Chiller Tripping Due to Current Imbalance 4

[CtrlAltDel]

Hi guys,

My questions are: How do I know if the contactors used in the motor starter is properly matched to the load? Could an under-rated contactor aggravate the situation of voltage imbalance at the contactor terminals thereby causing yet higher current imbalance to the motor?

Let me explain:

A customer reported recurring occasional and failure to start of it's chillers due to "current imbalance." The set point for current imbalance is 15%. At 1.1% utility voltage imbalance, this current imbalance setpoint is reached thus tripping the chiller.

When they run the chiller on in-house generators with 0.5% voltage unbalance, no tripping occurrs. There was current imbalance of 3.5% which will not trip the chillers.

Utility voltage however, is allowed by the Regulatory body to vary up to 2.5% maximum at the point of common coupling (metering point) in which case the previous 1.1% is still within allowable range.

The latest event was that, even at utility voltage unbalance of only 0.4% the chillers already tripped. Voltage unbalance at the motor terminals was not known.

My intial survey of the site revealed that the motors are Y-Delta started with rated line Amps of 708A, at 380Volts, 60Hz. this would make the each of the two delta contactor (running) carry a full load of 708A/1.732 or 408Amps.

The delta contactors have this rating:

AC-1: 1000V, 500A
AC-3: V 240 400 415
kW 187 312 324

by ratio, kW at 380V= 296.4kW

I have read before that AC-3 should be considered for inductive load such as this motor but how do I make the comparison since the AC-3 ratings are in kW? Is this contactor under-rated?

I suspect that the contactors might not be under-rated, but I am not sure. But one of them had stuck relay contacts as of the latest event causing the upstream breaker to trip. I might be barking up the wrong tree here as far as the contactors are concerned but could an under-rated contactor aggravate the current imbalance condition?

Please give your comments. I would appreciate your thoughts about this.

Thanks.

 

[mc5w]

Unfortunately some wye delta motor controllers are indeed undersized because there was no accounting for the matter that 1 contactor is breaking 1/3 of locked rotor ( about twice the AC-3 rating for typical motors ) leading to short contactor life. The delta run contactor typically has terrible making duty because of residual motor voltage. If starting and stopping only takes place 4 times a day then short contactor life is acceptable - if turning on and off more than that you need a heftier contactor.

The contactor manufacturer should have life versus current broken for different size contactors. One of the older editions of the SquareD digest is highly prized because of the life versus load tables in it for Telemecanique contactors.

The excess wear on the contacts eventually leads to high contact resistance. You also need to check for high resistance due to copper oxide in wiring joints, fuse clips, fuse blades, and so forth.

 

[CtrlAltDel]

thanks mc5w! Just a follow up question:

could a three-phase contactor (3 poles) be used for single-phase application, given that the single-phase load does not exceed the current carrying capacity of the contactor?

I ask because one of the arguments proposed by the chiller technicians is that unbalanced current (though not exceeding rated current) can destroy contactors.

In my understanding is that unbalanced 3-phase currents (where the highest phase current does not exceed nameplate rating of the contactor) does not directly affect the integrity of the contactor except when such current causes the contactor to make-and-break frequently due to protective action of the curent imbalance relay.

Comments are welcome!

 

[jraef]

No, those contactors do not appear to be undersized for that application.

Yes, contactors can theoretically contribute to current impalance when they either a) don't close one of the contacts, or b) become so pitted and worn that there is increased resistance across them. Both scenarios should be painfully obvious if they were happening. Slight changes in contact resistance would not contribute much in the way of a voltage imbalance imbalance. Besides, if that were your problem it would exist in the generator supply scenario as well.

More than likely your utlity supplied system cannot handle the transition spike that occurrs when your starter changes over from Y to Delta, causing a severe spike or dip that is causing the trip.

Current imbalance protection is provided to protect the windings from the increased heating effect caused by the negative sequence currents that result from the imbalance. Small imperfections in the motor windings can also cause diferences in negative sequence currents, and get amplified at lower stator-to-rotor frequency ratios present during acceleration. With DOL starting you would pass through that so quickly that it is hardly noticed. With Reduced Voltage starting such as you have, it gets exacerbated. So slightly imballanced current during any type of RV start is not that unusual. Add to that the contribution from the voltage imbalance from your utility and you end up going over the threshold. Waiting until the motor is at full speed before checking for imbalanced current will eliminate this sort of nuisance tripping. 2 things I would try before sinking money into this (assuming everything else were working fine). They are both similar.

