Transformer sizing goes wrong in two directions, and both cost real money.

Undersizing is the obvious one. A unit that runs continuously near its rating gets hot, and heat is what ages insulation. A transformer running permanently at the edge of its capacity will not fail this year, which is why the mistake is so easy to make, but it will fail years earlier than it should and it will fail without warning.

Oversizing is the mistake nobody talks about. A transformer that is far larger than its load still has no load losses running twenty four hours a day, every day, for its entire life. You paid more for the unit and then you pay again every month in losses on capacity you are not using.

Start with the actual load, not the connected load

The number most people use is the sum of everything on the schedule. That figure is almost never reached, because not everything runs at once. What you need is the maximum demand, which is the connected load reduced by a realistic diversity factor for the type of building or plant.

Get this from measurement where an existing installation is being replaced. A clip on meter over a working week tells you more than any calculation, and it is the single most useful hour anyone can spend before ordering a transformer.

Then add headroom for growth, deliberately

Add growth on purpose rather than by accident. A factory planning a second line in two years should size for it now, because replacing a transformer later costs far more than the difference between two ratings today. An estate that is fully built out does not need the same allowance.

As a rule of thumb, loading a distribution transformer to around seventy five to eighty percent of its rating at maximum demand leaves sensible headroom without paying for capacity you will never use. Below about fifty percent, you are almost certainly oversized.

Do not forget what the load actually is

Motor starting current, welding plant and large air conditioning loads all behave differently from lighting and sockets. If your load is motor heavy, the starting behaviour matters as much as the steady state figure and the calculation changes.

The practical answer

Bring us the load schedule, or better still a week of measured demand, and tell us what you expect to add in three years. We will tell you what we would specify and why, and if the rating you asked for is wrong in either direction we will say so. That conversation is free and it takes about twenty minutes.

See our transformer range, or send us your load figures.