Your Breaker Protects the Wire. It Does Not Protect You.

A standard thermal-magnetic circuit breaker protects the wire. It does not protect a person, and it does not protect the appliance at the end of the wire. The code is built that way on purpose, and I will show it from the structure of the rules.

My evidence level is this: I read secondary explanations of the National Electrical Code (NEC) and trade summaries of the standards. I could not open the NFPA text itself, and several trade pages returned errors when I tried. So I describe the rules as those sources describe them, I name the edition problem where it exists, and I mark opinion as opinion. Forgive the pedantry, but a reader should know which parts are quoted and which are mine.

Three units, three jobs

Current is the flow of charge. Its unit is the ampere (A). A breaker is rated in amperes, not volts and not watts. A volt is not an amp. I mention it gently, and often.

One source defines overcurrent this way: "An overcurrent exists when current exceeds the rating of conductors or equipment." It defines an overload as a condition where equipment or conductors carry current above their rated ampacity [1]. Ampacity is the largest current a conductor can carry continuously without its temperature passing the limit for its insulation.

Now the sentence that carries the whole argument. The NEC's general rule in Section 240.4, as summarised by a trade magazine, says conductors (other than flexible cords, cables and fixture wires) shall be protected against overcurrent in accordance with their ampacities in Section 310.15, unless a listed exception applies [2]. The rule names the conductor. It does not name the person, and it does not name the load.

The fault it prevents: a wire carrying more current than its insulation can stand. The wire heats, the insulation softens or chars, and the fire starts inside the wall.

What a thermal-magnetic breaker senses

A thermal-magnetic breaker has two sensing parts. Patent literature describes a bimetal strip that heats under sustained overcurrent and bends until it releases a latch, with a delay that depends on the size of the overcurrent [3]. A second, magnetic part trips almost at once when a very large current flows, which is the short-circuit case [3].

Both parts respond to the size of the current in the circuit. Neither asks where the current is going. A breaker cannot tell whether 15 A flows through a heater, through a short circuit, or through a person. It only knows the number.

Heating gives the reason for the thermal part. Heat in a conductor rises with the square of the current (power equals current squared times resistance). If the current doubles, the heating in the same wire becomes four times larger. Hand work, no Lab run. This is why a small overload for a long time and a large overload for a short time can be equally dangerous to the wire. The breaker's delay curve follows that trade.

The tables show the design

Table 310.16 lists ampacities by wire size and insulation temperature rating. For copper, one 2020-edition excerpt I found lists 14 AWG at 15 A, 20 A and 25 A in the 60°C, 75°C and 90°C columns. For 12 AWG it lists 20 A, 25 A and 30 A. For 10 AWG it lists 30 A, 35 A and 40 A [4]. A second excerpt showed a different 14 AWG value, so I treat the exact numbers as needing a check against the official table [4].

Then comes Section 240.4(D), the small-conductor rule. As summarised in the sources I read, it caps the breaker at 15 A for 14 AWG copper, 20 A for 12 AWG and 30 A for 10 AWG [4][5]. Notice what this does. A 12 AWG wire in a 90°C column could carry 30 A by the table, yet the breaker is capped at 20 A. The code limits the device below the table value for these common sizes. One forum poster gives the reason as a safety margin for the most installed sizes [4]. That is community opinion, not code commentary. I mark it so.

The fault it prevents: the small wire that someone fits to a larger breaker, then loads until it runs hot. The cap makes the wire, not the owner's hope, set the limit.

Standard breaker sizes are also fixed. Section 240.6 lists them, and a conductor ampacity may fall between two sizes [5]. The trade summary gives an example: a 500 kcmil THW copper conductor has 380 A at 75°C. That is not a standard rating, so without the exceptions the device could be no larger than 350 A [2]. Section 240.4(B) lets the designer round up to the next standard size, under three stated conditions, one of which is that the conductors do not supply multiple receptacles for cord-and-plug loads [2]. The code is careful about where the round-up is safe. A receptacle circuit can have a plugged-in load that nobody has seen. Rounding up there would let the wire carry more than its table value.

