Pre-commissioning tests on a distribution or power transformer: what to witness at works and what to repeat on site

Witness the routine tests at the manufacturer’s works, because that is the only place the transformer’s losses, impedance and dielectric strength can be proven at full test levels. On site, repeat the tests that detect what transport, erection and oil filling can change: insulation resistance, ratio at every tap, winding resistance, vector group, oil condition and, on larger units, frequency response. Compare every site reading against the works figure, not against a generic pass mark. New India Electricals Ltd (NIEL) supplies distribution transformers from 15 to 2500 kVA and power transformers from 2500 to 100,000 kVA at primary voltages up to 220 kV.

TL;DR

• IS 2026 (Part 1): 2011 lists six routine tests: winding resistance, voltage ratio and phase displacement, impedance and load loss, no-load loss and current, dielectric routine tests, and on-load tap changer tests.

• Lightning impulse is a type test for windings with Um up to 72.5 kV and a routine test above 72.5 kV under IS 2026 (Part 3). Below that line, ask for the type test report or order it.

• Insulation resistance and tan delta at works are special tests. IS 2026 (Part 1) calls them “reference values for comparison with later measurement in the field.” Order them, or site readings have nothing to be compared against.

• The CEA Safety Regulations, 2023 require site tests as per relevant standards before any new apparatus above 650 V is commissioned, with records kept for comparison.

• IS 1866 sets minimum breakdown voltage for new oil in a new transformer at 40, 50 and 60 kV for equipment below 72.5 kV, 72.5 to 170 kV, and above 170 kV.

• Repeating full dielectric withstand tests on a new transformer means 100 percent of test level under IS 2026 (Part 3), so it is not a casual site check.

• Frequency response analysis baselines taken at works and repeated on site catch winding movement from transport that no electrical test at power frequency will show.

What is the difference between routine, type and special tests on a transformer?

A routine test is made on every transformer. A type test is made on one transformer representative of a design. A special test is any other test agreed between manufacturer and purchaser. These are the definitions in IS 2026 (Part 1): 2011, the Indian standard for power transformers, and they decide what a purchaser gets by default.

The practical consequence sits in the third row. Insulation resistance and dissipation factor are the two measurements most site engineers rely on, and IS 2026 (Part 1) treats both as special tests. The standard notes that these readings are “reference values for comparison with later measurement in the field,” and sets no limits for them. If the purchase order does not ask for them, the works has no obligation to measure them, and the site team has no baseline.

On type tests, the standard allows a type test to count for a transformer that is “fully identical in rating and construction,” and for minor deviations only by agreement. Ask which unit the type test report covers and whether its rating matches yours.

Which tests should you witness at the manufacturer’s works?

Witness the full routine test sequence, and above all the loss and impedance measurements and the dielectric tests. These cannot be reproduced at full level on site. They also settle the guaranteed figures that the commercial terms of the order depend on.

Losses and impedance. IS 2026 (Part 1) requires the short-circuit impedance and load loss test at rated frequency, with supplied current “equal to the relevant rated current (tapping current) but shall not be less than 50 percent thereof.” Check the current actually used and how the result was corrected. No-load loss runs for every hour the transformer is energised, and load loss rises with the square of loading. NIEL’s guide to sizing a distribution transformer explains why the two must be guaranteed as separate figures. The works test is where those guarantees are measured.

Winding resistance. For oil-immersed units, IS 2026 (Part 1) requires the transformer to have been “under oil without excitation for at least 3 h.” The winding temperature is then taken as the average of top and bottom oil temperature. Record that temperature with the reading. Without it, the site measurement cannot be corrected to the same basis.

Dielectric routine tests. IS 2026 (Part 3): 2009 states that insulation tests are generally made at the supplier’s works, with the transformer at approximately ambient temperature but at least 10 degrees C. The separate source AC withstand test applies the full test voltage for 60 s, and “the test is successful if no collapse of the test voltage occurs.” The short-duration induced test follows it.

