IEC 60076-3:2013 Explained: A Clause-by-Clause Guide to Power Transformer Dielectric Testing

When a power transformer arrives at site, the test report is mostly numbers. LI, LIC, SI, AV, IVW, IVPD, LTAC. Test voltages in kV. PD readings in pC. Clearances in mm. If you didn’t write the spec, you’re reading someone else’s interpretation of a standard you might not have in front of you.

That standard, for most of the world, is IEC 60076-3. Edition 3 was published in July 2013 and is the current version. It replaced the 2000 edition and introduced enough changes that a spec written under the old rules doesn’t quite work under the new ones. The induced voltage test is now based on Ur instead of Um. The AC withstand test was renamed (LTAC, not ACSD). Switching impulse levels are now defined for all Um above 72.5 kV. The k-factor for impulse waveshape correction is new.

This article walks the standard clause by clause. Not the test procedures themselves — those are in IEC 60076-4 and IEC 60060-1. This is the reference for what IEC 60076-3 actually requires, what it leaves to the purchaser, and what trips people up when they read it for the first time.

What This Standard Covers and What It Doesn’t

Clause 1 sets the boundary. IEC 60076-3:2013 applies to power transformers as defined in IEC 60076-1. It specifies:

The applicable dielectric tests and minimum test levels
Recommended minimum external clearances in air

The clearances are recommendations that apply only when the purchaser hasn’t specified their own. The test levels are minimums — the purchaser can specify higher.

What’s out of scope:

External insulation of the transformer itself for liquid-immersed and gas-filled units. The standard covers internal insulation. External insulation requirements are agreement-based.
Bushings. Tested separately under IEC 60137. The dielectric tests in IEC 60076-3 verify the bushings are correctly applied and installed, but they don’t substitute for bushing tests.
Tap-changers. Specified, designed and tested under their own standards (IEC 60214 series). Same logic as bushings.
Insulation resistance testing. Not in IEC 60076-3. Different standard, different test method.
Power factor / dissipation factor testing. Not in IEC 60076-3.
DC withstand testing. Not in IEC 60076-3.
Frequency response analysis (FRA). Not in IEC 60076-3.
Special standards for specific transformer types (instrument transformers, traction transformers, converter transformers, etc.) — IEC 60076-3 only applies where these other standards specifically cross-reference it.

The standard is narrow: AC, impulse, and PD dielectric tests for power transformers, plus external air clearances. That’s it.

A Short History

The first edition of IEC 60076-3 dates to the 1980s, replacing the older IEC 76-3. The second edition was published in 2000. The third edition — IEC 60076-3:2013 Edition 3.0, July 2013 — is the current one.

The main changes in the 2013 edition, listed in the foreword:

Three categories of transformer clearly identified by Um with corresponding test requirements (Table 1). This is the master organizing structure of the new edition.
Switching impulse levels defined for all Um > 72.5 kV. Previously these were defined only for higher voltage ranges.
IVPD procedure revised to ensure adequate phase-to-phase test voltages.
AC withstand test renamed. Now called LTAC (Line Terminal AC withstand) instead of ACSD (AC Short Duration).
Induced voltage tests now based on Ur instead of Um. This is a substantive change — Ur is the rated voltage of the winding, Um is the highest voltage for equipment. They’re not the same number.
k-factor introduced for impulse waveshape evaluation (cross-referenced to IEC 60060-1).
Test level tables merged and aligned with IEC 60071-1:2010.
Additional test levels added for Um > 800 kV. Ultra-high-voltage equipment is now covered.
New Annex E sets out the engineering principles behind the test levels and clearances.

If you’re reading a spec written before 2013, expect old terminology and old voltage bases. The 2013 edition doesn’t grandfather old specs — it just supersedes the old standard.

Clause 2: Normative References

Other documents IEC 60076-3 leans on. Knowing these matters because they get cited downstream:

IEC 60050-421 — International Electrotechnical Vocabulary, transformers and reactors chapter. Where terminology comes from.
IEC 60060-1 — High-voltage test techniques, general definitions and test requirements. This is the standard that defines what a “lightning impulse” actually is, what tolerances apply, and so on.
IEC 60060-2 — High-voltage test techniques, measuring systems. Tells you what measurement equipment qualifies.
IEC 60071-1 — Insulation co-ordination, definitions, principles and rules. Where the test levels in Table 2 come from.
IEC 60076-1 — Power transformers, Part 1: General. The foundation document.
IEC 60137 — Insulated bushings for AC voltages above 1000 V. Bushing-specific standard.
IEC 60270 — High-voltage test techniques, partial discharge measurements. Defines how PD is measured and calibrated.

If your spec references IEC 60076-3, it implicitly references all of these.

Clause 3: Definitions That Matter

A handful of definitions control how everything in the standard reads.

