MatthewDB
Electrical
- Sep 20, 2011
- 142
Hello,
We produce a power electronic converter that operates at 480V 3ph input. The device is 3 wire only, no neutral connection. For operation on 600V and 690V systems we utilize a 3 PH auto-transformers. The auto-transformer is integrated into the same enclosure, so it falls under our UL listing (not under the customer's wiring local AHJ).
The transformer was selected by my predecessor. It is Y connected, a 3 phase 3 leg unit. The neutral is not connected as the load is solely 3 wire. In fact it is buried and not normally accessible. Had I selected this unit, I would have included a delta tertiary winding to provide a zero sequence / triplen path. I was concerned that we could block ground faults and / or have winding over-voltages on ground faults.
Being curious, I did some testing on the system. First I accessed the buried neutral and measured voltage waveforms. There was not a significant 3rd harmonic present on the neutral as I expected. I also tried 5A and 10A of zero sequence current - again, no significant impedance noted. The voltage only dropped a fraction of a percent.
That's when I realized that DUH, a 3 leg core doesn't act like three single phase units. Un-canceled MMF of one phase couples to other phases and can induce currents to flow in the other phases. My concern though is that this means that there is MMF top-to-bottom on the core. This MMF no doubt could cause heating in adjacent steel of the enclosure.
It makes me wonder if I really should push for a delta winding. I would need to pitch "hey we should make the transformer more expensive" without something very solid to back it up. I'm also not sure how I should size this winding. All of the literature I'm finding is talking about huge utility units for going between transmission system voltages. Here the wye point is solidly or impedance grounded.
I'm thinking for sizing there are two requirements - indefinite up to 10A zero sequence when we're applied on a HRG, and "don't burn off" when coordinated with our input breaker on a maximum fault current source.
We produce a power electronic converter that operates at 480V 3ph input. The device is 3 wire only, no neutral connection. For operation on 600V and 690V systems we utilize a 3 PH auto-transformers. The auto-transformer is integrated into the same enclosure, so it falls under our UL listing (not under the customer's wiring local AHJ).
The transformer was selected by my predecessor. It is Y connected, a 3 phase 3 leg unit. The neutral is not connected as the load is solely 3 wire. In fact it is buried and not normally accessible. Had I selected this unit, I would have included a delta tertiary winding to provide a zero sequence / triplen path. I was concerned that we could block ground faults and / or have winding over-voltages on ground faults.
Being curious, I did some testing on the system. First I accessed the buried neutral and measured voltage waveforms. There was not a significant 3rd harmonic present on the neutral as I expected. I also tried 5A and 10A of zero sequence current - again, no significant impedance noted. The voltage only dropped a fraction of a percent.
That's when I realized that DUH, a 3 leg core doesn't act like three single phase units. Un-canceled MMF of one phase couples to other phases and can induce currents to flow in the other phases. My concern though is that this means that there is MMF top-to-bottom on the core. This MMF no doubt could cause heating in adjacent steel of the enclosure.
It makes me wonder if I really should push for a delta winding. I would need to pitch "hey we should make the transformer more expensive" without something very solid to back it up. I'm also not sure how I should size this winding. All of the literature I'm finding is talking about huge utility units for going between transmission system voltages. Here the wye point is solidly or impedance grounded.
I'm thinking for sizing there are two requirements - indefinite up to 10A zero sequence when we're applied on a HRG, and "don't burn off" when coordinated with our input breaker on a maximum fault current source.