Estimating transformer in rush.

[haze10]

Because of the age and make of my transformers I can no longer get information from the manufacturer. I'm installing some new protective relays on the feeders to my building transformers. I'd like to set Inst at 2x no load inrush current. But I can't find what the in rush is. Is there a way to estimate it. The transformers are 1500 KVA padmount 5.0% impd, 4.16kv to 480V, delat/wye. Any best way to guess at the inrush?

 

[GroovyGuy]

Howdy Haze;
I would think that the energization inrush is somewhere in the 7 to 10 X IFL for a 1500KVA padmount. This of course assumes that the terminal voltage remains at, or close to, 1.0puV during the energization period.
GG

"I have not failed. I've just found 10,000 ways that won't work." Thomas Alva Edison (1847-1931)

 

[stevenal]

According to Cooper:

25XFLA for 0.01s and 12XFLA for 0.1s. This is for "hot load" inrush, so to get to no load inrush subtract 1XFLA.

 

[RRaghunath]

Agree with Stevenal.
IEC/IEEE also mention the same.
For Instt O/C element pickup setting on transformer primary, the industry norm is: 12xFLC or 130% of transformer through fault current, whichever is higher.

Rompicherla Raghunath

 

[stevesummers]

Hi,

I am from IEC world, so things are a little different.

the general inrush multipliers i use are 12x for small transformers <500kVA and then dropping down steadily to around 6x for larger say 10MVA transformers. in your case i would use a figure of around 8xIn.

I am not sure stevenal / RRgahunath are correct at 25xFLC for 0.01s.. I have not seen this figure used before anywhere.

You also need to consider the inrush duration - generally higher inrush figures of 12x In decay more quickly (2 or 3 cycles) - while smaller inrush figures of 6x decay much more slowly over several hundred ms.

 

[stevenal]

I am not sure stevenal / RRgahunath are correct at 25xFLC for 0.01s.. I have not seen this figure used before anywhere.

Exactly what the forum is for; learning new things. See Electrical Distribution-System Protection Third Edition, from Cooper Power Systems, page 90.

 

[davidbeach]

First step would be to agree what FLC means. If FLC is based on self-cooled it will be a much lower number than if FLC is based on top-nameplate. Max inrush is the same value in amps in either case but will have a higher multiplier in one case than in the other. ANSI bases things on self-cooled; IEC seems to base things on top-nameplate.

 

[stevesummers]

Hi,

I can't seem to find the Cooper book (can someone give me a link to amazon so i know what i am looking for)..

I had a quick trawl through my library and see the figures that stevenal came up with, so they seem to be valid.. just surprising as i had not heard of the 25x at 0.01s before.

Based on my use of SKM and ETAP and other protection books, it seems to be standard practice to go with a nominal figure of (8-12x FLA) at 0.1s. the 25x figure seems to be more applicable with fuses, where i guess it could actually melt, whilst a relay would not respond so quickly.

 

[stevenal]

In my experience 1500 kVA padmounts are self-cooled. If multiple kVA ratings were involved, I would use the self cooled rating.

I'm not sure where to find the Cooper book anymore. Mine's from 1990, and I believe there are newer editions. Might try contacting your Eaton rep.

You don't want to trip at 25X, 0.01s; you want to be secure for it. Join the two points with a straight line on your log-log graph, and you have a lower bound for your protection curve. Cooper adds a few more points for cold load for a more complete lower bound.

 

[RRaghunath]

davidbeach, good point you raised.
Since the inrush is based on the transformer active materials, I think it is logical to consider FLC to be base rating (ONAN), for the purpose of inrush current computation.


Rompicherla Raghunath

 

[prc]

Some salient features of transformer magnetizing inrush current:
1) Inrush current peak (kA) depends on
- Point on voltage wave when switching on (max at voltage zero)
- Amount of residual flux in core( max with max residual flux)
- Air core reactance(AR) of winding that is energized (inrush peak less with increased AR- AR varies with rated kVA,% impedance and winding dia HV/LV : AR increase with rating,impedance and diameter of winding)
- Applied voltage peak ( inrush more for higher kV)
2) Time constant for decay of inrush current of modern power transformers -5-10sec (more for higher rating)
3) When referring to times FLC,consider peak of FL or RMS value?
4) Inrush current only in two phases in opposite directions ( Either positive or negative in both positive and negative voltage half cycles) Neutral current in opposite directions in positive and negative cycles of voltage.

