ungrounded wye power transformer?
ElectricPete
One more question I should have included:
3 - Has anyone seen an ungrounded wye power transformer> Did it operate
reliabily, or were there problems?
electricpete
ElectricPete
"startup" transformer for a generating plant. The configuration of the four
windings are
345kv H winding - solid grounded wye.
13.8 kv Z winding - resistance grounded wye.
2 x 4.6 kv X and Y windings- UNGROUNDED WYE.
The two 4.16kv windings feed two ungrounded 4.16kv buses (radial feed).
This is a very unusual setup to the best of my knowledge in that ungrounded
systems are normally fed from delta-transformer windings. Wye transformer
windings, if present should be grounded at least thru a resistance. In
fact, the nameplate on this transformrer states that the neutrals of the
wyes should be solidly or resistance grounded to prevent "excessive neutral
to ground surge voltages or low frequency voltages". (we obviously did not
follow this recommendation when the transformer was designed/installed 18
years ago).
The transformer operated without problem for 18 years. Recently routine
power factor testing revealed a high power factor at the X0 bushing (neutral
of one of the ungrounded wye windings). Visual inspection revealed a 1/4"
hole clear through the lower porcelain (the part inside the transformer)
such that oil flowed freely between the bushing and the transformer. I
speculate that this hole may have been caused by the excessive neutral to
ground surge or low frequency voltages warned of in the nameplate. But if
so they must have been very high magnitude... the BIL of the failed neutral
bushing is 150kv.
It appears the design is obviously wrong in that a neutral resistance ground
is required. However we cannot add one without performing a edetailed
review of the 4160v ungrounded system relaying and fault current
interrupting, which will take awhile.
One other aspect which may compound the problem is that there is a very long
(1/4 mile) 4.16kv cable run between the low side and the 4160 bus,
introducing the possiblity of ferroresonance, although we have not observed
anything (other than the failed bushing) which might suggest we have this
problem.
My questions are:
1 - Can anyone explain the nature of the expected surge and low frequency
neutral to ground voltages in this configuration?
By low frequency, do we mean harmonics? I know that the design has some
problems in terms of handling the any 3rd harmonics currents (which act like
zero sequence), since there is no path for 3rd harmonic or zero sequence
currents to flow thru the 4160 windings. I would expect the result to be
that any 3rd harmonic loads would result in increased voltage distortion on
the bus, but I can't quite see how we get excessive voltages on the neutral
(particularly on the order of 150kv!).
As far a surge voltages, it's worth mentioning that each of the low side and
high side phase bushings are protected by a surge arrestor. It is difficult
to see how a surge can get in to the neutral. But I guess I can visualize
that if a zero-sequence surge below the surge arrestor rating passes into
the winding, then all 3 surges will meet and add together at the neutral
(similar to doubling in magnitude of a traveling wave at the open end of a
transmission line). Is this the correct scenario to explain the surge
concerns?
2 - We are considering as an interim measure (until we can evalute 4160v
system concerns) adding a 5kv or 10kv station class arrestor at the neutral.
Can anyone foresee any problems with this approach?
Any other comments or suggestions would be welcome.
Thanks!
electricpete
Chico
We just setup one of our major customers with a 600V transformer by
using 3 single phase 347V transformers. (I think the high side is
25kV L-L).
I think the history behind this customer was as follows:
The customer was fed from a 600V delta transformer which was beginning
to show its age. The customer was not confident of the reliability of
this transformer, and so we set out to replace the whole bit. We no
longer provide 600V delta winding transformers, and so didn't have any
of them in stock. Also, we would still have the problem of not having
any spare equipment should anything ever go wrong. The recommended
idea of getting around this was to use 3 single phase, 347V
transformers connected in an ungrounded delta configuration.
I cannot comment on the reliability of this setup, but the senior
engineers involved didn't seem too worried about that. They were more
worried about not being able to detect ground faults, which could
present a potentially dangerous situation.
Sorry I can't help with your problem. I just thought I'd point out
that maybe ungrounded why transformers may not be all that rare.
Ryan
NSPower
Andrew Mitchell
We use ungrounded Y secondaries on the transformers used to feed 60Hz to
ships(11kV/460V, Dyn11). We cannot ground the neutral as it interferes with
the earth fault detection schems on board ships.
There are a couple of problems with this setup...
1). A fault between the hv winding and the secondary cannot be detected
and will cause the secondary to be at least 5kV above earth. (this has never
occured - but it will one day !)
