Phase Sequence - Phase Failure (47) relay

[JoeStillman]

I am using a 47 relay in an auto transfer scheme for a power system with a lot of voltage distortion (14% thd). I suspect that the poor voltage waveform is faking out my 47 relay and causing false trips. I wonder if anyone here has had this problem.

Comments?

 

[aelectricalman]

I am using a 47 relay in an auto transfer scheme for a power system with a lot of voltage distortion (14% thd). I suspect that the poor voltage waveform is faking out my 47 relay and causing false trips. I wonder if anyone here has had this problem.

Comments?

They are very sensitive to undervoltage and imbalanced voltages. They don't do well with a lot of Delta systems due to single phasing.

 

[aelectricalman]

I am using a 47 relay in an auto transfer scheme for a power system with a lot of voltage distortion (14% thd). I suspect that the poor voltage waveform is faking out my 47 relay and causing false trips. I wonder if anyone here has had this problem.

Comments?

I would check my phase sequence. That is really the only problems I've ever seen out of those.

I've heard of problems when tapped from a capacitve source but its mostly in the form of draining capacitors.

 

[jim dungar]

I am using a 47 relay in an auto transfer scheme for a power system with a lot of voltage distortion (14% thd). I suspect that the poor voltage waveform is faking out my 47 relay and causing false trips. I wonder if anyone here has had this problem.

Comments?

Not an un-common occurrence. There is no easy solution.
Ideally you could get a copy of the voltage waveform and send it to the 47 relay manufacturer for assistance in developing an appropriate line filter or to find an alternative relay.
Practically, you will probably keep substituting products until you find a working one or you give up.

FWIW, my preferred source has been products from Taylor Electronics.

 

[iceworm]

I am using a 47 relay in an auto transfer scheme for a power system with a lot of voltage distortion (14% thd). I suspect that the poor voltage waveform is faking out my 47 relay and causing false trips. I wonder if anyone here has had this problem. ...

No, not that specific problem. Ansi 47 is negative sequence overvoltage. I have only looked at the analysis of one. So keep in mind: this is stastical sample of one. And not exactly my area of expertise. (Disclaimer - in case you missed that)

The relay is looking at both phase angle offset and phase voltage imbalance. The phase angle offset is generally not an issue, so you are thinking there is a voltage imbalance caused by the voltage distortion?

If so:
If the fundamental voltage is balanced, I would expect the three phases, distorted as they are, would look about the same. Assuming I'm not clueless on the relay algolrithm analysis, that should not give a negative sequence. So, how does the fundamental voltage balance look? Or, are the loads such that the distortion is different on the three phases? You may want to dig into the relay calculation to see exactly how it is figuring the negative sequence.

The issue I dealt with was a meduim heavy 480V single phase load (got to love them designers) causing a bit of voltage imbalance. The 480V imbalance reflected back to the 13.8kv bus and was tripping the 13.8 incommer. I ended up pushing out the negative sequence trip to where in was essentially a phase loss or a phase reversal relay. We did not happen to have any heavily loaded 480V motors so a more than minor voltage imbalance (~5%) wasn't going to over heat anything.

I'm sure there are members more knowledgable than I. Hopefully one will answer and I'll get some education as well

ice

 

[aelectricalman]

No, not that specific problem. Ansi 47 is negative sequence overvoltage. I have only looked at the analysis of one. So keep in mind: this is stastical sample of one. And not exactly my area of expertise. (Disclaimer - in case you missed that)

The relay is looking at both phase angle offset and phase voltage imbalance. The phase angle offset is generally not an issue, so you are thinking there is a voltage imbalance caused by the voltage distortion?

If so:
If the fundamental voltage is balanced, I would expect the three phases, distorted as they are, would look about the same. Assuming I'm not clueless on the relay algolrithm analysis, that should not give a negative sequence. So, how does the fundamental voltage balance look? Or, are the loads such that the distortion is different on the three phases? You may want to dig into the relay calculation to see exactly how it is figuring the negative sequence.

The issue I dealt with was a meduim heavy 480V single phase load (got to love them designers) causing a bit of voltage imbalance. The 480V imbalance reflected back to the 13.8kv bus and was tripping the 13.8 incommer. I ended up pushing out the negative sequence trip to where in was essentially a phase loss or a phase reversal relay. We did not happen to have any heavily loaded 480V motors so a more than minor voltage imbalance (~5%) wasn't going to over heat anything.

I'm sure there are members more knowledgable than I. Hopefully one will answer and I'll get some education as well

ice

You sure about that? I thought it was designed for undervoltage protection only? Thought it was specifically an undervoltage relay designed to lock out under single phasing (lost phase)

 

[iceworm]

You sure about that? I thought it was designed for undervoltage protection only? Thought it was specifically an undervoltage relay designed to lock out under single phasing (lost phase)

Well, first, pay attentin to my disclaimer.

