A transformer is one of the most efficient machines ever created, with an efficiency that exceeds 98 percent.
That makes it sound as if losses are hardly worth discussing. But those few percentage points operate without interruption, every hour of the day, for thirty years or more, and by the end of the transformer’s service life they often exceed its purchase price.
Transformer losses come in two forms.
No-load losses occur in the magnetic core and exist whenever the transformer is energised, completely независимо of whether anyone is actually consuming electricity. A factory may be closed for holidays, but the core still consumes power.
Load losses occur in the windings and increase with the square of the current, which means they depend on how heavily the transformer is loaded.
This distinction is crucial for the calculation: no-load losses are paid for 8,760 hours per year, while load losses are paid only when the transformer is actually supplying load.
Let us take a simple illustration.
A difference of just one kilowatt in no-load losses between two transformers means 8,760 kilowatt-hours per year, every year, without exception.
Multiply that by the energy price and by three decades of service life, then add the load losses, and it becomes clear why serious investors do not ask only how much a transformer costs to buy, but also how much it consumes.
This is precisely why European ecodesign regulations impose increasingly strict limits on losses for new transformers, and why the difference between an older unit from the 1990s and a modern transformer is greater than many people assume.
There is also an opposite pitfall.
A transformer selected with excessive spare capacity operates permanently at low load, where no-load losses dominate. As a result, the machine may consume almost as much as if it were running at full capacity, while delivering relatively little useful work.
Optimal sizing is neither overly cautious nor overly ambitious. It should be based on the actual load profile, selecting a transformer that operates most of the time in the range where the relationship between useful power and losses is most favourable.
For a transformer already in operation, the calculation is straightforward: measure or estimate its losses, compare them with those of a modern unit of the same rated power, and determine the annual value of the difference.
For older machines with high no-load losses, replacement can have a shorter payback period than most people would expect, especially when it coincides with the need for higher capacity or with the transition of the network to 20 kV.
Losses are a silent cost. They are not visible on the machine itself, but they appear regularly on every electricity bill, which is why they deserve a place in every serious analysis of a company’s energy costs.
