Dry Type Transformer Efficiency: Load Factor and Losses

Efficiency Should Not Be Judged from One Percentage

A quoted efficiency of 98.8%, 99.0% or 99.2% tells very little unless the operating conditions behind the figure are known.

For an EPC engineer evaluating dry type transformer efficiency, the more useful information is normally:

• rated capacity;
• actual operating load;
• no-load loss;
• load loss;
• power factor;
• expected daily or annual load profile;
• temperature basis of the loss data.

Two 1000 kVA dry-type transformers can have the same rated capacity yet produce different annual energy losses.

One design may have very low core loss but slightly higher full-load loss. Another may have higher core loss but lower load loss. Which transformer consumes less energy depends heavily on how many hours the transformer spends lightly loaded, moderately loaded or near rated current.

That distinction is why engineering specifications should normally examine guaranteed losses rather than relying only on a catalogue efficiency percentage.

Rated Capacity, Load Ratio, Losses and Efficiency Are Different

Four quantities need to be separated.

Rated capacity

Transformer capacity is normally stated in kVA.

A 1000 kVA transformer is designed around a specified apparent-power rating under defined service and cooling conditions. The rating does not mean the transformer continuously delivers 1000 kW.

The real output power depends partly on load power factor.

For example:

1000 kVA × 0.9 power factor = approximately 900 kW of active output at rated apparent load.

Operating load ratio

For the calculations in this article, the load ratio is:

x = actual transformer kVA / rated transformer kVA

A 1000 kVA transformer carrying 500 kVA therefore operates at:

x = 0.50

or 50% loading.

This use of load ratio should not be confused with the utility-system definition of load factor, which can refer to average demand divided by peak demand over a period.

Transformer losses

The two main groups relevant to efficiency calculation are:

No-load loss P₀

and

Load loss Pₖ

Transformer efficiency

Efficiency expresses how much input active power reaches the load:

η = Pout / (Pout + Ploss) × 100%

The important point is that Ploss does not remain constant as the transformer load changes.

No-Load Loss vs Load Loss

Understanding these two loss components explains most of the relationship between transformer loading and efficiency.

No-load loss

No-load loss is mainly associated with the energized magnetic core.

It exists whenever rated voltage and frequency are applied, even if almost no secondary load is connected.

For practical operating calculations, no-load loss is often treated as approximately constant when voltage and frequency remain close to the design conditions.

This creates an important procurement issue.

A transformer that remains energized 8,760 hours per year may accumulate significant no-load energy consumption even if its average load is low.

That makes no-load loss particularly important for:

• lightly loaded distribution transformers;
• redundant transformers;
• commercial buildings with large night-time load reductions;
• facilities where two transformers share a load that could sometimes be supplied by one unit.

Load loss

Load loss is associated primarily with winding resistance and other current-dependent losses.

The resistive component follows:

P ∝ I²R

For a simplified engineering estimate, load loss at loading ratio x can therefore be represented as:

Pload ≈ x² × Pₖ

where Pₖ is the specified load loss at rated current and the applicable reference temperature.

At 50% rated current:

0.5² = 0.25

so the simplified current-dependent load loss is about 25% of the rated-load value.

At 80% current:

0.8² = 0.64

or approximately 64%.

This squared relationship is one reason sustained high loading changes transformer losses much faster than many buyers expect.

Actual calculations can require corrections for winding temperature, stray losses, harmonics and other design conditions. Guaranteed values should therefore be taken from the approved technical datasheet and applicable factory test documentation.

How to Calculate Dry Type Transformer Efficiency

For a simplified three-phase transformer operating under normal sinusoidal conditions:

Pout = x × Sr × PF

where:

• x = operating load ratio;
• Sr = rated transformer capacity in kVA;
• PF = load power factor.

Approximate transformer loss is:

Ploss = P₀ + x²Pₖ

Efficiency can therefore be estimated by:

η = [x × Sr × PF] / [x × Sr × PF + P₀ + x²Pₖ] × 100%

Illustrative 1000 kVA example

Assume a hypothetical transformer with:

Rated capacity: 1000 kVA
No-load loss P₀: 2.0 kW
Rated load loss Pₖ: 10.0 kW
Power factor: 0.90

These figures are for calculation only and are not guaranteed values for a particular Huamei transformer.

Loading	Output Power	Estimated Loss	Efficiency
25%	225 kW	2.625 kW	98.85%
50%	450 kW	4.50 kW	99.01%
75%	675 kW	7.625 kW	98.88%
100%	900 kW	12.0 kW	98.68%

The result appears counter-intuitive at first.

Efficiency increases from 25% to around 50% loading, then begins to fall even though more useful power is being delivered.

The reason lies in the balance between constant core loss and current-dependent load loss.

At What Load Is a Transformer Most Efficient?