Um — highest voltage for equipment applicable to a transformer winding. The highest RMS phase-to-phase voltage in a three-phase system that the winding’s insulation is designed for. Note the wording: it’s a system-side number, not a winding-side number. A winding might be rated 220 kV but be installed in a system with Um = 245 kV.

Ur — rated voltage of a winding. The voltage assigned to be applied or developed at no-load. For three-phase windings, this is the line-to-line voltage. For single-phase windings to be connected star in a three-phase bank, it’s the phase-to-phase voltage divided by √3 (typically written 400/√3 kV).

The distinction between Um and Ur matters because 2013 edition test calculations sometimes use one, sometimes the other. The induced voltage test enhancement level is (1.8 × Ur)/√3. The applied voltage test value comes from Table 2 indexed by Um. Mix them up and you get the wrong test.

Rated insulation level. The set of rated withstand voltages that characterise the dielectric strength of the insulation. Expressed on the rating plate as Um / SI / LI / LIC / AC with associated values.

Rated withstand voltage. The value of the test voltage applied during the standard dielectric test that proves compliance. This is what shows up in Table 2.

Uniform insulation of a transformer winding. Insulation where all ends of the winding are connected to terminals with the same rated insulation level.

Non-uniform insulation of a transformer winding. Insulation where the winding has a neutral terminal designed for connection to earth (direct or indirect), and is designed with a lower insulation level than the line terminal. Also called “graded insulation.”

This last distinction drives whole sections of the standard. Non-uniformly insulated windings need different tests — the applied voltage test only stresses the insulation to the neutral level, so you need an LTAC or SI test to verify the line terminal insulation.

Clause 4: General Requirements

This clause sets the framing for everything that follows. A few key points buried in here that get missed:

Internal insulation only. For liquid-immersed and gas-filled transformers, the dielectric tests cover internal insulation. External insulation is agreement-based.

Bushings tested separately. IEC 60137. The transformer tests verify correct application and installation only.

Insulation temperature minimum. The insulation system temperature shall not be less than 10 °C during tests. Higher temperatures from IEC 60076-1 may be used.

Complete assembly required. The transformer shall be assembled as it will operate, including all elements that might influence dielectric strength. Coolers normally don’t need to be assembled. Insulating liquid or gas isn’t normally circulated during tests.

Free gas detection. Any equipment to detect free gas from internal faults shall be installed and monitored during the tests. If free gas appears, investigate and agree on further action.

Same liquid as service. Liquid-immersed transformers shall be tested with the same type and specification of insulating liquid (mineral, ester, silicone) as they will contain in service.

Altitude correction. For operation above 1000 m, clearances shall be designed accordingly. Bushings may need higher insulation levels than otherwise required.

Non-linear elements declaration. If the manufacturer uses non-linear protection elements (surge arresters, spark gaps) built into the transformer or tap-changer, this must be declared to the purchaser at tender and order, and shown on the rating plate.

DGA before and after. Note 2 mentions common practice: take oil samples for dissolved gas analysis before and after dielectric tests on larger transformers. Not required by the standard, but standard industry practice.

Clause 5: Highest Voltage for Equipment and Rated Insulation Level

This is short but defines how Um gets assigned and how the rating plate is structured.

A value of Um is assigned to both the line and neutral end of each winding. When tests for different windings conflict, the rules for the highest Um value win.

For series windings in autotransformers and phase-shifting transformers, where the rated voltage of the winding is less than the rated voltage of the system, Um corresponds to the rated voltage of the highest-voltage system the winding is connected to.

Standard Um values come from Table 2. Unless otherwise specified, the value used is the one equal to or nearest above the winding’s rated voltage.

The rated insulation level is written as:

Um / SI / LI / LIC / AC with the associated values, where:

SI = rated switching impulse withstand voltage (line terminals of the highest-Um winding)
LI = rated lightning impulse withstand voltage (each terminal of each winding)
LIC = rated chopped wave lightning impulse withstand voltage (if LIC was performed)
AC = highest rated AC withstand voltage to earth

If a level doesn’t apply, it’s omitted. For neutral terminals or windings without SI/LIC, it shortens to Um / LI / AC.

Standard’s Example 1 (66/11 kV transformer):

HV: Um 72.5 / LI 325 / AC 140 kV
LV: Um 12 / LI 75 / AC 28 kV

Standard’s Example 2 (220 kV transformer with non-uniform HV insulation):

HV: Um 245 / SI 750 / LI 950 / LIC 1045 / AC 395 kV
HVN: Um 52 / LI 250 / AC 95 kV
MV: Um 72.5 / LI 325 / AC 140 kV
LV: Um 24 / LI 125 / AC 50 kV

That’s the format every rated plate follows.

Clause 6: Re-connectable Windings

Short clause. Windings that can be connected in more than one configuration shall be tested in each configuration, unless otherwise specified. So a winding that can be reconnected star-delta gets two test sequences.

