Analysis of Factors Affecting Dielectric Loss Factor in Electrical Insulation Systems

The dielectric loss factor (tan δ) is one of the most widely used parameters for evaluating the insulation condition of high-voltage equipment such as transformers, bushings, cables, and rotating machines. This article presents a systematic analysis of the key factors that influence dielectric loss factor measurements, including temperature, moisture, frequency, test voltage, insulation aging, contamination, and measurement conditions. A clear understanding of these influencing factors is essential for obtaining reliable test data and making correct insulation diagnostics.

Technical Overview

Definition and Physical Significance of Dielectric Loss Factor

The dielectric loss factor, also known as the dissipation factor or loss tangent (tan δ), is defined as the ratio of the resistive (loss) current component to the capacitive current component that flows through a dielectric when an alternating voltage is applied. In an ideal loss-free dielectric, the current leads the voltage by exactly 90 degrees; in a real dielectric, conduction and polarization losses reduce this angle by the dielectric loss angle δ. The value of tan δ therefore quantifies the fraction of stored electrical energy that is dissipated as heat in each cycle. Because tan δ is a ratio rather than an absolute quantity, it is largely independent of insulation geometry, which makes it an effective global indicator of the overall condition of an insulation system.

Components of Dielectric Loss

Total dielectric loss under AC stress consists of three main components:

• Conduction loss, caused by leakage current flowing through the volume of the dielectric and across its surfaces.
• Polarization loss, caused by the lag of dipole orientation and interfacial polarization behind the alternating electric field.
• Partial discharge loss, generated by ionization in gas-filled voids and interfaces when the local field stress exceeds the inception value.

Summary of the Main Influencing Factors

The measured value of tan δ is not a fixed material constant; it varies significantly with operating and testing conditions. The principal factors are summarized below and analyzed individually in the following sections.

• Temperature: the most significant external factor, especially for moist insulation.
• Moisture: the most damaging internal contaminant for oil-paper and resin insulation.
• Frequency: different polarization mechanisms respond at different frequencies.
• Test voltage: voltage dependence reveals partial-discharge-active defects.
• Aging and deterioration: thermal and oxidative aging byproducts increase losses.
• Contamination: conductive particles, dust, salts, and surface films.
• Measurement conditions: humidity, interference, stray capacitance, and grounding quality.

Effect of Temperature

For most practical insulating systems, tan δ increases as temperature rises, but the shape of the temperature characteristic depends strongly on moisture content. Dry, well-processed oil-paper insulation shows a gentle and predictable increase with temperature, whereas damp insulation exhibits a steep rise because water sharply increases ionic conduction and interfacial polarization. For this reason, comparing tan δ measured at two different temperatures, for example before and after loading a transformer, is a classical method for detecting moisture. All field results should be corrected to the reference temperature of 20 °C using the correction tables provided in applicable standards, and trend analysis should always compare values at equivalent temperatures.

Effect of Moisture and Humidity

Water is the single most harmful contaminant in electrical insulation. Even a small increase in the moisture content of cellulose paper, for example from 1% to 2%, can multiply tan δ by several times. Moisture raises both the conduction loss and the interfacial polarization loss, and it migrates between the oil and the solid paper as temperature changes, so the measured tan δ of a transformer depends on its recent load history. High ambient humidity during a field test can also create surface leakage films on bushings and cable terminations; guard circuits and clean, dry surfaces are required to prevent false high readings.

Effect of Frequency

Each polarization mechanism has a characteristic frequency response. At power frequency of 50 or 60 Hz, dipolar and interfacial polarization dominate in composite oil-paper systems, while ionic conduction becomes increasingly important at low frequency. This frequency dependence is deliberately exploited in frequency domain spectroscopy (FDS), in which tan δ is measured from millihertz to kilohertz in order to quantify the moisture content of solid insulation. Comparison of results is only valid when the test frequency is identical, so routine measurements are standardized at power frequency.

Effect of Test Voltage

Healthy, homogeneous insulation exhibits a tan δ that is essentially constant as the test voltage increases. If gas voids, delamination, or severe contamination are present, partial discharges ignite above an inception voltage and the loss rises with voltage. Recording tan δ at several voltage levels, for example 10%, 50%, and 100% of rated voltage, produces the well-known tip-up characteristic (Δtan δ), which is a standard diagnostic for stator winding insulation and high-voltage cables. A significant positive tip-up indicates partial-discharge-active defects even when the absolute tan δ value remains acceptable.

Effect of Insulation Aging

Thermal aging of cellulose depolymerizes the molecular chains and generates water, organic acids, and other polar byproducts, all of which increase conduction and polarization losses. Oxidative aging of mineral insulating oil produces acids, sludge, and dissolved polar compounds that raise the oil dissipation factor and, in turn, the loss of the complete oil-paper system. As equipment approaches the end of its thermal life, a steadily rising tan δ trend is often the earliest measurable symptom of irreversible deterioration.

Effect of Contamination

Conductive particles suspended in insulating oil, carbon deposits, metallic debris from tap changer operation, and salt or industrial pollution deposited on external surfaces all increase dielectric losses. Contamination typically produces an abnormal tan δ even when the capacitance remains unchanged, which is why capacitance and tan δ should always be evaluated together.

