GFCI BREAKER ON A DELTA SYSTEM

Well, 2 of 3 legs are under 150V to ground.

I wonder if Square D's existing 208/120V three phase hardware would actually be suitable for the OP if the manufacturer took the time to get it tested/ listed for the application.
 
winnie said:
Well, 2 of 3 legs are under 150V to ground.
Click to expand...
Sure, but the OP refers to a "3 phase, 40 amp circuit". The voltage to ground of such a circuit is the highest of the voltages to ground of the 3 legs. So the circuit has a voltage to ground in excess of 150V, and the AHJ has erred in "wanting" the circuit to be GFCI protected.

winnie said:
I wonder if Square D's existing 208/120V three phase hardware would actually be suitable for the OP if the manufacturer took the time to get it tested/ listed for the application.
Click to expand...
Seems like if one wants to add GFCI voluntarily, rather than because of a code requirement, the only question is whether it would work or not, and that listing would not be required.

Cheers, Wayne
 
hornetd said:
I used to do that very thing to one of every 10 circuit breakers we installed in prebuilt communications shelters at Integrated Power Corporation (IPC). We just applied the loads that the manufacturer gave as 2 points on the trip curve, I don't remember which 2 points we used, and timed the opening with a laboratory multimeter with a timing function and memory. What was missing from that testing? Can you explain it so that a factory floor test and installation technician will understand?
Click to expand...
In no way do I see that as practical testing. Sure, you can do that for some special applications, but as I understand the post I replied to, the poster was asking for some built in method of testing the thermal magnetic trip, much like you test a GFCI by pushing a button.
 
tortuga said:
...

My understanding of UL 943 is that they use a formula that states 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 7000 ms (7 second ) trip time for a 5 mA ground fault. ...
Click to expand...
While that is the maximum permitted time to trip, they actually trip much faster than that. If the tripped as the maximum permitted time you would have to hold the test button in for ~ 4 seconds. I have never seen one that did not trip almost instantly when you push the test button.
However you are correct that at the same ground fault current level the RCDs it will trip faster than a GFCI.

However I still fail to understand why the global manufacturers of both RCDs and GFCIs have not submitted proposals to UL 943. It would be in their interest to only have to have one product that could be used world wide. Eaton, Siemens, and Schneider all have people on the UL 943 Technical Panel.
 
don_resqcapt19 said:
While that is the maximum permitted time to trip, they actually trip much faster than that. If the tripped as the maximum permitted time you would have to hold the test button in for ~ 4 seconds. I have never seen one that did not trip almost instantly when you push the test button.
However you are correct that at the same ground fault current level the RCDs it will trip faster than a GFCI.
Click to expand...
Yeah I have never measured how many mA a test button is, It similar with RCD's I have herd they often trip at 20-23mA.
don_resqcapt19 said:
However I still fail to understand why the global manufacturers of both RCDs and GFCIs have not submitted proposals to UL 943.
Click to expand...
I dont think there is a way to know if they had, In my (dated) experience UL is totally closed off to the public, it acts more like a private for profit corporation, unless your on a panel or have a complaint their internal processes are opaque.
NFPA for example has a research library in Quincy MA where anyone from the public can do pretty extensive historical research. Years back I spent some time there researching a code changes from 1969, and it was quite fascinating they have very detailed records available down to meeting minutes.
 
tortuga said:
Yeah I have never measured how many ma a test button is, It similar with RCD's I have herd they often trip at 20-23ma.

I dont think there is a way to know, In my (dated) experience UL is totally closed off to the public, it acts more like a private for profit corporation, unless your on a panel or have a complaint their internal processes are opaque.
NFPA for example has a research library in Quincy MA where anyone from the public can do pretty extensive historical research. Years back I spent some time there researching some code changes from 1969, and it was quite fascinating they have very detailed records available down to meeting minutes.
Click to expand...
Not sure about researching the past with UL, but their standards are ANSI standards and anyone can submit a proposal for a change much like the NEC, but it is even easier for a member of the technical panel to submit proposals for change and 3 global manufacturers have people on the 943 panel. Also, like the NEC, all submitted proposals must be acted on.

As far as the test button resistor, it must cause the device to trip at 85% of rated voltage, but not cause more than 9ma of current to flow at rated voltage, so my guess is that the test resistor is around 14,500 ohms.
 
don_resqcapt19 said:
In no way do I see that as practical testing. Sure, you can do that for some special applications, but as I understand the post I replied to, the poster was asking for some built in method of testing the thermal magnetic trip, much like you test a GFCI by pushing a button.
Click to expand...
Could you say more about that? What aspect of what I described do you find impractical?
 
hornetd said:
Could you say more about that? What aspect of what I described do you find impractical?
Click to expand...
The cost and time to do that to ordinary residential circuits, and even if there was code rule that required such testing there would be no way to enforce such a rule.
 
I may have missed it in previous posts, but as I was taught, the DIFFERENCE in RCD vs GFCI protection levels is a difference in the definition of “safe”. The muscle lock-on current threshold is 10-25mA, but the cardiac fibrillation level is around 50mA. So an RCD might NOT protect you from muscle lock-on, meaning you can’t let go, but will keep you from having it stop your heart. The 5mA GFCI threshold will not even result in a lock-on situation.

