The electrician who replaced our range circuit spent longer on the last thirty seconds of the job than I expected. He had the wire in the breaker, the breaker on the bus, everything obviously done — and then he pulled a stubby wrench out of his bag, set something on the handle of it, and turned the terminal screw until the thing clicked. One click. He put it away and moved on. I asked whether the screw had been loose. "No," he said, "it was tight. Tight isn't the spec." That sentence stuck with me, because until then I had assumed the goal of tightening a screw was to tighten the screw.
The failure that has no moving parts
A breaker panel is a box full of stationary metal. Nothing in it slides or spins, and there is no fluid to leak, which makes it easy to assume that a panel installed correctly stays installed correctly. The one thing that does change over years is the pressure at each connection point, and pressure is the entire mechanism by which a terminal works. A conductor is held against a metal surface hard enough that current crosses the boundary without noticing it's there. Reduce that pressure and the boundary starts to matter.
This is why the connections are the part of a panel that fails, and why they fail slowly enough to be invisible. Nothing breaks on the day of installation. What happens instead is that a joint which was merely snug rather than correctly clamped develops a small amount of resistance, and resistance turns current into heat every time the circuit is used.
Why a connection that is merely snug gets hot
A trade bulletin from the National Electrical Manufacturers Association lays out the chain plainly, and it is worth reading in the order it happens. Loose connections from improper tightening lead to increased resistance over time at the termination, which causes a temperature rise there, and the added heat degrades both the termination and the insulation around it. Aluminum and copper exposed to air at a poorly clamped joint can also oxidize, which adds resistance of its own and feeds the same loop. Where there are actual gaps between conductive parts, arcing becomes possible, and that leaves behind a high-resistance connection that runs hotter still. There is a fourth effect that is less dramatic and more common: voltage drop across the bad joint, which leaves equipment running below the voltage it expects and shortens its life without ever tripping anything.
None of that is a sudden event. It is a slow feedback loop in which heat makes the joint worse and the worse joint makes more heat, and the panel's protective devices are no help, because a breaker watches how much current a circuit draws and has no way to notice that some of that current is being wasted at its own terminal screw.
The rule that put a wrench spec in the Code
The requirement to follow a manufacturer's torque figure is not new. Section 110.3(B) of the National Electrical Code, which first appeared in the 1975 edition, requires that equipment be installed according to the instructions included in its listing or labeling — and the torque values a manufacturer publishes are part of those instructions. In practice that rule was widely ignored, and the number of termination failures made it obvious. The 2017 Code responded by adding 110.14(D), an explicit requirement to use a torque tool where the equipment manufacturer provides a numeric value.
The 2020 edition refined it. Now titled Terminal Connection Torque, the section requires that proper torque be achieved by an approved means wherever the manufacturer has specified a value, and an informational note offers examples of what counts: torque tools, and devices such as shear bolts or breakaway-style fasteners with visual indicators. The through-line across all three of those revisions is the same idea my electrician expressed in four words. A number exists; hitting it is the job.
Where the number comes from
The torque value belongs to the equipment, not to the screw. Panel and breaker manufacturers print theirs on the device, on a label inside the enclosure, or in the instruction sheet that came in the box — which is one small argument for keeping panel paperwork with the house records instead of throwing it away with the packaging. When a manufacturer's value genuinely cannot be found, the Code carries a fallback in Annex I, a set of torque tables drawn from the UL 486A-486B standards for wire connectors. Those tables are explicitly a last resort, applicable only when the equipment's own values are unavailable and cannot be obtained. The order of preference matters: the device's number first, the generic table only when there isn't one.
Why going around re-tightening everything is the wrong instinct
Learning about torque produces an immediate and bad idea, which is to open the panel and snug up every screw in it. Set aside that this means working in an energized panel, which is not homeowner territory under any circumstances. It is also the wrong maintenance procedure. NEMA's guidance is that re-torquing terminations after the initial correct application is not recommended, because the second application of force does not restore a joint and can distort one that was fine. The recommended way to find bad connections is not mechanical at all: infrared scanning and visual inspection, which detect the heat and the discoloration a failing joint produces without touching it. Where a mechanical check is genuinely necessary, the approach is a torque tool set to ninety percent of the specified value, applied to a de-energized circuit, purely to detect looseness rather than to fix it.
There is a maintenance document behind this too. NFPA 70B, the Recommended Practice for Electrical Equipment Maintenance, addresses threaded connections and terminations in its section 8.11, covering both the initial installation and what to do afterward. It's written for people who maintain electrical systems professionally, which is a decent hint about who should be opening the cover.
What a homeowner can actually notice
You cannot see a terminal screw with the panel cover on, which is the correct state for the cover to be in. What you can notice from outside is the heat that a bad joint eventually produces:
- A spot on the panel cover that is warm to the back of your hand when the rest of the cover is at room temperature, especially if it stays warm at the same place.
- A smell of hot plastic or a faint scorched odor near the panel, at any strength, at any time.
- A buzzing or crackling that comes from the panel itself rather than from a device on a circuit.
- Lights on one circuit that dim noticeably when a large appliance starts, when other circuits are unaffected — a classic signature of voltage drop at a single bad connection.
- A breaker that trips with nothing unusual plugged in, or one that feels warm compared with its neighbors.
- Any discoloration, browning, or soot visible at the edge of the cover or around the panel door.
The last one and the smell are the two that mean stop reading and make a call.
What to ask when someone works in your panel
The useful move here is a question, not a task. When an electrician is adding a circuit, swapping a breaker, or replacing the panel, it is entirely reasonable to ask whether they torque terminations to the manufacturer's values and what they use to do it. A qualified one will treat that as a normal question and may well show you the tool, the way mine did without being asked. It is also worth asking after a service upgrade or a panel replacement whether the terminations were torqued to spec, and keeping the answer with your records. You are not inspecting anyone's work — you are signaling that you know the difference between a screw that feels tight and a connection that was made to a number, and that difference is quietly responsible for a lot of what goes wrong inside a panel that otherwise looks perfect.