Clause 7: Dielectric Tests — The Master Test Plan

This is the longest and most-cited clause. Two pieces: the test definitions in 7.1, and the test requirements matrix in 7.2 (Table 1).

7.1 The Tests Defined

Eleven distinct tests are named in the standard. Each one has a purpose:

LI — Full wave lightning impulse test for line terminals. Verifies the winding’s ability to withstand fast rise time transients from lightning. Standard 1.2/50 µs waveshape (defined in IEC 60060-1).

LIC — Chopped wave lightning impulse test for line terminals. Same intent as LI plus verification against high-frequency phenomena. The impulse is chopped on the tail to produce a very high rate of change. Test peak value is higher than LI.

LIN — Lightning impulse test for the neutral terminal. Verifies impulse withstand of the neutral and its connected winding to earth and other windings. Only applicable when the neutral isn’t directly earthed in service.

SI — Switching impulse test for the line terminal. Verifies withstand against slow rise time transients from switching operations. Single-phase test. Voltage distributes inductively through all windings. Line terminals open circuit during the test.

AV — Applied voltage test. Verifies AC withstand of line and neutral terminals to earth and other windings. Voltage is applied to all terminals of the winding simultaneously, including the neutral — so there’s no turn-to-turn voltage stress, only ground-wall stress.

LTAC — Line terminal AC withstand voltage test. Verifies AC withstand of each line terminal to earth. Allows non-uniformly insulated transformers to be AC-tested at the line terminal level (which the AV test can’t reach because it’s limited by the neutral insulation).

IVW — Induced voltage withstand test. Verifies AC withstand of each line terminal and its winding to earth and other windings, plus along-the-winding and phase-to-phase withstand. Transformer is connected as for service. Symmetrical voltages appear at all line terminals with no voltage at the neutral.

IVPD — Induced voltage test with partial discharge measurement. Same connection as IVW. Verifies the transformer will operate without harmful partial discharge under normal service conditions. The PD acceptance criteria are the heart of the standard’s quality control.

AuxW — Auxiliary wiring insulation test. Verifies the insulation of auxiliary wiring not connected to the main windings.

LIMT — Lightning impulses applied to multiple terminals simultaneously. “Double-ended lightning impulse test.” Only for special transformer types — phase-shifting transformers with on-load bypass, or where multiple terminals see simultaneous impulses in service.

7.2 Test Requirements Matrix (Table 1)

This is the table that defines what’s routine, what’s a type test, what’s a special test, and what doesn’t apply — based on the Um of the highest-voltage winding.

The three categories:

Category A: Um ≤ 72.5 kV. Distribution transformers and smaller power transformers.

LI: Type test
LIC: Special
LIN: Special
SI: Not applicable
AV: Routine
IVW: Routine
IVPD: Special
LTAC: Not applicable
AuxW: Routine

Category B: 72.5 kV < Um ≤ 170 kV (uniform insulation).

LI: Routine
LIC: Special
LIN: Special
SI: Special
AV: Routine
IVW: Routine
IVPD: Routine
LTAC: Special
AuxW: Routine

Category B (non-uniform insulation): 72.5 kV < Um ≤ 170 kV.

LI: Routine
LIC: Special
LIN: Special
SI: Special
AV: Routine
IVW: Routine
IVPD: Routine
LTAC: Routine (can be replaced by SI by agreement)
AuxW: Routine

Category C: Um > 170 kV (uniform and non-uniform).

LI: Not applicable (included in LIC)
LIC: Routine
LIN: Special
SI: Routine
AV: Routine
IVW: Not applicable
IVPD: Routine
LTAC: Special
AuxW: Routine

A few patterns worth noting:

AV and AuxW are routine for everything. No transformer ships without these.
LI moves from type to routine to absorbed-into-LIC as voltage climbs.
IVPD moves from special to routine at the 72.5 kV boundary. Below that, IVPD is only done if asked for.
IVW disappears at Um > 170 kV — the IVPD enhancement voltage covers it.
LIN is always special. Never routine. Only done when the neutral isn’t directly earthed and the purchaser asks.

7.2.2 Test Voltage Levels (Tables 2 and 3)

Table 2 gives the standard test voltage levels — the heart of the standard, indexed by Um. For each Um value there can be several rows representing different insulation levels. The purchaser picks the appropriate row.

A few representative entries from Table 2:

Um (kV)	LI (kV)	LIC (kV)	SI (kV)	AV/LTAC (kV)
12	75	83	—	28
24	125	138	—	50
36	170	187	—	70
72.5	325	358	—	140
123	450	495	375	185
245	950	1045	750	395
420	1300	1430	1050	570
800	2100	2310	1675	880
1200	2250	2475	1800	—

Key rule from 7.2.2: “All test voltages are phase to earth.” So when you see AV = 395 kV at Um = 245 kV, that’s the line-to-earth voltage applied during the AV test, not phase-to-phase.