So if you TOSS a toaster into the bath with you, both will prevent you from cardiac arrest. But if you GRAB a faulty toaster with one foot in the bathtub, the muscle spasms allowed by the RCD might make you slip and crack your skull on the edge of the tub, whereas the GFCI would have prevented that.

In a litigious society like ours, I can understand why nobody wants to take responsibility for raising the trip threshold. Lawyers would have a field day with that…

Side note: avoid making toast near bathtubs…
 
don_resqcapt19 said:
The cost and time to do that to ordinary residential circuits, and even if there was code rule that required such testing there would be no way to enforce such a rule.
Click to expand...
I'll give you that it might well be unwarranted to do it in a residential situation so I probably should not have raised it in this thread.

I know it is enforceable by using testing instruments that have a memory function and can download that stored information to a computer. In our case we were doing certification of contract compliance so we presented a list of which breakers in which of the transportable prebuilt communications shelters we had made. The people doing acceptance testing for the client only had to pick one breaker in each shelter and check it against our list. If the results were essentially identical they were assured that our testing was adequate to support acceptance. The acceptance testers usually chose a couple of as yet untested breakers to test as an additional quality check.
jim dungar said:
And it is definitely not for the untrained much less 'Joe Homeowner'.
Click to expand...
I certainly didn't mean to imply that testing like this should be done by anyone but a licensed electrician.
 
wwhitney said:
Sure, but the OP refers to a "3 phase, 40 amp circuit". The voltage to ground of such a circuit is the highest of the voltages to ground of the 3 legs. So the circuit has a voltage to ground in excess of 150V, and the AHJ has erred in "wanting" the circuit to be GFCI protected.


Seems like if one wants to add GFCI voluntarily, rather than because of a code requirement, the only question is whether it would work or not, and that listing would not be required.

Cheers, Wayne
Click to expand...
Likely would work, potential problems being GFCI control circuit is likely 120 volts so you need to be sure to not place which ever pole it were derived from on the high leg, second thing may be that it is a 120/240 volt and not a straight 240 volt rated breaker?
 
hornetd said:
I certainly didn't mean to imply that testing like this should be done by anyone but a licensed electrician
Click to expand...
Simply having a license does not make a person qualified to do breaker testing.
It is not rocket science, but some training or certification would be expected.
 
Hello to all and thanks for the replies!





In an interesting twist the inspector is requiring me to obtain a letter from the electrical engineer on the project to back up my interpretation of the code. The inspector wants this letter on file to, in his words “absolve us of liability”.





When I spoke to the engineer about this he pointed out that we have GFCI protection on the single phase pump motor for the dishwasher and the 3 phase circuit in question is actually feeding the boost water heater for the unit which is and accessory for the dishwasher and most importantly in his take on the situation is a water heater and therefore, is not required to be GFCI protected per NEC 2020.





The engineer also suggested if the inspector can’t get passed this it would be possible to install a delta to wye dry type transformer to change the circuit voltage and then set a disconnect with a GFCI breaker down stream of the TX. The main problem with this idea is the restaurant owner and GC don’t want to pay for this scenario as they also back up my interpretation of the code.





I understand the importance of GFCI protection, and if it was just as simple as purchasing, and installing the breaker, I would do so, and have no issue with it, even if I had to eat the cost of the circuit breaker. I am certainly not trying to stand my ground just to prove the Inspector wrong. I believe in most situations, we would all rather just do what the inspector asks for and move onto the next job.





So, stay tuned for future development :)
 
DJK said:
When I spoke to the engineer about this he pointed out that we have GFCI protection on the single phase pump motor for the dishwasher and the 3 phase circuit in question is actually feeding the boost water heater for the unit which is and accessory for the dishwasher and most importantly in his take on the situation is a water heater and therefore, is not required to be GFCI protected per NEC 2020.
Click to expand...
One more reason to not require GFCI, it is not a dishwasher. Someone may very well mentioned this sooner but you did not tell us this is for the booster heater until now.
 
tortuga said:
Yeah I have never measured how many mA a test button is, It similar with RCD's I have herd they often trip at 20-23mA.

I dont think there is a way to know if they had, In my (dated) experience UL is totally closed off to the public, it acts more like a private for profit corporation, unless your on a panel or have a complaint their internal processes are opaque.
NFPA for example has a research library in Quincy MA where anyone from the public can do pretty extensive historical research. Years back I spent some time there researching a code changes from 1969, and it was quite fascinating they have very detailed records available down to meeting minutes.
Click to expand...
I have been a member of the UL 943 panel for a number of years.
 
hornetd said:
I'll give you that it might well be unwarranted to do it in a residential situation so I probably should not have raised it in this thread.

I know it is enforceable by using testing instruments that have a memory function and can download that stored information to a computer. In our case we were doing certification of contract compliance so we presented a list of which breakers in which of the transportable prebuilt communications shelters we had made. The people doing acceptance testing for the client only had to pick one breaker in each shelter and check it against our list. If the results were essentially identical they were assured that our testing was adequate to support acceptance. The acceptance testers usually chose a couple of as yet untested breakers to test as an additional quality check.

I certainly didn't mean to imply that testing like this should be done by anyone but a licensed electrician.
Click to expand...
In general, nothing in the code that requires action after the final inspection is enforceable.
 
You must log in or register to reply here.
Top