow and operate a clapper style contact to shunt or block the protective functions accordingly.

Figure 3: Example of an Electro-Mechanical Relay Polarizing Element

1

2

3

4

5

6

7

8

TOC SI TOC

SI

D

TOC WOUNDSHADING COILS

TOC

D OPER

C2

C1D POT POL

P2

P1

R2 R1LINE

* * *

SI =SEAL INTOC = TIME OVERCURRENT UNITD = DIRECTIONAL UNIT

* = SHORT FINGERS

Figure 4: Typical Polarizing Element Electrical SchematicAs electro-mechanical relays are largely dependent on the interaction between

magnets and mechanical parts, their primary problems are shared with all mechanical

Page 7/ Werstiuk

devices. Dirt, dust, corrosion, temperature, moisture, and nearby magnetic fields can affect relay operation. The magnetic relationship between devices can also deteriorate over time and cause the pick-up and timing characteristics of the relays to change or drift without regular testing and maintenance.

These relays usually only have one or two output contacts and auxiliary devices are often required for more complex protection schemes. Figure 5 depicts a simple overcurrent protection scheme using electromechanical relays for one feeder. Notice that four relays are used to provide optimum protection and any single-phase relay can be removed for testing or maintenance without compromising the protection scheme.

125VdcTRIP CIRCUIT

ICS

A PH51CO

A PHICS

A PH50IIT

1 2

10

A PH CO RLY 50/51

ICS

1 2

10

B PH CO RLY 50/51

ICS

1 2

10

C PH CO RLY 50/51

ICS

1 2

10

GND CO RLY 50/51

TRIPX-1

+ POS

- NEG

A PH51CO

A PHICS

A PH50IIT

A PH51CO

A PHICS

A PH50IIT

A PH51CO

A PHICS

A PH50IIT

Figure 5: Typical Electromechanical Overcurrent Trip Schematic

B) Solid State Relays

As technology progressed and electronic components shrunk in size, solid-state relays began to appear. The smaller, lighter, and cost-effective solid-state relays were designed to be direct replacements for the electro-mechanical relays. However, this generation of relays introduced new, unforeseen problems including; power supply failures and electronic component failures that prevented relays from operating, and sensitivity to harmonics that caused nuisance trips. Protective relays are the last line of defense during an electrical fault, and they must operate reliably. Unfortunately, early solid-state

Protective Relay Handbook6

C) Microprocessor Based RelaysMicroprocessor based relays are computers with preset

programming using inputs from the analog-to-digital cards (converts CT and PT inputs into digital signals), digital in-puts, communications, and some form of output contacts. The digital signals are analyzed by the microprocessor using algo-rithms (computer programs) to determine operational param-eters, pick-up, and timing based on settings provided by the end user.

The microprocessor relay, like all other computing devices, can only perform one task at a time. The microprocessor and will ana-lyze each line of computer code in predefined order until it reaches the end of the programming where it will begin analyzing from the beginning again. The relay scan time refers to the amount of time the relay takes to analyze the complete program once. A simplified program might operate as follows:

1. Start

2. Perform self-check

3. Record CT inputs

4. Record PT inputs

5. Record digital input Status’

6. Overcurrent pick-up? If yes, start timer

7. Instantaneous Pick-up? If yes, start timer

8. Any element for OUT101 On (1)? If yes turn OUT101 on.

9. Any element for OUT102 On (1)? If yes, turn OUT102 on.

10. Back to Start

Microprocessor relays also evaluate the input signals to de-termine if the analog input signal is valid using complex analy-ses of the input signal waveforms. These evaluations can re-quire significant portions of a waveform or multiple waveform cycles to properly evaluate the input signal which can increase the microprocessor relay’s operating time.

Electrical faults must be detected and cleared by the relay and circuit breaker as quickly as possible because an elec-trical fault can create an incredible amount of damage in a few cycles. The microprocessor relay’s response time is di-rectly related to the amount of programming and its processor speed. Early microprocessor speeds were comparatively slow, but they were also simple with small programs. As each ad-ditional feature or level of complexity is added, the processor speed must be increased to compensate for the additional lines of computer code that must be processed or the relay response time will increase. (see figures 6 and 7)

Figure 6: Simple Microprocessor Operation Flowchart

Figure 7: Simple Microprocessor Internal Schematic

D) Simple Digital RelaysEarly microprocessor based relays were nothing more than

direct replacements for electro-mechanical and solid state re-lays. Most were simple multiphase, single function relays with limited outputs. These relays were typically cheaper than com-parable relays from previous generations and added additional benefits including:

1. More sensitive settings,

2. Multiple time curve selections

3. Metering functions,

4. Remote communications,

5. Self-test functions that monitored key components to operate an LED on the front display or operate an output contact.

6. Simple fault recording

These relays were relatively simple to install, set, and test as they had limited functions and limited contact configurations. The General Electric MIF/MIV Series or Multilin 735/737 are good examples of simple digital relays.

/ Werstiuk

Electrical faults must be detected and cleared by the relay and circuit breaker as quickly as possible because an electrical fault can create an incredible amount of damage in a few cycles. The microprocessor relay’s response time is directly related to the amount of programming and its processor speed. Early microprocessor speeds were comparatively slow, but they were also simple with small programs. As each additional feature or level of complexity is added, the processor speed must be increased to compensate for the additional lines of computer code that must be processed or the relay response time will increase.

ANALOGTO

DIGITAL

CT

PT

MICROPROCESSOROUTPUTRELAYS

Perform self checkRecord CT inputsRecord PT inputsRecord digital input Status'Overcurrent pick up?If Yes, start timerInstantaneous Pick up?If yes, start timerTurn OUT101 On (1)?Turn OUT102 On (1)?Back to start

DIGITALINPUTS COMMS OTHER

COMMS

Figure 6: Simple Microprocessor Operation Flowchart ANALOGTO

DIGITALCONVERSION MICROPROCESSOR OUTPUT RELAYS


Page 9/ Werstiuk

Electrical faults must be detected and cleared by the relay and circuit breaker as quickly as possible because an electrical fault can create an incredible amount of damage in a few cycles. The microprocessor relay’s response time is directly related to the amount of programming and its processor speed. Early microprocessor speeds were comparatively slow, but they were also simple with small programs. As each additional feature or level of complexity is added, the processor speed must be increased to compensate for the additional lines of computer code that must be processed or the relay response time will increase.

ANALOGTO

DIGITAL

CT

PT

MICROPROCESSOROUTPUTRELAYS

Perform self checkRecord CT inputsRecord PT inputsRecord digital input Status'Overcurrent pick up?If Yes, start timerInstantaneous Pick up?If yes, start timerTurn OUT101 On (1)?Turn OUT102 On (1)?Back to start

DIGITALINPUTS COMMS OTHER

COMMS

Figure 6: Simple Microprocessor Operation Flowchart ANALOGTO

DIGITALCONVERSION MICROPROCESSOR OUTPUT RELAYS



E) Multi-Function Digital RelaysAs technology improved and microprocessors became faster and

more powerful, manufacturers began to create relays with the all-in-one-box philosophy we see today. These relays were designed to provide all the protective functions for an application instead of a protective element as seen in previous relay generations. Instead of installing four overcurrent (50/51), three Undervoltage (27), three overvoltage (59), two frequency (81), and 1 synchronizing (25) relays; just install one feeder management relay such as the SEL-351 or GE Multilin 750 relay to provide all these functions in one box for significantly less cost than any one of the previous relays. As a bonus, you also receive metering functions, a fault recorder, oscillography records, remote communication options, and additional protection functions you probably haven’t even heard of. Because all the protective functions are processed by one microprocessor, individual elements become interlinked. For example, the distance relay functions are automatically blocked if a PT fuse failure is detected.

The all-in-one-box philosophy caused some problems as all protection was now supplied by one device and if that device failed for any reason; your equipment was left without protec-tion. Periodic testing could easily be performed in the past with minimal risk or system disruption because only one element was tested at a time. Periodic testing with digital relays is a much more intrusive process as the protected device must either be shut down or left without protection during testing. Relay manu-facturers downplay this problem by stating that periodic testing is not required because the relay element tests will not change over time because of the digital nature of microprocessor relays and the relays have many self-check features that will detect most problems. However, output cards, power supplies, input cards, and analog-to-digital converters can fail without warning or detection and leave equipment or the system without protec-tion. Also, as everyone who uses a personal computer can attest, software can be prone to unexpected system crashes and digital relays are controlled by software.

As relays became more complex, relay settings became increas-ingly confusing. In previous generations a fault/coordination study was performed and the relay pick-up and time dial settings were determined then applied to the relay in secondary amps. Today we can have multiple elements providing the same protection but now have to determine whether the pick-up is in primary values, secondary values, or per unit. There can also be additional settings for even the simplest overcurrent (51) element including selecting the correct curve, voltage controlled or restraint functions, reset intervals, etc. Adding to the confusion is the concept of program-mable outputs where each relay output contact could be initiated by any combination of protective elements and/or external inputs, and/or remote inputs via communications. These outputs are pro-grammed with different setting interfaces based on the relay mod-el or manufacturer with no standard for schematic drawings.

Multifunction digital relays have also added a new problem through software revisions. The computer software industry ap-pears to be driven by the desire to add new features, improve opera-tion, and correct bugs from previous versions. Relay programmers from some manufacturers are not immune from these tendencies. It is important to realize that each new revision changes the relay’s programming and, therefore creates a brand new relay that must be tested after every revision change to ensure it will operate when required. In the past, the relay manufacturers and consumers, spe-cifically the utility industry, extensively tested new relay models for months — simulating various conditions to ensure the relay was suitable for their systems. The testing today is either faster or less stringent as the relays are infinitely more complex, and new revisions or models disappear before some end users approve their replacements for use in their systems.

Examples of multifunction digital relays include the Schweitzer Laboratories product line (SEL), the GE Multilin SR series and the Beckwith M Series.

F) The Future of Protective RelaysA paradigm shift occurred when relay designers realized that

all digital relays use the same components (analog-to-digital cards, input cards, output cards, microprocessor, and commu-nication cards) and the only real differences between relays is programming. With this principle in mind, new relays are being produced that use interchangeable analog/digital input/output, communication, and microprocessor cards. Using this model, features can be added to existing relays by simply adding cards and uploading the correct software. These protective relays will have the same look and feel as their counterparts across the prod-uct line.

These relays will be infinitely configurable but will also be in-finitely complex, requiring specialized knowledge to be able to operate and test. Also, software revisions will likely become more frequent. Another potential physical problem is also created if the modules are incorrectly ordered or installed.

Examples of this kind of relay include the Alstom M series, General Electric UR series, and the ABB REL series.

G) Digital Relay ConsiderationsWhile digital relays are more powerful, flexible, effective, and

loaded with extras, they are also exponentially more complex than their predecessors and can easily be misapplied. As the relay ac-complishes more functions within its programming, many of the protective functions within an electrical system become invisible and are poorly documented.

The electrical protection system in the past was summarized on a schematic drawing that was simple to understand and almost all of the necessary information was located in one place. If informa-tion was missing, you could look at the control cabinet wiring and trace the wires.


Today’s protection system is locked within a box and poorly docu-mented at best; this makes it nearly impossible for an operator or plant electrician to troubleshoot. It can even be difficult for a skilled relay technician due to communication issues, complex logic, cyber securi-ty, and difficult software. A one-character mistake can be catastrophic and can turn thousands of dollars’ worth of protective relay into a very expensive Christmas decoration. Some problems may never be dis-covered because the relay may only need to operate one percent of the time and testing periods are rare due to manufacturer’s claims that me-tering and self-diagnosis tests are all that’s required for maintenance.

There is also a lack of standardization between engineering com-panies and relay manufacturers regarding how control, annunciation, and protection should be documented. While electro-mechanical or solid-state relay schemes would be well documented in the schematic diagrams for the site, few engineering companies provide an equiva-lent drawing of internal relay output logic and often reduce a com-plex logic scheme into “OUT 1/TRIP” or simply “OUT1.” This kind of documentation is almost useless when trying to troubleshoot the cause of an outage, especially if the relay has cryptic annunciation on its display panel. An operator usually does not need the extra hassle of finding a laptop, RS-232 cable and/or RS-232 to RS-485 adapter, cable adapter, only to sort through the various communication prob-lems and menu layers only to determine that an overcurrent (51) ele-ment caused the trip. In the past, the operator could simply walk to the electrical panel and look for the relay target and read the label above the relay to determine what caused the trip. (see figures 8 through 11)


Figure 9: Typical Digital Relay Trip Schematic

Figure 10: Electro-Mechanical Generator Protection Panel

Figure 11: Digital Generator Protection Panel

A properly applied microprocessor relay, on the other hand, will usually provide protection that is more sensitive and can be a god-send when you are looking for those elusive, quirky problems. As with all aspects of life, there are trade-offs for every benefit. The dif-ference between electro-mechanical relays and digital relays can be compared to the carburetor and fuel injection in a car. The modern fuel injector is more efficient with less moving parts than a carbure-tor, but you would probably prefer a carburetor when you are broken down in the middle of the desert with no help in sight, because you could at least try something with the carburetor to get you home.

Page 13/ Werstiuk

There is also a lack of standardization between engineering companies and relay manufacturers regarding how control, annunciation, and protection should be documented. While electro-mechanical or solid-state relay schemes would be well documented in the schematic diagrams for the site, few engineering companies provide an equivalent drawing of internal relay output logic and often reduce a complex logic scheme into “OUT 1/TRIP” or simply “OUT1.” This kind of documentation is almost useless when trying to troubleshoot the cause of an outage, especially if the relay has cryptic annunciation on its display panel. An operator usually does not need the extra hassle of finding a laptop, RS-232 cable and/or RS-232 to RS-485 adapter, cable adapter, only to sort through the various communication problems and menu layers only to determine that an overcurrent (51) element caused the trip. In the past, the operator could simply walk to the electrical panel and look for the relay target and read the label above the relay to determine what caused the trip.

125VdcTRIP CIRCUIT

ICS

A PH51CO

A PHICS

A PH50IIT

1 2

10

A PH CO RLY 50/51

ICS

1 2

10

B PH CO RLY 50/51

ICS

1 2

10

C PH CO RLY 50/51

ICS

1 2

10

GND CO RLY 50/51

TRIPX-1

+ POS

- NEG

A PH51CO

A PHICS

A PH50IIT

A PH51CO

A PHICS

A PH50IIT

A PH51CO

A PHICS

A PH50IIT


Page 14/ Werstiuk

125VdcTrip Circuit

OUT1

OUT2

3 PHASE MICROPROCESSOR RELAY

43CSLOCALTRIP

94T1REMOTE

TRIP

TC

52a

+ POS

- NEG

OUT3

OUT4

OUT5

OUT6

OUT7

OUT8

52-211TC#2

52a

SCADA EVENTRECORDER

52-211TC#1

Figure 9: Typical Digital Relay Trip Schematic

Figure 10: Electro-Mechanical Generator Protection Panel

Figure 11: Digital Generator Protection Panel

A properly applied microprocessor relay, on the other hand, will usually provide protection that is more sensitive and can be a godsend when you are looking for those


3. REASONS FOR RELAY TESTINGIn the past, relay testing options were limited by the test equip-

ment available and the simple (relatively speaking) electro-me-chanical relays to be tested. Today’s relays are highly sophisticated with an incredible number of settings that are tested with equally powerful test equipment and test systems. Complicated electro-mechanical relays required only a day to calibrate and test all func-tions. A digital relay could require weeks to test all of its functions. Relay testing is expensive and intelligent choices regarding what we test and how the test must be made. The best answers to these questions come from an evaluation of our goals and options.

Some of the reasons for relay testing include:

A) Type TestingType testing is a very extensive process performed by a manu-

facturer or end user that runs a relay through all of its paces. The manufacturer uses type testing to either prove a prospective relay model (or software revision), or as quality control for a recently manufactured relay. The end user, usually a utility or large corpo-ration, can also perform type tests to ensure the relay will operate as promised and is acceptable for use in their system.

This kind of testing is very involved and, in the past, would take months to complete on complicated electro-mechanical relays. Every conceivable scenario that could be simulated was applied to the relay to evaluate its response under various conditions. Today, all of these scenarios are now stored as computer simulations that are replayed through advanced test equipment to prove the relay’s performance in hours instead of weeks.

Type testing is very demanding and specific to manufacturer and/or end-user standards. Type testing should be performed before the relay is ordered and is not covered in this book. However, indepen-dent type testing is a very important part of a relay’s life span and should be performed by end users before choosing a relay model.

B) Acceptance TestingThere are many different definitions of acceptance testing. For

our purposes, acceptance testing ensures that the relay is the cor-rect model with the correct features; is operating correctly; and has not been damaged in any way during transport. This kind of testing is limited to functional tests of the inputs, outputs, metering, com-munications, displays, and could also include pick-up/timing tests at pre-defined values. Acceptance test procedures are often found in the manufacturer’s supplied literature.

C) CommissioningCommissioning and acceptance testing are often confused with

each other but can be combined into one test process. Acceptance testing ensures that the relay is not damaged. Commissioning con-firms that the relay’s protective element and logic settings are ap-propriate to the application and will operate correctly. Acceptance tests are generic and commissioning is site specific.

Commissioning is the most important test in a digital relay’s lifetime.

The tests recommended in this book are a combination of accep-tance testing and commissioning and includes the following items:

• Functional tests of all digital inputs/outputs

• Metering tests of all current and voltage inputs

• Pick-up and timing tests of all enabled elements

• Functional checks of all logic functions

• Front Panel display, target and LEDs

This battery of tests is a combination of acceptance testing and commissioning to prove:

• The relay components are not damaged and are acceptable for service

• The specific installation is set properly and operates as expected.

• The design meets the application requirements.

• The relay works in conjunction with the entire system.

D) Maintenance TestingMaintenance tests are performed at specified intervals to

ensure that the relay continues to operate correctly after it has been placed into service. In the past, an electro-mechanical relay was removed from service, cleaned, and fully tested using as-found settings to ensure that its functions had not drifted or connections had not become contaminated. These tests were necessary due to the inherent nature of electro-mechanical relays.

Today, removing a relay from service effectively disables all equipment protection in most applications. In addition, digital relay characteristics do not drift and internal self-check func-tions test for many errors. There is a heated debate in the in-dustry regarding maintenance intervals and testing due to the inherent differences between relay generations.

In my opinion, the following tests should be performed annually on all digital relays:

• Perform relay self-test command, if available.

• Check metering while online and verify with external metering device, or check metering via secondary injection. (This test proves the analog-to-digital converters)

• Verify settings match design criteria.

• Review event record data for anomalies or patterns.

• Verify all inputs from end devices.

• Verify all connected outputs via pulse/close command or via secondary injection. (Optionally, verify the complete logic output scheme via secondary injection.)


E) TroubleshootingTroubleshooting is usually performed after a fault to determine

why the relay operated or why it did not operate when it was sup-posed to. The first step in troubleshooting is to review the event re-corder logs to find out what happened during the fault. Subsequent steps can include the following, depending on what you discover in the post-fault investigation.• Change the relay settings accordingly.• Change the event record or oscillography initiate commands.• Re-test the relay.• Test the relay’s associated control schemes.• Replay the event record through the relay or similar relay to see

if the event can be replicated.

4. EVOLUTION OF RELAY TESTINGThis section will outline the evolution of relay testing to better

understand the choices available to the relay-tester when testing modern digital relays.

A) Electromechanical Relay Testing TechniquesElectromechanical relays operated based on mechanics and

magnetism and it was important to test all of the relay’s charac-teristics to make sure that the relay was in tolerance. Various tests were applied to ensure all of the related parts were functioning correctly and, if the relay was not in tolerance; the relay resistors, capacitors, connections, and magnets were adjusted to bring the relay into tolerance. With enough patience, almost any relay could be adjusted to acceptable parameters.

Relay testing in the electromechanical age was very primitive due to the limitations of the test equipment available. The most advanced test equipment available to the average relay-tester would include a variac for current output, another variac for voltage signals, a built in timer with contact sensing, and a phase shifter for more advanced applications. With this equipment, detailed test plans and connec-tion diagrams, and a hefty dose of patience; the relay-tester was able to test the pick-up, timing, and characteristics of the electro-mechanical relays installed as well as most solid state relays. Test plans and connections for currents and voltages often had very little resemblance to the actual operating connections because the limited test equipment could not create simulations of actual system condi-tions during a fault. Electro-mechanical relays were also built with inter-related components that needed to be isolated for calibration.

The following techniques were used when testing electrical-mechanical relays:

i) Steady StateSteady state testing is usually used for pick-up tests. The in-

jected current/voltage/frequency is raised/lowered until the re-lay responds accordingly. Steady state testing can be replaced by jogging the injected value up/down until the relay responds. (see figure 12)

Figure 12: Steady State Pickup Testingii) Dynamic On/Off TestingDynamic on/off testing is the simplest form of fault simulation and was the first test used to determine timing. A fault condition is suddenly applied at the test value by closing a switch between the source and relay or activating a test-set’s output. (see figure 13)

Figure 13: Dynamic On/Off Waveform

iii) Simple Dynamic State TestingSome protective elements such as under-frequency (81) and under-voltage (27) require voltages and/or current before the fault condition is applied or the element will not operate correctly. Simple dynamic state testing uses pre-fault and/or post-fault values to allow the relay-tester to obtain accurate time tests. A normal current/voltage/frequency applied to the relay suddenly changes to a fault value. The relay-response timer starts at the transition between pre-fault and fault, and the timer ends when the relay operates. Simple dynamic state testing can be performed manually with two sources separated by contacts or switches; applying nominal signals and suddenly ramping the signals to fault levels; or using different states such as pre-fault and fault modes.

Figure 14: Simple Dynamic Test Waveform

Page 18/ Werstiuk

4. Evolution of Relay Testing This section will outline the evolution of relay testing to better understand the choices

available to the relay-tester when testing modern digital relays.

A) Electromechanical Relay Testing Techniques

Electromechanical relays operated based on mechanics and magnetism and it was important to test all of the relay’s characteristics to make sure that the relay was in tolerance. Various tests were applied to ensure all of the related parts were functioning correctly and, if the relay was not in tolerance; the relay resistors, capacitors, connections, and magnets were adjusted to bring the relay into tolerance. With enough patience, almost any relay could be adjusted to acceptable parameters.

Relay testing in the electromechanical age was very primitive due to the limitations of the test equipment available. The most advanced test equipment available to the average relay-tester would include a variac for current output, another variac for voltage signals, a built in timer with contact sensing, and a phase shifter for more advanced applications. With this equipment, detailed test plans and connection diagrams, and a hefty dose of patience; the relay-tester was able to test the pick-up, timing, and characteristics of the electromechanical relays installed as well as most solid state relays. Test plans and connections for currents and voltages often had very little resemblance to the actual operating connections because the limited test equipment could not create simulations of actual system conditions during a fault. Electro-mechanical relays were also built with inter-related components that needed to be isolated for calibration.

The following techniques were used when testing electrical-mechanical relays:

i) Steady State

Steady state testing is usually used for pick-up tests. The injected current/voltage/frequency is raised/lowered until the relay responds accordingly. Steady state testing can be replaced by

jogging the injected value up/down until the relay responds.

STEADY-STATE PICKUP TEST

1 A

2 A

3 A

4 A

5 A ELEMENT PICK-UP

PICKUPSETTING

JOGGING PICKUPTEST

40 A

ELEMENT PICK-UP

PICKUPSETTING

20 A

60 A

Figure 12: Steady State Pickup Testing


B) Solid State Relay Testing TechniquesSolid state relays were primarily created to be direct replace-

ments for electro-mechanical relays and the same test techniques were used for these relays. However, these solid state relays were constructed with silicon chips, digital logic, and mathematical for-mulas instead of steel and magnetism so the test plans were the same but the results were often very different. When an electro-mechanical relay was found out of tolerance, there were resistors or springs to adjust. Solid state relays did not have many adjust-ments besides the initial pick-up setting and these relays either operated correctly or did not operate at all. Relay-testers typically replaced entire cards instead of adjusting components when the relay failed.

Test equipment did not evolve much during this period and the typical test-set changed all of its analog displays to digital and made the previously described tests easier to perform.

C) Microprocessor Relay Testing TechniquesSimple microprocessor relays are almost identical in opera-

tion to the solid state relays they replaced and the test tech-niques for these relays are identical to the techniques previ-ously described.

Complex microprocessor relays included a large number of set-tings and interlinked elements which created confusion in the re-lay testing industry because a relay-tester could spend an entire week testing one relay and barely scratch the surface of the re-lay’s potential. The confusion increased when relay manufacturers claimed that relay testing was not required because the relay per-formed self-check functions and the end user would be informed if a problem occurred. Some manufacturers even argued that the relay could test itself by using its own fault recording feature to perform all timing tests. Eventually a consensus was reached where the relay-tester would test all of the enabled features in the relay. Relay-testers began modifying and combining their electro-mechanical test sheets to account for all of the different elements installed in one relay but the basic fundamentals of relay testing didn’t change very much.

One of the first problems that a relay-tester experiences when testing microprocessor relay elements is that different elements inside the relay often overlap. For example, an instantaneous (50) element set at 20A will operate first when trying to test a time-overcurrent (51) element at 6x (24A) its pick-up setting (4A). The relay-tester instinctively wants to isolate the element under test and usually changes the relay settings to set one output, preferably an unused one, to operate only if the element under test operates. Now they can perform that 6x test without interference from the 50-element. While these techniques will give the test technician a result for their test sheet, the very act of changing relay settings to get that result does not guarantee that the relay will operate correctly when required because the in-service relay settings and reactions have not been tested.

Relay-testers often use the steady-state and simple-dynamic test procedures described previously to perform their element tests on microprocessor relays which create another problem. These complex relays are constantly monitoring their input signals to determine if those signals are valid. The steady-state and simple-dynamic test procedures are often considered invalid system con-ditions by the relay and the protection elements will not operate to prevent nuisance trips for a perceived malfunction. For example, if a relay-tester tries to perform a standard electromechanical im-pedance test (21) on a digital relay, the relay will likely assume that there is a problem with a PT fuse and blocked the element; or the switch-on-to-fault (SOTF) setting could cause the relay to trip instantaneously. Relay-testers who encountered this problem often disable those blocking signals to perform their tests and, hopeful-ly, turned the blocking settings back on when they were complete. Again, the act of changing settings is fine if you need a number for a test sheet but will not guarantee that the relay will operate cor-rectly when it is required.

Modern test equipment allows the relay-tester to apply several different techniques to overcome any of the situations described above. When an instantaneous element operates before a time ele-ment timing test can operate, that is usually a good thing if the relay is programmed correctly. Instead of modifying the settings to get a result, the technician can modify their test plan to ensure that all time-element tests fall below the instantaneous pick-up. If a ground element operates before the phase element you are trying to test, ap-ply a realistic phase-phase or three-phase test instead and the ground element will not operate. If the loss of protection element prevents a distance relay from operating, apply a balanced 3-phase voltage for a couple of seconds between each test to simulate real life condi-tions. If switch-on-to-fault operates whenever you apply the fault condition; use an output to simulate the breaker status, apply pre-fault current, or lower the fault voltage so that you can lower the fault current when testing impedance relays. All of these possibili-ties are easily applied with modern test equipment to make our test procedures more intelligent, realistic, and effective.

Modern test equipment also allows the following additional test methods.

i) Computer-Assisted TestingBecause modern test equipment is controlled by electronics, computer-assisted testing became available. Standard test techniques could be repetitive on relays that were functioning correctly. Computer programs were created that would ramp currents and voltages at fixed rates in an effort to make relay testing faster with more repeatable results because every test would be performed identically. Computer-assisted testing has evolved to the point where the software will:

• connect to the relay• read the relay settings• create a test plan based on the enabled settings• modify the settings needed to isolate an element and prevent

interference


• test the enabled elements• restore the relay settings to as found values

By following the steps above, computer-assisted relay testing can replace the relay-tester on a perfectly functioning relay and can theoretically perform the tests faster than a human relay-tester can. This type of testing works extremely well when performing pick-up and timing tests of digital relays because these relays are computer programs themselves.

However, it is very unlikely that the basic test procedures described by computer-assisted testing, whether initiated by computers or humans, will discover a problem with a digital relay. Most digital relay problems are caused by the settings engineer and not the relay. If a computer or human reads the settings from the relay and regurgitates them into the test plan, they will not realize that the engineer meant to enter a 0.50A pickup but actually applied 5.0A which will make the ground pickup larger than the phase pickup. Each element will operate as programed and, when tested in isolation, will create excellent test results but may not be applied in the trip equation which was probably not tested by the automated program. An excellent relay technician could create additional tests to perform the extra steps necessary for a complete test; but will that technician be more capable or less capable as they rely more heavily on automation to perform their testing?

ii) State SimulationState simulations allow the user to create dynamic tests where the test values change between each state to test the relay’s reaction to changes in the power system. Multiple state simulations are typically required for more complex tests such as frequency load shedding, end-to-end tests, reclosing, breaker fail, and the 5% under/over pick-up technique described later in this chapter.

iii) Complex Dynamic State TestingComplex dynamic state testing recognizes that all faults have a DC offset that is dependent on the fault incidence angle and the reactance/resistance ratio of the system. Changing the fault incidence angle changes the DC offset and severity of the fault and can significantly distort the sine wave of a fault as shown in Figure 15. This kind of test requires high-end test equipment to simulate the DC offset and fault incidence angle and may be required for high speed and/or more complex state-of-the-art relays. (see figure 15)

Figure 15: Complex Dynamic Waveform

iv) Dynamic System Model Based TestingDynamic system model based testing uses a computer program to create a mathematical model of the system and create fault simulations based on the specific application. These modeled faults (or actual events recorded by a relay) are replayed through a sophisticated test-set to the relay. Arguably, this is the ultimate test to prove an entire system as a whole. However, this test requires specialized knowledge of a system, complex computer programs, advanced test equipment, and a very complex test plan with many possibilities for error. My biggest concern regarding this kind of testing is the level of expertise necessary for a successful test. In the age of fast-track projects and corporate downsizing, many times the design engineer is barely able to provide settings in time for energization without relying on the designer to provide test cases as well.

Figure 16: System Modeling Waveform

v) End-to-End TestingEnd to end testing is performed when two or more relays are connected together via communication channels to protect a transmission line. These relays can transfer status or metering information between each other to constantly monitor the transmission line in order to detect faults and isolate the faulted transmission line more quickly and reliably than single relay applications. The relays can communicate to each other through a wide range of possibilities including telecom equipment, fiber optic channels, or wave traps that isolate signals transferred over the transmission line.

Testing these complicated schemes in the past was limited to functional tests of the individual components with a simplified system test to prove that the basic functions were operating correctly. For example, a relay-tester would configure and test the relay, configure and test the communication equipment, then inject a fault condition into the relay. A relay-tester at the other location would verify that they received the signal and they would repeat the process at the remote end. This procedure tested the base components of the system but they often failed to detect problems that occurred with faults in real time. For example, a fault on parallel feeders could change direction in fractions of cycles when one breaker in the system operated that often caused the protection schemes to mis-operate.


Relay-testers could only test one end at a time because there was no way to have two test-sets at remote locations start at exactly the same moment. If the test-sets do not provide coordinated currents and voltages with a fraction of a cycle, the protection scheme would detect a problem and mis-operate. Global Positioning System (GPS) technology uses satellites with precise clocks to communicate with equipment on earth which allowed test-set manufacturers to synchronize test-sets at remote distances. After the test-sets are synchronized, test plans could be created with simulated faults for each end of the transmission line. To perform a test, the relay-testers synchronize their test-sets, load the same fault simulation with the values for their respective ends, set the test-sets to start at the exact same moment, and initiate the countdown. The test-sets will inject the fault into the relays simultaneously and they should respond as if the fault occurred on the line. The relays’ reactions are analyzed and determined to be correct before proceeding to the next test. Any mis-operations are investigated to see where the problem originates and corrected.

End-to-end testing is typically considered to be the ultimate test of a system and should ideally perform using Dynamic System Model Testing to ensure that the system is tested with the most comprehensive test conditions. Simpler end-end schemes such as line differential schemes can be tested using Simple Dynamic State testing.

5. RELAY TEST PROCEDURES – PICKUP TESTINGThere are several methods used to determine pick-up, and we

will review the most popular in order of preference. You must re-member that we strive for minimum impact or changes when test-ing. If there is a method to determine pick-up without changing a setting…use it!

A) Choose a Method to Detect Pick-upi) Front Display LEDs

Many relays will have LEDs on the front display that are predefined for pick-up or can be programmed to light when an element picks up. Choose or program the correct LED and change the test-set output until the LED is fully lit. Compare the value to the manufacturer’s specifications and record it on the test sheet.

• SEL relays often allow you to customize LED output configurations using the “TAR” or “TAR F” commands. Some SEL LEDs can only be controlled via the relay’s front panel.

• Monitor Beckwith Electric relay pick-up values by pressing and holding the reset button. The appropriate LED will light on pick-up. Some elements share a single LED, and you must use alternate methods to determine pick-up.

• GE UR relays allow you customize the front display LEDs using the “Product Set-up” “User Programmable LEDs” menu.

• Most of the GE/Multilin element pick-ups can be monitored via the pre-defined pick-up LED.

ii) Front Display Timer IndicationSome relays, including the Beckwith Electric models, have a menu item on the front panel display that shows the actual timer value in real-time. After selecting the correct menu item, change the test-set output until the timer begins to count down. Compare the value to the manufacturer’s specifications and record it on the test sheet.

iii) Status Display via CommunicationSome relays provide a real-time status display on an external computer or other device via communication. While this is the most unobtrusive method of pick-up testing, the accuracy of this method is limited by the communication speed. Some relays will also slow down the communication speed during events that can further decrease accuracy.

Slowly change the test-set output and watch the display to get a feel for the time between updates. Change the relay test-set output at a slower rate than the update rate until the relay display operates. Compare the value to the manufacturer’s specifications and record it on the test sheet.

iv) Output ContactThis method requires you to assign an output contact to operate if the element picks up. Choose an unused output contact whenever possible and monitor the contact with your relay test-set or external meter. You can often hear the output contact operate, but you must be sure that you are listening to the correct relay. Another element in the control logic could be operating while you are performing your test.

If you are unable to assign a pick-up element to an output contact, you can set the element time delay to zero. This method is obviously not a preferred method. Always test the time delay after the pick-up to ensure the time delay was returned to the correct value.

B) Pick-up Test ProcedureAfter selecting a pick-up method, apply the current/volt-

age/frequency necessary for pick-up and make sure a pick-up is indicated by whichever method you have selected. Slowly decrease the relay test-set output until the pick-up indication disappears. Slowly increase the test-set output until pick-up is indicated. If the test current/voltage is higher than the input rat-ing, only apply test value for the minimum possible duration. See individual element testing for tips and tricks for individual elements. (see figure 17)


Figure 17: Graph of Pick-up Test

Always document setting changes and return the settings to their original values before proceeding!

6. RELAY TEST PROCEDURES – TIMING TESTSTiming tests apply a test input at a pre-defined value in the pick-

up region and measures the time difference from test initiation until the relay output-contact operates. This is the dynamic on/off testing method described earlier. Some elements like undervoltage (27) or under-frequency (81U) require pre-fault, non-zero values in order to operate correctly.

The output contact used to turn the timing set off is preferably the actual output contact used while in-service. The timing test con-tact can be a spare output contact if another element interferes with the element timing, but the actual element output contacts must be verified as well. Some outputs are designed to operate at different speeds. Always use the high-speed output if a choice is available.