For the simplified loss model:

Total loss = P₀ + x²Pₖ

Maximum efficiency occurs approximately when:

x²Pₖ = P₀

or:

x = √(P₀ / Pₖ)

Using the illustrative values above:

x = √(2 / 10)

x ≈ 0.447

The theoretical maximum-efficiency point is therefore around 45% loading under the stated assumptions.

This does not mean every dry-type transformer should be operated at 45%.

It means that the maximum-efficiency loading point depends on the relationship between that transformer's no-load loss and rated load loss.

If a transformer has very low no-load loss and relatively high load loss, its efficiency peak may occur at a lower loading percentage.

If load loss is comparatively low, the peak can move higher.

The engineering objective is also rarely to operate continuously at the mathematical maximum-efficiency point. Capacity margin, redundancy, fault current, future expansion, thermal conditions and operational resilience can be more important.

High Percentage Efficiency Does Not Always Mean Low Energy Cost

Suppose Transformer A operates at 99.0% efficiency and Transformer B operates at 98.9%.

The 0.1 percentage-point difference looks small.

Across a large continuous load and many operating hours, however, the energy difference can become commercially relevant.

The opposite problem also occurs.

A transformer may display a very high instantaneous efficiency during a load test while still consuming unnecessary annual energy because it remains energized at low load for thousands of hours.

This is why annual loss analysis should use the load profile, not only one operating point.

A simplified annual calculation is:

Annual Energy Loss ≈ P₀ × energized hours + Σ(Pₖ × x² × operating hours)

If forced-air cooling is used, relevant fan and auxiliary consumption should also be considered in system-level energy analysis.

For installations with strongly changing demand, the engineer should evaluate several loading bands rather than selecting one assumed average percentage.

For example:

night load;
normal daytime load;
production peak;
seasonal peak.

A single “average load” can underestimate I²-related losses because squaring the average current is not mathematically equivalent to averaging the squared current.

How Power Factor Affects Efficiency

Transformer loading is often expressed in kVA, while useful energy is billed and consumed largely in kW.

This distinction matters.

At the same kVA loading, reducing power factor reduces useful active-power output while transformer current can remain similar.

Current-dependent losses therefore remain significant while useful kW output falls.

The calculated efficiency based on active output power consequently becomes lower.

This is one reason a statement such as:

“Transformer efficiency is 99%”

is incomplete unless its load and power-factor basis are defined.

For tender comparison, purchasers should verify whether competing efficiency figures were calculated under equivalent conditions.

Harmonics Can Change the Simple Loss Relationship

The x² loss equation is useful for basic engineering estimates, but non-linear loads require more care.

Typical harmonic-producing equipment includes:

• variable-frequency drives;
• UPS systems;
• rectifiers;
• data-center power supplies;
• industrial electronic loads.

Harmonic currents can increase winding and stray losses beyond what would be predicted from fundamental-frequency current alone.

A transformer carrying 70% of rated kVA under a highly distorted current waveform cannot automatically be assumed to have the same thermal and loss performance as a transformer carrying a clean sinusoidal 70% load.

Where substantial harmonics are expected, the RFQ should identify the load characteristics rather than asking the manufacturer to select the transformer only from total kVA.

Temperature Also Affects Load Loss

Winding resistance changes with conductor temperature.

Load-loss values therefore need a defined temperature basis when transformers from different suppliers are compared.

Comparing:

“Supplier A: 9.2 kW load loss”

with:

“Supplier B: 8.8 kW load loss”

is not controlled engineering comparison if the figures have been stated or corrected using different temperature conditions.

The approved technical datasheet and test report should use the project-agreed basis.

IEC-based dry-type transformer projects commonly reference IEC 60076-11 within its applicable scope. Energy-performance evaluation may also involve IEC TS 60076-20 depending on purchaser requirements and local regulation. IEC 60076-19-1 addresses uncertainty in measurement of transformer no-load and load losses during routine tests.

Does AN or AF Cooling Improve Electrical Efficiency?

AN and AF primarily describe cooling arrangements rather than transformer efficiency grades.

AN means natural air cooling.

AF means forced-air cooling using fans.

Fans can increase thermal capability in a transformer specifically designed for an AF rating, but the existence of fans does not automatically mean lower electrical losses.

At the same winding current, the electromagnetic losses are determined mainly by the transformer design and operating condition.

Lower winding temperature can influence resistance, but project evaluation must consider the complete thermal design rather than treating AF as an efficiency upgrade.

Fan auxiliary power should also be considered when evaluating total system energy consumption.

Any AN/AF ratings should be confirmed from the manufacturer's thermal design and approved technical datasheet.

Why Oversizing Can Increase Energy Loss

Specifying a much larger transformer than the project requires may provide capacity reserve, but it is not automatically an energy-efficiency improvement.

Consider a load requiring only 300 kVA for most of its life.