Effect of Measurement Conditions

Several test-side factors can distort results: electromagnetic interference from nearby energized equipment, inadequate grounding, high contact resistance at test connections, stray capacitance of test leads, and surface leakage in humid weather. Modern instruments apply anti-interference algorithms, shielding, and guard techniques, and verification measurements should be repeated whenever the environment changes.

Testing Samples

Power Transformers and Reactors

The composite oil-paper insulation of transformer windings is the most common test object. Measurements are performed between each winding and ground and between winding pairs to assess the overall insulation system.

Insulating Liquids

Mineral insulating oils, natural esters, and synthetic esters are tested in standardized cells to evaluate oxidation, contamination, and moisture-related degradation.

Bushings

Oil-impregnated paper, resin-impregnated paper, and resin-bonded paper bushings are tested between the main conductor and the measuring tap, which is one of the most sensitive ways to detect internal bushing deterioration.

High-Voltage Cables and Accessories

Cross-linked polyethylene, ethylene propylene rubber, mass-impregnated, and oil-filled cables, together with their joints and terminations, are tested to detect water trees, voids, and poor workmanship.

Rotating Machine Windings

Stator bars and complete windings insulated with epoxy-mica systems are tested with tip-up measurements to reveal voids and delamination caused by thermal cycling and loose wedging.

Solid Insulating Materials

Films, laminates, pressboard, and composite boards are tested during product development and quality control to characterize the loss behavior of the material itself.

Testing Items

A dielectric loss test program typically includes the following measurable quantities:

• Capacitance Cx and tan δ of the main insulation at power frequency, normally at 10 kV and corrected to 20 °C.
• Insulation power factor in percent, which is numerically equivalent to tan δ for small loss angles.
• Tan δ tip-up (Δtan δ) between specified voltage levels for machines and cables.
• Temperature characteristic of tan δ, used to evaluate moisture sensitivity.
• Dissipation factor of insulating liquids measured at 90 °C in accordance with IEC 60247.
• Trend comparison of all values against factory data and previous routine results.

Testing Methods

Schering Bridge Method

The classical Schering bridge balances the test object against a loss-free reference capacitor, achieving very high accuracy with the aid of a guard circuit that diverts surface leakage away from the measurement. It remains a reference method in laboratories and calibration facilities.

Digital Bridge and Automated Testers

Modern portable instruments use ratio-arm or fully digital measurement with automatic balancing, built-in standard capacitors, frequency tracking, and interference suppression, allowing reliable measurements even in energized substations.

Measurement Configurations

The grounded specimen test (GST) measures the total insulation between the energized part and ground, while the ungrounded specimen test (UST) isolates a specific insulation section, such as the gap between two windings, for individual assessment. Combining both configurations allows the losses of each insulation barrier to be separated.

Liquid Insulation Testing

Insulating liquids are measured in a three-terminal guarded cell heated to 90 °C in a temperature-controlled bath, with the test voltage applied across a defined electrode gap in accordance with IEC 60247. Strict cleaning and drying of the cell are essential because trace contamination dramatically affects the result.

Temperature Correction and Trending

Field readings are converted to the 20 °C reference using standard correction factors, and conclusions are drawn primarily from the trend of corrected values over the service life rather than from a single absolute number.

Applications

Analysis of the dielectric loss factor and its influencing factors supports a wide range of engineering activities:

• Factory routine and type testing of transformers, bushings, cables, and machine windings.
• Commissioning baseline measurements that anchor all future trend analysis.
• Condition-based and predictive maintenance programs for aging fleets of high-voltage assets.
• Moisture assessment of transformer solid insulation to justify drying or oil treatment.
• Detection of oil oxidation, contamination, and partial-discharge-active defects.
• Research, development, and incoming quality control of insulating materials and liquids.
• End-of-life evaluation and asset management decisions for critical equipment.

FAQ

What is an acceptable value of dielectric loss factor?

Limits depend on equipment type, voltage class, and applicable standards. As typical guidance, new transformer insulation normally shows tan δ below 0.5% at 20 °C, new bushings below approximately 0.7%, and new mineral oil below 0.5% at 90 °C. A significant increase relative to previous results is generally more meaningful than the absolute value alone.

Why does moisture increase tan δ so strongly?

Water is a highly polar substance that raises ionic conductivity and intensifies interfacial polarization, so it increases the loss current component directly while the capacitive component barely changes.

At what temperature should the measurement be performed?

Solid insulation tests are corrected to a reference of 20 °C, while insulating liquids are standardized at 90 °C. Measuring at two different temperatures provides additional diagnostic information about moisture.

Can a high tan δ identify the exact location of a defect?

No. Tan δ is a global indicator of the whole insulation volume and cannot localize a defect. However, the voltage tip-up test can confirm the presence of partial-discharge-active voids, and UST configurations can narrow the problem to a specific insulation barrier.

Why do repeated field measurements sometimes differ?

Differences in winding temperature, ambient humidity, interference levels, and moisture migration between oil and paper are the most common causes. Temperature correction, shielding, and repetition under stable conditions improve repeatability.

How often should dielectric loss tests be carried out?

For most high-voltage assets, an annual or biennial measurement combined with trending is sufficient. More frequent testing is recommended after unusual events such as overloads, nearby short circuits, or ingress of water.