A82-Transformer Factory Acceptance Testing-Sequence of carrying out Tests.

When you order a new transformer, the manufacturer will be conducting a series of tests (it may count more than thirty in the case of large special transformers) These tests, as specified in IEC /IEEE standards, can be grouped into the following headings.?

A)???Low voltage tests to check transformer connections are correct and transformer is fit for dielectric tests.

1.????Ratio

2.????Polarity and Phase Relationship

3.????Winding Resistance

4.????Oil testing- BDV, Moisture Content and DGA

5.????Insulation Resistance and Polarity Index

6.????Core Insulation resistance test

7.??Winding and Bushing Capacitance and Power Factor

8. ?Single Phase low voltage excitation current at rated conditions

B)???Loss Measurements & Temperature rise test to assure the guaranteed performance parameters are met.

1.????No-load loss and excitation current at 100 % and 110% of rated voltage-to be repeated after all-dielectric tests.

2.????Load Loss and Impedance Measurement

3.????Operational tests of accessories and instruments-Tap-Changer, pumps, cooling fans, Temperature indicators, oil level gauge, cooler control cubicle, OLTC control cubicle.

4.????Auxiliary loss

5.????Zero Sequence Impedance

6.????Temperature Rise or short time temperature rise

7.????Oil DGA Test

C)???Short circuit withstand capability test (Special Test) to assure transformer can withstand line faults on secondary side distribution lines.

D)???Dielectric Tests to confirm transformer can withstand overvoltages and surges that may come into service.

1. Power-frequency test on auxiliary devices and control and current transformer ?circuit

2. Lightning Impulse Test (full and chopped)

? ?? 3. Lightning Impulse- Front of wave (for IEEE standard)

??? 4. Switching Impulse Test

? 5. Power Frequency tests – Long-term and short-term with PD measurement

? 6. Oil DGA

? 7. No-load loss and excitation current at 100 % and 110% rated voltage

? ? 8. Single Phase low voltage excitation current at rated conditions

9. SFRA (Sweep Frequency Response Analysis) Test

?E) Mechanical Tests

???1. Lifting and pulling device – either by test or by mathematical calculations

???2. Checking electrical air clearances from bushings

???3. Operation of OLTC and Cooler controls

???4. Operation and checking of protective and indicative accessories

???5. Pressure Test with leak check

???6. Weight of Transformer as arranged for transport

These tests are to be done in a particular sequence for optimum results.?IEC 60076 for Power Transformers is not specifically giving a test sequence for the factory tests. But IEEE C57.12.90-2015 - Test Code for Power Transformers, stipulates as below, vide

Clause 4.3, “the resistance, polarity, phase relation, ratio, no-load loss, and excitation current, impedance, and load loss tests (and temperature-rise tests, when applicable) should precede dielectric tests. Using this sequence, the beginning tests involve voltages and currents, which are usually reduced as compared with rated values, thus tending to minimize problematic effects to the transformer.”

Clause 10.1.5.1 specifies the test sequence for dielectric tests as below:

“Impulse tests, when required, shall precede the low-frequency tests. The dielectric tests should be performed in the preferred sequence as follows: a) Lightning impulse tests (on all terminals) b) Switching impulse tests c) Applied voltage test. d) Induced voltage test.

By agreement between the manufacturer and purchaser, switching impulse tests (when required) may be performed before lightning impulse tests. The induced voltage test shall be the last dielectric test performed”

Let me share some guidelines that I used to recommend:

1. It is better to conduct the full temperature rise test or short time temperature rise test before the dielectric tests. Any defect will be aggravated during the temperature rise test and will come out during the dielectric tests. As mentioned above, the latest IEEE standard specifies such a sequence for temperature rise tests.

My company made a 100 MVA 220/132/33 kV auto-transformer in the early 1970s -one of the very first units of such rating to be made in the country. The on-load tap-changer was on the 132 kV line end. The regulating winding was made with three large rectangular conductors in parallel in the radial direction. The regulating winding was in between the series and common windings, an area of high leakage flux density. The transposition of the three conductors was in the middle of the winding. The operator missed the conductor transposition in one phase of the winding. Even in those days, our Japanese mentors (from Hitachi) insisted on conducting temperature rise tests before the high voltage tests, that we meticulously followed. During the temperature rise test, thickly insulated conductors of the regulating winding without transposition got burnt heavily due to the circulating current among parallel conductors but still withstood rated voltage. During the induced power frequency high voltage test, the regulating winding insulation broke down. Had we done the temperature rise test after the high voltage test, the transformer with a defect in winding would have left the factory and failed after some days or months of service.

