The definition of GFCI

[cmckeehan]

In the 2017 edition of the NEC the definition of a GFCI includes, it should trip in the established time, what is the established time or where can I find it? Everything I read points to immediately or one-tenth of a second, however, I cannot find it in the code book.

 

[Dennis Alwon]

I am not sure where that is but it is not in the code book. I suspect it may be in the info with the gfci or it is an ANSI regulation

 

[roger]

I believe one-tenth is supposed to be the maximum time. This OSHA article says they can trip in as little as one-fortieth of second.

Roger

 

[Dennis Alwon]

I believe one-tenth is supposed to be the maximum time. This OSHA article says they can trip in as little as one-fortieth of second.

Roger

So the question remains-- it this an OSHA reg.Z? or do we know what org is responsible for the under 1/10 of a sec. requirement

 

[roger]

So the question remains-- it this an OSHA reg.Z? or do we know what org is responsible for the under 1/10 of a sec. requirement

I'm not sure where the requirements come from except what was established to be somewhat safe. I recall reading something in an article about the evolution of the GFCI but can't remember where it was.

Roger

 

[tom baker]

It would be in the UL standard.

 

[synchro]

The UL 943 standard for GFCI's specifies a curve of the time duration vs. current below which the device must trip. But apparently most manufacturers don't implement such a curve and trip it about just as quickly at low currents as high currents. This is more conservative and if anything safer than required, but probably leads to more nuisance trips.

 

[Dennis Alwon]

It would be in the UK standard.

Huh, why in the UK standard? They use RCD not gfci

 

[tom baker]

Huh, why in the UK standard? They use RCD not gfci

Sorry typo, UL

 

[don_resqcapt19]

If they took the maximum time permitted by the UL standard to trip, you would have to hold the test button in for about 5 seconds or so. The requirement from UL 943 says:
"The maximum permitted time to trip in seconds is equal to the quantity (20/fault current in milliamps) raised to the 1.43 power. The application of this formula would permit a 7 second trip time for a 5 mA ground fault."

 

[jim dungar]

"The maximum permitted time to trip in seconds is equal to the quantity (20/fault current in milliamps) raised to the 1.43 power. The application of this formula would permit a 7 second trip time for a 5 mA ground fault."

I thought they had a maximum trip time and the formula was used for faults >5mA, but it has been years since I looked at the standard.

 

[Jraef]

If they took the maximum time permitted by the UL standard to trip, you would have to hold the test button in for about 5 seconds or so. The requirement from UL 943 says:
"The maximum permitted time to trip in seconds is equal to the quantity (20/fault current in milliamps) raised to the 1.43 power. The application of this formula would permit a 7 second trip time for a 5 mA ground fault."

It should be noted though that the MAXIMUM allowable is just what it says. Most of the Class A GFCI breakers on the market are actually much faster than that, as in 100msec. to react + opening time, which for a typical breaker is about 1.5 cycles, so 25msec, for a total of 125msec total clearing time, regardless of the fault current level. I don't know if GFCI receptacles are faster, I only looked at the test data on breakers when I was at Siemens. But when they established test bench criteria, 150msec clearing time was considered a fail.

Still though, even that "maximum allowable" is not dangerous. 10mA is the "let-go" threshold, which is why we (in North America) insist on a margin of safety by targeting 5mA as the goal (4mA minimum, 6mA maximum). 40mA is the average cardiac fibrillation threshold, so even if you go get hit with 6mA for 7 seconds, it's not going to kill you or make you lock on.