3ph motor drawing more current as bus load increases

[Delta]

I saw this today and have no idea what is causing it.

A 3 phase motor drives a centrifugal separator on a fuel purifier skid. The current draw on each phase increases from around 3.5A to 5.3A as engine load increases.

This is on a drilling rig, equipped with 6 1500kW Caterpillar Diesel generators, running a 690v.  Two delta - delta/star transformers supply 4 rectifiers, one per secondary, onto a common DC bus.  12 inverters then power 12 motors of around 500 - 800kW rating.

There are transformers from 690 down to 480 for the two main busses.

The motot in question is 480v.

We have 4 engines online, at about 50% load.

Two of the VFD motors are experiencing cycling loads with a period of around 8 seconds.  This causes the engine load to swing between 48 - 52%.

As the engines throttle up, the current draw from the separator motor increases to 5.3A, dropping back to 3.5A as the load drops off.

There is no change of mechanical load on the purifier skid.  It just pumps over a tank.

The voltage is solid at 468v. The engine speed is maintained at 1200rpm plus or minus 1rpm, so no frequency swing.

What on earth could cause this?

Could large harmonics caused by the VFDs cause the induction motor to attempt to follow the harmonic frequency?

[jbb]

That's odd.

Just a quick check to confirm: you've got a common DC bus, and a VFD connected between this bus and your motor?

[Delta]

The motor in question is a small 480v DOL induction motor.

I mentioned the hefty VFD lineup as I wonder if it could be significant.

[duak]

I'm assuming the separator motor is driven from its own VFD.  If not, then where is it getting power?  How is the separator motor current measured?

Here's a few ideas:

1.) could EMI from the large inverter/motors be coupled to the separator VFD/motor through the DC bus as a common mode signal that then confuses the separator VFD's current sense or control/drive circuitry? 

2.) if any of the motors or wiring have leakage or a short to ground/earth/deck it would likely manifest as something unusual in the smallest or most sensitive element.

3.) could there be a feedback or coupling path that is not electrical? ie., engines throttle up, air demand increases, local air pressure drops and pump has to work harder to maintain flow.

Look at the DC bus voltage to ground/earth/deck with a 'scope to see if something unusual is happening.

Best o' luck

[jbb]

Interesting.

I have a guess: it could be that the extra VFD load is drawing extra harmonic currents from the genset. Because the genset has some impedance, you might get harmonic voltages on the 480V bus. Harmonic voltages can make induction motors quite unhappy.  To diagnose you would need some kind of power quality meter (Fluke make some nice ones for in-field use).

Alternative guess (less likely): if the phase currents drawn from the genset and transformers are unbalanced, this can cause voltage imbalance. Voltage imbalance can also make induction motors quite unhappy. To diagnose you can connect three (quality!) meters in delta configuration (phase a-b, phase b-c, phase c-a). If they give different V readings you could have a balance issue. Alternatively you could use a power quality meter...

Whatever you do, be bloody careful of the AC and DC busses. The sort of installation you describe can have huge potential for arc flash (depending on projection relays) so you should speak to the site electrician before poking around.

As to whether or not the situation is a problem for the DOL motor: it’s hard to say. What’s it’s rated current?

[Rerouter]

Vfd's are horribly noisy things, it's likely what you are seeing is reactive/apparent current instead of real current drawn by the motor.

At a guess your vfd's are very closely in step with each other's frequency. But not perfectly. Making a beat frequency of about 8 hz. The harmonic noise of the 2 then adds and subtracts as they drift between aligning and opposing. Causing reactive fluctuations in current while not changing the real current.

[drussell]

...
The current draw on each phase increases from around 3.5A to 5.3A as engine load increases.
...
As the engines throttle up, the current draw from the separator motor increases to 5.3A, dropping back to 3.5A as the load drops off.

How are you measuring this current draw? 

Is the voltage truly sinusoidal?  Probably not.  Are you at least somehow taking basic RMS measurements?

