Feature

Output Logic Testing

by Chris Werstiuk
Valence Engineering, Inc.
Introduction

O
ne of the most confusing aspects of modern relay testing is deciding 4. Question the design engineer if
what to test. Testing earlier relay generations was relatively straight- the logic doesnt make sense to
you.
forward, as these devices had limited scope and fixed configurations.
5. Use the operating description
Pickup and timing tests were performed to make sure the device was mechani- instead of interpreting the relays
cally sound and all the parts operated correctly. Todays relays are infinitely more settings when testing complex
logic. The relay is going to do
complex, and our testing philosophy should be modified accordingly. The com-
what its programmed to do.
plexity of these relays is due to the flexibility of their outputs. LEDs, output You want to make sure that its
relays, virtual outputs to remote devices, and displays on the relays front panel programmed correctly. If an op-
erating description doesnt exist,
can customize to communicate nearly every aspect of the electrical system. These ask for one.
outputs are controlled by their internal logic which actually defines the relays
functionality. Simple Output
Output logic errors are the most prevalent problems I discover during relay Logic Testing Example
testing and occur in 80 percent of the relays. Because of this experience, I feel The output logic in Figure 1 is
that output logic testing is, by far, the most important part of relay testing. This from a typical feeder relay and is
article discusses output logic testing philosophies to help discover these problems shown in the most common logic
before its too late. formats and translated into elec-
Think of the relays output logic as an electrical schematic that must be com- trical schematics. This logic was
missioned with the same attention to detail that one would apply to any other compared to the dc schematic and
schematic. The first step to output testing is to translate the output logic into a performs the following functions.
checklist, electrical schematic, operating description, or any other representation
that you understand. The logic should be broken
down into its base components or combinations,
R1 - TRIP R2 - CLOSE R3 - AUX R7 - AUX
and each combination should be checked off as
it is proven. Following are some rules for output 50P + 51PT 27 50P + 51PT + 27
logic testing. 50P 27T
50P
OR R1 OR R3 51PT OR R7
1. Always use the as-left relay settings when test- 51PT 27T
ing output logic. Why test something you are
not going to use?
2. All setting changes must be completed before 27T
50P 51PT 50P 51PT 27T
performing output logic tests. Why would you CLOSE
test something that youre going to change in PUSHBUTTON
the future? One-character mistakes in changed
relay logic can be disastrous. R1 R3 R7
3. Use end devices (breaker/disconnect switch sta-
tus) whenever possible. Your goal is to simulate
real life conditions to ensure the logic operates Figure 1 Example Output Logic
correctly while in-service.

www.netaworld.org Winter 2006-2007 NETA WORLD 

 Output R1 trips a lockout relay that trips the circuit One way to document these tests is to create a spreadsheet
breaker. with each output contact in a column as shown in Figure 2.
Output R2 closes the circuit breaker. Each row under the output designations represents an OR
gate. Remove the highlight as each step is performed.
Output R3 trips the circuit breaker directly.
Output R7 provides trip annunciation to the SCADA/
DCS/alarm panel. Complex Relay logic Example
Complex logic schemes require more thought. The
Testing this logic is relatively straightforward. After all breaker fail protection commonly installed on SEL relays
pickup and timing tests are completed, perform the fol- is a great example of a slightly more complex logic scheme.
lowing steps: The output logic for this example will be OUT101=SV1T.
1. Verify that the rules discussed above are followed and In order to test this output, we need to find out how SV1T
energize as much of the circuit as possible. If the station and its related settings are defined. SV1T is derived from
is in operation and there are not sufficient test switches to the settings in Figure 3.
isolate all the desired trips, record the reasons for untested
logic. (See rule #3.) Relay Settings Application Review
2. Connect relay test set inputs to Output R1 and inject a OUT101 = SV1T IN201 = TRIPS FROM OTHER
50P timing test. Verify: SV1 = (SV1 + IN201) * (50P1 + 50G1) RELAYS
SV1PU = 10
Output R1 operates within the correct time. SV1DO = 10
The relay annunciates the correct trip and phase. 50P1P = 0.25
50G1P = 0.25
The lockout relay trips.
3. Reset the lockout and inject a 51 timing test and verify
Figure 3 Example #2 Settings
everything in step 2.
Break down all of the possible combinations required for
OUT101 to operate before testing. The first step is to expand
FINAL OUTPUT CHECKS the equation until all of the brackets have been removed and
SOLID STATE R1 R2 R3 R5 R6 R7 we are left with a series of OR functions. SV1 becomes SV1
TRIP
* 50P1 + SV1 * 50G1 + IN201 * 50P1 + IN201 * 50G1.
NA 51PT CLOSEPB 27T NA NA 27T
Figure 4 displays the different logic schemes and Figure 5
50P 51PT is an example test sheet.
50P

COMMENTS:
MATH LOGIC
SV1 = (SV1 + IN201) * (50P1 + 50G1)
SV1 = (SV1 * 50P1) + (5SV1 * 50G1) + (IN201) * (50P1 + 50G1)
YES
NO SEE NOTES SV1 = (SV1 * 50P1) + (5SV1 * 50G1) + (IN201 * 50P1) + (IN201 * 50G1)

Figure 2 Example Test Sheet for Output Logic GATE LOGIC SCHEMATIC
SV1
OR
IN201
AND SV1

4. Connect the test set input across Output R2 and push 50P1
OR
10 CYCLES
SV1 IN
201
SV1

the close button. Verify the contact operates and closes 50G1

the circuit breaker.. SV1

AND
50P1
5. Connect the relay test set input to Output R3 and inject SV1 50G 50N
a 27 timing test. Verify: 50G1
AND
OR SV1
Output R3 operates within the correct time. IN201
AND
10 CYCLES
50G1 SV1 SV1T 10 CYCLES
The relay annunciates the correct trip and phase. IN201
AND
The circuit breaker trips. 50G1

6. Repeat steps 2, 3, and 5 with the test set inputs con-

Figure 4 Example Breaker Failure Logic
nected across R7 and verify DCS/SCADA/Annunciator
operation.

