×
INTELLIGENT WORK FORUMS
FOR ENGINEERING PROFESSIONALS

Log In

Come Join Us!

Are you an
Engineering professional?
Join Eng-Tips Forums!
  • Talk With Other Members
  • Be Notified Of Responses
    To Your Posts
  • Keyword Search
  • One-Click Access To Your
    Favorite Forums
  • Automated Signatures
    On Your Posts
  • Best Of All, It's Free!
  • Students Click Here

*Eng-Tips's functionality depends on members receiving e-mail. By joining you are opting in to receive e-mail.

Posting Guidelines

Promoting, selling, recruiting, coursework and thesis posting is forbidden.

Students Click Here

Jobs

Is there a benefit in testing OC element drop-offs?

Is there a benefit in testing OC element drop-offs?

Is there a benefit in testing OC element drop-offs?

(OP)
Hi

When testing OC relays I've always faithfully checked that the element drops out with the specifications of the relay manufacturer. I can appreciate its purpose when considering relays in series. So let's say I have relay A upstream and B downstream. Maximum load current at A is say 100A. If A was an old EM relay it's reset ratio would typically be 90%. I would thus need to ensure that effective pickup setting > 100/0.9 = 111A. (Note effective does not mean pickup setting. For EM relays effective pickup could be anywhere from 5% to 30% of pickup of I remember correctly). Thus if my system is heavily loaded and B clears a downstream fault for which A has picked up, then A will not time out and trip due to the load current being above its drop-out value.

Anyway, the point being if I have a generator EF relay is there any benefit in testing the dropout of the EF element? Generator is connected to a delta-star step-up trfr (delta winding on generator side). There is no other EF relay to grade with. Load current is not an issue.

I can extend this further to is there any benefit in verifying drop-offs for EF relays since generally load is not a factor?

Am interested in other views/comments regarding this.

RE: Is there a benefit in testing OC element drop-offs?

EF?

RE: Is there a benefit in testing OC element drop-offs?

(OP)
EF = Earthfault (or groundfault as it is more commonly known in the States). Actually the heading should have been, "Is there a benefit in testing EF element drop-off's?"

RE: Is there a benefit in testing OC element drop-offs?

What kind of relay?

Ancient electromechanical relays? Absolutely, test the curve performance in every relay; often.

Modern numeric relays? Don't waste your time.

Repeatibility, over time, of electromechanical relays is a very legitimate concern. On numeric relays I expect full repeatibility from the out of box test to the last time the relay fully functions prior to the ultimate decommission.

As far as I can tell, element testing of numeric relays should be confined to product acceptance (type) testing. If one works properly following the factory tests they all well. (At least for anybody that I might be willing to consider as reputable.)

It's a mindset change for the Relay Techs, but we're much more interested in whether or not the whole package - relay, relay settings, wiring, etc. - does what a relay in that position ought to do than in whether or not each element performs per the factory test. Remember, you're testing code, not anything in the manufacturing (unless you suspect they may have loaded the wrong code at the factory).

For numeric relays, our entire testing assumes that a relay that doesn't have a fail alarm meets factory specs and instead functional tests attempt to prove that it was wired and programmed properly. Ideally the T&E Engineer doesn't even look at the setting while preparing the test plan, but rather looks at what a relay on that terminal of that line should do based on our setting criteria and then select test values based on desired performance. Failure to test properly becomes a QA check on the settings. Assuming the setting are correct and then testing elements will miss a whole heck of a lot of really bad programming.

RE: Is there a benefit in testing OC element drop-offs?

(OP)
Hi David

It depends on the context. Here I am looking at factory acceptance testing (FAT) of a suite of protection relays for a very large generator. Generator relay is a modern digital one from a very reputable supplier. I am not a commissioning engineer though I used to be in a former life. I still compile relay test plans and review test results. For FAT I am of the school of thought that all threshholds (whether they be overcurrent pickups or timers) be tested and verified against the tolerances as per the OEM.

I tend to disagree that one is just testing code. There are digital input modules and output relays as well as communication interfaces to consider as well, besides AC inputs. Then there is the display, LED's, relay faceplate pushbuttons, etc. as well. Certainly they are all dependent on the code for correct operation and the code is of paramount importance. I've also encountered drift, where a distance relay characteristic has significantly shifted from the original. This was picked up with a maintenance test which means even maintenance testing needs to be fairly detailed.

