We are using Single phase power transformers connected in Delta / Star (3-phase bank) as Generator Step Up transformers. We rewound one of those and the winding capacitance for this rewound transformer is almost 50% higher than existing other two transformers. Will it be any impact on the performance, if the transformers with unequal winding capacitances are connected in Delta / Star for a star winding generator of 210MW capacity? All other parameters are matching with existing transformers.

 

Krish... there should be no ill effects at the power frequency. But, the difference could affect the ability to withstand transient phenomena caused by very high dv/dt waves initiated by lightning and switching surges.

The data supplied is insufficient to provide a quantitative analysis. I suggest you ask the original manufacturer to conduct a transient analysis to rule out ferroresonance!

Regards, Phil Corso

 

Hello Phil Corso,

thanks a lot. I fully agree with you as far as ferro-resonance point of view is concerned. But, we worry about the probable unequal voltage distribution with respect to ground at LV side (Generator side) due to unequal distributed capacitance.

<System description>:
210MW, 3 phase, 15.75kV Double star Turbo-Generator with reactance grounding (via neutral grounding transformer) is connected with 3 nos. single phase GSU transformers through Isolated Bus Duct arrangement. Power evacuation is at 230kV level

<Turbo-Generator>:
210MW, 247MVA, 3-phase, 15.75kV, 50Hz, Double star, 3000 rpm
Neutral Grounding Transformer (NGT) = 15.75kV / 240V, 50kVA
Reactor connected at secondary of NGT = 0.986 ohm

<GSU transformers>:
3 x 84MVA, 15.75kV / 230/sqrt.3 kV, Oil filled two winding, % impedance = 14.4%
Single phase units connected in Delta / Star (Yd1) configuration
230kV side star neutral - solidly grounded

<Capacitances>:
Generator = 0.22 micro farad / phase
Bus duct = 6 nano farad / phase
GSU transformer HV - Ground with LV guarded (Old / New) = 1.6 / 2.2 nano farad
GSU transformer LV - Ground with HV guarded (Old / New) = 3.5 / 5.8 nano farad
GSU transformer HV - LV with ground guarded (Old / New) = 4.7 / 7.0 nano farad

<Problems faced already>:
When we excite the Generator with existing GSU transformers at mid tap position, we experience unequal voltage distribution with respect to ground at generator terminals (instead of 9kV it was around 8.5kV (R), 9.62kV (Y), 9.6kV (B)). But, other than mid tap position, it is Ok. Since a change (tap position) in HV side affects the LV side, we give due importance for winding as well as inter-winding capacitance of the Single phase transformers in our station.

<Rewinding>:
It was not done by Original manufacturer, since this type has become obsolete now.

<Technical Snags>:
1. If we connect 2nos. old transformers and 1no. new transformer in Delta / Star configuration, will it be any unequal voltage problem at generator terminals with respect to ground?

2. With such mismatching in capacitance values, what would be the consequences on Generator and GSU transformer due to ferro-resonance?

If you need still more data for quantitative analysis, I would furnish. Requesting your valuable analysis and outcome.
With warm regards
-Krish

 

Krish... please provide additional data:

1) The phase-to-ground voltage values yield a Voltage-Unbalance-Factor (VUF) of 7.8 and 7.7 based upon NEMA and IEC stds, respectively. This value is considered high. But before panic sets in, please provide associated R-Y, Y-B, and B-R line-to-line values.

2) The Generator's capacitance-to-ground value wipes out the other elements, essentially making them ineffective in the ph-grnd unbalance noted above. How were the values determined... by measurement? Or "text" derived?

3) Although you have provided gen'r capacitance-to-grnd values, what are the comparable values of the GSU 15.75 kV windings?

4) How are the 15.75kV PT secondaries connected... open-delta or grounded-wye?

5) Was failure-mode of single-ph GSU determined?

6) Can you send a simplified SLD to me?

7) More on effects of high-side nF's later!

Enough for now, Phil Corso (cepsicon[at]aol[dot]com)

 

Krish... further to earlier e-mail:

Remy Q3, sorry, I overlooked 3.5/5.8 value!

Q8) Which phase is represented by "new" xfmr... R-Y, Y-B, or B-R?

Q9) Is the NGT comprised of a single 3-ph distribution type xfmr, three 1-ph xfmrs, or a zig-zag transformer (NGZ)? Rated ph-ph, or rated ph-ph/sqrt(3)?

Q10) What are the phase-to-ground capacitances of the surge arrestors, if installed?

Q11) When phase-to-grnd voltages were measured, was NGT connected?

Q12) If answer to Q11 is yes, what was current "flow" through neutral?

Q13) Aprox age of system?

Phil

 

Thanks Phil.. my answers are...
Additional Data

1) Please discard the previous values since I got some good data from my old data base. The PT (15.75kV / 110V) voltages measured in unit-1 of our power plant, at Tap No. 9b. under generator no load condition.

R-N = 68.3V, Y-N = 65.7V, B-N = 58.6V
R-Y = 110.2V, Y-B = 110V, B-R = 110V

2) Could there be any difference between source side and load side distributed / winding capacitances?

All capacitance values are determined by Automatic numerical type tan delta test kit at 7.5kV

3) Already furnished

4) 15.75kV PT secondaries are connected in grounded-wye. No open delta secondary is available. Auxiliary PTs are used for open delta protection.

5) Not able to understand. Please give one example

6) SLD and other data available in the following link
http://www.4shared.com/file/180949814/af99b384/GSU_data.html

7) Thanks. Eagerly expecting

8) New xtmr means the “rewound” one. This has not yet been connected in Delta and is the questionable one,now we are discussing.

9) NGT is single phase, distribution type, dry, two winding xtmr.

