A turns ratio out of tolerance almost always has a simple explanation: the tap changer is on a different tap than assumed, the test set was connected or configured with a different vector group, or the result was compared against the wrong nameplate value. When the error is on a single phase and that phase's excitation current is high, the probable cause is shorted turns, and the transformer must not be energized until that is ruled out.
Symptom
The turns ratio test (TTR) of a transformer gives values out of tolerance on one or all phases, on one tap or several. It can happen at receipt of new equipment, at commissioning, during maintenance or after a trip, when you must decide whether the transformer can be re-energized.
The turns ratio depends on the number of turns of each winding and on the tap in service. A real change in ratio means the number of effective turns changed, for example because some are shorted, or that the tap is not the one you think. But most out-of-tolerance results come from the method: connection, test set configuration or reference value.
Possible causes, from the most to the least likely
| Cause | Why it happens | What gives it away |
|---|---|---|
| Tap different from the assumed one | The off-circuit or on-load tap changer is in another position, or its indicator does not match | The measured value matches the ratio of another tap |
| Test connection or configuration | Vector group set wrongly, wrong terminals on a delta winding, crossed leads | Error on all three phases, often close to a factor of 1.73; inconsistent phase angle |
| Wrong reference value | Comparison with the nameplate of another tap or another unit | The error disappears when the nameplate is checked |
| Tap changer contact on one phase only | Selector misalignment or a high-resistance contact | One phase differs on some taps, not all; winding resistance altered on those taps |
| Shorted turns | Turn-to-turn insulation damage from a fault, overvoltage or aging | One phase out of tolerance with a high excitation current |
| Open winding or broken internal connection | Mechanical or thermal damage | Unstable or impossible reading; infinite winding resistance |
What to check first
- Real tap changer position at the mechanism, not just on the remote indicator, and its match with the nameplate. See on-load tap changer.
- Test set configuration: vector group, nominal ratio entered and terminals connected on each phase.
- Reference value: ratio calculated from the nameplate for that tap.
- Error pattern: all phases equally suggests method or tap; a single phase suggests a transformer problem.
- Excitation current of each phase, which many TTR sets measure together with the ratio. In a three-phase core-type transformer, the center phase usually has a lower excitation current than the outer phases; what is abnormal is a phase that departs from that pattern or from its history. See excitation current test.
- Is the error the same on all three phases?
- Yes: Does the value match the ratio of another tap?
- Yes: Tap different from assumed: verify the real tap changer position
- No: Connection, vector group or reference value error
- No: Is the excitation current of that phase high?
- Yes: Possible shorted turns: do not energize until confirmed
- No: Tap changer contact on that phase: compare the ratio on all taps
- Yes: Does the value match the ratio of another tap?
How it is confirmed
- Repeat the test with the checked configuration and on all taps: a tap changer problem shows up on some taps; a winding problem on all of them. Also check the phase angle the set reports: it must match the nameplate vector group.
- Winding resistance per phase and per tap: high-resistance contacts and open windings show up clearly. See winding resistance test.
- Excitation current per phase: shorted turns raise it on the affected phase.
- Dynamic tap changer test when the problem follows certain taps. See dynamic resistance test of the tap changer.
- Frequency response analysis and dissolved gas analysis, if there was a fault or a through-fault. See frequency response analysis test.
How to keep it from happening again
- Record the tap in every test and verify its position at the mechanism.
- Use the same configuration and the same reference as in the factory test, so the comparison is direct.
- Measure the ratio on all taps at receipt and commissioning, as a reference for the future.
- Maintain the tap changer according to the number of operations and the time the manufacturer indicates.
What not to do
- Do not energize a transformer with one phase out of tolerance and a high excitation current without investigating: shorted turns can develop into an internal fault.
- Do not move an off-circuit tap changer with the transformer energized.
- Do not conclude there is damage without first checking the test set configuration and the nameplate value.
- Do not compare results from different test sets without verifying that they use the same ratio convention.