Even though it may seem insignificant, I am still reluctant to forego verifying the reset levels of EF elements. Particularly in a relay with a new version of software or a new relay altogether which does not have a proven track record. Software are programmed by people and to err is human (some say!). It is hoped that all relays are put through a rigorous testing regime prior to being sold, but this is not always the case - and in most cases there are bugs the OEM is not aware of. A few years ago, whilst working for a utility, I came across a relay from a very well known supplier. It was used for a directional OC application and the elements used were actually recommended by the OEM. However, numerous spurious trips later it was discovered that the directional element did not reset as per the specifications. It took much longer to reset and at a much lower level than what it should have. The application was parallel lines and all of them were lost for a fault on one.

When testing a biased differential characteristic which has two slopes, I advocate 5 points of testing these same point should be verified everytime the relay is being tested. These should be the initial pickup and then testing the two turning points, one point each side of the turning points. Point 5 also proves the diff hi-set. This proves that the restraint characteristic is as per the settings (of course there's all the harmonic blocking and stuff to be included as well). And yes, some modern relays have dynamic characteristics such as the SEL-487 which makes things a bit trickier. I got stuck with an OEM recently as he tends to do two or three ad hoc tests points inside and outside the operating region to purely prove stability and operation. Not good enough I say for FAT or maintenance testing. My gripe with this is that ad hoc test points are not easy to replicate for future testing and they do not verify the actual restraint characteristic.

Anyway, that's my twopence worth.





RE: Is there a benefit in testing OC element drop-offs?

I guess it all depends on what "Factory Acceptance Testing" is. If the relays (only a small handful are needed today) are what you've used on your last many generator protection projects then you probably don't need to be testing the element performance. If they're relays that you've never used before, then you need to be doing type testing in addition to the functional testing.

Over the life of the use of a particular relay product, I see the testing as:
* Type/qualification testing - ideally bench testing, prove to yourself, and your company, that the relay does what the manufacturer says it does.
* Installation checkout - basically wire checkout, does each field device poke the correct input, does each output have the desired action, does each analog value show up in the right place?
* Functional testing - test engineer looks at the application and produces test values based on the setting criteria without reference to the settings. This should include cases in all zones of protection, all operational modes (reclose blocked and enabled in line relays for example). The test should verify everything, that the right elements caused the tripping, that the right outputs asserted, that the right SCADA alarms were received in the control center, that the right things showed up on the local SCADA HMI, that the right display points appeared on the relay, that the right target LEDs illuminated, and anything else that matters. This test should be performed with the settings intended to be left in service following the test. If the test requires "testing settings" it is meaningless; those tests should have been performed under the type testing. This is only a valid test if the in-service condition is the as-tested condition; minor allowances made for the need to inject test currents/voltages at the test switches and a breaker simulator to avoid excessive wear and tear on the real breakers. Finish the testing by actually tripping each breaker/lockout involved and verifying correct status inputs.
* In service checks - put the position in service and check all analog inputs. Do the phase angles make sense? Do the magnitudes make sense? Use load below the minimum operate point of any differentials and check operate vs. restraint currents for each differential (can be several if the relay does both phase and sequence differentials).
* Maintenance testing - verify that as-found settings match setting of record, verify that relay self tests are all satisfactory, verify all metering values, verify that all inputs and outputs still operate correctly. No need to run a bunch of test cases on modern relays, the metering checks will catch any uncompensated A/D drift.

Testing a relay with one set of settings and then putting it into service with another set of settings is, in my opinion, less than useless - it instills unjustifiable confidence in the actual performance of the installed system.

Where it really, really has to be right, there may be a very detailed functional test performed on an RTDS system, followed by a more abbreviated functional test following actual installation. The RTDS testing might consider thousands of cases and run autonomously with summary result reviewed; testing for instance all 10 fault cases at multiple locations with the protection of multiple lines being simultaneously tested. Then the as-installed functional testing only needs to test enough cases to verify all of the protection system responses in the as-installed condition. Many utilities use RTDS for all 500kV systems.

Red Flag This Post

Please let us know here why this post is inappropriate. Reasons such as off-topic, duplicates, flames, illegal, vulgar, or students posting their homework.

Red Flag Submitted

Thank you for helping keep Eng-Tips Forums free from inappropriate posts.
The Eng-Tips staff will check this out and take appropriate action.

Reply To This Thread

Posting in the Eng-Tips forums is a member-only feature.

Click Here to join Eng-Tips and talk with other members!


Resources