1) Try increasing the imbalance trip delay time on your protective relay. It should have one. If not, it is too cheap of a relay. A good imbalance relay should have a trip delay for this exact reason. I typically go 20 or even 30 seconds.

2) If your relay does not have a trip delay, mask out the imbalance trip altogether during starting by using an aux. contact of the Delta contactor in series with the trip contact of the protection relay (depending upon logic configuration).



"Our virtues and our failings are inseparable, like force and matter. When they separate, man is no more."
Nikola Tesla

 

[CtrlAltDel]

thanks jraef. that made me think.

I already suggested to the chiller service people to either increase the trip delay or increase the imbalance threshold to 25% (instead of 15% provided that none of the phase currents exceed the full load current.) It is also my belief that current imbalance is only temporary and that once the motor has reached rated speed it will stabilize.

Unfortunately, they will have none of it, saying 15% is their standard, although they can manually change that setting higher. They are very adamant on this even though the chiller had operated and started normally for 10 straight months on utility voltage before this problem. I should note that the chiller has resumed operating normally after that problem two weeks ago.

I didn't mention before but the chiller tripped while on utility voltage 10 seconds after starting, giving credence to your suggestion to increase trip delay by 20sec or 30sec.

thanks again!

I welcome more comments!

 

[Marke]

Hello CtrlAltDel

If you operate an induction motor under very light load conditions, the current is primarily magnetising current and this is very voltage dependent. A small imbalance in applied voltage will result in a much greater imbalance in no load current. As the load is increased, the current balance improves and at full load it is close to the voltage imbalance.

All motors have an inherent imbalance in their magnetic circuit due to the grain orientation of the steel used in the laminations. If you rotate the phases around the input to the motor (or starter) you will find that one connection will maximise the imbalance and another will minimise it.
I have used this technique often to minimise the current imbalance at light load.

If the incoming phases are labled as P1, P2 and P3 and the input to the starter is L1, L2 and L3, then connection 1 would be P1 - L1, P2 - L2 and P3 - L3. Connection 2 would be P1 - L2, P2 - L3 and P3 - L1, and connection 3 would be P1 - L3, P2 - L1 and P3 - L2.

The imbalance under light load is not generally a problem for the motor provided that the current balances up as the load is applied. In your case, what you really need is a protection relay that is more tolerant to imbalance at light load than at high load. I have built protection schemes like this in the past.

Try rotating the phases around to see if you can reduce the light load imbalance. Also ivestigate whether the relay can have a desensitised current imbalance at light load.

Best regards,

Mark Empson

http://www.lmphotonics.com

 

[CtrlAltDel]

Hello Marke,

Thanks for the post. Your suggestions make sense. My only problem now is how to convince the chiller operator. They can easily desensitize the current imbalance relay by increasing its tolerance or increasing the trip time delay. My previous monitoring of this motor showed that at only 1.1% voltage imbalance, the chiller would already trip due to current imbalance at around 18%. Utility voltage can not, (by regulatory mandate) and have not yet, breached the 2.5% voltage imbalance limit.

One of my colleagues has this problem on a large water pump motor of a customer which also had nuisance start-up tripping due to current imbalance. When they followed his suggestion to increase current imbalance protection to tolerate up to 20% imbalance, the problem was solved.

Rotating the phases as you said is also an option. At present I have three monitoring instruments installed at the site to determine long term voltage profile and to capture the tripping event should it take place again. If (not when, hehehe) I resolve this problem I will let you know.

Thanks guys, for all your posts! You've been a lot of help!

Best regards.

 

[RRaghunath]

If you are looking to find out whether current imbalance is caused truly by the voltage imbalance or is there a problem in the motor windings, the tip as below may be of help.

"The cause of current unbalance can be quickly identified in three phase systems. The process is referred to as ‘rotating phases.’ The trick is to identify the current readings of each phase, such as: A = 10 Amps; B = 12 Amps; and, C = 15 Amps. You then switch the conductors such that phase A is moved to phase B, phase B is moved to phase C and phase C is moved to phase A. This preserves the rotation of the motor. Restart the motor and re-measure current.

If the current balance remains the same, with the A, B and C phase currents remaining low, medium and high, then the unbalance is due to the motor. If the unbalance rotates such that A = 15, B = 10 and C = 12 (or close), then the current unbalance is due to the power supply.