The lowest value given in Table 2 for a particular Um is the minimum. The purchaser can specify higher, preferably standard values for coordination, but not necessarily from a single row. Mixing rows is permitted but may result in over-design.

Table 3 gives lower values for special cases where Table 2 minimums are considered too high. These can only be used with extensive studies or proven existing practice. This is an escape valve for unusual installations, not a routine option.

7.2.3 Test Sequence

Tests are performed in this sequence:

Lightning impulse tests (LI, LIC, LIN, LIMT)
Switching impulse (SI)
Applied voltage test (AV)
Line terminal AC withstand test (LTAC)
Induced voltage withstand test (IVW)
Induced voltage test with PD measurement (IVPD)

By agreement, SI can come before LI. If IVPD is performed, the sequence can vary by agreement except that IVPD must be last. This is critical — IVPD is the final acceptance test because its sensitivity depends on a stable thermal and dielectric state.

7.3 Test Requirements for Specific Transformers

Sub-clauses 7.3.1 through 7.3.3 give the detailed test procedures for each of the three categories. Each repeats the routine, type, and special test designations from Table 1 with the specific voltage formulas.

For IVPD in particular, the formulas to remember:

Standard IVPD enhancement voltage: (1.8 × Ur)/√3 Standard one-hour PD measurement voltage: (1.58 × Ur)/√3

Higher levels may be specified. Common alternative: enhancement at (√3 × Um)/√3 = Um/√3 × √3… wait, the formula is enhancement at (√3 × Um)/√3 and PD measurement at (1.5 × Um)/√3 if higher.

These two formulas — Ur-based versus Um-based — are the practical face of the 2013 edition’s shift to Ur. If your spec was written before 2013 using Um throughout, it’s not wrong, but it’s higher voltage than the current minimum.

7.4 Neutral Terminal Voltages

For Um ≤ 72.5 kV, the neutral is tested at the line terminal AV level. Uniform insulation is assumed.

For Um > 72.5 kV with directly earthed neutral (or through a CT, no added impedance), the AV test voltage shall be at least 38 kV when Um ≥ 17.5 kV.

For Um > 72.5 kV with non-directly-earthed neutral, the Um and test voltages for the neutral are specified by the purchaser. Um shall in no case be less than 17.5 kV. Annex D walks the calculation of neutral insulation levels for transformers earthed through an impedance.

Clause 8: Tests on Transformers That Have Been in Service

A clause people overlook. It covers two scenarios:

As-new condition. Following a warranty repair or complete rewind that restores the transformer to “as new” — all routine tests at 100 % of original test voltage. The transformer is treated as new.

Repair to restore functionality. After a breakdown following service. Tests at 80 % to 100 % of original test voltage level. New parts at 100 %, used parts at 80 % may be adequate. IVPD always at 100 %. PD acceptance criteria may be modified by agreement.

This is the standard’s framework for retesting after repair. Different from new-machine testing.

Clause 9: Auxiliary Wiring (AuxW)

Power and control circuitry. Test conditions:

General auxiliary wiring: 2 kV AC for 1 minute, applied to earth. Test passes if no voltage collapse.
Current transformer secondary wiring: 2.5 kV AC for 1 minute. If the CT knee-point voltage exceeds 2 kV, test at 4 kV.
Wiring disconnected for transport: Retested at site after erection, either at 2 kV AC repeat or 1 kV DC IR measurement (minimum 1 MΩ).
Motors and apparatus: Tested per their relevant IEC standards (often lower than the wiring values, so may need to be disconnected).
Solid state and microprocessor devices: Excluded from the test circuit.
Three-phase undervoltage relays and withdrawable devices: Removed from the test circuit.

Note in the standard: it’s normal practice to check all auxiliary wiring on site at 1 kV DC for 1 minute before energisation.

Clause 10: Applied Voltage Test (AV)

The simplest of the dielectric tests. Procedure:

Full test voltage applied for 60 seconds.
All accessible terminals of the winding under test connected together.
All other windings, core, frame, and tank connected to earth.
Voltage approximately sinusoidal, frequency at least 80 % of rated.
Measure peak voltage. Peak divided by √2 equals the test value.
Voltage starts at not more than one-third of test value, increased rapidly to test value.
At end, voltage reduced rapidly to less than one-third before switching off.

Test passes if no voltage collapse occurs.

For non-uniformly insulated windings, the test is at the neutral terminal voltage. This is why the line terminal needs a separate test (LTAC or SI) on these.

Clause 11: Induced Voltage Tests (IVW and IVPD)

The induced voltage tests are run with the transformer connected as for service. All neutral terminals and other normally-earthed terminals are earthed. For three-phase transformers, a symmetrical three-phase voltage is used.

The voltage is induced from one winding through the magnetic circuit to the others. Frequency must be sufficiently above rated to avoid excessive magnetising current. Either an external generator at higher frequency, or a step-up transformer can be used.

