Scenario Background: The Sudden Gas Alarm

A 40 MVA, 138/13.8 kV, YNd1 power transformer, in service for approximately 15 years at a utility substation, triggered a sudden Buchholz gas alarm and a sudden pressure relay alarm during a period of high load. The Dissolved Gas Analysis (DGA) from an immediate oil sample revealed a significant spike in combustible gases, particularly hydrogen (H₂), ethylene (C₂H₄), and traces of acetylene (C₂H₂). This gas profile indicated a high-energy electrical fault (arcing) combined with thermal overheating. The transformer was taken offline immediately. The site crew was dispatched to perform emergency electrical diagnostics to locate and characterize the fault before further decisions could be made. The Transformer Turns Ratio (TTR) meter was the first and most critical tool deployed.

Initial Suspicions: Given the DGA results and the sudden onset, the primary suspects were a flashover between windings or components, or a severe case of shorted turns.

Field TTR Testing Procedure and Initial Findings

Following strict safety isolation and grounding procedures, the technicians connected a modern three-phase digital TTR meter to the HV (138 kV delta) and LV (13.8 kV wye) bushings. The transformer had an off-circuit tap changer on the HV winding.

Test Sequence Executed:

1. Performed an automated three-phase TTR test at the nominal tap position.

2. Manually tested all HV tap positions for Phase A only, as an initial survey.

3. Recorded ratio, excitation current, and phase angle for all measurements.

Immediate Red Flags: The automated three-phase test at nominal tap revealed a stark anomaly:

• Phase A (A-a): Ratio = 9.92 | Excitation Current = 85 mA

Phase B (B-b):

• Ratio = 10.00 (Nameplate: 10.00) | Excitation Current = 12 mA

Phase C (C-c):

• Ratio = 10.01 | Excitation Current = 11 mA

The data showed a 0.8% negative ratio deviation on Phase A (9.92 vs. 10.00) and, more strikingly, an excitation current approximately 7 times higher than the healthy phases. Phase B and C were within normal limits (±0.1% ratio, low excitation current).

Deep Dive Analysis and Fault Isolation

The combination of a lower-than-expected ratio and a drastically elevated excitation current on a single phase is a classic textbook signature of shorted turns in the HV winding. The shorted turns act as a single-turn secondary, drawing excessive magnetizing current and reducing the effective number of turns in the primary (HV) winding.

Tap-Changer Correlation Test: To rule out a tap changer issue and isolate the fault to the winding itself, the team performed TTR tests on Phase A across all HV tap positions. The results were telling:

• The 0.8% ratio error and high excitation current were present on every tap position.

• The error magnitude was constant, not varying with tap.

Winding Resistance Confirmation: To solidify the diagnosis, a DC winding resistance test was performed. The results showed:

• HV Winding Resistance (Phase A to Phase B): 0.105 Ω (Previous baseline: 0.115 Ω)

HV Winding Resistance (Phase B to Phase C):

• 0.116 Ω

HV Winding Resistance (Phase C to Phase A):

• 0.106 Ω

Diagnostic Conclusion and Correlation with DGA

The TTR and winding resistance findings provided a consistent and unambiguous diagnosis:

Fault: A significant short circuit between turns within the High-Voltage winding of Phase A.

Mechanism: The high-energy arcing within the short generated the acetylene and hydrogen detected by DGA. The continuous circulating current in the short caused localized severe overheating, generating the ethylene and other thermal gases. The fault was electrical and thermal in nature, as reflected in both the electrical tests and the fluid analysis.

Severity: The magnitude of the ratio shift and current increase suggested the short involved multiple turns, representing a serious fault that would rapidly worsen under load.

Outcome and Action Taken

Based on this conclusive diagnostic evidence, the utility's engineering team made the following decisions:

1. Immediate Condemnation: The transformer was declared unfit for service. Returning it to operation risked a complete winding failure, potential tank rupture, and fire.

2. Repair vs. Replace Analysis: Given the transformer's age and the extent of the HV winding damage (requiring a complete rewind), a cost-benefit analysis favored replacement over repair.

3. Contingency Activation: A mobile substation was deployed to restore power to the affected load while a new transformer was procured.

4. Forensic Investigation: The failed unit was shipped to a repair facility for teardown. The internal inspection confirmed a localized, carbonized short between several turns in the middle of the Phase A HV winding, validating the TTR diagnosis precisely.

Key Takeaway: This case study demonstrates the unparalleled effectiveness of the TTR test as a first-line diagnostic. Within hours of the outage, the fault was not only detected but accurately isolated to a specific winding and phase. The quantitative data from the TTR meter provided the irrefutable evidence needed to make a swift, multi-million dollar capital decision with confidence, preventing a more catastrophic failure and ensuring grid reliability.