2. When the transformer has two or three ratings based on the cooling method – ONAN/ONAF-1/ONAF-2 or ONAN/ONAF/OFAF – it is better to start the temperature rise test with the highest rating i.e. ONAF-2 or OFAF rating. When conducting a temperature rise test on large rating transformers esp. with a high current LV side, a shutdown may be taken after sending the full current for half an hour. The oil inside the tank will be still cool but tank parts subjected to leakage flux or circulating current will heat up quickly and attain the final temperature. This way hot spots in the metal can be detected early and easily as the underlying oil will be still very cool. In case of any unacceptable hot spots, the test can be discontinued till suitable countermeasures are taken. Remember temperature rise test is a costly test as it consumes a huge amount of energy and any repeat test shall be avoided as far as possible.

3. Reason for the dielectric test sequence

Even though not mentioned in any standard, it is always better to start with a PD test gradually increasing the induced overvoltage up to 1.3 Ur. This may be done before applying the no-load loss test or dielectric tests. It will avoid any damage to windings from test bed failures as the test can be immediately stopped if high PD is detected during this investigative testing. The transformer can be returned for reprocessing if a high PD is detected at this stage of testing.

Failure detection during impulse or switching impulse testing is more difficult than during power frequency overvoltage testing. By keeping the AC tests after other dielectric tests, it is ensured that there is no chance of the transformer leaving the test bed with undetected faults from the earlier tests. Any incipient fault will be aggravated during the final AC test with PD monitoring.

4. Type tests and short circuit tests.

As per IEC standards, type tests and impulse tests need not be done before the short circuit test. Considering that the short circuit test is so expensive (almost the cost of the transformer) it is better to conduct all-dielectric tests before the short circuit test also and repeat after the short circuit test. This will prevent the transformer from going to the short circuit lab that may not pass the high voltage tests later.

5. Variation in no-load loss

5.1 Final no-load test after dielectric tests

?After the dielectric tests, no-load loss measurement is repeated to make sure that the transformer is not leaving the factory with any hidden defect from the dielectric tests. It also demagnetises the core to measure the single-phase low voltage excitation current, to be used as the benchmark in service life. At this stage, no-load losses may be measured 1-3 % more than the first measurement taken in the beginning. This increase is thought to be due to the reduction in core interlaminar resistance from the effect of the impulse test. Loss increase can be more than the above range in cores with low interlaminar resistance. (Clause 3.2.11 of IEEE standard C57.123-2010 -Guide for Transformer Loss Measurement) But for guarantee purposes, the first measurement should be considered as this loss increase usually diminishes with time and reach its original value after several hours. This phenomenon was first noticed and analysed by transformer engineers in the early1950s. Please see “Effect of Impulse Testing on Transformer Iron Loss”- J H Mc Whirter & others (Westinghouse Electric Corporation, Sharon, Pa Plant), AIEE paper,1957 February, Pages 1275-1278.

5.2 No-load loss after short circuit test.

Small changes in the excitation current and core loss can be expected after a short-circuit test. Hence, the IEEE test code generally allows for a maximum increase of 5% for transformers with stacked cores. However, in distribution transformers with wound cores, an increase of up to 25% in the magnitude of the excitation current could occur due to small distortions of the core, even in the absence of a winding failure (Clause 3.2.13 of C57.12. 123)

5.3 Temperature Correction for No-load Loss.

No-load loss decreases with an increase in core temperature. This is due to the reduced eddy loss component as the core eddy current comes down with the increase in temperature as lamination resistivity goes up with temperature. This reduction is about 0.065% per degree C (1 % for 15 C degree variation). Due to the difficulty in accurately determining core temperature, this correction is not applied for Power Transformers.

6. Importance of the single-phase low voltage excitation test.

Along with DGA, this is the most important parameter for confirming the healthiness of transformers in service. Hence this benchmark excitation current readings must be taken as the last test and should be entered in the test report.?