Insulation of power transformers is a complex systems, comprised of various elements in terms of functionality and design.
Transformer insulation is generally divided into two main groups:

  • Internal insulation
  • External insulation

External insulation is, for example, tap cover insulation in direct contact with the atmosphere, air insulation between taps of a give coil, or between taps of different coils and earthed parts.

Internal insulation (oil, gas or mold) is divided into main and axial coil insulation, tap assembly insulation, switch insulation etc.

Main coil insulation is the insulation stretching from the coil to the grounded parts of the magnetic conductor, the tank or other coils (including other phases).

Longitudinal insulation is the insulation between different points of a single coil: between windings, layers and cores.

The internal oil filled insulation uses the following:

  • Solid (normally cellulose) insulation. This is the insulation between insulated conductors in immediate vicinity, coils and taps.
  • Purely oil insulation: in certain cases these are the gaps between the coils and the tank, inlet screen and the tank, a tap and the tank.
  • Combined (oil-barrier) insulation: oil gaps separated with barriers – intercoil insulation, interphase insulation, the insulation between the coil and the core, etc.

In the process of operation, the transformer insulation endures operational voltage for unlimited amounts of time, as well as brief surges: thunderstorm surges (impulses of several to several dozen microseconds), commutation surges (impulses with extended attenuation, lasting up to several thousand microseconds) and quasi-stationary surges (an increase of the operating frequency several hours long). Coordination of the internal transformer insulation must facilitate electric strength in all of these cases.

The insulation tested for compliance to the requirements of coordination by a series of high voltage tests, which includes the following:

  • A one minute test of industrial frequency;
  • A long term (lasting one hour) test of industrial frequency while measuring the intensity of partial discharges equal to 130 – 150% of operating voltage;
  • Commutation impulse with a front of at least 100 microseconds and duration of at least 1000 microseconds;
  • A full thunderstorm impulse with a front of 1.2 microseconds and duration of 50 microseconds;
  • A short 2 – 3 microsecond thunderstorm impulse.

Previously, when selecting overload limits for standardization, the primary values defining the gaps within the internal transformer insulation were the values of short test overload surges: thunderstorm commutation impulses and one minute of industrial frequency voltage. However, with developments in the area of extremely high voltages, a possibility emerged to deeply limit overloads in traditional voltage class transformers, therefore significantly limiting the difference of voltages of test surges versus normal operating voltage. That means that normal operating voltage of the transformers may become the decisive factor when selecting the size of the transformer’s internal insulation.

Most of the components in this type of insulation utilize combined insulation system, comprised of liquid (mineral transformer oil) and solid (cellulose) phases. As this combined insulation is subjected to industrial frequency voltage and impulses, the intensity in the oil channels is higher than that in solid insulation due to lower dielectric capacity of the oil (2.3 and 4.5); the dielectric strength of the oil is also lower than that of solid insulation. Therefore, the dielectric strength of most components of transformer insulation is defined by the dielectric strength of the oil channel with the highest load.

The main task when creating a method of calculating the electric insulation system is sound selection of the criteria to define its dielectric strength. As far as composite transformer insulation is concerned, a criteria of defining the dielectric strength of the oil needs to be selected.

At this time there is no quantitative physical theory of electric breakdown in pure transformer oil, therefore, it is not possible to theoretically prove the criteria of electric strength of oil gaps.

Therefore, when selecting the criteria of transformer oil electric strength, experimental electric data should be used, with consideration of primary factors influencing occurrence and development of insulation breakdown process.

There are a number of physical theories which rely on macroscopic mechanisms for the explanation of electric breakdowns in liquids. This group of theories is based on experiments with technically pure liquids with relatively high conductivity currents and extended voltage application. Such conditions hold true for most of industrial liquid dielectric applications, e.g. in transformers. Results of such studies are taken into consideration when selecting the electric strength criteria. The initial physical processes in the macroscopic mechanism of electric breakdown facilitate discharge development. The most significant aspects of the problem are the sources of free electrons in the liquid, the process of their solvation and movement, the energy connection to the environment, the influence of molecular structure, conditions of formation and proliferation of electron avalanches and the amount of energy created when electrons move. The development of the processes to the level when they can be detected and measured sets the point of their transformation from elemental physical processes into a macroscopic phenomenon, which can be measured in normal manner.
The moment of translation of the elemental physical processes into the macroscopic scale is accepted as the criteria of the system’s electric strength. Applied to the internal transformer insulation, this criteria is quantitatively defined by two parameters: the intensity of the electric field which causes the initial macroscopic stage of the discharge (partial discharge) and the intensity thereof.