A 2000 kVA transformer may operate safely, but its no-load loss continues whenever the unit is energized.

A smaller properly engineered transformer could potentially produce lower total energy loss over the actual load cycle.

This does not mean transformers should be selected with minimal spare capacity.

An EPC study should still consider:

• future expansion;
• motor starting;
• cyclic peaks;
• redundancy philosophy;
• ambient temperature;
• harmonic loads;
• permissible temperature rise;
• reliability requirements.

The point is narrower:

Transformer capacity and transformer energy efficiency should be evaluated together rather than assuming that a larger rating is always better.

How to Compare Two Transformer Quotations

For efficiency-sensitive projects, do not compare only:

1000 kVA
11 kV / 0.4 kV
Dyn11
6% impedance

Ask each supplier to state the guaranteed loss values under an agreed technical basis.

A useful comparison table includes:

Parameter	Supplier A	Supplier B
Rated capacity
HV / LV voltage
Frequency
Vector group
Impedance
Winding material
No-load loss
Load loss
Loss reference temperature
Cooling
AN / AF rating
Temperature rise
Efficiency calculation basis

Low no-load loss versus low load loss

Consider two hypothetical designs:

Transformer A:
lower no-load loss, higher rated load loss.

Transformer B:
higher no-load loss, lower rated load loss.

Transformer A may be more attractive for a lightly loaded transformer that remains energized continuously.

Transformer B may perform better where the transformer spends most operating hours near high load.

This is why the lowest value in one column does not automatically identify the most efficient transformer for the project.

The purchaser needs the operating load profile.

Guaranteed Losses Matter More Than Marketing Efficiency Claims

An efficiency figure on a product page is mainly useful when the calculation conditions are defined.

For procurement, the stronger document set is:

• approved technical datasheet;
• guaranteed no-load loss;
• guaranteed load loss;
• impedance;
• reference conditions;
• routine test report;
• FAT records when FAT is contractually required.

Loss values can then be compared with the agreed project specification.

Routine Test, Type Test, Special Test and FAT should not be treated as interchangeable terms.

The precise test program and acceptance requirements should be agreed before production according to the applicable standard and approved project specification.

What Should an Efficiency-Focused RFQ Include?

An RFQ should not simply state:

“Please quote your most efficient 1000 kVA dry transformer.”

A more controlled specification should include the parameters that affect both electrical design and operating evaluation:

Rated capacity
Primary voltage
Secondary voltage
Frequency
Phase
Vector group
Tap range
Required impedance
Expected continuous loading
Expected peak loading
Typical load profile
Power factor
Harmonic information if applicable
AN or AF cooling requirement
Ambient temperature
Altitude
Enclosure / IP requirement
Specified loss limits, if any
Applicable standard
Required test documentation
Quantity

If loss capitalization or lifecycle-energy evaluation is part of the tender, state that methodology before suppliers prepare their offers.

That allows different transformer designs to be evaluated on a consistent basis.

FAQ

1. What is a good efficiency for a dry-type transformer?

There is no universal efficiency percentage that should be used for every rating and project. Compare guaranteed no-load and load losses at equivalent rated conditions and evaluate them against the real load profile.

2. Is a dry-type transformer most efficient at full load?

Not necessarily. In a simplified model, maximum efficiency occurs when load-dependent loss equals no-load loss. The corresponding load percentage depends on the transformer's specific P₀/Pₖ relationship.

3. Why does transformer efficiency change with load?

No-load loss remains relatively constant while load loss increases approximately with the square of current. Their relative contribution therefore changes as transformer loading changes.

4. Does a transformer consume power when there is no load?

Yes. An energized transformer has no-load loss even when secondary load is very small or zero.

5. Does power factor affect transformer efficiency?

Yes, when efficiency is expressed using active output power. At the same kVA/current loading, lower power factor means lower useful kW output while current-related losses can remain significant.

6. Should I compare efficiency or losses when buying a transformer?

For technical procurement, guaranteed no-load loss and load loss are usually more useful than a standalone efficiency percentage because the buyer can calculate performance under the project's actual operating profile.

7. Can an oversized transformer be less energy-efficient?

It can be under a lightly loaded duty cycle because the transformer still incurs no-load loss whenever energized. Oversizing must therefore be justified by capacity margin, future expansion, redundancy and other project requirements.

8. What transformer loss information should be requested before ordering?

Request guaranteed no-load loss, load loss, the applicable temperature/reference conditions, rated capacity, impedance, cooling method and the required factory test documentation.

CTA

If you are evaluating dry-type transformers for an EPC or industrial project, provide the rated capacity, primary and secondary voltage, frequency, vector group, impedance, expected load profile, cooling method, enclosure requirement and any specified loss limits.

The transformer design can then be reviewed against both the required kVA rating and the project's actual operating-loss profile before quotation.