Past that, here is what I recall:

[h=6]ANSI 47 ? Negative sequence overvoltage[/h]Protection against phase unbalance resulting from phase inversion, unbalanced supply or distant fault, detected by the measurement of negative sequence voltage.

The one I looked at performed the following calculation:

Vns = 1/3 x (Va + aVb + (a^2)Vc)

where:
Vns is the negative sequence voltage
Va, Vb, Vc are the three phase voltage vectors
"a" is a 120deg phase shift operator
"a^2" is a 240deg phase shift operator

And an Ansi 46 looks at the current

[h=6]ANSI 46 ? Negative sequence / unbalance[/h]Protection against phase unbalance, detected by the measurement of negative sequence current:

• Sensitive protection to detect 2-phase faults at the ends of long lines
• Protection of equipment against temperature build-up, caused by an unbalanced power supply, phase inversion or loss of phase, and against phase current unbalance

And I'm pretty sure you already know all this.

ice

 

[JoeStillman]

Thanks for the input guys.

The 60 Hz voltage is well balanced, but there are a LOT of harmonics. The 5th, 11th, 17th and 23rd are negative sequence harmonics, which I suspect is tricking the relay into thinking I have a phase reversal.

I'm going to try and find a low-pass filter that can go upstream of the relay.

 

[kingpb]

SEL uses the 47 as Phase Reversal.

After the motor is installed, phase reversal is not normally necessary, in most applications. I would recommend disabling the function, unless the process needs it.

 

[JoeStillman]

SEL uses the 47 as Phase Reversal.

After the motor is installed, phase reversal is not normally necessary, in most applications. I would recommend disabling the function, unless the process needs it.

Believe me, I would do away with it if I could. It is part of our auto-transfer scheme. I get a gen start and initiate transfer on phase-unbalance or phase reversal. I've never seen a transfer scheme without these functions.

 

[iceworm]

... I get a gen start and initiate transfer on phase-unbalance or phase reversal. I've never seen a transfer scheme without these functions.

Hummm... I have only worked about two that have a utility connection and significant generation. As I recall, the utility connection is set to open on negative sequence, (as well as several other conditions). The generator(s) start and transfer on dead bus.

Your spec says to start and transfer on negative sequence and the bus is still live - but the relays are sensing a signature expected to cause emminent failure ? I've not seen that before.

Or not. You are saying the negative sequence does trip the main and that initiates the gen start and transfer?

So educate me: I assume you can't just set the trip outside of the normal operating envelope - or you would. So why not?

ice

 

[JoeStillman]

Hummm... I have only worked about two that have a utility connection and significant generation. As I recall, the utility connection is set to open on negative sequence, (as well as several other conditions). The generator(s) start and transfer on dead bus.

Your spec says to start and transfer on negative sequence and the bus is still live - but the relays are sensing a signature expected to cause emminent failure ? I've not seen that before.

My intention is to protect against 1) utility restoring with a phase reversal (it could happen.) 2) motor regeneration on 1? failure - enough regeneration to sustain my UV relays, but in reverse phase.

I assume you can't just set the trip outside of the normal operating envelope - or you would. So why not?

The THD is 14%. The widest operating envelope I can get is 15%. I'm a little uncomfortable with the 1% margin.

 

[ATSman]

Sorry if I missed it but can you give us the mfg. and model#
of relay?

 

[JoeStillman]

Basler BE3-47N. I emailed them but so far, just hearing crickets.

 

[kingpb]

But if you use a sync check relay it, would be looking at angle and thus a phase reversal (180 deg out) would be covered, due to being way outside the angle limits.

Thus eliminating the 47.

 

[JoeStillman]

I'm pretty sure the sync checks are 1?, but that's a good point.

 

[iceworm]

I was pretty sure you knew anything I knew - and I was right.

My intention is to protect against 1) utility restoring with a phase reversal (it could happen.) 2) motor regeneration on 1? failure - enough regeneration to sustain my UV relays, but in reverse phase. ....

Yes and yes. As I understand, a negative sequence OV is what you want.

You are dealing with an issue where the normal operating signature is very close to the failed signature - As you have said - hard to tell them apart.

... The THD is 14%. The widest operating envelope I can get is 15%. I'm a little uncomfortable with the 1% margin.

Yes

I'm pretty sure the sync checks are 1?, but that's a good point.

Any I've seen are 1ph.

Basler BE3-47N. I emailed them but so far, just hearing crickets.

Bummer, keep wacking at them - but you already knew that.

Have you started looking for another relay - one with a bigger window? I suspect you have.

... I'm going to try and find a low-pass filter that can go upstream of the relay.

Short of finding another relay, I like it.

ice

 

[JoeStillman]

We widened the window on the 47 relays and they are holding fine. Although the THD is lousy at 14%, only a portion of it is negative sequence. I think with the wider tolerance, we are over the total negative seq. harmonic levels.