250HP Induction Motor - Greater Phase to Phase current Imbalance 4

[friend81]

Hi Folks,

Hope someone can help me on the below issue,

We have installed a 250HP, 380V, 60Hz Submersible pumset at 340m deep in a Borewell . The motor is a 6 lead motor & all the 6 leads were taken out of the well & connected in DELTA as a 3 wire motor before conencting to the VFD.

After switch-on the pumpset the Amps measured in each phase of VFD output are 376A / 436A / 540A in running voltage of 374V/376V/373V. The nameplate current of the motor is 430A. When measured through the Power Input cable to VFD the current is stable as 376A/374A/375A.

The current imbalance is measured only in Cables between the VFD output & Motor. After inverting the motor leads in VFD output same phase of the motor continues with the higher current. The insulation resistance of motor is 40M Ohms & Winding resistance 0,66 Ohms are stable in each phase. So problem in the motor is ruled out. Further if the problem is in the Motor the current will be same in Input as well as Output cables.

Could anyone say what could cause this type of Current Imbalance in VFD Output. No filters are installed yet along with VFD.

Could it be caused by Impedence in longer drop cable tied to the Steel column pipe ? Or anything wrong with the VFD ?

 

[jraef]

My first thought is that you are attempting to measure current and voltage on the output of a VFD with a simple hand-held clamp on meter, which is likely incapable of interpolating the complex harmonic rich current flow going to the motor. My rule of thumb (which you will have to do monetary conversion on for your country) is that if you spent less than $3000 for the meter, it is pointless to use it on the output side of a VFD. The information you get will not accurately reflect what the motor really sees. At that HP, I would hope that the VFD has the ability to provide you with individual output motor current amps. That would be more trustworthy if available. Also hopefully your VFD will provide for phase current imbalance protection going to the motor, which most large drive do, because it has critical motor heating ramifications that need to be protected against.

The input current to the VFD is meaningless with regard to current balance going to the motor, there is no direct connection because the input power is just the raw material from which the VFD makes DC, so the input AC data is otherwise irrelevant. It might only be relevant as an indicator of line voltage imbalance.

"Will work for (the memory of) salami"

 

[oldfieldguy]

What are you using to measure? Many VFD's have the function to measure and display the output current. One would think that the VFD would be set up to properly measure this. Comparing these values with another measuring device might be interesting.


old field guy

 

[Skogsgurra]

I agree with the others, very tricky to realiably measure current with modern clamp-on meters.

If you can find an old Crompton non-electronic moving iron meter, that could be used. Or a Messwandlerbau clamp together with a moving iron Multizet or similar.

I wouldn't trust one of Yokogawa's or Agilents modern meters either - even if they are way above Jeffs USD 3000 limit. I am not against electronics. Not at all. But, for field measurements, I prefer good old transformers and electromechanical principles.

BTW, are the cables separate or "cabled"? If separate, you can easily have an impedance unbalance. Especially when cables are 340 m long.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[LionelHutz]

It sounds like there is a motor or cable problem. I doubt it's the VFD.

 

[friend81]

The values 376A / 436A / 540A are measured by a hand held Clamp Meter. The VFD displays only average current as 450A. When the overload protection is enabled in VFD the protection switches off the pump displaying "Imbalance Current Protection".

The extension cable is a 185 Sq.mm single lead double insulated copper conductor tied along the steel column pipe. 6 Cable leads are tied together as Pairs U1+W2, V1+U2 & W1+V2 to make DELTA connection as 3 Wire in the VFD Output.

If the problem is with the motor either one of phases should show IR value less than others or Winding Resistance should also be differentiated. Both are stable in all 3 phases. And if the Imbalance is caused due to the defect in the motor the load of 540A will not be reflected in the Power Input cable too !!!

This is the fist time i am handling an VFD installation hence some questions may appear stupid, bear with me.

If the unbalance is caused by Impedance how can it be eliminated ? Is it advisable to run the motor with Impedance Imbalance or to be specific the actually current drawn by the motor is really lesser than what has been measured as 540A by the clamp meter ?

 

[Marke]

What is the rated full load current of the motor?
What frequency are you running at when these measurements are made?

I have seen similar when the cables were mis-identified and one winding was reversed. The measurements were at very low frequency and the drive refused to ramp up.

best regards,
Mark

Mark Empson
Advanced Motor Control Ltd

 

[jraef]

And if the Imbalance is caused due to the defect in the motor the load of 540A will not be reflected in the Power Input cable too !

!!

No, absolutely not. The incoming power is just converted to DC, so nothing that happens on the load phases passes through to the individual line phases. The line only sees total power, evenly distributed across the incoming phases by the converter bridge rectifier.

The extension cable is a 185 Sq.mm single lead double insulated copper conductor tied along the steel column pipe.

Hmmm... 6 separate unshielded conductors tied to a steel column, no conduit?



"Will work for (the memory of) salami"

 

[friend81]

The Rated Full Load Current of motor is 436A. The motor was running at 60Hz when these current was measured. When reduced to 54Hz the measurments were 220A/360A/420A.

The cables have been verified for continuty to avoid misconnection of the leads.

Normally in Submersible pump installtion no conduits are in practice. Generally it would be 3 core sheathed cables or double insulated individual leads.

 

[Marke]

Hi friend81

So you are getting that level of imbalance when running and the pump is doing some work.
Sounds like something seriously wrong. I would not run the pump like that.

I would suggest that you
a) separate the three windings and measure the resistance and inductance of each winding
b) recheck the polarity of each winding. If the start and finish of one winding are reversed, you can get some interesting currents.