If the time delay is a constant value such as zero seconds or one second, apply the input at 110% of the pick-up value and record the time delay. Determine if the manufacturers specified time de-lays are in seconds, milliseconds, or cycles and record test results. (see figures 18 and 19)

Figure 18: Simple Off/On Timing Test

Figure 19: Dynamic On/Off Testing

When the digital relay time delay is zero or very small (less than two seconds), the actual measured time delay can be longer than ex-pected. There is an inherent delay before the relay can detect a fault plus an additional delay between fault detection and relay output op-eration. Some relays use error checking features that can also increase the expected time delay because most dynamic faults involve sudden gaps in the current/voltage waveforms that would not occur during a real fault. These delays are very small (less than five cycles) and are insignificant with time delays greater than two seconds.

The first delay exists because the relay is a computer, and com-puters can only perform one task at a time. The relay evaluates each line of programming, one line at a time, until it reaches the end of the program, and then returns to the start to scan the entire program again. As more features are added, more program lines are added. Large program sizes are offset by faster processors that spend less time evaluating each line. If a fault occurs just after the relay processes the line of code that detects that particular fault, the relay has to run through the entire program one more time before the fault is detected. This time delay is usually a fraction of a cycle. The “Timer Accuracy” specifications in Figure 20 detail this time delay.

Figure 20: SEL-311C Relay Element Specifications

The second time delay occurs after the relay has detected the fault and issues the command to operate the output relays. There is another fraction of a cycle delay to evaluate what output contacts should operate and then the actual contact operation can add up to an additional cycle depending on relay manufacturer, model, etc. “Pick-up Time” in the Figure 21 represents this delay for the specified relay.

Page 27/ Werstiuk

Slowly change the test-set output and watch the display to get a feel for the time between updates. Change the relay test-set output at a slower rate than the update rate until the relay

display operates. Compare the value to the manufacturer’s specifications and record it on the test sheet.

iv) Output Contact

This method requires you to assign an output contact to operate if the element picks up. Choose an unused output contact whenever possible and monitor the contact with your relay test-set or external meter. You can often hear the output contact operate, but you must be sure that you are listening to the correct relay. Another element in the control logic could be operating while you

are performing your test.If you are unable to assign a pick-up element to an output contact, you can set the element time

delay to zero. This method is obviously not a preferred method. Always test the time delay afterthe pick-up to ensure the time delay was returned to the correct value.

B) Pick- up Test Procedure

After selecting a pick-up method, apply the current/voltage/frequency necessary for pick-up and make sure a pick-up is indicated by whichever method you have selected. Slowly decrease the relay test-set output until the pick-up indication disappears. Slowly increase the test-set output until pick-up is indicated. If the test current/voltage is higher than the input rating, only apply test value for the minimum possible duration. See individual element testing for tips and tricks for individual elements.

STEADY-STATE PICKUP TEST

1 A

2 A

3 A

4 A

5 A ELEMENT PICK-UP

PICKUPSETTING

Figure 17: Graph of Pick-up TestAlways document setting changes and return the settings to their original values

before proceeding!

Page 28/ Werstiuk

6. Relay Test Procedures – Timing Tests Timing tests apply a test input at a pre-defined value in the pick-up region and measures

the time difference from test initiation until the relay output-contact operates. This is the dynamic on/off testing method described earlier. Some elements like undervoltage (27) or under-frequency (81U) require pre-fault, non-zero values in order to operate correctly.

The output contact used to turn the timing set off is preferably the actual output contact used while in-service. The timing test contact can be a spare output contact if another element interferes with the element timing, but the actual element output contacts must be verified as well. Some outputs are designed to operate at different speeds. Always use the high-speed output if a choice is available.

If the time delay is a constant value such as zero seconds or one second, apply the input at 110% of the pick-up value and record the time delay. Determine if the manufacturers specified time delays are in seconds, milliseconds, or cycles and record test results.

1 2 3 4 5 6 70

TIME INCYCLES

60V

90V

120V

150V

TEST IN PROGRESS

PICKUP

30V


1 2 3 4 5 6 70

TIME IN SECONDS

30V

60V

90V

120V

150V TEST IN PROGRESS

PICKUP

Figure 19: Dynamic On/Off TestingWhen the digital relay time delay is zero or very small (less than two seconds), the actual

measured time delay can be longer than expected. There is an inherent delay before the relay

Page 28/ Werstiuk

6. Relay Test Procedures – Timing Tests Timing tests apply a test input at a pre-defined value in the pick-up region and measures

the time difference from test initiation until the relay output-contact operates. This is the dynamic on/off testing method described earlier. Some elements like undervoltage (27) or under-frequency (81U) require pre-fault, non-zero values in order to operate correctly.

The output contact used to turn the timing set off is preferably the actual output contact used while in-service. The timing test contact can be a spare output contact if another element interferes with the element timing, but the actual element output contacts must be verified as well. Some outputs are designed to operate at different speeds. Always use the high-speed output if a choice is available.

If the time delay is a constant value such as zero seconds or one second, apply the input at 110% of the pick-up value and record the time delay. Determine if the manufacturers specified time delays are in seconds, milliseconds, or cycles and record test results.

1 2 3 4 5 6 70

TIME INCYCLES

60V

90V

120V

150V

TEST IN PROGRESS

PICKUP

30V


1 2 3 4 5 6 70

TIME IN SECONDS

30V

60V

90V

120V

150V TEST IN PROGRESS

PICKUP

Figure 19: Dynamic On/Off TestingWhen the digital relay time delay is zero or very small (less than two seconds), the actual

measured time delay can be longer than expected. There is an inherent delay before the relay

Page 29/ Werstiuk

can detect a fault plus an additional delay between fault detection and relay output operation. Some relays use error checking features that can also increase the expected time delay because most dynamic faults involve sudden gaps in the current/voltage waveforms that would not occur during a real fault. These delays are very small (less than five cycles) and are insignificant with time delays greater than two seconds.

The first delay exists because the relay is a computer, and computers can only perform one task at a time. The relay evaluates each line of programming, one line at a time, until it reaches the end of the program, and then returns to the start to scan the entire program again. As more features are added, more program lines are added. Large program sizes are offset by faster processors that spend less time evaluating each line. If a fault occurs just after the relay processes the line of code that detects that particular fault, the relay has to run through the entire program one more time before the fault is detected. This time delay is usually a fraction of a cycle. The “Timer Accuracy” specifications in Figure 20 detail this time delay.

Instantaneous/Definite-Time Overcurrent Elements

Pickup Range: OFF, 0.25 - 100.00 A, 0.01 A steps (5 A nominal)OFF, 0.05 - 20.00 A, 0.01 A steps (1 A nominal)

Steady State Pickup Accuracy: +/- 0.05 A and +/-3% of Setting (5 A nominal)+/- 0.01 A and +/-3% of Setting (1 A nominal)

Transient Overreach: < 5% of Pickup

Time Delay: 0.00 - 16,000.00 cycles, 0.25-cycle steps

Timer Accuracy: +/- 0.25 cycle and +/-0.1% of setting

Figure 20: SEL-311C Relay Element SpecificationsThe second time delay occurs after the relay has detected the fault and issues the command

to operate the output relays. There is another fraction of a cycle delay to evaluate what output contacts should operate and then the actual contact operation can add up to an additional cycle depending on relay manufacturer, model, etc. “Pick-up Time” in the Figure 21 represents this delay for the specified relay.

Output Contacts: 30 A Make6A continuous carry at 70 C; 4 A continuous carry at 85 C50A for one secondMOV Protected: 270 Vac, 360 Vdc, 40 ; Pickup Time: <5ms.

Figure 21: SEL-311C Output Contact SpecificationsYour test-set also adds a minor delay to the test result as shown by the “Accuracy”

specifications in Figure 22:MANTA MTS-1710 TIME MEASUREMENT SPECIFICATIONS


Figure 21: SEL-311C Output Contact Specifications

Your test-set also adds a minor delay to the test result as shown by the “Accuracy” specifications in Figure 22:

Figure 22: Manta Test Systems MTS-1710 Technical Specifications

What does all this mean? With a time delay of zero, the time test result for a SEL 311C relay, using a Manta MTS-1710 test-set, could be as much as 28.0 ms or 1.68 cycles as shown in Figure 23.

Figure 23: 50-Element Maximum Expected Error

An SEL-311C relay instantaneous overcurrent element is set for one cycle and the timing test result was 2.53 cycles. That equals 153% error using the percent error formula and the relay appears to fail the timing test. But we determined that the relay/test-set combination could add 1.68 cycles to the time delay. The expected time (one cycle) plus the maximum “50-element expected Error” (1.68 cycles) equals 2.68 cycles. The test result is lower than the maximum expected time calculated in the previous sentence and, therefore, the relay passes the timing test.

7. RELAY TEST PROCEDURES – LOGIC TESTINGThe relay testing methods described so far have limitations

when applied to microprocessor relays and are more suited to ac-ceptance testing because they only prove that the analog inputs (voltage and current signals) are operating correctly, at least one output is operating correctly, and the relay will do what it is pro-grammed to do when elements are isolated. Logic testing attempts to apply a more holistic testing approach that tests the relay with-out changing relay settings and monitoring the contacts that will actually operate when the relay is in service.

There are some serious flaws when you use traditional test methods to perform commissioning tests. The goal of a commis-sioning test is to ensure that the relay will operate correctly when applied to a specific application using the installed settings. This requires testing with the applied settings and ensuring the relay has been properly configured. Are you really performing a com-

missioning test of as-left settings when you are changing settings to test? If OUT101 is connected to your trip circuit and all of your testing is performed on OUT107, how do you know that OUT101 is operating correctly? Does your output logic equation include all of the enabled elements? Are all of the enabled elements in your trip equation?

Almost all problems found in the field with microprocessor re-lays have absolutely nothing to do with the actual relay and occur because of drawing and/or relay setting mistakes. Here are some examples of some common mistakes found during relay testing.

• A differential relay element tests correctly on all three phases when isolated but a GE T-60 relay’s output setting is “XFMR PCNT DIFF OP A” which will only operate if an A Phase fault is detected. B and C differential protection is effectively disabled. The correct element was “XFMR PCNT DIFF OP”. A one character mistake could have made a B or C phase differential fault much worse than it could have been.

• The 50N1 (Instantaneous Overcurrent on IN input) setting is in the trip equation but 50N1 is off in the element settings. 50G1 (Residual Instantaneous Overcurrent) is on in the element settings but missing from the trip equation. All ground protection is disabled.

• A generator step-up transformer differential element is to be disabled by the lower voltage starting breaker 52a signal when the generator is run up to speed. However, a 52b breaker signal is actually sent which disabled differential protection when the generator is online. The relay will never trip and could cause millions of dollars in damage and lost revenue for a year waiting for a replacement transformer.

None of the examples described above would have been dis-covered using traditional testing techniques. Several of these problems were found several years after the relay was placed into service during maintenance testing by a different relay-tester.

Logic testing starts by comparing all of the onsite documen-tation to the settings and making sure all drawings and the relay settings match. You can create your test plan based on the relay settings once all of the site documentation is reviewed. It is important to look at the settings objectively and look for incon-sistencies inside the settings themselves. Look for impossible logic conditions and make sure that an element that is enabled and setup is also found in the output logic. Look for elements in the output logic that aren’t turned on or set. If there are no obvi-ous errors, note the logic for each output, including signals sent over communication channels and LED or front panel displays. Once you have a comprehensive list of all of the output logic, create a checklist for each output broken down into simple OR statements. For example, a simple SEL overcurrent relay might have the following settings:

Page 29/ Werstiuk














Page 29/ Werstiuk














Page 30/ Werstiuk

Auto ranging Scale: 0 – 99999 secAuto ranging Scale: 0 – 99999 cyclesBest Resolution: 0.1 ms / 0.1 cycles

Two wire pulse timing modeAccuracy: 0 – 9.9999 sec scale: +/-0.5ms +/- 1LS digit

all other scales: +/- 0.005% +/- 1LS digitFigure 22: Manta Test Systems MTS-1710 Technical Specifications

What does all this mean? With a time delay of zero, the time test result for a SEL 311C relay, using a Manta MTS-1710 test-set, could be as much as 28.0 ms or 1.68 cycles as shown in

Minimum Time Test ResultRelay Operate Time: 0.25 cyclesRelay Timing Accuracy: 0.10 cycles (0.1% of setting, because setting is zero

and next setting is 1 cycle, use 0.1 cycles)

Relay Operate Time: 0.30 cycles (< 5 ms)Test-set : 0.03 cycles (+/-0.005%)

1.00 cycles (+/- 1LS digit)1.68 cycles or 0.28 ms

Figure 23: 50-Element Maximum Expected Error An SEL-311C relay instantaneous overcurrent element is set for one cycle and the timing

test result was 2.53 cycles. That equals 153% error using the percent error formula and the relay appears to fail the timing test. But we determined that the relay/test-set combination could add 1.68 cycles to the time delay. The expected time (one cycle) plus the maximum “50-element expected Error” (1.68 cycles) equals 2.68 cycles. The test result is lower than the maximum expected time calculated in the previous sentence and, therefore, the relay passes the timing test.

Page 30/ Werstiuk

Auto ranging Scale: 0 – 99999 secAuto ranging Scale: 0 – 99999 cyclesBest Resolution: 0.1 ms / 0.1 cycles

Two wire pulse timing modeAccuracy: 0 – 9.9999 sec scale: +/-0.5ms +/- 1LS digit

all other scales: +/- 0.005% +/- 1LS digitFigure 22: Manta Test Systems MTS-1710 Technical Specifications

What does all this mean? With a time delay of zero, the time test result for a SEL 311C relay, using a Manta MTS-1710 test-set, could be as much as 28.0 ms or 1.68 cycles as shown in

Minimum Time Test ResultRelay Operate Time: 0.25 cyclesRelay Timing Accuracy: 0.10 cycles (0.1% of setting, because setting is zero

and next setting is 1 cycle, use 0.1 cycles)

Relay Operate Time: 0.30 cycles (< 5 ms)Test-set : 0.03 cycles (+/-0.005%)

1.00 cycles (+/- 1LS digit)1.68 cycles or 0.28 ms

Figure 23: 50-Element Maximum Expected Error An SEL-311C relay instantaneous overcurrent element is set for one cycle and the timing

test result was 2.53 cycles. That equals 153% error using the percent error formula and the relay appears to fail the timing test. But we determined that the relay/test-set combination could add 1.68 cycles to the time delay. The expected time (one cycle) plus the maximum “50-element expected Error” (1.68 cycles) equals 2.68 cycles. The test result is lower than the maximum expected time calculated in the previous sentence and, therefore, the relay passes the timing test.


• TRIP = 51P1T + 51N1T + 50P1 + 50N1

• (Trip Breaker) OUT101 = TRIP

• (Scada/Remote Trip Indication) OUT107 = TRIP

• (Front Panel Display) 52A = IN101, DP1 = 52A, DP_1 = Breaker Closed, DP_2 = Breaker Open

For those unfamiliar with SEL logic, a brief description of the SEL codes above includes:

• 51P1T—Phase Inverse Time Overcurrent Trip

• 51N1T—Neutral Inverse Time Overcurrent Trip

• 50P1—Phase Instantaneous Overcurrent Pickup

• 50N1—Neutral Instantaneous Overcurrent Pickup

• DP_1—Front Panel Display Point One

• DP_2—Front Panel Display Point One

If you wish to combine traditional pick-up and timing testing with logic testing, your test plan could look like the test plans described below. It is important to note that you must apply the appropriate fault simulation to ensure the element operates dur-ing the timing test to prevent interference by other elements. If a phase-related element is enabled, a Phase-Phase or 3-Phase test should be applied. If a ground related element is applied, a Phase-Ground test should be applied.

A) Test Plan #11. Perform 51P1T pick-up test using steady state technique and

use pick-up LED/Display/computer to determine pick-up (rec-ommended) or assign unused output for pick-up indication (not recommended)

2. Perform 51P1T timing test at 2xpick-up and use OUT101 for timer stop.






8. Perform 51N1T pick-up test using steady state technique and use pick-up LED/Display/computer to determine pick-up (recommend-ed) or assign unused output for pick-up indication (not recommended)

9. Perform 51N1T timing test at 2xpick-up and use OUT101 for timer stop.






15. Perform 50P1 pick-up test using steady state technique and use pick-up LED/Display/computer to determine pick-up (recom-mended) or assign unused output for pick-up indication (not recommended)

16. Perform 50P1 timing test at 1.1xpick-up and use OUT101 for timer stop.


18. Perform 50N1 pick-up test using steady state technique and use pick-up LED/Display/computer to determine pick-up (rec-ommended) or assign unused output for pick-up indication (not recommended)

19. Perform 50N1 timing test at 1.1xpick-up and use OUT101 for timer stop.


21. Check breaker status and compare to front panel display. (If breaker is open, then display should indicate open.)

22. Change breaker status and compare front panel display.

B) Test Plan #2 (Streamlined using alternating outputs)1. Perform 51P1T pick-up test using steady state technique and

use pick-up LED/Display/computer to determine pick-up (rec-ommended) or assign unused output for pick-up indication (not recommended).




5. Perform 51N1T pick-up test using steady state technique and use pick-up LED/Display/computer to determine pick-up (rec-ommended) or assign unused output for pick-up indication (not recommended).





9. Perform 50P1 pick-up test using steady state technique and use pick-up LED/Display/computer to determine pick-up (recom-mended) or assign unused output for pick-up indication (not recommended).



12. Perform 50N1 pick-up test using steady state technique and use pick-up LED/Display/computer to determine pick-up (recom-mended) or assign unused output for pick-up indication (not recommended).



15. Check breaker status and compare to front panel display. (If breaker is open, then display should indicate open).


C) Test Plan #3 (Streamlined using multiple inputs and timers)

1. Perform 51P1T pick-up test using steady state technique and use pick-up LED/Display/computer to determine pick-up (rec-ommended) or assign unused output for pick-up indication (not recommended).

2. Perform 51P1T timing test at 2xpick-up and verify OUT101 and OUT107 operates.



5. Perform 51N1T pick-up test using steady state technique and use pick-up LED/Display/computer to determine pick-up (rec-ommended) or assign unused output for pick-up indication (not recommended).

6. Perform 51N1T timing test at 2xpick-up and verify OUT101 and OUT107 operates.



9. Perform 50P1 pick-up test using steady state technique and use pick-up LED/Display/computer to determine pick-up (recom-mended) or assign unused output for pick-up indication (not recommended).

10. Perform 50P1 timing test at 1.1xpick-up and verify OUT101 and OUT107 operates.

11. Perform 50N1 pick-up test using steady state technique and use pick-up LED/Display/computer to determine pick-up (recom-mended) or assign unused output for pick-up indication (not recommended).

12. Perform 50N1 timing test at 1.1xpick-up and verify OUT101 and OUT107 operates.

13. Check breaker status and compare to front panel display. (If breaker is open, then display should indicate open).


Notice that we do not simulate the breaker when performing the logic test for the 52A (IN101) front panel display. You should always use the actual end device to prove input status and logic to make sure the actual device status contact:

• uses the correct status indication

• is connected correctly

• uses the correct input voltage. Different relays use an internally supplied voltage source or external source to determine input status. Some relays can use both methods and an easily be connected incorrectly.

Relay logic is often more complex than the previous example and more complicated logic schemes should be broken down to a simple OR statement. For example, a breaker failure logic scheme could be written as:

• SV1 = (50P2 [0.5 A] + 50N2 [0.5 A]) *(SV1T [Seal-in] + TRIP [Initiate]) [Breaker fail operate logic]

• SV1PU = 15 cycles [Breaker Failure Timer]

• OUT102 = SV1T [Breaker Fail Signal = Current is still flowing through the breaker 15 cycles after the trip signal is sent and will stay closed until the current is lower than 0.5A. Send trip signal to the next upstream breaker)

This logic can be broken down into the following logic equations:

• OUT102 = 50P2 * TRIP

• OUT102 = 50N2 * TRIP

• OUT102 = 50P2 * SV1T

• OUT102 = 50N2 * SV1T

For those unfamiliar with SEL logic, some brief descriptions of the SEL codes above include:

• 50P2—Phase Instantaneous Overcurrent Pickup

• 50N2—Neutral Instantaneous Overcurrent Pickup


• SV1—SELogic Variable #1

• SV1PU—SELogic Variable #1 Time Delay to Operate

• SV1T—SELogic Variable #1 Operated

Broken down into is base components, we can now test each of these equations using the following test plan.

D) Breaker Fail (OUT102) Test Plan1. OUT102 = 50P2 * TRIP. Perform 51P1T timing test at 2xpick-

up and set timer to start when OUT101 operates and to stop when OUT102 operates.

2. OUT102 = 50N2 * TRIP. Perform 51N1T timing test at 2xpick-up and set timer to start when OUT101 operates and to stop when OUT102 operates.

3. OUT102 = 50P2 * SV1T. Perform 51P1T timing test at 2xpick-up. Do not stop the test after OUT101 and OUT102 operates. Lower fault current below 51P1 pick-up setting but greater than 50P2 setting. OUT101 should open but OUT102 should still be closed. Lower fault current below 50P2 setting. Both outputs should now be open.

4. OUT102 = 50N2 * SV1T. Perform 51N1T timing test at 2xpick-up. Do not stop the test after OUT101 and OUT102 operates. Lower fault current below 51N1 pick-up setting but greater than 50N2 setting. OUT101 should open but OUT102 should still be closed. Lower fault current below 50N2 setting. Both outputs should now be open.

Applying logic testing will not find every problem but it will al-low the relay-tester to feel reasonably confident that the relay has been set correctly, there are no obvious logic errors, and the relay will operate when required and is connected properly.

8. RELAY TEST PROCEDURES – COMBINING PICKUP, TIMING TESTS, AND LOGIC TESTINGModern test equipment uses a minimum of three voltage and

three current outputs with the ability to independently vary the phase angles between any of the outputs. With this equipment, you can use different states to create more complicated dynam-ic tests which can make relay testing more realistic, effective, and efficient.

A microprocessor relay element does not fall out of calibra-tion…It either works correctly or it doesn’t. Using this principal, the pick-up and timing test can be combined into one test. The simplest element to use as an example is the instantaneous over-current (50) element. If the applied current is greater than the pick-up setting, the element will operate. If our example 50-element pickup setting is 25A, the element will not operate if the current is less than 25A, and will operate instantaneously if the current is greater than 25A in an ideal world.

The test plan to test the 50-element in one test is shown on the following chart. (see figures 24 and 25)

Figure 24: 50-Element Ideal Combination Test

We do not live in an ideal world and this test would probably fail due to accuracy errors of the relay and the test-set. In digital relays and modern test equipment, the combined error is usually less than 5%. We can modify our test-set to allow for the inherent error in relay testing using the following chart.

Figure 25: 50-Element Combination Test

This method will work for all relay elements, including ele-ments with time delays such as the time overcurrent (51) ele-ment.

Multiple timing tests were performed on electromechanical relays to ensure the relay magnets and mechanical time dials were in the proper location. If the relay timing did not match the manufacturer’s curve, a magnet or time dial setting was adjusted to bring the relay into tolerance. There are no possible adjust-ments to a microprocessor relay so the timing will be correct, or the relay is set wrong and it will be incorrect.

The following test plan will allow the user to test the pick-up and timing of a microprocessor based 51-element with a pick-up of 5A.

Page 38/ Werstiuk

The test plan to test the 50-element in one test is shown on the following chart.Pre-Fault Fault 1 Fault 2

Nominal Current (4.0 A) for 2 seconds

24.99 A for 1 second 25.00 AStart timerStop timer when relay output operates. Time should be instantaneous.

1 2 3 4 5 6 7TIME IN SECONDS

5 A10 A15 A20 A25 A

PRE-FAULTPICK UP

FAULT 1

FAULT 2

TRIP

Figure 24: 50-Element Ideal Combination TestWe do not live in an ideal world and this test would probably fail due to accuracy errors of

the relay and the test-set. In digital relays and modern test equipment, the combined error is usually less than 5%. We can modify our test-set to allow for the inherent error in relay testing using the following chart.

Pre-Fault Fault 1 Fault 2Nominal Current (4.0 A) for 2 seconds

23.75A for 1 second(25A - 5% = 25 - (25 x 0.05) = 25 - 1.25 = 23.75A)

26.25A (25A + 5% = 25 + (25 x 0.05) =25 + 1.25 = 26.25A)Start timerStop timer when relay output operates. Time should be less than 5 cycles


5 A10 A15 A20 A25 A

PRE-FAULTPICK UP

FAULT 1

FAULT 2

TRIP5% ERROR


Page 38/ Werstiuk





5 A10 A15 A20 A25 A

PRE-FAULTPICK UP

FAULT 1

FAULT 2

TRIP




23.75A for 1 second(25A - 5% = 25 - (25 x 0.05) = 25 - 1.25 = 23.75A)



5 A10 A15 A20 A25 A

PRE-FAULTPICK UP

FAULT 1

FAULT 2

TRIP5% ERROR


Page 38/ Werstiuk





5 A10 A15 A20 A25 A

PRE-FAULTPICK UP

FAULT 1

FAULT 2

TRIP




23.75A for 1 second(25A - 5% = 25 - (25 x 0.05) = 25 - 1.25 = 23.75A)



5 A10 A15 A20 A25 A

PRE-FAULTPICK UP

FAULT 1

FAULT 2

TRIP5% ERROR


Page 38/ Werstiuk





5 A10 A15 A20 A25 A

PRE-FAULTPICK UP

FAULT 1

FAULT 2

TRIP




23.75A for 1 second(25A - 5% = 25 - (25 x 0.05) = 25 - 1.25 = 23.75A)



5 A10 A15 A20 A25 A

PRE-FAULTPICK UP

FAULT 1

FAULT 2

TRIP5% ERROR



You should notice that the timing test current was equal to the multiple of current without adding 5% to compensate for test-set and relay error. The 5% error is used when comparing the time in fault…not the applied current. If there was a problem with the applied settings, the measured time delay would be significantly shorter than the expected time delay to indicate the problem.

Another problem would become evident if you were testing this relay if the 50 and 51-elements were assigned to the same trip coil. The 50-element (25A) would operate when the relay-tester tried to perform the 6x timing test (30A). The relay-tester could isolate the 51-element to another relay output to perform their 6x test without in-terference…or they could step back and review their procedure. Two different issues come into play when protective elements overlap.

The microprocessor relay’s 51-element does not have any pos-sible adjustments, so is a third test really necessary to prove the characteristic curve if two other tests are successful? If the third test is required, we can change the test current of the third test. The pri-mary reason for choosing whole numbers for 51-element tests in electro-mechanical relays is that it is easier to determine the expect-ed result on the graph using whole numbers. Most relay-testers are using the formulas to determine expected values when testing mi-croprocessor relays, so changing the third test current to 24A or 4.8x pick-up should be no problem when calculating the expected result.

The second issue in play is commissioning vs. acceptance testing. Testing without changing settings proves that the settings have been applied correctly. If 30 Amps are applied to the relay in-service, would the 51 or 50 element operate? Is there any advantage to test-ing an element at a test point where it will never operate in-service?

This technique works for all protective elements including more complicated relay elements such as distance protection (21). In fact, if you are unable to apply this test procedure and achieve a successful test, there is probably something wrong with the relay settings.

21-elements use the measured impedance and angle between the current and voltage to detect a fault on a transmission line or other electrical apparatus. The most typically applied characteristic is a MHO circle. If the measured impedance falls within the circle, the 21-element operates after a pre-set time delay. Our Zone 2 element is set at 3.4 Ω @ 87º with a 20 cycle time delay as shown in the following figure.

We can start the test by applying a prefault state using nom-inal conditions with an impedance near the x-axis, far away

from the circle. We then apply the Fault 1 impedance just out-side of the circle followed by the Fault 2 impedance applied just inside the circle. Don’t forget that there is also a 5% error to account for and that is shown by the shaded band around the original circle. If the measured time between fault 2 and the relay output is 20 cycles +/- the relay tolerance, the test is successful. You could use the same technique to plot the entire circle, but it is extremely unlikely you will find a problem with the relay’s programming which should have been tested in the Type Testing process. (see figure 26)


This test technique can also be applied to overlapping zones of protection. Our example relay has a Zone 1 distance protection element set at 1.5 Ω @ -87º with no intentional delay. We can modify the previous test plan for the new impedance to use the following test parameters: (see figure 27)

Figure 27: 21-Element Dual Zone Combination Test

The previous test would fail because the Fault 1 impedance falls within the Zone 2 circle and Zone 2 will trip in 20 cycles, which is a shorter time than the Fault 1 duration.

A simple modification of the Fault 1 time to a value greater than the Zone 1 time but less than the Zone 2 time delay will make

Page 41/ Werstiuk

We can start the test by applying a prefault state using nominal conditions with an impedance near the x-axis, far away from the circle. We then apply the Fault 1 impedance just outside of the circle followed by the Fault 2 impedance applied just inside the circle. Don’t forget that there is also a 5% error to account for and that is shown by the shaded band around the original circle. If the measured time between fault 2 and the relay output is 20 cycles +/- the relay tolerance, the test is successful. You could use the same technique to plot the entire circle, but it is extremely unlikely you will find a problem with the relay’s programming which should have been tested in the Type Testing process.

Pre-Fault Fault 1 Fault 2Voltage (69.28V)Current (3.0 A @ -30º)(23.09 Ω @ 30 º) for 2 seconds

Voltage (30.0V)Current (8.40A) @ -87º(3.57Ω @ 87 º) for 0.5 seconds

Voltage (30.0V)Current (9.29A) @ -87º(3.23Ω @ -87 º) Start timerStop timer relay output operates. Time should be 20 cycles +/- 5%

FAULT 1 - NO TRIP

FAULT 2 - TRIP

PRE-FAULT ---->

X

R

ZONE 2 21 ELEMENT


Page 42/ Werstiuk

This test technique can also be applied to overlapping zones of protection. Our example relay has a Zone 1 distance protection element set at 1.5 Ω @ -87º with no intentional delay. We can modify the previous test plan for the new impedance to use the following test parameters:

Pre-Fault Fault 1 Fault 2Voltage (69.28V)Current (3.0A) @ -30º(23.09Ω @ 30 º)

for 2 seconds

Voltage (20.0V)Current (12.73A) @ -87º(1.57Ω @ 87 º)

for 1 seconds

Voltage (30.0V)Current (14.08A) @ -87º(1.42Ω @ -87 º) Start timerStop timer relay output operates. Time should be less than 5 cycles

FAULT 1 = 1 s = TRIP @ 20 cy

FAULT 2 - TRIP

PRE-FAULT ---->

X

R

ZONE 1 21 ELEMENT

ZONE 2 21 ELEMENT

Figure 27: 21-Element Dual Zone Combination TestThe previous test would fail because the Fault 1 impedance falls within the Zone 2 circle

and Zone 2 will trip in 20 cycles, which is a shorter time than the Fault 1 duration.

Page 39/ Werstiuk

This method will work for all relay elements, including elements with time delays such as the time overcurrent (51) element.

Multiple timing tests were performed on electromechanical relays to ensure the relay magnets and mechanical time dials were in the proper location. If the relay timing did not match the manufacturer’s curve, a magnet or time dial setting was adjusted to bring the relay into tolerance. There are no possible adjustments to a microprocessor relay so the timing will be correct, or the relay is set wrong and it will be incorrect.

The following test plan will allow the user to test the pick-up and timing of a microprocessor based 51-element with a pick-up of 5A.


4.75 A for 5 second(5 A - 5% = 5 - (5 x 0.05) = 5 - 0.25 = 4.75 A)

10.0 A (2x nominal pickup)Start timerStop timer when relay output operates. Time should be expected result for 2xpickup +/- 5%


4.75 A for 5 seconds(5 A - 5% = 5 - (5 x 0.05) = 5 - 0.25 = 4.75 A)



4.75 A for 5 seconds(5 A - 5% = 5 - (5 x 0.05) = 5 - 0.25 = 4.75 A)


You should notice that the timing test current was equal to the multiple of current without adding 5% to compensate for test-set and relay error. The 5% error is used when comparing the time in fault…not the applied current. If there was a problem with the applied settings, the measured time delay would be significantly shorter than the expected time delay to indicate the problem.

Another problem would become evident if you were testing this relay if the 50 and 51-elements were assigned to the same trip coil. The 50-element (25A) would operate when the relay-tester tried to perform the 6x timing test (30A). The relay-tester could isolate the 51-element to another relay output to perform their 6x test without interference…or they could step back and review their procedure. Two different issues come into play when protective elements overlap.

The microprocessor relay’s 51-element does not have any possible adjustments, so is a third test really necessary to prove the characteristic curve if two other tests are successful? If the third test is required, we can change the test current of the third test. The primary reason for choosing whole numbers for 51-element tests in electro-mechanical relays is that it is easier to determine the expected result on the graph using whole numbers. Most relay-


this a practical test. Change the Fault 1 time delay to 10 cycles (Zone 1 time (5 cycles) < Fault 1 time < (Zone 2 time (20 cycles) - Zone 1 time (5 cycles)) = 5 cycles < Fault 1 time < 15 cycles = Fault 1 time = 10 cycles) (see figure 28)

Figure 28: 21-Element Dual Zone Combination Test

Adding logic testing to this step is as simple as connecting all of the relay outputs used by the relay settings and monitoring all contacts during the test. If OUT101 and OUT107 are supposed to operate when a 50 element operates, make sure both elements operate during the 50-element test. Or, you could perform the test twice using each output for timing.

When this test technique is applied to the relay in the Logic Testing example, the test plan has fewer steps and is more compre-hensive. We will use the following relay settings in our examples.

• SV1PU = 15 cycles [Breaker Failure Timer]• TRIP = 51P1T + 51N1T + 50P1 + 50N1• SV1 = (50P2 + 50N2) *(SV1T + TRIP ) [Breaker fail operate logic]• OUT101 = TRIP [Trip Breaker]• OUT102 = SV1T [Trip upstream breaker]• OUT107 = TRIP [Scada/Remote Trip Indication]• 52A = IN101• DP1 = 52A [Front Panel Display]• DP_1 = Breaker Closed• DP_2 = Breaker OpenA) Test Plan1. Perform 51P1T combination test at 2xpick-up and verify

OUT101 and OUT107 operates.

2. Perform 51P1T combination test at 4xpick-up and verify OUT101 and OUT107 operates.

3. Perform 51N1T combination test at 2xpick-up and verify OUT101 and OUT107 operates.

4. Perform 51N1T combination test at 4xpick-up and verify OUT101 and OUT107 operates.

5. Perform 50P1 combination test at 1.1xpick-up and verify OUT101 and OUT107 operates.

6. Perform 50N1 combination test at 1.1xpick-up and verify OUT101 and OUT107 operates.

7. Perform 51P1T combination test at 2xpick-up and set timer to start when OUT101 operates and stop the timer and test when OUT102 operates.

8. Perform 51N1T combination test at 2xpick-up and set timer to start when OUT101 operates and stop the timer and test when OUT102 operates.

9. Perform 51N1T combination test at 2xpick-up and set timer to start when OUT101 operates and stop the timer OUT102 oper-ates. Do not stop the test when OUT102 operates. Switch to an additional state with the neutral current set at 0.525A and veri-fy OUT102 is still in the trip state. Switch to an additional state with the neutral current set at 0.475A and verify that OUT102 has changed state and create a timer to start timing when you enter the last state and stops when the contact opens. Compare the second timer to the SV1DO setting.

