Possible 3-Phase GFCI Solution

Status
Not open for further replies.
Hi All,

Brand new to this forum and hoping I can get some feedback on an idea I'm working on for a GFCI implementation for a 3-phase heating setup to use on a pilot brewing system in a commercial environment. I'm an EE that spends the majority of my time working with low voltage DC applications, so not a lot of time working with AC, particularly 3-phase. I know (or believe I do) that GFCI isn't required for a 3-phase setup, but I'd still like to implement if possible for a reasonable cost. That said I mean something significantly less expensive than a Bender LifeGuard 3-phase GFCI.

My first question is regarding the use of 2-pole GFCI breakers in a 120V/208Y 3-phase load center. Am I correct in believing that a 2-pole 240V style GFCI breaker will function as expected in such a panel, providing 208V single phase between either L1/L2, L2/L3 or L3/L1? If not then I know my idea won't work, but from what I can gather online it appears it should work.

So if the answer to my first question is "yes" then please refer to the image I've included below for what I'm trying/proposing to do. I would like to use 3 individual single-phase heating elements rated at 6000W at 240V, so about 4500W at 208V. These elements would be used to fire a brew kettle for making 1 barrel test and specialty batches in our otherwise 10 barrel sized brewery. If desired the system could be used in a single phase residential setting with 2 of the 3 elements providing 12kw power plugged into two separate circuit 240V 30A GFCI protected outlets. But for the primary application all 3 elements would be plugged into an enclosure which has been wired to create a 3-phase delta configured load on a 3-phase wye power source. The 3 elements would be powered on and off with 3 separate SSRs that are switched by the same controller.

That said my assumption is that if the answer to my first question is "yes" regarding using a 2-pole GFCI on 208V single phase from the load center, then the schematic I've shown should also work since I'm basically powering 3 separate loads through the 3 GFCI breakers.

Any feedback is greatly appreciated! If I'm totally out to lunch with this idea that's fine, but if nothing else it's been fun to think about.

Thanks, Kevin
 

Attachments

  • 3-Phase-GFCI.jpg
    3-Phase-GFCI.jpg
    17.5 KB · Views: 4

winnie

Senior Member
Location
Springfield, MA, USA
Occupation
Electric motor research
You will need to check the specifications of the individual GFCI devices. They derive power from the connected supply, and therefore must be designed to work with the intended supply. Check the datasheet for the GFCIs that you are considering.

As long as each GFCI sees a totally independent load, then the system will function as you expect. You could not connect a true 'delta' load to your set of GFCIs. But as you have drawn things you have 3 separate single phase loads connected 'delta' upstream of the GFCIs.

-Jon
 
You will need to check the specifications of the individual GFCI devices. They derive power from the connected supply, and therefore must be designed to work with the intended supply. Check the datasheet for the GFCIs that you are considering.

As long as each GFCI sees a totally independent load, then the system will function as you expect. You could not connect a true 'delta' load to your set of GFCIs. But as you have drawn things you have 3 separate single phase loads connected 'delta' upstream of the GFCIs.

-Jon

Hi Jon,

So here's a follow up question regarding the overall 3-phase circuit loading. While I'm not connecting a true delta load as you've pointed out, in my non-3-phase experienced mind the circuit analysis for "phase/element" current and line currents on the 3-phase power lines in would still be calculated based on what I showed as the equivalent circuit. In other words each element at 9.6-ohm creates a 21.67A "phase" current, the sum of the 3 phases is 65A, and the line current through the main 3-phase 50A breaker would be 37.5A for each of the lines L1, L2 and L3.

Does that sound correct?

Thanks,

Kevin
 

winnie

Senior Member
Location
Springfield, MA, USA
Occupation
Electric motor research
As a general rule, don't try to add up individual phase currents to get some sort of 'total' current. It just causes confusion. Yes, the concept of 'total aggregate current' has meaning; it is just too easy to confuse the meaning with the normal 3 phase current value that most people use.

You are far better just figuring the normal 3 phase current, and then if you want to describe the total aggregate load, use total VA or total W.

Your calculations look correct, my check below:

You have 21.67A going through each of 3 single phase loads.

Each single phase load consumes 4500VA.

Your entire system consumes 13500VA.

The line current is 21.67 * sqrt(3) = 37.5A

Since everything is symmetric that is 37.5A on each of the 3 phase lines.

And as a double check, using the normal 3 phase calculation: 37.5 * 208 * sqrt(3) = 13500

-Jon
 

rlundsrud

Senior Member
Location
chicago, il, USA
I am certain that you realize this, but I thought I would point out that if a fault occurs on any one of the phases (A, B, or C) you will have 2 of the GFI's drop out. I assume that to heat whatever batch you are running, you need all 3 sets of elements to be operational any, so this wouldn't be an issue. If however, one of the elements are for redundancy, this would be a problem as it would eliminate said redundancy. With that said, this is a well thought out system that should work as designed.
 

winnie

Senior Member
Location
Springfield, MA, USA
Occupation
Electric motor research
Actually, as the system is drawn out, each of the three heaters is an independent load on an independent GFCI. So a failure on a single heater should not trip the other GFCIs.

-Jon
 

romex jockey

Senior Member
Location
Vermont
Occupation
electrician
Why not use one of these.....?

41Nz0i35jcL.jpg


~RJ~
 

kwired

Electron manager
Location
NE Nebraska
I went to the Leviton website, but I cannot find a blank face 250 volt GFCI. Where did you find that?

Lowes and Home Desperate do have the 125 V GFCI blank face on the shelf around here. Popular for bathroom spa type tubs.
Look at the one pictured carefully - is not a typical device box GFCI instead it has a CT loop on the back side, you should be able to place any voltage conductors through there, as long as all incoming and return current passes through the CT, the output will be satisfied.
 

TwoLeftHands

Member
Location
San Jose CA
Not sure if you tried this but you might want to google "RCD" instead of GFCI as that is the name commonly used in Europe where ~240V systems are used.

In addition to GFCI, I'd make sure the heating elements/wiring are "double insulated" (see UL definition) from any liquid and that all cable harnesses have a "drip loop" on them and are potted. The thing with liquid and condensation in humid environments is that the copper strands in cable act like a very efficient wick. Make sure you meet creepage & clearance and use a listed thermal fuse (you can probably find one used for electric water heaters) thermally bonded to product.

I'd also recommend some way of driving any metal of your aparatus slightly DC negative to ground by using a zinc slug so it doesn't corrode.
 

kwired

Electron manager
Location
NE Nebraska
Size of the CT opening could be a limiting factor for OP - he likely needs about 8 AWG conductors the way it sounds on his main supply anyway.
 
Status
Not open for further replies.
Top