ABB VFD delta high leg

hillbilly1 said:
Maybe I missed something, but didn’t the op say the phase loss moved with the changing of the phase? Sounds like it’s in the drive, possibly surge protection on the front end blown and open.
Click to expand...

From the OP's post #1:
"I moved the wire on A and the no current "moved" to the new location, C. "

From the OP's post #20:
"I swapped A and C on the VFD input, and the no current "followed" to the new location. "

From the OP's post #24:
"I was back this morning and rotated the phases on the input to the VFD. The result was the same input phase loss fault. I rotated again and still the same result. I measured current and the no current "moves" as well."

In posts 20 and 24, I read that as having the zero current measurement move to a different VFD input terminal when the phases are swapped or rotated. In post 1, I think he is referring to A and C as terminals on a VFD since he calls them a "location" instead of an input phase.

It would be good if it were clarified whether the A phase is the one consistently having a zero current.
 
synchro said:
From the OP's post #1:
"I moved the wire on A and the no current "moved" to the new location, C. "

From the OP's post #20:
"I swapped A and C on the VFD input, and the no current "followed" to the new location. "

From the OP's post #24:
"I was back this morning and rotated the phases on the input to the VFD. The result was the same input phase loss fault. I rotated again and still the same result. I measured current and the no current "moves" as well."

In posts 20 and 24, I read that as having the zero current measurement move to a different VFD input terminal when the phases are swapped or rotated. In post 1, I think he is referring to A and C as terminals on a VFD since he calls them a "location" instead of an input phase.

It would be good if it were clarified whether the A phase is the one consistently having a zero current.
Click to expand...
If the swapping leads results in the "no current" following the swap then something in the supply conductors is the problem? Is possible to still measure voltage with no load, but put load on it and then check for voltage. Might need some temporary test load just to test the supply circuit if the drive won't run long enough to measure before fault code shuts it down. Resistance in that conductor will drop the voltage, but only when there is load on it. Voltage won't drop with no current flow.
 
tortuga said:
So with the power off the imbalance goes away, that leads me to suspect the customer is incorrectly loading the bank or you have some hi resistance failing connections. Some photos would help.
When I have fixed this exact problem it was by either moving the loads around or having the POCO close the delta;
Consider there are 4 main types of loads you can have on a typical utility 240 hi-leg delta made from 3 transformers on a pole
(In theory 5 types but I digress)
  • 120V loads (two wire),
  • 120/240 3-split phase loads (three wire),
  • single phase 240 loads (two wire)
  • three phase 240 loads (three wire).
Where B is the wild leg;
A-C is the split phase lighting pot,
A-B is a power pot; and B-C is a power pot
the first two kinds of loads
120V loads and 120/240 3-wire split phase loads
are only on the A-C winding so I calculate them differently

Where as the third type, straight 240 single phase loads, can be on any of the pots A-B, A-C, or B-C,
these loads used to be less common but are becoming very common, loads such as server racks or single phase 230V equipment.

The last type 240 three phase loads (the most common) evenly load all three pots.

Now if we eliminate the B-C, pot and make it a open hi-leg delta the kVA rating of the entire bank degrades.
Any non-three phase loads that use both transformers such as single phase 240 loads connected 'B-C' across the open set or the 5th type a 208V load running B-Neutral degrade the kVA rating even more and can cause voltage instability.
I have never seen the 208V L-N done intentionally but I do come across single phase 240 loads on the open set (B-C), moving those loads to the power transformer A-B improves the voltage stability.
Click to expand...
I have meter main service. The first 400A sub panel (B1) has a bus system with 10 3-phase circuits (30 spaces). One 100A 3-phase feeder to sub panel (B3). Starting at spaces 31/32 to 53/54, the bus is A-C only. There is a 1-phase 100A feeder to sub panel (B2) that straddles "both buses", connected C-A.

If I understand your post correctly, B-C 1-phase loads cause voltage instability. I don't have any.


