If you're going to change wire, I'd go with #3 and a 100A breaker... and call it a primary only protection scheme for the transformer.
Breaker size to reverse feed a 30 kva transformer?
- Thread starter zinsco
- Start date
- Status
zinsco
Member
- Location
- San Diego, CA
Oh well, I totally spaced that I can get 3-#3's in the existing 1" conduit. I knew I could pull some #4's without having to replace the conduit and have that wire changed in like 5 min so I all ready got the 80 amp breaker and #4 wire to get this done tomorrow morning.
Shouldn't really make that much diff anyway according to the guy at Square D. He said that an 80, 90, 100 amp QO breaker will handle essentially the same 710 to 1500 amps for 0.15 seconds.
As far as over current on the 480v side, there is a disconnect with 20 amp fuses. After the inrush of the transformer, that 80 amp breaker should be more then enough to handle the load of the machine.
zinsco
Member
- Location
- San Diego, CA
Did the Job this morning and the first time I turned on the 80 amp breaker it held. All is well in the world. Thank all of you, you have all been great and this forum is so useful.
I do have a question about the ground, though. There is a ground wire coming from the 120/208v panel landed to the case of the transformer and then another ground wire from the same place running with the 480v wires and landed to the disconnect.
Why is there no reading from hot to ground at the disconnect on the 480v side? It must have something to do with the XO not being used and not bonded, but I have never reverse fed a transformer before so this is strange to me.
I do have a question about the ground, though. There is a ground wire coming from the 120/208v panel landed to the case of the transformer and then another ground wire from the same place running with the 480v wires and landed to the disconnect.
Why is there no reading from hot to ground at the disconnect on the 480v side? It must have something to do with the XO not being used and not bonded, but I have never reverse fed a transformer before so this is strange to me.
Great. You goal for the pri. OCPD is to use the highest rated breaker as allowed by the NEC 450 when a sec OVPD has been provided. You want that mag trip pickup to be as high as you can get the thermal element is of no consequence.
Glad to hear it held...
...but the first time comment made me a bit leery... as in it didn't hold a different time...:blink:
Sounds like you have an ungrounded secondary. It is permitted, but a ground detector must be installed.
How many fuses at the disconnect?
Voltage to ground of all three lines?
zinsco
Member
- Location
- San Diego, CA
Hahahaha, I should have just said, "It held." It's all good, to tripping.
3 fuses, and that's what I'm saying, 0 to ground for all three lines on the 480v side at the disconnect. But there is a ground wire from the 208v panel to the transformer and from the transformer to the 480v side disconnect at the machine.
Hmm... 0 volts to ground for all three lines seems strange... but a good strange.
What are you using to measure?
Regardless, there has to be a ground detector installed on this system.
See 250.20(B) and 250.21.
zinsco
Member
- Location
- San Diego, CA
I just used my knopp tester (wiggy to use slang) No digital or multi tester.
Two questions:
First: how is this not a grounded system if a ground wire is ran from the panel to the transformer, to the disconnect and to the machine?
Second: How does 250.20(B) apply since it mentions 3-phase, 4-wire, but I have 3-wire on the primary and the secondary, no neutral for both? Is it considered a 3-phase, 4-wire system since the panel it comes from is 3-phase, 4-wire, wye?
- Location
- Placerville, CA, USA
- Occupation
- Retired PV System Designer
Grounding the transformer shell does not make the secondary side a grounded system. It takes a ground connection to one of the windings to make it grounded.
If you do not ground any of the X terminals, including X0, nor any of the wires connected to them, you will measure zero volts from any of them to ground. There is no return path to allow current to flow through the meter. (Yes, there is current through a voltmeter! But it is a smaller current for a high impedance meter.)
If the voltmeter current is small enough, then the capacitance to ground of the other words, including the transformer windings, will carry enough current to allow a meter reading close to the nominal voltage.