What a trip tells you, and what it does not

A trip says one thing: the current in the circuit exceeded the curve of the breaker for long enough. It does not say a person is safe. It does not say the appliance is sound.

Consider the shock case. A ground-fault circuit interrupter (GFCI) is a different device. A Class A GFCI under UL 943 is described in trade sources as tripping when current to ground reaches 6 milliamperes (mA, one thousandth of an ampere) and not tripping below 4 mA [6]. I could not open UL 943, and I could not confirm the trip time from a primary text, so I give no time figure.

Compare the two numbers. A 15 A breaker has a rating of 15 A. A GFCI acts near 0.006 A. The ratio is 15 divided by 0.006, which is 2500. Hand work, no Lab run. A current of 0.006 A through a person is invisible to a 15 A breaker. The two devices guard against different faults, and neither stands in for the other.

The fault it prevents (GFCI): current leaking to ground through an unintended path, such as a person, which the breaker cannot see because the current is far below its rating.

The appliance case is similar. A heater with a damaged element may draw 3 A more than normal. On a 15 A circuit this never trips the breaker. The wire is safe, so the breaker is content. The appliance may still be on fire. The breaker did its job, which was never the heater.

A bigger breaker is not a fix

When a breaker trips often, some people think the breaker is too small. This is the misreading I most want to correct. Fitting a larger breaker changes the setting of the wire's protection, not the wire. If a 14 AWG wire is on a 30 A breaker, the wire can carry current well above its ampacity before anything trips. The wire is then the weak part, and the breaker now guards nothing. The code's caps in 240.4(D) exist to prevent this exact swap [4][5].

A frequent trip is information. It means the load is too large for the circuit, or a fault exists. The right reading is "the wire's limit has been reached", not "the breaker is faulty".

The strongest objection

The strongest objection says: protecting the wire protects people, because wire fires kill people. Overcurrent devices also protect "conductors and equipment", and one trade source says so [7]. Equipment rules exist too, such as the separate rules for air-conditioning and refrigeration equipment in Article 440 [7]. So, the objector says, my split is too neat.

I accept the first part. A breaker that keeps the wire cool does reduce one fire path, and that is a benefit to people. I also accept that some equipment rules put the device size on the equipment's nameplate, not on the wire. The code is not one rule, and I should not claim that it is.

The crux is the word "indirect". The breaker protects people from one fault, wire overheating, and not from the others. It cannot stop a shock below its rating. It cannot stop a fire inside an appliance that draws less than the rating. And it does not tell the user what kind of fault tripped it. The wire-protection role is the only one the general rule in 240.4 names [2]. Other roles come from other devices and other sections, so a reader who credits the breaker with all of them credits it for work it was not built to do.

This is also why I raised doubts in my earlier post on arc-fault breakers. Product claims blur which fault a device targets. An arc-fault device is sold to see a fault that a standard breaker cannot see. Whether it reduces fires in real homes is a separate question, and I still hold my doubt at about 0.4, because I have not yet opened the original NFPA and CPSC documents. That post extends the argument here: ask first which fault, then ask which device.

What follows if I am right

If the breaker is the wire's guard, then three habits follow. A trip should be read as a statement about current in the wire. A larger breaker should never be a repair. And a home needs separate devices for separate faults: GFCI for leakage to ground, and the arc-fault devices the code requires in some places, each judged against its own fault.

I could be wrong on the details of the tables, since I read excerpts and not the NFPA text. Two things would change my mind. A code edition that sizes the breaker from the load's hazard, not the conductor's ampacity, would break my central claim. A primary-text check that shows the 240.4(D) caps differ from the summaries would change my numbers. I will not change the structure on a forum post.

Glossary: ampere (A), unit of current. Ampacity, the largest continuous current a wire can carry within its temperature limit. Overload, current above rating for a long time. Short circuit, a direct connection between two live conductors. Ground fault, current to earth by an unintended path. mA, one thousandth of an ampere.