A transformer doesn’t usually fail from the inside out. It fails from the bushings.
Bushing failures account for a disproportionate share of transformer in-service failures. Industry studies and utility reliability surveys have reported figures of approximately 15% to 40%, varying with transformer population, voltage class, and age profile — and the failures tend to be catastrophic: explosive porcelain rupture, oil fire, tank rupture, secondary damage to adjacent equipment. A transformer can be perfectly healthy internally and still suffer a serious event because one of its bushings failed. The diagnostic question isn’t whether the transformer is in good condition; it’s whether the bushings are.
That makes bushing testing one of the highest-leverage activities in transformer maintenance. The tests are quick, the instruments are the same ones used for transformer power factor work, and the failure mechanisms are reasonably well-understood. Catching a degrading bushing early — before the tan delta has run away or the capacitance has shifted — saves the transformer, the substation, and arguably the people working nearby.
This is a practical guide to bushing testing. C1, C2, hot collar, and how to read what each one tells you. It assumes familiarity with transformer power factor testing — the bushing tests use the same instrument and the same GST/UST/GSTg modes covered there, applied to a different test object. The value here is in what bushings specifically reveal, where they fail, and what each measurement diagnoses.
Table of Contents
What a Bushing Actually Is
A transformer bushing is the insulated path that carries the high-voltage conductor from outside the tank to the winding inside. It looks simple from the outside — a porcelain or composite housing with a terminal on top — but inside it’s a precisely engineered capacitor.
The modern condenser-type bushing has a central conductor surrounded by alternating layers of insulation (paper or resin) and conductive foil. The foil layers form a series of capacitors that grade the electric field uniformly from the high-voltage conductor at the center to the grounded mounting flange at the outside. Without this grading, the field would concentrate near the conductor and cause partial discharge or breakdown.
Two main types dominate modern transformer service:
OIP (Oil-Impregnated Paper) bushings use kraft paper as the dielectric, impregnated with mineral oil. Older design, still common in service. Class A thermal rating (105°C). Lower initial cost. The oil-paper system is moisture-sensitive — water ingress through failed gaskets is the dominant failure mechanism. IEEE C57.19.01 provides a reference tan delta value of approximately 0.5% for OIP bushings at 20°C, commonly used as an assessment threshold.
RIP (Resin-Impregnated Paper) and RIS (Resin-Impregnated Synthetic) bushings use paper or synthetic fiber impregnated with epoxy resin. Newer design, increasingly standard on new equipment. Class E thermal rating (120°C). Higher initial cost but more rugged, less moisture-sensitive, and safer (no oil to ignite). IEEE C57.19.01 provides a reference tan delta value of approximately 0.85% for RIP/RIS bushings at 20°C — higher than OIP because the dielectric properties differ. Some manufacturers specify tighter limits.
A few older transformers still have solid-type bushings — bulk insulation without the condenser layers. These don’t have test taps and can only be assessed via the hot collar method.
The condenser-type bushing’s most important external feature for testing is the test tap (sometimes called the voltage tap, capacitance tap, or potential tap depending on its rated voltage). The tap connects to the outermost foil layer of the condenser stack. In normal operation it’s grounded through a small cap. For testing, the cap is removed and the tap is connected to the test instrument.
The two main capacitances you can measure through the tap:
C1 — the main insulation, between the central conductor and the test tap. This is most of the bushing’s insulation volume and the primary indicator of overall bushing health.
C2 — between the test tap and the mounting flange (ground). A small capacitance, representing the outermost layers of insulation plus the tap compartment.
Why the distinction matters comes down to where bushings fail.
Where Bushings Fail and Why It Matters for Testing
The dominant failure mechanism for OIP bushings is moisture ingress. Water enters through a failed gasket — usually at the top of the bushing, where temperature cycling stresses the seal — and accumulates in the tap compartment. From there, it attacks the bushing’s insulation from the outside in, starting with the outermost foil layers.
This pattern has a direct diagnostic consequence: the C2 measurement, which covers exactly those outermost layers, often catches moisture ingress before the C1 measurement does. A bushing in the early stages of moisture degradation can show normal C1 tan delta but elevated C2 tan delta. By the time C1 is showing significant change, the moisture has progressed deep into the main insulation and the bushing is well past its early-warning window.
This is the single most important practical point about bushing diagnostics that most online content underweights: C2 is the early warning, C1 is the confirmation. A test program that measures only C1 misses bushings in the early stages of moisture-driven failure — the stage where the bushing is still recoverable or, more often, can be replaced before it fails in service.