 NETA WORLD Winter 2006-2007 www.netaworld.org

 FINAL OUTPUT CHECKS Modern test equipment allows one to simulate almost
OUT 101 OUT 102 OUT 103 OUT 104 OUT 105 OUT 106 OUT 107
any real-life fault and be able to perform a timing test on
nearly every element without interference from other ele-
SV1 * 50P1 NA NA NA NA NA NA
ments. If one is unable to obtain a timing test for an element
SV1 * 50G1
due to interference from another element, the settings are
IN201 * 50P1
redundant (except 87-differential elements, of course) and
IN201 * 50G1 should be noted on the test sheets and brought to the design
COMMENTS: engineers attention. Waiting until the end and proving all
of the output logic at once is the least confusing method,
YES
NO SEE NOTES but it can be inefficient due to the need to reconfigure the
test set between each test, a relatively minor inconvenience
Figure 5 Example Test Sheet for Breaker Failure Logic for computer operated testers. This method should always
be used, if setting changes are performed during testing, to
After the logic has been reduced to its simplest compo- ensure the relay will operate when required.
nents, output testing is a breeze using the following steps. Another method incorporates the element testing into
the output testing. This method is more efficient and should
Apply IN201 and nothing should happen. only be used if setting changes during testing are nonex-
You could apply 0.3 ampere in any phase to operate istent or carefully planned. With this method, perform
the relay but this will operate both the 50P and 50G each timing test on another output until all of the assigned
elements. First apply three-phase current higher than outputs are used. In the example, the first 51PT timing test
0.25A with IN201 applied. OUT101 should close after (A-phase, for example) can be performed using R1 and the
10 cycles has passed. You can now mark 50P1*IN201 as next timing test (B-phase, for example) is performed on
completed. R7. After all of the 51PT timing tests are completed, the
first 50-element timing test can be performed on R7 and
Remove IN201 and if the output remains closed, mark the next test will be performed on R1. This method is more
50P1*SV1 as completed because the SV1 seal-in function efficient but requires careful attention to make sure that no
has operated correctly. element/output combinations are missed.
Turn the current off and the output should open.
Reapply current and nothing should happen. Conclusion
A common mistake for relay testers is to use spare out-
Apply 0.20 amperes at 0 in each phase (equals 0.6
puts, displays, and/or LEDs for their pickup and timing tests
amperes ground fault) and nothing should happen until
and ignore the in-service output logic, believing that they
you apply IN201. If the output operates 10 cycles after
are using the same elements in their test equations as the
applying IN201, mark 50G1*IN201 as completed.
final logic. Following are two real life examples of mistakes
Remove IN201 and you can mark 50P1*SV1 as com- made when testing the output logic last.
pleted if the output remains closed.
Example 1
Output Logic Testing Methods After testing the differential element using a spare con-
tact on a GE T60 relay using XFMR PCNT DIFF OP
There are many ways to prove the output logic settings. as the test logic, I went back to test each of the outputs.
The tester can perform all other tests and wait until the Everything went fine until I came to the third output and
end before performing a timing test for every element on decided to apply a B-phase differential fault and nothing
every output contact as shown in the example. To reduce happened! I reviewed the setting and it was set to operate
test set changes, connect the test set contact sensing to when XFMR PCNT DIFF OP A operated. One character
output R1 and perform a 51PT timing test, then move the made the difference between an A-phase differential trip
contact sensing to R7 and repeat the test. Then perform a and a differential trip on all phases.
50P test, move to R1 and perform the test again. If the test
set has multiple inputs and/or event reporting, connect all Lesson learned: One character mistakes are easy to miss
relay outputs to the test set inputs and watch to make sure and potentially disastrous!!
all the correct relay outputs operate. After proving that the
contacts actually operate at least once, monitor their status
via the software or open the event records to make sure the
correct outputs operate.

www.netaworld.org Winter 2006-2007 NETA WORLD 

Example 2
I finished testing all of the elements on an SEL-300G
relay and everything was fine. I started testing the output
logic and came across the highlighted bit of logic of this
equation: 64G1T + (64G2T * !27PP1 * 60LOP) + 87R +
87U + (81D2T+ 81D3T) * 52A + 81D5T + 81D6T. 64G2T
is 3rd harmonic undervoltage protection, 27PP1 is under-
voltage protection, and 60LOP is fuse failure protection.
64G2T was disabled by this equation because !27PP1 *
60LOP is an impossible condition and can never happen.
The setting was changed to !27PP1 * !60LOP and operates
correctly. These settings were supplied by a major genera-
tor manufacturer from a boilerplate template. How many
generators out there have 64G2T protection disabled and
no one knows it?

Lesson learned: Always perform output logic tests and

question logic that doesnt make sense!

Chris Werstiuk graduated from the Electrical Engineering Technol-

ogy program at the Northern Alberta Institute of Technology and is a
Registered Engineering Technologist and Journeyman Power System
Electrician. He has performed low-, medium-, and high-voltage ac-
ceptance testing, commissioning, and maintenance testing at various
locations throughout the Americas. His experience is varied, and he
is comfortable in mining, utility, generating, distribution, production,
institutional, plant, pipeline, and commercial environments. Chris is
currently a Project Manager for Valence USA, a full service field service
and electrical training organization.

 NETA WORLD Winter 2006-2007 www.netaworld.org