Negative Seq. Relay setting for Xfmr 3

[SMB1]

I have 69/13.8KV Xfmr Delta(HV)/Way(LV)
I want to set the –ve seq. relay to coordinate it with OC protection in the LV side

If the L-G fault is 4500 A. What is the value of the –ve sequence in the HV side
I got the result from Aspen Oneliner but I could not get the same result by hand.
So, what is the the formula shall I use?

 

[jbartos]

Suggestion: There are two aspects involved:
1. The fault current expressed in symmetrical components has a positive sequence component, a negative sequence component and a zero sequence component.
2. The neutral may have a system-grounding resistor limiting the L-G fault current.

 

[SMB1]

Thank you jbartos for your reply.

Yes in the LV side i have the +ve , -ve and Zero sequence but in the HV side i have +ve and -ve only, there is no zero component.

Now, I want to get the -ve seq. in the HV side.
What i did is that:

I got the (-ve seq. current LV side)*(Xfmr ratio)=-ve sequence in the HV side.

Unfortunately, I got different value by using the program.

Is my formula correct?

 

[ScottyUK]

SMB1,

Does your transformer have a phase shift eg Dy11, etc, or is it a Yy0 for example which has no shift. Is this the source of your error?


Scotty.

 

[busbar]

SMB1 — Negative-sequence protection is typically applied to rotating machines. Can you please explain its purpose in a transformer-protection application?

 

[SMB1]

I have 69/13.8KV Xfmr Delta(HV)/Way(LV)

ScottyUK: My Xfmr is Dyn1

busbar: During L-G fault in 13.8KV outgoing feeders, as you know, there will be a -ve seq. component in the delta side(HV). So, I can coordinate the GOC relay at LV side's feeders with the -ve seq. relay in the HV side.

I hope it is clear
(All numerical O/C relays having -ve seq. setting feature)and I want to make use of it.

 

[dpc]

Your equation is correct. The negative sequence current seen on the primary is equal to secondary side negative sequence adjusted by the effective turns ratio.

If you are off by 1.732, you need to account for the delta-wye connection. The actual turns ratio of the two windings is not the same as the effective turns ratio at the transformer terminals.

busbar,

With advent of digital relays, negative sequence current OC relaying is essentially "free". It can be used to some advantage, especially for protection against line-line faults, since these will have neg. seq current while 3-phase faults do not. Because the neg.seq element is insensitive to balanced load current, it can be set below transformer full load current, unlike the phase OC elements.

And the neg seq on the primary of the delta-wye transformer will also detect SLG faults on the secondary. The benefit of this is questionable, since the same relay will also provide residual OC protection against ground faults from the secondary CTs.

On the other hand, I've never actually put the neg seq elements into service on any of the relays I've set. Generally, you want some other OC element to trip first if the fault is outside the transformer differential zone, so you end up needing to coordinate with phase OC elements anyway.

I think there is a tendency to think that we must use every feature on these relays because the features are there. But the more trip functions that are active, the greater the risk of a false operation.

 

[busbar]

dpc, I appreciate your comments. A concern would be that a 69kV negative-sequence overcurrent function would suitably coordinate with various overcurrent functions on the 13.8kV side. A 69kV breaker operation would have to be secure from 13.8kV faults in industrial or utility settings where (mis-)coordination with feeder fuses might be involved.

Timely restoration efforts following power-transformer outages initiated by primary protection seem to bring about added fear and loathing [hopefully without career-changing consequences] in the often tense “discovery/evaluation” period.

I suppose that I had too narrow a view that primary-side negative-sequence overcurrent protection could erroneously be interpreted as general transformer or {potentially conflicting with} lowside-feeder protection.

I wonder if Professor Schweitzer in his basement workshop imagined the vast new aspects {or can of worms to some} in system protection he was creating,

 

[DEPU]

How did you calculate your ratio?
you must use L-N voltage for 13.8kv
for line to ground fault the ratio=(LL)/(LN)

 

[jbartos]

Suggestion: It appears that the secondary side of the transformer, 13.8kV, 4500A L-G fault is very high and potentially destructive. S=13.8kV x 4.5kA/sqrt3=36MVA. Which system grounding is there?

 

[PowerSim]

Negative sequence relay can be connected in the primary supply to the transformer and set as sensitively as required to protect for secondary phase-to-ground or phase-to-phase faults. This relay will also provide better protection than phase overcurrent relays for internal transformer faults. The relay should be set to coordinate with the low-side phase and ground relays for phase-to-ground and phase-to-phase faults. The relay must also be set higher than the negative sequence current because of unbalanced loads.

