Arc faults

[mshields]

My understanding is that Equipment ground fault is only required on 480Y277V (for all intent and purpose) but not for 208Y120V circuits BECAUSE in a 208Y120V system, any arc produced will be extinguished at the zero crossing of the voltage. Put another way, 120V cannnot sustain an arc.

So why then, do we have these things called AFCI circuit breakers. Clearly these are designed to detect just such an arc. What am I missing

 

[mbrooke]

My understanding is that Equipment ground fault is only required on 480Y277V (for all intent and purpose) but not for 208Y120V circuits BECAUSE in a 208Y120V system, any arc produced will be extinguished at the zero crossing of the voltage. Put another way, 120V cannnot sustain an arc.

So why then, do we have these things called AFCI circuit breakers. Clearly these are designed to detect just such an arc. What am I missing

If you were to re-name sputtering or delayed fault clearing as arcing, you have the "arc" in AFCI breakers.

Once upon a time it was realized that a short circuit clearing on thermal trip vs magnetic trip persisted for just a little while longer. It was theorized that the increased incident energy at the fault point was behind residential electrical fires.

As such the magnetic pickup for residential circuit breakers was lowered from 25x down to 7-10x. This provides arc fault protection for the whole circuit in most cases.

Then comes along the poorly substantiated claim that some rural property out in the boonies might have only 500 amps of fault current at the panel, which would limit the short circuit current on a circuit down to 75 amps.

A 75 amp amp magnetic pickup breaker was proposed, but soon realized normal inrush would trip it. Ironically to prevent nuisance tripping the modern AFCI was born.

 

[xptpcrewx]

My understanding is that Equipment ground fault is only required on 480Y277V (for all intent and purpose) but not for 208Y120V circuits BECAUSE in a 208Y120V system, any arc produced will be extinguished at the zero crossing of the voltage. Put another way, 120V cannnot sustain an arc.

So why then, do we have these things called AFCI circuit breakers. Clearly these are designed to detect just such an arc. What am I missing

Arcing faults exist for all voltage levels. It doesn’t just stop happening below a certain nominal voltage. AFCI and GF protection operate on different principles. GF isn’t looking for arcing faults, only zero sequence current.

 

[mbrooke]

Arcing faults exist for all voltage levels. It doesn’t just stop happening below a certain nominal voltage.

Below about 235 volts to ground arcing ceases at typical atmospheric pressures.


There is operational experience from utilities that arcing can not be sustained at 120 volt (please forgive the size, I don't know how to shrink them)







Testing done by UL has confirmed the same unless the path is heavily carbonized, typically by applying 15kv first as done in 1699 testing.

 

[tom baker]

My understanding is that Equipment ground fault is only required on 480Y277V (for all intent and purpose) but not for 208Y120V circuits BECAUSE in a 208Y120V system, any arc produced will be extinguished at the zero crossing of the voltage. Put another way, 120V cannnot sustain an arc.

So why then, do we have these things called AFCI circuit breakers. Clearly these are designed to detect just such an arc. What am I missing

Thats what I learned is 120 V can't sustain an arc but 277 V can, which is why there is special bonding for ckts over 250 to ground. AFCI breakers detect the characteristic signature of a arc fault, parallel, series, I guess they do, but its hotly debated.

 

[synchro]

My understanding is that Equipment ground fault is only required on 480Y277V (for all intent and purpose) but not for 208Y120V circuits BECAUSE in a 208Y120V system, any arc produced will be extinguished at the zero crossing of the voltage. Put another way, 120V cannnot sustain an arc.

So why then, do we have these things called AFCI circuit breakers. Clearly these are designed to detect just such an arc. What am I missing

AC arc welders typically run at about 30 arc volts and so arcs can certainly exist at these lower voltages. Arcs often have a negative resistance region in their I vs. V characteristic, and so welders effectively provide a series impedance (e.g., reactance) that is larger in magnitude than this negative resistance so that it stabilizes and sustains the arc. Without a sufficient series impedance the current may quickly get high enough that it extinguishes the arc by blowing it out with magnetic forces or destroying the electrode. Fluorescent and HID lamps have plasma arc discharges that exhibit a negative resistance and therefore use a ballast to stabilize their operation.

I think in the ground faults we are talking about above there is a very low source impedance and so arcs are likely to be very short lived unless the voltage is high enough to sustain a relatively large arc that will have a higher resistance, and therefore limit the current to a sustainable value. Apparently 120V may not be enough to do this in some cases. Another factor is that arcs can be "quenched" by basically taking away too much of their heat for them to be sustained, for example by distributing the arc over a relatively large surface of metal. This method was sometimes used in "arc" transmitters of more than a century ago to extiguish the arc between the "dots" and "dashes" of Morse code. So perhaps 480V arcs can be long enough that most of the arc is further away from metal so that it's not taking as much heat energy away. This is speculation based on what I know, but I haven't seen any real comparative data on it.

And so I think "series" type arcs where the load limits the current could be sustained relatively easily at 120V. But so called "parallel" arcs less so. I think that's why the UL arc fault tester needed a neon sign transformer to initiate the arc.

 

[romex jockey]

And so I think "series" type arcs where the load limits the current could be sustained relatively easily at 120V. But so called "parallel" arcs less so. I think that's why the UL arc fault tester needed a neon sign transformer to initiate the arc.

exactly why the original NEMA-afci task force disbanded, with it's then chair Joe Englel forwarding an explanation to CMP-2 in the '12 cycle, along with NFPA EE's like>

~RJ~

 

[ramsy]

So why then, do we have these things called AFCI circuit breakers. Clearly these are designed to detect just such an arc. What am I missing

Outlet bombs like this <= 5-Amps of computer loads with unlisted appliances.

