Contact surface area & household power transmission

Axial heat transfer.

Well they do sometimes :) I had a 15A outlet melt once. A copy machine was plugged in further down the chain.

Mostly it is a case of I squared R where R is very small. Hopefully the breaker will trip before the junction acts like a fuse.

Problem with that theory: The current seen by the circuit breaker never exceeds full load current (this isn't a short). Unfortunately, the junction temperature can rise to levels that can result in a fire.

This is why the NEC requires Arc Fault protection. As these junctions begin to fail, they start arcing.

| I've been pondering a question lately that I figure you guys can help me | with. | | There are, based on my casual observation, many places in household wiring | where the connections within a circuit have a much smaller contact area than | the cross-sectional area of the wire that is required for the same circuit. | | A good example would be the 'quick connect' slots in the back of a | receptacle. Love them or hate them, they are allowed by code in many (most?) | areas, and it seems to me the contact area is extremly small--perhaps | several orders of magnitude smaller than the cross-sectional area of a 14 | gauge wire. | | Another example would be a 100A breaker feeding a subpanel. The breaker | feeds cable the size of a pencil, but clips to the hot bus bar with the same | tiny tabs as a 15A breaker (which seem to have a comparatively small contact | area when compared to the diameter of 2 AWG cable). | | So, it seems to me that, in a circuit running at capacity, there would be a | high concentration of current flowing through the very small contact area in | these apparent 'bottlenecks'...why don't they overheat?

They do get hotter. It's called a hotspot. It can also happen with wire, especially stranded. However, when the heat is sourced at just a spot like that, it can be carried along the bus, wire, or tab, and dissipate into the surrounding environment a lot easier than if the entire length were heating due to a big overload.

Problem with *that* is that the load is still in series with the arc, and the arc current is still limited to the load current of the device. As I understand it, current arc fault interrupters are designed to detect only the high-current arcing you can get with a line to neutral (or ground) fault, not low-current arcs you get with the load still in series with the arc.

Dave

This is true of the older units. The latest generation if AFCIs is supposed to respond to hi-z series arcs.

As the article says, "combination" AFCIs are required as of 1-1-08 by the 2005 NEC. Combination AFCIs detect arcs at a 5A level and will detect series arcs. (Previously the detection level was 75A.) (Old and new AFCIs also have ground fault detection required at 50mA, but usually is at 30mA.)