Why negative dissipation factor physically impossible?
As we know, the dissipation factor (tan δ) is defined as the ratio of dielectric (active) losses to the reactive power of the insulation system, expressed as:
A negative dissipation factor implies that either the power losses (P) or the reactive power (Q) is negative. In such a case, as illustrated in the phasor diagram below, the insulation would appear to behave like a negative resistance, or there may be an inductive component present in the test circuit. This inductance could draw a relatively large inductive current compared to the capacitive current of the insulation, misleading the measurement.
In both scenarios, a negative tan δ value is physically meaningless and indicates an abnormal condition. Therefore, the test setup and the equipment under test should be thoroughly investigated to identify the root cause.
Below are some common causes that can lead to negative dissipation factor measurements in high-voltage equipment.
High leakage currents along bushing surfaces
One of the causes of negative dissipation factor measurements is surface leakage current along bushings and insulators. These surface currents are typically caused by the presence of moisture and contamination on the insulation surface and return through the test set’s measurement circuit.
As shown in the diagram below, this creates a parallel path to the main insulation, which can significantly increase the measured dissipation factor or even result in a negative value.
To prevent this, the insulation surface must be thoroughly clean and completely dry. It is also recommended to avoid performing dissipation factor tests during rainy or heavily overcast weather conditions.
In some cases, using a semi-conductive tape wrapped around the insulator can help redirect a portion of the surface current to ground. However, the most effective solution is to ensure the insulation surface is clean and dry prior to testing.
Windings not-short-circuited
One of the key considerations in transformer and bushing dissipation factor testing is short-circuiting the transformer windings, as shown in the diagram below.
If the transformer is not short-circuited, the presence of magnetizing inductance in the test circuit can lead to negative dissipation factor readings or cause abnormally high values at certain frequencies. For this reason, the windings must always be short-circuited during the dissipation factor test of the transformer and its bushings.
Additionally, when testing transformer bushings, it is recommended that—for example, if the HV bushings are under test—the LV and tertiary (TV) sides (if present) be grounded. This is because the bushings on the non-tested sides can become energized through inter-winding capacitances (CHL and CLG), posing a safety hazard to personnel.
Presence of an electrostatic shield
In high-voltage power transformers as well as transformers used in renewable energy power plants, an electrostatic shield is placed between the primary and secondary windings. This shield is connected to ground and prevents high-frequency overvoltages caused by lightning strikes and network switching on the high-voltage side from transferring to the low-voltage side.
These overvoltages are capacitively coupled from the high-voltage to the low-voltage side. By grounding the electrostatic shield, as shown in the diagrams below, the capacitive coupling between the high-voltage and low-voltage windings is effectively eliminated. Consequently, all overvoltages are diverted through the capacitance between the windings and ground, rather than passing directly from high-voltage to low-voltage.
If an electrostatic shield is used inside the transformer, it is indicated on the transformer nameplate as shown in the figure below with a dashed line.
In three-winding transformers, shielding effectively occurs during the testing of one of the capacitances between the windings. For example, consider the winding arrangement of a transformer as shown below. When measuring the capacitance between the LV and TV windings, the HV winding acts as a guard. This causes the HV winding to behave like an electrostatic shield.
In this case, the LT capacitance essentially consists of stray capacitances between bushings, the transformer structure, and other parts, amounting to several hundred picofarads. This can cause the dissipation factor to be measured as either negative or abnormally high.
In both scenarios—whether an electrostatic shield exists between two windings or in a three-winding transformer—there is no need to evaluate the stray capacitance condition between windings. Additionally, measurements taken at different times are not directly comparable because many factors influence these stray capacitances.
Electromagnetic Interference (EMI)
One of the main challenges in dissipation factor testing at energized high-voltage substations arises from electromagnetic induction. For example, when measuring the dissipation factor of a bushing, transformer, or CVT in a live substation, leakage currents induced by busbars and energized lines flow through the equipment terminals.
These leakage currents are primarily at the power frequency. Therefore, if the dissipation factor test is performed exactly at the network frequency (50 or 60 Hz), the results may sometimes be negative.
In such cases, the test should be conducted at frequencies slightly offset from the network frequency, such as 47 or 53 Hz (for a 50 Hz system) or 57 or 63 Hz (for a 60 Hz system). Additionally, the terminals of the equipment under test should be fully isolated as much as possible, and reliance solely on power switches or disconnectors should be avoided.
This is because the longer the conductors connected to the equipment terminals, the greater the induced voltage from nearby energized equipment becomes.
Wirings
As a final and very important recommendation, always carefully inspect the test circuit wiring and connection points. In dissipation factor testing, the metal body of the equipment, which is normally grounded, must be connected to the ground, and the test device must also be properly connected to it.
Any loose connection in the injection path, measurement path, or grounding can introduce unwanted inductance and resistance in the current path. This may cause an increase in the measured dissipation factor or even result in a negative value at certain frequencies.
Therefore, before making the test device connections, make sure to clean the connection points thoroughly with a wire brush to remove any paint, contamination, or oxidation.
