2 MVA transformer Excitation Current Issue 2

[rcw retired EE]

One of two new, identical, 2 MVA, 4.16kV-480/277V delta-wye, KNAN transformers has unusual excitation current tests results prior to energization. Both transformers passed factory tests and most of the usual NETA field tests: insulation power factor (Doble), insulation resistance, TTR, winding resistance, oil test etc.

Following standard Doble procedures, the single phase excitation current was measured. One transformer had no issues but the test set could not build current on unit #2 when trying to energize H1-H3 or H3-H2. Test set would trip. H1-H2 was OK.

A transformer degaussing test set was used to demagnetize the core with limited success. The Doble excitation test set was only able to build up to 100V before tripping. Prior to demagnetization, it could not get even get to 10V.

Our testing company thinks the transformer is going to blow when energized. The manufacturer, a very reputable supplier, says to energize the transformer and that the excitation current test is of no concern. They point out that individual excitation current tests were not specified and are not part of the typical IEEE standard tests. The factory engineers say they have seen this before where two sister units have different excitation currents but were placed in service with no problems. They do not have an explanation as to why there is a difference.

There is no indications of physical impact or damage that might have moved the core or caused magnetic anomalies.

On the good unit, typical excitation current readings at 1.25kV:
H1-H2: 0.49 A, 3.6 kW. H2-H3: 1.05A, 7.97kW. H3-H1: 1.1 A, 7.9kW

Suspect unit tests at 1.25 kV and 0.10 kV:
@1.25 kV: H1-H2: 0.47 A, 3.5 kW. H2-H3: tripped. H3-H1: tripped
@0.100kV: H1-H2: 1.13 A, 4.37 kW. H2-H3: 5.96A, 52.7 kW. H3-H1: 2.56A, 8.61 kW

Factory tests results show 3-phase no load loss at 2kW and excitation current at 0.2% = 4.8A @ 4.16 kV (about 1.45 A @ 1.25 kV)

I tend to go with the manufacturer, they are guaranteeing it. But it would be nice to see an explanation before closing the breaker.

Have you seen any anomalies like this on new units? Could single phase versus three phase tests be skewing the results?

 

[prc]

There can be slight variation in low voltage excitation current values between units, but not the order of 100-400 % or worst tripping out! Definitely the transformer is faulty -from inter turn or some shorting somewhere like tap changer connections. Please don't energise it. Fault must be in H2-H3 winding. Conduct test from LV side also and compare. You will be able to locate the faulty winding easily. Please see clause 7.2.11.2.1 of IEEE Std C57.152-2013 Diagnostic Field tests for more details.
This is the most reliable and consistent test available for transformer engineers to check the healthiness of transformers since Doble standardized it during 1960's.

 

[edison123]

Can you energize HV side with 480 V and do the measurements? Such low voltage most likely won't blow up the trafo but will show imbalance.

Muthu

http://www.edison.co.in

 

[ScottyUK]

After following prc's advice first -

Do you have access to a 3-phase Variac? It doesn't need to be a big one, 10A or so per phase is sufficient. Feed the HV winding with the balanced 3-phase supply from the variac and increase gently from zero. Connect an ammeter in at least two phases, and for preference in all three phases in case there's a path to earth. Observe what happens as you increase voltage.

 

[waross]

I wonder if the single phase test on one winding is leaving the other cores magnetized.
I realise that it may not be possible to do further testing, but have you considered degaussing the core and then doing your single phase test starting on a different winding.
That is, see if the issue is related to the order in which the phases are tested individually. Sort of like rolling the connections to isolate a problem.

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

 

[TestMonkey]

Something is not right. Usually when it is me testing and I get results like that, I probably did something wrong. Connections or wrong programming of test equipment.

Tested following standard Doble procedures? Forgive me for assuming the test set used was Doble M4000.

Post indicates tested at 1.25kV. Maybe it was calcultated to 1.25kV?
Doble says: "Excitation-current measurements should be performed at the highest test
voltage possible within the range of the set, but should never exceed the
voltage rating of the winding across which the test voltage is applied."
Doble emphasizes that statement in bold.

Is there any kind of tap changer? DETC or LTC?
According to Doble there should be tests for NEW equipment as follows:
"Load Tap Changers
New - 1L, N, 1R, 2R, … , 15R, 16R (18 tests/phase)
De-Energized Tap Changers
New - A, B, C, D, E (5 tests/phase)"

Perhaps you omitted the information to shorten the post or I am out of date with new Doble procedures.

