The fire that starts while nobody is home
Here is the uncomfortable part about aluminum branch-circuit wiring: the connection that eventually starts a fire was tightened correctly. A careful electrician torqued that outlet screw to spec in 1970, the receptacle worked for decades, and no one touched it again. And that is exactly the problem. The failure didn't come from a mistake. It came from the metal doing what aluminum does, slowly, in the dark, every time you ran the toaster and every time the house cooled off at night.
Most people assume a wiring fire means a wire got overloaded and melted. That is almost never what happens with aluminum. The wire itself is fine. It's the joint — the half-inch where aluminum meets a brass screw or a copper pigtail — that overheats. Understanding why means understanding a handful of ordinary physical facts that, stacked together, turn a good connection into a bad one without anyone lifting a finger.
Aluminum grows and shrinks more than the screw holding it
Every metal expands when it heats and contracts when it cools. Aluminum does it more than copper, and much more than the brass and steel of a terminal screw. Its coefficient of thermal expansion is roughly 30–40% higher than copper's. That difference sounds academic until you remember what a wire connection actually experiences.
Run a hair dryer on that circuit and current heats the conductor. The aluminum swells against the screw. Turn the dryer off and the aluminum shrinks back — but it doesn't return to exactly where it was. It was clamped under pressure, and softer metals under sustained pressure creep: they flow, very slowly, out from under the load. Electricians call it cold flow. Each heating cycle nudges the aluminum a little further out from under the screw, and each cooling cycle leaves the clamp a hair looser than before. Thousands of cycles later, the connection that was torqued perfectly is now loose enough to arc.
The invisible oxide layer that fights the current
There's a second mechanism working alongside the first, and it's arguably worse. The moment bare aluminum meets air, it grows a skin of aluminum oxide. This happens in seconds and you cannot stop it. Aluminum oxide is hard, it's tenacious, and — critically — it is an electrical insulator.
Copper's oxide is a mediocre conductor and mostly harmless. Aluminum's oxide is essentially glass. So current at an aluminum joint has to find its way through microscopic breaks in that oxide film. Fewer contact points means higher resistance at the joint, and higher resistance means more heat for the same current. More heat means more thermal expansion, which means more cold flow, which loosens the joint, which shrinks the contact area, which raises resistance again. It is a self-reinforcing loop — a slow-motion thermal runaway that can push a connection past 150°F, then far higher, while the breaker never sees enough current to trip. A breaker protects the wire from overload. It has no idea a single terminal is glowing.
Dissimilar metals and a little bit of moisture
Stack one more factor on top. When two different metals touch in the presence of even trace moisture, you get a tiny galvanic cell — the same electrochemistry as a battery. Aluminum is anodic to copper and brass, meaning at that junction the aluminum preferentially corrodes. That corrosion builds more resistive gunk exactly where you need clean metal-to-metal contact. This is why simply twisting an aluminum wire together with a copper one under a standard wire nut is a genuinely dangerous repair, and why devices made for the job are marked for it.
Why modern aluminum feeders are fine — and old branch circuits aren't
If aluminum is this troublesome, why does the utility feed your whole house with it, and why do electricians still pull big aluminum feeders to sub-panels and ranges every day? Because the problem lives almost entirely in small, solid, single-strand aluminum used for 15- and 20-amp branch circuits, mostly installed in North American homes between roughly 1965 and 1973 when copper got expensive.
Large aluminum feeders use a different, more forgiving alloy (AA-8000 series), they terminate on lugs designed and rated for aluminum, they get treated with antioxidant compound, and they're torqued to a spec on a device built to hold them. Done that way, aluminum is safe and completely code-compliant. The danger isn't the element. It's small vintage aluminum landing on outlets and switches that were never designed to hold it.
How to tell what you've got
You don't need to open a wall. Pull the cover plate off a receptacle and look at the cable jacket or the conductor. Solid aluminum branch wiring is often printed with AL or ALUMINUM right on the sheathing, and the exposed strands under a terminal will be silvery-white rather than the salmon-pink of copper. Homes built or rewired in that late-'60s-to-early-'70s window are the ones to check. Warm cover plates, outlets that have stopped working, a faint smell of hot plastic, or flickering that isn't the bulb are all reasons to stop using a circuit and get eyes on it.
Your next moves
- Date your wiring. Find out when your home was built or last rewired. If it lands between 1965 and 1973 in North America, treat aluminum branch wiring as likely until proven otherwise.
- Do the cover-plate check today. Kill the breaker, remove one or two receptacle plates, and look at the conductors. Silvery-white and stamped AL means aluminum. Do not start unscrewing terminals if you're not qualified — just look.
- Feel your outlets and plates. Warm faceplates, discolored receptacles, buzzing, or the smell of hot plastic mean stop using that circuit now and call a licensed electrician.
- Insist on the right repair, not a wire nut. Proper fixes are CO/ALR-rated devices, a COPALUM crimp, or a listed connector like the AlumiConn — with antioxidant compound where the manufacturer calls for it. Twisting aluminum to copper under a standard wire nut is not a fix.
- Verify torque on any accessible connection. For large aluminum feeders and lugs, connections should be torqued to the device's printed spec — not "tight," but the actual number — because guessing is how cold flow gets its head start.
When the number is the safeguard
Everything above comes down to getting a few specifics exactly right: whether a device is rated CO/ALR, what antioxidant the lug requires, what the ampacity of that conductor really is at temperature, and the precise torque value stamped on the terminal. Those aren't things to hold in your head on a hot attic afternoon, and they're not things to approximate. That's the quiet work Voltly is built for — ampacity tables, terminal and torque references, and the NEC lookups an electrician actually reaches for, all offline, right when your hands are already full. When the difference between a safe joint and a slow fire is a single spec, it helps to have the number in your pocket instead of your memory.
If you'd rather carry the reference than the doubt, take a look at Voltly.