FRA and Winding Resistance: Two Essential Tests for Electrical and Mechanical Integrity
DC winding resistance (DCR) measurement has been a cornerstone of transformer testing for over a century. The Transformer Frequency Response Analyzer is a relative newcomer. Both methods assess aspects of winding condition, but they detect different failure modes and respond to different physical phenomena. Understanding their complementary roles allows engineers to build a complete diagnostic picture without redundant testing.
Physical Principles Compared
DCR testing injects a direct current (typically 1–50 A) through the winding and measures the steady-state voltage drop, calculating resistance per Ohm's Law. This reveals:
Turn-to-turn shorts (reduces total resistance proportionally to the number of shorted turns)
Open circuits (infinite resistance)
Poor connections or contact resistance (tap changer, terminal lugs)
Conductor degradation (corrosion, pitting)
FRA, by contrast, injects a low-voltage AC signal sweeping from 10 Hz to 25 MHz. It responds to:
Geometric changes (winding displacement, spacer shift) that affect distributed inductance and capacitance
Core movement (affects low-frequency inductive coupling)
Lead and bushing issues (affects high-frequency response)
Shorted turns (produces broadband damping)
Sensitivity to Different Failure Modes
A failure mode matrix illustrates the complementary nature:
| Failure Mode | DCR Sensitivity | FRA Sensitivity |
|---|---|---|
| Single shorted turn (1 turn in 1000) | Very low (0.1% change, often below test repeatability) | High (broadband damping, clear spectral change) |
| Multiple shorted turns (10+ turns) | Moderate (1-2% change, detectable with precision instrument) | Very high (dramatic amplitude reduction, new notches) |
| Open winding (broken conductor) | Very high (infinite or extremely high resistance) | Moderate (response becomes capacitive-only, non-physical) |
| Axial winding displacement (spacers shifted) | None (resistance unchanged) | High (mid-band resonant shifts) |
| Radial buckling (2-3 mm) | None | High (peak amplitude reduction, frequency shift) |
| Tap changer contact erosion | High (increased resistance, phase imbalance) | Moderate (high-frequency anomalies) |
| Core movement or loose clamping | None | High (low-frequency band changes) |
Detection Timing: Incipient vs. Advanced Failure
FRA detects mechanical damage earlier than DCR:
A transformer that experiences a through-fault may have winding displacement of 5 mm with no shorted turns. FRA shows clear deviation; DCR remains normal. The condition is repairable by re-clamping.
If the same transformer is returned to service without repair, friction and vibration may eventually cause the displaced winding to short to an adjacent turn. At that point, DCR will show a measurable change—but now the damage may require rewinding rather than simple clamping.
Thus, FRA provides earlier warning, enabling less expensive, less invasive repairs.
Case Example: Normal DCR, Abnormal FRA – Correct Decision
A 30 MVA transformer was struck by lightning on the HV bushing. Post-event DCR showed all phases balanced within 0.3% of factory values—well within acceptable limits. However, FRA testing revealed on Phase B:
Mid-frequency CC of 0.73 compared to baseline
4 dB amplitude drop in the 20–60 kHz band
Based on FRA evidence, the utility performed an internal inspection. Phase B showed radial buckling over 15% of the winding circumference. The damage was repaired for $150,000. If the transformer had been returned to service, the buckled turns would have eventually shorted, likely causing a catastrophic failure costing over $1.5 million. The normal DCR result would have been dangerously misleading without FRA.
Case Example: Abnormal DCR, Normal FRA – Tap Changer Issue
Annual DCR testing on a 50 MVA transformer showed Phase C resistance 4% higher than Phases A and B. FRA testing showed all phases within normal correlation (CC > 0.96 across all bands). The normal FRA ruled out winding damage. Inspection of the de-energized tap changer found a pitted contact on Phase C. The tap changer was serviced, and DCR returned to normal. Without FRA, the utility might have suspected winding damage and performed an unnecessary internal inspection.
Recommended Combined Testing Protocol
For comprehensive assessment, use both tests in sequence:
Perform DCR annually as a quick screening tool. If DCR is normal and stable, proceed to routine FRA every 3–5 years.
If DCR shows imbalance >1% or change from baseline >1%, perform FRA immediately to determine if the cause is winding-related (FRA abnormal) or tap-changer-related (FRA normal).
After any fault event (through-fault, lightning, shipping), perform FRA regardless of DCR results. FRA detects mechanical damage that DCR misses.
If FRA shows deviation but DCR is normal, investigate for winding displacement without shorted turns. Schedule repair before shorting occurs.
If both FRA and DCR are abnormal, the transformer has likely experienced both mechanical deformation and electrical damage—highest priority for immediate outage.
Economic Optimization of Dual Testing
Some utilities consider eliminating DCR in favor of FRA, or vice versa. Data from 500+ transformer failure analyses shows:
Relying only on DCR would miss 35% of mechanical faults (axial displacement, radial buckling without shorted turns).
Relying only on FRA would miss 10% of electrical faults (tap changer contact erosion, open circuits, gross conductor corrosion).
The optimal strategy is to perform both tests, but at different frequencies: DCR annually (low cost, fast), FRA every 3–5 years or after events (higher cost, longer test time). This balances diagnostic coverage with operational budget.
The Transformer Frequency Response Analyzer does not replace winding resistance measurement—it complements it. Together, they provide a complete picture of transformer electrical and mechanical integrity, catching faults that either test alone would miss.