Which tolerance applies, and to what
The tolerance reference is always the nameplate ratio for the tap being measured. IEC 60076-1 sets it on the principal tapping as the lesser of ±0.5 % of the declared ratio and one tenth of the measured short-circuit impedance percentage, and gives its own tolerance for the other tappings. In IEEE practice ±0.5 % of the calculated nominal ratio is used; IEEE C57.12.90 requires the ratio to be determined on all taps and for the full winding. A contract or specification may demand a tighter tolerance than the standard.
The tolerance applies to the voltage ratio, not to the phase angle, which is verified separately against the nameplate vector group. A correct ratio with an incorrect angle points to test or transformer connections different from those assumed.
How the patterns are read
| What the test shows | Probable reading |
|---|---|
| Same error on all three phases and on all taps | Reference, vector group or test set connection |
| Same error on all three phases on some taps only | Tap changer position different from the indicated one, or nameplate error |
| One phase out, on all taps | Problem in the winding of that phase: shorted turns or internal connection |
| One phase out on some taps only | Tap changer contact or selector on that phase |
| Unstable or impossible reading on one phase | Open winding or a poor contact |
A short circuit between a few turns changes the ratio only by a small fraction, which can stay within tolerance. That is why the ratio is never interpreted alone: the excitation current, the winding resistance and, after an event, the FRA complete the picture. A transformer with every ratio within tolerance but an excitation current outside the pattern of its phases is still suspect.
Quirks that cause false alarms
- Tap changers with a reversing switch or with coarse and fine selectors. Some positions share the ratio with others, so two different taps give the same value. Know the tap changer diagram before interpreting the tap list.
- Tertiary. If the transformer has a tertiary winding, ratios are measured by pairs of windings, and the winding not being measured must be left with no external connection so it does not distort the result.
- Delta or zigzag windings. The line voltage ratio and the turns ratio differ by factors that depend on the connection; it is the usual source of the factor close to 1.73 mentioned in this guide’s FAQ.
- Core excitation. A core magnetized by an earlier DC test, such as winding resistance, can give a high excitation and an unstable ratio. When a test set detects an anomaly on a phase, repeat the reading before concluding.
Illustrative example
After receipt of a power transformer, the ratio comes out of tolerance on all three phases on a few adjacent taps and correct on the rest. The pattern (all three phases, only some taps) points to the tap changer and not to the winding: when the real position was checked at the mechanism, the remote indicator was one position off from the mechanical position. The ratio matched the adjacent tap. There was no damage; the difference was in the indicator. Without the ratio measured on all taps, that finding would have reached operation.
When to call us
If your transformer’s ratio came out of tolerance and you need to know whether it is the method, the tap changer or the winding, we carry out power transformer tests. They include ratio on all taps, excitation current, winding resistance, dynamic tap changer test and FRA (frequency response analysis), with an interpretation of the whole set. Field testing is done in Colombia, or elsewhere on request. Contact us.
Related guides
- Turns ratio test
- Excitation current test
- Winding resistance test
- On-load tap changer
- Dynamic resistance test of the tap changer
- Frequency response analysis test
- Transformer with abnormal noise or vibration
References
- IEC 60076-1, Power transformers – Part 1: General.
- IEEE C57.12.00, Standard for General Requirements for Liquid-Immersed Distribution, Power, and Regulating Transformers.
- IEEE C57.12.90, Standard Test Code for Liquid-Immersed Distribution, Power, and Regulating Transformers.
- IEEE C57.152, Guide for Diagnostic Field Testing of Fluid-Filled Power Transformers, Regulators, and Reactors.
Frequently asked questions
What is the tolerance of the transformer turns ratio?
IEC 60076-1 sets, on the principal tapping, the lesser of ±0.5 % of the declared ratio and ±1/10 of the measured short-circuit impedance percentage; for the other tappings the tolerance is agreed. IEEE C57.12.00 requires the voltages to be within 0.5 % of the nameplate values.
What does it mean when only one phase has a ratio out of tolerance?
Something distinguishes that phase: shorted turns, a tap changer contact in another position or with high resistance, or a different test connection on that phase. If its excitation current is also high compared with the other phases, shorted turns are the most probable cause.
Why does the measured ratio differ by a factor close to 1.73?
Because a phase voltage ratio was compared with a line voltage ratio. This happens when the test set is configured with a different vector group or the wrong terminals are used on a delta winding. It is a method error, not a transformer fault.