Tip provided by Howard W Penrose, Ph.D.
T-Solutions, Inc.

http://www.tsoln-inc.com"

The above tip I have come across in the newsletter of maintenance forums.com website.


raghunath_n00@rediffmail.com

 

[CtrlAltDel]

hello rraghunath,

thanks for your post. I talked to the chiller operators and they do not approve of this experiment. No matter. I have already started monitoring the site and as of this writing, the initial data I got showed current unbalance of 7% with utility voltage unbalance of about 0.6%. No problem yet. In case anything comes up I'll post it here.

Thanks again and best regards!

 

[mc5w]

I had an instance where a dirty fuse clip burned up a motor by causing it to run partially single phase. After rewinding the motor refused to get up to full speed at no load and would trip out the overload relays after a few minutes. Doing some amp and voltage checks revealed excess voltage drop in the one fuse clip. Cleaning all 3 fuse clips with silicon carbide abrasive paper and replacing the now gone plating with Ilsco Deox(R) solved the problem. This motor controller was rather ancient but in an indoor location.

I also encountered another instance of nuisance blowing of fuses that where cleaning and greasing the fuses and clips took care of the problem. I also had an instance where copper oxide on some 30 amp fuses impaired the amount of field current in a DC motor. Merely cleaning the fuse barrels made so much difference that I had to do a substantial readjustment of the field resistor.

Yes, 3 pole contactors are often used for single phase motor applications. If the motor overload relay has single phase differential then all 3 poles of the motor overload relay need to be connected in series.

In an instance of electronic ballasts causing prolonged arcing in 277 volt light switches with a resulting loud pop, I recommended hooking up some definite purpose motor contactors with all 3 poles in series. One problem was that the ballasts would work on 120-277 volts nominal without changing taps and somebody else's measurements showed that the ballasts was trying to run on 120 volts ( via the voltage drop in the contact arc ) which prolonged the arc an extra half cycle or full cycle. I have not heard back as to whether this scheme is making a difference but I can assure you that some loads do indeed have abnormal making or breaking duty which requires heftier switchgear.

In other words, the rating of a switching device is for typical applications and if you have a nontypical application expect to spend more money to get a switching device to handle that application.

 

[CtrlAltDel]

Hello mc5w,

thanks for the info. I'll keep those in mind should I come up with similar problems.

best regards!

 

[CtrlAltDel]

Hi fellows,

Just some follow-up developments on my monitoring of the chiller compressor motors. I thought also that perhaps the startup current were too high and caused voltage sags making the imbalance current worse. But this isn't exactly the case.

1. Both motors are connected to the same 380V bus.
2. Only one of them can run at a time.
3. The changeover from motor A to B can occur at any time.
4. Before a running motor is stopped, the reliever is first accelerated.
5. Eight times out of ten, the reliever can successfully start and take over load.
6. Two times out of ten, the reliever motor will trip due to current imbalance about 10 seconds from Start command.
7. No matter which motor is the reliever and it tripped during initial acceleration, its startup tripping does not affect the one already running: no voltage sags on the bus, no flicker, and no tripping of the one already running. I have equipment to capture these fast voltage events. There was no voltage event during any startup tripping.
8. Once any of the motor is already running normally, it has never tripped on current imbalance.
9. My instrument has captured utility voltage sags and in all cases but one, the running chiller rode-through it.

It would seem to me that the default 15% current imbalance with 5 second delay is adequate protection during normal running condition. But in this case, it simply isn't realistic protection during startup.

The current imbalance protection during normal running should not be the same during startup. I have also read in my research that some chillers have this current imabalance protection as part of their high-tech protection, while others have none.

Curiosly, our office have two different chiller systems and connected to the same utility circuit as my subject, yet none of our chillers have ever tripped on startup. Of all the plethora of hvac, mechanical, and electrical protection, our chillers have no current imbalance protection.

Your comments and thoughts are most welcome!

 

[rbulsara]

I would find a way to defeat the imbalanace current function during the start (20 sec or so)..or have longer delay

 

[jraef]

I agree. As I said before, current imbalance protection is mainly used to compensate for the increased heating effects of negative sequence currents if the overload protection scheme does not do so. 10 or 20 seconds one way or the other is not going to make much difference.

If your chiller manufacturer is still being rigid about that issue however, you might try "rolling" your conductors. That means changing the connections without changing the rotation sequence, shifting the leads to the left or right together. Here is a website describing this practice and why it sometimes helps. Franklin Pump Link . It becomes a bit trickier when dealing with Y-Delta starters, but the concept is the same.