 

[unclebob]

Always thought that inrush current was independent of load.

 

[waross]

If you have the %impedance and the XR ratio, you can extract the resistance of the transformer.
You can use the resistance to calculate an upper limit of the surge. This will be on the high side as it neglects the air core inductive reactance of the transformer.
The value may be of interest as the inrush will be lower.


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[krisys]

The above statement is universally true.

Sometimes it is prudent to set at a reasonable value (industry standard value), with a time delay of 100 ms (five cycles). The potential sub-cycle inrush will die down at 5th cycle. So generally no risk of tripping.

As such the 25x is the generally a sub-cycle inrush, and the relay is not sensitive to pick up at this time. As the sub-cycle inrush might be rich in DC component. DC restraint, second harmonic blocking, fifth harmonic (normal for over voltage)protection etc. are common in the modern relays.

As such OP is installing the new relay. I presume that the new relay will be having all these features. So in my view there should not be any issue with the inrush.

A practical note:
The issue is only during the transformer energisation. Even if the transformer trips, so what? As you have a doubtful setting for the inrush, just switch on again and if you feel, just increase the setting a little bit.
Switching on the transformer is normally starting up of some system. Hence, you are not adding any additional disturbance to the system.

 

[prc]

waross,I am afraid your judgment on inrush from fundamentals seems not correct,though you are always correct in this forum ! Air core reactance will always be there and you cannot avoid it. Variation in inrush will happen,depending on the point on voltage wave where you switch in -max at voltage zero and minimum at voltage peak.
Quote-"The issue is only during the transformer energisation. Even if the transformer trips, so what?" [Krisys, can you tell me how you put the quote notation like that in your posting?] A lot esp with MV transformers switched in by VCB. VCB switching voltage surges are maximum under no-load and worst consequences if inrush is broken. I had to handle tens of transformer failures esp with dry type transformers where tripping happened during inrush time. Many times during relay commissioning. Forgot to activate second harmonic restraining feature in relay, wrong settings, trial and error setting procedure as suggested by krisys. Even if not caused a failure, such trippings will do irreversible damage to MV transformer insulation.

 

[itsmoked]

Greetings PRC!
Are you saying it's imperative to make sure the protection always rides thru long enough to get down to only a couple of FLA before opening? So if one is going to field adjust start long and work back to short?


-----------------------------------------------------------------------

Krisys, can you tell me how you put the quote notation like that in your posting?

Use this for the quoting function:

You can skip the "who" if you want to by hitting ENTER.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[waross]

Greatest calculated value of inrush. Based on resistance alone.
Next greatest calculated value of inrush. Based on impedance of resistance and air core reactance.
If the value calculated with resistance alone is 20X FLA then there is no need to use a setting of 25X FLA.

Will this work?
Use the transformer % impedance to calculate the available short circuit current.
Multiply by √2 to convert from RMS to peak current.
Multiply by 2 for a fully offset wave form.


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[prc]

itsmoked, thank you for the tip. Yes. With VCB and MV transformers(esp dry type), it is always good to avoid the tripping of inrush and no-load current. Make use of data of similar transformers or ask OEM for the calculated values instead of doing trial and error at site. This was discovered in early 1970's by GE engineers and also by Mr Greenwood.

 

[stevenal]

Even if the transformer trips, so what?

See the OP, these are feeder relay settings. Probably don't want to trip the feeder while energizing a single transformer.
And even if the subject were transformer protection, a trip on energization should be investigated, since the first assumption should always be that a fault is present.

 

[krisys]

In the modern relays, the inrush detection and blocking function is available. By enabling this function, I believe that the identification of inrush and the fault current can be identified. In turn a fool proof setting to take care of the transformer energisation inrush and to trip during the energisation of the faulty transformer.

Can relay experts in the forum come in at this stage?