10. Perform 51P1T combination test at 2xpick-up and set timer to start when OUT101 operates and stop the timer OUT102 oper-ates. Do not stop the test when OUT102 operates. Switch to an additional state with the phase current set at 0.525A and verify OUT102 is still in the trip state. Switch to an additional state with the phase current set at 0.475A and verify that OUT102 has changed state and create a timer to start timing when you enter the last state and stops when the contact opens. Compare the second timer to the SV1DO setting.

11. Check breaker status and compare to front panel display. (If breaker is open, then display should indicate open.)


This technique, when applied correctly:

• Can be faster than traditional testing techniques

• Is more efficient than traditional testing techniques

• Is more comprehensive because it tests the pick-up, timing, and logic in one step

• Provides true commissioning results because no settings are changed and tests are more realistic

• Can make maintenance testing simple if the tests are saved and re-played at maintenance intervals.

9. RELAY TEST PROCEDURES – SYSTEM TESTINGLogic testing combined with dynamic testing is a very powerful

and effective test method when applied by an experienced relay-tester who has a good understanding of the relay elements and the system the relay protection is applied to. However, there is a fatal flaw when performing relay testing based on supplied setting files…do the set-tings match the engineer’s intent? As mentioned before, a modern microprocessor relay will perform the tasks that it is instructed to per-form and cannot determine if the engineer has understood the relay’s operation or not. This problem was coined Garbage in = Garbage Out when computers were first implemented in society but we appear to

Page 43/ Werstiuk

A simple modification of the Fault 1 time to a value greater than the Zone 1 time but less than the Zone 2 time delay will make this a practical test. Change the Fault 1 time delay to 10 cycles (Zone 1 time (5 cycles) < Fault 1 time < (Zone 2 time (20 cycles) - Zone 1 time (5 cycles)) = 5 cycles < Fault 1 time < 15 cycles = Fault 1 time = 10 cycles)

Pre-Fault Fault 1 Fault 2Voltage (69.28V)Current (3.0A) @ -30º(23.09Ω @ 30 º) for 2 seconds

Voltage (20.0V)Current (12.73A) @ -87º(1.57Ω @ 87 º) for 10 cycles

Voltage (30.0V)Current (14.08A) @ -87º(1.42Ω @ -87 º) Start timerStop timer relay output operates. Time should be less than 5 cycles

FAULT 1 = 10 cy = NO TRIP

FAULT 2 - TRIP

PRE-FAULT ---->

X

R

ZONE 1 21 ELEMENT

ZONE 2 21 ELEMENT

Figure 28: 21-Element Dual Zone Combination TestAdding logic testing to this step is as simple as connecting all of the relay outputs used by

the relay settings and monitoring all contacts during the test. If OUT101 and OUT107 are supposed to operate when a 50 element operates, make sure both elements operate during the 50-element test. Or, you could perform the test twice using each output for timing.

When this test technique is applied to the relay in the Logic Testing example, the test plan has fewer steps and is more comprehensive. We will use the following relay settings in our examples.

SV1PU = 15 cycles [Breaker Failure Timer] TRIP = 51P1T + 51N1T + 50P1 + 50N1 SV1 = (50P2 + 50N2) *(SV1T + TRIP ) [Breaker fail operate logic] OUT101 = TRIP [Trip Breaker] OUT102 = SV1T [Trip upstream breaker] OUT107 = TRIP [Scada/Remote Trip Indication] 52A = IN101 DP1 = 52A [Front Panel Display] DP_1 = Breaker Closed


become far more trusting as computers became part of our daily life. Neither the relay-tester nor the relay can determine whether the pick-up setting is intended to be 0.5 instead of the 5.0 amps the engineer accidently applied unless they have the engineer’s notes or a coor-dination study or it is an obvious error. Testing a relay to its applied settings with no comparison to intent or common sense will almost al-ways create a successful test unless there are gross mechanical or set-ting and test plan errors. A relay’s mechanical problems can be more easily detected by simply applying voltage and current and perform-ing a meter test followed by exercising each digital input and output. Gross setting errors can be detected by a combination of dynamic and logic testing. But what happens when the logic is too complex to de-cipher like this real world example of a capacitor bank control circuit:

Opening the Capacitor BankSV8 =(RB15 * !LT6 + PB7 * LT5) * LT10 + RB13 * !LT10 + /SV3T

Closing the Capacitor Bank SwitchesSV9 =(RB16 * !LT6 + PB8 * LT5) * LT10 * SV10T + RB14 * !LT10 * SV10T

This logic doesn’t look that complicated until you realize that any word bit that begins with RB is logic from another device that has over 100 lines of programming. If the logic was expanded to represent just what is inside this one relay, it would look like:

Opening the Capacitor BankSV8 =(RB15 * ! (PB10 * !LT6 * (!LT5 * PB5)) + PB7 * (!LT5 * PB5)) * (!LT10 * (PB6 * LT5 + RB12 * !LT6)) + RB13 * ! (!LT10 * (PB6 * LT5 + RB12 * !LT6)) + /3P27 * !50L * 52A

Closing the Capacitor Bank SwitchesSV9 =(RB16 * ! (PB10 * !LT6 * (!LT5 * PB5)) + PB8 * (!LT5 * PB5)) * (!LT10 * (PB6 * LT5 + RB12 * !LT6)) * SV10T + RB14 * ! (!LT10 * (PB6 * LT5 + RB12 * !LT6)) * IN104

It turns out that the testing this logic was quite simple after the engineer was contacted. This logic translates into the following bullet points:

Closing the Capacitor Bank Switches1. If the capacitor switches are open and the capacitor control is

in “Manual”, close the capacitor switches when the “Close Ca-pacitor” button is pushed.

2. If the capacitor switches are open and the capacitor control is in “Auto”, the capacitor switches will close if:

• If the phase to phase voltage is below 6.84 kV.

• If there is a lagging power factor and the load is above 10 MW.

• If there is a leading power factor between 0.99 and 1.00 and the load is above 15 MW.

Opening the Capacitor Bank3. If the capacitor switches are closed and the capacitor control

is in “Manual”, open the capacitor switches when the “Open Capacitor” button is pushed.

4. If the capacitor switches are closed and the circuit breaker opens, open the capacitor switches.

5. If the capacitor switches are closed and the capacitor control is in “Auto”, the capacitor will close if:

• If the phase to phase voltage is above 7.74 kV.

• If there is a leading power factor of 0.96 or less.

A very complex logic equation was translated into simple, easy to simulate conditions and all of the settings worked perfectly. If the logic had been tested without understanding the engineer’s in-tent, it could take hours or days to reverse engineer and unless something obvious went wrong, no errors would have been de-tected because the relay would perform as programmed.

The ideal testing scenario would occur if the settings engineer created a description of operation that will allow us to test the relay based on their intent instead of their settings. An example of the setting engineer’s description of operation for our example relay could include:

1. The Zone 2 element should operate OUT101 and OUT107 at 3.4Ω @ 87º after a 20 cycle delay.

2. The Zone 1 element should operate OUT101 and OUT107 at 1.5Ω @ -87º in less than 5 cycles.

3. The 51PT element should operate at OUT101 and OUT107 at 10A in 10.4 seconds.

4. The 51PT element should operate at OUT101 and OUT107 at 20A in 7.3 seconds.

5. The 51NT element should operate at OUT101 and OUT107 at 2A in 7.4 seconds during a Phase-Ground fault.

6. The 51NT element should operate at OUT101 and OUT107 at 4A in 3.6 seconds during a Phase-Ground fault.

7. The 50PT element should operate at OUT101 and OUT107 at 25A in less than 5 cycles.

8. The 50NT element should operate at OUT101 and OUT107 at 6A in less than 5 cycles during a Phase-Ground fault.

9. OUT102 will operate if the breaker remains closed 15 cycles after a trip signal is sent. The breaker is considered closed if the phase or residual current is greater than 0.5A.

With these instructions, we can build a test plan similar to the test plans described in this section to ensure that the relay will per-form as the relay engineer intended instead of regurgitating what the relay was programmed to do. It is important that this descrip-tion be created when the engineer designs the system and the re-lay’s intended operation is fresh in their mind.


10. RELAY TEST PROCEDURES - DYNAMIC SYSTEM-MODEL TESTING

Dynamic system model based testing uses a computer program to create a mathematical model of the electrical system and create fault simulations based on the specific application. These modeled faults (or actual events recorded by a relay) are replayed through a sophisticated relay test-set to the relay and, if performed correctly, is the ultimate test to prove an entire protection system as a whole. Dynamic System Model based testing can also provide more real-ism by creating waveforms that can incorporate real system condi-tions such as DC offset, transients, or CCVT distortions as shown in the example waveform in Figure 29.

Figure 29: System Modeling Waveform

This test is typically limited to type testing or end-end testing because it requires specialized knowledge of a system, complex computer programs, advanced test equipment, and a very complex test plan with many possibilities for error.

11. CONCLUSIONMicroprocessor based relays have become increasingly com-

plex but modern test equipment with 3 or more voltage/current channels and multi-state controls provide the tools needed to per-form relay testing with greater efficiency and, more importantly, effectiveness. Relay tester should first ask the question “Why am I testing this relay?” then apply all of the tools in their tool belt to determine what combination of test techniques to use. For ex-ample, the following test plans could be performed on the same relay with different reasons for testing:

REFERENCEST. Giuliante, ATG Consulting; M. Makki, Softstuf; Jeff Taffuri, Con Edison. “New Techniques For Dynamic Relay Testing.”

Kenneth Tang, Manta Test Systems Inc. “Dynamic State & Other Advanced Testing Methods For Protection Relays Address Chang-ing Industry Needs.”

SEL-311C – Protection and Automation System – Instruction Manual. Copyright SEL 1999, 2000, 2001 Schweitzer Engineer-ing Laboratories.

SEL-300G – Protection and Automation System – Instruction Manual. Copyright SEL 1998-2001 Schweitzer Engineering Lab-oratories.

GEC Alstom T&D, Protective Relays Application Guide (Blue Book) GEC Alstom. Third Edition Reprinted March 1995.

Protective Relaying Theory and Applications (Red Book) Edited by Walter A. Elmore - “Portions of this book were originally pub-lished as ‘Applied Protective Relaying,” edited by J.L. Blackburn, Westinghouse Electric Corporation, Coral Springs, Florida, Copy-right 1982.” Copyright 1994 by ABB.

Walter A. Elmore, Protective Relaying Theory and Applications Second Edition Revised and Expanded. Copyright 2004 by ABB Power T&D Company Inc.

J. Lewis Blackburn , Protective Relaying: Principles and Applica-tions – Second Edition. Copyright 1998 Marcel Dekker, Inc.

Instruction Book – M-3310 Transformer Protection Relay. Beckwith Electric Co Inc.

Instruction Book – M-3420 Generator Protection Relay.Beckwith Electric Co Inc.

Instruction Book – M-3425 Generator Protection Relay.Beckwith Electric Co Inc.

D-60 Line Distance Relay Instruction Manual. General Electric

750/760 Feeder Management Relay Instruction Manual .GE Multilin

745 Transformer Management Relay Instruction Manual.GE Multilin

489 Generator Management Relay Instruction Manual.GE Multilin

FT Flexitest Family Switches, Covers, and Test Plugs. ABB,STATES® Type FMS, AVO INTERNATIONAL

Page 49/ Werstiuk

11. Conclusion Microprocessor based relays have become increasingly complex but modern test

equipment with 3 or more voltage/current channels and multi-state controls provide the tools needed to perform relay testing with greater efficiency and, more importantly, effectiveness. Relay tester should first ask the question “Why am I testing this relay?” then apply all of the tools in their tool belt to determine what combination of test techniques to use. For example, the following test plans could be performed on the same relay with different reasons for testing:

Acceptance Testing

1. Perform a metering check on all analog current and voltage channels2. Perform a function test of all digital inputs and outputs.3. Perform a relay self-test

Commissioning 1. Perform a metering check on all analog current and voltage channels2. Perform a function test of all digital inputs and outputs.3. Perform a relay self-test4. Test each element enabled in output logic using:

a. Steady State for pickup testing, simple dynamic testing for timing, and logic testing, OR

b. Combine Pickup, Timing Tests, and Logic Testing, ORc. System Testing, ORd. Dynamic System-Model Testing

Maintenance Testing

1. Perform a metering check on all analog current and voltage channels2. Perform a function test of all digital inputs.3. Perform a relay self test

a. Perform a trip test for each output by applying one logic test for each enabled output, OR

b. Perform Logic Testing, ORc. Combine Pickup, Timing Tests, and Logic Testing, ORd. System Testing, ORe. Dynamic System-Model Testing


ABSTRACTThe paper describes the use of Hall-effect sensors for adding

digital recording and harmonic analysis capabilities to major sub-station equipment such as electromechanical relays, circuit break-ers, and power transformers. Three case studies are presented de-scribing the use of these sensors as tools for diagnosing problems and identifying root causes of equipment failures. The first case study describes how the sensors were used in a generating plant to identify the cause of a transformer differential relay trip opera-tion. The second study describes how the sensors were used in a switching station to measure harmonic content during capacitor bank switching operations. And, the third case study describes how a potential hazard was inadvertently discovered while using the sensors to capture trip and secondary current signatures in a distribution substation.

The paper also describes the unique characteristics of the Hall-effect sensors and the process of preparing them for use in the sub-station environment. The types of enclosures used and the needed recording requirements are also discussed. The intent of the au-thors is to present the reader with a novel tool that is truly helpful for identifying problems with major substation equipment.

BACKGROUND Hall-effect sensors use small, current-to-voltage transducer that

respond to magnetic fields and are therefore useful for monitor-ing both direct and alternating currents (AC and DC). These types of transducers have seen widespread use in industrial-process and automotive applications. A typical transducer with an 2 applied magnetic field is shown in Figure 1. The transducer produces a voltage output that is proportional to the magnitude of the applied

field. The response time is in the 10 microseconds range making the transducer capable of measuring high order harmonics from 50 and 60 Hz sources. As for sensitivity, the transducer output is -2.5 to 2.5 volts. In a well shielded environment and with the current carrying conductor touching the transducer surface (as shown in Figure 1), the transducer output measures 1 millivolt for every 20 milliamps of induced current making it capable of sensing currents up to 50 amps.

With the above capabilities and with the proper enclosure, re-corder, and data formats (as described in the next section), the Hall-effect transducer is ideal for a wide range of equipment monitoring applications including, but not limited to, capturing electromechanical relay targets, monitoring DC control circuits, recording breaker trip signatures, measuring inrush currents, and monitoring current transformers (CTs).

INTRODUCTION A novel Hall-effect sensor with a non-intrusive, clothespin-like

enclosure is shown in Figure 2. The actual transducer is visible in the center of the sensor and is covered by a curved strip of mu-metal used for shielding against external magnetic fields and for amplifying internal ones. The voltage output from the transducer is provided over a shielded RJ45 cable. This type of enclosure pro-vides for simple installation on live wires in harsh environments without the necessity for removing equipment from service.

Figure 2: Hall-effect sensor with shielded clothespin-like enclosure

USING HALL-EFFECT SENSORSTO ADD DIGITAL RECORDING CAPABILITY

TO ELECTROMECHANICAL RELAYSPowerTest 2011

by Amir Makki,Softstuf, Inc; Sanjay Bose, The Consolidated Edison Company of New York; Tony Giuliante, ATG Consulting; and John Walsh, Sean Breatnach Technical Services (SBTS) LLC

Figure 1: Hall-effect transducer with applied magnetic field


With “proper recording”, the voltage outputs from the sensor become accurate representations of the currents being monitored. The term “proper recording” means, among other things, minimi-zation of measurement and timing errors induced by digitization and signal conditioning methods. Good accuracy can be achieved when recording with a resolution of 16-bits and at a sampling rate near 3 or above 2,400 Hz. An off-the-shelf recorder that provides such performance is shown in Figure 3. The recorder has 8 chan-nels for connecting sensors and samples simultaneously on all channels (each channel has its own controller and the controllers are synchronized with a skew-factor under 1 degree).

Figure 3: Off-the-shelf recorder with one clothespin sensor shown

In an effort to validate the use of Hall-effect sensors for moni-toring substation equipment, a benchmark test was conducted. The test utilized a power system simulator to play back an “A” phase to ground digital fault record into a numerical relay. The Hall-effect sensors were mounted on the phase currents of the power system simulator as depicted in Figure 4. The digital fault record was cap-tured by a quality digital fault recorder (DFR) and the measured fault magnitude was 60 amps which was also ideal for challenging the 50 amp range of the Hall-effect transducer being used (other models support higher ranges).

Figure 4: Hall-effect sensors benchmark test (DFR record play back)

After the test, the resulting fault records from the relay and Hall-effect sensors were compared with the original DFR record. The results were remarkable in that the fault records were almost ex-act replicas even though they were captured by different instru-ments having different resolutions and sampling rates. Figure 5 shows the captured “A” phase waveforms from each instrument. The Hall-effect Digital Fault Record Power System Simulator Numerical Relay a b c Hall-effect Recorder 4 waveform clearly shows that the sensor flat-lines at the 50 amps range which is a desirable outcome because the actual peak can be calculated using sinusoidal interpolation (it does not collapse as in the case with CT saturation).

Figure 5: DFR, numerical relay, and Hall-effect waveforms (“A” phase)

As for data formats, the term “proper recording” also means full utilization of IEEE Standards C37.111-1999 (COMTRADE) and C37.232-2007 (Naming Convention for Time Sequenced Data Files) for all of the needed data storage, management, analysis, and exchange requirements. This entire experiment with Hall-effect sensors was designed around these standards. The level of simplicity and commonality in analysis as shown throughout the next sections would not have been possible without the full utiliza-tion of these standards.

The combination of Hall-effect sensors with proper recording and IEEE data formats provides a robust and friendly system for monitoring major equipment in the substation. The system can be mounted inside a relay panel for real-time monitoring or can be used with a laptop as a portable diagnostic instrument for trou-bleshooting control circuits, CTs, motor currents, and so forth. A number of case studies were conducted with the Hall-effect sensors and the results are very telling. Three of these cases are outlined in the next sections. The first case investigates the root cause of a transformer differential relay trip operation in a generating plant, the second identifies harmonic content during capacitor bank op-erations in a switching station, and the third case was conducted to monitor and time breaker operations in a distribution substation.

3

or above 2,400 Hz. An off-the-shelf recorder that provides such performance isshown in Figure-3. The recorder has 8 channels for connecting sensors andsamples simultaneously on all channels (each channel has its own controller andthe controllers are synchronized with a skew-factor under 1 degree).

Figure-3, Off-the-shelf recorder with one clothespin sensor shown

In an effort to validate the use of Hall-effect sensors for monitoring substationequipment, a benchmark test was conducted. The test utilized a power systemsimulator to play back an “A” phase to ground digital fault record into a numerical relay. The Hall#effect sensors were mounted on the phase currents of the power system simulator as depicted in Figure#4. The digital fault record was captured by a quality digital fault recorder (DFR) and the measured fault magnitude was 60amps which was also ideal for challenging the 50 amp range of the Hall#effect transducer being used (other models support higher ranges).

Figure-4, Hall-effect sensors benchmark test (DFR record play back)

After the test, the resulting fault records from the relay and Hall-effect sensorswere compared with the original DFR record. The results were remarkable in thatthe fault records were almost exact replicas even though they were captured bydifferent instruments having different resolutions and sampling rates. Figure-5shows the captured “A” phase waveforms from each instrument. The Hall#effect

Digital FaultRecord

Power SystemSimulator

NumericalRelay

ab

c

Hall-effectRecorder

4

waveform clearly shows that the sensor flat-lines at the 50 amps range which is adesirable outcome because the actual peak can be calculated using sinusoidalinterpolation (it does not collapse as in the case with CT saturation).

Figure-5, DFR, numerical relay, and Hall-effect waveforms (“A” phase)

As for data formats, the term “proper recording” also means full utilization of IEEE Standards C37.111#1999 (COMTRADE) and C37.232#2007 (Naming Convention for Time Sequenced Data Files) for all of the needed data storage, management, analysis, and exchange requirements. This entire experiment withHall#effect sensors was designed around these standards. The level of simplicity and commonality in analysis as shown throughout the next sections would not have been possible without the full utilization of these standards.

The combination of Hall#effect sensors with proper recording and IEEE data formats provides a robust and friendly system for monitoring major equipment in the substation. The system can be mounted inside a relay panel for real#time monitoring or can be used with a laptop as a portable diagnostic instrument for troubleshooting control circuits, CTs, motor currents, and so forth. A number of case studies were conducted with the Hall#effect sensors and the results are very telling. Three of these cases are outlined in the next sections. The first case investigates the root cause of a transformer differential relay trip operation in a generating plant, the second identifies harmonic content during capacitor bank operations in a switching station, and the third case was conducted to monitor and time breaker operations in a distribution substation.

CASE I – TRANSFORMER INRUSH

At a generating plant, an attempt was made to energize a generator auxiliarypower transformer. When the transformer was energized, it tripped on “A” phase

Relay

DFR

Hall-effect

50 Amps


CASE I – TRANSFORMER INRUSH At a generating plant, an attempt was made to energize a gen-

erator auxiliary power transformer. When the transformer was energized, it tripped on “A” phase Relay DFR Hall-effect 50 Amps 5 differential relay. After this event, the differential re-lay was tested and the results were satisfactory. The connected buses were tested to determine the area of the fault and they too tested satisfactorily. The protective relay wiring and CTs were also tested with satisfactory results. After a considerable amount of testing and analysis, the root cause of the trip could not be determined.

At this point, the differential relay was replaced, a number of Hall-effect sensors were connected to the relay circuits (at the panel), and the transformer was again energized. The transformer tripped again, but this time the Hall-effect sensors captured the waveform data shown in Figure 6. The waveform data shows typi-cal inrush current signature for about 3 cycles then degrades over the next 9 cycles during which the differential relay operates to trip the circuit breaker.

Figure 6: 345 kV transformer differential relay operation

Figure 7: Typical 345 kV transformer inrush current signature

Further analysis of the data reveals that the root cause was CT saturation due to remanent flux. The saturated CT failed to pro-vide the 2nd harmonic content required by the differential relay for restraint, so the relay operated. Corrective actions were then taken and the transformer was restored to service. For comparison, Figure 7 shows a typical inrush current waveform from a similar Prefault Data Typical Inrush Typical Inrush Low 2nd Harmonic Breaker Opens Prefault Data 6 transformer during energization. The shown waveform contains strong 2nd harmonic content (above 30% of fundamental) and the magnitude slowly decays towards zero (which could take over 10 seconds to complete).

CASE II – CAPACITOR BANK RINGING Power systems designed to function at 50 and 60 Hz are prone to

experiencing failures when subjected to voltages and currents that contain high harmonic frequency content. Very often, the opera-tion of electrical equipment may seem normal, but under a certain combination of conditions, the impact of harmonics is enhanced and with damaging results. The only means of determining the magnitude and type of harmonics is through “careful monitoring”. Once sufficient data are collected and analyzed then the proper mitigation strategies can be defined and implemented. Here is a good example of careful monitoring:

At a 345 kV switching station, capacitor bank switching was causing damage to solid state protective relay equipment and the capacitor banks themselves were also experiencing failures. After considerable testing and analysis, the root cause of the damage and failures could not be determined. At this point, a number of Hall-effect sensors were installed at the capacitor bank breaker CTs to capture the harmonic content at the time of switching (at both cut-in and cut-out times). A few days later, the switching sta-tion was visited and the captured data was retrieved for inspection and analysis.

Figure 8: Capacitor bank cut-in waveform at 32 samples per cycle

The retrieved data revealed the presence, during cut-in, of high harmonic spikes with large current magnitudes as shown in Figure 8. This type of phenomenon is called “capacitor bank ringing” and a good mitigation strategy is to install zerocrossing detectors and use them to cut-in the capacitor bank when the individual phase volt-ages are at zero (this prevents the occurrence of large discontinui-ties in current magnitudes). As for “careful monitoring”, the shown spikes are typical signatures of under-sampling. Using a Fourier filter, the calculated frequency of the spikes is 660 Hz (the 11th har-monic). Knowing that the installed sensors were being sampled at 32 samples per cycle, and seeing the asymmetric, saw-tooth like signature of these spikes, it is clear that the actual frequency should be a number above the 16th harmonic. Figure 9 shows the same cut-in event but with a sampling rate of 320 samples per cycle. Clearly, the capacitor banks were not ringing at the 11th harmonic, they were ringing at the 21st harmonic. Careful monitoring requires a solid un-derstanding of the events being observed and of the nature of the waveform signatures being captured.

5

differential relay. After this event, the differential relay was tested and the resultswere satisfactory. The connected buses were tested to determine the area of thefault and they too tested satisfactorily. The protective relay wiring and CTs werealso tested with satisfactory results. After a considerable amount of testing andanalysis, the root cause of the trip could not be determined.

At this point, the differential relay was replaced, a number of Hall-effect sensorswere connected to the relay circuits (at the panel), and the transformer was againenergized. The transformer tripped again, but this time the Hall-effect sensorscaptured the waveform data shown in Figure-6. The waveform data shows typicalinrush current signature for about 3 cycles then degrades over the next 9 cyclesduring which the differential relay operates to trip the circuit breaker.

Figure-6, 345 kV transformer differential relay operation

Figure-7, Typical 345 kV transformer inrush current signature

Further analysis of the data reveals that the root cause was CT saturation due toremanent flux. The saturated CT failed to provide the 2nd harmonic contentrequired by the differential relay for restraint, so the relay operated. Correctiveactions were then taken and the transformer was restored to service. Forcomparison, Figure-7 shows a typical inrush current waveform from a similar

PrefaultData

TypicalInrush

TypicalInrush

Low 2nd

HarmonicBreakerOpens

PrefaultData

5

differential relay. After this event, the differential relay was tested and the resultswere satisfactory. The connected buses were tested to determine the area of thefault and they too tested satisfactorily. The protective relay wiring and CTs werealso tested with satisfactory results. After a considerable amount of testing andanalysis, the root cause of the trip could not be determined.

At this point, the differential relay was replaced, a number of Hall-effect sensorswere connected to the relay circuits (at the panel), and the transformer was againenergized. The transformer tripped again, but this time the Hall-effect sensorscaptured the waveform data shown in Figure-6. The waveform data shows typicalinrush current signature for about 3 cycles then degrades over the next 9 cyclesduring which the differential relay operates to trip the circuit breaker.

Figure-6, 345 kV transformer differential relay operation

Figure-7, Typical 345 kV transformer inrush current signature

Further analysis of the data reveals that the root cause was CT saturation due toremanent flux. The saturated CT failed to provide the 2nd harmonic contentrequired by the differential relay for restraint, so the relay operated. Correctiveactions were then taken and the transformer was restored to service. Forcomparison, Figure-7 shows a typical inrush current waveform from a similar

PrefaultData

TypicalInrush

TypicalInrush

Low 2nd

HarmonicBreakerOpens

PrefaultData

6

transformer during energization. The shown waveform contains strong 2nd

harmonic content (above 30% of fundamental) and the magnitude slowly decaystowards zero (which could take over 10 seconds to complete).

CASE II – CAPACITOR BANK RINGING

Power systems designed to function at 50 and 60 Hz are prone to experiencingfailures when subjected to voltages and currents that contain high harmonicfrequency content. Very often, the operation of electrical equipment may seemnormal, but under a certain combination of conditions, the impact of harmonics isenhanced and with damaging results. The only means of determining themagnitude and type of harmonics is through “careful monitoring”. Once sufficient data are collected and analyzed then the proper mitigation strategies can be defined and implemented. Here is a good example of careful monitoring:

At a 345 kV switching station, capacitor bank switching was causing damage to solid state protective relay equipment and the capacitor banks themselves were also experiencing failures. After considerable testing and analysis, the root cause of the damage and failures could not be determined. At this point, a number of Hall-effect sensors were installed at the capacitor bank breaker CTs to capture the harmonic content at the time of switching (at both cut-in and cut-out times). A few days later, the switching station was visited and the captured data was retrieved for inspection and analysis.

Figure-8, Capacitor bank cut-in waveform at 32 samples per cycle

The retrieved data revealed the presence, during cut-in, of high harmonic spikeswith large current magnitudes as shown in Figure-8. This type of phenomenon iscalled “capacitor bank ringing” and a good mitigation strategy is to install zero-crossing detectors and use them to cut-in the capacitor bank when the individual phase voltages are at zero (this prevents the occurrence of large discontinuities

a

b

c

n

Calculated Harmonic = 11th

Spikes


Figure 9: Capacitor bank cut-in waveform at 320 samples per cycleCASE III – BREAKER TIMING

At a distribution substation, the 27 kV capacitor bank vacuum tube breakers were experiencing failures. The breakers are typi-cally operated twice a day for controlling voltage and reactive power and over time they accumulate a significant number of op-erations. To help troubleshoot the failures, a number of Hall-effect sensors were installed to monitor both the trip coils of the breakers (DC captures) and the CTs of the capacitor bank feeder (AC cap-tures). A few days later, the distribution substation was visited and the waveform data was retrieved. The DC captures indicated that the breaker timing was within specification but the AC captures showed considerable transients immediately after the breaker opened as shown in Figure 10.

Further analysis of the transients revealed that the vacuum tube bottles were breaking down. Such breakdowns pose a danger of damage to surrounding equipment and more importantly they pose a safety concern for operations personnel. Racking out a failed vacuum tube circuit breaker from its cell is a serious arc-flash-haz-ard to operations personnel. Clearly, such potentially life threaten-ing events can be avoided with proper monitoring.

Figure 10: Vacuum bottle breakdown (27 kV capacitor bank breakers)CONCLUSIONS

The paper presented a number of successful case studies us-ing Hall-effect sensors for monitoring major substation equip-ment. The studies highlighted the benefits of using the Hall-effect sensors to diagnose problems and identify root causes of equip-ment failures. The studies were conduct at a generating plant, at a switching station, and at a distribution substation. The capability

of the sensors to accurately monitor DC control circuits, as well as high AC harmonics from secondary phase currents, was also suc-cessfully demonstrated.

The sensors are non-intrusive and inexpensive. They can be de-ployed in a timely manner and without having to remove equip-ment from service. With proper enclosures and recording instru-ments and with careful selection of resolution and sampling rates, Hall-effect sensors are useful for a wide range of monitoring appli-cations including capturing relay targets, timing circuit breakers, measuring inrush currents, and diagnosing DC control circuits. Other exciting applications include embedding Hall-effect sensors directly into equipment designs and especially in electromechani-cal relay designs.

Using Hall-effect sensors to monitor major substation equip-ment is also in line with today’s “Smart Grid” initiatives. Such monitoring provides enhancements in maintenance and engi-neering that are extra ordinary. Imagine the benefits of learn-ing about equipment failures upon occurrence, or even bet-ter, imagine the benefits of catching potential failures before they occur. Access to such knowledge helps utilities increase grid reliability, reduce maintenance costs, a b c Relay Contact Breaker Time = 2.25 cycles “Arcing” 9 restore lines faster, and extend the service life of major equipment. Clearly, in conclu-sion, using Hall-effect sensors helps make our “legacy” substa-tions similar to our newest substations.

ACKNOWLEDGMENTS The Hall-effect sensor is a patented design by Con Edison, Soft-

stuf, and ATG Consulting. The first case study was a collabora-tive effort including personnel from Softstuf, Exelon, and ABB. The second and third case studies were efforts including personnel from Softstuf and Con Edison. The clothespin sensors and record-ers are manufactured by TIS Labs, LLC. The waveforms shown are screen captures from the WavewinTM software (trademark by Softstuf, 1991-current). The authors extend their thanks to all per-sons involved in the three case studies presented in this paper. Es-pecial thanks are extended to:

• Maria Makki (software development),

• George & Hamid Semati (hardware development),

• Mark Taylor (product development),

• Frank Rothweiler (product testing),

• Shi Jiang (case studies at Con Edison),

• Dwight Smith (case studies at Exelon), and

• Bob Wilson (publication planning).

7

in current magnitudes). As for “careful monitoring”, the shown spikes are typical signatures of under#sampling. Using a Fourier filter, the calculated frequency of the spikes is 660 Hz (the 11th harmonic). Knowing that the installed sensors werebeing sampled at 32 samples per cycle, and seeing the asymmetric, saw-toothlike signature of these spikes, it is clear that the actual frequency should be anumber above the 16th harmonic. Figure-9 shows the same cut-in event but witha sampling rate of 320 samples per cycle. Clearly, the capacitor banks were notringing at the 11th harmonic, they were ringing at the 21st harmonic. Carefulmonitoring requires a solid understanding of the events being observed and ofthe nature of the waveform signatures being captured.

Figure-9, Capacitor bank cut-in waveform at 320 samples per cycle

CASE III – BREAKER TIMING

At a distribution substation, the 27 kV capacitor bank vacuum tube breakers wereexperiencing failures. The breakers are typically operated twice a day forcontrolling voltage and reactive power and over time they accumulate asignificant number of operations. To help troubleshoot the failures, a number ofHall-effect sensors were installed to monitor both the trip coils of the breakers(DC captures) and the CTs of the capacitor bank feeder (AC captures). A fewdays later, the distribution substation was visited and the waveform data wasretrieved. The DC captures indicated that the breaker timing was withinspecification but the AC captures showed considerable transients immediatelyafter the breaker opened as shown in Figure-10.

Further analysis of the transients revealed that the vacuum tube bottles werebreaking down. Such breakdowns pose a danger of damage to surroundingequipment and more importantly they pose a safety concern for operationspersonnel. Racking out a failed vacuum tube circuit breaker from its cell is a

a

b

c

n

Calculated Harmonic = 21st

Ringing

8

serious arc-flash-hazard to operations personnel. Clearly, such potentially lifethreatening events can be avoided with proper monitoring.

Figure-10, Vacuum bottle breakdown (27 kV capacitor bank breakers)

CONCLUSIONS

The paper presented a number of successful case studies using Hall-effectsensors for monitoring major substation equipment. The studies highlighted thebenefits of using the Hall-effect sensors to diagnose problems and identify rootcauses of equipment failures. The studies were conduct at a generating plant, ata switching station, and at a distribution substation. The capability of the sensorsto accurately monitor DC control circuits, as well as high AC harmonics fromsecondary phase currents, was also successfully demonstrated.

The sensors are non-intrusive and inexpensive. They can be deployed in a timelymanner and without having to remove equipment from service. With properenclosures and recording instruments and with careful selection of resolution andsampling rates, Hall-effect sensors are useful for a wide range of monitoringapplications including capturing relay targets, timing circuit breakers, measuringinrush currents, and diagnosing DC control circuits. Other exciting applicationsinclude embedding Hall-effect sensors directly into equipment designs andespecially in electromechanical relay designs.