Lightning impulse. Table 1 of IS 2026 (Part 3) decides whether impulse testing is routine or type:

That split matters for NIEL’s range. Every distribution transformer at 11, 22 or 33 kV falls under the first row. There, impulse withstand is proven by a type test report, not on the unit delivered. Power transformers at 110, 132 or 220 kV primary fall under the second or third row, where impulse is routine on every unit.

Tap changer and accessories. Witness the on-load tap changer operating through its range. For the power transformer range, NIEL lists a Buchholz relay with alarm and trip contacts, an oil temperature indicator with alarm and trip contacts, and a magnetic oil level gauge. Confirm at works that each of these alarms and trips is wired to the marshalling box terminals shown on the drawing.

Which special tests are worth ordering as a baseline for site?

Order insulation resistance, tan delta and capacitance measurements at works on any transformer where site condition monitoring matters. On power transformers, also order a frequency response measurement. None of these has a pass mark under IS 2026 (Part 1). Their value is the comparison: a works fingerprint turns “looks acceptable” into “unchanged since the factory.”

Frequency response analysis (FRA, often called SFRA) is the clearest case. The IEC abstract of IEC 60076-18:2012, reproduced by iTeh Standards, says the standard covers frequency response measurement “either on-site or in the factory either when the test object is new or at a later stage.” A frequency response trace responds to the geometry of windings and core. A transformer that has shifted on its clamping during a rough road or sea journey can show normal ratio, resistance and insulation resistance at power frequency. The change appears only when the site trace is laid over the works trace, measured with the same connections.

For cast resin units, the equivalent baseline is partial discharge. NIEL’s comparison of cast resin and oil filled transformers explains that partial discharge measurement is a special test by agreement under IS 11171. The acceptance level has to be written into the order in picocoulombs.

On power transformers, zero-sequence impedance helps where earth fault settings depend on it, and sound level where the substation sits near offices or housing. Ask for both in the enquiry, not after dispatch.

What must be repeated on site before a transformer is energised?

Repeat every test that transport, handling, erection and oil filling can change. The Central Electricity Authority (Measures relating to Safety and Electric Supply) Regulations, 2023 state that “no new apparatus, cable or supply line of voltage exceeding 650 V shall be commissioned unless such apparatus, cable or supply line are subjected to site tests as per relevant standards.”

The same regulations require records of all tests to be “duly kept in such a way that these records can be compared with the past records.” They also let the Electrical Inspector require further tests before or after charging.

A practical site programme for an oil filled unit:

On auxiliary circuits, IS 2026 (Part 3) specifies a 1 minute, 2 kV AC test on auxiliary wiring at works. It notes that large transformers have their auxiliary equipment dismantled for shipment, so “after completion of erection on site, a 1 000 V megaohm meter test is recommended.” It adds that electronic equipment with a withstand voltage below 1000 V should be removed before that test.

Where the supply voltage exceeds the voltage notified by the state, the regulations require inspection and testing by the Electrical Inspector before supply starts. At or below the notified voltage, the owner self-certifies and submits the report.

Which works tests should not be repeated casually on site?

Do not treat the full separate source AC or induced withstand tests as routine site checks on a new transformer. IS 2026 (Part 3) states that repeating dielectric tests to prove that a new, factory-tested transformer continues to meet the standard “is always performed at 100 % of test level.” That is a works-level stress applied to a unit that has already passed it once.

The 80 percent level that engineers sometimes quote applies to a different case. IS 2026 (Part 3) allows dielectric tests at 80 percent of the original values for transformers “which have already been in service and have been refurbished or serviced,” provided the internal insulation has not been modified. Even then, the long-duration induced test with partial discharge measurement is always repeated at 100 percent.

On site, confirm insulation condition with insulation resistance, dissipation factor and oil tests, and compare against works baselines. Reserve full dielectric withstand tests for a formal acceptance agreement or for investigating a specific doubt. Agree those tests in writing with the manufacturer, because the test source, connections and pass criteria all need to match the works method.

What oil test results should a new transformer meet before energising?