11.2 IVW — Induced Voltage Withstand

Test time at full voltage:

At test frequency up to 2× rated: 60 seconds, unless otherwise specified.
At test frequency above 2× rated: 120 × (rated frequency / test frequency), but not less than 15 s.

So a 50 Hz transformer tested at 200 Hz: 120 × 50/200 = 30 seconds. Tested at 400 Hz: 120 × 50/400 = 15 seconds (the minimum).

Pass criterion: no voltage collapse.

11.3 IVPD — Induced Voltage Test with PD Measurement

This is where the standard gets specific about acceptance criteria for partial discharge.

Test duration at enhancement voltage:

Um ≤ 800 kV: 60 seconds at test frequency up to 2× rated.
Um > 800 kV: 300 seconds at test frequency up to 2× rated.

At higher test frequencies, duration scales the same way as IVW.

Test sequence (a 14-step procedure summarised):

Switch on at not higher than (0.4 × Ur)/√3
Raise to (0.4 × Ur)/√3, record background PD
Raise to (1.2 × Ur)/√3, hold for at least 1 minute, measure PD
Raise to one-hour PD measurement voltage, hold 5 minutes minimum, measure PD
Raise to enhancement voltage, hold for test time
Reduce without interruption to one-hour PD measurement voltage
Measure PD
Hold at one-hour PD measurement voltage for at least one hour
Measure PD every 5 minutes during the one hour
Reduce to (1.2 × Ur)/√3, hold at least 1 minute, measure PD
Reduce to (0.4 × Ur)/√3, measure background PD
Reduce below (0.4 × Ur)/√3
Switch off

PD measured continuously on at least one channel throughout. Inception and extinction voltages of significant PD noted to aid evaluation.

PD measurement requirements (11.3.4):

PD measured per IEC 60270.
Each PD channel including bushing or coupling capacitor calibrated in apparent charge (pC).
PD given in pC, referring to highest steady-state repetitive impulses.
Occasional bursts may be disregarded.
For each step, PD measured on all line terminals equipped with bushings with Um ≥ 72.5 kV. If more than six such terminals exist, only the six highest-voltage terminals need to be measured.

Acceptance criteria (11.3.5):

The test is valid only if background PD doesn’t exceed 50 pC at both start and end (100 pC for shunt reactors).

The test passes if all of:

a) No collapse of test voltage. b) None of the PD levels during the one-hour period exceed 250 pC. c) PD levels during the hour don’t show a rising trend and no sudden sustained increase in the last 20 minutes. d) PD levels during the hour don’t increase by more than 50 pC. e) PD at (1.2 × Ur)/√3 after the hour doesn’t exceed 100 pC.

If c) or d) fail, the one-hour period may be extended, and they’re considered met if fulfilled for any continuous one-hour period.

The standard explicitly notes the test is non-destructive (as long as no breakdown occurs). Failure of the PD criteria doesn’t warrant immediate rejection — it leads to consultation between purchaser and manufacturer about further investigations. Annex A gives guidance on what to do next.

Clause 12: LTAC — Line Terminal AC Withstand

The test arranged so test voltage appears between the tested terminal and earth. Each line terminal of the tested winding tested in turn. Test time, frequency and voltage application as for IVW (clause 11.2).

For non-uniformly insulated lower voltage windings with taps, the tap position is selected to bring the lower-voltage terminal voltage as close as possible to the required value.

Pass: no voltage collapse.

The standard’s note in this clause is worth reading: LTAC is intended only as a withstand test for the line terminal to earth. It’s not testing phase-to-phase or turn-to-turn. The test arrangement can be done any convenient way, including with voltage at the neutral to reduce turn-to-turn stress. Normally done as three single-phase tests.

If a switching impulse test is performed, LTAC may be omitted by agreement.

Clause 13: Lightning Impulse Tests (LI, LIC, LIN, LIMT)

The longest clause. Lightning impulse is the dominant dielectric stress test for high-voltage windings.

13.1 General Requirements

Polarity. For liquid-immersed transformers, normally negative polarity (reduces external flashover risk). Positive may be specified by the purchaser. Mixed polarities require additional reference impulses and agreed test sequence.

Tap positions:

Tapping range ±5 % or less AND rated power ≤ 2500 kVA: test on principal tapping only.
Tapping range > ±5 % OR rated power > 2500 kVA: test on two extreme tappings and the principal tapping — one tapping per phase of a three-phase transformer.
Special cases (single-phase, multiple tap-changers, non-symmetrical range): tap position giving highest internal voltages, determined by calculation or low-voltage impulse measurement.

Records. Applied voltage recorded per IEC 60060-2. Records show wave shape (front time, time-to-half, amplitude). At least one additional measurement channel — typically neutral current or capacitive probe current. The exact detection method is agreed between manufacturer and purchaser.