My first guess would be that the phase NOT under test is not properly grounded or the neutral (X0) on the LV side is not properly grounded when performing the tests. Or it could be an internal or cable problem with the M4000.

If everyone is confident it was tested correctly and the manufacturer says its ok to energize, be sure to have some kind of written ok from the manufacturer.

Not sure that it is possible or would even help but a SFRA test is a good way to "look" inside a transformer. And having a sister unit offers an excellent comparison.

Best of luck.

 

[TestMonkey]

Forgot to mention, typically during the Doble excitation test the measurements will be in mA and Watts. But every situation is different.
The factory results are not to be considered for comparison, there is a published whitepaper dealing with that.
Look for a copy of Doble Test Procedures, 72A-2244 Rev A. The new version is on their website is password protected.
And, Transformer Excitation Current Test Procedures - Doble Engineering Diagnostics Toolbox

 

[rcw retired EE]

Thank you for all the suggestions. We are pulling a 480V, 3-phase temp power source to the transformer to try energizing it 3 phase at reduced voltage. The manufacturer has provided a written guarantee that the unit is suitable for energization.

Our testing contractor is very experienced with Doble testing and did all tests per the procedures: all DETC taps, etc. I didn't post the full test results, just a summary and converted the mA & W to A & kW. The results were sent to Doble for analysis. Their engineer suggested further testing, SFRA & Leakage Reactance. SFRA tests would not tell us much unless we could also dismantle the bus duct and HV cabling on the sister unit to get a comparison test. I'm investigating what test equipment is needed for a leakage reactance test.

Our management is evaluating the cost and time delays of further testing against the risks of energizing a unit that the supplier guarantees and says they can replace in three weeks. (Removal and replacement time not included). If it does fail, the other unit can supply the project's power needs for commissioning. Cost of testing is approaching the cost of the transformer when air freight of equipment, travel time, support labor etc is included. Also, if the testing contractor is retesting the transformer, his engineers are not testing the other equipment needed for energization. I'll know more when I get to site tomorrow.

Thank again for the background and advice. Safety has to be priority #1.

 

[rcw retired EE]

Energized the transformer at 480V, three phase. (11.5% of nominal voltage): 0.2 A, 5.2A & 5.1A
No other indication of problems. Transformer was isolated, but X0 was grounded. Test was only 10-15 minutes. We were unable to get access to the LV terminals to try LV excitation tests.

We repeated the test on the identical transformer which only drew 0.2A on all three phases.
Details:
H1: 278.0V @ 0°, 0.20 Amps @ -60°, 0.03 kA, 0.13 KVA
H2: 279.V @ -240°, 5.2 Amps @ -275°, 0.63 kW, 0.71 kVA
H3: 277.8 V @ -120°, 5.1 A @-90°, 0.66 KW, 0.72 kVA
Total Input: 1.32 kW, 1.86 kVA.
LV Terminals: 55.4 V phase-phase, 33 V phase-neutral.

That 1.32 kW loss translates to about 11.2 kW at full voltage (4.16 kV) if everything is linear. No load loss would be around 60% of the anticipated full load losses. I expect that much heating would eventually do some damage.

We will see what the supplier wants to do.

 

[TestMonkey]

Thanks for the update.

Hopefully you or the owner can reject the unit. Or the supplier does the right thing and replaces it without question.

 

[edison123]

At that low voltage on HV side, I expect the no-load current to be less than 1 A.

Reversed polarity of one or two phases?

A baker surge comparison test might be informative.

Muthu

http://www.edison.co.in

 

[rcw retired EE]

Supplier asked us to pull the transformer after seeing the 480V test data. They will take it to a local repair facility for tests prior to untanking. If they find nothing wrong, we pay for shipping and testing costs.

Looking at the data from the LV test, I am thinking there is a turn-turn short or other issue in the H2-H3 or X3 winding. Currents into H2 & H3 are 180 apart and essentially in phase with the H2-H3 voltage. Normal excitation current would lag voltage by 90 like the current into H1 lags the H1-H3 voltage by 90 degrees. (Test voltage connections ended up phase rotation H1, H3, H2).

TTR tests look OK except one winding's ratio is high by about 0.004 or 0.02% compared to others but still within stadard limits. Not enough data to support the turn-turn fault theory.
(Ratio = 14.253, read 14.301, 14.301, 14.306. Ratio = 15.001, read 15.006, 15.006, 15.008. Ratio = 15.390, read 15.400, 15.401, 15.404.)