I suspect you are not actually reliably measuring the real current and therefore the actual true power going to the motor...  Stick a scope on it, would always be my answer when you can't seem to rationalize what is going on with the power input.  That will usually make it absolutely obvious. 

[ConKbot]

Since this is workplace stuff, I'd buy/rent a 3 phase power analyzer.  Like a fluke 435/437, 4 current transformer probes, 4 voltage inputs, so you can safely get a full picture of what's going on as far as real and reactive power. If a single motor or drive is kept from burning out, you're paid for the rental.

[Jeroen3]

Could large harmonics caused by the VFDs cause the induction motor to attempt to follow the harmonic frequency?

If you have a 50% current increase in a simple motor it could be more damaging to other equipment.
Have it looked at.

But first determine if the measurement you have right now is valid. If you concluded this using the analog panel meters, you should measure it again with a true RMS clamp meter.

[tpowell1830]

I have been on drilling rigs and sometimes the systems are thrown together just to get things working. I am not saying that your system is this way, it was just an observation in several drilling rigs that I have seen. With that said, there are so many inductive loads, I wonder if you have capacitor banks to correct power factor?

The harmonics from the VFDs should be isolated from your system through the VFD controls unless there is some type of ground loop effect going on, in which case, that is very dangerous.

Just some thoughts...

[Delta]

Thanks for all the replies.

To clarify, the motor in question is NOT on a VFD, it is direct-on-line from the main 480v board. The 480v board is supplied from a transformer off the 690v generator board (the 690v board also supplies the VFD rectifiers via two transformers)

I noticed the current variation on a panel meter, then confirmed it with a current clampmeter.

The rig is only a few years old, and this is all original equipment.

I will take some more measurements tomorrow, I have a Fluke scopemeter with 300v Cat III inputs, and 10:1 probes, but am not comfortable connecting this to a 480v bus...

The rig's power factor is quite shit, around .6, but there is no power factor correction.
But surely each load determines it's own PF, it's not possible for the system to "force" reactive currents through this motor.

[tpowell1830]

Thanks for all the replies.
...
The rig's power factor is quite shit, around .6, but there is no power factor correction.
But surely each load determines it's own PF, it's not possible for the system to "force" reactive currents through this motor.

Hmmm, on the grid, land based, the power companies have many substations that are specifically there to handle phase correction, power factor, etc. However, on a genset, I have to remind myself that it is a load as well. In this case, I would take an o'scope, make a coil (xformer) sensor and look at output line of each genset to see if there are strange anomalies in the waveform when the current rise is active. I would not be interested in the voltage level, only that the correct frequency is there and the waveform is nice and sinusoidal and the p-p is consistent. If all is well, I would then go to each motor and do the same.

If all is still well and good, 
Think some more...

[Jeroen3]

0.6 PF isn't very good when the majority of your load is a VFD. Are there any non-vfd motors that are slipping constantly?
Or is the kvar sharing not working as intended?

Distortion of the waveform of the generator will be very low with these size generators. Unless your are really pushing the limits on the amps range.
Small or lightweight generators are much worse. These are not lightweight.

Are the generators themselves CAT as well? Or are those only the engines?

[Delta]

Didn't get a chance to do any testing today....

The complete generators are Cat. The alternators are self excited from a PMG unit, with a rotating rectifier providing the field current to avoid the need for slip rings.

There are a LOT of direct-on-line 480v motors on board.

Am I safe connecting a 300v rated isolated scope meter with 10:1 probes to a 480v supply? I don't think it is...
I will at least probe phase-neutral on the 220v board (127v) and post a screenshot, but I'll be going through three transformers at that point so it might hide some of the nastiness.

[expinkolator]

As the engines throttle up, the current draw from the separator motor increases to 5.3A, dropping back to 3.5A as the load drops off.

speed is maintained at 1200rpm plus or minus 1rpm, so no frequency swing.