Using Hall-effect sensors to monitor major substation equipment is also in linewith today’s “Smart Grid” initiatives. Such monitoring provides enhancements in maintenance and engineering that are extra ordinary. Imagine the benefits of learning about equipment failures upon occurrence, or even better, imagine the benefits of catching potential failures before they occur. Access to such knowledge helps utilities increase grid reliability, reduce maintenance costs,

a

b

c

RelayContact Breaker Time = 2.25 cycles

“Arcing”


REFERENCES The following listing of books, articles, and standards is provided

as a source for additional information:

[1] IEEE Std C37.111-1991, Standard Format for Transient Data Exchange

[2] IEEE Std C37.111-1999, Standard Format for Transient Data Exchange

[3] IEEE Std C37.118-2005, Standard for Synchrophasors for Power Systems

[4] IEEE Std C37.232-2007, Recommended Practice for Naming TSD files

[5] S, Bose et al, “US Patent Application 12/249,547,” Current Measuring Device

[6] S, Bose et al, “US Patent Application 12/505,947,” Relay Be-havior Patterns

[7] Blackburn and Domin, “Protective Relaying,” Third Edition; CRC Press, 2007 10

Amir Makki is the Chairman and Co-founder of Softstuf, Inc (1991-current). His professional contributions include over 50 publications and patents. He holds BS and MS degrees in Electri-cal Engineering from Tennessee Tech University and pursued his Ph.D. studies in Software Engineering at Temple University. Amir is a senior member of IEEE Standards Association and is an active member of the Protection Systems Relay Committee (PSRC). He is the past Chairman of the H8 working Group which produced IEEE Std. C37.232 (Recommended Practice for Naming Time Sequence Data Files), and currently serves as Chairman of the Cyber Secu-rity Task Force for Protection Related Data Files.

Sanjay Bose is employed by the Consolidated Edison Company of New York (1986-current). He is currently the Vice President for Substation Operations. In this capacity, he is responsible for the overall operation, maintenance, repair and upgrade of all Transmission and Distribution Substations and of all Public Util-ity Regulating Stations. Prior to this, he held various positions at the Company including General Manager for Protective Systems Testing, Senior System Operator at the Energy Control Center, and Design Engineer supporting substation construction activi-ties. Sanjay holds a BS degree in Power Electrical Engineering from Birla Institute of Technology, Ranchi, India (1985), and is an active member of EPRI and the IEEE.

Tony Giuliante is the President and Founder of ATG Consulting. Prior to forming his company in 1995, Tony was executive Vice President of GEC ALSTHOM T&D Inc. - Protection and Con-trol 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 relay-ing education and their analysis in operational environments.” He has authored over 50 technical papers and is a frequent lecturer

on all aspects of protective relaying, including electromechanical, solid state and digital based equipment. Tony is a past Chairman of the IEEE Power System Relaying Committee 1993-1994, and past Chairman of the Relay Practices Subcommittee. He has de-grees of BSEE and MSEE from Drexel University (1967 & 1969).

John Walsh is the President and Founder of Sean Breatnach Tech-nical Services (SBTS) LLC, a consulting firm that provides solu-tions in managing legacy infrastructure to utility companies and utility contractors. He has degrees in Electronic Technology and Industrial Management from The College of Staten Island, New York. He has worked as a Technician, Supervisor and Manager in electrical substation commissioning, relay protection testing and maintenance for 35 years. He consults regularly with companies on substation automaton, product development, substation testing protocols and project management. Prior to his current position, he managed a Protective Systems Testing team at The Consoli-dated Edison Company of New York.


Over the years, the methodology for testing distance relays has been slowly progressing from steady state to simulating real sys-tem conditions. Steady-state testing is a tried and true method for distance relays, but it does not address the system’s source imped-ance. The source impedance of the system influences the dynamic expansion of the mho characteristic and changes the apparent reach of the relay which is not observed in steady-state testing. By taking into account the system’s source impedance, the true behavior of the relay can be observed.

Testing distance relays is often seen as a difficult procedure. One of the difficulties for test technicians is choosing the correct method of testing distance elements. However, definition of cor-rect is subjective and can change depending on who is speaking. Some may insist that a constant-voltage method is the way to go, while others will argue that a constant current is best. Still others may suggest calculating the sequence components of the expect-ed fault and deriving the secondary test currents so that constant source impedance is kept. All three methods are in wide general use throughout the electrical industry. This paper will discuss those methods and outline a simplified method of testing using the sequence components.

When testing distance relays, one should have a general under-standing of the power system the relays are protecting. For this article, the power system discussed is modeled in Figure 1.

Figure 1: Power System Model

Unlike the systems considered in most papers on this topic, which use two sources, the system modeled here uses a single source. Using a single-source model helps to simplify the calcula-tions required to determine the secondary voltages and currents from the test equipment. This can be advantageous when testing in the field where different system scenarios are encountered. The system shown in Figure 1 has the single source feeding a transmis-sion line that terminates at a bus. The distance relay shown is the

zone 1 protection for the line. The relay can be either electrome-chanical or microprocessor based.

In North America, it is common for the phase protection of transmission line relays to utilize the mho element. Figure 2 shows the mho element with typical test points. The vertical axis is the reactive impedance, and the horizontal axis is the resistive imped-ance. The line angle of the relay is represented by θL, which is measured in degrees, while the reach of the relay is represented by the length, ZLMag, and is measured in ohms.

Figure 2: Mho Protective Element in a Distance RelayCOMMON TEST PROCEDURES

As stated earlier, there are various methods that can be used to test the distance element. They are referred to as constant voltage, constant current, and constant impedance methods. The reach set-ting of the distance element determines the method most likely to be used. For long lines that have a moderate to high magnitude of impedance, a constant voltage method may be used to determine the trip points. This method involves keeping the fault voltage constant and ramping the magnitude of current so that the ratio of voltage to current is equal to the impedance pickup point. In contrast, the constant current method is accomplished by keeping the fault level of current constant while ramping the voltage. The voltage is ramped until the ratio of voltage to current is equal to the impedance pickup point. The constant current method is generally used for short lines.

Determining what constitutes a short line or a moderate level of impedance is often a subjective decision. The determining factor is the relay test equipment being used. As much as electrical tech-

AN IMPROVED CONSTANT SOURCEIMPEDANCE TESTING METHOD FOR MHO

DISTANCE RELAYSNETA World, Spring 2012 Issue

by Jason Buneo and Rene Aguilar, Megger


nology has progressed in the last one hundred years, it is still not possible to manufacture a test set that can output infinite voltage and infinite current and be suitable for every single test scenario. Consequently, it is usually the output current limit of the test set that determines the method of test for distance relays. If output current of the test set is the determining factor, a short line would be one in which the constant voltage method requires the test set to exceed its output current capability. The constant current method could then be used to overcome this shortcoming. By keeping the current at a fault level that can be handled by the relay test equip-ment, the user can ramp the voltage until the desired impedance is seen by the relay.

These two methods have been the most common methods of testing distance relays for many years. The test procedures have been developed mainly from using homespun components such as variable autotransformers and phase shifters to produce appropri-ate voltages and currents. After this homespun approach, came the first generation of test sets, but these were little more than discrete homespun components in a single box. Later, the test equipment became more and more sophisticated, but the test methods did not evolve. In the early days, it took significant effort to obtain cur-rent lagging voltage by 60 degrees. Today, it is not a problem to obtain current lagging voltage by 80.21 degrees. Just as test tech-nicians used the old equipment to approximate fault values to the best of its capability, they should continue to use this approach for modern equipment. When early methods of testing are used with modern test equipment, they are commonly referred to as steady-state testing whereby one quantity remains steady while the other is varied. With today’s modern equipment, a new method of test-ing known as dynamic testing has been developed. This allows both the voltage and current to be varied during a test. In order to understand why this is significant, one must first understand what a relay is actually seeing when a steady-state test is performed.

When steady-state testing is performed on a relay, either current or voltage is varied. By changing only one quantity, the character-istic of the mho circle will change and expand. This expansion is caused in part by the equivalent source impedance behind the pro-tective relay. By varying either voltage or current but not both, the test technician allows the source impedance to change for each test point on the characteristic curve. Figure 3 shows the steady-state characteristic as the dotted line. During testing by a steady-state method, the actual characteristic varies and is represented by the overlapping circles. Note how the trip points of the overlapping circles intersect with the steadystate characteristic. This usually happens around the line angle of the relay. If the test points are far from the line angle, then the overlapping circles and the steady-state characteristic trip points will no longer intersect. It is at this point that the test technician may notice overreaching of the relay when using steady-state test methods. In order to test the true char-acteristic of the relay, dynamic methods must be employed.

Figure 3: Dynamic Expansion of mho Characteristic

DYNAMIC TESTINGIn recent years, a method of testing has been develped that tries

to more accurately represent true system conditions while testing. This is referred to as dynamic testing. One of the characteristics of dynamic testing is that the equivalent source impedance seen by the relay remains constant. In other words, the mho circle does not change continuously when dynamic test methods are used. Mod-ern relay test equipment has made dynamic testing possible. One of the most popular methods of testing distance relays is utilizing COMTRADE files captured by digital fault recorders or by the protective relays themselves. These files allow the test technician to play back actual faults that have occurred in the field and verify that the relay operates correctly. By default, these files will contain a constant source impedance as well as other parameters, such as dc offset, which will affect the behavior of the relay. The down-side to this method is that a fault must occur before the file can be captured and played back.

Simulation software offers an alternative to waiting for faults to occur. Depending on the software package, very detailed models of transmission systems are possible and various fault conditions can be simulated and exported to COMTRADE files. The software simulation packages can also include effects from parallel lines, series compensated lines, and multisource line models, among others. The test scenarios possible with many of the software pro-grams are limitless. This is ideal for commissioning before the re-lays are placed in service, as the test technician can be sure that the different scenarios will lead to correct relay operation. Unfor-tunately, after commissioning, most testing crews do not have a limitless amount of time available to test the relays once they are in service. The number of tests performed is therefore reduced, but they still need to demonstrate that the relay will operate correctly if a fault occurs.

In this situation, the relay needs to be tested in a relatively short time. Instead of attempting to test the endless scenarios, many companies opt to scale down testing to include only the most likely situations. These situations reflect the types of faults that are most likely to occur. Examples are a three-phase bolted fault, a line-to-line fault, a lineto- ground fault, and a double line-to-ground fault.


By calculating these faults using their sequence components, the test technician can produce fairly accurate testing scenarios. How-ever, it should be noted that this method is not as accurate as a full system simulation.

SEQUENCE COMPONENTSNo discussion of fault conditions would be complete without

a brief consideration of symmetrical components. Symmetrical components were first described in a paper by Dr. C. L. Fortes-cue entitled, Method of Symmetrical Coordinates Applied to the Solution of Polyphase Networks. Dr. Fortescue stated that unbal-anced three-phase voltages or currents could be transformed into three sets of balanced three-phase components. To achieve this, the symmetrical components method uses a variable called the “a” operator. This variable is a vector with a magnitude of 1 and an angle of 120°. In a three-phase system, the “a” operator is applied to the reference voltage or current. By applying the “a” operator to the reference voltage and current phases, they are each rotated counter clockwise by 120°. To get a 240° displacement, the “a” operator is squared. This is shown in Figure 4 below where the B-phase and C-phase voltages are expressed in terms of A-phase voltage. This same procedure also applies to current.

Figure 4: Application of the “a” Operator

After each of the phases is expressed in terms of a common refer-ence, they can be broken down further into their actual sequence components. Each of the three phases will contain three components called positive, negative, and zero sequence as shown in Figure 5.

Figure 5: Positive, Negative, and Zero Sequence Components

Mathematically, the sequence components can be broken down into the following equations:

Eq. 1 VA = VA1 + VA2 + VA0

Eq. 2 VB = VB1 + VB2 + VB0 = a2∙VA1 + a∙VA2 + VA0

Eq. 3 VC = VC1 + VC2 + VC0 = a∙VA1 + a2∙VA2 + VA0

To put these equations in perspective, one must consider what each of the sequence components tells us about the condition of the power system. The power system from Figure 1 can be bro-ken down into an equivalent circuit diagram where each of the sequence components is represented in a way that depends on the conditions the system is experiencing. Figure 6 illustrates the se-quence component network under normal operating conditions.

Figure 6: Equivalent Sequence Component Circuit of Balanced Power System

From Figure 6, the closed path of the top circuit shows that only the positive sequence components are present during normal bal-anced operation. The circuits for negative and zero sequence are left open because there are no imbalances present. However, it should be noted that, in reality, there will be small values of nega-tive and zero sequence components even during normal condi-tions. For the purposes of this paper, the system is being treated as an ideal system. With a three-phase bolted fault, only positive sequence values will be present since the system is still balanced.

When the power system experiences imbalance, such as a phase-to-phase fault, the equivalent circuits change. The negative sequence component is introduced and needs to be accounted for. With a phase-to-phase fault, the fault currents of the two faulted phases will be flowing into one another. This configuration will also make their sequence components equal and opposite in direc-tion. Figure 7 shows the inclusion of the negative sequence net-work due to a phase-to-phase fault.


Figure 7: Equivalent Sequence Component Circuit of a Phase-to-Phase Fault

Finally, when the power system experiences an imbalance due to a line-to-ground fault, the zero sequence component is also taken into account. In a single line-to-ground fault, there is an un-balance between the phases, and a current path to ground. The representation of this type of fault is shown in Figure 8. Unlike the phaseto- phase fault, all sequence component current values flow in the same direction.

Figure 8: Equivalent Sequence Component Circuit of a Ground FaultTYING IT ALL TOGETHER

After the sequence components have been defined for the system, the secondary test currents can be calculated. A line-to-ground fault will be calculated so that positive, negative, and zero sequence com-ponents will be present. We can begin by determining the source im-pedance of the system. If the source impedance is unknown, then an educated guess can be made to simulate different types of systems. It is easiest to determine the source impedance in terms of its ratio to line impedance because this is a quantity that will probably be known. The smaller the percentage of source impedance to line impedance, the stronger the system and the less the expansion. The opposite is true for a weak system. The larger the percentage of source impedance to line impedance, the greater the expansion. By controlling the magnitude of the source impedance, one controls how much expansion the mho

circle will experience in the relay. Figure 9 illustrates how the expan-sion of the mho circle is tied to the source impedance. The red circle represents the dynamic expansion. A weak system usually has a large source impedance, so its mho expansion is big. A strong system has a small source impedance. Its mho expansion is smaller and approaches the size of the steady-state characteristic.

Figure 9: Influence of Source Impedance on mho Expansion

To begin calculating the secondary fault values, one must define the parameters needed. To illustrate this example, the fault values for a zone 1 characteristic are shown in Table 1.

In Table 1, the only quantities arbitrarily chosen are the source impedance, ZS , and the source angle, ZsAng , other quantities should be available from the relay settings. The source impedance, Zs, is determined by using 1.6 percent of the positive sequence line im-pedance. The source impedance angle was chosen to be equivalent to the positive sequence line impedance so that the system would remain homogeneous. By using a homogeneous system, one sim-plifies the calculation of the mho expansion. Figure 10 illustrates how the expansion of the mho circle can be calculated.

Figure 10: Mho Circle Expansion

74 • SPRING 2011 NETAWORLD • 75AN IMPROVED CONSTANT SOURCE IMPEDANCE TESTING METHOD FOR MHO DISTANCE RELAYS AN IMPROVED CONSTANT SOURCE IMPEDANCE TESTING METHOD FOR MHO DISTANCE RELAYS

When the power system experiences imbalance, such as a phase-to-phase fault, the equivalent circuits change. The negative sequence component is introduced and needs to be accounted for. With a phase-to-phase fault, the fault currents of the two faulted phases will be flowing into one another. This configuration will also make their sequence components equal and opposite in direction. Figure 7 shows the inclusion of the negative sequence network due to a phase-to-phase fault.

Finally, when the power system experiences an imbalance due to a line-to-ground fault, the zero sequence component is also taken into account. In a single line-to-ground fault, there is an unbalance between the phases, and a current path to ground. The representation of this type of fault is shown in Figure 8. Unlike the phase-to-phase fault, all sequence component current values flow in the same direction.

Tying iT all TogeTherAfter the sequence components have been defined for the system, the secondary test currents can be calculated. A line-to-ground fault will be calculated so that positive, negative, and zero sequence components will be present. We can begin by determining the source impedance of the system. If the source impedance is unknown, then an educated guess can be made to simulate different types of systems. It is easiest to determine the source impedance in terms of its ratio to line impedance because this is a quantity that will probably be known. The smaller the percentage of source impedance to line impedance, the stronger the system and the less the expansion. The opposite is true for a weak system. The larger the percentage of source impedance to line impedance, the greater the expansion. By controlling the magnitude of the source impedance, one controls how much expansion the mho circle will experience in the relay. Figure 9 illustrates how the expansion of the mho circle is tied to the source impedance. The red circle represents the dynamic expansion. A weak system usually has a large source impedance, so its mho expansion is big. A strong system has a small source impedance. Its mho expansion is smaller and approaches the size of the steady-state characteristic.

Figure 9. Influence of Source Impedance on mho Expansion


In Table 1, the only quantities arbitrarily chosen are the source impedance, ZS , and the source angle, ZsAng , other quantities should be available from the relay settings. The source impedance, Zs, is determined by using 1.6 percent of the positive sequence line impedance. The source impedance angle was chosen to be equivalent to the positive sequence line impedance so that the system would remain homogeneous. By using a homogeneous system, one simplifies the calculation of the mho expansion. Figure 10 illustrates how the expansion of the mho circle can be calculated.


Using geometric principles, one can determine the dynamic reach of the mho circle using the settings of the protective relay. The zone 1 reach, ZLMag, and source impedance, ZS, are plotted as two vector quantities. When the source impedance angle, ZSang, and the line impedance angle, ZLang, equal one another, a straight line is formed through the center of the mho circle. Because the expansion is due to a line-to-ground fault, the ground impedance also needs to be considered. There will be a change in the apparent reach and apparent line angle as seen by the relay. The k0 factor assists in calculating how the ground impedance will affect the reach and angle of the apparent impedance seen by the relay. The zero sequence magnitude and angle compensation are expressed in equations 4 and 5.

The zero sequence magnitude and angle compensation are then applied to the original line impedance and angle. The source impedance magnitude is chosen as a percentage of the line impedance and the line and source angles are set equal to one another as shown in equations 6 – 9.

Table 1. System Parameters Positive Sequence Line Impedance: ZL 0.89 Ω Positive Sequence Line Angle: ZLAng 81.89 Zone 1 Reach: ZLMag 0.71 Ω Nominal Voltage: Vn 69.28 V Nominal Voltage Angle: VnAng 0 Source Impedance: Zs 0.0142 Ω Source Angle: ZsAng 81.89 Zero Sequence Magnitude Compensation: K0M 0.558 Zero Sequence Angle Compensation: K0A -12.72

Figure 8: Equivalent Sequence Component Circuit of a Ground Fault


))cos(*(1

)sin(*tan

))sin(())cos(1(

00

001

200

200

AM

AMAng

AMAMMag

kk

kkZeroComp

kkkkZeroComp

Eq.5

Eq.4

LAngNewsAng

LinesMag

AngLAngLAngNew

MagLineNew

ZZ

ZZ

ZeroCompZZ

ZeroCompZZ

9 Eq.

8 Eq.

7 Eq.

6 Eq.

*016.0

*

Table 1: System Parameters


Using geometric principles, one can determine the dynamic reach of the mho circle using the settings of the protective relay. The zone 1 reach, ZLMag, and source impedance, ZS, are plotted as two vec-tor quantities. When the source impedance angle, ZSang, and the line impedance angle, ZLang, equal one another, a straight line is formed through the center of the mho circle. Because the expansion is due to a line-toground fault, the ground impedance also needs to be consid-ered. There will be a change in the apparent reach and apparent line angle as seen by the relay. The k0 factor assists in calculating how the ground impedance will affect the reach and angle of the appar-ent impedance seen bythe relay. The zero sequence magnitude and angle compensation are expressed in equations 4 and 5.

The zero sequence magnitude and angle compensation are then applied to the original line impedance and angle. The source im-pedance magnitude is chosen as a percentage of the line imped-ance and the line and source angles are set equal to one another as shown in equations 6 – 9.

Because the line and source impedances are vector quantities, they can be broken down into their real and imaginary compo-nents. Then, the center point along the mho circle can be found as shown in equations 10 – 15.

After the center point of the circle has been found, the radius of the circle and compensation angle can be determined. The com-pensation angle is the apparent line angle that the relay sees during the ground fault. Next, an adjustment for the position of the mho circle is determined. Finally, the new total apparent reach, Ztot and line angle, Zang, can be determined. This is shown in equations 16 – 23.

When using these equations, one must note that a correction to the line angle, Zang, may be necessary to keep the line angle in the correct quadrant. This correction is determined after solving for X and Y. The following relations show what corrections are necessary:

Using the relay settings and the previous equations, one can de-termine the apparent reach, ZTot, and line angle, ZAng, are 1.095Ώ and 77.2° respectively. After the new reach and line angle are known, the secondary currents and voltages to be applied to the relay can be computed. First, the K factor needs to be determined:

The following expression is then derived by adding all of the impedances of the sequence network in Figure 8:

This can be simplified to:

From here, the fault currents can be expressed:

By solving for the sequence components in each phase, one can calculate the final secondary currents. These are shown in Eqs. 28 – 30:

The voltages are found in a similar manner. The fault voltages can be expressed by the following equations:














))cos(*(1

)sin(*tan

))sin(())cos(1(

00

001

200

200

AM

AMAng

AMAMMag

kk

kkZeroComp

kkkkZeroComp

Eq.5

Eq.4

LAngNewsAng

LinesMag

AngLAngLAngNew

MagLineNew

ZZ

ZZ

ZeroCompZZ

ZeroCompZZ

9 Eq.

8 Eq.

7 Eq.

6 Eq.

*016.0

*














))cos(*(1

)sin(*tan

))sin(())cos(1(

00

001

200

200

AM

AMAng

AMAMMag

kk

kkZeroComp

kkkkZeroComp

Eq.5

Eq.4

LAngNewsAng

LinesMag

AngLAngLAngNew

MagLineNew

ZZ

ZZ

ZeroCompZZ

ZeroCompZZ

9 Eq.

8 Eq.

7 Eq.

6 Eq.

*016.0

*


Because the line and source impedances are vector quantities, they can be broken down into their real and imaginary components. Then, the center point along the mho circle can be found as shown in equations 10 – 15.

After the center point of the circle has been found, the radius of the circle and compensation angle can be determined. The compensation angle is the apparent line angle that the relay sees during the ground fault. Next, an adjustment for the position of the mho circle is determined. Finally, the new total apparent reach, Ztot and line angle, Zang, can be determined. This is shown in equations 16 – 23.


Using the relay settings and the previous equations, one can determine the apparent reach, ZTot, and line angle, ZAng, are 1.095Ώ and 77.2° respectively. After the new reach and line angle are known, the secondary currents and voltages to be applied to the relay can be computed. First, the K factor needs to be determined:






These quantities are converted to each of the phase quantities:

The final calculated test voltages and currents are shown in Table 2.

Table 2: Calculated Secondary Voltages and Currents

The values calculated will simulate a system with constant source impedance precisely at the edge of the mho characteristic. To find the values that are slightly outside and slightly inside the operating characteristic, one must perform the previous series of calculations for each new test point. Fortunately, software programs that can perform these calculations quickly for the user are readily available.

ConClusionsWith the increasing capabilities of today’s relay test equipment and software, steady-state methods are becoming replaced by more accurate forms of testing. Computer simulations and appropriate equipment can quickly and easily provide realistic scenarios that will thoroughly test protective relays before they are put into service. However, once in service, these same relays must be periodically tested. The dynamic methods are scaled back from a full simulation to values calculated from the systems sequence components. This article has illustrated a method by which this can be performed.

reFerenCesC.L. Fortescue, Method of Symmetrical

Co-ordinates Applied to the Solution of Polyphase Networks, Annual Convention of the American Institute of Electrical Engineers, Atlantic City, NJ, 1918

A. T. Guilante, Dynamic Relay Testing, ATG Consulting

2

2

)sin(*

)cos(*

)sin(*

)cos(*

ImIm

ReRe

Im

Re

Im

Re

sNewY

alsalNewX

LAngNewNewNew

LAngNewNewalNew

sAngsMags

sAngsMagals

ZZC

ZZC

ZZZ

ZZZ

ZZZ

ZZZ

15 Eq.

14 Eq.

13 Eq.

12 Eq.

11 Eq.

10 Eq.

X

YZ

YXZ

ZYTestAngleRadiusY

ZXTestAngleRadiusX

CompRadiusY

CompRadiusX

CZ

CZComp

CZCZRadius

Ang

tot

sadjustmentMag

alsadjustmentMag

AngMagadjustment

AngMagadjustment

XalNew

YNewAng

YNewXalNewMag

1

22

Im

Re

Re

Im1

2Im

2Re

tan

)sin(*

)cos(*

)sin(*

)cos(*

tan

)()(

23 Eq.

22 Eq.

21 Eq.

20 Eq.

19 Eq.

18 Eq.

17 Eq.

16 Eq.

If X > 0 and Y > 0 then the angle is tan-1

Y/X If X < 0 and Y > 0 then you need to add 180 to ZAng. If X < 0 and Y < 0 then you need to add 180 to ZAng. If X > 0 and Y < 0 then you need to add 360 to ZAng.

t

nfff Z

VIII 0 27 Eq.

Lin eSLin eSLin eSt ZKZKZZZZZ 25 Eq.

)2()( KZZZ SLinet 26 Eq.

ffff

c

ffff

b

fffff

a

IIII

IIII

IIIII

20

20

0 3

30 Eq.

29 Eq.

28 Eq.

)( Sfnf ZIVV 31 Eq.

)(

)(0 KZIV

ZIV

Sff

Sff

33 Eq.

32 Eq.

ffff

a VVVV 0 34 Eq. fff

fb VVVV 20 35 Eq.

ffff

c VVVV 20 36 Eq.

1)3( 0 kK 24 Eq.

Rene Aguilar received his B.S. in Electrical Engineering from the University of Texas at Austin. He worked on APPDS (Automatic Protection Detection System) used for detecting coordinating issues between devices in a distributed generated system. In 2006, he joined Megger as an application engineer in the technical support group. Rene is in charge of developing automatic testing for numerical relays as well as the implementation of IEC 61850 on various Megger products. Rene has extensive experience in the testing and commissioning of electrical schemes and multivendor device applications of IEC 61850. He is a member of IEEE and and an active member of the Power System Relay Committee.

Jason Buneo received his B.S and M.S in Electrical Engineering from the University at Buffalo in 2001 and 2005, respectively. In 2005, he joined GE Energy Services as a field service engineer. He specialized in arc-flash and coordination studies, protective relay testing and calibration, and low- and medium-voltage switchgear repair. In 2008 he joined Megger as an Applications Engineer specializing in protective relay evaluation and testing. He is also a participating member of the IEEE Power Systems Relaying Committee.



















2

2

)sin(*

)cos(*

)sin(*

)cos(*

ImIm

ReRe

Im

Re

Im

Re

sNewY

alsalNewX

LAngNewNewNew

LAngNewNewalNew

sAngsMags

sAngsMagals

ZZC

ZZC

ZZZ

ZZZ

ZZZ

ZZZ

15 Eq.

14 Eq.

13 Eq.

12 Eq.

11 Eq.

10 Eq.

X

YZ

YXZ

ZYTestAngleRadiusY

ZXTestAngleRadiusX

CompRadiusY

CompRadiusX

CZ

CZComp

CZCZRadius

Ang

tot

sadjustmentMag

alsadjustmentMag

AngMagadjustment

AngMagadjustment

XalNew

YNewAng

YNewXalNewMag

1

22

Im

Re

Re

Im1

2Im

2Re

tan

)sin(*

)cos(*

)sin(*

)cos(*

tan

)()(

23 Eq.

22 Eq.

21 Eq.

20 Eq.

19 Eq.

18 Eq.

17 Eq.

16 Eq.



t

nfff Z

VIII 0 27 Eq.



ffff

c

ffff

b

fffff

a

IIII

IIII

IIIII

20

20

0 3

30 Eq.

29 Eq.

28 Eq.

)( Sfnf ZIVV 31 Eq.

)(

)(0 KZIV

ZIV

Sff

Sff

33 Eq.

32 Eq.

ffff

a VVVV 0 34 Eq. fff

fb VVVV 20 35 Eq.

ffff

c VVVV 20 36 Eq.

1)3( 0 kK 24 Eq.





















2

2

)sin(*

)cos(*

)sin(*

)cos(*

ImIm

ReRe

Im

Re

Im

Re

sNewY

alsalNewX

LAngNewNewNew

LAngNewNewalNew

sAngsMags

sAngsMagals

ZZC

ZZC

ZZZ

ZZZ

ZZZ

ZZZ

15 Eq.

14 Eq.

13 Eq.

12 Eq.

11 Eq.

10 Eq.

X

YZ

YXZ

ZYTestAngleRadiusY

ZXTestAngleRadiusX

CompRadiusY

CompRadiusX

CZ

CZComp

CZCZRadius

Ang

tot

sadjustmentMag

alsadjustmentMag

AngMagadjustment

AngMagadjustment

XalNew

YNewAng

YNewXalNewMag

1

22

Im

Re

Re

Im1

2Im

2Re

tan

)sin(*

)cos(*

)sin(*

)cos(*

tan

)()(

23 Eq.

22 Eq.

21 Eq.

20 Eq.

19 Eq.

18 Eq.

17 Eq.

16 Eq.



t

nfff Z

VIII 0 27 Eq.



ffff

c

ffff

b

fffff

a

IIII

IIII

IIIII

20

20

0 3

30 Eq.

29 Eq.

28 Eq.

)( Sfnf ZIVV 31 Eq.

)(

)(0 KZIV

ZIV

Sff

Sff

33 Eq.

32 Eq.

ffff

a VVVV 0 34 Eq. fff

fb VVVV 20 35 Eq.

ffff

c VVVV 20 36 Eq.

1)3( 0 kK 24 Eq.





















2

2

)sin(*

)cos(*

)sin(*

)cos(*

ImIm

ReRe

Im

Re

Im

Re

sNewY

alsalNewX

LAngNewNewNew

LAngNewNewalNew

sAngsMags

sAngsMagals

ZZC

ZZC

ZZZ

ZZZ

ZZZ

ZZZ

15 Eq.

14 Eq.

13 Eq.

12 Eq.

11 Eq.

10 Eq.

X

YZ

YXZ

ZYTestAngleRadiusY

ZXTestAngleRadiusX

CompRadiusY

CompRadiusX

CZ

CZComp

CZCZRadius

Ang

tot

sadjustmentMag

alsadjustmentMag

AngMagadjustment

AngMagadjustment

XalNew

YNewAng

YNewXalNewMag

1

22

Im

Re

Re

Im1

2Im

2Re

tan

)sin(*

)cos(*

)sin(*

)cos(*

tan

)()(

23 Eq.

22 Eq.

21 Eq.

20 Eq.

19 Eq.

18 Eq.

17 Eq.

16 Eq.



t

nfff Z

VIII 0 27 Eq.



ffff

c

ffff

b

fffff

a

IIII

IIII

IIIII

20

20

0 3

30 Eq.

29 Eq.

28 Eq.

)( Sfnf ZIVV 31 Eq.

)(

)(0 KZIV

ZIV

Sff

Sff

33 Eq.

32 Eq.

ffff

a VVVV 0 34 Eq. fff

fb VVVV 20 35 Eq.

ffff

c VVVV 20 36 Eq.

1)3( 0 kK 24 Eq.





















2

2

)sin(*

)cos(*

)sin(*

)cos(*

ImIm

ReRe

Im

Re

Im

Re

sNewY

alsalNewX

LAngNewNewNew

LAngNewNewalNew

sAngsMags

sAngsMagals

ZZC

ZZC

ZZZ

ZZZ

ZZZ

ZZZ

15 Eq.

14 Eq.

13 Eq.

12 Eq.

11 Eq.

10 Eq.

X

YZ

YXZ

ZYTestAngleRadiusY

ZXTestAngleRadiusX

CompRadiusY

CompRadiusX

CZ

CZComp

CZCZRadius

Ang

tot

sadjustmentMag

alsadjustmentMag

AngMagadjustment

AngMagadjustment

XalNew

YNewAng

YNewXalNewMag

1

22

Im

Re

Re

Im1

2Im

2Re

tan

)sin(*

)cos(*

)sin(*

)cos(*

tan

)()(

23 Eq.

22 Eq.

21 Eq.

20 Eq.

19 Eq.

18 Eq.

17 Eq.

16 Eq.



t

nfff Z

VIII 0 27 Eq.



ffff

c

ffff

b

fffff

a

IIII

IIII

IIIII

20

20

0 3

30 Eq.

29 Eq.

28 Eq.

)( Sfnf ZIVV 31 Eq.

)(

)(0 KZIV

ZIV

Sff

Sff

33 Eq.

32 Eq.

ffff

a VVVV 0 34 Eq. fff

fb VVVV 20 35 Eq.

ffff

c VVVV 20 36 Eq.

1)3( 0 kK 24 Eq.





















2

2

)sin(*

)cos(*

)sin(*

)cos(*

ImIm

ReRe

Im

Re

Im

Re

sNewY

alsalNewX

LAngNewNewNew

LAngNewNewalNew

sAngsMags

sAngsMagals

ZZC

ZZC

ZZZ

ZZZ

ZZZ

ZZZ

15 Eq.

14 Eq.

13 Eq.

12 Eq.

11 Eq.

10 Eq.

X

YZ

YXZ

ZYTestAngleRadiusY

ZXTestAngleRadiusX

CompRadiusY

CompRadiusX

CZ

CZComp

CZCZRadius

Ang

tot

sadjustmentMag

alsadjustmentMag

AngMagadjustment

AngMagadjustment

XalNew

YNewAng

YNewXalNewMag

1

22

Im

Re

Re

Im1

2Im

2Re

tan

)sin(*

)cos(*

)sin(*

)cos(*

tan

)()(

23 Eq.

22 Eq.

21 Eq.

20 Eq.

19 Eq.

18 Eq.

17 Eq.

16 Eq.



t

nfff Z

VIII 0 27 Eq.



ffff

c

ffff

b

fffff

a

IIII

IIII

IIIII

20

20

0 3

30 Eq.

29 Eq.

28 Eq.

)( Sfnf ZIVV 31 Eq.

)(

)(0 KZIV

ZIV

Sff

Sff

33 Eq.

32 Eq.

ffff

a VVVV 0 34 Eq. fff

fb VVVV 20 35 Eq.

ffff

c VVVV 20 36 Eq.

1)3( 0 kK 24 Eq.





















2

2

)sin(*

)cos(*

)sin(*

)cos(*

ImIm

ReRe

Im

Re

Im

Re

sNewY

alsalNewX

LAngNewNewNew

LAngNewNewalNew

sAngsMags

sAngsMagals

ZZC

ZZC

ZZZ

ZZZ

ZZZ

ZZZ

15 Eq.

14 Eq.

13 Eq.

12 Eq.

11 Eq.

10 Eq.

X

YZ

YXZ

ZYTestAngleRadiusY

ZXTestAngleRadiusX

CompRadiusY

CompRadiusX

CZ

CZComp

CZCZRadius

Ang

tot

sadjustmentMag

alsadjustmentMag

AngMagadjustment

AngMagadjustment

XalNew

YNewAng

YNewXalNewMag

1

22

Im

Re

Re

Im1

2Im

2Re

tan

)sin(*

)cos(*

)sin(*

)cos(*

tan

)()(

23 Eq.

22 Eq.

21 Eq.

20 Eq.

19 Eq.

18 Eq.

17 Eq.

16 Eq.



t

nfff Z

VIII 0 27 Eq.



ffff

c

ffff

b

fffff

a

IIII

IIII

IIIII

20

20

0 3

30 Eq.

29 Eq.

28 Eq.

)( Sfnf ZIVV 31 Eq.