13.2 Full Wave LI Test

Wave shape per IEC 60060-1: 1.2 µs front time, 50 µs time-to-half. Tolerances per IEC 60060-1.

The standard distinguishes between transformers with and without non-linear elements (surge arresters, spark gaps):

Without non-linear elements:

One reduced-amplitude reference impulse (50–75 % of full test voltage)
Three impulses at full test voltage
Records compared between reference and full-voltage impulses

With non-linear elements:

Series of impulses at increasing levels
Records compared, considering the non-linear element’s operating threshold

Pass criterion: no significant differences between the voltage and current records at reference and full test voltages, beyond what can be explained by minor voltage variations.

13.3 Chopped Wave LIC Test

The full wave impulse is augmented with impulses chopped on the tail. Time to chop typically 2–6 µs.

Test sequence example for transformers without non-linear elements:

a) Reduced-amplitude reference impulse (≤ 75 %) b) Full wave at 100 % c) Full wave at 100 % d) Chopped wave at reduced amplitude e) Chopped wave at full level (LIC peak) f) Chopped wave at full level g) Full wave at 100 % h) Full wave at 100 % i) Reduced-amplitude reference

Pass criterion same principle as LI — no significant differences beyond explainable variation.

The chopped wave test is more onerous because the chop produces very high frequency components. The peak voltage is higher (LIC values in Table 2 are 10 % higher than LI). This test is the routine test for Um > 170 kV; for lower voltages, it’s special.

13.4 LIN — Lightning Impulse on Neutral Terminal

Wave shape per 13.2.1 except front duration may be up to 13 µs (compared to standard 1.2 µs). The neutral sees slower wave fronts in service due to the long path along the winding.

Test sequence per the line terminal procedure. Test criteria same.

Done only when specified by the purchaser, and normally only when the neutral isn’t directly earthed in service.

Clause 14: Switching Impulse Test (SI)

Distinctive from LI in several ways.

Waveshape: Time to peak (Tp) at least 100 µs (vs 1.2 µs for LI). Time above 90 % (Td) at least 200 µs. Time to zero (Tz) minimum 1000 µs. This is the standard transformer switching impulse, different from the general 250/2500 µs in IEC 60060-1 because transformers have saturable cores.

The time to zero requirement triggers the standard’s note about core saturation. Considerable flux develops in the magnetic circuit during the test. Voltage can be sustained only until the core saturates. To get long enough time to zero, reverse polarity priming impulses are used before each full test impulse to reset the core.

Connection: Impulses applied directly to a line terminal of the highest-voltage winding, or to a lower-voltage winding for inductive transfer. Single-phase test on three-phase transformers (phase by phase).

Star windings with neutral brought out: Neutral earthed directly or through small impedance. Voltage of opposite polarity and about half amplitude appears on remaining line terminals.

Delta windings: Terminal at end of phase under test earthed. Other terminals open. Different terminal earthed for each phase test.

Test sequence: One reference impulse at 50–70 % of full voltage. Three impulses at full voltage. Reverse polarity priming impulses (≤ 70 % full level) before each full impulse to set core magnetisation.

Pass criterion: No sudden voltage collapse or discontinuity in voltage or current oscillograms. Successive oscillograms may differ due to magnetic saturation effects.

Clause 15: Action Following Test Failure

If the transformer fails any dielectric test, the complete sequence is repeated at full level after repair. Parts fully tested and clearly not involved in the failure may be excluded at the purchaser’s discretion. Particular attention to whether internal transients or contamination might have damaged other parts.

If failure is in a bushing and purchaser is satisfied the transformer wasn’t affected, the bushing may be replaced and testing continued.

If failure was external (e.g., flashover external to the transformer), repeat the particular test; if successful, complete the sequence without repeating earlier successful tests.

Clause 16: External Clearances in Air

This clause applies when clearances aren’t specified by the purchaser. When specified, the manufacturer may use higher values if required for testing.

“Clearance in air” is defined as the shortest distance between any metallic part of the bushing terminal and any part of the transformer, taking a line that doesn’t pass through the bushing insulator. The bushing’s own length isn’t part of these clearances.

Three clearance categories:

Phase-to-earth and phase-to-neutral. Largest of LI-based and SI-based values.
Phase-to-phase between phases of the same winding. Based on conductor-to-conductor configuration with phase-to-phase divided by phase-to-earth ratio of 1.5.
Between line terminals of HV winding and LV winding terminals. Phase-to-earth values apply.

Selected values from Table 4:

Um (kV)	LI (kV)	SI (kV)	Line-earth (mm)	Phase-phase (mm)
24	125	—	220	220
72.5	325	—	630	630
123	450	375	900	900
245	950	750	1700	2300
420	1425	1175	3100	4200
800	2050	1675	5600	7700

Altitude correction: above 1000 m, increase clearances by 1 % per 100 m of additional altitude.