Other failure modes:
- Capacitive layer shorting — adjacent foil layers short together, reducing the effective number of capacitors and increasing the measured capacitance. Often results from electrical stress, vibration, or aging of the inter-layer insulation.
- Internal partial discharge — voids or poor impregnation cause PD activity that progressively erodes the insulation.
- Thermal aging — long-term overheating accelerates dielectric degradation. Often correlated with overload history or cooling problems.
- Tap connection problems — corrosion or looseness at the test tap connection, sometimes producing capacitance decreases that mimic other faults.
- Surface contamination — salt, dust, moisture films on the bushing porcelain produce surface leakage that elevates power factor without indicating internal problems.
Each of these has a characteristic signature on the test results, which is what makes the test diagnostic rather than just confirmatory.
The Three Bushing Tests: C1, C2, and Hot Collar
A complete bushing assessment uses three tests, each looking at different parts of the bushing’s insulation system.
C1 Test (Main Insulation)
What it measures: the capacitance and tan delta of the main bushing insulation between the central conductor and the test tap.
Setup:
- Test mode: UST (Ungrounded Specimen Test)
- Test voltage applied to: the central conductor (HV terminal of the bushing)
- Test tap connected to: the UST measurement input
- All other connected equipment: grounded
- Test voltage: typically 10 kV at line frequency
- Other bushings on the same voltage level: isolated and shorted; untested bushings grounded
The UST mode isolates the C1 path because the measurement input is connected to the test tap. Any current that flows from the conductor to ground via paths other than C1 (such as through C2 to the flange) bypasses the measurement.
What you read: capacitance in pF and tan delta (or power factor) in %.
Typical values:
- New OIP bushings from reputable manufacturers: 0.2-0.4% tan delta
- New RIP/RIS bushings: typically ≤0.35% tan delta
- Capacitance: matches the bushing nameplate within manufacturer tolerance (usually ±5% or better)
C2 Test (Tap Insulation and Outermost Layers)
What it measures: capacitance and tan delta between the test tap and the flange (ground), covering the tap compartment insulation and the outermost layers of the main core.
Setup:
- Test mode: GSTg (Grounded Specimen Test with Guard)
- Test voltage applied to: the test tap
- Central conductor (and main bushing terminal): connected to guard
- All other connected equipment: grounded
- Test voltage: 500 V to 2000 V (much lower than C1 because the C2 insulation is thinner and isn’t rated for 10 kV)
- Frequency: line frequency
The GSTg mode isolates the C2 path. The HV conductor is guarded out, so any current flowing through C1 doesn’t appear in the measurement. Only current flowing from the tap through the tap compartment insulation to ground is captured.
What you read: same — capacitance and tan delta.
Typical C2 values vary more by bushing design than C1, so comparison against the nameplate or factory record is essential. Manufacturer documentation provides the expected C2 capacitance for each bushing model.
Why C2 matters more than its small magnitude suggests: the C2 test interrogates exactly the part of the bushing that fails first in moisture-driven aging. A rising C2 tan delta on an OIP bushing, even if C1 still looks fine, is one of the strongest early indicators of gasket failure and moisture ingress. The literature reports cases where C2 changed by 100% or more before C1 showed any deviation.
Hot Collar Test
For bushings without test taps (solid-type bushings, some older designs), C1 and C2 measurements aren’t possible directly. The hot collar test is the alternative — and it’s also valuable as a supplementary test on tapped bushings.
Setup:
- A flexible conductive collar is wrapped around the bushing’s porcelain at a specific height
- Test mode: GST (Grounded Specimen Test)
- Test voltage applied to: the collar
- Bushing conductor: grounded
- Frequency: line frequency
- Test voltage: typically 10 kV (or lower for shorter bushings)
The hot collar applies test voltage to a localized section of the bushing exterior and measures the current flow through that section of insulation. Multiple collar positions along the bushing length build up a picture of the insulation condition at different depths.
What it reveals:
- Local deterioration in specific sections of the bushing
- Surface contamination (showing up as elevated power factor across all collar positions)
- Low oil or compound levels in the upper portion of the bushing (the upper collar positions show higher capacitance from the missing dielectric)
- Voids in the compound (irregular readings)
Hot collar is more sensitive to localized problems than C1/C2 because the latter average across the whole bushing. A small fault in one section may not move the C1 number enough to flag, but the hot collar at that height will catch it.
The limitation of hot collar is that it doesn’t have the same standardized acceptance criteria as C1/C2 — interpretation relies more on comparison between positions on the same bushing, comparison between sister bushings, and trending over time.