)(

)(0 KZIV

ZIV

Sff

Sff

33 Eq.

32 Eq.

ffff

a VVVV 0 34 Eq. fff

fb VVVV 20 35 Eq.

ffff

c VVVV 20 36 Eq.

1)3( 0 kK 24 Eq.





















2

2

)sin(*

)cos(*

)sin(*

)cos(*

ImIm

ReRe

Im

Re

Im

Re

sNewY

alsalNewX

LAngNewNewNew

LAngNewNewalNew

sAngsMags

sAngsMagals

ZZC

ZZC

ZZZ

ZZZ

ZZZ

ZZZ

15 Eq.

14 Eq.

13 Eq.

12 Eq.

11 Eq.

10 Eq.

X

YZ

YXZ

ZYTestAngleRadiusY

ZXTestAngleRadiusX

CompRadiusY

CompRadiusX

CZ

CZComp

CZCZRadius

Ang

tot

sadjustmentMag

alsadjustmentMag

AngMagadjustment

AngMagadjustment

XalNew

YNewAng

YNewXalNewMag

1

22

Im

Re

Re

Im1

2Im

2Re

tan

)sin(*

)cos(*

)sin(*

)cos(*

tan

)()(

23 Eq.

22 Eq.

21 Eq.

20 Eq.

19 Eq.

18 Eq.

17 Eq.

16 Eq.



t

nfff Z

VIII 0 27 Eq.



ffff

c

ffff

b

fffff

a

IIII

IIII

IIIII

20

20

0 3

30 Eq.

29 Eq.

28 Eq.

)( Sfnf ZIVV 31 Eq.

)(

)(0 KZIV

ZIV

Sff

Sff

33 Eq.

32 Eq.

ffff

a VVVV 0 34 Eq. fff

fb VVVV 20 35 Eq.

ffff

c VVVV 20 36 Eq.

1)3( 0 kK 24 Eq.







The values calculated will simulate a system with constant source impedance precisely at the edge of the mho characteris-tic. To find the values that are slightly outside and slightly inside the operating characteristic, one must perform the previous series of calculations for each new test point. Fortunately, software pro-grams that can perform these calculations quickly for the user are readily available.

CONCLUSIONSWith the increasing capabilities of today’s relay test equipment

and software, steady-state methods are becoming replaced by more accurate forms of testing. Computer simulations and appro-priate equipment can quickly and easily provide realistic scenarios that will thoroughly test protective relays before they are put into service. However, once in service, these same relays must be peri-odically tested. The dynamic methods are scaled back from a full simulation to values calculated from the systems sequence com-ponents. This article has illustrated a method by which this can be performed.

REFERENCESC.L. Fortescue. “Method of Symmetrical Co-ordinates Applied to the Solution of Polyphase Networks.” Annual Convention of the American Institute of Electrical Engineers, Atlantic City, NJ, 1918.

A. T. Guilante. Dynamic Relay Testing, ATG Consulting.

Rene Aguilar received his B.S. in Electrical Engineering from the University of Texas at Austin. He worked on APPDS (Automatic Protection Detection System) used for detecting coordinating is-sues between devices in a distributed generated system. In 2006, he joined Megger as an application engineer in the technical sup-port group. Rene is in charge of developing automatic testing for numerical relays as well as the implementation of IEC 61850 on various Megger products. Rene has extensive experience in the testing and commissioning of electrical schemes and multivendor device applications of IEC 61850. He is a member of IEEE and and an active member of the Power System Relay Committee.

Jason Buneo received his B.S and M.S in Electrical Engineering from the University at Buffalo in 2001 and 2005, respectively. In 2005, he joined GE Energy Services as a field service engineer. He specialized in arc-flash and coordination studies, protective relay testing and calibration, and low- and medium-voltage switchgear repair. In 2008 he joined Megger as an Applications Engineer spe-cializing in protective relay evaluation and testing. He is also a par-ticipating member of the IEEE Power Systems Relaying Committee.



















2

2

)sin(*

)cos(*

)sin(*

)cos(*

ImIm

ReRe

Im

Re

Im

Re

sNewY

alsalNewX

LAngNewNewNew

LAngNewNewalNew

sAngsMags

sAngsMagals

ZZC

ZZC

ZZZ

ZZZ

ZZZ

ZZZ

15 Eq.

14 Eq.

13 Eq.

12 Eq.

11 Eq.

10 Eq.

X

YZ

YXZ

ZYTestAngleRadiusY

ZXTestAngleRadiusX

CompRadiusY

CompRadiusX

CZ

CZComp

CZCZRadius

Ang

tot

sadjustmentMag

alsadjustmentMag

AngMagadjustment

AngMagadjustment

XalNew

YNewAng

YNewXalNewMag

1

22

Im

Re

Re

Im1

2Im

2Re

tan

)sin(*

)cos(*

)sin(*

)cos(*

tan

)()(

23 Eq.

22 Eq.

21 Eq.

20 Eq.

19 Eq.

18 Eq.

17 Eq.

16 Eq.



t

nfff Z

VIII 0 27 Eq.



ffff

c

ffff

b

fffff

a

IIII

IIII

IIIII

20

20

0 3

30 Eq.

29 Eq.

28 Eq.

)( Sfnf ZIVV 31 Eq.

)(

)(0 KZIV

ZIV

Sff

Sff

33 Eq.

32 Eq.

ffff

a VVVV 0 34 Eq. fff

fb VVVV 20 35 Eq.

ffff

c VVVV 20 36 Eq.

1)3( 0 kK 24 Eq.





















2

2

)sin(*

)cos(*

)sin(*

)cos(*

ImIm

ReRe

Im

Re

Im

Re

sNewY

alsalNewX

LAngNewNewNew

LAngNewNewalNew

sAngsMags

sAngsMagals

ZZC

ZZC

ZZZ

ZZZ

ZZZ

ZZZ

15 Eq.

14 Eq.

13 Eq.

12 Eq.

11 Eq.

10 Eq.

X

YZ

YXZ

ZYTestAngleRadiusY

ZXTestAngleRadiusX

CompRadiusY

CompRadiusX

CZ

CZComp

CZCZRadius

Ang

tot

sadjustmentMag

alsadjustmentMag

AngMagadjustment

AngMagadjustment

XalNew

YNewAng

YNewXalNewMag

1

22

Im

Re

Re

Im1

2Im

2Re

tan

)sin(*

)cos(*

)sin(*

)cos(*

tan

)()(

23 Eq.

22 Eq.

21 Eq.

20 Eq.

19 Eq.

18 Eq.

17 Eq.

16 Eq.



t

nfff Z

VIII 0 27 Eq.



ffff

c

ffff

b

fffff

a

IIII

IIII

IIIII

20

20

0 3

30 Eq.

29 Eq.

28 Eq.

)( Sfnf ZIVV 31 Eq.

)(

)(0 KZIV

ZIV

Sff

Sff

33 Eq.

32 Eq.

ffff

a VVVV 0 34 Eq. fff

fb VVVV 20 35 Eq.

ffff

c VVVV 20 36 Eq.

1)3( 0 kK 24 Eq.





















2

2

)sin(*

)cos(*

)sin(*

)cos(*

ImIm

ReRe

Im

Re

Im

Re

sNewY

alsalNewX

LAngNewNewNew

LAngNewNewalNew

sAngsMags

sAngsMagals

ZZC

ZZC

ZZZ

ZZZ

ZZZ

ZZZ

15 Eq.

14 Eq.

13 Eq.

12 Eq.

11 Eq.

10 Eq.

X

YZ

YXZ

ZYTestAngleRadiusY

ZXTestAngleRadiusX

CompRadiusY

CompRadiusX

CZ

CZComp

CZCZRadius

Ang

tot

sadjustmentMag

alsadjustmentMag

AngMagadjustment

AngMagadjustment

XalNew

YNewAng

YNewXalNewMag

1

22

Im

Re

Re

Im1

2Im

2Re

tan

)sin(*

)cos(*

)sin(*

)cos(*

tan

)()(

23 Eq.

22 Eq.

21 Eq.

20 Eq.

19 Eq.

18 Eq.

17 Eq.

16 Eq.



t

nfff Z

VIII 0 27 Eq.



ffff

c

ffff

b

fffff

a

IIII

IIII

IIIII

20

20

0 3

30 Eq.

29 Eq.

28 Eq.

)( Sfnf ZIVV 31 Eq.

)(

)(0 KZIV

ZIV

Sff

Sff

33 Eq.

32 Eq.

ffff

a VVVV 0 34 Eq. fff

fb VVVV 20 35 Eq.

ffff

c VVVV 20 36 Eq.

1)3( 0 kK 24 Eq.







■■ Transformers

■■ Batteries











■■ Transformers

■■ Batteries








We previously discussed the problem of low-ratio CTs used on systems with high fault current and mentioned the IE working group report on this subject. Because of the emphasis in this report on making sure that CTs do not saturate, a number of people ex-pressed concern about the operation of instantaneous ground fault relays connected to zero-sequence, or core balance, CTs. Because of this concern, Powell recently ran a series of tests to check the operation of typical CTrelay combinations. Two different relays were tested with each of two CTs. The relays were the GE HFC and the AB IT. Electromechanical relays were chosen for the test because their higher burden places a greater load on the CTs. The CTs used were both made by ITI. The first was Model 141-500, 50/5, C10 accuracy. The second was Model 143-500, 50/5, C20 accuracy. The test results are presented in the table below.

Both of these relays operated correctly and reliably with both CTs. However, we also tested a third relay, the AB ITH, a high dropout version of the IT. We found that this relay was not reli-able in this service. It picked up at quite low values, and operated well with primary currents up to about 150 amperes. At currents of 600 amperes and higher, it chattered badly and did not close its contacts long enough to operate a circuit breaker. Upon asking around, we found that this relay had been recommended for 50 GS service in the past, but its manufacturer (Westinghouse at that time) changed the recommendation when the chattering problem was discovered. Based on this information and the tests, Powell strongly recommends that the ITH relay not be used as a 50GS relay. Summarizing, both the HFC and the IT work quite well at primary ground fault currents up to 1800 amperes, even though the CTs are badly saturated at that current level. This circuit, with these CTs and relays, should not be used on solidly-grounded sys-tems with high ground fault current. For these systems, residually-connected relays should be used, or the zero-sequence CTs should have higher ratios.

Baldwin Bridger is retired Technical Director of Powell Electrical Manufacturer Co., Houston, Texas. He has worked as an engineer and engineering manager in the design of low- and medium-volt-age 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.

INSTANTANEOUS GROUND FAULT RELAYS (50GS)AND ZERO-SEQUENCE CTS

POWELL TECHNICAL BRIEF #68NETA World, Spring 2012 Issue

by Baldwin Bridger, Powell Electrical Manufacturer Co.


An extremely powerful tool for analyzing unbalanced three-phase circuits is the method of symmetrical components intro-duced by Charles Legeyt Fortescue. In 1918, Fortescue presented a paper entitled Method of Symmetrical Coordinates Applied to the Solution of Polyphase Networks, which demonstrated that any set of n unbalanced phasors could be expressed as the sum of n symmetrical sets of balanced phasors. The n phasors of each set of components are equal in length, and the angles between adjacent phasors of the set are equal.

In a normally balanced three-phase system, unbalanced fault conditions generally cause unbalanced currents and voltages to exist in each of the phases. Calculating unbalanced voltage and current levels through the elements of a power system is very com-plex. Using Fortescue’s method, the unbalanced system conditions can be mathematically changed into three easy to calculate bal-anced systems. After the three balanced systems are calculated, the result is then, with simple addition, changed back into the ac-tual unbalanced levels at each location of the power system. In fact, most numerical relays operate from symmetrical component quantities.

By organizing the separate equivalent components into net-works according to the interconnections of the elements, we arrive at the concept of three sequence networks. Solving the balanced sequence networks for the fault conditions is relatively easy. The solution gives symmetrical current and voltage components which can be summed together to reflect the effects of the original un-balanced fault on the overall system. According to Fortescue’s method, there are three equivalent circuits for each element of a three-phase system. Figure 1 assumes ABC system rotation. The balanced sets of components for a three phase system are the fol-lowing:

1. Positive-sequence components consisting of three phasors equal in magnitude, displaced from each other by 120° in phase, and having the same phase rotation sequence as the original phasors.

2. Negative-sequence components consisting of three phasors equal in magnitude, displaced from each other by 120° in phase, and having the phase sequence rotation opposite to that of the original phasors.

3. Zero-sequence components consisting of three phasors equal in magnitude and with zero phase displacement from each other.

FIGURE 1: Positive-sequence, negative-sequence, and zero-sequence components

The three sets of symmetrical components are designated by a subscript 1 for the positivesequence components, a subscript 2 for the negative-sequence components, and a subscript 0 for the zero-sequence components. The positive-sequence set is the only one present during perfectly balanced system operation. The presence of negative-sequence and zerosequence components indicates unbalanced operation of the power system and/or power system faults.

All three-phase quantities in a power system can be represented by the sum of the symmetrical components. Let’s take voltages as an example. These three-phase voltages can be expressed in terms of the sequence components as shown in equations (1), (2), and (3).

The unknown quantities can be reduced by making use of the fact that the positive and negative sequence components always have exactly 120° between them. So we define an operator Using this “a” operator, a = 1 120º. we simplify equations (2) and (3) in equation (4) below.

We also know by definition that V0A=V0B=V0C

MODERN PROTECTIVE RELAY TECHNIQUES:USING A 94-YEAR-OLD CONCEPT TO PROTECT

ELECTRICAL EQUIPMENTNETA World, Spring 2012 Issue

by Suparat Pavavicharn, Basler Electric Company

The three sets of symmetrical components are designated by a subscript 1 for the positive-sequence components, a subscript 2 for the negative-sequence components, and a subscript 0 for the zero-sequence components. The positive-sequence set is the only one present during perfectly balanced system operation. The presence of negative-sequence and zero-sequence components indicates unbalanced operation of the power system and/or power system faults.


The unknown quantities can be reduced by making use of the fact that the positive and negative sequence components always have exactly 120° between them. So we define an operator Using this “a” operator, we simplify equations (2) and (3) in equation (4) below.

We also know by definition that V0A = V0B = V0C.

Substituting equation (4) in equations (1), (2), and (3)

The quantity Adding VA, VB, and VC and solving for each of the symmetrical components, with A-phase quantities yields

The A-phase components are the reference; therefore, the suffix “A” is dropped and the components are denoted by V0, V1, and V2. Currents can be determined in a similar manner; the resulting sequence currents are I0, I1, and I2.

SEQUENCE NETWORKS IN MODERN NUMERICAL PROTECTIVE RELAYS Zero-sequence components and negative-sequence components are both measurable indications of unbalanced conditions. Obtaining zero-sequence quantities with electromechanical relays has not been a problem because the component does not require phase shifting by the “a” operator. See equation (8). The sum of the phase quantities is proportional to the zero-sequence component. Conversely, the negative-sequence component has been more difficult and costly to obtain with electromechanical relays due to the high cost of techniques required to filter negative-sequence components.

Modern numerical relays include a number of advantages and functions associated with simple, accurate calculation of symmetrical components from phase quantities. The numerical relay turns the analog voltage and current levels into digital values with a magnitude and phase angle. After the A/D conversion, determining the symmetrical components is accomplished by the mathematical capabilities of the modern microprocessor. As a result, numerical relays can derive easily the positive-, negative-, and zero-sequence quantities. Numerical relays also provide the tools for the relay engineers to analyze sequence components recorded in oscillography files.

POLARIZING DIRECTIONAL OVERCURRENT RELAYS WITH SYMMETRICAL COMPONENTS Directional overcurrent relays (67) can use the phase relationship of voltage and current to determine the direction to a fault. To determine the direction to a fault, the angular relationship is the only concern. Usually, the angle setting of the directional relay is referred to as the maximum torque angle (MTA). The simple concept of detecting fault direction in these relays is that there is an approximate 180° difference of calculated sequence impedances Z0, Z1, and Z2 for faults in the two directions from the relay location in faulted conditions. This high variation in phase angle is a reliable indication of the direction to a fault.

The three-phase voltage drop equation for a system that can be represented by voltages at two defined locations, system voltage VS and fault voltage VF, is the following:

When the impedances are closely balanced, equations (11), (12), and (13) can be expressed in symmetrical component quantities as shown in equations (14), (15), and (16).

Modern numerical protective relays use the angular relationships of symmetrical component currents and voltages and the resultant angular nature of sequence impedance Z0 , Z1, and Z2 to determine the direction to a fault. These three impedances are used to create three directional assessments; 67Z E R O , 67POS, and 67NEG respectively, in modern numerical protective relays. The common process in these relays is to measure the impedance angle and compare it to a window of the MTA ±90° as forward or reverse.

ZERO-SEQUENCE CURRENT FOR GROUND FAULT PROTECTION

Figure 2: Zero-sequence current components used for ground fault protection

A ground fault current can be considered a zero-sequence current as it passes through the system to reach the location of the ground fault. (See Figure 2.) This zero-sequence current may be supplied from different sources and delivered to the particular zone involving ground fault. A zero-sequence current also can be caused by an imbalance in impedances and sources within the power system that has multiple groundings or it could be a triplin (3, 6, 9, 12, etc.) harmonic current. These currents would not necessarily involve a ground fault. In these circumstances, we would not want ground fault protective relays to cause false tripping.

With numerical protective relays, protection against a phase-to-ground fault is less difficult. A ground fault protective relay must detect all phase-to-ground faults within its defined zone of protection. The well-known method of detecting ground faults is to use a numerical relay that responds only to the zero-sequence current of the system.

58 • SPRING 2012

FEATURE

Modern Protective relay techniques: using a 94-year-old concePt to Protect electrical equiPMent

NETAWORLD • 59Modern Protective relay techniques: using a 94-year-old concePt to Protect electrical equiPMent

VA = V0A + V1A + V2A (1)

VB = V0B + V1B + V2B (2)

VC = V0C + V1C + V2C (3)

V1a = V2a

V1B = a2 ∙ V1A V1C = a ∙ V1A (4)

V2B = a ∙ V2A V2C = a2 ∙ V2A

VA = V0A + V1A + V2A (5)

VB = V0A + a2 ∙ V1A + a ∙ V2A (6)

VC = V0A + a ∙ V1A + a2 ∙ V2A (7)

V0A = ⅓ (VA + VB + VC) (8)

V1A = ⅓ (VA + a ∙ VB + a2 ∙ VC) (9)

V2A = ⅓ (VA + a2 ∙ VB + a ∙ VC) (10)

VA,S – VA,F = ZAA ∙ IA + ZAB ∙ IB + ZAC ∙ IC (11) VB,S – VB,F = ZBA ∙ IA + ZBB ∙ IB + ZBC ∙ IC (12) VC,S – VC,F = ZCA ∙ IA + ZCB ∙ IB + ZCC ∙ IC (13)

V0,S – V0,F = Z0 ∙ I0 (14) V1,S – V1,F = Z1 ∙ I1 (15) V2,S – V2,F = Z2 ∙ I2 (16)

a = 1120º.

1 + a + a2 = 10º + 1120º + 1240º = 0

FEATUREThe three sets of symmetrical components are designated by a subscript 1 for the positive-sequence components, a subscript 2 for the negative-sequence components, and a subscript 0 for the zero-sequence components. The positive-sequence set is the only one present during perfectly balanced system operation. The presence of negative-sequence and zero-sequence components indicates unbalanced operation of the power system and/or power system faults.

















58 • SPRING 2012

FEATURE



VA = V0A + V1A + V2A (1)

VB = V0B + V1B + V2B (2)

VC = V0C + V1C + V2C (3)

V1a = V2a

V1B = a2 ∙ V1A V1C = a ∙ V1A (4)

V2B = a ∙ V2A V2C = a2 ∙ V2A

VA = V0A + V1A + V2A (5)

VB = V0A + a2 ∙ V1A + a ∙ V2A (6)

VC = V0A + a ∙ V1A + a2 ∙ V2A (7)

V0A = ⅓ (VA + VB + VC) (8)

V1A = ⅓ (VA + a ∙ VB + a2 ∙ VC) (9)

V2A = ⅓ (VA + a2 ∙ VB + a ∙ VC) (10)



a = 1120º.

1 + a + a2 = 10º + 1120º + 1240º = 0

FEATURE


















58 • SPRING 2012

FEATURE



VA = V0A + V1A + V2A (1)

VB = V0B + V1B + V2B (2)

VC = V0C + V1C + V2C (3)

V1a = V2a

V1B = a2 ∙ V1A V1C = a ∙ V1A (4)

V2B = a ∙ V2A V2C = a2 ∙ V2A

VA = V0A + V1A + V2A (5)

VB = V0A + a2 ∙ V1A + a ∙ V2A (6)

VC = V0A + a ∙ V1A + a2 ∙ V2A (7)

V0A = ⅓ (VA + VB + VC) (8)

V1A = ⅓ (VA + a ∙ VB + a2 ∙ VC) (9)

V2A = ⅓ (VA + a2 ∙ VB + a ∙ VC) (10)



a = 1120º.

1 + a + a2 = 10º + 1120º + 1240º = 0

FEATURE



The quantity 1 + a + a2 = 1 0º + 1 120º + 1 240º = 0 Adding VA, VB, and VC and solving for each of the symmetrical compo-nents, with A-phase quantities yields


SEQUENCE NETWORKS IN MODERN NUMERI-CAL PROTECTIVE RELAYS

Zero-sequence components and negativesequence components are both measurable indications of unbalanced conditions. Ob-taining zero-sequence quantities with electromechanical relays has not been a problem because the component does not require phase shifting by the “a” operator. See equation (8). The sum of the phase quantities is proportional to the zerosequence compo-nent. Conversely, the negativesequence component has been more difficult and costly to obtain with electromechanical relays due to the high cost of techniques required to filter negative-sequence components.

Modern numerical relays include a number of advantages and functions associated with simple, accurate calculation of sym-metrical components from phase quantities. The numerical relay turns the analog voltage and current levels into digital values with a magnitude and phase angle. After the A/D conversion, determin-ing the symmetrical components is accomplished by the math-ematical capabilities of the modern microprocessor. As a result, numerical relays can derive easily the positive-, negative-, and zero-sequence quantities. Numerical relays also provide the tools for the relay engineers to analyze sequence components recorded in oscillography files.

POLARIZING DIRECTIONAL OVERCURRENT RELAYS WITH SYMMETRICAL COMPONENTS

Directional overcurrent relays (67) can use the phase relation-ship of voltage and current to determine the direction to a fault. To determine the direction to a fault, the angular relationship is the only concern. Usually, the angle setting of the directional relay is referred to as the maximum torque angle (MTA). The simple con-cept of detecting fault direction in these relays is that there is an approximate 180° difference of calculated sequence impedances Z0, Z1, and Z2 for faults in the two directions from the relay loca-

tion in faulted conditions. This high variation in phase angle is a reliable indication of the direction to a fault.



Modern numerical protective relays use the angular relation-ships of symmetrical component currents and voltages and the resultant angular nature of sequence impedance Z0 , Z1, and Z2 to determine the direction to a fault. These three impedances are used to create three directional assessments; 67ZERO, 67POS, and 67NEG respectively, in modern numerical protective relays. The common process in these relays is to measure the impedance angle and compare it to a window of the MTA ±90° as forward or reverse.


FIGURE 2: Zero-sequence current components used for ground fault protection


















58 • SPRING 2012

FEATURE



VA = V0A + V1A + V2A (1)

VB = V0B + V1B + V2B (2)

VC = V0C + V1C + V2C (3)

V1a = V2a

V1B = a2 ∙ V1A V1C = a ∙ V1A (4)

V2B = a ∙ V2A V2C = a2 ∙ V2A

VA = V0A + V1A + V2A (5)

VB = V0A + a2 ∙ V1A + a ∙ V2A (6)

VC = V0A + a ∙ V1A + a2 ∙ V2A (7)

V0A = ⅓ (VA + VB + VC) (8)

V1A = ⅓ (VA + a ∙ VB + a2 ∙ VC) (9)

V2A = ⅓ (VA + a2 ∙ VB + a ∙ VC) (10)



a = 1120º.

1 + a + a2 = 10º + 1120º + 1240º = 0

FEATURE


















58 • SPRING 2012

FEATURE



VA = V0A + V1A + V2A (1)

VB = V0B + V1B + V2B (2)

VC = V0C + V1C + V2C (3)

V1a = V2a

V1B = a2 ∙ V1A V1C = a ∙ V1A (4)

V2B = a ∙ V2A V2C = a2 ∙ V2A

VA = V0A + V1A + V2A (5)

VB = V0A + a2 ∙ V1A + a ∙ V2A (6)

VC = V0A + a ∙ V1A + a2 ∙ V2A (7)

V0A = ⅓ (VA + VB + VC) (8)

V1A = ⅓ (VA + a ∙ VB + a2 ∙ VC) (9)

V2A = ⅓ (VA + a2 ∙ VB + a ∙ VC) (10)



a = 1120º.

1 + a + a2 = 10º + 1120º + 1240º = 0

FEATURE


















58 • SPRING 2012

FEATURE



VA = V0A + V1A + V2A (1)

VB = V0B + V1B + V2B (2)

VC = V0C + V1C + V2C (3)

V1a = V2a

V1B = a2 ∙ V1A V1C = a ∙ V1A (4)

V2B = a ∙ V2A V2C = a2 ∙ V2A

VA = V0A + V1A + V2A (5)

VB = V0A + a2 ∙ V1A + a ∙ V2A (6)

VC = V0A + a ∙ V1A + a2 ∙ V2A (7)

V0A = ⅓ (VA + VB + VC) (8)

V1A = ⅓ (VA + a ∙ VB + a2 ∙ VC) (9)

V2A = ⅓ (VA + a2 ∙ VB + a ∙ VC) (10)



a = 1120º.

1 + a + a2 = 10º + 1120º + 1240º = 0

FEATUREThe three sets of symmetrical components are designated by a subscript 1 for the positive-sequence components, a subscript 2 for the negative-sequence components, and a subscript 0 for the zero-sequence components. The positive-sequence set is the only one present during perfectly balanced system operation. The presence of negative-sequence and zero-sequence components indicates unbalanced operation of the power system and/or power system faults.

















58 • SPRING 2012

FEATURE



VA = V0A + V1A + V2A (1)

VB = V0B + V1B + V2B (2)

VC = V0C + V1C + V2C (3)

V1a = V2a

V1B = a2 ∙ V1A V1C = a ∙ V1A (4)

V2B = a ∙ V2A V2C = a2 ∙ V2A

VA = V0A + V1A + V2A (5)

VB = V0A + a2 ∙ V1A + a ∙ V2A (6)

VC = V0A + a ∙ V1A + a2 ∙ V2A (7)

V0A = ⅓ (VA + VB + VC) (8)

V1A = ⅓ (VA + a ∙ VB + a2 ∙ VC) (9)

V2A = ⅓ (VA + a2 ∙ VB + a ∙ VC) (10)



a = 1120º.

1 + a + a2 = 10º + 1120º + 1240º = 0

FEATURE


















58 • SPRING 2012

FEATURE



VA = V0A + V1A + V2A (1)

VB = V0B + V1B + V2B (2)

VC = V0C + V1C + V2C (3)

V1a = V2a

V1B = a2 ∙ V1A V1C = a ∙ V1A (4)

V2B = a ∙ V2A V2C = a2 ∙ V2A

VA = V0A + V1A + V2A (5)

VB = V0A + a2 ∙ V1A + a ∙ V2A (6)

VC = V0A + a ∙ V1A + a2 ∙ V2A (7)

V0A = ⅓ (VA + VB + VC) (8)

V1A = ⅓ (VA + a ∙ VB + a2 ∙ VC) (9)

V2A = ⅓ (VA + a2 ∙ VB + a ∙ VC) (10)



a = 1120º.

1 + a + a2 = 10º + 1120º + 1240º = 0

FEATURE


















58 • SPRING 2012

FEATURE



VA = V0A + V1A + V2A (1)

VB = V0B + V1B + V2B (2)

VC = V0C + V1C + V2C (3)

V1a = V2a

V1B = a2 ∙ V1A V1C = a ∙ V1A (4)

V2B = a ∙ V2A V2C = a2 ∙ V2A

VA = V0A + V1A + V2A (5)

VB = V0A + a2 ∙ V1A + a ∙ V2A (6)

VC = V0A + a ∙ V1A + a2 ∙ V2A (7)

V0A = ⅓ (VA + VB + VC) (8)

V1A = ⅓ (VA + a ∙ VB + a2 ∙ VC) (9)

V2A = ⅓ (VA + a2 ∙ VB + a ∙ VC) (10)



a = 1120º.

1 + a + a2 = 10º + 1120º + 1240º = 0

FEATURE


















58 • SPRING 2012

FEATURE



VA = V0A + V1A + V2A (1)

VB = V0B + V1B + V2B (2)

VC = V0C + V1C + V2C (3)

V1a = V2a

V1B = a2 ∙ V1A V1C = a ∙ V1A (4)

V2B = a ∙ V2A V2C = a2 ∙ V2A

VA = V0A + V1A + V2A (5)

VB = V0A + a2 ∙ V1A + a ∙ V2A (6)

VC = V0A + a ∙ V1A + a2 ∙ V2A (7)

V0A = ⅓ (VA + VB + VC) (8)

V1A = ⅓ (VA + a ∙ VB + a2 ∙ VC) (9)

V2A = ⅓ (VA + a2 ∙ VB + a ∙ VC) (10)



a = 1120º.

1 + a + a2 = 10º + 1120º + 1240º = 0

FEATURE


A ground fault current can be considered a zerosequence current as it passes through the system to reach the location of the ground fault. (See Figure 2.) This zero-sequence current may be supplied from different sources and delivered to the particular zone involv-ing ground fault. A zero-sequence current also can be caused by an imbalance in impedances and sources within the power system that has multiple groundings or it could be a triplin (3, 6, 9, 12, etc.) harmonic current. These currents would not necessarily in-volve a ground fault. In these circumstances, we would not want ground fault protective relays to cause false tripping.

With numerical protective relays, protection against a phase-to-ground fault is less difficult. A ground fault protective relay must detect all phase-to-ground faults within its defined zone of protec-tion. The well-known method of detecting ground faults is to use a numerical relay that responds only to the zero-sequence current of the system.

The relay simply measures the sum of the three-phase currents. See equation (17). Under balanced conditions, the sum equates to zero. Note that zero-sequence current is also known as residual current.

Ground fault protection elements should not be set more sen-sitive than the normal system unbalance. There are two types of current transformer connections to measure zerosequence current in the system as shown in Figure 3 and Figure 4.

FIGURE 3: Residual CT connection

The setting of the ground fault protective relay in Figure 3 should be above the line maximum unbalance current to avoid nui-sance tripping and to accommodate for initial asymmetrical motor starting currents.

FIGURE 4: Core-balance CT connection

The connection in Figure 4 is also known as flux summation. The conductor cables pass through the center hole of the core-balance current transformer. Unbalanced flux in the A, B, and C conductors will induce current flow in the secondary of the CT. This produces secondary current in the system known as zerose-quence current.

ZERO-SEQUENCE VOLTAGE FOR GROUND FAULT PROTECTION IN UNGROUNDED SYSTEM

The available fault current for a single-phaseto- ground fault is very limited for ungrounded systems. That makes it difficult to detect zerosequence current in a ground fault protective relay that responds to only zero-sequence current. In this case, sensitive voltage relays can be used to detect ground fault where the fault currents are very small.

A set of voltage transformers is wired with a grounded wye on a primary side and a broken delta on a secondary side. The 59N element of the relay is connected across the broken delta. It is nec-essary to connect a resistor across the broken delta to avoid fer-roresonance. See Figure 5.

FIGURE 5. Zero-sequence voltage components used for ground fault protection in ungrounded 3-phase, 3-wire system


Grounded wye/broken delta voltage transformers act as zero-sequence filters. The relay measures the sum of the three-phase voltages. See equation (18). Under balanced conditions, the sum equates to zero. When a ground fault occurs, the relay will detect the presence of the secondary zero-sequence voltage (3V0).

NEGATIVE-SEQUENCE CURRENT FOR MOTOR AND GENERATOR PROTECTIONIn the case of motors, the primary cause of motor failure is excessive heat. If sustained over a long period of time, this excessive heat condition will result in motor burn out and also reduce the life span of the motor. Usually, overheating is caused by overcurrent conditions, which can be overloads, locked-rotor conditions, undervoltage, phase failure, repetitive starts, or phase unbalance.

Unbalance in the feeder phase voltages or motor winding impedance causes unbalanced currents to flow to the motor. Current unbalance is a major factor in temperature increase in motors. Current unbalance in a motor can be represented by the presence of an excessive negative-sequence component in the motor current. The negative-sequence current (I2) from unbalance causes rotor heating and additional copper losses in the stator windings. It is necessary to protect motors against excessive negative-sequence overcurrent.

Current unbalance (46) measurement in numerical protective relays is easy to implement. Measuring algorithms includes the true negative-sequence measurement and the difference between the maximum and minimum phase currents. The current unbalance measuring elements have an I2

2 • t = K characteristic that makes the time delay settings easier to apply than with the voltage element. The worst-case unbalance occurs for an open phase at full load. In this case, the negative-sequence current equates to the positive-sequence current. If K = 40, the time dial should be set to cause tripping

in 40 seconds. It is common to use current unbalance to provide an alarm first so that corrective action can be taken before removing the motor from service.

There is an inverse-time overcurrent element (51) that uses negative-sequence current as the operating quantity. The curve characteristic should be a straight line on log-log scales, corresponding to I2

2 • t = K and should be set to match the motor running mode characteristic.

For generators, protection is based on the same principle as motors where excessive heating from unbalanced stator currents will result in generator damage. In addition, the negative-sequence current produced by unbalanced conditions induces double-frequency currents into the rotor, causing overheating. The continuous unbalance current capability of a cylindrical-rotor synchronous generator is defined in IEEE/ANSI C50.13 and of a salient pole generator is defined in IEEE/ANSI C50.12, as shown in Table 1.

There are several causes of unbalanced three-phase currents in a generator. The most common causes are system asymmetries, unbalanced loads, unbalanced system faults, and open circuits. The highest source of negative-sequence current is the generator phase-to-phase fault.

60 • SPRING 2012


Ground fault protection elements should not be set more sensitive than the normal system unbalance. There are two types of current transformer connections to measure zero-sequence current in the system as shown in Figure 3 and Figure 4.

The setting of the ground fault protective relay in Figure 3 should be above the line maximum unbalance current to avoid nuisance tripping and to accommodate for initial asymmetrical motor starting currents.

The connection in Figure 4 is also known as flux summation. The conductor cables pass through the center hole of the core-balance current transformer. Unbalanced flux in the A, B, and C conductors will induce current flow in the secondary of the CT. This produces secondary current in the system known as zero-sequence current.

ZERO-SEQUENCE VOLTAGE FOR GROUND FAULT PROTECTION IN UNGROUNDED SYSTEMThe available fault current for a single-phase-to-ground fault is very limited for ungrounded systems. That makes it difficult to detect zero-sequence current in a ground fault protective relay that responds to only zero-sequence current. In this case, sensitive voltage relays can be used to detect ground fault where the fault currents are very small.

A set of voltage transformers is wired with a grounded wye on a primary side and a broken delta on a secondary side. The 59N element of the relay is connected across the broken delta. It is necessary to connect a resistor across the broken delta to avoid ferroresonance. See Figure 5.