All 3 prongs of appliance cord & plug were black. No UL, ETL, or NRTL markings of any kind, anywhere on surge-protecting power strips, purchased from amazon.com

Some kind of dissimilar metals caused chemical reactions that eventually increased resistance & burned up.

The originally listed bakelite firebox was the only thing that saved this mobile home tinderbox from burning down.

After repair, and complying with replacement code, NEC 406.4(D)(4), the new AFCI outlet stands guard ready to detect unqualified persons tampering with the electrical.

When it smells un-qualified tampering, it trips & won't reset untill qualified persons arrive.

 

[don_resqcapt19]

AC arc welders typically run at about 30 arc volts and so arcs can certainly exist at these lower voltages. Arcs often have a negative resistance region in their I vs. V characteristic, and so welders effectively provide a series impedance (e.g., reactance) that is larger in magnitude than this negative resistance so that it stabilizes and sustains the arc. Without a sufficient series impedance the current may quickly get high enough that it extinguishes the arc by blowing it out with magnetic forces or destroying the electrode. Fluorescent and HID lamps have plasma arc discharges that exhibit a negative resistance and therefore use a ballast to stabilize their operation.

I think in the ground faults we are talking about above there is a very low source impedance and so arcs are likely to be very short lived unless the voltage is high enough to sustain a relatively large arc that will have a higher resistance, and therefore limit the current to a sustainable value. Apparently 120V may not be enough to do this in some cases. Another factor is that arcs can be "quenched" by basically taking away too much of their heat for them to be sustained, for example by distributing the arc over a relatively large surface of metal. This method was sometimes used in "arc" transmitters of more than a century ago to extiguish the arc between the "dots" and "dashes" of Morse code. So perhaps 480V arcs can be long enough that most of the arc is further away from metal so that it's not taking as much heat energy away. This is speculation based on what I know, but I haven't seen any real comparative data on it.

And so I think "series" type arcs where the load limits the current could be sustained relatively easily at 120V. But so called "parallel" arcs less so. I think that's why the UL arc fault tester needed a neon sign transformer to initiate the arc.

That arc is manually maintained by moving the rod. Without movement the arc goes out or the rod welds itself to the work.

 

[don_resqcapt19]

View attachment 2556796
Outlet bombs like this <= 5-Amps of computer loads with unlisted appliances.

All 3 prongs of appliance cord & plug were black. No UL, ETL, or NRTL markings of any kind, anywhere on surge-protecting power strips, purchased from amazon.com

Some kind of dissimilar metals caused chemical reactions that eventually increased resistance & burned up.

The originally listed bakelite firebox was the only thing that saved this mobile home tinderbox from burning down.

After repair, and complying with replacement code, NEC 406.4(D)(4), the new AFCI outlet stands guard ready to detect unqualified persons tampering with the electrical.

When it smells un-qualified tampering, it trips & won't reset untill qualified persons arrive.

There is no reason to suspect that is a picture of the effects of an arcing fault. The joule heating from a high resistance connection will create the same damage. There is nothing on the marker that will directly detect a high resistance, high heat producing connection.

 

[mbrooke]

View attachment 2556796
Outlet bombs like this <= 5-Amps of computer loads with unlisted appliances.

All 3 prongs of appliance cord & plug were black. No UL, ETL, or NRTL markings of any kind, anywhere on surge-protecting power strips, purchased from amazon.com

Some kind of dissimilar metals caused chemical reactions that eventually increased resistance & burned up.

The originally listed bakelite firebox was the only thing that saved this mobile home tinderbox from burning down.

After repair, and complying with replacement code, NEC 406.4(D)(4), the new AFCI outlet stands guard ready to detect unqualified persons tampering with the electrical.

When it smells un-qualified tampering, it trips & won't reset untill qualified persons arrive.

Sparking and arcing is the END stage of joule heating. I see nothing in that picture to indicate arcing was the culprit ie carbon path.

Like the industry wants people to believe you've inadvertently come to believe the above is arcing and will be detected by an AFCI.

 

[romex jockey]

When it smells un-qualified tampering, it trips & won't reset untill qualified persons arrive.

LOL, then most of my state (which subscribed to 210.12 in the '98 cycle) should be in the dark w/flashlight in phonebooks ramsy

~RJ~

 

[ramsy]

Remodelers remove AFCI's after inspection, and there are no whistleblowers.

210.12 can't be enforced in this sector, until AHJ is called back for some casualty claim, and cancels insurance for negligence.

My clients demand "Additionally Insured" certificates from their contractor's General Liability policy, or file performance claims against their license to replace AFCI's removed after inspection.

Certificates of additionally insured remain in affect, long after remodel contractors go AWAL, bankrupt, or change corporate names.

 

[romex jockey]

Remodelers remove AFCI's after inspection, and there are no whistleblowers.

210.12 can't be enforced in this sector, until AHJ is called back for some casualty claim, and cancels insurance for negligence.

My clients demand "Additionally Insured" certificates from their contractor's General Liability policy, or file performance claims against their license to replace AFCI's removed after inspection.

Certificates of additionally insured remain in affect, long after remodel contractors go AWAL, bankrupt, or change corporate names.

I often feel penalized for being licensed ramsy

~RJ~