9
\$\begingroup\$

So aluminium is 3.3 times lighter than copper (8.96 vs 2.70 g/cm3). And its resistance is only 55% bigger. So in weight/resistance ratio it is actually better than copper. Also it is cheaper.

Why aluminium is not a primary choice in cables? Does it have something to do with corrosion? Does it solder differently?

\$\endgroup\$
11
  • 3
    \$\begingroup\$ Skim this, for example. First link from a google search. \$\endgroup\$ Commented Jul 25, 2023 at 7:15
  • 8
    \$\begingroup\$ @RiDi did you read periblepsis link? There are like 100 reasons in there. Also from the handful of times I had to solder aluminium, I can safely say it sucks. Like real hard. \$\endgroup\$
    – Max
    Commented Jul 25, 2023 at 7:57
  • 4
    \$\begingroup\$ @micropyre They do not use the same size wire. They use 55% more wire so the resistance is the same, and the cable still ends up being lighter and cheaper because aluminium is so light and cheap. You could even use 100% more wire, making it less resistive, lighter and cheaper. \$\endgroup\$ Commented Jul 25, 2023 at 8:01
  • 4
    \$\begingroup\$ For overhead power lines the cables are often made from aluminum, the conductivity per weight is better. A core of steel wires is often used to increase overall conductor strength. But soldering copper is easy and aluminum pretty difficult. The surface of an aluminum wire is covered with a thin oxide layer which is not solderable. \$\endgroup\$
    – Uwe
    Commented Jul 25, 2023 at 8:01
  • 5
    \$\begingroup\$ Termination of aluminium cables are much more difficult than copper. In underground and overhead lines where this only happens once and can be done with hydraulic tools or even explosives to crimp splices, this is less of a concern. Thermal expansion is also a problem. \$\endgroup\$
    – winny
    Commented Jul 25, 2023 at 8:14

6 Answers 6

18
\$\begingroup\$

It is the primary choice in HV power transmission cables- aluminum with a steel core for strength.

As well as the well-known issue of cheaply and easily making reliable field connections with small gauge simple aluminum wires (eg. house wiring), the higher resistivity means things like transformers and motors would need more iron. They may still be somewhat more economical, all things considered. From here is a comparison of 2.5MVA electrical transformers made with copper vs. hybrid vs. aluminum:

enter image description here

\$\endgroup\$
4
  • \$\begingroup\$ The question does specify cables, rather than motors or transformers \$\endgroup\$ Commented Jul 25, 2023 at 14:47
  • 1
    \$\begingroup\$ I'm confused. Why is aluminum showing higher efficiency and lower losses than copper? \$\endgroup\$ Commented Jul 25, 2023 at 16:43
  • 7
    \$\begingroup\$ @TylerShellberg Looks like they designed them all to have 99.53% efficiency at 50% load. That can't be a coincidence. The variation is probably core loss vs. conductor loss at different loads. \$\endgroup\$ Commented Jul 25, 2023 at 17:03
  • \$\begingroup\$ @user253751 Good point. \$\endgroup\$ Commented Jul 25, 2023 at 17:04
10
\$\begingroup\$

In addition to the other answers, Aluminum has a 30% greater co-efficient of thermal expansion than copper. This is not a problem per se. However, in some cases where aluminum wire has been used, proper account for the greater thermal expansion has not been taken into account. As a result, with repeated thermal cycling, termination connections have been loosened to the point where the connection resistance has become excessive and resulting in charring and fires.

enter image description here (image from Grace Electrical Services

As mentioned, the thermal coefficient of expansion does not, in itself, make aluminum inappropriate for uses, such as building wiring. However, it does require that appropriate termination techniques and materials be applied. Aluminum wiring is still permitted by the US National Electrical Code for building wiring. (citation here )