In addition to the reasons for doing it mentioned in the Franklin paper, Y-Delta starters have another issue. When transitioning from Y to Delta, there is a corresonding phase shift that occurrs, and that shift can have either a leading or lagging power factor. If lagging, there is additional loss of motor power during the transition, causing additional circuit disturbance. That disturbance may be behind your imballance trips, and that may also explain why it doesn't happen elsewhere or on the motors already running. Rolling just the Y or Delta connection pattern can change that power factor shift from lagging to leading.

Someone used to sell a little device called a Leading Phase Indicator that lit up a small light when the shift was leading or lagging, but I can't find it any longer. Probably the advent of lower cost soft starters has made Y-Delta less popular and so the demand for fixing their problems less important as well. I know who made it and I am contacting them to see if they still want to sell it, I'll post back if I have any success.

"Our virtues and our failings are inseparable, like force and matter. When they separate, man is no more."
Nikola Tesla

 

[CtrlAltDel]

jraef and albursara,

Thanks for sharing your thoughts. You're right about the soft starters as a now more viable option for starting the motor.

By the way, York (one of our chiller brands) implements current imbalance protection on their more expensive units (meaning with factory installed solid-state soft starter; ours is started Wye-Delta) this way:

"... The current balance is only checked atter the
motor has been running for a minimum of 45 seconds
and the motor current is 80% FLA or greater. If the
current in any phase deviates from the average (a+b+c)/3
current by greater than 30% for a minimum of 45 consecutive
seconds, a shutdown is initiated.

If
IøA = 200A
IøB = 200A
IøC = 118A
Then
IAV = 200 + 200 + 118
IAV = 173A
IACCEPTABLE = 173 ± 30% = 121A or 225A
Therefore
Since IøC = 118A which is less than the acceptable
121A, the chiller would shut down if this unbalance
exists tor 45 consecutive seconds."

(taken from York Millennium Document FORM 160.48-O1, page 26)

Now this is what I call flexibility! Please correct me if misunderstood this. First, the motor should be running for at least 45 seconds AND with at least 80% load before testing for current balance. If these conditions are met, balance is tested and if it is more than 30% the motor will not yet trip. It will count yet another 45 seconds to make sure the unbalance stayed and only then will it trip the chiller. These folks seem to know what they are talking about.

Contrast this with my subject chillers with default 15%, and 5 seconds protection and a solution to the problem is quite apparent. It's a pity since this setting is adjustable to 40% and 10 seconds.

On a lighter note, you know, in my typical workday, I often realize that equipment problems are not that hard to solve. They merely follow the law of nature and electromagnetic phenomena. When I gained insight into those phenomena (when and if, hehehe) the solution is quite easy to see.

The most difficult problems, in my experience, are the people behind the equipment, hahaha! No guys, I'm not soliciting moral support. It's true.

 

[jraef]

Yes, CtrlAltDel, you got that correct, and it backs up what we have been saying all along. York engineers obviuosly know why they have IB protection, the other supplier probably has no electrical engineers on staff any longer, so their tech support people just rotely regurgitate whatever was left over from when they used to, and are incapable of thinking it through intellegently.

But now I am ranting too... and you can probably see that I am with you on the issue of the worst problems being with people rather than than equipment.

"Our virtues and our failings are inseparable, like force and matter. When they separate, man is no more."
Nikola Tesla

 

[justinn4]

York scheme makes a lot of sense. Main concern with unbalanced current phases is motor rotor overheat and some thermal degradation of stator winding insulation.
Motor starting last only a few seconds, if the motor accelerates to full speed in reasonable time the heat developed will be proportional to i^2*R*t in the windings in spite of unbalance..
When the motor is running at full load or close to full load, negative sequence due to current un-balance will be more critical since the rotor cage will see the negative sequence at double of the line frequency, the negative sequence voltage induced in the rotor cage is high, in spite of the reduced negative sequence flux, and a parasitic current with negative torque (brake) will overheat the rotor close to destruction when current unbalance exceeds around 40% unless the load is reduced.
The main concern for unbalanced currents is at operating steady condition rather than at starting. May be a minimum acceleration relay will be a better protection while the motor is starting.

 

[CtrlAltDel]

Thanks again guys!

I will keep you updated.

Meanwhile, I'll go over and review my engineering books to refresh myself about all those phase sequence components.

Good luck to you on your work and have a nice day!