How to Select a Dry-Type Transformer: Capacity, Class and Losses

What Counts as a Dry-Type Transformer

A dry-type transformer has no liquid dielectric: the winding insulation, the resin or the air itself does the electrical insulation work, and cooling is by natural or forced air. The absence of oil removes the fire load, the bunding requirement and the oil analysis routine, which is why dry-type units are the standard choice inside buildings, substations on upper floors, tunnels, hospitals and food or pharmaceutical plants. IEC 60076-11 is the reference standard for these units, with GB 1094.11 as the comparable national specification. Because the standard leaves several choices to the purchaser, selecting a unit means writing those choices into the order rather than accepting whatever is offered.

Capacity: Start From the Load Profile, Not the Nameplate

Capacity selection should follow a real load profile: connected load, load factor, diversity between loads, daily and seasonal shape, and the harmonic content of the largest drives or rectifiers. A transformer that runs at more than about 80% of rating on a sustained basis cannot follow its normal cyclic loading without exceeding its temperature rise limits, and one sized for a diversity figure that is never realised simply carries higher no-load losses for its whole life. The practical approach is to calculate the expected demand, add the planned expansion already permitted by the site, and then check the result against the standard cyclic loading guidance rather than against a rule of thumb.

Voltage, Vector Group and Impedance

The rated voltage of each winding follows the network, including the maximum system voltage at the point of connection and the expected fluctuation band. The vector group must match what the protection scheme and any parallel operation require; a common distribution group is Dyn11 on a three-phase unit, and two transformers operated in parallel must share the same vector group, phase displacement and turns ratio. Impedance is a design trade-off: a low value limits voltage drop and improves motor starting, while a higher value reduces the short-circuit current that downstream switchgear and busbars have to withstand. Distribution units are commonly supplied in the range of a few percent impedance, and the value chosen must be coordinated with the short-circuit rating of the equipment behind it.

Insulation Class and Temperature Rise

The temperature class of the insulation system is defined by IEC 60085, and IEC 60076-11 fixes the maximum average winding temperature rise that applies to each class. The gap between the two figures is the thermal margin the designer keeps for hot spots, overload and a warm site. A Class H unit running at Class F rise lasts longer than the same unit pressed to its own limit, so the rise actually guaranteed on the test certificate matters as much as the class letter. Normal ambient conditions for rating assume a maximum air temperature of 40 degC, and an installation above 1000 m altitude requires derating.

Enclosure, Cooling and Site Conditions

Enclosure protection is selected against the environment and not against habit: a clean, dry electrical room needs only basic protection, while dusty or damp areas need a higher degree of protection under IEC 60529 and often forced cooling with filtered air. Forced-air cooling raises the usable capacity of a given frame but adds fans and controls that must be maintained, and once fans are fitted the guaranteed capacity without them is the figure the load should be planned around. Ventilation of the room, clearance to walls, cable entry direction and seismic or transport loads all belong in the specification because they are expensive to change after the unit is cast.

Losses, Efficiency Grade and Harmonic Duty

Energy efficiency is regulated rather than optional in many markets. GB 20052-2020 sets minimum allowable values of energy efficiency and the efficiency grades for dry-type transformers, with Grade 1 as the highest level, so a unit offered without a declared grade cannot be compared objectively. No-load loss is driven by core material and flux density and is present 8760 hours a year, while load loss scales with the square of the load, which is why the choice between amorphous and grain-oriented cores depends on the load factor. Where a significant part of the load is electronic, the transformer should be specified for harmonic duty using a K-factor or equivalent derating, because eddy losses in the windings rise steeply with frequency and cannot be removed by a larger cooling fan.

Frequently Asked Questions

Q: Which standard governs dry-type transformer selection?
A: IEC 60076-11 is the core standard, supported by GB 1094.11 in China, IEC 60085 for the insulation system and IEC 60529 for enclosure protection.

Q: What temperature rise should be specified for a Class H unit?
A: A maximum average winding rise of 125 K is the limit for Class H under IEC 60076-11, but many prudent specifications call for a lower rise to preserve margin for hot spots and overload.

Q: Does a higher impedance make a transformer better?
A: No. Higher impedance reduces short-circuit current but increases voltage drop and losses. The correct value is the one coordinated with the downstream switchgear rating and the motor starting duty.

Q: Can two dry-type transformers be operated in parallel?
A: Yes, if their rated voltages and turns ratios match, their vector groups and phase displacement are identical, and their impedances are close enough that the load divides near to proportion. Impedance mismatch of more than about 10% is usually avoided.

Q: How much maintenance does a dry-type transformer need?
A: Less than an oil-filled unit, but not none. Winding cleanliness, terminal torque, fan operation, dust on insulation surfaces and periodic insulation resistance measurement are the routine items, and partial discharge testing is used where the standard requires it.