B1: (A, B, C)
9, 3-phase circuits
4, 1-phase 240V circuits A-C only
11, 1-phase 120V circuits

B2: (A, C)
17, 1-phase 120V circuits
2, 1-phase 240V circuits

B3: (A, B, C)
2, 3-phase 240V circuits
9, 1-phase 120V circuits
 
synchro said:
From the OP's post #1:
"I moved the wire on A and the no current "moved" to the new location, C. "

From the OP's post #20:
"I swapped A and C on the VFD input, and the no current "followed" to the new location. "

From the OP's post #24:
"I was back this morning and rotated the phases on the input to the VFD. The result was the same input phase loss fault. I rotated again and still the same result. I measured current and the no current "moves" as well."

In posts 20 and 24, I read that as having the zero current measurement move to a different VFD input terminal when the phases are swapped or rotated. In post 1, I think he is referring to A and C as terminals on a VFD since he calls them a "location" instead of an input phase.

It would be good if it were clarified whether the A phase is the one consistently having a zero current.
Click to expand...
I am moving the conductors to the vfd input. This is on 3 different brands of VFD. "A" phase moved to either L1, L2, or L3 has zero current. I do not see this problem on non-vfd motor 3-phase circuits. Though I am seeing voltage imbalance fluctuate on VFD/non-VFD circuits.
 
It's puzzling that the current is only on phases B and C feeding the VFDs, given that the measured B-C voltage is consistently less than the A-B and A-C voltages. The output of diode bridge rectifiers like those in VFDs is a function of such L-L input voltages. The diodes conduct current whenever each of the three L-L instantaneous voltages exceeds the DC bus voltage across the VFD's capacitor by the forward drop of the rectifier diodes. When this diode current conduction occurs, it will be near the peaks of the L-L voltage, and not during the entire L-L voltage waveform which a meter is typically measuring. And so it's possible that the B-C voltage waveform is more "peaky" than the A-B, A-C waveforms (or the A-B, A-C voltage peaks are flattened), such that the D-C bus voltage is established only by the B-C voltage peaks. This would then prevent the diodes connected to line input A from conducting, even though the associated A-B and A-C voltages measured by an RMS or averaging meter are higher than that across B-C.

And so I think only 'scope waveforms could resolve this for sure.

One test that could provide some insight would be to measure the DC bus voltage. Then, after disconnecting the line input B, measure the bus voltage, and also the A and C line currents. I would expect significant current now on the A input, assuming what I mentioned above is happening (also assuming that an error is not detected which prevents the VFD from running). Also, I'd expect the DC bus voltage to be somewhat less than with B connected.

To be extra cautious, you could run with a reduced load or a lower speed to reduce the overall current drawn. However, I wouldn't expect the DC bus voltage to drop much when fed only with A-C, and so any increase in current drawn should be relatively small compared to what you already have with single phasing.
 
You should put a power monitor on the service and one for the feeder to the panel that feeds the VFD. You will want all 5 wires. neutral, 3 phases and the ground bus (if down stream of the MBJ). I would also talk to their neighbors to see if they are also experiencing the problem.
 
Here's what I did yesterday.

I used a Fluke T5-600 averaging meter and a Fluke 325 RMS meter.

RMS meter
A to N 120.6V
B to N 208V
C to N 120.7V

phase - phase
AB 241.6V
AC 241.2V
BC 238.3V

Avg meter
A to N 122V
B to N 205V
C to N 122V

phase - phase
AB 245V
AC 245V
BC 227V
 
wncjefftino said:
I swapped A and C on the VFD input, and the no current "followed" to the new location. I'm going to use the original VFD at a different location (still 240 delta system) to see if it's an ABB VFD issue.
Click to expand...
Wait
Motor parameters entered wrong cause spurious trip, double check Motor Rated Current (A), Motor Rated Voltage (V), Motor Rated Frequency (Hz), Motor Rated Speed (RPM), and Motor Rated Power (kW/hp) and current limit set correct
EMC filter disconnect
 
wncjefftino said:
I have meter main service. The first 400A sub panel (B1) has a bus system with 10 3-phase circuits (30 spaces). One 100A 3-phase feeder to sub panel (B3). Starting at spaces 31/32 to 53/54, the bus is A-C only. There is a 1-phase 100A feeder to sub panel (B2) that straddles "both buses", connected C-A.

If I understand your post correctly, B-C 1-phase loads cause voltage instability. I don't have any.