Acceptance Criteria
The thresholds for bushing testing are tighter than for transformer-level power factor because bushings are more sensitive components with well-characterized failure modes. The criteria below are widely cited reference points, not universal acceptance limits — actual criteria depend on bushing type, manufacturer specification, and the asset owner’s standards.
Tan Delta Limits
IEEE C57.19.01 reference values at 20°C, commonly applied as assessment thresholds:
- OIP bushings: approximately 0.5%
- RIP/RIS bushings: approximately 0.85%
IEC 60137 sets similar values with some variation by bushing class. Some manufacturers specify tighter limits for their products.
These are guidance thresholds, not normal operating values. New bushings should read well below these — typically 0.2-0.4% for OIP and ≤0.35% for RIP/RIS. A bushing operating at 0.45% on commissioning has very little margin for degradation.
The widely-applied doubling/tripling rule (from IEEE C57.19.01 and CIGRE TB 445):
- Tan delta exceeding 2× the nameplate or commissioning value: investigate. May indicate developing problems. Increase monitoring frequency.
- Tan delta exceeding 3× the nameplate value: consider the bushing compromised. Replacement is generally warranted.
These multiplicative criteria are often more sensitive than the absolute thresholds because they catch a bushing that has degraded significantly from its baseline while still being below 0.5%. A new bushing at 0.2% that has drifted to 0.45% has more than doubled its tan delta — under the doubling rule it warrants investigation even though it remains below the absolute threshold.
Capacitance Change Limits
Capacitance changes from the nameplate or commissioning value are interpreted on a tiered basis:
- Capacitance increase greater than approximately 5%: often indicates shorted condenser layers or other structural changes within the capacitive grading system. The field grading may have been disrupted and the bushing’s withstand capability reduced. Investigation warranted.
- Capacitance increase greater than 10%: critical. The bushing should not remain in service until the cause is identified.
- Capacitance decrease greater than 3%: typically indicates a test tap connection problem (corroded, loose) rather than a bushing fault, but warrants confirmation by repeat measurement with cleaned connections.
Capacitance change is often a stronger indicator of mechanical or structural change than tan delta. Tan delta tracks dielectric losses (a chemistry/moisture question); capacitance tracks the bushing’s geometry (a structural question).
Important Qualifications
These thresholds should be applied with the same caution that applies to transformer-level power factor testing:
Never act on a single measurement. An elevated bushing tan delta in a single test should be confirmed by a repeat measurement under controlled conditions, supported by visual inspection, hot collar testing, oil sampling (if the bushing has a separate oil sample point), and review of the transformer’s history. Surface contamination, weather effects, and instrument issues can produce false elevations. A consistent finding across multiple diagnostics supports a real decision.
Acceptance criteria depend on the bushing type and manufacturer. The 0.5% and 0.85% figures are reference points from IEEE C57.19.01. Specific bushings may have tighter manufacturer specifications. Older OIP designs may have somewhat different acceptable ranges. Verify against the bushing manufacturer’s documentation before applying any threshold as a hard pass/fail.
Trend matters more than absolute value. A bushing trending from 0.25% to 0.42% over five years is a developing problem regardless of whether 0.42% is below the absolute threshold. The rate of change is often the most useful diagnostic.
Environmental and Setup Factors That Distort Results
Bushing testing is more sensitive to setup conditions than transformer-level testing because the measured quantities are smaller and the bushings are more exposed.
Weather. Don’t test in rain, fog, or relative humidity above 75% per IEEE C57.152. Surface moisture on porcelain produces leakage paths that elevate the measured tan delta substantially. A bushing tested in damp conditions can read 1% or more when its actual internal condition is well within limits. If marginal weather can’t be avoided, recognize that the results may be contaminated and plan to repeat under better conditions.
Surface cleanliness. Clean the porcelain and the test tap connection area before testing. Years of accumulated salt and dust create surface leakage that mimics internal degradation. Manufacturer test reports are usually done on clean bushings; field comparison requires comparable conditions.
Temperature. Bushing tan delta is strongly temperature-dependent. OIP bushing dielectric losses increase significantly with temperature, but the exact relationship depends on design and condition — rules of thumb like “doubles every 20°C” are rough approximations only. This is why the older generic correction tables are no longer recommended by IEEE C57.12.90-2010, and why modern DFR-based individual temperature correction methods are preferred over generic correction factors. Test at consistent temperatures when trending. For absolute comparison against thresholds, DFR-based temperature correction is more reliable than table-based correction.
Test tap connection integrity. A poor connection at the test tap produces erratic and elevated readings. Clean the tap surface, use a proper test tap adapter, and verify the connection before energizing.
