Tan Delta Test for Transformers: Dissipation Factor Guide
Insulation does not fail without warning. Long before a winding breaks down, its insulation quietly changes: moisture creeps in, oil ages, contamination builds up, and the tiny losses inside the insulation grow. The tan delta test — also called the dissipation factor or power factor test — is designed to catch exactly that slow deterioration. It measures how lossy the insulation has become, and because it is sensitive to moisture and ageing, it often moves long before other tests show anything wrong.
This guide explains tan delta testing in plain language: what the test measures, how it is set up, how to interpret the results, and why temperature matters so much that ignoring it can make good insulation look bad and bad insulation look good. As with everything on this site, always work to your project specification, the manufacturer's instructions and the applicable standards for your region; the general guidance here never overrides those documents. If you are new to transformer testing as a whole, start with our power transformer testing and commissioning guide, which puts tan delta in context alongside the other insulation tests.
What tan delta actually measures
Think of a transformer winding and its insulation as a capacitor: two conductors separated by an insulating material. If the insulation were perfect, applying an alternating voltage would draw only a capacitive current, exactly ninety degrees out of phase with the voltage, and no energy would be lost. Real insulation is never perfect. It always draws a small resistive current in phase with the voltage, representing energy lost as heat inside the insulation — losses caused by moisture, ionic contamination, ageing products and polar materials in the oil and paper.
Tan delta is the ratio of that resistive loss current to the capacitive current. In healthy, dry insulation the loss current is tiny compared with the capacitive current, so tan delta is small — typically well under one percent for a transformer winding at normal temperature, though the exact expectation depends on the insulation system, voltage class and the manufacturer's guidance. As insulation deteriorates, the loss current grows and tan delta rises. That is the whole idea in one sentence: tan delta turns the invisible condition of insulation into a single number that can be trended over the life of the transformer.
Tan delta, dissipation factor and power factor: the same idea
Engineers meet three names for what is essentially one measurement, and the confusion is understandable. Tan delta is the tangent of the loss angle — the small angle by which the total current departs from the ideal ninety-degree capacitive position. Dissipation factor is the same quantity, and is the term most test-set manufacturers use. Power factor, in this context, is the cosine of the angle between the applied voltage and the total current. For the small loss angles found in healthy insulation, tan delta and power factor are numerically almost identical, and many test sets display both. The practical point is simple: whichever term your report uses, you are looking at the same diagnostic — how much energy the insulation is wasting.
The test is normally performed with the transformer de-energised, isolated, earthed and discharged, following the site safety rules for high-voltage work. A test voltage, usually up to around ten kilovolts depending on the equipment and the winding under test, is applied between the winding and earth or between windings, and the test set measures the resulting current and its phase relationship. Because the test voltage is far below the operating voltage, the test does not stress the insulation; it simply interrogates it.
Test setup: windings, guard and grounding
A transformer has several insulation systems — high-voltage winding to earth, low-voltage winding to earth, and high-voltage to low-voltage winding — and tan delta is measured for each of them in turn. The standard technique uses a guarded measurement: the winding under test is energised, the winding being measured against is connected to the measuring circuit, and any other winding is connected to the guard terminal so that its capacitance does not contaminate the result. Getting the guard connections right is essential; a wrong guard arrangement measures the wrong insulation system and produces a meaningless number.
As with the insulation resistance test, cleanliness matters. Dirty, damp bushings and contaminated surfaces create leakage paths that add to the measured loss and inflate tan delta, so bushings should be clean and dry before testing. The tap changer position should be recorded, because it affects the winding configuration being measured, and the oil and winding temperatures should be recorded too — temperature correction is not optional, as the next section explains. Safety discipline is the same as for any high-voltage test: barriers, earthing, discharge after each measurement, and no one working on the transformer while voltage is applied.
Interpreting results: trends beat single numbers
A single tan delta value is hard to judge on its own, because acceptable values vary with insulation type, voltage class, age and temperature. What matters most is the trend and the comparison. A winding whose tan delta is stable year after year is almost certainly healthy, even if its absolute value is not the lowest in the fleet. A winding whose tan delta is rising steadily, or one phase that stands clearly apart from the other two, deserves attention even if its value has not yet crossed any published limit. And a sudden jump between two tests is a reason to investigate now, not at the next scheduled outage.
