Self-study notes. This page is static reading. It does not ask you to sign in. It does not certify a product, a factory, or a person. Use the edition of each standard named in the contract. Questions: hello@transformerpath.com.
Module 3 ends when the unit is in service and the first records are filed. This module is how those records are used after something changes: a gas alarm, a through-fault, a strange noise, or a trace that no longer matches. The decision at the end is one of three practical outcomes. Return the unit to service. Repair it where it stands. Or return it to a factory or a workshop that can dry it and test it again. The notes below are a way to assemble the evidence. They are not a rule that replaces the procedure or the manufacturer, with the records in front of them.
One dissolved-gas sample is not a diagnosis
Dissolved-gas analysis measures gases dissolved in the insulating liquid. IEC 60599 is the IEC guide to interpreting dissolved and free gases. IEEE C57.104 is the IEEE guide for gases generated in mineral-oil-immersed transformers. Both expect a history. A single report, read against a limit remembered from an older edition, is not the method in either guide.
Identify the sample before you read the numbers. Serial number, date, sampling point, whether the unit was on load, and whether a known event came first: a through-fault, a tap-changer operation, a trip, or oil handling. A sample from the wrong valve answers a different question. Free gas from a gas-operated relay is not the same specimen as gas dissolved in the bulk oil. An on-load tap changer that interrupts current in oil, in its own compartment, produces gases that must not be read as if they came from the main tank. Confirm which compartment was sampled.
The qualitative picture in both guides is stable enough to learn, and too coarse to name a failed part. Hydrogen is associated with partial discharge and with other low-energy electrical activity, and it can also rise from stray gassing that is not an active fault. Methane and ethane are associated with lower-temperature thermal faults in the oil. Ethylene becomes more prominent as the temperature of a thermal fault rises. Acetylene is associated with arcing and with temperatures high enough to take seriously even at a low concentration. Carbon monoxide and carbon dioxide are associated with cellulose, not only with a fault in the oil, and they also rise with normal aging. Oxygen and nitrogen say something about leaks and about how the sample was handled. They are not fault gases in the same sense.
Acetylene means a high-temperature event has occurred in the liquid that was sampled. It does not, by itself, name the lead, the changer, or the winding. IEEE C57.104 was revised so that interpretation uses status, rates of change, and the combination of gases, rather than one table of parts per million applied to every unit. IEC 60599 uses ratios and typical values, and it says when a ratio is meaningful. Do not apply a ratio when the concentrations are too low for it to mean anything. Do not copy a parts-per-million threshold from an old slide into a specification. If a contractual attention level is required, cite the edition and the table or the status definition in that edition. Levels have changed between editions.
A result that is stable over several samples is a different situation from the same number reached in a week. A unit just energized after a dry-out can show a rise that then flattens. A unit that shows acetylene after a relay trip needs the electrical investigation as well as another sample. The second sample checks that the first was not contaminated, and it starts the rate. A satisfactory gas result does not measure moisture, acidity, or breakdown voltage. Those are IEC 60422 and IEEE C57.106. It does not measure winding movement. That is the frequency-response comparison. Natural and synthetic esters are not mineral oil. Do not paste a mineral-oil limit onto an ester report.
Compare the new trace with the baseline
The baseline is the factory trace from Module 2, confirmed on site in Module 3, taken to IEC 60076-18. The new trace has to use the same connections, the same tap, and a comparable liquid fill. IEEE C57.149 is the interpretation guide. A difference is investigated. It is not, by itself, a decision to scrap the unit, and a trace that still overlays the baseline is not, by itself, a decision to energize.
Read the frequency bands in the guide you named rather than from memory. The low-frequency end of an open-circuit trace is sensitive to the core. The middle of the trace is where bulk winding movement usually appears. The high-frequency end is sensitive to leads, bushing connections, and tap leads, and also to how the test leads were draped. Repeat the setup before you repeat the alarm. Ratio, winding resistance, and excitation current answer different questions and belong in the same file. A shorted turn moves ratio. A poor contact moves resistance. Bulk movement can move the frequency-response trace while ratio stays inside tolerance.
After a through-fault, protection records say how large the current was and how long it lasted. IEC 60076-5 is about withstand at the design stage. It is not a field tool for deciding that this particular fault was harmless. If the current was within the nameplate withstand and the traces, the ratio, and the resistance match the baseline, the usual course is to return the unit to service and to plan the next oil sample. If the traces have changed in the winding band, the usual course is to keep the unit isolated and investigate. Those are usual courses. They are not a requirement that overrides the procedure.
Three outcomes, and the evidence each one needs
Re-energizing is the outcome when the identity is certain, the new dissolved-gas sample is explained by the history, acetylene is absent or is explained by the compartment that was actually sampled, the frequency-response overlay matches the baseline in the bands the guide treats as the winding, ratio and resistance match, and the protective devices that tripped have been proved. Record the tap, the temperatures, and the time, as in Module 3. Plan the next sample close enough to see whether the rate flattens.
A repair where the unit stands is the outcome when the evidence points at a part that can be replaced or repaired without opening the winding assembly: a bushing whose capacitance has moved, a cooler valve, a current transformer, a tap-changer mechanism, or a lead that the resistance test and the high-frequency end of the trace both point at. The repair is confirmed by repeating the tests that found it. A bushing change is confirmed by a new capacitance reading and a new frequency-response trace, because the bushing is part of the circuit that was measured. Do not return the unit to service on the strength of the replacement alone.
A return to a factory, or to a workshop that can dry the insulation and repeat dielectric tests, is the outcome when the evidence points inside the winding assembly and the site cannot restore it. Examples are acetylene in a sample that is confirmed to be from the main tank, a frequency-response change in the winding band that a repeated measurement still shows, a ratio error, or a reactance change of the kind IEC 60076-5 treats as damage, together with a fault current that was large. Internal inspection, vapour-phase or other drying, and a new set of factory tests are then the work. A site crew that cannot do those tests should not be asked to "try voltage and see."
Online gas monitors are useful for rate. They are not automatically equivalent to a laboratory chromatograph. When a monitor alarms, the usual next step is a laboratory sample from a known point, read with the history. Keep the reports. A file with several samples is interpretable. A file with one highlighted cell is not.
Check your reading
- Why is a single dissolved-gas report not a diagnosis under IEC 60599 or IEEE C57.104?
- What must you confirm before you treat acetylene as a main-tank event?
- What has to match before a new frequency-response trace can be compared with the baseline?
- Which result can stay normal while the winding has still moved?
- What extra check follows a bushing replacement before the unit is energized?
Answers. Both guides interpret history, rates, and the combination of gases. One row of numbers has no rate. Acetylene has to be tied to the compartment that was sampled, because an oil-switching tap changer in its own compartment produces gas that is not the main tank. The new trace needs the same connections, the same tap, and a comparable fill, measured so that IEC 60076-18 and IEEE C57.149 can be applied. Voltage ratio can remain inside tolerance when the frequency-response trace has changed, because ratio depends on turns. After a bushing change, measure the new capacitance and take a new frequency-response trace. The old overlay no longer describes the circuit.
Continue
- Module 3 — Site checks and energization
- How to read a dissolved-gas result
- Frequency-response traces and winding movement
- Site tests before first energization
- Engineer Track