GFCI Protection For VFD

Devin_Palmer

Member
Location
Raleigh NC
Occupation
Electrician
I am working on an estimate for the installation of a VFD for a fluid transfer pump at a brewery. The ground is constantly being hosed down, and it is a very wet location. The only GFCI breakers I can find trip at 6mA of leak current. I have called the manufacturer, and they have told me that the VFD will trip the GFCI unless it is a 20 mA trip setting. 208V three-phase 1.5HP Pump. What should I do?
 
Not sure what you are trying to do.
Why does this pump need ground fault protection?
There are special purpose GFCIs that trip at 20 mA, but are many $$$ and are stand alone devices, and a fault on the output side of the VFD will not trip the SPGFCI or GFCI.
Also GFCIs and SPGFCIs don't work on the load side of the VFD because of the varying voltage.
 
Gotcha, it is in a setting where the workers will be in standing water while touching the machine. That was the main reason I was concerned about ground fault protection. Is it really not needed?
 
Gotcha, it is in a setting where the workers will be in standing water while touching the machine. That was the main reason I was concerned about ground fault protection. Is it really not needed?
GFCI nuisance was replaced with systems described in 210.8(B)(4) Exception #2, even if your equipment was supplied by receptacles outdoors.

Worker's compensation claims for the trip hazard of standing water, is more likely to find Mgmt. liable for lack of drainage, or elevated platforms.
 
I am working on an estimate for the installation of a VFD for a fluid transfer pump at a brewery. The ground is constantly being hosed down, and it is a very wet location. The only GFCI breakers I can find trip at 6mA of leak current. I have called the manufacturer, and they have told me that the VFD will trip the GFCI unless it is a 20 mA trip setting. 208V three-phase 1.5HP Pump. What should I do?
I'd use a three phase RCD or RCBO;
 
GFCI nuisance was replaced with systems described in 210.8(B)(4) Exception #2, even if your equipment was supplied by receptacles outdoors.

Worker's compensation claims for the trip hazard of standing water, is more likely to find Mgmt. liable for lack of drainage, or elevated platforms.
Thanks man! I missed that when I glanced through the code.
 
Gotcha, it is in a setting where the workers will be in standing water while touching the machine. That was the main reason I was concerned about ground fault protection. Is it really not needed?
Is there receptacles involved that might trigger portions of 210.8 to require GFCI protection?

Outside of that and totally my thoughts, VFD's are pretty sensitive at detecting ground faults on the output side of the drive and possibly better protection than the typical GFCI devices anyway. The drive is protecting itself from fault current, but inherently is going to limit user exposure to any fault current as well.

A GFCI on input side is not going to detect a ground fault on output side either.
 
A GFCI on input side is not going to detect a ground fault on output side either.
I feel like I have seen 'ground fault alarm' on a VFD before, you guys know way more about VFD's than me but dont they have some kinda built in magic that detects ground faults on the motor side?
 
I feel like I have seen 'ground fault alarm' on a VFD before, you guys know way more about VFD's than me but dont they have some kinda built in magic that detects ground faults on the motor side?
Many do have built in ground fault detection on the output side, however the brand that was commonly used at the plant I did a lot of work at was set at 50% of the rated output of the drive. Not adjustable...you either enabled or disabled the ground fault protection when you set the drive parameters.
 
I feel like I have seen 'ground fault alarm' on a VFD before, you guys know way more about VFD's than me but dont they have some kinda built in magic that detects ground faults on the motor side?
Yes they do, and in most cases if there is a ground fault on the load side circuit, the drive will shut down with "ground fault" for a fault code. If it is a motor you are trying to start, it will trip before the motor even has had any noticeable acceleration as it will pick up this fault at a pretty low voltage level during the starting ramp up.

ETA: introduce sudden low impedance ground fault while running and it possibly can be damaging to the drive.
 
Many do have built in ground fault detection on the output side, however the brand that was commonly used at the plant I did a lot of work at was set at 50% of the rated output of the drive. Not adjustable...you either enabled or disabled the ground fault protection when you set the drive parameters.
On A-B drives, the GF trip is factory set at I think 38% of the VFD current rating... You would be LONG dead by the time that trips.

I don't know of any VFDs that will not trip a Class A GFCI (5mA), which is what you need for personnel protection. But for "other than dwelling" occupancies, you only need Class A for 15A and 20A 120V outlets. So just don't use the 120V powered VFDs plugged into an outlet if there is water involved.

If they WANT some added protection, you can get "Equipment GF protection" that trips at 30mA on many circuit breakers. Elsewhere in the world, 30mA is considered safe for protecting personnel, it's just that it is above the "let go" threshold (but below the cardiac arrest threshold). So they will not be able to let go until it trips, but it will not stop their heart. (Probably). That would be better than nothing. Most VFDs will be OK with a 30mA trip, so long as the grounding is all done properly and carefully.

Just don't call it "GFCI" protection, that is a defined term referring to Class A (5mA) protection. Just call is "Ground Fault" protection.
 
Elsewhere in the world, 30mA is considered safe for protecting personnel, it's just that it is above the "let go" threshold (but below the cardiac arrest threshold). So they will not be able to let go until it trips, but it will not stop their heart. (Probably).
I have looked into why this is, such as RCD protected public swimming pools in Europe.
RCD's are more time sensitive than UL 943 devices down to 30ma, I modeled the standards IEC 61008 to UL 943 at various currents to test UL 943 Class A compliance above 30ma (not not a GFPE breaker but a actual IEC 61008 RCD) UL 943 defines class A as allowed to trip in T = (20/I)^1.43 seconds

At 45ma (150% of 30ma for a ground fault) a UL 943 device must trip by .313 seconds or 314 milliseconds (ms)
30ma RCD breaker (RCBO) trips in 300ms = pass.