10) Surge arresters not installed due to high winding capacitance of Generator. Lightning arresters are available. Leakage current is 0.461mA at 10kV.

11) Yes, phase-to-grnd voltages were measured with NGT connected. In the attached file, you can find the residual voltages measured directly on 15.75kV Bus at 3000 rpm.

12) Current could not be measured, but the voltage across the reactor was around 8.18V (instead of 3V normal). Could be derived as 126 mA at primary with an assumption that current is fully reactive.

13) Various units between 20 - 30 years.

With warm regards,
Krish

 

Krish, analysis thus far:

o Voltage Unbalance. Data derived from PT readings shows Ph-Ph VUF is less than 0.2 pct while the Ph-Grnd VUF is about 8.8 pct. Thus indicating harmonics are not an issue. And because GSU+Duct capacitance, ignoring unbalance, are considerably less than that of the Gen, then unbalance is in the 15kV system.

o Zero-seg current is more than double that expected.

o Is LA leakage-current the total? Or just that of one arrestor. Either way it is much more than that of gen.

o Remy Q5) It questioned if the mode-of-failure was determined for the faulted xfmr, i.e., insulation breakdown, over voltage, tap mechanism, etc. The cause of failure together with age could could rule-out lightning or switching induced surges.

My preliminary conclusion... Look at LA's. Since the SLD shows that they can be isolated, I suggest they be Meggered!

Phil

 

Krish, further to my 25-Dec-09 (13:21) post...

I should have said measure capacitance, when possible, of all 15kV components, such as LA's, PT's, excitation xfmr, and aux load xfmrs, using your Tan-delta test equip.

Phil

 

OK,

Gentelmen that's Excellent brainstorming I believe it so much.

Thank you

 

Thanks a lot Phil... I am giving some more information for your quantitative analysis.

a) 0.461mA leakage current at 10kV is for a single Lightning arrester. Leakage currents are equal for all the three lightning arresters. Magnetising currents of all PTs are also equal.

b) Q5) The rewound xfmr was originally kept as spare in idly energised condition (230kV side energized keeping 15.75kV side open). It got fire due to failure of OIP bushing. Hence, we had to re-wind. Before kept as spare, it had been in service in one the units. Some couple of months ago, a 30 year old GSU xfmr got failed due to same kind of OIP bushing failure. But there was no damage to winding.

c) <Detailed History> Out of 5 sets of GSU transformers in our power plant, 4 sets have experienced with 9b tap (mid tap) problem. If the GSU tap is kept at 9b, Generator trips on Stator Earth Fault (95% stator Earth fault & 15.75kV bus open delta). We used to test (Hi-pot etc.) all the components in 15.75kV system, but nothing could reveal. So we would restart the unit, but once again it would get tripped on the same Stator earth fault protection with a disturbance record of unequal voltages in 15.75kV system with respect to ground.

Then, as a routine measure, we would change the GSU tap to 10 or 8 (other than 9b) to solve this problem. The problem in 15.75kV side gets solved just by changing the GSU tap position ie) change in 230kV side influences the 15.75kV side. It is a strange for us. So, We concluded that inter-winding / winding capacitance of GSU might be the cause. That is why, we are much concern about the variation in the winding capacitance of the rewound xfmr.

d) When the OEM was referred for 9b problem, they simply advised us to change the tap position. No fruitful solution yet.

e) Unequal voltages could be realised even at 3000 rpm (no load) due to residual magnetism of the rotor

f) We have faced this problem with all kinds of protection as well (Electromagnetic (Alstom), Static (ABB), Numerical (Beck-with)

g) Is there any other parameters / conditions to be matched to form Delta / Star GSU xfmr bank using single phase GSU xfmrs except voltage ratio, capacity, tap positions, % impedance?

h) THD of our Generators are around 3%

Please help us if you can give some valuable suggestions.
With warm regards,
Krish

 

Krish.. a perfect case of "catching the mice, and allowing the elephants to go free!"

If physical inspection revealed nothing suspicious with taps, then I suggest a rather simple test.

However, you are probably aware of my idiosyncrasy regarding anonymity, please contact me off-list!

Regards, Phil Corso (cepsicon [at] aol [dot] com)

 

Krish... An analysis of the Table you provided showing the test results of six component configurations, repeats almost exactly the VUFs I forwarded to you. Following is an abridged version of your table with comments/and questions:

A) Gen, Bus, VTs, w/o GSUs.
Are you able to excite gen from alternate source, because Excit'n xfmr not connected?

B) Gen, Bus, VTs, w/3 GSUs.
Is "new" GSU included as one of the three?

C) Gen, Bus, VTs, w/2 GSUs.
Is the unconnected GSU the "new" one?

D) Gen, Bus, VTs, w/3 GSUs, Tap 9b.
No comments at this time!

E) gen, Bus, VTs, w/3 GSUs, Tap 10.
No comments at this time!

F) Gen, Bus, VTs; w/3 GCUs, Tap 10.
No comments at this time!

Q14) Have you recorded voltage measurements showing Line and Phase voltages for each configuration?

Q15) Are 9b Tap cases responsive when Tap transition is from: a) Tap 8 or Tap 10 to Tap 9b; b) while on Tap 9b; or c) the transition from tap 9b to Tap 8 or Tap 10?

Q16) Can you send GSU xfmr name-plate data showing primary voltage vs tap?

Q17) Were trips caused by either Stator EF or neutral current, or both, concurrently? Is stator protection REF, that is Restricted-Earth-Fault?

Q18) You mentioned that trips occurred regardless of whose "Black-Box" was used. Were trips always cause by the same "designated" trip function?

Q19) Finally, how long has the problem persisted?

Regards, Phil Corso (Cepsicon [at] aol [dor] com)