Best regards
Mark.

Mark Empson
Advanced Motor Control Ltd

 

[waross]

Possibly an impedance issue with the 1000 Ft plus of unsymmetrical supply cables.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[LionelHutz]

I'd agree with what Bill just posted. I bet you didn't make any attempt at maintaining symmetry in those motor cables.

 

[jraef]

That was behind my comment on there being 6 separate conductors strapped to a steel pole. Any twist around the pole (i.e. "Maypole" fashion) over that length will create a reactor.

"Will work for (the memory of) salami"

 

[Marke]

Good point Jeff

I have seen 80 - 90 amps flowing in the earth cable due to current induced in the riser when one or more cables rotate around the riser on the way up.

Mark Empson
Advanced Motor Control Ltd

 

[Skogsgurra]

Right, but then some of the cables should be twisted around the tube and some not. That is why I asked if the cables were "cabled" or separate.

If separate, the imbalance can be worse than if the cores are in one or two cables.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[LionelHutz]

I wasn't at all meaning the issue was the wires wrapped around the stand pipe. Just tying them into a bundle the wrong way could be causing it, due to the wire configuration in the bundle. The best configuration probably would be to twist the wires up into 2 groups of 3-phase conductors and then tie those to the stand pipe.

 

[waross]

I saw once a suggested configuration for inverter feeds. Seven wires, six wrapped around one. The outer wires were to be configured A,B,C,A,B,C and the grounding conductor was in the center.
Comments?
At that length, the wires don't have to be wrapped around the pipe and wrapping all the wires around the pipe should affect all wires equally.
If one wire is wrapped around the pipe, even a plastic pipe, the separation from the other wires alone will cause problems. The proximity of the iron will help.
If the wires are tied flat on the pipe, the outer wires will not enjoy the same amount of field cancelling as the center wires and will have more impedance. The proximity of the iron will help.
Looking more and more like a cabling issue. Again, the proximity of the iron will help.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[friend81]

Hi, MARKE - The measurements were made during the intial startup of the system after completing installation, when obsereved this imbalance the same has been shutdown. We are planning to pullout the pumpset next week.

The resistance is good & uniform in all 3 phases. The polarity is also ok. So problem with motor is ruled out being brand new & factory tested.

I have consulated one of our known installer who install biggers pumps like 400HP & above, he suggested that the separate cable leads has to be crossed soon after the drop cable joint to avoid impedence. That is individual cables of U1+V1, V2+W2, W1+U2 & tied flat to column pipe.

 

[Marke]

Hello friend81

Lifting the pump is an expensive option, it is in NZ anyway, and I would be looking at what can be achieved before lifting it if possible.

There can be a problem with the motor, I have seen issues with brand new motors, especially operating on VFDs with long cables to the motor. Submersible motors typically are sensitive to voltage transients as the insulation rating is quite low. Many manufacturers require dv/dt filtering to minimise the rate of rise of voltage to less than 500V/uS. Very Long cables may require the use of a sinusoidal filter on the output of the vfd. It is possible for the voltage transients to cause a shorted turn in the motor and if it is only one turn, the motor will continue to operate with a large imbalance and a high temperature rise, so do not discount the motor itself.

You have measured the resistance and that is balanced. Have you measured the inductance? that can give an indication of a shorted turn problem. I would measure each winding separately and look for equal inductances.

There is a possibility that the lay of the cables as the pump has gone down has resulted in an imbalance in the turns around the riser, effectively adding a transformer/reactor in series with one or more conductors. - Hard to determine, but can result in a high current in the earth conductor going down to the motor.

I would suggest that you look at the polarity of the three windings and ensure that you do not have a start and finish transposed.
I have done this before without lifting the pump, the experts were all set to pull the pump, but we did some tests and saved a fortune.
Most of the problems that I have encountered have been with six wire double star motors where the cables have been identified at the top incorrectly.
From your description, this is a six wire delta motor, so I would try using a 24V transformer with about a 10 amp secondary rating.
Connect the 24VAC to one winding, say U1-U2. Then connect the other two windings in series v2 to w1 and measure the voltage between v1 and w2. next revers one winding, i.e. connect v2 to w2 and measure between v1 and w1. If you can measure an appreciable voltage difference, you can use this to determine the polarity of the windings. You need to do the test twice, connecting the 24V to 2 separate windings.

Try it in the workshop on a known motor first so that you can see what affect you have and then know what to look for.

Another way that I have used to confirm the polarity, is to first reconnect the motor in star and use a low voltage three phase applied to the windings. Measure the current in each phase and reverse either the high or the low phase to see if you can balance it up.
Be very careful, submersible pumps are very unforgiving and can be easily damaged by incorrect wiring and high voltage testing.
Good luck,
Mark.

Mark Empson
Advanced Motor Control Ltd

 

[ozmosis]

Friend81,
As Mark highlights, I'm surprised that no sine wave filter has been incorporated. Most borehole pump manufacturers would specify the dv/dt levels and this would often necessitate an output filter. In doing this, and with the length of motor cable involved, it would also be important to evaluate if a step up transformer would be required. A sine wave filter will 'add' to the volt drop issue at the motor end and needs to be considered for the reliability of the motor.

Regarding borehole pumps. A large Danish pump company are often telling me that the major cause of motor failure and (prior to failure) the resultant odd measurements of volts & current, are due to inferior termination and glanding at the motor end. If not carried out by a (recognised) installer or the manufacturer, this is by far the biggest cause of early life failure.
The use of a VFD will typically 'find' weaknesses in poor termination as it will in motor insulation.