Drop it down to the 30ma threshold of the RCD UL 943 goes up to 560ms;
RCD trips in 300ms = pass

At 60ma (2x) UL 943 is 207ms. RCD tripping time is 150 ms = Pass.

And so on.
The only thing that a RCD does not pass UL 943 is (of course) is below 30ma UL 943 will allow a trip time of 1 second (1000ms) at 20ma.
A 5ma ground UL 943 trip time is 7260ms or about 435 cycles of 60Hz.
as we know most industrial installs have good equipment grounding a so 5-29 ma ground fault potential can be mitigated by a EGC.

In Europe and Asia they have about every possible type of 230V (220 - 240V) you can imagine (grounded, ungrounded, delta, wye ... etc) so a RCD is also good on any 240 60 hz system we have like hi-delta where a UL 943 is not for over 150V to ground.
 
My first job was at a large slaughter house then a large candy plant where both had not only nightly wash downs but standing water in many areas. We always pulled a copper ground wire to motors and nobody ever received a shock while touching a running motor. In the basement of slaughter house they had thousands of salted wet cow hides and salt built up was everywhere. Would get shocked replacing a burnt out 200 watt incandescent light bulb due to moist salt on glass but never received a shock from motors there. Was told that most pumps are triple grounded due to ground wire, being bolted to a grounded steel base and grounded piping. I can see no reason for GFCI protection but would take it up with the AHJ..
 
But for "other than dwelling" occupancies, you only need Class A for 15A and 20A 120V outlets.
That ended with 2014 NEC Now it is 50 amp receptacles 150 volts to ground and less and includes three phase receptacles up to 100 amp and 150 volts or less to ground. I can see this causing problems with portable machines that are cord and plug connected that have a VFD within them.

I do see that 210.8(B) in 2023 does not include the words "class A". Not sure if that was what they intended or not, particularly for 15/20 amp 125 volt receptacles or even nearly any single phase receptacle that fits in this section. I could see non class A protection being something they intended for three phase applications.
 
That ended with 2014 NEC Now it is 50 amp receptacles 150 volts to ground and less and includes three phase receptacles up to 100 amp and 150 volts or less to ground. I can see this causing problems with portable machines that are cord and plug connected that have a VFD within them.

I do see that 210.8(B) in 2023 does not include the words "class A". Not sure if that was what they intended or not, particularly for 15/20 amp 125 volt receptacles or even nearly any single phase receptacle that fits in this section. I could see non class A protection being something they intended for three phase applications.
It doesn't but that is covered by the definition any time the term GFCI is used
A device intended for the protection of personnel that functions to de-energize a circuit or portion thereof within an established period of time when a ground-fault current exceeds the values established for a Class A device.
 
On A-B drives, the GF trip is factory set at I think 38% of the VFD current rating... You would be LONG dead by the time that trips.

I don't know of any VFDs that will not trip a Class A GFCI (5mA), which is what you need for personnel protection. But for "other than dwelling" occupancies, you only need Class A for 15A and 20A 120V outlets. So just don't use the 120V powered VFDs plugged into an outlet if there is water involved.

If they WANT some added protection, you can get "Equipment GF protection" that trips at 30mA on many circuit breakers. Elsewhere in the world, 30mA is considered safe for protecting personnel, it's just that it is above the "let go" threshold (but below the cardiac arrest threshold). So they will not be able to let go until it trips, but it will not stop their heart. (Probably). That would be better than nothing. Most VFDs will be OK with a 30mA trip, so long as the grounding is all done properly and carefully.

Just don't call it "GFCI" protection, that is a defined term referring to Class A (5mA) protection. Just call is "Ground Fault" protection.
We had hundreds of great Danfoss VFD'S. At least a couple times a month we would get a call one tripped out on what they called " Earth Fault ". Over 95% of the time it was a false event. Would megger the 480 volt motors with a 1,000 volt megger, inspect the splices inside pecker head and readings were good. Might be one to three year span before next false ground fault trip. Did have maybe eight out of ten outdoor Danfoss exhaust fan drives trip.out on earth fault frequently but all had great measurements. Finally asked our great Danfoss tech why we were receiving so many false earth fault trips. Told me to for a short manufacturing period they had faulty ribbon cable that came off the control boards. He replaced them for free and problem disappeared. Both Danfoss & ABB drive techs told me the ground fault pick up on their drives is different then the plain Jane GFCI residential receptacles. One told me they use an algorithm to calculate when to trip out on ground fault.
 
Yes, the Class A, is not needed in any code rule because that is covered by the definition of Ground Fault Circuit Interrupter.
'Class A' definition relies solely on a outside (extra legal) definition in UL 943 and it really only has one problem it allows 420 cycles of the full 60Hz when your getting zapped in the 6- 29ma (North American range). Is 7 seconds really protecting anybody from anything in that range?
If insulation is breaking down or wet and a equipment ground present I am honestly not sure what kinda fault would stay at 6-29ma and not quickly ramp up?
 
'Class A' definition relies solely on a outside (extra legal) definition in UL 943 and it really only has one problem it allows 420 cycles of the full 60Hz when your getting zapped in the 6- 29ma (North American range). Is 7 seconds really protecting anybody from anything in that range?
If insulation is breaking down or wet and a equipment ground present I am honestly not sure what kinda fault would stay at 6-29ma and not quickly ramp up?
While the product standard does allow that long time to trip for lower level ground faults, in reality they trip pretty much instantaneously. If they actually tripped at the maximum permitted trip time you would have to hold the test button in for over 5 seconds.
 
Top