\$\endgroup\$
4
  • 1
    \$\begingroup\$ The image doesn't add anything other than as an attention-grabber \$\endgroup\$
    – pipe
    Commented Jul 25, 2023 at 13:09
  • 4
    \$\begingroup\$ One lab I worked at experienced electrical noise problems when F4 Phantom jets took off, especially with afterburners. Turned out to be aluminum mains wiring rattling about. Every year, all wire connections to the outlets were tightened to prevent issues. \$\endgroup\$
    – qrk
    Commented Jul 25, 2023 at 14:55
  • 5
    \$\begingroup\$ Aluminum is allowed when used with terminals that are rated for it (marked "Cu-Al"). As a further precaution, practice was for Al wire is to be coated with anti-oxidation paste (Ox-guard or similar) where it makes connections, although that seems to be going away now. In normal houses, Al is only used for high-current runs; normal receptacles use copper. Mobile homes used to use Al wire for receptacles, but switched over to copper to prevent the kind of thing shown in your photo. (Sadly, a good friend of ours died in a fire from such an incident.) \$\endgroup\$ Commented Jul 25, 2023 at 18:43
  • \$\begingroup\$ @pipe -- I disagree that the picture is just an attention grabber. The house my parents built had a fire because of the aluminum wiring, so to me, the picture is warranted as an effective, reasonable warning about a common problem. \$\endgroup\$ Commented Oct 28, 2023 at 3:32
10
\$\begingroup\$

The surface of an aluminum object will oxidize when exposed to air, producing a thin layer of aluminum oxide.

This is great for manufacturing, because it protects the material from further degradation. It's not so great for conductors, because the oxide layer is an electrical insulator. (You may have seen small white plates of aluminum oxide form between components and heatsinks.)

Terminating aluminum cables requires specialty components and anti-oxidizing paste. Installation therefore requires more effort and is thus more expensive and fault-prone. So aluminum is only used for long stretches of cable where it will eventually be cheaper than copper.
For example, many underground cables are aluminum.

You may have seen copper-clad aluminum cables. You can connect these using standard terminals, but the aluminum is only present for mechanical strength, not for conductivity.

\$\endgroup\$
11
  • 2
    \$\begingroup\$ I'm not sure that "rust" is the right word. "Rust" usually means something bad. I'm pretty sure that materials scientists have another word or phrase that they use to describe a self-healing, protective coating that naturally grows on a metal surface, but I can't seem to recall it at this moment. \$\endgroup\$ Commented Jul 25, 2023 at 12:53
  • 1
    \$\begingroup\$ Oxidize might be better word \$\endgroup\$
    – Marla
    Commented Jul 25, 2023 at 13:04
  • 4
    \$\begingroup\$ Instead of "rust" you may talk of a "passivation layer". \$\endgroup\$
    – user107063
    Commented Jul 25, 2023 at 14:08
  • 5
    \$\begingroup\$ Copper-clad aluminum is most certainly used for its conductivity as well as lightness and handling but not its mechanical strength (pure copper is much less prone to fatigue). You may be confusing this with copper-clad steel. \$\endgroup\$
    – user107063
    Commented Jul 25, 2023 at 15:53
  • 1
    \$\begingroup\$ Aluminum is not very strong. Aluminum alloys can be quite strong, and are very frequently used in construction because of it, but the pure aluminum used in aluminum wires (including copper-clad aluminum) is not much of an improvement over copper. (it might even be worse--I don't recall.) You can't use aluminum alloys for cables because they're much less conductive than pure aluminum. \$\endgroup\$
    – Hearth
    Commented Jul 26, 2023 at 0:44
9
\$\begingroup\$

Aluminum is marvelous, underrated stuff

And North American AC electrical wiring has achieved mastery of using it, including the unfortunate exception which proves the rule.

Never mind copper being twice the weight for the same ampacity. It's 10 times the cost for the same ampacity!

Copper had to come first, because it exists primordially - you can find copper nuggets in nature, so it was one of the first worked metals in human history. The affordable process for refining aluminum requires staggering amounts of electricity, so by definition, aluminum enters a world that is already fully built-out in copper.

What are the tricks to working with it?

In practical application in house wiring, there are two "golden rules" called out on the first page of the Code book.

Use terminals designed and rated for aluminum wire (NEC 110.3). Generally, the simplest way to do this is to use a lug terminal which is itself made of aluminum. Panel neutral bars, large lugs, Alumiconn and Polaris style connectors all use this method.

