Copper vs aluminium winding is usually debated as a matter of principle. It is not. It is arithmetic, plus three site conditions.
The conditions are short-circuit duty, efficiency class and freight limits. Get them wrong and the cheaper metal becomes the expensive choice. Get them right and aluminium is a legitimate engineering decision, not a shortcut.
This article sets out the arithmetic first, then the three places where aluminium pays back the money it saved.
The One Number That Decides Everything
Conductivity decides the rest.
Commercial copper sits at 100% IACS, the International Annealed Copper Standard. Electrical-grade aluminium sits at about 61%. Both are handbook values, not marketing figures.
That gives one ratio. To match the resistance of a copper conductor, aluminium needs about 1.64 times the cross-section. One divided by 0.61 is 1.64, and you can check it on any calculator.
Now apply density. Copper is 8.96 g/cm³. Aluminium is 2.70. Multiply 1.64 by 2.70, divide by 8.96, and the equal-resistance mass ratio lands near 0.49.
The same winding duty therefore takes about 1.6 times the space and about half the weight.
That pair of numbers carries the whole commercial argument. Aluminium cuts the metal bill and the freight. It costs window space.
One caution about the ratio. Conductivity is quoted at 20 °C. Both metals lose conductivity as they heat, and they do not lose it at the same rate. The 1.64 figure is a design starting point, not a promise about operating temperature.
Everything after this point is a consequence of that trade.
Where the Savings Actually Go
A cheaper winding does not automatically produce a cheaper transformer. The copper vs aluminium winding comparison only holds if the rest of the design is allowed to change with it.
Start with the window. A conductor 1.64 times larger needs more room. The core has to grow to serve it, and a larger core carries more no-load loss. That loss runs for 8,760 hours a year, at zero output.
Then the tank. A bigger core and a longer mean turn push the tank outward. That adds oil, steel and cubic metres on the vessel.
Then the joints. Aluminium takes more preparation, more tooling and more inspection. That is labour, and labour does not shrink when the metal price falls.
A worked example shows the shape of it. Take a 630 kVA unit and set the copper winding mass at 100 units. An aluminium design matched for resistance needs about 164 units of section and about 49 units of mass.
For illustration, assume the core grows by a few percent. No-load loss follows core mass closely, so the unit now pays more for every hour it stands energised.
The material bill falls. The owning cost may not.
The example is deliberately unitless. Put in your own winding mass and your own valuation of a loss kilowatt, and the comparison becomes specific to your project. The mechanism does not change.
Total owning cost has four parts: purchase price, capitalised loss, maintenance, and failure risk. Only the first one appears on a quotation.
The Joint Is Where Aluminium Fails
Aluminium grows an oxide layer within minutes of being cleaned. Aluminium oxide is an insulator.
So a joint must be prepared mechanically or chemically, then sealed at once. A copper joint forgives a slow afternoon. An aluminium joint does not.
Aluminium also creeps. Under sustained bolt pressure at operating temperature, it flows slowly. Preload falls year by year. A joint that measured tight at factory testing can be loose at year five.
Thermal expansion adds to the problem. Aluminium moves about 23 × 10⁻⁶ per kelvin, copper about 17. Every load cycle works the joint a little more.
Direct copper-to-aluminium contact needs a bimetallic transition. Joined without one, moisture turns the pair into a galvanic cell, and the aluminium side corrodes first.
This is why aluminium failures cluster. They concentrate at joints and terminal connections rather than spreading through the winding.
A copper winding tends to age evenly. An aluminium winding tends to fail at one point, and that point is usually a joint someone stopped watching.
Two checks catch most of this before the tank closes. Torque marking on every bolted joint, photographed and recorded. And a thermographic scan at rated load, if your contract includes a temperature rise test.
Short-Circuit Forces Punish the Weaker Metal
Short-circuit force scales with current squared. This is the part of the copper vs aluminium winding question that price sheets never show.
A through-fault at ten times rated current produces about a hundred times the force on the winding. Nothing in the insulation system sees that; the bracing does.