The engines wouldn't throttle up without a change in frequency. Centrifuges are usually high inertia loads so a small change in frequency may result in a large transient change in load.

[Delta]

As the engines throttle up, the current draw from the separator motor increases to 5.3A, dropping back to 3.5A as the load drops off.

speed is maintained at 1200rpm plus or minus 1rpm, so no frequency swing.

The engines wouldn't throttle up without a change in frequency. Centrifuges are usually high inertia loads so a small change in frequency may result in a large transient change in load.

The frequency swing is tiny. 2/1200 rpm = 0.1Hz. Even if it was a few 0.1s of Hz, would it really cause such a change in current draw?

On second thoughts, it is not a big frequency swing, but it is a fast one, so hmmmmm.....

[Kalin]

Have you verified that the seperator has been cleaned recently and is not just struggling with the increased load? From my understanding the fuel flow through the seperator would be proportional to the generator speed and that would be the first place I would look.

Sent from my SM-G930W8 using Tapatalk

[expinkolator]

The frequency swing is tiny. 2/1200 rpm = 0.1Hz. Even if it was a few 0.1s of Hz, would it really cause such a change in current draw?

If you are measuring the centrifuge rpm you are not measuring the frequency. Load depends on the difference between rpm and frequency - the slip. Seeing little rpm change helps to explain the load change.

Is it spinning a large heavy rotor? how long does it take to spin up from rest?

[Delta]

Have you verified that the seperator has been cleaned recently and is not just struggling with the increased load? From my understanding the fuel flow through the seperator would be proportional to the generator speed and that would be the first place I would look.

Sent from my SM-G930W8 using Tapatalk

The separator does NOT supply fuel to the engines, it simply sucks from the bottom of a vented tank, and discharges back into the top of it.

[Delta]

The frequency swing is tiny. 2/1200 rpm = 0.1Hz. Even if it was a few 0.1s of Hz, would it really cause such a change in current draw?

If you are measuring the centrifuge rpm you are not measuring the frequency. Load depends on the difference between rpm and frequency - the slip. Seeing little rpm change helps to explain the load change.

Is it spinning a large heavy rotor? how long does it take to spin up from rest?

That is engine rpm, so directly proportional to frequency, I have no way of measuring separator frequency.


Thanks for all the replies, I didn't get a chance to do any more testing, and I'm now waiting for the helicopter to take me to the pub.  I will pick this up again next trip!

[Kalin]

Oh ok. Sorry about that. Not like the setup I had in mind

[duak]

I think I may have something on the problem.  I work with small (1-2 KW) 3 phase systems in CNC machine tools but don't have first hand experience with bigger systems or systems with this particular problem.  I have seen weird problems in line current waveforms in one machine when the neighboring machine is decelerating and regenerating energy back on the the line.  I used a clamp on ammeter probe connected to a 50 ohm load resistor and monitored the line waveform with a scope.

Three phase distribution systems can have problems with the harmonics in distorted line voltage waveforms caused by DC bus rectifiers.  If memory serves, the 3rd harmonic of the line frequency causes something called a Negative Sequence current to flow in 3 phase machines connected to the circuit.  This current is a phasor that rotates counter to the direction of the line frequency phasor and in an induction motor tries to turn it the opposite way.  I believe it will cause the line current to the motor to increase and cause it to heat up more than it should.

I would try to monitor the separator motor's line currents and see if they are distorted.  I'd also check to see if the motor starter overloads are set correctly.  If higher than they should be, then perhaps someone adjusted them to solve a tripping problem.

I don't know if this is a problem - ie., are motors mysteriously burning out or are the overloads tripping at odd times?  If there is a problem, I think the ultimate solution might be adding or increasing the size of line reactors to the bus rectifiers or some sort of magic transformer that blocks the negative sequence currents.  I recall something about delta-wye transformers doing this, but I'm not a power guy.

Cheers,