Figure 5. Zero-sequence voltage components used for ground fault protection in ungrounded 3-phase, 3-wire system

FEATURE


Table 1. Continuous Unbalance Current Capability of Generators

Type of Generator Permissible I2 (% of stator rating)

Salient Pole With connected amortisseur windings 10 With non-connected amortisseur windings 5 Cylindrical Rotor Indirectly cooled 10 Directly cooled (to 960 MVA) 8 Directly cooled (961 to 1200 MVA) 6 Directly cooled (1201 to 1500 MVA) 5

3I0 = IA + IB + IC (17)

3V0 = VA + VB + VC (18)

Figure 3: Residual CT connection

Figure 4: Core-balance CT connection

FEATURE


Grounded wye/broken delta voltage transformers act as zero-sequence filters. The relay measures the sum of the three-phase voltages. See equation (18). Under balanced conditions, the sum equates to zero. When a ground fault occurs, the relay will detect the presence of the secondary zero-sequence voltage (3V0 ).

NEGATIVE-SEQUENCE CURRENT FOR MOTOR AND GENERATOR PROTECTION

In the case of motors, the primary cause of motor failure is ex-cessive heat. If sustained over a long period of time, this excessive heat condition will result in motor burn out and also reduce the life span of the motor. Usually, overheating is caused by overcur-rent conditions, which can be overloads, locked-rotor conditions, undervoltage, phase failure, repetitive starts, or phase unbalance.

Unbalance in the feeder phase voltages or motor winding im-pedance causes unbalanced currents to flow to the motor. Current unbalance is a major factor in temperature increase in motors. Cur-rent unbalance in a motor can be represented by the presence of an excessive negative-sequence component in the motor current. The negativesequence current (I2) from unbalance causes rotor heating and additional copper losses in the stator windings. It is necessary to protect motors against excessive negative-sequence overcurrent.

Current unbalance (46) measurement in numerical protective relays is easy to implement. Measuring algorithms includes the true negativesequence measurement and the difference between the maximum and minimum phase currents. The current unbal-ance measuring elements have an I2

2 • t = K characteristic that makes the time delay settings easier to apply than with the voltage element. The worst-case unbalance occurs for an open phase at full load. In this case, the negative-sequence current equates to the positive-sequence current. If K = 40, the time dial should be set to cause tripping in 40 seconds. It is common to use current unbal-ance to provide an alarm first so that corrective action can be taken before removing the motor from service.

There is an inverse-time overcurrent element (51) that uses neg-ative-sequence current as the operating quantity. The curve char-acteristic should be a straight line on log-log scales, corresponding to I2


For generators, protection is based on the same principle as mo-tors where excessive heating from unbalanced stator currents will result in generator damage. In addition, the negativesequence cur-rent produced by unbalanced conditions induces double-frequency currents into the rotor, causing overheating. The continuous unbal-ance current capability of a cylindrical-rotor synchronous genera-tor is defined in IEEE/ANSI C50.13 and of a salient pole generator is defined in IEEE/ANSI C50.12, as shown in Table 1.

TABLE 1: Continuous Unbalance Current Capability of Generators

There are several causes of unbalanced threephase currents in a generator. The most common causes are system asymmetries, unbalanced loads, unbalanced system faults, and open circuits. The highest source of negative-sequence current is the generator phase-to-phase fault.

The short time (unbalanced fault) negativesequence capability of a generator is defined in ANSI C50.13 as shown in Table 2 be-low.

TABLE 2: Short Time Negative-Sequence Capability of

Generators

Modern numerical protective relays provide negative-sequence inverse-time protection using an I2

2 • t characteristic shaped to match the short-time withstand capability of the generator. The relay time units can be set to protect generators with K values of 10 or less. An alarm setting associated with these relays can provide detection for negative-sequence currents as small as 0.3 percent of a generator’s rating. The trip pickup can be set at the continuous negativesequence capability of the generator operating at full output. Normally, the protection against negativesequence current is connected to trip the main generator breaker(s) and the field breaker. An alarm should be provided for indicating when the generator’s continuous capability is exceeded.

SUMMARY

Modern numerical relays include a number of advantages and functions associated with the simple and accurate calculation of symmetrical components from phase quantities. Hardware se-quence filters in predecessor technologies have been replaced with simple mathematical techniques that implement the equations for symmetrical components with the mathematical capabilities of the microprocessor and the digitized quantities. In particular, zero-sequence and negative-sequence based relay elements extend protection options for finding fault direction and providing both ground fault and motor/generator protection.












60 • SPRING 2012








FEATURE





3I0 = IA + IB + IC (17)

3V0 = VA + VB + VC (18)



FEATURE












60 • SPRING 2012








FEATURE





3I0 = IA + IB + IC (17)

3V0 = VA + VB + VC (18)



FEATURE

62 • SPRING 2012 Modern Protective relay techniques: using a 94-year-old concePt to Protect electrical equiPMent

2200 NORTHWOOD AVE, EASTON, PA 18045 USA(610) 515-8775 • Fax - (610) 258-1230

For our complete inventory visit our website: www.belyeapower.com

The short time (unbalanced fault) negative-sequence capability of a generator is defined in ANSI C50.13 as shown in Table 2 below.

Modern numerical protective relays provide negative-sequence inverse-time protection using an I2

2 • t characteristic shaped to match the short-time withstand capability of the generator. The relay time units can be set to protect generators with K values of 10 or less. An alarm setting associated with these relays can provide detection for negative-sequence currents as small as 0.3 percent of a generator’s rating. The trip pickup can be set at the continuous negative-sequence capability of the generator operating at full output.

Normally, the protection against negative-sequence current is connected to trip the main generator breaker(s) and the field breaker. An alarm should be provided for indicating when the generator’s continuous capability is exceeded.

SUMMARYModern numerical relays include a number of advantages and functions associated with the simple and accurate calculation of symmetrical components from phase quantities. Hardware sequence filters in predecessor technologies have been replaced with simple mathematical techniques that implement the equations for symmetrical components with the mathematical capabilities of the microprocessor and the digitized quantities. In particular, zero-sequence and negative-sequence based relay elements extend protection options for finding fault direction and providing both ground fault and motor/generator protection.

Suparat Pavavicharn is a Senior Application Engineer with Basler Electric Company. Pavavicharn is a 1997 graduate of Khon Kaen University, Thailand, with a Bachelor of Science Degree in

Electrical Engineering. She also received a Master of Science in Sustainable Energy and Management from Flensburg University, Germany, in 2003 and is presently working towards a Masters in Business Administration at Webster University. Pavavicharn brings 14 years of experience in power station, electric utility, and industrial sectors. Her experience over the last eight years has focused on all facets of protection including design, fault studies, setting and testing, and on-site commissioning.

Table 2. Short Time Negative-Sequence Capability of Generators

Type of Generator K (permissible I22∙t) Salient Pole 40 Cylindrical rotor Indirectly cooled 30 Directly cooled (0 – 800 MVA) 10 Directly cooled (801 – 1600 MVA) 10 – (0.00625)(MVA-800)

FEATURE


Suparat Pavavicharn is a Senior Application Engineer with Basler Electric Company. Pa-vavicharn is a 1997 graduate of Khon Kaen University, Thailand, with a Bachelor of Sci-ence Degree in Electrical Engineering. She also received a Master of Science in Sustain-able Energy and Management from Flensburg

University, Germany, in 2003 and is presently working towards a Masters in Business Administration at Webster University. Pa-vavicharn brings 14 years of experience in power station, electric utility, and industrial sectors. Her experience over the last eight years has focused on all facets of protection including design, fault studies, setting and testing, and on-site commissioning.


Multi-function protection relays calculate and use symmetrical components to enhance their performance during system faults. Three examples are presented in this article:

• Zero-sequence current elimination for transformer differential protection

• Positive-sequence voltage polarization for mho phase distance elements

• Negative-sequence current detection to inhibit out-of-step blocking

ZIGZAG TRANSFORMER INSIDE TRANSFORM-ER DIFFERENTIAL ZONE

A shunt-connected zigzag transformer provides a zero-sequence current sink forungrounded systems such as that shown in Fig-ure 1 by establishing a connection fromground to neutral. Zero-sequence current (I0) flows up through the neutral of the zigzag transformer during ground faults. Therefore, it is simple to apply non-directional overcurrent protection for the detection of single line-to-ground faults. If the system is left ungrounded, high volt-age appears on the unfaulted phases during single line-to-ground faults and conventional ground overcurrent protection is useless.

A zigzag transformer consists of three 1:1 ratio transformers. Each leg of the zigzag trans-former consists of two windings that are 120 degrees out of phase. Windings are wound around the core such that zero-sequence cur-rent flows through the bank when there is system unbalance (i.e., ground fault). Only exciting cur-rent flows through a zigzag trans-former during balanced system conditions. The grounding trans-former appears as the leakage re-actance of the core when a ground fault occurs and zero-sequence current flows splitting evenly into the three phases.

FIGURE 1: Zigzag Transformer Three-Line Diagram

A grounding resistor is sometimes used since the transformer alone equates to reactance grounding. Typically, the zigzag trans-former is sized such that its impedance is 100 percent on its own base. For the 10-second rating, 400 amps primary is commonly applied in the United States.

FIGURE 2: Ungrounded System with Zigzag Transformer for Ground Current

If the zigzag transformer is located inside the zone of transform-er differential protection,such as for the deltaconnected windings shown in Figure 2, then the zero-sequence current contribution during external ground faults must be eliminated or else a misop-eration can occur. Here is one method in which numerical trans-former protection relays can reliably remove the ground current.

The currents shown are taken directly from the CT secondary and have been divided by the tap setting for the delta winding to convert them into per unit. If zero-sequence current elimination is selected (see Figure 3 as an example), then the relay calculates the ground current as follows:

MULTI-FUNCTION NUMERICAL PROTECTIONRELAYS USING SYMMETRICAL COMPONENTS

FOR MORE RELIABLE AND SECURE PROTECTIONNETA World, Spring 2012 Issue

by Steve Turner, Beckwith Electric Co., Inc.

Figure 5 shows the A-Phase-to-B-Phase loop measurement for a phase mho distance element. Note that there are corresponding loops for B-Phase-to-C-Phase and C-Phaseto-A-Phase as well. The mho phase distance element measures the impedance up to the point of the fault (F) for that loop. IAB and VAB are the faultloop current and fault voltage measured by the relay during the fault.

figUre 5: Phase A-to-Phase B Fault Loop

The corresponding mho phase distance element is generated as follows using two internally-calculated potentials:

An A-Phase-to-B-Phase fault is internal with respect to the mho element when the absolute value of Ѳ is 90 electrical degrees or less (see Figure 6).

VAB1 is the voltage recorded prior to the fault (i.e., pre-fault) and is stored in a buffer. If there is no reference signal (e.g., bolted three-phase fault at the secondary terminals of the voltage transformer), then the mho distance element can misoperate during an external fault. This is one of the main reasons for using pre-fault memorized voltage as the polarizing quantity. Note that only positive-sequence quantities are present during balanced system conditions such as pre-fault load flow. The length of the buffer to store the memory voltage should be short enough that a skew is not introduced in Ѳ which can also lead to misoperation. Typically 30 cycles is sufficient. Ѳ is the angle between the two voltage signals and whenever this is equal to 90°, it describes a point along a circle as shown in Figure 7. Some other advantages to using positive-sequence memory voltage as the reference signal are as follows:

•Greaterexpansionofthemhocharacteristicalong the resistive axis •Reliableoperationforclose-inzerovoltagefaults •Securityforclose-inreversephase-to-phasefaults •Securityduringsinglepoletripping

Figure 7 shows both the static and expanded dynamic mho distance operating characteristics in the voltage plane which is obtained by multiplying the impedance diagram by the fault current. Expansion along the resistive axis is a direct function of the apparent source impedance behind the relay; i.e., the weaker the source, the greater the expansion.

figUre 7: Static and Expanded Mho Distance Operating Characteristics

A grounding resistor is sometimes used since the transformer alone equates to reactance grounding. Typically, the zigzag transformer is sized such that its impedance is 100 percent on its own base. For the 10-second rating, 400 amps primary is commonly applied in the United States.

figUre 2: Ungrounded System with Zigzag Transformer for Ground Current

If the zigzag transformer is located inside the zone of transformer differential protection,such as for the delta-connected windings shown in Figure 2, then the zero-sequence current contribution during external ground faults must be eliminated or else a misoperation can occur. Here is one method in which numerical transformer protection relays can reliably remove the ground current.


figUre 3: Zero-Sequence Current Elimination

Figure 4 shows the transformer differential operating characteristic. Point A is the filtered operating point for an external ground fault and Point B is unfiltered.

figUre 4: Transformer Differential Operating Characteristic

disTance ProTecTion Pre-faUlT PosiTive-seqUence memory volTage PolariZaTionDistance elements applied for the protection of high-voltage transmission lines often use positive-sequence voltage as the polarizing signal or reference. These two equations show a classic method to create a mho distance element:

32 • SPRING 2012 Multi-Function nuMerical Protection relays using syMMetrical coMPonents For More reliable and secure Protection

FEATURE FEATURE

Multi-Function nuMerical Protection relays using syMMetrical coMPonents For More reliable and secure Protection

NETAWORLD • 33

IA = IB = IC

IA = TAP

I 0 IB = TAP

I 0 IC =

TAP

I 0

IG = IA + IB + IC

IG = TAP

I 03

IA’, IB’ and IC’ are the internally-compensated currents:

IA’ = IA + 3GI

IB’ = IB + 3GI

IC’ = IC + 3GI

IA’ = 3GI

- 3GI

= 0 IB’ = 0 IC’ = 0

VAB = IAxZ1L -IBxZ1L IA = -IB

VAB = 2IAxZ1L IAB = IA – IB = 2IA

The loop impedance is calculated as follows:

ZAB = AB

AB

I

V =

A

LA

I

ZxI

2

2 1 = xZ1L

Vop = xZ1LIAB – VAB (Operating Voltage) Vpol = VAB1 (Polarizing Voltage)

Where = Vop - Vpol

Vop = xZ1LIF - VF (Operating Voltage) Vpol = V1 (Polarizing Voltage)

Where

x = per unit reach or per unit distance up to the point of the fault (F) Z1L = positive-sequence line replica impedance or line impedance IF = fault current VF = fault voltage V1 = prefault positive-sequence memory voltage

figUre 6: Mho Distance Element Directional Diagram


















FEATURE FEATURE


NETAWORLD • 33

IA = IB = IC

IA = TAP

I 0 IB = TAP

I 0 IC =

TAP

I 0

IG = IA + IB + IC

IG = TAP

I 03


IA’ = IA + 3GI

IB’ = IB + 3GI

IC’ = IC + 3GI

IA’ = 3GI

- 3GI

= 0 IB’ = 0 IC’ = 0




ZAB = AB

AB

I

V =

A

LA

I

ZxI

2

2 1 = xZ1L


Where = Vop - Vpol


Where




FIGURE 3: Zero-Sequence Current Elimination

Figure 4 shows the transformer differential operating character-istic. Point A is the filtered operating point for an external ground fault and Point B is unfiltered.

FIGURE 4: Transformer Differential Operating Characteristic

DISTANCE PROTECTION PRE-FAULT POSITIVE-SEQUENCE MEMORY VOLTAGE POLARIZATION

Distance elements applied for the protection of high-voltage transmission lines often use positive-sequence voltage as the po-larizing signal or reference. These two equations show a classic method to create a mho distance element:


FIGURE 5: Phase A-to-Phase B Fault Loop


An A-Phase-to-B-Phase fault is internal with respect to the mho element when the absolute value of θ is 90 electrical degrees or less (see Figure 6).

FIGURE 6: Mho Distance Element Directional Diagram

VAB1 is the voltage recorded prior to the fault (i.e., pre-fault) and is stored in a buffer. If there is no reference signal (e.g., bolted three-phase fault at the secondary terminals of the voltage trans-former), then the mho distance element can misoperate during an external fault. This is one of the main reasons for using pre-fault memorized voltage as the polarizing quantity. Note that only positive-sequence quantities are present during balanced system


















FEATURE FEATURE


NETAWORLD • 33

IA = IB = IC

IA = TAP

I 0 IB = TAP

I 0 IC =

TAP

I 0

IG = IA + IB + IC

IG = TAP

I 03


IA’ = IA + 3GI

IB’ = IB + 3GI

IC’ = IC + 3GI

IA’ = 3GI

- 3GI

= 0 IB’ = 0 IC’ = 0




ZAB = AB

AB

I

V =

A

LA

I

ZxI

2

2 1 = xZ1L


Where = Vop - Vpol


Where




















FEATURE FEATURE


NETAWORLD • 33

IA = IB = IC

IA = TAP

I 0 IB = TAP

I 0 IC =

TAP

I 0

IG = IA + IB + IC

IG = TAP

I 03


IA’ = IA + 3GI

IB’ = IB + 3GI

IC’ = IC + 3GI

IA’ = 3GI

- 3GI

= 0 IB’ = 0 IC’ = 0




ZAB = AB

AB

I

V =

A

LA

I

ZxI

2

2 1 = xZ1L


Where = Vop - Vpol


Where




















FEATURE FEATURE


NETAWORLD • 33

IA = IB = IC

IA = TAP

I 0 IB = TAP

I 0 IC =

TAP

I 0

IG = IA + IB + IC

IG = TAP

I 03


IA’ = IA + 3GI

IB’ = IB + 3GI

IC’ = IC + 3GI

IA’ = 3GI

- 3GI

= 0 IB’ = 0 IC’ = 0




ZAB = AB

AB

I

V =

A

LA

I

ZxI

2

2 1 = xZ1L


Where = Vop - Vpol


Where




















FEATURE FEATURE


NETAWORLD • 33

IA = IB = IC

IA = TAP

I 0 IB = TAP

I 0 IC =

TAP

I 0

IG = IA + IB + IC

IG = TAP

I 03


IA’ = IA + 3GI

IB’ = IB + 3GI

IC’ = IC + 3GI

IA’ = 3GI

- 3GI

= 0 IB’ = 0 IC’ = 0




ZAB = AB

AB

I

V =

A

LA

I

ZxI

2

2 1 = xZ1L


Where = Vop - Vpol


Where




conditions such as pre-fault load flow. The length of the buffer to store the memory voltage should be short enough that a skew is not introduced in θ which can also lead to misoperation. Typi-cally 30 cycles is sufficient. θ is the angle between the two volt-age signals and whenever this is equal to 90°, it describes a point along a circle as shown in Figure 7. Some other advantages to using positive-sequence memory voltage as the reference signal are as follows:

• Greater expansion of the mho characteristic along the resistive axis

• Reliable operation for close-in zero voltage faults

• Security for close-in reverse phase-to-phase faults

• Security during single pole tripping

Figure 7 shows both the static and expanded dynamic mho dis-tance operating characteristics in the voltage plane which is ob-tained by multiplying the impedance diagram by the fault current. Expansion along the resistive axis is a direct function of the appar-ent source impedance behind the relay; i.e., the weaker the source, the greater the expansion.

FIGURE 7: Static and Expanded Mho Distance Operating CharacteristicsNEGATIVE-SEQUENCE CURRENT DETECTION TO INHIBIT PROTECTION

Certain protection functions are typically only intended to oper-ate during balanced three-phase conditions:

• Rate-of-change-of-frequency (81R)

• Loss-of-field (40)

• Out-of-step tripping|blocking (78)

• Under-voltage load shedding

If negative-sequence current (I2) is detected, this is an indi-cation that an unbalanced disturbance is in progress. Figure 8 shows the oscillographic record captured by a relay for the sim-ulation of a large generator that slipped three poles; the event then evolved into a resistive ground fault. Figure 9 shows two large generating units at one plant that are interconnected one substation away via the transmission system. If either machine was to go unstable, then the desired sequence of events is to trip only the runaway generator while blocking the distance relays protecting the local transmission line terminals. This criterion prevents cascading outages from occurring and preserves the in-tegrity of offsite power sources to the plant. For this particular case, the desired end result would be to unblock the distance protection so as to clear the ground fault as shown in Figure 9.

Figure 10 shows the trajectory of an unstable swing that passes through both the generator out-of-step (OST) tripping charac-teristic and a transmission line relay Zone 1 operating charac-teristic. Note that the point O in Figure 10 corresponds to the generator voltage transformer terminals. Such a swing could operate both the line protection and the generator outof- step tripping function–thereby illustrating one case why power swing blocking is required for the transmission line protection. Phase distance protection can also operate during a stable swing–also shown in Figure 10.

FIGURE 8: Pole Slip Evolves into Single Line-to-Ground Fault

Figure 11 is a simple logic diagram that illustrates how to use negative-sequence current detection (46) to unblock the distance protection (21) after it has been inhibited by power swing blocking logic (78_PSB).


FIGURE 9: Two Large Generating Units Interconnected via the Transmission Grid

FIGURE 10: Unstable Swing passes through Zone 1 and OST

FIGURE 11: Negative-Sequence Current Unblocking

Another post-swing disturbance that is even harder to detect is if a three-phase fault were to occur on one of the transmission lines terminated at the plant following the first pole slip since there is no negative-sequence current present. Figure 12A shows a typical power swing blocking (PSB) characteristic. If the im-pedance trajectory passes relatively slowly (e.g., three cycles or more) through the outer characteristic to the inner characteristic, then a power swing is detected and the phase distance protection is blocked. If there is a fault, then the measured impedance will quickly (e.g., two cycles or less) jump from the pre-fault location to a point on the faulted power system. A pair of inner blinders is

required to detect when a swing evolves into a three-phase fault as shown in Figure 12B. If a balanced line fault occurs, then the mea-sured impedance drops in between the blinders and remains there. Thus, the blinders can detect this condition and quickly unblock the phase distance protection.

FIGURE 12A: Power Swing Blocking Characteristic

FIGURE 12B: Power Swing Blocking using Inner Blinder

Out-of-step tripping and blocking characteristics operate on the measured positivesequence impedance since a power swing is a balanced three-phase phenomenon. The positivesequence imped-ance can be calculated as follows:

Another post-swing disturbance that is even harder to detect is if a three-phase fault were to occur on one of the transmission lines terminated at the plant following the first pole slip since there is no negative-sequence current present. Figure 12A shows a typical power swing blocking (PSB) characteristic. If the impedance trajectory passes relatively slowly (e.g., three cycles or more) through the outer characteristic to the innercharacteristic, then a power swing is detected and the phase distance protection is blocked. If there is a fault, then the measured impedance will quickly (e.g., two cycles or less) jump from the pre-fault location to a point on the faulted power system. A pair of inner blinders is required to detect when a swing evolves into a three-phase fault as shown in Figure 12B. If a balanced line fault occurs, then the measured impedance drops in between the blinders and remains there. Thus, the blinders can detect this condition and quickly unblock the phase distance protection.

Out-of-step tripping and blocking characteristics operate on the measured positivesequence impedance since a power swing is a balanced three-phase phenomenon. The positive-sequence impedance can be calculated as follows:

conclUsionsThis article presented several examples that illustrate how multi-function protection relays calculate and use symmetrical components to enhance their performance during system faults. The use of symmetrical component quantities help to provide protection that is more reliable and secure.

Steve Turner is a Senior Applications Engineer at Beckwith Electric Company. His previous experience includes working as an application engineer with GEC Alstom, an application engineer in the international market for SEL, focusing on transmission line protection applications. Steve worked for Progress Energy, where he developed a patent for double-ended fault location on transmission lines.Steve has both a BSEE and MSEE from Virginia Tech University. He has presented

at numerous conferences including: Georgia Tech Protective Relay Conference, Western Protective Relay Conference, ECNE and Doble User Groups, as well as various international conferences. Steve is a senior member of IEEE.

36 • SPRING 2012

FEATURE


RENT RELAY TESTERSFROM ELECTRO RENT

Megger SMRT36

Doble F6150

Manta MTS-5000

Noram-SMCMentor 12-36

800-304-4744 Visit Our New Informative Website! Industrial.ElectroRent.com

• Cost effective daily, weekly and monthly rentals.

• Electro Rent’s inside application representatives have 15+ years average experience and are ready to help you get what you want.

Delivering Results You Can Trust Since 1965

Large inventory includes equipment from motor testers to borescopes

V1 = 3

1(Va + aVb + a2Vc)

I1 = 3

1(Ia + aIb + a2Ic)

Z1 = 1

1

I

V

Where

a = 1 120 a2 = 1 240

figUre 12b: Power Swing Blocking using Inner Blinder

figUre 12a: Power Swing Blocking Characteristic


CONCLUSIONS This article presented several examples that illustrate how multi-

function protection relays calculate and use symmetrical compo-nents to enhance their performance during system faults. The use of symmetrical component quantities help to provide protection that is more reliable and secure.

Steve Turner is a Senior Applications Engineer at Beckwith Electric Company. His previous experience includes work-ing as an application engineer with GEC Alstom, an application engineer in the in-ternational market for SEL, focusing on transmission line protection applications. Steve worked for Progress Energy, where he developed a patent for double-ended fault location on transmission lines.

Steve has both a BSEE and MSEE from Virginia Tech Univer-sity. He has presented at numerous conferences including: Geor-gia Tech Protective Relay Conference, Western Protective Relay Conference, ECNE and Doble User Groups, as well as various international conferences. Steve is a senior member of IEEE.


PRO

TECT

IVE

RELA

Y

HANDBOOK

Published By

VOLUME 3 SERIES II

IntellirentSponsored by

PROTEC

TIVE RELA

Y HA

ND

BOO

K VOLU

ME 3 SERIES II

Published by NETA

- The InterNational Electrical Testing A

ssociation

PRealy V3 Cover.indd 1 2/6/14 11:07 AM


T 972.317.0479 | www.intellirentco.com | contact us at [email protected]

Test

Equipment

Rental

Shermco Industries recently completed a project that upgraded multiple generator control systems and main and feeder breaker control relays as well as provided an upgraded Supervisory Con-trol and Data Acquisition (SCADA) system for an electric mu-nicipal utility. To meet customer requirements, the system had to be capable of communicating with and controlling several dif-ferent protective devices while retaining the potential to expand in the future. In addition to communicating with various protec-tive devices for data input, the SCADA system also had to be user friendly with control being accomplished through a simple touch-screen monitor.

SYSTEM DESCRIPTION The distribution system for this electric municipal utility con-

sists of one main 15 kV distribution bus fed by a 69 kV to 15 kV transformer through a main breaker. The main bus is also fed individually by multiple generator sets, each with separate gen-erator controls and a generator breaker. Feeder breakers are also tied to the bus for distributing electricity to the town. The two buses then distribute power to the community via 15 kV feeders. Adding complexity to the system is the recent addition of a single wind turbine to one of the distribution feeders. For this project, the utility required the capability of automating the 15 kV main and feeder breakers, as well as the various generators, through the SCADA system while using the SCADA system as a monitoring system for their generation capability.

COMMUNICATIONS DESCRIPTION The overall system architecture consists of a variety of SEL

relays that control and monitor the 15 kV class breakers. Ad-ditionally, a combination of SEL relays and Woodward genera-tor controls are provided for the generators including the wind turbine. To facilitate communication with all the devices, an Ethernet switch and an SEL Real-Time Automation Controller (RTAC) were installed in a star pattern (see Figure 1) for the system communication. By configuring the system in this star pattern, each relay has its own channel to the system communi-cations controller as opposed to a multidrop network where each node must wait in turn to send data. Utilizing the RTAC with IEC 61850 protocol, each SEL relay can broadcast information to the entire SCADA system, thus enhancing data access throughput and speeding up response times.

Figure 1: Typical System Communication Diagram

The SEL RTAC provides a real-time operating system with the functionality of sophisticated communication and data han-dling required for advanced power system integration projects. The RTAC features secure communications, advanced data concentration, high-speed logic processing, flexible engineer-ing access, and protocol conversion capabilities between mul-tiple built-in client/server protocols. The RTAC also gives in-tegrators the necessary tools to easily integrate and concentrate information from the wide variety of microprocessor-based de-vices found in today’s substations.

Breaker protection and control is provided using the SEL-451 relay for the main breaker and SEL-751A relays for the feeder breakers. The SEL-451 relay has capabilities to monitor power, including thermal or rolling interval demand, as well as peak demand on positive-, negative-, and zero-sequence current. It also provides sufficient programming capabilities to eliminate the need for a separate programmable logic controller (PLC) to control the various operating scenarios associated with the generation system. The SEL-751A relays provide complete feeder protection with overcurrent, overvoltage, undervoltage, and frequency elements. Both relays also accommodate Eth-ernet-based communication with IEC 61850 communication protocol for the SCADA system. The SEL-451 accommodates hard-wired digital and analog inputs to communicate with the Woodward LS-4 device.

NEWER SOLID-STATE RELAYSOFFER ENHANCED FLEXIBILITY FOR SCADA

SOLUTIONS IN THE MUNICIPAL ENVIRONMENTNETA World, Spring 2012 Issue

by Lynn Hamrick and Owen Wyatt, Shermco Industries


Generator protection and control is provided using a combi-nation of SEL-700G relays and Woodward’s GCP-30 series of generator controllers. The SEL-700G provides a complete pro-tection and synchronization solution for synchronous generators. They also accommodate Ethernet-based communication with IEC 61850 communication protocol for the SCADA system. The SEL-700G also accommodates hard-wired digital and analog inputs to communicate with the Woodward generator controllers.

For coordinated generator operation and control, the Woodward LS-4 was utilized. The LS-4 is a breaker control and protection module that enables the user to control, synchronize, and protect the main bus in multiple generator applications. It accommodates an automatic adjustment in frequency, voltage, and load for multi-ple GCP-30 series generator controllers in a networked application using CANBus communication. Woodward’s GCP-30 Series gen-erator controllers coordinate information with the engine control-lers and provide load management features including automatic base loading/peak shaving, import/export control, and emergency power/backup power generation.

HUMAN MACHINE INTERFACE (HMI) Operator interface and control is provided either through the

front panel of each device or through a human machine interface (HMI) that consists of a PC-based workstation. The workstation also interfaces and communicates with the SCADA system via Ethernet connection. The HMI package that was implemented was the Wonderware InTouch software system. Utilizing a tool inside the Wonder-Ware InTouch software, an electrical historian is inte-grated into the SCADA system for each feeder and main intercon-nection point that records the amperes, volts, kVA, kW and kvar. A method of manipulating the historian chart has been developed that allows the user to zoom in and out of specific sections of time, allowing the characteristics of each feeder to be analyzed over minutes, days, and even weeks.

As part of the control features of the SCADA system, the ability to operate the breakers and circuit generators is also included on the individual breaker screens. After selecting a distribution system device and requesting a device operation to trip or close the break-er, the SCADA system queries the user to go through a verification process prior to actually operating the device to mitigate accidental operations. Once verified that the operation is intended, the SCA-DA master then sends a permissive command using IEC 61850

communication protocol and the associated relay initiates the op-eration when the protective features, such as synch-check, will al-low. This utility chose to also provide automated synch-check for their generation breakers; therefore, a permissive-to-close bit was developed within the generator breaker relays to ensure that the re-lays will not initiate a close out-of-synchronization or into an exist-ing fault. The SCADA system is a very effective way to operate de-vices without having personnel directly next to the device while it is operating. This eliminates the need for a remote operating panel.

Another feature added to the system utilized the alarming capabil-ity included within the SEL relays themselves. When an alarm condi-tion occurs, the SCADA system flags the alarm and alerts the utility’s personnel of the situation. By looking at the main one-line screen of the SCADA system, the utility’s personnel can quickly identify the specific feeder and relay that indicated the alarm condition. As an example, the SCADA alarms are triggered in the event of a breaker tripping due to any one of the protective features. With this alarming feature, the utility is made aware of any breaker operations that have occurred while its personnel were not present at the main substations.

Using available remote desktop applications, the system can be in-terrogated and operated remotely through a secure internet connection from a wireless laptop computer. Due to the enhanced capabilities of today’s smart phones, the SCADA system can also be interrogated and operated remotely via cell phone with full function capabilities.

CONCLUSIONSTechnological advances in microprocessor-based protective

relays and generator controllers have resulted in enhanced flex-ibility for implementing SCADA systems for utilities. A SCADA system allows the utility’s personnel to quickly react to transient

situations. The system also assists personnel in performing their daily activities. With the protective relays that have the capability to communi-cate with other devices, data can eas-ily be gathered and analyzed in order to mitigate or respond to problems within the distribution system. Utility personnel can oversee the entire elec-trical grid for a municipality remotely using a laptop or a smartphone.


Lynn Hamrick brings over 25 years of work-ing knowledge in design, permitting, construc-tion, and startup of mechanical, electrical, and instrumentation and controls projects as well as experience in the operation and maintenance of facilities. Lynn is a Professional Engineer, Cer-

tified Energy Manager and has a BS in Nuclear Engineering from the University of Tennessee.

Owen Wyatt is a Level 2 NETACertified Test Technician and is a licensed professional engi-neer in the State of Iowa. Owen has experience in performing design activities associated with electrical substations, protective relay systems, SCADA systems, and electrical infrastructure

systems in accordance with NEC requirements. Owen has also performed numerous power system studies to include fault cur-rent, protective device coordination and arc flash analysis. Addi-tionally, he is experienced in commissioning electrical systems in accordance with NETA specifications. These commissioning ac-tivities include relay testing, medium-voltage switchgear testing, and associated control system testing to NETA specifications.


In the area of protective relay maintenance, it is well understood that electromechanical relays require servicing to remain within expected pickup and operating specifications. Although electro-mechanical relays have protected electrical systems for many years, the physical integrity of relays can deteriorate over time. The required frequency of relay calibration and maintenance de-pends on the type of environment to which the protection is sub-jected. In harsher environments, the induction disc may need to be cleaned more often so as not to impede the electromechanical action of the magnet on the disc and affect the operation of the relay during a fault.

Photo 1: An example of a 50/51 electromechanical relay.

In years past, it was common to calibrate relays and their vari-ous components. This meant adjusting springs, cleaning contacts, and adjusting magnets to ensure the relay operated as designed. This work is still required on existing systems where older, vintage equipment is in service.

With the advent of solid state relays, one might question the value in performing maintenance testing of a protective relay that does not require any form of calibration. Although the manuals for relays from various manufacturers state that specific routine tests are not required, the operation standards for facilities may require regular testing. Most relay manufacturers have various self-testing functions that report any software issues within the relay. As such,

routine relay testing is required for the purposes of validating reli-ability, possible liability, and risk assessment evaluations for insur-ance purposes. The condition of the output contacts of a solidstate relay cannot necessarily be assessed by the relay’s self-checking feature. The self-check programming within a relay looks at in-ternal PCB trace voltages, processor chip checks, etc. These error checks do not analyze the state of a damaged output contact that could either be welded closed from a previous dc arc or a perma-nently open contact due to damage to the relay. Obviously these types of hardware failures can severely compromise the integrity of a protection system, and verification of correct operation of the outputs and inputs of a solidstate relay is well warranted.

Although bench testing of a relay can prove that the protection elements are operating according to manufacturer’s specifica-tions, it is extremely important to perform functionality testing in the field before energizing. Additional testing of relay communi-cations and custom designed control logic can reveal deficiencies in the desired protection scheme. A proactive approach would be to perform testing on the relay to definitively know that all sys-tems are functioning correctly. This is in contrast to assuming that all relay protection systems are operating perfectly and possibly being subjected to a rare relay failure that could remain unnoticed until after a catastrophic fault that was not cleared by the same failed protection relay.

Photo 2: Remember the SR-51? A fabulous machine set up to test a variety of electrical protection components; a bit of a blast from the past for us older techs.

