Dry-Type Transformer Winding Resistance Imbalance
During acceptance testing of a 500 kVA dry-type transformer, 480 V delta primary and 208Y/120 V secondary, a technician measures primary winding resistance with a 10 A four-wire micro-ohmmeter. Nameplate primary current is 601 A. At 28°C, after 60 seconds of stabilization, the readings are A-B 0.00910 Ω, B-C 0.00930 Ω, and C-A 0.00960 Ω. The test leads were zeroed before the test. A second set after 5 minutes gives A-B 0.00912 Ω, B-C 0.00933 Ω, and C-A 0.00961 Ω. The transformer is on tap 1, nominal. The manufacturer's factory values at 75°C are A-B 0.01070 Ω, B-C 0.01075 Ω, and C-A 0.01072 Ω. The technician wants to accept the transformer because every reading is within 10% of the factory values.
1. Calculate the phase imbalance and temperature-corrected winding resistances. Does the transformer pass?
Read the worked answer
Average the three readings: (0.00910 + 0.00930 + 0.00960) / 3 = 0.009333 Ω. The C-A reading deviates from the average by (0.00960 - 0.009333) / 0.009333 = 2.86%. The maximum-to-minimum spread is (0.00960 - 0.00910) / 0.00910 = 5.49%. For transformer winding resistance, phase-to-phase comparison on the same tap is typically expected within about 2%; this transformer fails that screening criterion.
Correct to 75°C using the copper factor (234.5 + 75) / (234.5 + 28) = 309.5 / 262.5 = 1.179. The corrected values are approximately A-B 0.01073 Ω, B-C 0.01096 Ω, and C-A 0.01132 Ω. Comparing C-A with the factory value of 0.01072 Ω gives an error of about 5.6%. The fact that all readings are within 10% of factory values is not sufficient; the phase balance is the more sensitive indicator. The transformer should not be accepted until the C-A reading is explained.
2. What are plausible causes of the C-A reading, and what follow-up tests should be performed?
Read the worked answer
A high resistance on one phase can come from a poor test connection, lead placement, contact resistance at the tap selector, a loose lug or busbar connection, or a damaged winding connection. A shorted turn would usually lower resistance, not raise it, so the C-A high reading is more consistent with a connection or contact problem. The stable repeat readings make a moving test lead less likely, but contact quality still needs to be checked.
Repeat the measurement with the current and potential leads reversed and with the micro-ohmmeter connected directly at the winding terminals, not through temporary jumpers. Inspect and torque the tap-changer connections and busbars. Measure all taps to see whether the imbalance follows the tap or remains on the C-A phase. Perform turns-ratio, excitation-current, and insulation-resistance tests to screen for winding damage. If the imbalance persists after connections are verified, involve the manufacturer and consider a more detailed winding assessment before acceptance.
3. Evaluate the test current and measurement setup. What safety and measurement precautions apply?
Read the worked answer
The 10 A test current is only about 1.7% of the 601 A rated primary current, which is appropriate for a low-resistance measurement because it limits heating and avoids changing the resistance during the test. The four-wire method is correct because it removes lead resistance from the measurement, and the leads were zeroed before testing.
Before testing, the transformer must be de-energized, locked out, and verified absent of voltage. Because the primary is delta-connected and the secondary may have stored energy or induced voltage, discharge and ground all terminals before connecting the meter. Use a meter with an appropriate CAT rating for the available fault current, keep the test current leads securely connected, and do not open a current loop while current is flowing. Allow the reading to stabilize, record the winding temperature, and repeat each phase consistently. After testing, discharge the winding and remove the test leads before restoring the transformer.