Adjacent grounded equipment. Stray capacitance to surrounding grounded structures can affect the measurement. Modern test sets handle this through guard circuits and noise rejection, but the setup should still minimize close proximity of test leads to grounded surfaces.
Variable frequency capability. As with transformer-level PF testing, modern instruments can run narrowband DFR sweeps across multiple frequencies (typically 1-500 Hz) for individual temperature correction and earlier defect detection. For bushings particularly, the DFR sweep can distinguish between moisture in the insulation and surface contamination — two different problems that produce similar single-frequency readings.
Diagnostic Patterns: Reading What the Tests Reveal
The diagnostic value of bushing testing comes from combining the C1, C2, and (where applicable) hot collar results. The patterns:
C1 normal, C2 elevated tan delta. Early moisture ingress, classic gasket failure pattern. Catch this before C1 starts to move. Action: investigate seal integrity, plan for bushing replacement or refurbishment, increase monitoring frequency.
C1 and C2 both elevated. Moisture or contamination that has progressed deeper into the main insulation. Later stage of the moisture failure mode. Action: replacement generally warranted.
C1 tan delta elevated, C2 normal. Internal degradation not driven by external moisture ingress. Could be thermal aging, internal PD, or contamination. Action: hot collar test to localize, oil sampling if separately oiled.
Capacitance increase >5%, tan delta normal or moderately elevated. Shorted condenser layers. Mechanical/structural problem rather than chemistry. Action: replacement.
Capacitance decrease, tan delta erratic. Test tap connection problem. Action: clean and retest. Don’t assume a fault until tap connection integrity is confirmed.
Hot collar reading elevated at one specific position. Localized internal problem at that depth. Could be void formation, local contamination, or oil-level issue. Action: combine with C1/C2 results and consider replacement.
Hot collar uniformly elevated across all positions, C1/C2 normal. Surface contamination. Action: clean and retest.
All readings within limits but tan delta trending upward over multiple tests. Slow degradation, often the most common pattern on aging bushings. Action: continue monitoring, plan for replacement before the trend hits the threshold.
Single elevated reading inconsistent with history. Confirm before acting. Could be weather, contamination, instrument issue, or genuine degradation. Repeat under controlled conditions.
How Bushing Testing Fits With Other Diagnostics
Bushing testing is part of a broader transformer condition assessment. The interactions matter:
With transformer power factor testing. Bushings are part of the insulation system measured during transformer-level PF tests. A transformer with elevated overall PF can be caused by bushing problems rather than internal issues — running bushing-specific tests (C1, C2, hot collar) localizes the problem. Conversely, a transformer with normal bushings and elevated PF points the diagnosis toward internal insulation.
With DGA. Bushings can produce combustible gases if internally arcing or overheating. On transformers where the bushings communicate with the main tank oil (common in older designs), bushing problems show up in main tank DGA. Elevated hydrogen, methane, or in some cases acetylene without a clear winding-related source may warrant investigation of the bushings. Some utilities perform separate DGA on bushing oil samples when accessible.
With infrared thermography. Bushings with internal degradation often run hotter than healthy ones. Periodic IR scans, especially under load, can identify a problematic bushing before electrical testing is scheduled.
With visual inspection. Look at the bushing. Cracked porcelain, oil leaks at gaskets, discoloration of the upper terminal area, or external contamination tell you things no electrical test can. The cheapest diagnostic is often the most useful.
The combination is more powerful than any single test. A bushing with normal C1, slightly elevated C2, slightly elevated DGA hydrogen, and a visible gasket oil weep is telling a consistent story — early moisture ingress — and warrants action even though no single measurement crosses a hard threshold.
The Takeaway
Bushing testing is one of the highest-value activities in transformer condition assessment because bushings fail more often than the transformers they’re attached to, and their failures are disproportionately consequential. The tests are quick, the instruments are the same ones used for transformer-level PF work, and the failure modes are well-characterized.
The discipline that makes the test diagnostic rather than just confirmatory: measure C2 routinely, not just C1, because C2 catches moisture ingress earlier; apply the doubling/tripling rules against commissioning baselines, not just the absolute thresholds; trend rather than checking single measurements; use hot collar to localize problems that C1 and C2 average out; control for weather and surface contamination before believing an elevated reading; and combine the electrical results with DGA, infrared, and visual inspection for a complete picture.
Done well, bushing testing catches the failures that would otherwise become the transformer’s most likely catastrophic event. Done as a once-every-few-years C1 measurement with no follow-up on borderline results, it provides false confidence. The instrument doesn’t change between those two approaches — the discipline does.