B1: (A, B, C)
9, 3-phase circuits
4, 1-phase 240V circuits A-C only
11, 1-phase 120V circuits

B2: (A, C)
17, 1-phase 120V circuits
2, 1-phase 240V circuits

B3: (A, B, C)
2, 3-phase 240V circuits
9, 1-phase 120V circuits
Click to expand...
What is the size of the two transformers in kVA on the pole?
It would be interesting to see a load calc for an open delta, if take those circuits and presume some low loads
then compute the minimum required size of each transformer in a open delta bank;
  • 180VA per 120V circuit ,
  • 600 VA per 1-phase 240V circuit,
  • 4800 VA per three phase circuit
Presuming all the 1-phase is really on lighting transformer (A-C).
If I am remembering right three phase loads as seen on each open delta transformer are VA x 2 not 1.732.
For the open delta bank I come up with a 114kVA for the lighting transformer and 106kVA for the power transformer.
So if that is even in the ballpark (its a wild guess), you should see two 100kVA or larger transformers on the pole, and thats if they are the only customer on that bank.
 
tortuga said:
What is the size of the two transformers in kVA on the pole?
It would be interesting to see a load calc for an open delta, if take those circuits and presume some low loads
then compute the minimum required size of each transformer in a open delta bank;
  • 180VA per 120V circuit ,
  • 600 VA per 1-phase 240V circuit,
  • 4800 VA per three phase circuit
Presuming all the 1-phase is really on lighting transformer (A-C).
If I am remembering right three phase loads as seen on each open delta transformer are VA x 2 not 1.732.
For the open delta bank I come up with a 114kVA for the lighting transformer and 106kVA for the power transformer.
So if that is even in the ballpark (its a wild guess), you should see two 100kVA or larger transformers on the pole, and thats if they are the only customer on that bank.
Click to expand...
The transformers are 75kVA and 25kVA. And they are shared.
 
wncjefftino said:
The transformers are 75kVA and 25kVA. And they are shared.
Click to expand...
Ok they are not just for your building then thats too bad,
just to give you an idea
Given 75kVA 'lighting' transformer for the split phase A-C or 'single phase services'
and a 25kVA transformer A-B
If I recall correctly your balanced three phase load capacity is limited to the smaller transformer 25kVA;
Transformer AB = 25 kVA + 25kVA of Transformer AC
= 50 total kva open delta.
Then factor 0.866 x (50kVA) = 43.3 kVA or about 104A @ 240V for three phase load capacity.

For the all split phase loads;
75kVA - (three phase load 25kVA) = 50 kVA or about 208 Amps @ 240V remains on A-C for the split phase loads
That is assuming good power factor etc.

Granted utility xfromers on a pole have a 1/2 hour rating of something like 2-3x the nameplate, so you may be able to have a peak demand of ~312A three phase 240 for a brief time.
If I am way off somone on here will correct me.
 
Are there any power factor correction capacitors on the pole or inside the facility? If so, perhaps they were originally intended for only across-the-line motor loads. Then if VFDs were added (either external or internal to the equipment), the inductive load would decrease and the capacitors may be larger than they need to be. Perhaps the system then would be closer to resonance at a harmonic frequency, and therefore enhancing the response to harmonic currents. And the VFDs would be producing such harmonic currents.

The presence of harmonics could explain the discrepancy between the RMS and average voltage measurements. Harmonics could also contribute to a higher peak voltage on B-C, therefore causing it to dominate the load current from the rectifiers in the VFDs, and resulting in no current being drawn from phase A.
 
synchro said:
Are there any power factor correction capacitors on the pole or inside the facility? If so, perhaps they were originally intended for only across-the-line motor loads. Then if VFDs were added (either external or internal to the equipment), the inductive load would decrease and the capacitors may be larger than they need to be. Perhaps the system then would be closer to resonance at a harmonic frequency, and therefore enhancing the response to harmonic currents. And the VFDs would be producing such harmonic currents.

The presence of harmonics could explain the discrepancy between the RMS and average voltage measurements. Harmonics could also contribute to a higher peak voltage on B-C, therefore causing it to dominate the load current from the rectifiers in the VFDs, and resulting in no current being drawn from phase A.
Click to expand...
Confirm it, disconnect pfc equipment if possible and see the issue resolved
Check increased current in pfc equipment for resonce before proceed to disconnect it
 
You must log in or register to reply here.
Top