The construction power voltage was varying during the test, making it difficult to get high accuracy voltage readings on the LV windings. With the repair shops steady and controllable test voltage we should be able to do more troubleshooting.

Meanwhile we are energizing the facility on the other unit for now.

 

[unclebob]

Ahhh TTR's...

I hope you got the cranky one, because electronics ones with no knobs to adjust the ratio tend to give different readings in two consecutive tests.

 

[rcw retired EE]

Update. Manufacturer finally got transformer into their local service shop. Our site engineer's comments after witnessing the test:

"1. By time I got there they had done ratio test on Tap C . (H1H3 – X0X1 15001, H1H2-X0X2 15003, H2H3-X0X3, 14.993)
2. We removed the tops and inspected the inside, nothing stood out, oil is very Green.
3. We energized using an Aux Wye-Delta utility transformer to hi side. (Used another transformer to step up the voltage applied to H1, H2, H3)
Step 1 : they applied 35 Volts to low side of aux transformer, measured input current : 0.6 A , 1 A , 2 A .
Step 2 : they applied 60 volts to low side of aux transformer, measured input currents : 15 A, 17 A , 30 A. Phase C continued climbing and smoke started to come out .
4. They drained the oil half way below connections and energized it again to spot where smoke is coming from, it appeared that lead connection to tap changer on Phase C was the source of smoke."

As many of you pointed out, there was an issue in the tap changer connection. (DETC). Should be back in service next week.

Thanks for the input!

 

[waross]

After this position, I hope you got a thank you for your perseverance.

Our testing company thinks the transformer is going to blow when energized. The manufacturer, a very reputable supplier, says to energize the transformer and that the excitation current test is of no concern. They point out that individual excitation current tests were not specified and are not part of the typical IEEE standard tests. The factory engineers say they have seen this before where two sister units have different excitation currents but were placed in service with no problems. They do not have an explanation as to why there is a difference.

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

 

[prc]

wilson,you must insist from manufacturer for a written report for the real mode and cause of failure. It is not the connection that is causing the problem but a closed loop is occurring somewhere.

 

[rcw retired EE]

Thanks for the comment PRC. We are hounding the supplier for an explanation of what type of shorted tap or shorted winding was found when they dug into it. Just a loose connection would not draw much power. There had to be a shorted turn somewhere.

After the on site test at 10% voltage (480V 3 phase on 4.16kV winding) showed a relatively large in-phase current on the H2-H3 winding, we were pretty certain we had a turn-turn or tap-tap fault, most likely at the DETC tap changer or the cable termination block.

I'll post what we find.

 

[rcw retired EE]

Service shop that is repairing the transformer just reported "the connection to the tap changer was bad and destroyed the windings." A full rewind is in progress.

No details were provided on how the tap changer connection was "bad" or the actual failure mode. If it wasn't damaged during transit or during installation, how did the unit pass the factory tests?

More information to follow.

 

[rcw retired EE]

UPDATE- On C phase winding, a tap lead shorted through the turn insulation to the HV turn wrapped on top. See attached report from repair facility with pictures of failure.

On site tests after repairs show good excitation currents but a >10% variation in winding resistance measurements and marginal insulation power factor. Repair shop test results show the resistance of individual HV coils were very similar before they were wired in delta. Another connection issue?

We are concerned that the same FR3 oil was apparently used without processing. It now has a few ppm of acetylene and high moisture content. Insulation power factor is high, probably due to the contamination and water in the oil. (I hadn't realized that FR3 oil pulls moisture out of the cellulose windings unlike "normal" transformer oil that puts the moisture in the windings).

Our client is concerned about having a "new" transformer with marginal quality oil and suspect winding resistance readings. We are still fighting the vendor, who claims since equal resistance, or minimum quality of oil, or power factor tests were not specified, the unit meets our specification. I'll let our lawyers work that out.

I appreciate the advice we got on this site. Had we not followed the advice and energized it, that failure would have been a lot larger. (But maybe I would have gotten a new transformer?)

I don't believe I want to buy from this supplier again, at least not this facility.

 

[itsmoked]

A not so reputable vendor after all. They appear to have now earned and now deserve you telling us who they are. They are, after all, screwing your company in a now social media world.

Keith Cress
kcress -

http://www.flaminsystems.com