These are minimum nominal values. Actual clearances must be at least equal to specified, accounting for manufacturing tolerances. Stated on the outline drawing as design clearances.

Annex A: PD Measurement Application Guide

Informative annex covering practical aspects of PD measurement that don’t fit cleanly into the main clauses.

Calibration. PD measurement channels must be calibrated in apparent charge (pC) per IEC 60270. Calibration uses either the bushing test tap (for condenser-type bushings) or a high-voltage coupling capacitor.

Wide-band vs narrow-band. Wide-band measurement preferred (bandwidth not less than 100 kHz) for greater chance of detecting PD. Narrow band may be required to eliminate interference. Centre frequency choice matters for narrow-band — must give reasonable sensitivity to PD in the transformer.

Interference reduction. Shielding, band-stop filters, low-pass filters on supply leads, electrostatic shielding outside the transformer (provisional, for test only).

Procedure after unsuccessful test. The annex lists eight types of investigation:

Determine if indications are correlated to the test sequence or coincident irrelevant sources.
Check whether PD is transmitted from the supply.
Determine whether the source is internal or external to the transformer.
Locate the source in terms of the electrical circuit diagram.
Acoustic or ultrasonic location within the tank.
UHF electromagnetic sensors introduced into the tank.
Determine probable physical nature from voltage variation, hysteresis, pulse pattern, time development.
Re-process the transformer (additional drying or oil impregnation) and repeat test.

The annex notes specifically that limited exposure to high PD can cause temporary local oil cracking with reduced extinction and re-inception voltages, but the original conditions may self-restore in hours. Limited variation of PD with voltage, a pattern typical of a floating particle, and no time-development may be acceptable evidence the transformer is suitable for service.

This is important — IEC 60076-3 doesn’t treat a PD failure as automatic rejection. It triggers investigation and agreement.

Annex B: Transferred Overvoltage

Informative. Covers overvoltages transferred from HV to LV windings when the LV winding is loaded or open in service.

Two mechanisms:

Capacitive transfer. Voltage division between transfer capacitance Ct (between windings) and loading capacitance Cs. Order of magnitude: Ct ≈ 10⁻⁹ F. Loading the secondary with switchgear, short cables, or added capacitors (few nF) reduces the transferred peak. Longer cables represented by characteristic impedance.

Inductive transfer. Depends on surge current flow in the HV winding. Without external loading, voltage transient usually has a superimposed damped oscillation at a frequency set by leakage inductance and winding capacitances. Damped by resistive surge diverters or by capacitive loading (typically tenths of microfarads, which also eliminate the capacitive component).

Also covers power-frequency transferred overvoltage when an LV winding adjacent to an energised HV winding is left without earth connection. Resonance conditions can arise. Tertiary windings and stabilising windings face the same risk. Stabilising windings should normally be permanently earthed (externally or internally to tank).

The annex doesn’t impose requirements — it explains the physics and the practical mitigation methods.

Annex C: Information to Supply with Enquiry and Order

Informative. Lists what the purchaser should specify when ordering, by Um category.

All cases (any Um):

Um value
Ur value
AV test level
LI test level (or for Um > 170 kV, also SI)

Special cases as needed:

PD measurement requirements (with IVW or separate, on which units, duration if less than one hour)
LIC requirement and which units
LIN requirement and which units
LIMT requirement
Clearances if different from Clause 16
LTAC for uniform insulation (Um > 72.5 kV)
Alternative higher IVPD voltages
Whether IVPD substitutes for IVW
Switching impulse for 72.5 kV < Um ≤ 170 kV uniform insulation

The annex strongly recommends choosing Um, LI, AV, and other values from a single line of Table 2 as they’re coordinated. Mixing values from different lines may result in over-design.

For complex transformers — non-uniformly insulated HV windings, high-power LV windings, special neutral arrangements — the annex recommends discussing test connections and procedures at order placement or design review stage.

Annex D: Neutral Insulation Voltage Level Calculation

Informative. Walks through how to determine the minimum withstand voltage for a non-directly-earthed neutral terminal.

Key points:

When the neutral isn’t directly earthed, an overvoltage protective device shall be installed between the neutral and earth to limit transient voltages.
Purchaser selects the device, determines its impulse protection level, and specifies the corresponding impulse withstand for the neutral.
AC withstand voltage shall be higher than the maximum overvoltage from system fault conditions.
Margin between neutral impulse level and protective device operating voltage. Both above maximum fault voltage.

The annex gives formulas for calculating maximum neutral fault current and resulting maximum neutral voltage for two-winding three-phase transformers (single-phase earth fault on a star-connected winding). The math involves positive, negative, and zero-sequence impedances. This is the engineering needed to set neutral test voltages for non-directly-earthed configurations.

Annex E: Basis for the Test Levels (New in 2013)

This is the most useful annex if you want to understand why the standard says what it does. Added in the 2013 edition specifically to document the engineering rationale.