2). Any load connected between a phase and the floating neutral causes
all sorts of weird voltages to occur on the other phases. The worst case is
instead of about 440Vll, you get 40Vll on one phase and 800Vll on the
others.... (this has happened many times....causing a variety of expensive
damage, the ussual culprit is our own electricains... whose excuse is - "i
thought it was like the 50Hz system, and I needed 240v to run this xxxx")
The first problem is the most hazardous, I am thinking of grounding the
neutral via some sort of non linear device that will cause the overcurrent
protection to operate....
Regards
Andrew M
ElectricPete <electr...@technologist.com> wrote in message
news:857kj1$ss5$1...@bgtnsc01.worldnet.att.net...
> > My questions are:
>
> One more question I should have included:
>
> 3 - Has anyone seen an ungrounded wye power transformer> Did it operate
> reliabily, or were there problems?
>
> electricpete
>
>
s falke
> Our utility uses a 4-winding 60MVA 345 - 13.8/4.26kv transformer as a
> "startup" transformer for a generating plant. The configuration of the
four
> windings are
> 345kv H winding - solid grounded wye.
> 13.8 kv Z winding - resistance grounded wye.
> 2 x 4.6 kv X and Y windings- UNGROUNDED WYE.
>
> The two 4.16kv windings feed two ungrounded 4.16kv buses (radial feed).
>
ground
> is required. However we cannot add one without performing a edetailed
> review of the 4160v ungrounded system relaying and fault current
> interrupting, which will take awhile.
You might investigate installing a resistor on the secondary of a
4160:120/240 xfmr with high side connected X0-Grd. Monitor resistor voltage
or current. The added stuff will damp transients and may provide a cure.
This application has been discussed extensively as 'high-resistance
grounding' in IEEE/Industry Applications over the last 20 years. I can't
imagine that the add-on would have an effect on the existing relay scheme to
any significant degree. The usual '59G' voltage relay (IAV51/CV-8/ITE59G)
for this has taps of 1-40V but can handle continuous 208V when a phase is
faulted to ground. I would think that an arrestor connected X0-Grd could
fail violently for any sort of sustained overvoltage. I think the new
'hole' in
the X0 bushing may indicate such a problem.
On a (short) refernce about ferroresonance, you might look over--
http://standards.ieee.org/reading/ieee/std_public/description/dtransformers/
C57.105-1978_desc.html 11-Jan-00
C57.105-1978 (R1999) IEEE Guide for Application of Transformer Connections
in Three-Phase Distribution Systems
Section 7. Ferroresonance
7.1 Qualitative Description
7.2 Transformer Connections Highly Susceptible to Ferroresonance
7.3 Measures to Minimize the Probability of Ferroresonance with Ungrounded
Primary Connection
7.4 Transformer Connections Least Susceptible to Ferroresonance
7.5 Measures to Minimize the Probability of Ferroresonance with Grounded
Primary Connections
7.6 Bibliography
HTH
--s falke
ElectricPete
> 4160:120/240 xfmr with high side connected X0-Grd. Monitor resistor
voltage
> or current. The added stuff will damp transients and may provide a cure.
> This application has been discussed extensively as 'high-resistance
> grounding' in IEEE/Industry Applications over the last 20 years. I can't
> imagine that the add-on would have an effect on the existing relay scheme
to
> any significant degree. The usual '59G' voltage relay (IAV51/CV-8/ITE59G)
> for this has taps of 1-40V but can handle continuous 208V when a phase is
> faulted to ground. I would think that an arrestor connected X0-Grd could
> fail violently for any sort of sustained overvoltage. I think the new
> 'hole' in
> the X0 bushing may indicate such a problem.
All good points. I'm starting to reach the same conclusion with regard to
relaying.
As a T&D engineer, I'm not too familiar with the plant's relaying, but it
appears
to be simple stuff (51 and 51n relays) which doesn't care about ground. I
still want
the plant people to confirm this for me. Also there's the matter of breaker
interrupting capacity.
Maybe this is also a no-brainer, but I'm not sure if previous interrupting
analyses would still be applicable/conservative.
I'll have to pull out my textbook and review.