New oil in a new transformer should meet the limits in Table 1 of IS 1866: 2000, the Indian code of practice for supervision of mineral insulating oil in equipment. Breakdown voltage and water content tighten as the equipment voltage rises, so a 220 kV power transformer faces stricter limits than an 11 kV distribution unit.

Two notes in the standard deserve attention. For transformers below 72.5 kV class, the maximum water content “should be agreed between supplier and user.” And a high dissipation factor “may indicate excessive contamination or the misapplication of solid materials used in manufacture and should be investigated.” In other words, a poor dissipation factor on new oil is a question for the manufacturer, not just the oil filtration contractor.

Record the oil temperature at sampling. Breakdown voltage is the headline figure, but water content says more about whether the active part picked up moisture during erection.

How do the tests differ between a distribution transformer and a power transformer?

The test list is similar; the depth changes with voltage and rating. A 33 kV distribution transformer is proven for impulse by type test and usually commissioned on insulation resistance, ratio, resistance, vector group and oil BDV. A 132 kV or 220 kV power transformer adds routine impulse testing at works, partial discharge measurement, frequency response baselines and fuller protection checks on site.

Both ranges carry an IS 2026 compliance statement on the NIEL product pages. The distribution range lists vector group Dyn11. The power range lists Dyn11, YNd11 and a star-star connection, which is exactly what the site vector group check has to confirm. For factory-assembled containerised secondary substations, the transformer, switchgear and interconnections are wired and tested together at works. That reduces the site programme but does not remove the CEA requirement for site tests.

Who carries out site testing, and what paperwork should be handed over?

The owner is responsible for site tests. The work is done by the supplier, the erection contractor or a testing agency under the owner’s control, and above the notified voltage the Electrical Inspector inspects and tests before supply starts. The handover pack should let every site reading be compared against the works test certificate for that serial number.

NIEL’s turnkey substation scope covers design, procurement, erection, testing and commissioning, and NIEL has erected more than 40 substations up to 230 kV class. That work is described in its guide to substations for industrial plants. Where the transformer sits inside a substation package, one party can own the sequence from works test to energisation. That closes the gap between a supplier who tested at works and an erection contractor who never saw the report.

The handover pack should contain:

• Works routine test certificate for the serial number delivered, signed by the witness

• Type test report for the design, with the rating it covers

• Special test results ordered as baselines: insulation resistance, tan delta, capacitance, FRA traces

• Site test records on the same basis and temperature as the works results

• Oil test reports before energisation

• Rating plate photograph, general arrangement and schematic drawings, and a spares list

What should a transformer test clause in the purchase order include?

Write the test programme into the order, because anything not agreed defaults to the standard’s minimum.

• Standard the transformer is built and tested to, with year: IS 2026, IS 1180 for distribution units, or IEC 60076

• Routine tests to be witnessed, with notice period for the inspection call

• Type test report to be supplied, and the unit and rating it covers

• Lightning impulse as a routine or special test on units with Um up to 72.5 kV, if required on the unit delivered

• Special tests as baselines: insulation resistance, tan delta, capacitance, FRA, sound level, zero-sequence impedance

• Partial discharge acceptance level in picocoulombs for cast resin units

• Guaranteed no-load and load loss figures with tolerance and reference temperature

• Oil test limits before energising, by reference to IS 1866 Table 1

• Site tests to be performed, by whom, and the comparison basis

• Protection device functional checks and the schematic they are checked against

• Handover documents and the format of test records

Planning the test programme for a new transformer?

NIEL supplies distribution transformers from 15 to 2500 kVA and power transformers from 2500 to 100,000 kVA at primary voltages up to 220 kV, and has erected more than 40 substations up to 230 kV class. Send us the rating, voltage class and site details, and our engineers will review the works witness and site test clauses before the order is placed.

Find the right transformer for your project:

• Distribution transformers: 15 to 2500 kVA, built to IS 2026.

• Power transformers: 2500 to 100,000 kVA at primary voltages up to 220 kV.

• Containerised secondary substations: transformer, switchgear and interconnections wired and tested together at works.

Call +91 80 42434343, email support@newindiaelectricals.com, or contact our team with your transformer details.

^{[1][2][3][4][5][6]}