Why three categories? Distribution and small power transformers (Um ≤ 72.5 kV) are produced in large quantities, often of one design. Testing should remain quick and affordable. All are uniformly insulated, so AV proves line-to-earth and neutral-to-earth simultaneously. IVW only needs to check turn-to-turn. LI is type test (not routine) to limit testing equipment requirements. IVPD is special — its goals are mostly covered by IVW and AV at this voltage range.

Middle range (72.5 kV < Um ≤ 170 kV) covers higher-power units, often built to order. Both uniform and non-uniform insulation possible. IVW and IVPD both routine — but can be combined into one test by agreement. LTAC routine for non-uniform insulation.

Large transmission and generation transformers (Um > 170 kV) get the full battery: LIC routine, SI routine, IVPD with its full one-hour PD hold routine. IVW dropped because the IVPD enhancement covers it. AV always routine.

Why these test voltage ratios? The standard documents the engineering relationships:

Switching impulse withstand (SI) is typically 0.8–0.85 of lightning impulse withstand (LI). IEEE C57.12.00 uses 0.83.
Induced voltage withstand (IVW) line-to-earth is about 50 % of SI, i.e. 40–43 % of LI.
PD measurement level (1.58 × Ur)/√3 equals 1.5 × Ur + 5 %. Same formula used in IEEE C57.12.00-2010.
Enhancement level (1.8 × Ur)/√3 equals 1.7 × Ur + 5 %, harmonised with IEEE C57.12.00-2010.

This is why the standard’s values look like odd numbers. They’re derived from these ratios, then rounded.

Why the clearances? Line-to-earth from IEC 60071-1 rod-to-structure for LI ≤ 750 kV, conductor-to-structure for higher LI. Phase-to-phase from conductor-to-conductor re-based to a ratio of 1.5 (transformer with delta winding). Higher of LI-derived and SI-derived values controls.

What’s Not in IEC 60076-3 (and What Standards Cover Them)

To avoid citing this standard for things it doesn’t address:

Insulation resistance (DC) testing — not in IEC 60076-3. No specific IEC standard for transformer IR; relies on manufacturer practice and IEEE 62 family.
Power factor / dissipation factor — not in IEC 60076-3. IEEE C57.152 covers this for field testing.
Frequency response analysis (FRA) — IEC 60076-18.
Sweep frequency response analysis — IEC 60076-18.
Sound level — IEC 60076-10.
Short-circuit withstand — IEC 60076-5.
Temperature rise — IEC 60076-2.
Loss measurements — IEC 60076-1.
Bushing tests — IEC 60137.
Tap-changer tests — IEC 60214-1.
Dissolved gas analysis — IEC 60599 (interpretation), IEC 60567 (sampling).
PD location / identification — IEC 60076-3 Annex A is informative only; CIGRE technical brochures cover field practice.
HV test technique fundamentals — IEC 60060-1 (definitions), IEC 60060-2 (measuring systems), IEC 60060-3 (on-site testing).
Insulation coordination — IEC 60071-1 (principles), IEC 60071-2 (application).

IEC 60076-3 is one piece of a larger family. A complete transformer test spec references several standards in this list.

Using the Standard in Practice

How to actually work with IEC 60076-3:2013:

Writing acceptance specs. Cite the standard with the edition year: “IEC 60076-3:2013.” Specify the routine and special tests required by referencing the test designations (LI, LIC, SI, AV, IVW, IVPD, LTAC). Specify the Um values for each winding. Specify whether non-routine tests (LIC, LIN, LIMT, SI for Category B, LTAC for uniform insulation) are required. Annex C is the checklist.

Reading a factory test report. Test designations should match the standard. PD acceptance criteria should reference 11.3.5 — 250 pC max, 100 pC at 1.2×Ur/√3 after the hour, no rising trend. Background PD should be reported at start and end (must be ≤ 50 pC for valid test, ≤ 100 pC for shunt reactors).

Investigating a PD failure. The standard explicitly contemplates that PD criteria failure isn’t automatic rejection. Annex A is the framework for the investigation that follows. Document the agreement with the manufacturer about further actions.

Specifying clearances. Default to Clause 16/Table 4. Specify only if your installation requires different. Add altitude correction if above 1000 m.

Acceptance after repair. Clause 8 is the controlling section. As-new (100 % test voltage) or functional repair (80–100 %) are different cases.

Audit defense. Know which tests are routine, type, and special for each Um category. Know the difference between Um and Ur (and which the test formula uses). Know that Annexes A through E are informative, not normative — they don’t impose requirements, they explain rationale and give guidance.

The standard is roughly 56 pages in English (the document also contains a French version, doubling the page count). It’s not light reading, but it’s the controlling document for power transformer dielectric testing in most of the world. If you write specs, audit factory tests, witness factory acceptance tests, or interpret reports — worth having on hand.