Therefore, the accepted criteria for the primary oil barrier insulation defining the electric strength, is such intensity in the oil channel which causes partial discharges with apparent charge of 10-8 – 10-7 C, inducing irreversible damage on the barrier’s surface.

The intensity of the commencement of the initial stage of the discharge is the main, although not the only, factor influencing the processes initiation and development. The process is influencedby a number of factors, of which the following are worthy of consideration.

  1. Oil’s chemical structure.

The chemical composition of the oil has certain influence on the beginning and development of the initial breakthrough processes; for example, the amount of aromatic hydrocarbons in the oil’s molecule determines if the liquid will absorb or emit gas when subjected to the electric field. The chemical composition of the oil influences the development of molecular dissociation and secondary reactions which form gaseous substances. The presence and quantity of aromatic hydrocarbons influences impulse strength of the transformer oil in a nonhomogeneous electric field. Catalytic hydrocarbonyl auto-oxidation may influence the chemical composition of the oil and reduce its electric strength in the process of aging.

  1. Movement of liquid.

Electro-hydrodynamic forces acting on the insulation liquid as a result of the applied electric field, move the liquid. Besides, in large transformers the liquid is circulated by external pumps and under the influence of temperature difference in various liquid layers. Pressure may, in some cases, influence the liquid’s electric strength as a result of agitation of submerged contaminant particles, cause cavitation or promote generation and spread of discharges created by excitation of the liquid.

  1. Solid contaminants.

The harmful influence of solids has long been noted, and it is known that thorough purification of transformer oil can significantly increase its dielectric strength.

  1. Moisture.

Dielectric strength of mineral transformer oil is heavily influenced by its moisture content, especially in combination with solid contaminants. When determining dielectric strength, the more important parameter is the relative moisture content ratio rather than the absolute moisture content. The former depends on the oil’s chemical composition, temperature and the degree of oil’s aging.

  1. Temperature.

Many factors influencing the electric strength of transformer oil depend on temperature. For instance, the significant correlation of the oil’s viscosity and surface tension one the one hand, and its temperature on the other means that the related processes of cavitation and liquid movement are highly influenced by the oil’s temperature. Similarly, temperature changes influence relative moisture saturation of the oil and directly affect the oil’s dielectric strength. Indeed, results of tests where dielectric strength of the oil increases proportionately to temperature, usually indicate unacceptable concentration of water in the oil.
In a composite insulation (oil and cellulose), the estimation of the influence of temperature on the oil’s electric strength is highly complicated by moisture migration, requiring significant time to reach equilibrium between moisture content in the oil and in the cellulose insulation.

The above factors must define the conditions for application of the selected dielectric strength criteria.

Estimation of electric strength of actual systems requires the primary geometrical factor to be defined; that is the “size”, which determines electric strength. Although at this time it is not completely proven which of the two geometrical factors (square area or volume) should be considered the system’s “size”, most specialists prefer the tensioned volume, i.e. the volume of oil limited by the surface of the electrode and 80 or 90% equi-gradient surface area. From the physical viewpoint, the larger the volume of the insulation, the more probable is occurrence of a vulnerability, which may initiate breakdown in the area of high electric field intensity. For example, if the breakthrough is caused by solid particles, then a larger volume is a source of a larger number of contaminating particles, which may enter the area of high field intensity and initiate the breakdown.

When barrier system is used, travel of particles within the volume is limited and the dielectric strength of the system increases.

If assemblies with equal electrode surface area in homogeneous or quasi-homogeneous field are used, then the width of the oil gap becomes the geometric parameter defining its dielectric strength.