This works because aluminum has a different coefficient of expansion than copper. When the aluminum lug envelops a copper wire, this thermal expansion works favorably.

The exception that proves the rule is the 1970s experiment with small aluminum branch circuit wires in North America. This was done in response to a copper shortage, and the government ordered the certification agency (UL) to fast-track approval of switches, receptacles and splices. UL hastily certified ordinary devices made of brass without adequate testing. This caused frequent failure, and the postmortem revealed the rule I state above. Design standards were significantly revised (to CO-ALR), and those devices have terminals made of indium, a soft, conformant metal that mates well with aluminum.

Torque terminals to spec with a torque wrench (NEC 110.14). This was a universal "best practice" that electricians did on large high-current terminals, and those have proven reliable in both metals. However, for "the small stuff" (15-50A wires), no one bothered - they just cranked them down "gud-n-tite" with no particular standard for what that means.

And totally unrelated to aluminum, this was found to be a major cause of failure for copper connections, and of course there's no reason to think mistorqued aluminum connections were performing any better. This was coupled with some trade-show testing that proved pro electricians couldn't set torques "by feel" any better than their spouses. It turns out the entire industry had been in the bottom quartile of the Dunning-Kruger curve, and many are still in denial.

Use the proper alloy. The 1970s effort involved taking AA-1350 alloy "outdoor transmission line" wire and shrinking it down. This did not perform well. A new AA-8000 alloy was specially developed for interior electrical wire.

So with those application techniques mastered, aluminum is proving to be a winner for architectural electrical distribution. It has always dominated the skies of utility distribution.

That said, in residential the stigma of the 1970s mistake lingers, and don't wire with aluminum smaller than 50A if you want to get good money when you sell the house.

How do you design for aluminum?

A worthy question. And please - if you are designing an EVSE, accommodate aluminum wire. Almost none do, and it greatly increases installation cost (6 AWG copper NM vs 6 AWG aluminum SER/SEU).

The gold standard in the electrical industry is an aluminum-bodied lug connector. So have a copper lead or bar come out of your machine and land on an aluminum bodied lug with two screw terminals - one for your copper lead/bar, and one for the customer supplied wire connection. Torque your side to spec and have a label reminding the installer to do the same. It also helps to use 75C thermal connections, which allows a smaller wire size to be used.

If you are attaching these lugs to a PCB or plastic molding, they do not need to be insulated and can be quite inexpensive.

Reason for no: Tokamaks

If you're trying to run a cyclotron or magnetic isotopic separator, you need the maximum possible magnetic flux in the least possible space. This can be important for motors, too, if you need to cram an extremely high power motor in the limited space between railroad rails, for instance.

For that, copper can't be beat.

Reason for no: you want to sell your house.

NEC has no quarrel with use of aluminum wire for small branch circuits, subject to the new rules.

However, if you do, the buyer's inspector will find it, mistake it for the problematic old stuff, and tell the buyer not to buy your house. They will demand a discount, remediation, or walk away.

Use for >50A circuits continues to be perfectly fine. A few superstitious jurisdictions raise that standard to 90A.

Reason for no: flexible cordage

For fine-threaded flexible cordage such as found between an appliance and its wall plug, aluminum is really too brittle. It does not like the constant flexure.

Reason for no: physical working

Aluminum has no fatigue limit so if it's in an application where it'll be carrying a lot of physical weight and moving a lot, such as an overhead line, it's going to accumulate metal fatigue and fail. Copper has the same problem, though.

The cure for this is ACSR, which is aluminum wire around a steel core. Now the steel core takes care of the physical strain of supporting the cable, and the aluminum is just along for the ride. This is not used in residential "service drops"; that's just regular duplex/triplex/quadplex.

