Short answer: If you can’t describe the network, you can’t choose the transformer.
Here’s the thing. “Distribution network” covers a lot of sins. Urban feeders. Rural overhead lines. Industrial parks. Microgrids that pretend they’re utilities.
Before you look at kVA or voltage, ask:
Is this a utility-owned distribution system or a private network?
Is the load mostly residential, commercial, or industrial?
Is power flowing one way—or sometimes back from solar or BESS?
Those answers change everything. I’ve seen the same “standard” transformer behave beautifully in a city grid and fail early in a weak rural network. Same model. Totally different reality.
Bottom line: Size for realistic future load, not today’s spreadsheet.
People love neat numbers. Peak demand today × 1.1. Done.
That works… until EV chargers show up or someone adds a small factory at the end of the line.
When I look at sizing, I ask:
What’s the actual peak, not the averaged one?
Will load grow in steps or slowly over time?
Is this transformer hard to replace once installed?
Base load (measured, not guessed)
Seasonal peaks (summer is a liar)
Load growth over 10–20 years
Duty cycle (constant vs. spiky)
If this unit feeds a critical part of the distribution network, slight oversizing is cheap insurance. Undersizing is not.
Quick answer: Nominal voltage is easy. Voltage variation is the real fight.
Distribution networks are rarely polite. Long feeders. Weak sources. Reactive power doing weird things at night.
Don’t just ask for “11 kV / 0.4 kV” and move on.
Ask:
What’s the maximum and minimum primary voltage?
Are voltage drops common at peak load?
Is there on-load or off-circuit tapping—and will anyone actually use it?
A transformer that can tolerate voltage swings will live longer. Full stop.
Yes, indoor installations often push people toward dry-type transformers. Yes, outdoor substations often use oil-filled units. But rules like that break down fast.
Fire risk is a big deal
Space is tight and enclosed
Maintenance access is limited
You need higher power density
Ambient temperature is brutal
Long-term thermal performance matters
I’ve seen dry-type transformers struggle in dusty distribution rooms and oil-filled units run for decades outdoors with basic care. Context wins.
Very important. And often ignored.
No-load losses run 24/7. Even when the feeder is quiet. Even at 3 a.m.
In a distribution network, that adds up fast.
When comparing options:
Look at core loss, not just load loss
Ask for loss capitalization, not just purchase price
Think in years, not invoices
Utilities obsess over this for a reason. Private networks should too.
IEC 60076. IEEE C57. Everyone meets them.
What I care about more:
Thermal margins
Insulation class choices
How conservative the design feels when you read the test data
A transformer can pass every test and still hate your network conditions. Ask how it behaves outside the lab.
Yes. More than most people admit.
A transformer isn’t just steel and copper. It’s support, drawings, responses when something feels off.
Before choosing:
Ask about similar distribution network projects
See how they answer uncomfortable questions
Check whether engineering talks or sales talks
If everything sounds polished and vague, that’s a red flag.
Define the distribution network first: urban, rural, industrial, or mixed
Size for growth, not just today’s demand
Voltage variation matters more than nominal ratings
Oil vs dry-type depends on site conditions, not habit
Losses cost real money over 20+ years
Standards are a baseline, not a guarantee
The supplier is part of the equipment
