Parallel Transformer Troubleshooting: Unequal Current, Temperature, Circulating Current, Neutral Current & Trips

Step 1 — prove the difference is real

Compare meters/CT ratios, measurement timing, ambient temperature, load condition and sensor accuracy. Data from different times or operating points can create a false comparison.

If the hotter unit is also carrying more load, the temperature difference may first be a load-sharing result rather than a cooling failure.

Step 2 — check taps and ratio for circulating current

Compare actual taps, primary/secondary voltage and any history of tap changes. Significant current difference at very light load, or a difference that changes with tap adjustment, points back toward source-voltage mismatch and circulating current.

Step 3 — use kVA and %Z to explain unequal loading

Calculate expected sharing using rated kVA and measured/guaranteed %Z, then compare with field current/kVA. If field behavior follows the predicted trend, the difference may be parameter-driven. If it is much larger than predicted, continue to voltage, phase, connection, measurement and equipment-condition checks.

Step 4 — review phase unbalance, neutral and zero sequence

Compare phase currents and neutral current. A changed zero-sequence path after paralleling can alter neutral and ground-fault behavior. Use the one-line, grounding points, CT locations and load unbalance to interpret it.

Step 5 — put protection events on the timeline

If the event occurs at energization or transfer, review magnetizing inrush, CT transient behavior and switching/protection logic. If it occurs under stable load, focus on actual faults, unbalance, overload or grounding conditions. Relay event reports and oscillography are more useful than a verbal “it tripped once.”

Step 6 — then inspect the transformer and cooling system

After system causes are screened, inspect fans/pumps/radiators, oil level/flow, terminal heating, bushings, winding or core condition as appropriate. This sequence does not ignore equipment faults; it prevents a normal load-sharing difference from being misdiagnosed as a cooling problem.

Build a parallel-operation baseline

Keep both nameplates, tap positions, %Z, ratio/connection test data, phase currents, kW/kVA, bus voltage, neutral current, oil/winding/ambient temperatures, relay events and cooling state from the initial operating period. Future “one unit is always hotter” questions become data comparisons instead of guesses.

Procurement / engineering execution checklist

Frequently Asked Questions

Does a hotter transformer automatically have a cooling problem?

No. Compare actual load, impedance-driven sharing, circulating current, ambient and measurement before blaming cooling.

Does unequal current always mean different %Z?

No. Ratio/tap mismatch, unbalanced load, phase/connection issues, measurement and equipment condition can also contribute.

Which historical data is most useful?

FAT/delivery baseline, %Z, ratio/connection, taps, first-parallel currents/temperatures/neutral current and relay events.

This article is an engineering and procurement guide. Final paralleling, protection, grounding and switching decisions must follow the project one-line, applicable standards, utility requirements and authorized engineering/commissioning procedures.