Comparison across the three phases is a powerful first check, since the phases of a healthy transformer should give similar results under the same conditions. Comparison with the factory or commissioning values is even better, and comparison with sister units of the same design can help when no baseline exists. Where tan delta is high, the pattern of results helps narrow the cause: high values on all insulation systems may point to general moisture or oil condition, which should be checked against oil test results, while a high value on one winding only points to a localised problem in that winding's insulation. Always interpret tan delta together with capacitance, because the two are measured simultaneously and a changing capacitance alongside a changing tan delta tells a different story from tan delta alone.
Temperature effects: the correction you must not skip
Tan delta rises with temperature, and the effect is large enough to mislead. A transformer tested on a hot summer afternoon can show a tan delta noticeably higher than the same transformer tested on a cold winter morning, with no change in insulation condition at all. That is why every tan delta result must be recorded with the insulation temperature at the time of test, and why results are corrected to a standard reference temperature — commonly twenty degrees Celsius — before comparison. Test sets and standards provide correction factors, and the correction method used should be stated in the report.
The practical consequences are worth spelling out. Never compare a summer result directly with a winter result without correction, and be suspicious of any "deterioration" that coincides with a hot test day. Conversely, a genuinely deteriorating winding can hide behind a cold test day if the raw numbers are compared without correction. Where possible, test at a similar temperature each time, or at least record the temperature carefully so the correction is reliable. Temperature discipline is what turns tan delta from a rough indicator into a precise trending tool.
The tip-up test and voltage dependence
One useful extension of tan delta testing is the tip-up test: measuring tan delta at two or more voltages and looking at how much it rises with voltage. In healthy insulation, tan delta changes little with applied voltage. In insulation containing voids or delaminated layers, the loss rises sharply once the voltage is high enough to start ionisation inside the voids, so the tan delta "tips up" with voltage. A large tip-up is therefore a warning of voiding or delamination, and it is one of the reasons tan delta is sometimes measured at several voltages rather than one.
Tip-up testing is not always performed, and it needs a test set capable of the required voltages and a clear procedure, but where it is used it adds diagnostic depth that a single-voltage test cannot give. It also connects naturally to partial discharge testing, since the voids that cause tip-up are also sources of partial discharge. If tan delta shows a strong tip-up, partial discharge investigation is a sensible next step.
When to test, and what to do with the results
Tan delta is typically measured at commissioning, to establish the baseline, and then periodically as part of a condition-assessment programme — often aligned with other insulation tests and oil sampling so that all the evidence is gathered at the same time. It is also run after events that could affect insulation, such as moisture ingress, a failed breather, or a long outage in humid conditions, and before returning a transformer to service after major work.
When results are concerning, the response should be proportionate and evidence-based. A rising trend calls for a shorter test interval, a check of oil moisture and acidity, and a review of the breather and sealing. A high value on one winding calls for corroboration from other tests and, where justified, further investigation before the next energisation decision. A stable, low value across all windings is a genuine vote of confidence in the insulation system. Record every result with its temperature, tap position and correction method in one file per transformer, alongside the other test records, and use our free transformer calculators for the load and rating checks that accompany any condition assessment. Done this way, tan delta becomes what it should be: an early warning system that gives you time to act before insulation failure forces the decision on you.
Frequently asked questions
What is a good tan delta value for a transformer?
There is no single universal limit: acceptable values depend on insulation type, voltage class, age and temperature. Healthy oil-paper insulation at reference temperature is typically well under one percent, but always judge against the manufacturer's guidance, the commissioning baseline and the trend, not against a number remembered from another transformer.
Can tan delta be measured on an energised transformer?
No. Tan delta testing is an offline test performed with the transformer de-energised, isolated, earthed and discharged. Online monitoring of dielectric losses exists on some large units, but it is a different technique from the standard offline test described here.
Why did tan delta rise if nothing changed?
The most common innocent cause is temperature: a hotter test day raises tan delta. Check the recorded temperature and the correction applied before concluding that the insulation has deteriorated, and repeat the test under comparable conditions if the result is surprising.
← Back to the transformer testing guide · SFRA testing → · Partial discharge testing → · CT and PT testing →