\$\endgroup\$
7
  • 3
    \$\begingroup\$ Copper can and regularly is beat for maximum flux in minimum space: large particle accelerators and MRIs typically use superconducting cable, because despite the enormous drawbacks (such as requiring a supply of liquid helium), it does beat out copper in terms of ampacity. \$\endgroup\$
    – Hearth
    Commented Jul 26, 2023 at 3:12
  • 1
    \$\begingroup\$ Hm, just a niggle about osmium -- it's a very hard metal. I also don't see any references using it for connector plating, at all (rhodium is usually the one for that), or in conjunction with aluminum wiring; though searching the internet is pretty awful for finding that sort of thing. I see a reference inspectapedia.com/aluminum/Aluminum_Wiring_Repair_COALR.php which suggests indium may be suitable, though? \$\endgroup\$ Commented Jul 26, 2023 at 4:02
  • 3
    \$\begingroup\$ @TimWilliams Osmium would also be a very bad choice for connector plating considering it spontaneously reacts with oxygen in the air to form the extremely toxic gas osmium tetroxide. Osmium is nasty stuff. \$\endgroup\$
    – Hearth
    Commented Jul 26, 2023 at 4:21
  • 1
    \$\begingroup\$ In regard to the maximum flux: the enrichment gear at Manhattan project used silver. They did this because copper was not available, but still... \$\endgroup\$
    – fraxinus
    Commented Jul 26, 2023 at 6:26
  • 1
    \$\begingroup\$ I could already recognize a Harper answer from the first few lines (reinstate neutral!). Did you mean aluminium is 10 times cheaper for the same ampacity? Otherwise nobody would use it \$\endgroup\$ Commented Jul 26, 2023 at 9:12
4
\$\begingroup\$
  1. Aluminium is soft. Copper isn't the toughest of metals, but it is when compared with aluminium.
  2. Aluminium oxidises on contact with air. The oxide layer is thin, but it's an insulator. You have to break through it to make contact.
  3. If you mix dissimilar metals, and damp gets in, then one of those metals will suffer from electrolytic corrosion.

Screw terminals for wires are often made of brass. It's tough, and it's mostly made of copper. So if you put a copper wire into a brass terminal, and do the screw up reasonably tight, it will make good contact for many years.

Put an aluminium wire in the same contact, and you'll have to do the screw up tight to break through the oxide. That crushes the soft wire, making it liable to snap. Then damp gets in, and the aluminium starts to corrode. It now gives a bad contact. So you do the screw up tighter, and the already corroded wire snaps.

You can't make the screw terminals out of aluminium instead; it's too soft.

Where aluminium was used for house wiring and telephone cables in the 1970's to 80's it's now failing. It's good for large cables when used with specialised terminals, particularly crimp terminals. But for smaller wire gauges, it's terrible stuff. UK wiring standards now ban it for the sizes of cables you would find in houses.

\$\endgroup\$
0
\$\begingroup\$

Some of the problem is history... bad history...

Aluminum wires are larger, softer, and more brittle than copper, leading to poorer connections and connection failures. Aluminum’s oxidation process also creates a surface coating with high electrical resistance. That resistance is one factor in creating excess heat buildup in the connection. It is this breakdown of the connections at outlets, light switches, and other junctions that can create enough heat to start a fire.

A study done by the U.S. Consumer Product Safety Commission (CPSC) revealed that “homes wired with aluminum wire manufactured before 1972 (‘old technology’ aluminum wire) are 55 times more likely to have one or more connections reach ‘Fire Hazard Conditions’ than a home wired with copper.” A typical home can have 200 or more connections, exponentially increasing the risk of fire. (Source)

Did you notice the phrase "'old technology' aluminum wire"? That's part of the problem. The industry has overcome most of the issues involving aluminum in electrical conductors, perhaps all of them... except one.

People have really long memories when it comes to bad experiences.

Why does the U.S. avoid nuclear energy? Because we all remember Three Mile Island and refuse to believe things have improved. Why isn't aluminum used in, e.g., house wiring? Because we all remember the bad days when houses either burned down or the wire became brittle and broke apart inside the insulation forcing expensive wire-pulling-and-replacing contractor visits.

Never underestimate a human being's ability to refuse to let go of the past.

\$\endgroup\$

Not the answer you're looking for? Browse other questions tagged or ask your own question.