VA-TEK Engineering Tan delta is the one insulation test that produces a number you can watch move. A withstand test passes or fails; a ratio test compares two windings; tan delta tells you how much energy the insulation is losing every cycle, and it does so before the insulation has failed. The difficulty is that the number means little on its own. Two transformers with identical capacitance can differ by a factor of three in dissipation factor, and the same transformer can move by a factor of two simply because the test was run at a different temperature or with the guard connection left off.
This article explains what the measurement is, why the test conditions matter as much as the result, and how to turn a single reading into a trend that supports a decision.
Key takeaways
- Tan delta (dissipation factor) is the ratio of resistive to reactive current in the insulation — a loss measurement, expressed as a percentage.
- The value depends on temperature, applied voltage and test mode. A result is only comparable with another result taken under the same conditions.
- Capacitance comes from the same measurement and answers a different question: a change in capacitance points to a change in geometry, not in chemistry.
- One reading is a reference; the trend across three or more tests at the same temperature is the diagnostic instrument.
- UST, GST and GSTg modes exist to remove surface leakage from the measurement. Choosing the wrong one can dominate the result.
1. What the measurement actually is
An insulation system behaves approximately like a capacitor with a small resistor in parallel. The ideal capacitor draws current exactly 90° out of phase with the applied voltage and dissipates nothing. Real insulation draws a current that leads by slightly less than 90°, because a small resistive component is in phase with the voltage. The angle by which the current falls short of 90° is the loss angle δ, and its tangent is the ratio of the resistive to the reactive component.
That ratio is why tan delta is expressed as a percentage and why it is nearly identical to insulation power factor for the small angles involved. A dissipation factor of 0.5% means that half a percent of the energy the insulation stores each cycle is being converted to heat. In a large transformer, half a percent of the energy in a 50 or 60 Hz field is a meaningful amount of heat, and a rising trend means more of it.
The measurement is made at power frequency with an applied voltage, and the instrument reports voltage, current, power, tan delta, capacitance and, usually, power factor from the same acquisition. The method of measurement is described in IEC 60247 for insulating liquids and applied to solid insulation through the same bridge principle; field practice for transformers is set out in guides such as IEEE C57.152. The TDC-3100 performs this measurement with an internal source adjustable from 0.5 to 12 kV (2000 VA), which is enough to stress the insulation of most distribution and medium-power transformers at a level above the interference floor.
2. Three things that move the number before the insulation does
Temperature. Dissipation factor rises with temperature in most insulation systems, and the slope differs between oil-impregnated paper, resin and solid dielectrics. A result taken at 45 °C and compared with a factory value taken at 20 °C will look like a deterioration that has not happened. Either test at a stable temperature and compare like with like, or apply the correction factors appropriate to the insulation type — and record the temperature with the result, every time.
Test mode. The three common connections are UST (ungrounded specimen test), GST (grounded specimen test) and GSTg (grounded specimen test with guard). UST measures only the insulation between the energised terminal and the measuring terminal, which is why it is preferred when the bushing or the test lead itself could contribute leakage. GST measures the insulation to ground including surface paths, and GSTg adds a guard connection that diverts surface leakage away from the measurement. A dirty bushing tested in GST mode can read several times higher than the same bushing in GSTg mode, and the difference is contamination, not degradation.
Voltage. Some defects only appear above a threshold, and some insulation shows a tip-up — a dissipation factor that rises with applied voltage. Testing at more than one voltage level, when the source allows it, separates a linear insulation system from one that is starting to discharge.
3. What a single result can and cannot tell you
A tan delta value on its own is a comparison waiting to happen. It can be compared with a specification limit, with the factory test report, or with the same unit measured last year. Of the three, the factory report is the most misused: it was taken at a different temperature, often with a different bridge, and sometimes on a different tap configuration.
| Evidence | What it supports | What it cannot support |
|---|---|---|
| Single tan delta value | A pass/fail against a stated limit, under stated conditions | Any conclusion about trend or rate of change |
| Tan delta trend at constant temperature | Deterioration of the insulation system over time | Locating the defect |
| Capacitance comparison | Physical movement, bushing contamination, winding deformation | Chemical ageing of the paper |
| Phase-to-phase comparison | A winding that behaves differently from its sisters | Absolute condition of the healthy phases |
| Tip-up test at two voltages | Whether the insulation starts to discharge under stress | The location or type of the discharge |
Capacitance deserves its own line. It comes out of the same measurement and, for a given winding geometry, is a stable quantity. A capacitance that has shifted beyond the repeatability of the instrument points to something physical: a bushing that has absorbed moisture, a winding that has moved, a connection that has changed. Reading capacitance and tan delta together separates a chemical problem from a mechanical one.
4. Where it sits in the commissioning sequence
Insulation power factor is measured after the insulation resistance and before the applied-voltage tests, because it is a low-stress measurement that establishes a baseline without stressing the insulation. In a typical transformer commissioning sequence it follows winding resistance and turns ratio, which confirm that the windings are intact and connected as designed, and it precedes the applied potential and induced potential tests. The order and the reasoning behind it are set out in transformer commissioning test sequence.
The same instrument covers bushings, breakers, surge arresters, CVTs and rotating machines, so the test is usually run as a set rather than on a single component. Where the object under test is a cable, the measurement at power frequency is not the usual route; cable insulation is evaluated with VLF tan delta at 0.1 Hz, which is a different measurement with different criteria and is covered in VLF tan delta testing.
For the wider programme, tan delta is one of five core electrical checks on a transformer, alongside winding resistance, turns ratio, insulation resistance and the applied-voltage tests. The transformer electrical testing guide explains how the results are read together, because no single test in the set is conclusive on its own.
5. Common questions
What is a normal tan delta value for a power transformer?
For oil-immersed distribution and medium-power transformers, insulation power factor values well below one percent at 20 °C are typical of a healthy winding, and values above one percent usually prompt investigation. The exact acceptance limit belongs to the specification or the utility’s standard, and a value that has doubled since the last test matters more than a value that has always been slightly high.
Can tan delta testing damage the insulation?
The test is performed at a fraction of the operating voltage and for a short duration, so it is not a stress test. It is normally run before the applied-voltage tests in a commissioning sequence, precisely because it is a low-stress measurement.
Why do two tests of the same transformer give different results?
Temperature, test mode, lead arrangement and the interference level at the substation all influence the reading. Recording temperature and connection mode with every result is what makes the next comparison meaningful; the TDC-3100 stores 200 records on board for that reason, and its interference suppression is intended for energised substation environments.
6. Getting a baseline you can trust
The first field test of a transformer has no history behind it, so it becomes the baseline for everything that follows. That makes the conditions of the first test worth recording in more detail than seems necessary: oil temperature, ambient temperature, test mode, applied voltage, capacitance, and the phase or terminal measured. A test performed in a hurry on a cold morning becomes next year’s reference whether or not it was representative.
The TDC-3100 covers the measurement range needed for transformers, bushings, breakers and CVTs, with UST, GST and GSTg modes, capacitance measurement to 60 nF at 12 kV, and a CVT ratio function. Its dissipation-factor accuracy is stated as ±(reading × 1% + 0.040%), which sets the smallest change worth interpreting. Specifications for the instrument, including the test modes and the source rating, are on the TDC-3100 product page.
If you are building a commissioning or maintenance programme and need to know which measurements apply to a specific transformer fleet, send the nameplate details and the test standard you work to through the contact form.