If there is an X0 terminal on the transformer secondary, it is a four wire wye system. But if there are only line to line loads it may not be required to run that fourth wire anywhere last the transformer.
zinsco
Member
- Location
- San Diego, CA
I found this article on Reverse Feeding Dry-Type Transformers:
https://www.geindustrial.com/sites/geis/files/gallery/XformerWhitepaper2.pdf
Grounding
When the secondary (wye) of a delta-wye transformer is energized instead of the primary (delta), then the wye side of the transformer is not a separately derived service. As such, the neutral should not be connected to building ground nor should it be bonded to the transformer enclosure. The delta side of the transformer becomes the output, which is the separately derived system. The output delta “B” phase should be tied to ground unless the facility distribution system utilizes a different grounding scheme.
Should this be done to ground the system?
https://www.geindustrial.com/sites/geis/files/gallery/XformerWhitepaper2.pdf
Grounding
When the secondary (wye) of a delta-wye transformer is energized instead of the primary (delta), then the wye side of the transformer is not a separately derived service. As such, the neutral should not be connected to building ground nor should it be bonded to the transformer enclosure. The delta side of the transformer becomes the output, which is the separately derived system. The output delta “B” phase should be tied to ground unless the facility distribution system utilizes a different grounding scheme.
Should this be done to ground the system?
kwired
Electron manager
- Location
- NE Nebraska
- Occupation
- EC
GD already answered, but I may try to explain in a little different manner.
The 480 volt side of the transformer in question is a separately derived system.
That means it is completely isolated from the 208/120 side. The magnetic field is what couples them, current flowing in primary side creates a magnetic field, that magnetic field creates a voltage in the (isolated from primary) secondary coils.
The only way this new system is grounded is if you connect one of the points of the new system to ground.
So your grounding conductor from the primary is likely grounded - but has no reference to the secondary system at all.
zinsco
Member
- Location
- San Diego, CA
Excellent explanation, thank you! I understood what GD said, but you paint the picture and help me to see it.
It would seem then, that I need to ground one of the legs on the 480v side. I already asked the question but let me ask it again so it doesn't get overlooked:
I found this article on Reverse Feeding Dry-Type Transformers:
https://www.geindustrial.com/sites/g...hitepaper2.pdf
"Grounding:
When the secondary (wye) of a delta-wye transformer is energized instead of the primary (delta), then the wye side of the transformer is not a separately derived service. As such, the neutral should not be connected to building ground nor should it be bonded to the transformer enclosure. The delta side of the transformer becomes the output, which is the separately derived system. The output delta “B” phase should be tied to ground unless the facility distribution system utilizes a different grounding scheme."
Should this be done to ground the system?
You are not require to ground the secondary system (480 output). That's why I pointed you to 250.20/21. Of course you will need a ground detector installed to keep such a system compliant. Many industrial facilities set up ungrounded systems for the advantage their connected equipment OCPD's will not trip/blow on first or subsequent same-phase ground faults. As a result, equipment continues to operate. It is highly recommended, if not absolutely required to track down and fix the faults post haste... and the no-trip/blow condition allows for an orderly shutdown if that is desired. If the faults are not remedied, and provided faults are not resistive enough to prevent it, a second-phase-to-ground fault will trip/blow the OCPd's.
You are permitted to ground the system, and you can ground any single "line" (there is no neutral connection on the secondary)... but generally accepted trade convention is to make a "Grounded B-Phase" system. You would have to rewire your system because 1) the grounded conductor would have to be white, gray, or one of the permitted ID methods, and 2) the grounded conductor cannot be single trip/blow (any OCPD in the grounded conductor must be common trip with the ungrounded conductor ocpd's).
zinsco
Member
- Location
- San Diego, CA
In other words, the disconnect with fuses would not work. Gotcha.
Not sure what type of ground detector to get and how I would connect it. I guess I could just ask the supplier and it would probably connect to the disconnect.
Well I don't have any particular product recommendation. Perhaps some other forum patrons do and would be willing to share their thoughts.
Plenty of info' on the web...
kwired
Electron manager
- Location
- NE Nebraska
- Occupation
- EC
OP has a delta secondary, I could be wrong but I believe use of HRG systems are limited to systems with a neutral per NEC. If not that seems to be what is most popular anyway, HRG for a wye system or ungrounded with ground detection for a delta system.
With a zigzag transformer and a resistor, okay?
- Location
- Massachusetts
With a zigzag transformer and a resistor, okay?
The transformer is exsiting, the OP is not starting from scratch.
Retrofit should be possible.
- Status