Practicable Limits to GFCI Recept Protection

[reddy kw]

What are limits of downstream protection provided by 20A GFCI receptacles? Is GFCI recept. protection limited by distance from GFCI recept. to protected outlet? I have tried manufacturer's websites but haven't found pertinent info.
I've done service calcs. (i.e. 180 v-a per single/multiple recept. per yoke, 90 v-a per receptacle in Plugmold, etc.). I've figured in branch-circuit load, wire size, voltage drops and OCPD's. I've read and re-read the 2008 NEC.
Excluding requirements by local codes, and issues of the local AHJ, nothing seems to specify the max. number of downstream receptacles I can protect with a 20 amp GFCI recept. in unfinished basements, garages, outbuildings, etc.

What is the practicable limit in terms of (1) How many downstream outlets can be protected and (2) How many feet from the GFCI can the last protected outlet be?

 

[roger]

What is the practicable limit in terms of (1) How many downstream outlets can be protected

As many as you want.

and (2) How many feet from the GFCI can the last protected outlet be?

This would be a manufacturers issue, if the manufacturer doesn't address it, then there is no distance however, at some length there will be enough leakage to make the decision for you.

Roger

 

[reddy kw]

As many as you want.

This would be a manufacturers issue, if the manufacturer doesn't address it, then there is no distance however, at some length there will be enough leakage to make the decision for you.

Roger

Thanks for the reply, Roger. That's what I though.
I've heard about limits on GFCI receptacles for years, in non-NFPA/NEC books. Many electricians seem to believe there are limits, but popular belief doesn't make it true.
It makes it an urban myth.

 

[480sparky]

I recall years ago there being a limit on both the number and the distance some manufacturers placed on their GFCIs.

 

[Jraef]

I wouldn't imagine any mfr is going to give you a hard number, because it depends on a lot of factors beyond their control. Even if a 1000' run worked today, 10 years from now as the insulation begins its long slow slide into oblivion (like we all do), the GFI may pick that leakage current up some day. And also, lets say you put 50 receptacles down stream. Plug in 50 incandescent light bulbs and no problem. But make 40 of them CFLs and the cumulative effect of the small amount of leakage in each little SMPS inside of them can be enough for the GFI to notice and trip. So if the mfr gives you a number and you have nuisances cause by one of the above, they may have to stand by that number.

 

[Electric-Light]

I wouldn't imagine any mfr is going to give you a hard number, because it depends on a lot of factors beyond their control. Even if a 1000' run worked today, 10 years from now as the insulation begins its long slow slide into oblivion (like we all do), the GFI may pick that leakage current up some day. And also, lets say you put 50 receptacles down stream. Plug in 50 incandescent light bulbs and no problem. But make 40 of them CFLs and the cumulative effect of the small amount of leakage in each little SMPS inside of them can be enough for the GFI to notice and trip. So if the mfr gives you a number and you have nuisances cause by one of the above, they may have to stand by that number.

CFLs do not matter. It is only connected between neutral and hot, so leakage between the two will balance out, thus not detected.

Problematic loads are grounded SMPS that uses triangular arranged filter capacitor

N-||-G-||-H

If the filter capacitor was 0.01uF, and you have 100 of these power supplies connected, you effectively have 1uF between hot and ground, which may conduct enough current to trip some GFCI.

Another issue is inductive load that kicks back, such as magnetic rapid start and electronic ballasts with passive inductive power factor correction on front end.

 

[ElectricianJeff]

What are limits of downstream protection provided by 20A GFCI receptacles?

I always thought it was limited to the number of stickers in the box.

 

[reddy kw]

I recall years ago there being a limit on both the number and the distance some manufacturers placed on their GFCIs.

480sparky-That's exactly why I started wondering about this. New GFCI instructions don't say anything about limits. I too remember seeing manufacturers suggesting limits years back (been on the job since '86, things have changed).
Until 2003, I worked under NYC Electrical Code (at time, an ammended version of 1987 NEC). I got used to having maximum numbers for everything, like number of boxes on a circuit. Thought maybe there are limits for GFCI's.
Had to re-learn everything when I moved upstate. Thanks.

 

[steve66]

I know this is for a breaker, but if a breaker has a limit, you would think a receptacle would be similar:

View attachment 4985

Steve

 

[reddy kw]

I know this is for a breaker, but if a breaker has a limit, you would think a receptacle would be similar:

View attachment 4985

Steve

STEVE-Thanks! I think that's the citation I recalled. I expect limit to length of wire run would be similar for GFCI recept.

 

[reddy kw]

I wouldn't imagine any mfr is going to give you a hard number, because it depends on a lot of factors beyond their control. Even if a 1000' run worked today, 10 years from now as the insulation begins its long slow slide into oblivion (like we all do), the GFI may pick that leakage current up some day. And also, lets say you put 50 receptacles down stream. Plug in 50 incandescent light bulbs and no problem. But make 40 of them CFLs and the cumulative effect of the small amount of leakage in each little SMPS inside of them can be enough for the GFI to notice and trip. So if the mfr gives you a number and you have nuisances cause by one of the above, they may have to stand by that number.

Jraef-Thanks for reply. I know what you mean about the long, slow slide. Give Sq D props for being up front about their GFCI's limitations.
I recall reading GFCI receptacles should be replaced after 15 years for max. reliability (don't recall where I read this). Reliability of circuit breakers is also finite. Calibration goes south. Unfortunately neither get replaced 'til they completely fail.
IMHO, insulation that starts to break down after 10 years is poor quality, or not installed correctly, or pulling too much current.
Side Note: In NYC you'll see 50-100 yr. old hi-rise apartment house electrical systems (run in RMC, in masonry walls) still working perfectly. Wiring only gets replaced when people need more power for A/C, IT equip., kitchen appliances, etc.. We'd pull new conductors into existing pipes using old conductors as pull lines. Old insulation would be in great shape, it was tough.
I haven't had issues with CFL's (or other electronics) on GFCI or AFCI circuits. I separate neutral and ground wires for each circuit. Each ungrounded cond. has dedicated grounding and neutral cond's. Splicing multiple circuits' grounding cond's. together every JB causes ground loops. Cross-connected neutrals make current flow out but not return on same path. Either condition trips GFCI's and AFCI's.
BTW, I'd never install forty or fifty outlets of any type on one circuit.

 

[mlnk]

Plastic romex staples vs. steel staples also makes a difference. I have seen a GFCI trip when showers were in use in 2 bathrooms. Added additional GFCI receptacle to solve problem.