Aluminium has lower yield strength and lower fatigue strength than copper. Under the same force it deforms earlier, and it does not spring back the way copper does.
Enlarging the section does not fix this. Stiffness comes from the support structure, and a larger conductor needs more of it: denser spacers, stronger end blocks, heavier clamping rings.
The bracing is not exotic. Spacers between disc pairs, end blocks under the clamping plates, tie rods holding the assembly against axial force. What changes with aluminium is how many there are, and how hard they are worked.
IEC 60076-5 is the proving standard. The type test is carried out on a new unit. After ten years of thermal cycling and a few through-faults, the unit on site is not the unit in the report.
This is where part of the aluminium saving goes back. The bracing is inside the tank, invisible on a quotation, and paid for either way.
Efficiency Rules Change the Math
Loss limits do not care what the winding is made of.
EU Regulation 2019/1783 and 10 CFR Part 431 in the United States set maximum losses by rating and application. An aluminium unit must meet the same figure as a copper one.
Higher efficiency tiers push conductor section up. Aluminium already needs 1.64 times the copper section; add a tighter loss limit and the window grows again. The core grows with it, and so does the tank.
The practical effect is a narrowing advantage. At low efficiency tiers, aluminium’s cost position is strong. At the top tiers, the extra material and the larger core can absorb most of the difference.
This is the part of the argument that rarely appears in a supplier’s comparison sheet. Aluminium is not uniformly cheaper. It is cheaper at some efficiency levels and roughly level at others.
When the Copper vs Aluminium Winding Decision Flips
Neither metal wins everywhere. The decision is a set of site conditions.
Read the table by the constraint that binds hardest, not by the average.
The Third Option: Copper-Clad Aluminium
Copper-clad aluminium is an aluminium conductor with a thin copper skin. It appears in low-cost offers, and it is worth understanding before it appears in yours.
The skin does almost nothing at power frequency. Current distributes across the section, not along the surface, so resistance follows the aluminium core. A CCA conductor rated as copper will run hotter than copper.
What the skin does change is appearance. The terminal looks like copper to a casual inspection, and the winding photographs like copper.
There is a legitimate use for CCA in high-frequency work, where the skin depth is thin and the copper layer carries the current. A 50 Hz transformer is not that application.
If a quotation says copper and the price sits far below the market, ask two things: the conductor certificate, and the measured resistance per kilometre. A copper conductor and a CCA conductor of the same nominal size do not measure the same.
What to Ask Before You Accept Either
Material is not the issue. Evidence is.
- Winding material certificate. Ask for it explicitly, and ask for copper-clad aluminium (CCA) to be excluded in writing. CCA looks like copper at the terminal and behaves like aluminium in the winding.
- Joint process. Welded, bolted, or brazed, and whether a bimetallic transition is used at every copper-to-aluminium interface.
- Short-circuit type test report, with the design reference that matches your drawing.
- Measured losses from the routine test, compared with the efficiency class claimed for your market.
- Winding mass and shipping mass. This is the number that tells you whether the design is what the quotation described.
Five items, one email. A supplier who answers all five is quoting a design. A supplier who answers two is quoting a price.
FAQ
Does an aluminium winding always mean a shorter life?
No. A well-made aluminium unit with proper joint preparation, bimetallic transitions and adequate bracing can serve a normal life. The copper vs aluminium winding decision turns on joints and bracing, not on the metal’s basic life.
Can copper and aluminium be mixed in one transformer?
In principle yes, and it happens in practice. Every interface then becomes a bimetallic joint, so the joint count rises and the inspection burden with it.
Why do some tenders specify copper only?
Usually for short-circuit duty, for maintenance access, or because the utility has standardised on copper after failures. It is a risk decision rather than a claim about conductivity.
Can an aluminium winding reach the top efficiency tiers?
Yes, but the section has to grow, which grows the core and the tank. Whether that still wins on price depends on the tier and on freight.