> On a (short) refernce about ferroresonance, you might look over--
>
>
http://standards.ieee.org/reading/ieee/std_public/description/dtransformers/
> C57.105-1978_desc.html 11-Jan-00
>
> C57.105-1978 (R1999) IEEE Guide for Application of Transformer Connections
> in Three-Phase Distribution Systems
>
> Section 7. Ferroresonance
Great idea. I'll look at it first thing in the morning (we have most of
those standards at work)
Thanks!
electricpete
s falke
Honestly don't know if there's a parallel with medium voltage, but something
that gets overlooked with low-voltage molded-case breakers with the typical
high-interrupting type(say, "65,000 amp" rating) only has a *single-pole*
rating of 8,660 amps. This isn't a problem for solidly-grounded (480Y/277V)
systems but gets crucial for, say, 480V ungrounded or resistance-grounded
systems.
--s falke
ElectricPete
> imagine that the add-on would have an effect on the existing relay scheme
to
> any significant degree. The usual '59G' voltage relay (IAV51/CV-8/ITE59G)
> for this has taps of 1-40V but can handle continuous 208V when a phase is
> faulted to ground.
As indicated in my previous post, that sounded reasonable. Today I talked
to another engineer
and realized that the existing relaying is all based on phase current...
there are no ground current trips (only
an alarm). So if we added a neutral resistor than ground current would have
to build up on the order of load current before we would generate a trip.
That would be a bad thing. Or more likely, the neutral resistor would limit
fault current to a low level such that you would never be able to have
enough ground fault current to trip the phase relaying..... also not a good
thing.
The fix would probalby involve adding a CT and 51N at the new neutral ground
point... plus assorted coordinate residual connected ground relays
downstream to provide for selective tripping. And definitely need to
consider fault current interruption because introduction of new ground on
the wye can increase available fault current.
I did pull the standard on transformer connections that you mentioned today.
It was very instructive.
There were discussions of 3rd harmonics, ferroresonance etc. Strangely
enough, neither of those give problems in this winding configuration (per
the standard). There was also an interesting discussion of capacitive
arcing/restrike phenomenon in ungrounded systems which can lead to high
voltages. The standard recommends that any transformer (wye or delta) which
feeds an ungrounded system should be specified an purhcased for ungrounded
use (basedon this capacitive arcing phenomenon). Unfortunately, there's no
mention of exactly what this means in terms of BIL etc... only that you have
to tell the manfufacturer apparently.
The windings are 75 kv bil. Apparently they are uniform 75kv bil and not
graded toward a lower level at the neutral, which is a good thing for this
application.
I also pulled the C57 standard on new power transformers (I forget the
name./number). There was a statement in there that neutral bushings which
were intended to be floating during normal operation should be floating
during BIL test. I can't figure out exactly how they would float a neutral
during BIL test.... if you need two points to apply voltage across, then
doesn't the neutral need to be one of them? Or do they apply impulse test
on phase-to-phase basis? Or do they apply it phase to ground even when the
winding has no ground connection (wouldn't make much sense) Does anyone
know enough about impulse testing to answer this question. (I looked at the
original manufacturer's impulse test report and there is no mention of the
configuration during the test).
electricpete
r_p_...@my-deja.com
"ElectricPete" <electr...@technologist.com> wrote:
> > You might investigate installing a resistor on the secondary of a
> > 4160:120/240 xfmr with high side connected X0-Grd. Monitor resistor
> voltage
>
regard to
> relaying.
> As a T&D engineer, I'm not too familiar with the plant's relaying,
but it
> appears
> to be simple stuff (51 and 51n relays) which doesn't care about
ground. I
that the "expert" consultants had designed the AUX system with high
resistance grounded neutral that limits max ground fault to 10 Amps
(6.9kv) and then set 51N relays to trip at 50 A.
???
--
R.P.Bayly, P.E.
Sent via Deja.com http://www.deja.com/
Before you buy.
s falke
"ElectricPete" <electr...@technologist.com> wrote
> > . . . The usual '59G' voltage relay (IAV51/CV-8/ITE59G)
> > for this has taps of 1-40V but can handle continuous 208V when a phase
> > is faulted to ground.
>
> As indicated in my previous post, that sounded reasonable. Today I talked
> to another engineer and realized that the existing relaying is all based
on
> phase current... there are no ground current trips (only an alarm).
> So if we added a neutral resistor than ground current would
> have to build up on the order of load current before we would generate a
> trip. That would be a bad thing. Or more likely, the neutral resistor
would
> limitfault current to a low level such that you would never be able to
have
> enough ground fault current to trip the phase relaying..... also not a
> goodthing.
1. That's why the lo-range overvoltage relay is desirable.
2. To (effectively?) parallel a resistor with a rating less than or equal
to the capacitive reactance on the 4kV circuit will damp phase-to-ground
transients to a moderate degree.
--s falke
. . .
> electricpete