 

[jbartos]

Suggestion: Visit

http://doc.newage-avkseg.com/doku.pdf/pl/xs2/xs2_e.pdf

for a negative sequence relay that has a universal application.
Its characteristics show that it will be applicable for time setting above one second. When it comes to the negative sequence current on the primary side, this relay might serve as a “second? third? ..” line of defense in the transformer protection.
IEEE Std 242 Buff Book does not address the negative sequence relay for the transformer protection. There will be other much faster relays that will be in operation first.

 

[SMB1]

Thanks for all of you
DEPU: I used same ratio as you wrote

jbartos: We have a solodily grounding system
4500A is the end of line fault
the colse in fault is about 9800A

 

[stevenal]

I'll disagree with DEPU. The negative sequence quantities are a balanced set. Unless you are looking for the negative sequence currents inside the delta,(an unlikely CT location) use the effective turns ratio from the line to line quantities, not the actual ratio.

Don't forget that your line to ground fault of 4500 amps has equal parts +, -, and 0 sequence. Your - sequence component then is 4500/3. How's 300 amps primary sound? (confirmed by Aspen V2001G.

Don't forget that some relays (SEL) input negative sequence quantities as 3I2, while others (ABB) use inputs of I2. Make sure you're using the right quantity in Aspen.

 

[dpc]

DEPU,

As stevenal says, I was referring to current that would be measured in a line CT at the transformer terminals, not in the delta.

So you just go by the nameplate voltage ratio for the transformer.

 

[DiscoP]

To use it effectively, you should set it just above the negative sequence current measured for a phase to phase fault on the LV side of the transformer.

If you set it lower and grade it with overcurrent protection, what is the point, wouldn't HV side overcurrent do the job.

This setting will not be very sensitive, but could clear a fault quicker than the back-up overcurrent protection on the HV side.

But I am not sure if it is worth the trouble

 

[rbulsara]

Pardon my ignorance if my thought is not valid.

Beaside havin the same question as busbar,I thought -ve sequence relays are usually set to some % of the rated currents with time delays in ‘Several’ seconds, while the secondary side LV GOC protection would be set at low currents and more importantly to trip within ‘0.5’ seconds.

This by itself should discriminate the primary from the secondary side trip on GF.

A very sensitive –ve sequence setting on primary will result in nuisance trips. What is the type of relay you have?

 

[stevenal]

Not much trouble. Most modern relays include the function, you just need to apply it. You can set it to back up low side protection for line to ground faults and line to line faults, without worrying about what load current might do. Tools like Aspen will allow you to quickly see how protection coordinates across the delta wye.

 

[SMB1]

Comments on stevenal (Electrical) Sep 2, 2003

//I'll disagree with DEPU. The negative sequence quantities are a balanced set. Unless you are looking for the negative sequence currents inside the delta,(an unlikely CT location) use the effective turns ratio from the line to line quantities, not the actual ratio.//

96/13.8KV Delt(HV)/Y(LV)
1. During LG fault in the LV side we will have L-N voltage not L-L. As a result, the Xfmr ratio will be LL/LN

2. Also, during the LG fault
(-ve seq. outside the delta = -ve seq. inside the delta)
this should be clear.
To prove this:
any LG fault in Y side will behave as LL fault in the Delta side.
The fault current will flow on one winding (Y side). After applying the (AmpsXTurns) concept the current in the Delta side will flow also on one winding.

So, in the (HV)Delta side during the LG fault in(LV)Y side

the line current= the phase current
i.e,
(-ve seq. outside the delta = -ve seq. inside the delta)

 

[stevenal]

SMB1,

Again, negative sequence currents are a balanced set, just like your positive sequence load currents. For this case, use effective turns ratio just as you do for load current.

Use actual turns ratio for phase quantities. Do this for a sanity check. 4500 Amps > 520 amps primary line to line. If the other two line to line currents are zero, then your line to line quantity is equal to the adjacent phase quantities. Answer 520 amps primary on the two affected phases.

Back to sequence quantities: Positive and negative primary phase currents are both 300 amps. Zero sequence is zero outside the delta. Add them up. One non-zero quantity is lagging the secondary current by 30 degrees, the other leading. The resultant is 300*2*cos(30)=300*sqrt(3)=520 amps primary. Sanity check complete.