One thing is for sure, relay protection system maintenance is performed considerably differently and more efficiently than in years past. Digital technology has brought about an entirely new

RELAY MAINTENANCE HAS CHANGED,HASN’T IT?

NETA World, Spring 2012 Issue by Kerry Heid, Magna Electric Corp.


way of performing maintenance and in some cases a different phi-losophy as well.

One of the major challenges is to remove multifunction relays from service. Single phase or individual functional relays of elec-tromechanical style can be safely removed with the power system on line. This is not necessarily the case with multifunction relays as every function is being defeated with the relay out of service.

Even with the advent of new technology, the importance of performing protection system maintenance has not changed. As-set management, system reliability, and personnel safety all hinge on the protection scheme operating according to the coordination study requirements. During power system faults, incident energy is in direct proportion to the amount of time the system takes to operate. Incident energy values are determined using an arc-flash hazard analysis and are based on the equipment working accord-ing to the time current characteristics. These TCCs are based on an operating value when the equipment is new from the factory. As time goes on, the maintenance of the protective devices becomes very critical to ensure the operating times found in the study match the times in the field. Service-aged equipment has been found to be unreliable and in some cases even inoperable which creates an obvious variance from the expected operation.

CONCLUSION Relay protection maintenance is critical to ensuring that major

equipment damage and injury to personnel are not encountered. Maintenance also drives reliability. The frequency and type of re-lay maintenance depends on a number of factors including critical-ity, environment, and type of relay system.

Kerry Heid is the President of Magna Elec-tric Corporation, a Canadian based electrical projects group providing NETA certified testing and related products and solutions for electri-cal power distribution systems. Kerry is a past President of NETA and has been serving on its board of directors since 2002. Kerry is chair

of NETA’s training committee and its marketing committee. Kerry was awarded NETA’s 2010 Outstanding Achievement Award for his contributions to the association and is a NETA senior certified test technician level IV.

Kerry is the chair of CSA Z463 Technical committee on Mainte-nance of Electrical Systems. He is also a member of the executive on the CSA Z462 technical committee for Workplace Electrical Safety in Canada and is chair of working group 6 on safety related maintenance requirements as well as a member of the NFPA 70E – CSA Z462harmonization working group.


While preparing an outline and gathering research for an article for the spring addition of World I received an email from one of our senior relay technician at AETCO. In the email was an at-tached Industry Advisory from NERC (North American Electric Reliability Corporation) titled “Reducing Human Performance Er-rors by use of Configuration Control Practices” at first I thought this advisory would be just another research document for the article. Upon further deliberation I felt the real life examples used in this advisory and the control practices recommended for mitigation of human error not only apply to Bulk Power Systems but are good tools to be applied to all facilities within our in-dustry when Acceptance and Maintenance Testing of Relays and protection systems employed on Transmission, Generation and Distribution system industry wide. Instead of paraphrasing the advisory I felt this advisory should be shared in its entirety with the NETA community.

ADVISORY Initial Distribution: November 08, 2011

NERC and the Regional Entities have observed inadequate con-figuration control procedures being employed during Protection System construction or maintenance activities. Entities can further reduce the bulk power system’s (BPS) exposure to these reliability risks by considering these examples and suggested barriers and if warranted, augmenting their existing configuration control prac-tices during construction and maintenance activities. This alert ap-plies only to non-cyber assets.

• Primary Interest Groups:

• System Operators

• System Operators – System Protection

• System Operations—Transmission Engineering

• Generation Engineering

• Transmission Planning

While the vast majority of Protection System construction and maintenance activities take place without negatively impacting the BPS, NERC and the Regional Entities are aware of situations where entities inadequately employed configuration control prac-tices, resulting in unnecessary BPS equipment outages. The impact of these situations highlights the need for improvement in con-

figuration control procedures during Protection System construc-tion and maintenance. Effective configuration control procedures include the evaluation, approval, and management of changes to an established equipment configuration. By developing and imple-menting proper configuration control, entities can reduce exposure to the inherent risk of human performance errors that occur during the maintenance and testing of Protection Systems.

This document does not intend to prescribe or define all aspects of a configuration control program. Instead, it is intended to high-light a few key elements of a configuration control program that, if properly implemented, could have prevented these incidents. The goal is to improve awareness of common industry practices that, if employed, can help reduce the risks associated with the construction and maintenance of Protection Systems. Below are a few real-world examples of incidents that emphasize the need for better configuration control procedures during Protection Sys-tem construction and maintenance. Each of these examples dem-onstrates the risk to BPS reliability when adequate configuration control procedures are not observed.

EXAMPLE 1 A relay technician performing scheduled maintenance on a pro-

tective relay system established the proper clearance and isolation procedures to perform the work. These procedures included open-ing several test switches that would provide an electrical barrier between the isolated equipment and any in-service equipment. Af-ter completion of the work, the technician began restoring the test switches to the closed position; however, he overlooked one of the switches in the process, leaving that test switch open. The open switch happened to block the only trip signal to one of the circuit breakers. The technician released his clearance on the equipment and exited the substation. Sometime later, a fault occurred on the BPS and the open test switch prevented a trip signal from tripping the circuit breaker and from initiating the breaker failure scheme. Because the circuit breaker did not trip, the fault continued to be fed through the closed breaker until it was cleared by remote, time-delayed backup relaying. The result was an undesirable increase in the scope of the BPS equipment outage.

EXAMPLE 2 During a substation project, the construction team failed to use

the latest version of a construction document to complete the in-

INDUSTRY ADVISORY – NERCNorth American Electric Reliability Corporation Reducing Human Performance Errors

By The Use Of Configuration Control Practices

RELAY SAFE WORK PRACTICESNETA World, Spring 2012 Issue

by Scott A. Blizard, American Electrical Testing Co. Inc.


stallation of a protective relay system. The most recent version of the document had incorporated a configuration change to the CT ratio for the protective relays. Because the team used outdated documentation, the incorrect CT ratio was configured for the re-laying. During commissioning, the team failed to detect the er-ror, since their testing reference was to the outdated documents. The Protection System equipment was placed into service with the wrong CT ratio and then sometime later tripped improperly during a system disturbance.

EXAMPLE 3 A relay technician is performing a preventive maintenance ac-

tivity on a transmission line protective relay and makes a tempo-rary setting change in order to perform a calibration check. Upon completion of the work, the technician fails to restore the tempo-rary setting to its original value. The equipment was inadvertently placed back in service with the temporary setting installed. The in-correct setting did not immediately produce an improper response; however, weeks later, the relay operated incorrectly for a fault on an adjacent transmission line. The improper line trip resulted in the outage of more BPS equipment than was necessary to clear the fault.

EXAMPLE 4 A relay technician is working inside a transmission line relay

panel, with the necessary work clearance and equipment isolation already established. The technician discovers a need for some ad-ditional documentation and steps out of the relay panel and walks over to a nearby file cabinet. Upon his return to the relay panel, the technician is distracted and inadvertently enters a different but identical panel with relaying protecting a BPS element that is in service. Unaware that he has entered the wrong panel, the techni-cian resumes working and eventually crosses two wires that sends a transfer trip signal to a remote substation, tripping an in-service 345Kv transmission line.

EXAMPLE 5 A technician accidentally opens the wrong current shorting

switch for one contribution to a differential relay protecting an in-service transformer, causing the transformer to trip.

The above cases are examples of human performance errors that may have been prevented had adequate barriers and configuration control been applied.

Below are examples of some configuration control practices that are being applied in the maintenance and testing of protection sys-tems. Employing these or similar practices can help entities reduce the risks of human performance errors.

• Maintenance Alteration Log (MAL) – A record of all manipulations of equipment during a construction or maintenance activity. This document requires the owner to initial each manipulation once when it is performed, and again when the item is restored to its normal state. Proper use of a MAL could have

prevented the human error incidents in Examples 1 and 3 above.

• Isolation Card – A laminated plastic card placed by technicians on the physical equipment at points of isolation during maintenance or testing activities. Each technician has a personalized set of numbered isolation cards. Individual cards are placed on the physical equipment at points of isolation (e.g. test switch, control switch, or control panel) and often in a one-to-one association with entries on the MAL. After the technician has completed the work, and all items on the MAL have been restored to normal, the full set of isolation cards should have been collected. If cards are missing, the technician works to resolve the discrepancy before releasing his clearance on the equipment. By employing practices that include proper use of a MAL and Isolation Cards, entities can reduce the risk of the human performance incidents such as in Examples 1 and 3.

• Barriers – Colored electrical tape or rubber blankets are examples of soft barriers used to cover or protect exposed, energized components to prevent undesired electrical connections during maintenance. A device used to deter the operation of a control switch during a maintenance activity is an example of a rigid barrier. Soft barriers, such as safety tape, can be used as a visual barrier and placed across the openings of in-service equipment panels during maintenance to help prevent personnel from inadvertently entering these panels during a maintenance activity. Use of visual barriers could have been used to help prevent the technician from inadvertently entering an in-service relay panel as in Example 4.

• Flagging – Signage, safety tape, or any device used to attract the attention of personnel. Flagging can be used to identify equipment that is within the technician’s zone of protection or to identify equipment that is outside the zone of protection. Flagging could have been used to help attract the attention of the technician prior to his entering the wrong relay panel, as in Example 4.

• Controls for distributing project documentation – Revised documentation should be distributed to personnel responsible for the construction, installation and testing, as well as those affected by the change. Old documents should be removed and filed or discarded, as appropriate. After documents have been approved, they should be available at all locations for which they are designated, used, or otherwise necessary, and all obsolete documents should be promptly removed from all points of use to prevent unintended use. Some entities apply such controls by establishing a single source of record for protection information. Proper document distribution controls, including timely distribution of updated documentation and destruction of outdated documentation, could have prevented the incident in Example 2.

• Equipment Isolation List – A detailed list of equipment isolation points used to electrically isolate the equipment under test during a maintenance or construction activity. Examples of items that would appear on an equipment isolation list are individual test switch poles, control switch positions, circuit breakers, etc. A technician should develop an equipment


isolation list and have a peer check it prior to starting the job.

• Peer Review/Peer Check – The peer check is an independent review, by qualified personnel, to validate the technicians’ equipment isolation list. The peer check should be provided by someone other than the technician performing the work or by members of a team that peer check each other’s work. Peer review may be effectively used in conjunction with other practices, such as when relay settings have been modified in the field for testing or installation purposes by downloading or documenting the setting left on the relay and having an independent reviewer compare the setting with the office record.

• Self Check – Self checking is the process of pausing to review one’s own actions prior to executing error-likely tasks. It is a four-step mental process to prevent errors, particularly on critical tasks or an irreversible procedure or step. Using the acronym STAR: Stop and take the time to eliminate external distractions, focus on the task at hand with 100% undivided and focused attention. Think, verify that no critical conditions have changed, consider the impact of your immediate action and question anything that you are have uneasy feelings about or are uncertain. Act, without losing physical or visual contact with the device, remain poised and attentive to your actions. Lastly, Review, verify that you got only the specific results that you expected and wanted. By implementing self-checking skills, the technician could have avoided opening the incorrect test switch in Example 5.

• Place-keeping – A physical marker, either temporary or permanent, that helps one keep his/her place when reviewing sequential lines or columns. Using a straight edge or consistent marking methods, one can mark sequential progress when executing long and detailed procedures. These methods are essential when interruptions or delays prevent fluid movement through a process. Consistent procedures in placekeeping allow smooth transitions and handoffs for events that involve multiple persons to interact on the same or related procedures. Detailed place-keeping also provides a historical record for procedures that occur over extended time periods.

• Pre-job briefing – A pre-job briefing, also referred to as a tailgate or tailboard meeting, is helpful for providing clarity prior to a job start. These are usually carried out by the supervisor or more experienced personnel who understand details of the work and can point out the potential perils personnel may encounter during construction or maintenance. Some entities document the pre-job briefing in writing and have the document signed by each employee or contractor present on the job site. Pre-job briefings may be appropriate on a daily basis or multiple times during the day depending on the complexity of the work being performed. The ES-ISAC estimates that the risk to BPS reliability from this vulnerability is HIGH, due to the daily exposure of the BPS to the adverse consequences of human performance errors during protection system maintenance and testing.

BACKGROUND: The analysis of BPS events frequently identifies human perfor-mance errors during protection system maintenance as a root cause or contributing cause of the event.

CONTACT: Earl Shockley Director of Reliability Risk Management Office: (404) 446-2570 [email protected]

To report any incidents related to this alert, contact: ES-ISAC 24-hour hotline (609) 452-1422 [email protected]

CONCLUSION Prior to testing relays and protection systems, a pre job brief

should be performed outlining not only the work to be performed, identify all hazards associated with the work, identify circuits that will be affected and identify the proper level of PPE required to be worn in accordance with NFPA 70E to perform the work. Test for the presence of current, ac voltage and dc voltages prior to servicing equipment.

Exercise extreme caution when performing modifications, maintenance, and testing in current transformer secondary cir-cuits. Current transformers act as constantcurrent sources to what-ever load is applied on the secondary. This means that the voltage changes to provide the same current, no matter what the imped-ance is in the secondary circuit. When the secondary is open cir-cuited the voltage becomes extremely large. This high voltage may destroy the insulation, causing a fault that can destroy the CT, damage other equipment, and be hazardous to personnel. Ex-treme care must be taken to ensure that a reasonable secondary burden is always present or that the CT secondary’s have been shorted to prevent high voltages when the primary is energized.

Be safe; when in doubt, always err on the side of caution.

Scott Blizard is the current Vice President-Chief Operations Offi-cer and the former head of Safety for American Electrical Testing Co. Inc. Scott is a master electrician and a NETA level 4 test tech-nician with over 30 years of experience in the electrical industry.


LINE IMPEDANCE TESTING Between 80-90 percent of all power system faults involve

ground. Many protective relaying schemes depend on ground dis-tance protection to accurately sense and locate ground faults on multiterminal subtransmission and transmission lines. In addition to the need for dependable ground fault detection, protective re-laying must provide adequate selectivity to avoid ovetripping for faults outside of its zone of protection and other undesired conse-quences, such as undertripping or unintended automatic reclosing initiation.

The problem has become more apparent due to recent major power system disturbances in North America such as the North-east blackout of 2003. Correct application and setting of protective devices, particularly distance relays, have become subject to heavy scrutiny lately. Validation of accurate distance relay settings is now a major topic of discussion by electric power utilities as well as professional technical committees such as the IEEE Power Sys-tems Relaying Committee. It becomes apparent very quickly that the accuracy of line parameter values may affect many people.

Although ground distance relay design, charac-teristics, and implementations vary, some of the typical parameters required to set a ground dis-tance relay include the following:

• Zone impedance reach and characteristic angle

• Blinder positions, resistive reaches, and angles

• Directional supervision limiting angle

• Polarizing current (3I0, I2)

• Supervising element (3I0)

• Z0/Z1 (zero-sequence compensation)

• Z0M/Z1 (zero-sequence mutual coupling compensation

Relay manufacturers have di°erent methods of calculating zero-sequence compensation, also known as the k factor, but generally it is deÿned as the ratio between the zero-sequence imped-ance Z0 and the positive-sequence impedance Z1 of a given transmission line. The k factor is used to correct the ground impedance calcu-la-tion so that the ground fault loop calculation can be simpliÿed and treated similarly to the phase-to-phase fault loop calculations performed in the protective device. Therefore, if the k factor is not accurate, fault reach (distance) will be calculated incorrectly. Transmission line impedances used for k factor are o˝en calculated

by line con-stants programs. Due to the large number of vari-ables required, line parameter calculations are prone to error, particular-ly in the zero-sequence impedance value of the line. For example, utili-ties often assume fixed soil resistivity values (10 Ωm, 100 Ωm, etc.) applied across their system models, even in cases where the transmission line may span types of soils different from those assumed in the line constants program. Due to the uncertainties related to soil resistivity and actual transmission tower grounding, the calculation of Z0 of a given line is more susceptible to error than its Z1. This is because the calculation of Z1 is independent of the ground path impedance. For parallel transmission lines, the accurate calculation of zero-sequence mutual impedance Z0M is also prone to the errors described above.

Such errors in the estimation and calculation of line parameters will affect accuracy of settings used in transmission line protective devices, particularly in distance and overcurrent relays, causing them to either underreach or overreach, resulting in a misopera-tion. In order words, relay sensitivity to detect ground faults will be affected.

Additionally, ZO and Zl are used as inputs by many digital re-lays to calculate the location from the line terminal to the fault. Accurate fault location data is needed by utility crews to promptly locate and remove foreign objects from the primary system, and repair damaged lines as quickly as possible. Moreover, short-cir-cuit and coordination studies also depend on accurate modeling

SOLVING RELAY MISOPERATIONS WITH LINE PARAMETER MEASUREMENTS

NETA World, Fall 2012 Issue Will Knapek, OMICRON

Figure 1: OMICRON CPC 100 +CP CU1


data to enable the protection engineer to set relays correctly.

The alternative to line parameter calculation is taking actual measurements on a given transmission line to accurately deter-mine its impedances and k factor. Measuring the line impedance using the correct techniques, equipment, and safety precautions provides the opportunity to eliminate the uncertainties described above. In the past, line parameter measurement was considered prohibitive and costly as it required large, high-power equipment to overcome nominal frequency interferences, since off-nominal frequency injection was not possible. With modern digital technol-ogy and ingenious design, OMICRON has overcome these chal-lenges with the CP CUI coupling unit, an extension to the CPC 100 (Figure 1).

Will Knapek is an Application Engineer far OMICRON electronics Corp, USA. He holds a BS ftom East Carolina University and an AS ftom Western Kentucky University, both in Industrial Technology. He retired ftom the US Army as a Chief Tfdrrant Officer after 20 years of service, 15 of which were in the pow-er field. Will Knapek has been active in the

testing field since 1995 and is certified as a Senior NICET Techni-cian and a farmer NETA Certified Test Technician,Level IV. Will is also a member of IEEE.


Your first end-to-end test can seem to be a daunt-ing task, but the actual test procedure is not very complicated and can be per-formed more quickly and effectively than traditional relay test tech-niques; if everything works correctly. That last statement is the hard part because you are typically working with a team at a remote location which makes it difficult to determine whether the problem lies with the test plan, relay settings, test-set configura-tion, or operator error. This article will introduce you to end-to-end testing and answer many of the questions you might have.

Providing protection for a transmission line is difficult because the typical transmission line is fed from multiple sources and can carry loads that vary significantly. Traditional protection schemes use an impedance relay on either side of the transmis-sion line that constantly monitors voltage and cur-rent to calculate impedance in real time. If a fault occurred on the transmission line, the measured

impedance would be less than the calculated line impedance and the relays would operate to isolate the fault from the system. gnfortu-nately, the relays cannot be set at the exact line impedance because the typical protection-class current transformer has a 10 percent er-ror factor and the calculations will make assumptions (conductor spacing, splices, conductor impedance, etc.) that could cause the re-lay protection to overreach and isolate the wrong transmission line, so most impedance relays are set at 75 to 90 percent of the transmission line impedance as shown in Figure 1. A time-delayed backup impedance element is often set to reach beyond the trans-mission line to protect for faults in that last 75 to100 percent of the protected line so that all faults will be isolated from the system eventually. This element will also provide backup protection for the adjacent transmission line as shown in Figure 2. Remember that the other relay is looking in the op-posite direction as shown in Figure 3.

AN INTRODUCTION TO END-TO-END TESTING

NETA World, Winter 2012 Issue by Chris Werstiuk, Manta Test Systems

Figure 2: Typical Relay’s Primary and Backup Zones of Protection

Figure 1: Typical Impedance Relay’s Zone of Protection


Figure 3: Typical Transmission Line Impedance Protection Including All Zones of Protection

Every misoperation will cost the utility labor and lost revenue, so 75-90 percent protection is often not good enough, even though both sides overlap to provide a reasonably good protection scheme. Mod-ern protection schemes apply a communication channel between the relays on either side of a transmission line that allows the relays to communicate with each other. The simplest scheme (not really, but easiest to understand) applies differential protection principles. Each relay measures the current entering or leaving the transmission line and shares its local current with the remote relay bidirectionally. If the current in does not match the current out, both relays will trip. These relays often have a backup scheme using traditional impedance protection that is only active if the relays are unable to communicate.

The other communication schemes have many names (DUTT, PUTT, POTT, DCUB, etc.) and operating parameters, but they all perform the same basic function. Both relays monitor the real-time impedance and current direction. If both relays agree the fault is between them, both relays will trip as soon as possible depending on the communication medium and scheme.

While it is possible to test each of the individual components of a communication scheme separately, many problems can only be detected when the entire scheme is tested as a whole. It is possible to test one side at a time which can give the tester a reasonable sense that the scheme will operate successfully on proven relay settings, but many problems with communication-assisted pro-tection occur when the fault changes direction or by incorrectly defined communication delays which are inherent in the system. These problems can only be detected by properly applied end-to-end testing or a review of an incorrect relay operation after a fault.

End-to-end testing was considered daunting a decade ago, but advances in relay testing technology and personal computers have reduced the complexity to a couple of extra steps for a reasonably experienced relay tester.

2. WHAT IS END-TO-END TESTING?End-to-end testing uses two or more test-sets at multiple loca-

tions to simulate a fault at each terminal of a transmission line simultaneously to evaluate the entire protective relay scheme as a whole. This test technique previously required specialized knowl-edge and equipment to perform, but modern test-sets make it a rel-atively simple task. Figure 4 represents an overview of the equip-ment and personnel required for a typical end-to-end test using a simple transmission line with two ends or, as they are sometimes called, nodes. It is possible to have a system with three or more nodes which simply adds another location to the test plan.

The following components are necessary to perform a success-ful end-to-end test:

1. A relay test-set for each location with a minimum of:

• three voltage channels• three current channels• at least one programmable output to simulate breaker status or

other external signals• at least one programmable input to detect trip or breaker status

signals• an internal GPS clock (Some test-sets allow for other time sig-

nal synchronizations such as IRIG) or an external GPS clock with output signal and an additional test-set start input


• waveform playback or fault state/state simulator with at least three states available.

2. Some test-sets require a computer to control the test-set playback or state functions.

3. A computer and software to download and display event records obtained from the relay after each test.

4. At least one relay tester at each location with some form of com-munication between the two locations such as telephone or over-network communication. It is possible, but not recommended, for one person to perform all tests if the relay, relay test-sets, and communication systems have all been configured properly.

5. A setting file, waveform, or detailed description of the specific test scenarios.

6. An understanding of the relay protection scheme and what the expected result for each test should be.

7. The design engineer who created the settings and test plans stand-ing by if any problems are detected that need correction.

3. HOW DOES IT WORK?Most system disturbances develop within one millisecond of their

initiation and modern protective relays must be able to detect faults within this time frame to be effective. Practical experience has shown that two test-sets must start within 10 microseconds of each other to provide reliable results. This causes a problem for multiple relay tes-ters at multiple locations because it is nearly impossible for them to press start within 10 microseconds of each other. The remote relay testers could use the power system to synchronize their test-sets, but this method could add up to one millisecond or 22° error to the test which is not within the ten microsecond tolerance required for con-sistent results. The GPS system time allows synchroni-zation within 1 microsecond which is within the maximum allowable time delay. A substation IRIG signal can also be used but synchronization errors can be as high as 10 microseconds.

Figure 4: Typical End-To-End Test Overview

Figure 5-1: Typical 3-Phase Comtrade Fault File for 1st breaker

Figure 5-2: Typical 3-Phase Comtrade Fault File for 2nd breaker


Figure 6: Typical State Simulation Instructions for 1st Breaker (SOURCE BUS) and 2nd Breaker (LOAD BUS)

Figure 7: Typical End-to-End Test Scheme

Once two test-sets have synchronized time sources, both test-sets simultaneously apply fault simula-tions that are calculated to be similar to the signals that would occur if a real fault occurred. If the test-sets are properly synchronized, and the test plans are cre-ated and implemented correctly, and the re-lays are set correctly, and the correct auxiliary con-nections are made, and the communi-cation scheme is working, then the test should be successful.

Fault simulations can be supplied as waveform files (Figure 5) such as IEEE C37-111 COMTRADE files. This is typically the simplest method for everyone involved, as long as everything works correctly. Bowever, waveform files can be difficult to trou-bleshoot if problems occur.

Fault simulations can also be supplied in a test-set proprietary format that can be loaded directly into the test-set or in a stan-dard spreadsheet format (Figure 6) to be entered using the state simulation feature in the test-set. State simulations are easier to understand and troubleshoot but can be prone to conversion errors.

It is also important to note that different test-set manufactur-ers and test-set models may be synchro-nized to the same time source and set to the same start time, but may not start outputting the test at the same moment. Different firmware revisions can also be problematic with some manufacturers. Always consult with the relay test-set manufacturers if two different models of test-sets will be used for end-to-end testing on one line to determine if a cor-

rection factor must be applied. Different models from the same manufacturer can produce different starting times without notice, and the correction factors should be verified at the same location, if possible, before performing any remote testing.

4. ON WHAT SCHEMES SHOULD I PERFORM END-TO-END TESTING

End-to-end testing should be performed whenever it would be beneficial to test an entire protection scheme in real time to make sure that all equipment will operate correctly when required. This test technique need not be limited to transmission lines and can be applied any time you wish to test coordination between different devices. For example, a number of test-sets can be connected to the protective relays for all feeders in a system. A fault simulation can be played into all relays simultaneously to ensure that complex blocking schemes work as intended.

5. WHEN SHOULD I PERFORM END-TO-END TESTING?

All new installations with remote communication between re-lays should be tested via end-to-end testing. This test technique can also be a useful and effective maintenance test, particularly if end-to-end testing is performed during commissioning that will provide expected results. There is no more effective way of prov-ing the entire protection scheme than replaying the same number


of tests into the protection system and observing the same results. Performed correctly, using this test technique for maintenance tests can be more efficient as well.

6. HOW DO I PERFORM AN END-TO-END TEST?The relay testers at each end of the line should perform the fol-

lowing steps when performing an end-to-end test:

1. Obtain all of the test cases for all sides of the test procedure from the design engineer be-fore the scheduled test date. This step can be skipped for line differential relays if the test is to be performed by personnel who have a good understanding of the relay’s intended operation and test-set functions. It is possible to create test plans for impedance schemes, but they will be a poor substitute for real simula-tion files and may be too simple to truly test the entire scheme. A typical end-to-end test procedure will have 8 to 15 tests that are on either side of each impedance zone as shown in Figure 7. The test cases should have a combina-tion of fault types and phases. (For example: Test Case #1 = 3 Phase Fault, #2 = A-N fault, #3 = A-B fault, etc.)

2. Review all sides of each test case on a split screen and make sure the test plans make sense and that you understand what is supposed to happen. Most test plan files have a description such as “Test1=50%fromBRK1=Zone1Trip.” Some common problems to look for between all ends of a single test case include:

• Are prefault times consistent?

• Are prefault voltages nearly identical?

• Are prefault current magnitudes nearly identical?

• Are prefault current angles 180° apart?

• Same fault times?

• Same fault type?

• Same faulted phases on current and voltage?

• Are fault current magnitudes in phase for line faults and 180° apart for external faults?

• Any postfault enabled?

3. After arriving on site, determine the test location, setup your test-set, install yourGPS antenna outside the building with good access to the sky, and a pply gPS time as your test-set reference. (Or use other reference such as IRIG, if required)

4. Always remember that the relays under test communicate with each other, and your actions could cause an unintended trip on the other sides. Communicate with all ends under test. If all locations agree, isolate the circuit breaker and relay under test at all locations.

5. Connect the appropriate digital inputs and outputs between your test-set and protection scheme. I always recommend using the circuit breaker under test as part of the test, if possible, instead of simulating the breaker contacts.

6. Connect your test-set to replace current/potential transformer connections using the site’s three-line drawings.

7. Prepare a metering test and communicate with remote testers. If they agree, apply a meter test on all sides and verify correct results. Make sure that A-phase from your test-set is A-phase in the relay and repeat for B- and C-phase. A three-phase balanced meter test will NOT prove this.

8. Communicate with all remote sites and determine which test plan will be used for the test. (Test Case #1 for example)

9. Load the selected test case into all test-sets. Every site should have a unique test to load.

10. Place all circuit breakers in the correct posi-tions or ensure circuit breaker contacts are properly simulated by the test-set.

11. Communicate with all remote sites and select a start time. Apply the start time to the test-set.

12. The test should start automatically.

13. Review targets for correct operation and download all event records. Review the event records for correct operation or no operation, as required by the test procedure. Clear event records in the relay between tests to make sure you get the correct event files.

14. Repeat Steps 8-13 for all test cases.

As you can see, the key to any end-to-end test procedure is correct preparation. Steps 1 and 2 of the test procedure are the hardest parts of any end-to-end test. If the design engineer has correctly specified the tests and their outcomes, and you have carefully reviewed the test cases, the actual test procedure usually runs very smoothly with enough practice. ghen you review the test results in the field, these two principles will apply to most communication schemes:

1. Faults inside the zone of protection (current at the same angle on both ends) will be isolated more quickly by both relays than they would without the communication scheme.

2. Connect A-phase voltage and current from both test sets into any phase of the relay.

3. Perform the end-to-end test procedure using the same test plan.

4. Review the event record waveform from the relay. The voltages and currents should be identical.

5. Repeat until you feel comfortable.

Chris Werstiuk is an Electrical Tech-nologist, Journeyman Electrician, Professional Engineer and author of the upcoming book, “digital Relay test-ing: A Practical Guide from the Field.” He is also the founder of “RelayTesting. net”, an online resource for testing tech-nicians who need

custom test leads, test sheets templates, step-by-step testing guides, or an online forum to exchange ideas and information.


With the introduction of multifunction relays, testing methods for these devices must be approached in a different fashion. Gone are the days when the need to verify the operating characteristics of a relay were the only tests performed.

The characteristic of an element is no longer dependent on the mechanical operation of a rotating disk, bearings, drag magnets, and capacitors. The characteristic is derived from the execution of a mathematical formula determined by various settings, applied to variables that are the output of analog-to-digital converters modi-fied by sequence filters and carried out by a microprocessor. This change in how the protection element works means a change in what is important to test. With the intelligent electronic devices (IEDs) monitoring the health of the power system instead of mul-tiple, single-function, electro-mechanical relays, internal logic verification is as important as verifying the individual protection elements.

Testing of the logic in an IED presents a plethora of challenges. It is now the test technician’s job to inject the proper voltages and currents to simulate various operating and fault scenarios and ob-serve the elements’ action to the overall logic scheme. With elec-

tromechanical relays this was fairly straight forward, as the tests were applied to one relay and one element. Now the challenge is to be able to simulate conditions that operate an element without another element operating at the same time.

One of the tools available to the modern testing technician is the ability to simulate faults in two relays at two different loca-tions that are synchronized by Global Positioning Satellites (GPS). With modern test sets, the capability is available to test the protec-tion scheme as a system. Two or more locations can be set up to inject a simulated fault into the local relay with the values that would be seen by that relay during an actual fault. These signals are synchronized by the GPS signals and are coordinated to allow for a true system test.

THE GLOBAL POSITIONING SYSTEMThe current GPS system consists of 31 satellites. Each satel-

lite has a highly accurate atomic clock on board which is used to send a synchronized time signal to earth. Using the mathematical principle of four equations with four unknowns, a receiver must receive the time signal of at least four satellites to determine its position (longitude, latitude, and height) as well as accurate time.

GPS TESTING, THE FUTURE OF TESTING

NETA World, Winter 2012 Issue by William Knapek, OMICRON electronics

Figure 1: Fault At 95% of Line Length


The position can be determined to an accuracy of 66 feet and the time to an accuracy of 14 nanoseconds.

Two types of GPS receivers are available. One is the naviga-tion unit which is even available in your cell phone and the other is the satellite clock. The satellite clock has the ability to output a selection of timing signals which can be used to synchronize third party units with each other. To be able to use the time signal for synchroniza-tion purposes, a GPS clock is needed, which pro-vides a number of timing signals. These signals are the IRIG-B, PPM, and a programmable pulse signal. The Inter-Range Instru-mentation Group time code is commonly known as IRIG-B. The IRIG-B signal is one pulse per second. The PPM is a pulse per minute and of course the program-mable signal is programmable to your needs. The PPM and programmable signals can be used to synchronize test equipment at two or more loca-tions to the closest accuracy around the world. The GPS system is thus well suited for this type of test due to its availability all over the world and due to the accuracy of the timing signal.

To synchronize a test set with a satellite clock, the PPM signal is used. This signal triggers each test set to start the testing sequence. Each test sequence will be injecting the currents and voltages at the same magnitudes and phase angles that the protective relay would see during the actual fault.

TESTING TRIP SYSTEMSAs modern relays often have complex logic schemes applied, the

actual operation of the protection system must be verified. Testing the logic as a system will preclude misoperations of the protection system. This is most important in schemes that involve multiple relays, such as breaker-failure schemes, differential schemes, and communication-based schemes.

A state sequencer routine can be set up on one test set to start on a GPS pulse, inject the ap-propriate analog signals, and then measure the response of an IED to the applied signals. A sec-ond test set is set up at a remote location to start on the exact same GPS pulse. It applies the ap-propriate analog signals to the remote relay with the values expected at that location based on the fault characteristics.

Using a time signal view to measure the response time of the outputs from the relay, the system performance can be evaluated. These signals can confirm that the protection system performs as designed. Figure 1 is an example of a time signal view that can be used to verify a communication scheme. This test was for a permissive overreaching transfer trip cheme (POTT). This relay was injected with variables that equated to a fault at a location of 95 percent of the line length.

Figure 2: Fault At 5% of Line Length


Figure 2 is a screen shot of the remote location set for a fault of five percent of the line length. As you can see, the time signal view shows that the system performs as intended.

In the case of a bus differential system, several test sets can be set up to initiate a fault at the same time and measure the response of the 87B device. This is a much more effective test than just verify-ing the pickup and timing of a differential relay.

It is possible to use different models of test sets to perform these tests. With the use of the GPS signal to start the test, which manu-facturer’s test set you use is not important as long as it produces a quality signal. However it must be noted that not all test sets will process the GPS signal at the same speed. This could cause a variance in the timing of the injected signals. This variance can skew the test results. When using multivendor test sets, you must perform a few trials, preferably in the lab, to see the difference in the signal start times. When this is determined then an adjustment to the faster test set can be programmed to get all test sets to inject at the same time.

PMU TESTINGAnother application of GPS testing is to verify the correct opera-

tion of phasor measurement units (PMU) or syncrophasors. These are appearing in more and more protection systems throughout the world. In typical applications PMU’s are located at various points in the power system network and synchronized by a GPS signal. Synchrophasors measure voltages and currents at diverse locations on a power grid and can output accurately time-stamped voltage and current phasors. Because these phasors are truly synchronized, synchro-nized comparison of two quantities is possible, in real time. These comparisons can be used to assess system conditions.

When testing PMU’s you must inject signals from various loca-tions and retrieve the results for analysis. The timing of the in-jected signals must be exact or the results will not be a true repre-sentation of the power system condition.

CONCLUSIONThe advent of GPS has led to a new way to look at the tra-

ditional testing methods. New possibilities are available in the commissioning and maintenance of protection systems. Use of these tools allows us to go well beyond the verification of pickup/dropout and curve verification of relays. GPS or end-to-end test-ing is the future of testing.

REFERENCES1. Alexander Dier, Heinrich Metzler; GPS synconized tests

at the Vorarlberger Illwerke AG; OMICRON User Confer-ence, Berlin, November 1995

2. Global Positioning System; Wikipedia.com

3. Phasor Measurment Units; Wikipedia.com

William Knapek received a BS Degree in Industrial Technology from East Carolina University in 1994. In 1995 he retired from the US Army as a Chief Warrant Officer after 20 years of service. During his time with the Army Corps of Engineers, he held positions as a power plant instrumentation specialist, a writer/instructor for the Army Engineer

School, and a facility engineer for a Special Operations com-pound. He has been active in the electrical testing industry since retiring in 1995. He worked for NETA companies in the Nashville, Tenessee, area until joining OMICRON electronics as an applica-tion engineer in April of 2008.

He is currently the Secondary Engineering Service and Custom-er Support Manager for OMICRON electronics Corp, USA. He is certified as a Senior NICET Technician and a former NETA Level IV Test Technician. Will is a member of IEEE.


NETA Accredited Companies Valid as of January 1, 2014

For NETA Accredited Company List Updates Visit www.netaworld.org

For additional information on NETA visit netaworld.org For additional information on NETA visit netaworld.org

Ensuring Safety and ReliabilityTrust in a NETA Accredited Company to provide independent, third-party electrical testing to the highest standard, the ANSI/NETA Standards.

NETA has been connecting engineers, architects, facility managers, and users of electrical power equipment and systems with NETA Accredited Companies since1972.

United StateSAlAbAmA

AMP Quality Energy Services, LLC 4220 West Schrimsher SW, Site W1 PO Box 526 Huntsville, AL 35804(256) 513-8255 [email protected] Rodgers Utility Service Corporation 4614 Commercial Dr. NWHuntsville, AL 35816-2201(256) 837-8400 Fax: (256) [email protected] D. Peterson

ArizonA

ABM Electrical Power Solutions 3602 East Southern Ave., Suite 1 & 2 Phoenix, AZ 85040(602) 796-6583www.abm.com Jeff Militello

American Electrical Testing Co., Inc. 12566 W. Indianola Ave.Avondale, AZ 85392(480) [email protected] Madaglia

Electric Power Systems, Inc.557 E. Juanita Ave., #4Mesa, AZ 85204 (480) 633-1490 Fax: (480) 633-7092www.eps-international.com

Electrical Reliability Services 1775 W. University Dr., Suite 128Tempe, AZ 85281(480) 966-4568 Fax: (480) 966-4569www.electricalreliability.com

Hampton Tedder Technical Services 3747 West Roanoke Ave. Phoenix, AZ 85009(480) 967-7765 Fax:(480) 967-7762 www.hamptontedder.com

Southwest Energy Systems, LLC 2231 East Jones Ave., Suite A Phoenix, AZ 85040(602) 438-7500 Fax: (602) 438-7501 [email protected] www.southwestenergysystems.com Robert Sheppard

Western Electrical Services, Inc.5680 South 32nd St. Phoenix, AZ 85040(602) 426-1667 Fax: (253) 891-1511carcher@westernelectricalservices.comwww.westernelectricalservices.com Craig Archer

CAliforniA

ABM Electrical Power Solutions720 S. Rochester Ave., Suite AOntario, CA 91761 (951) 522-8855 Fax: (909) 937-6798www.abm.comBen Thomas

Apparatus Testing and Engineering 7083 Commerce Cir., Suite H Pleasanton,CA 94588 (925) 454-1363 Fax: (925) [email protected] (Jerry) Carr

Apparatus Testing and EngineeringPO Box 984Folsom, CA 95763-0984(916) 853-6280 Fax: (916) [email protected] James Lawler

Applied Engineering Concepts1105 N. Allen Ave.Pasadena, CA 91104(626) 398-3052 Fax: (626) [email protected] Castonguay

Electrical Reliability Services5810 Van Allen WayCarlsbad, CA 92008 (760) 804-2972 www.electricalreliability.com

Electrical Reliability Services6900 Koll Center Pkwy., Suite 415 Pleasanton, CA 94566 (925) 485-3400 Fax: (925) 485-3436 www.electricalreliability.com

Electrical Reliability Services10606 Bloomfield Ave.Santa Fe Springs, CA 90670(562) 236-9555 Fax: (562) 777-8914www.electricalreliability.com

Hampton Tedder Technical Services 4571 State St.Montclair, CA 91763 (909) 628-1256 x214 Fax: (909) [email protected] Tedder

Industrial Tests, Inc.4021 Alvis Ct., Suite 1Rocklin, CA 95677(916) 296-1200 Fax: (916) 632-0300 [email protected] Poole

3

2

1

4

5

6

12

11

10

9

8

7

17

16

15

14

13

18

For additional information on NETA visit netaworld.org

Pacific Power Testing, Inc.14280 Doolittle Dr.San Leandro, CA 94577 (510) 351-8811 Fax: (510) 351-6655steve@pacificpowertesting.comwww.pacificpowertesting.comSteve Emmert

Power Systems Testing Co. 4688 W. Jennifer Ave., Suite 108Fresno, CA 93722 (559) 275-2171 x15 Fax: (559) [email protected] Huffman

Power Systems Testing Co.6736 Preston Ave., Suite ELivermore, CA 94551(510) 783-5096 Fax: (510) 732-9287 www.powersystemstesting.com

Power Systems Testing Co.600 S. Grand Ave., Suite 113Santa Ana, CA 92705-4152(714) 542-6089 Fax: (714) 542-0737 www.powersystemstesting.com

POWER Testing and Energization, Inc. 731 E. Ball Rd., Suite 100Anaheim , CA 92805(714) 507-2702 www.powerte.com

Tony Demaria Electric, Inc. 131 West F St.Wilmington, CA 90744(310) 816-3130 x111Fax: (310) [email protected] Demaria

ColorAdoElectric Power Systems, Inc.6753 E. 47th Avenue Dr., Unit DDenver, CO 80216(720) 857-7273 Fax: (303) 928-8020 www.eps-international.com

Electrical Reliability Services7100 Broadway, Suite 7E Denver, CO 80221-2915(303) 427-8809 Fax: (303) 427-4080 www.electricalreliability.com

Magna IV Engineering96 Inverness Dr. East, Unit REnglewood, CO 80112(303) 799-1273 Fax: (303) [email protected] Aric Proskurniak

Precision Testing Group5475 Hwy. 86, Unit 1 Elizabeth, CO 80107 (303) 621-2776 Fax: (303) [email protected] Glenn Stuckey ConneCtiCut

Advanced Testing Systems 15 Trowbridge Dr.Bethel, CT 06801 (203) 743-2001 Fax: (203) [email protected] www.advtest.com Pat MacCarthy

American Electrical Testing Co., Inc.34 Clover Dr.South Windsor, CT 06074(860) 648-1013 Fax: (781) 821-0771 [email protected] Poulin

EPS Technology29 N. Plains Hwy., Suite 12Wallingford, CT 06492(203) 679-0145 www.eps-technology.com

High Voltage Maintenance Corp.150 North Plains Industrial Rd.Wallingford, CT 06492(203) 949-2650 Fax: (203) 949-2646 www.hvmcorp.com

Southern New England Electrical Testing, LLC3 Buel St., Suite 4Wallingford, CT 06492(203) 269-8778 Fax: (203) 269-8775 [email protected] Asplund, Sr.

floridAC.E. Testing, Inc. 6148 Tim Crews Rd.Macclenny, FL 32063(904) 653-1900 Fax: (904) [email protected] Chapman

Electric Power Systems, Inc.4436 Parkway Commerce Blvd.Orlando, FL 32808(407) 578-6424 Fax: (407) 578-6408 www.eps-international.com

Electrical Reliability Services11000 Metro Pkwy., Suite 30 Ft. Myers, FL 33966 (239) 693-7100 Fax: (239) 693-7772 www.electricalreliability.com Industrial Electric Testing, Inc.201 NW 1st Ave.Hallandale, FL 33009-4029(954) 456-7020 www.industrialelectrictesting.com

Industrial Electric Testing, Inc.11321 West Distribution Ave.Jacksonville, FL 32256(904) 260-8378 Fax: (904) 260-0737gbenzenberg@bellsouth.netwww.industrialelectrictesting.com Gary Benzenberg

Industrial Electronics Group850369 Highway 17 South PO Box 1870Yulee, FL 32041(904) 225-9529 Fax: (904) [email protected] E. Teal

GeorGiAElectrical Equipment Upgrading, Inc. 21 Telfair Pl.Savannah, GA 31415(912) 232-7402 Fax: (912) [email protected] Miller

Electrical Reliability Services2275 Northwest Pkwy. SE, Suite 180 Marietta, GA 30067(770) 541-6600 Fax: (770) 541-6501 www.electricalreliability.com

19

34

27

2820

35

21

36

29

22

37

30

23

38

31

24

39

32

25

26

40

41

33

67


Electrical Testing, Inc.2671 Cedartown Hwy.Rome, GA 3016-6791(706) 234-7623 Fax: (706) 236-9028 [email protected]

Nationwide Electrical Testing, Inc.6050 Southard TraceCumming, GA 30040(770) 667-1875 Fax: (770) [email protected] B. Bagle

illinois

Dude Electrical Testing, LLC 145 Tower Dr., Suite 9Burr Ridge, IL 60527(815) 293-3388 Fax: (815) [email protected] Dude

Electric Power Systems, Inc.23823 Andrew Rd.Plainfield, IL 60585(815) 577-9515 Fax: (815) 577-9516 www.eps-international.com

High Voltage Maintenance Corp. 941 Busse Rd. Elk Grove Village, IL 60007 (847) 640-0005www.hvmcorp.com

PRIT Service, Inc. 112 Industrial Dr. PO Box 606 Minooka, IL 60447(815) 467-5577 Fax: (815) 467-5883 [email protected] Hageman

indiAnA

American Electrical Testing Co., Inc. 4032 Park 65 Dr.Indianapolis, IN 46254(317) 487-2111 Fax: (781) 821-0771 [email protected] Canale

Electrical Maintenance & Testing Inc. 12342 Hancock St.Carmel, IN 46032 (317) 853-6795 Fax: (317) [email protected] K. Borst

High Voltage Maintenance Corp.8320 Brookville Rd., #EIndianapolis, IN 46239(317) 322-2055 Fax: (317) 322-2056 www.hvmcorp.com

iowA

Shermco Industries2100 Dixon St., Suite CDes Moines, IA 50316(515) 263-8482 [email protected] Hamrick

Shermco Industries796 11th St. Marion, IA 52302(319) 377-3377 Fax: (319) [email protected] Hamrick

louisiAnA

Electric Power Systems, Inc.1129 East Hwy. 30Gonzalez, LA 70737(225) 644-0150 Fax: (225) 644-6249www.eps-international.com

Electrical Reliability Services14141 Airline Hwy., Building 1, Suite XBaton Rouge, LA 70817(225) 755-0530 Fax: (225) 751-5055www.electricalreliability.com

Electrical Reliability Services9636 St. Vincent, Unit AShreveport, LA 71106(318) 869-4244www.electricalreliabilty.com

Electrical Reliability Services121 E. Hwy108Sulphur, LA 70665(337) 583-2411 Fax: (337) 583-2410 www.electricalreliability.com

Tidal Power Services, LLC8184 Hwy. 44, Suite 105 Gonzales, LA 70737 (225) 644-8170 Fax: (225) 644-8215darryn.kimbrough@tidalpowerservices.comwww.tidalpowerservices.comDarryn Kimbrough

Tidal Power Services, LLC 1056 Mosswood Dr.Sulphur, LA 70663 (337) 558-5457 Fax: (337) 558-5305steve.drake@tidalpowerservices.comwww.tidalpowerservices.com Steve Drake

mAine

Electric Power Systems, Inc.56 Bibber Pkwy., #1Brunswick, ME 04011(207) 837-6527www.eps-international.com

Three-C Electrical Co., Inc.72 Sanford DriveGorham, ME 04038(800) 649-6314 Fax: (207) [email protected] Cialdea

mArylAnd

ABM Electrical Power Solutions 3700 Commerce Dr., #901- 903 Baltimore, MD 21227(410) 247-3300 Fax: (410) 247-0900www.abm.com Bill Hartman

ABM Electrical Power Solutions4390 Parliament Pl., Suite QLanham, MD 20706(301) 967-3500 Fax: (301) 735-8953 www.abm.com Frank Ceci

Harford Electrical Testing Co., Inc. 1108 Clayton Rd.Joppa, MD 21085 (410) 679-4477 Fax: (410) [email protected] Biondino

42 49

43 50 58

44

51

52

59

45

60

46

54

53

62

63

61

47

48

55

57

56

68


High Voltage Maintenance Corp.9305 Gerwig Ln., Suite BColumbia, MD 21046(410) 309-5970 Fax: (410) 309-0220 www.hvmcorp.com

Potomac Testing, Inc.1610 Professional Blvd., Suite ACrofton, MD 21114(301) 352-1930 Fax: (301) 352-1936 [email protected] Bassett

Reuter & Hanney, Inc.11620 Crossroads Cir., Suites D - EMiddle River, MD 21220 (410) 344-0300 Fax: (410) 335-4389 www.reuterhanney.comMichael Jester

mAssAChusettsAmerican Electrical Testing Co., Inc.480 Neponset St., Bldg. 6Canton, MA 02021-1970(781) 821-0121 Fax: (781) 821-0771 [email protected] A. Blizard

High Voltage Maintenance Corp. 24 Walpole Park South Dr. Walpole, MA 02081(508) 668-9205 www.hvmcorp.com

Infra-Red Building and Power Service152 Centre St.Holbrook, MA 02343-1011 (781) 767-0888 Fax: (781) 767-3462 [email protected] www.infraredbps.comThomas McDonald Sr.

Three-C Electrical Co., Inc.40 Washington StreetWestborough, MA 01581(508) 881-3911 Fax: (508) 881-4814 [email protected] Cialdea

miChiGAnDYMAX Service Inc.46918 Liberty Dr. Wixom, MI 48393 (248) 313-6868 Fax: (248) 313-6869 www.dymaxservice.comBruce Robinson

Electric Power Systems, Inc.11861 Longsdorf St.Riverview, MI 48193 (734) 282-3311www.eps-international.com

High Voltage Maintenance Corp. 24371 Catherine Industrial Dr., Suite 207Novi, MI 48375 (248) 305-5596 Fax: (248) 305-5579 www.hvmcorp.com

Northern Electrical Testing, Inc.1991 Woodslee Dr.Troy, MI 48083-2236(248) 689-8980 Fax: (248) 689-3418 [email protected] www.northerntesting.comLyle Detterman

POWER PLUS Engineering, Inc. 46575 Magallan Dr.Novi, MI 48377(248) 344-0200 Fax: (248) 305-9105 [email protected] Mancuso

Powertech Services, Inc.4095 South Dye Rd.Swartz Creek, MI 48473-1570(810) 720-2280 Fax: (810) 720-2283 [email protected] www.powertechservices.com Kirk Dyszlewski

Utilities Instrumentation Service, Inc.2290 Bishop Circle EastDexter, MI 48130 (734) 424-1200 Fax: (734) [email protected] www.uiscorp.comGary E. Walls

minnesotA

DYMAX Holdings, Inc.4751 Mustang Cir.St. Paul, MN 55112(763) 717-3150 Fax: (763) 784-5397 [email protected] www.dymaxservice.com Gene Philipp

High Voltage Service, Inc.4751 Mustang Cir.St. Paul, MN 55112(763) 717-3103 Fax: (763) 784-5397www.hvserviceinc.comMike Mavetz

missouri

Electric Power Systems, Inc.6141 Connecticut Ave.Kansas City, MO 64120(816) 241-9990 Fax: (816) 241-9992 www.eps-international.com

Electric Power Systems, Inc.21 Millpark Ct.Maryland Heights, MO 63043-3536 (314) 890-9999 Fax:(314) 890-9998 www.eps-international.com

Electrical Reliability Services348 N.W. Capital Dr. Lees Summit, MO 64086 (816) 525-7156 Fax: (816) 524-3274 www.electricalreliability.com

nevAdA

ABM Electrical Power Solutions6280 South Valley View Blvd., Suite 618Las Vegas, NV 89118(702) 216-0982 Fax: (702) 216-0983 www.abm.com Jeff Militello

Control Power Concepts353 Pilot Rd, Suite BLas Vegas, NV [email protected] Fettig

Electrical Reliability Services6351 Hinson St., Suite BLas Vegas, NV 89118 (702) 597-0020 Fax: (702) 597-0095 www.electricalreliability.com

65

66

67

69

68

70

85

71 78

72

79

73

80

74

81

75

76

77

82

84

83

64

69


Electrical Reliability Services1380 Greg St., Suite 217Sparks, NV 89431 (775) 746-8484 Fax: (775) 356-5488 www.electricalreliability.com

Hampton Tedder Technical Services 4920 Alto Ave.Las Vegas, NV 89115 (702) 452-9200 Fax: (702) 453-5412 www.hamptontedder.comRoger Cates

new hAmpshire

Electric Power Systems, Inc.915 Holt Ave., Unit 9Manchester, NH 03109 (603) 657-7371 Fax: (603) 657-7370 www.eps-international.com

new Jersey

American Electrical Testing Co., Inc. 50 Intervale Rd., Suite 1Boonton, NJ 07005(973) 316-1180 Fax: (781) 316-1181 [email protected] Somol

Eastern High Voltage11A South Gold Dr.Robbinsville, NJ 08691-1606(609) 890-8300 Fax: (609) 588-8090 [email protected] www.easternhighvoltage.comJoseph Wilson

High Energy Electrical Testing, Inc.515 S. Ocean Ave.Seaside Park, NJ 08752(732) 938-2275 Fax: (732) 938-2277 [email protected] www.highenergyelectric.com Charles Blanchard

Longo Electrical-Mechanical, Inc.1625 Pennsylvania Ave.Linden, NJ 07036(908) 925-2900 Fax: (908) [email protected] Longo

Longo Electrical-Mechanical, Inc.One Harry Shupe Blvd., Box 511Wharton, NJ 07855(973) 537-0400 Fax: (973) [email protected] Longo

M&L Power Systems, Inc.109 White Oak Ln., Suite 82 Old Bridge, NJ 08857(732) 679-1800 Fax: (732) 679-9326 [email protected] Bagle

Scott Testing Inc.1698 5th St.Ewing, NJ 08638(609) 882-2400 Fax: (609) 882-5660 [email protected] Sorbello

Trace Electrical Services & Testing, LLC293 Whitehead Rd.Hamilton, NJ 08619(609) 588-8666 Fax: (609) 588-8667 [email protected] Vasta

new mexiCo

Electric Power Systems, Inc. 8515 Cella Alameda NE, Suite A Albuquerque, NM 87113 (505) 792-7761 www.eps-international.com

Electrical Reliability Services8500 Washington Pl. NE, Suite A-6Albuquerque, NM 87113 (505) 822-0237 Fax: (505) 822-0217 www.electricalreliability.com

new york

A&F Electrical Testing, Inc.80 Lake Ave. S., Suite 10Nesconset, NY 11767 (631) 584-5625 Fax: (631) 584-5720 [email protected] Kevin Chilton

A&F Electrical Testing, Inc.80 Broad St., 5th FloorNew York, NY 10004 (631) 584-5625 Fax: (631) 584-5720 [email protected] www.afelectricaltesting.com Florence Chilton

American Electrical Testing Co., Inc. 76 Cain Dr.Brentwood, NY 11717 (631) 617-5330 Fax: (631) 630-2292 [email protected] Schacker

Elemco Services, Inc.228 Merrick Rd.Lynbrook, NY 11563 (631) 589-6343 Fax: (631) 589-6670 [email protected] O’Brien

High Voltage Maintenance Corp.1250 Broadway, Suite 2300New York, NY 10001 (718) 239-0359 www.hvmcorp.com

HMT, Inc.6268 Route 31Cicero, NY 13039 (315) 699-5563 Fax: (315) 699-5911 [email protected] Pertgen

north CArolinAABM Electrical Power Solutions3600 Woodpark Blvd., Suite GCharlotte, NC 28206(704) 273-6257 Fax: (704) [email protected] Goins

ABM Electrical Power Solutions5805 G Departure Dr. Raleigh, NC 27616(919) 877-1008 Fax: (919) 501-7492 www.abm.com Rob Parton

91

92

99

86 93 100

87

94 101

8895 102

89 96 103

90

97

104

98

105

106

70


ELECT, P.C.7400-G Siemens Rd. PO Box 2080 Wendell, NC 27591 (919) 365-9775 Fax: (919) [email protected] W. Tyndall

Electric Power Systems, Inc.319 US Hwy. 70 E, Unit EGarner, NC 27529(919) 322-2670www.eps-international.com

Electrical Reliability Services6135 Lakeview Road, Suite 500Charlotte, NC 28269(704) 441-1497www.electricalreliability.com

Power Products & Solutions, Inc.12465 Grey Commercial Rd. Midland, NC 28107 (704) 573-0420 x12 Fax: (704) 573-3693 [email protected] www.powerproducts.biz Ralph Patterson

Power Test, Inc.2200 Hwy. 49Harrisburg, NC 28075 (704) 200-8311 Fax: (704) 455-7909 [email protected] Walker

ohioCE Power Solutions, LLC4500 W. Mitchell Ave.Cincinnati, OH 45232 (513) 563-6150 Fax: (513) 563-6120 [email protected] Harris

DYMAX Service, Inc.4213 Kropf Ave. Canton, OH 44706 (330) 484-6801 Fax: (740) 333-1271 www.dymaxservice.comGary Swank

Electric Power Systems, Inc.2601 Center Rd., #101Hinckley, OH 44233 (330) 460-3706 Fax: (330) 460-3708 www.eps-international.com

Electrical Reliability Services610 Executive Campus Dr.Westerville, OH 43082 (877) 468-6384 Fax: (614) [email protected] www.electricalreliability.com

High Voltage Maintenance Corp.5100 Energy Dr.Dayton, OH 45414 (937) 278-0811 Fax: (937) 278-7791www.hvmcorp.com

High Voltage Maintenance Corp.7200 Industrial Park Blvd. Mentor, OH 44060 (440) 951-2706 Fax: (440) 951-6798 www.hvmcorp.com

Power Services, LLC998 Dimco Way, PO Box 750066Centerville, OH 45475 (937) 439-9660 Fax: (937) 439-9611 [email protected] Beucler

Power Solutions Group, Ltd.670 Lakeview Plaza Blvd. Columbus, OH 43085(614) [email protected] www.powersolutionsgroup.comStuart Spohn

Power Solutions Group, Ltd.425 W. Kerr Rd.Tipp City, OH 45371 (937) 506-8444 Fax: (937) 506-8434bwilloughby@powersolutionsgroup.comwww.powersolutionsgroup.com Barry Willoughby

oklAhomA

Shermco Industries1357 N. 108th E. Ave.Tulsa, OK 74116 (918) 234-2300 [email protected] Harrison

oreGon

Electrical Reliability Services4099 SE International Way, Suite 201 Milwaukie, OR 97222-8853(503) 653-6781 Fax: (503) 659-9733 www.electricalreliability.com

Taurus Power & Controls, Inc.9999 SW Avery St.Tualatin, OR 97062-9517 (503) 692-9004 Fax: (503) 692-9273 [email protected] www.tauruspower.comRob Bulfinch

pennsylvAniA

ABM Electrical Power Solutions710 Thomson Park Dr.Cranberry Township, PA 16066-6427 (724) 772-4638 Fax: (724) [email protected] (Pete) McKenzie

American Electrical Testing Co., Inc. Green Hills Commerce Center5925 Tilghman St., Suite 200Allentown, PA 18104(215) 219-6800 [email protected] Munley

Burlington Electrical Testing Co., Inc.300 Cedar Ave.Croydon, PA 19021-6051(215) 826-9400 x221 Fax: (215) [email protected] P. Cleary

Electric Power Systems, Inc.1090 Montour West Industrial Blvd. Coraopolis, PA 15108 (412) 276-4559www.eps-international.com

Electric Power Systems, Inc.2495 Boulevard of the Generals Norristown, PA 19403 (610) 630-0286www.eps-international.com

EnerG Test204 Gale Lane, Bldg. 2 – 2nd FloorKennett Square, PA 19348(484) 731-0200 Fax: (484) [email protected] Bleiler

High Voltage Maintenance Corp.355 Vista Park Dr.Pittsburgh, PA 15205-1206 (412) 747-0550 Fax: (412) 747-0554 www.hvmcorp.com

107 115 123

108

116

124

109

117

126

125

110

118

127

128

111

112

119

129113

120

130114

121

122

71


Longo Electrical-Mechanical, Inc.1400 F Adams RoadBensalem, PA 19020(215) 638-1333 Fax: (215) [email protected] Longo

North Central Electric, Inc. 69 Midway Ave.Hulmeville, PA 19047-5827 (215) 945-7632 Fax: (215) 945-6362 [email protected] Messina

Reuter & Hanney, Inc.149 Railroad Dr. Northampton Industrial ParkIvyland, PA 18974 (215) 364-5333 Fax: (215) 364-5365 [email protected] Michael Reuter

south CArolinA

Power Products & Solutions, Inc.13 Jenkins Ct.Mauldin, SC 29662(800) 328-7382 [email protected] www.powerproducts.bizRaymond Pesaturo

Power Solutions Group, Ltd.135 Old School House Rd. Piedmont, SC 29673 (864) 845-1084 Fax: (864) [email protected] www.powersolutionsgroup.com Frank Crawford

tenneseeElectric Power Systems, Inc.146 Space Park Dr.Nashville, TN 37211(615) 834-0999 Fax: (615) 834-0129 www.eps-international.com

Electrical & Electronic Controls6149 Hunter Rd.Ooltewah, TN 37363 (423) 344-7666 x23 Fax: (423) [email protected] Michael Hughes

Power & Generation Testing, Inc.480 Cave Rd. Nashville, TN 37210 (615) 882-9455 Fax: (615) 882-9591 [email protected] Ramieh

texAsAbsolute Testing Services, Inc.6829 Guhn Rd.Houston, TX 77040(832) 467-4446 Fax: (713) 849-3885 [email protected] www.texasats.comRichard Gamble

Electric Power Systems, Inc.4100 Greenbriar Dr., Suite 160 Stafford, TX 77477(713) 644-5400www.eps-international.com

Electrical Reliability Services1057 Doniphan Park Cir., Suite AEl Paso, TX 79922 (915) 587-9440 Fax: (915) 587-9010 www.electricalreliability.com

Electrical Reliability Services1426 Sens Rd., Suite 5 Houston, TX 77571 (281) 241-2800 Fax: (281) 241-2801 www.electricalreliability.com

Grubb Engineering, Inc.3128 Sidney BrooksSan Antonio, TX 78235(210) 658-7250 Fax: (210) [email protected] Robert D. Grubb Jr.

National Field Services649 Franklin St.Lewisville,TX 75057 (972) 420-0157 www.natlfield.comEric Beckman

Power Engineering Services, Inc.9179 Shadow Creek Ln.Converse,TX 78109 (210) 590-4936 Fax: (210) 590-6214 [email protected] R. Engelke

Saber Power Systems9841 Saber Power LaneRosharon, TX 77583(713) [email protected] Taylor

Shermco Industries33002 FM 2004Angleton , TX 77515(979) 848-1406 Fax: (979) [email protected] www.shermco.com Malcom Frederick

Shermco Industries1705 Hur Industrial Blvd. Cedar Park, TX 78613 (512) 267-4800 Fax: (512) [email protected] Ewing

Shermco Industries2425 E. Pioneer Dr.Irving, TX 75061(972) 793-5523 Fax: (972) 793-5542 [email protected] Widup

Shermco Industries12000 Network Blvd., Bldg. D, Suite 410San Antonio, TX 78249(512) 267-4800 Fax: (512) [email protected] Kevin Ewing Tidal Power Services, LLC4202 Chance Ln.Rosharon, TX 77583(281) 710-9150 Fax: (713) 583-1216 [email protected] www.tidalpowerservices.com Monty C. Janak

utAhElectrical Reliability Services3412 South 1400 West, Unit AWest Valley City, UT 84119 (801) 975-6461 www.electricalreliability.com

Western Electrical Services, Inc.3676 W. California Ave.,#C-106Salt Lake City, UT 84104rcoomes@westernelectricalservices.comwww.westernelectricalservices.com Rob Coomes

131 138 146

132

139

147

133

140

148

134

141 149

135

142

150

151

152

153

136

143

137

144

145

72


virGiniA

ABM Electrical Power Solutions814 Greenbrier Cir., Suite E Chesapeake, VA 23320(757) 548-5690 Fax: (757) 548-5417www.abm.comMark Anthony Gaughan, III

Electric Power Systems, Inc.827 Union St. Salem, VA 24153 (540) 375-0084 Fax: (540) 375-0094www.eps-international.com

Potomac Testing, Inc.11179 Hopson Rd., Suite 5 Ashland, VA 23005 (804) 798-7334 Fax: (804) 798-7456 www.potomactesting.com

Reuter & Hanney, Inc.4270-I Henninger Ct.Chantilly, VA 20151 (703) 263-7163 Fax: (703) 263-1478 www.reuterhanney.com

wAshinGton

Electrical Reliability Services2222 West Valley Hwy. N., Suite 160 Auburn, WA 98001 (253) 736-6010 Fax: (253) 736-6015 www.electricalreliability.com

POWER Testing and Energization, Inc.22035 70th Ave. SouthKent, WA 98032 (253) 872-7747 www.powerte.com

POWER Testing and Energization, Inc.14006 NW 3rd Ct., Suite 101Vancouver, WA 98685(360) 597-2800 Fax: (360) 576-7182 [email protected] www.powerte.comChris Zavadlov

Sigma Six Solutions, Inc.2200 West Valley Hwy., Suite 100 Auburn, WA 98001 (253) 333-9730 Fax: (253) 859-5382 [email protected] White

Taurus Power & Controls, Inc.6617 S. 193rd Pl., Suite P104Kent, WA 98032(425) 656-4170 Fax: (425) 656-4172 [email protected] Lightner

Western Electrical Services, Inc.14311 29th St. EastSumner, WA 98390(253) 891-1995 Fax: (253) 891-1511dhook@westernelectricalservices.comwww.westernelectricalservices.comDan Hook

Western Electrical Services, Inc.4510 NE 68th Dr., Suite 122 Vancouver, WA 98661 (888) 395-2021 Fax: (253) 891-1511 [email protected] www.westernelectricalservices.comTony Asciutto

wisConsin

CE Power Solutions of Wisconsin, LLC3100 East Enterprise Ave.Appleton, WI 54913(920) 968-0281 Fax: (920) 968-0282 [email protected] Fulton

Electrical Energy Experts, Inc.W129N10818, Washington Dr.Germantown,WI 53022 (262) 255-5222 Fax: (262) 242-2360 [email protected] www.electricalenergyexperts.com William Styer

Electrical Testing Solutions2909 Green Hill Ct.Oshkosh, WI 54904(920) 420-2986 Fax: (920) [email protected] www.electricaltestingsolutions.comTito Machado

Energis High Voltage Resources, Inc. 1361 Glory Rd.Green Bay, WI 54304(920) 632-7929 Fax: (920) [email protected] www.energisinc.comMick Petzold

High Voltage Maintenance Corp.3000 S. Calhoun Rd.New Berlin, WI 53151 (262) 784-3660 Fax: (262) 784-5124 www.hvmcorp.com

155

154

156

157

158

159

160

161

162 168

169

163

164

165

166

167

73


canada

Magna IV Engineering200, 688 Heritage Dr. SECalgary, AB T2H1M6 Canada(403) 723-0575 Fax: (403) 723-0580 [email protected] Virginia Balitski

Magna IV Engineering1005 Spinney Dr.Dawson Creek, BC V1G 1K1 Canada(780) 462-3111 Fax: (780) [email protected]

Magna IV Engineering1103 Parsons Rd. SW Edmonton, AB T6X 0X2 Canada(780) 462-3111 Fax: (780) [email protected] Balitski

Magna IV Engineering106, 4268 Lozells AveBurnaby, BC VSA 0C6Canada(604) 421-8020

Magna IV Engineering8219D Fraser Ave. Fort McMurray, AB T9H 0A2 Canada (780) 791-3122 Fax: (780) 791-3159 [email protected] Virginia Balitski

Magna IV Engineering1040 Winnipeg St. Regina, SK S4R 8P8 Canada(306) 585-2100 Fax: (306) 585-2191 [email protected] Frostad

Magna Electric Corporation 3430 25th St. NECalgary, AB T1Y 6C1 Canada (403) 769-9300 Fax: (403) 769-9369 [email protected] www.magnaelectric.com Cal Grant

Magna Electric Corporation3731-98 StreetEdmonton, AB T6E 5N2 Canada(780) 436-8831 Fax: (780) [email protected] Granacher

Magna Electric Corporation1033 Kearns Crescent, Box 995Regina, SK S4P 3B2 Canada(306) 949-8131 Fax: (306) 522-9181 [email protected] Heid Magna Electric Corporation851-58th St. EastSaskatoon, SK S7K 6X5 Canada(306) 955-8131 x5 Fax: (306) [email protected] Wilson

Magna Electric Corporation1375 Church Ave.Winnipeg, MB R2X 2T7 Canada (204) 925-4022 Fax: (204) [email protected] www.magnaelectric.comCurtis Brandt

Orbis Engineering Field Services Ltd. #300, 9404 - 41st Ave. Edmonton, AB T6E 6G8 Canada(780) 988-1455 Fax: (780) [email protected] Lorne Gara

Pacific Powertech Inc.#110, 2071 Kingsway Ave. Port Coquitlam, BC V3C 1T2 Canada(604) 944-6697 Fax: (604) 944-1271 [email protected] Josh Conkin

REV Engineering, LTD3236 - 50 Ave. SE Calgary, AB T2B 3A3 Canada(403) 287-0156 Fax: (403) [email protected] Nicholas Davidson, IV

BrUSSelS

Shermco IndustriesBoulevard Saint-Michel 47 1040 Brussels, Brussels, Belgium+32 (0)2 400 00 54 Fax: +32 (0)2 400 00 32 [email protected] Paul Idziak

chileMagna IV EngineeringAvenida del Condor Sur #590Officina 601Huechuraba, Santiago 8580676 Chile +(56) [email protected] Fuentes

PUerto rico

Phasor EngineeringSabaneta Industrial Park #216 Mercedita, Puerto Rico 00715 (787) 844-9366 Fax: (787) 841-6385 [email protected] Rafael Castro

172

181

173

182

174

183

175

176

177

178

184170

179

185

186

171

180

REV 01.14

74

ABOUT THE INTERNATIONAL ELECTRICAL TESTING ASSOCIATION

The InterNational Electrical Testing Association (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 Certified Technicians con-duct the tests that ensure this equipment meets the Association’s stringent specifica-tions. NETA is 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.

CERTIFICATIONCertification 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 ETT-2000 Standard for Certification of Electri-cal Testing Technicians.

QUALIFICATIONS OF THE TESTING ORGANIZATIONAn 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 associa-tion dedicated to setting world standards in electrical maintenance and acceptance testing. Hiring a NETA Ac-credited 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 ETT-2000 Standard for Certification of Electrical Testing Technicians.

• A Registered Professional Engineer will review all engineering reports

• All tests will be performed objectively, according to NETA specifications, using cali-brated instruments traceable to the National Institute of Science and Technology (NIST).

• The firm is a well-established, full-service electrical testing business.

Setting the Standard

neta handbook series ii - protective vol 3-pdf

Documents