Current imbalance in three phase chiller motor 7

[Run1on]

I have a current imbalance in a three phase chiller motor (480v 300 hp)Each phase: 1- 200 2 - 215 3 - 252. The voltage imbalance for this is less than three percent. We are also seeing other motors in the facility experiencing current imbalance. We also see the power factor readings for each leg can be more than 1o percent diffeent. If one motor suffered current imbalance I would suspect a motor problem but five tend to lead to a distribution problem. Is it possible several nonlinear loads (277 lighting, etc.) can be a contributing factor?

Run1on

 

[Togel]

A phase imbalance that significant may shorten the life of your motors. I have seen this before and it is generally a power source issue. Are you on generator power? Your energy provider is responsible for providing three phase power with a power factor with 0.95. As an immediate precaution, I recommend the purchase and installation of power factor correction at the motor starter itself. Additionally, ensure you have motor phase protection for your motors both source and load side of your contactors. Both can be purchased from your chiller manufacturer. With this in mind, I would check for severe wear on your contactors in the motor starter. Phase imbalance accross your contactors will cause the current to spike as the voltage drops at the point of brown out and circuit interruption. The current spike will most likely be oversize of your contactors and flash arrestors, and cause excessive melting and carbonation which will further cause a larger phase inbalance when the circuit is broken. If this is the case, order new/larger/vaccuum type contactors for backup purposes and limit the number of shutdowns as much as you can.

 

[Run1on]

Togel:

The customer has lost their contactors once already. I will schedule a meeting with the local utility and request they meter their incoming supply for minimal requirements. This customer is at the end of the line and I suspect the KVAR generation of the utility is not sufficient to support their needs. As such we may have to install the necessary KVAR correction.

Run1on

 

[Togel]

You will also note that if the power company refuses to assist with voltage correction, that correcting phase imbalance will also yield savings in power consumption. This savings can be sold to management of the facility to install automatic power factor correction with the future power savings. What is the nature of the facility? Hotel? Tell the power company to bring their PPA to the meeting so that the proper documents are present to clarify the power quality issue. I have even used that tactic of informing the power company that due to the over consumption, it is being cinsidered to install stand alone redundant generators to to run the facility independant of the power provide. Such action would also compensate for low KVAR and in-rush problems during start up of the chillers causing flicker in the rest of the facility.

 

[Run1on]

It is a school and we have found that another school within the grid of this utility has experienced similar problems.

 

[electricpete]

I've heard that current imbalance can be approx 6 times voltage imbalance. So the current imbalance you report could easily be the result of a voltage imbalance of 3% or less.

I would point out that it is not the "nonlinear" nature of the loads that cause an unbalance problem, but rather the fact that single-phase loads (like the 277v you mention) may not be properly balanced among the three phases.

I'm not sure that capacitive support is necessarily the solution to unbalanced voltage. It'll help raise the overall voltage... but I don't think it'll reduce the unbalance. The way I look at it you either need balanced single phased loads or else a lower impedance path from the source (possible add parallel lines) which would reduce the effect of a given imbalance.

 

[electricpete]

To add a little more discussion on ways of correcting imbalance.... load compensation settings of regulators may help to reduce the effect of load imbalance for people at far end of the line.

 

[Run1on]

One other item to mention: the controller for this chiller has display capabilities and as such we are able to watch the operating characteristics of the motor.
The chiller's current draw is not consistent some of which can be due to load demands and possibly due to a load limiting functioning programmed into the unit . The unit is drawing nowhere near the FLA high end of 400 amps between 200 and 270 is normal.

I want to thank everyone who has responded. It was a exciting moment to find a site where I can receive help from fellow engineers.

Kevin
Run1on

 

[Togel]

Voltage correction is a good solution as it will be a more cost effective way of mitigating damage to the motor windings by correcting power at the central plant. Performing a coordination study of the facility will/may solve the problem at a much higher dollar value. If this is pursued, I would involve the original design engineer or contractor as they may be culpable for replacement costs. Voltage regulation will help, but keep in mind the poor power factor and wear on the contactors. Without capacitence and some form of voltage regulation, i.e. correction of your power factor at the central plant, you will eventually burn out the motor windings and continue to eat contactors.

You find that your chiller FLA readings will continue to raise and lower. Check to see what the purge count on the chiller is. It only takes a few purges to knock your vanes out of calibration. This will cause a chiller to "run all over the place" trying to find a set point that satisfies the load condition. This will be especially apparent during low load conditions. I would have the service provider/chiller manufacturer return to re-calibrate the chillers. Is your chilled water temperature remaining stable? Are sufficiently removing humidity from the classrooms? Have ever experinced any flow problems with either chilled water or condensed water?

 

[peterb]

It doesn't seem reasonable to me that power factor correction will in any way solve a voltage unbalance problem, which will certainly give rise to the current unbalance noted and will show up in other motors, as noted. The source of this unbalance needs to found and corrected - this is a case for investigation along with the utility, as the source is quite likely to be external to the plant.
As regards a coordination study, I am at a loss to understand what this would have to do with correcting the unbalance situation. I believe that a power quality study may be in order to identify problem areas and to determine corrective action.
What is the overall power factor, supply transfomer rating, load and measured voltage level within the plant? This info would be a starting point to evaluating the need for/benefit of capacitor addition

 

[electricpete]

good answer, peterb.

 

[busbar]

peterb is on the right track. More numbers about the system at hand will help yield higher-quality advice. There could be a widely varying number of components between
‘the utility’ and a 300hp, 480V motor.

In the simplest sense, voltage is source based (the utility) and current is load based (the school/chiller). Without basic but systematic measurements and comparison to labeled ratings it is hard to draw reasonable conclusions.

At this point, consider the skilled use of a basic power analyzer–something that can look at waveforms, and not just the raw voltage and current quantities from a multimeter. A “walk around” by a person familiar with the site (and in possession of system drawings) and someone who is experienced in identifying the effects of electrical interactions can be time and money well spent. Because of the potential cost of chiller electrical damage, you may want to consider retaining–or at least consulting with–an electrical engineering firm or service to interpret the measurements if they’re not well understood by the immediate user. The technical person or firm should be able to identify the problem and prepare a report whose information will ultimately influence the possibly non-technical administrator that has budgetary authority. If the problem is allowed to persist and failure(s) occur, you can bet the administrator and his superiors are going to be asking some embarrassing questions about how their maintenance funds have been spent.

 

[electricpete]

Buzzard - good discussion. I can understand the process of rotating leads to determine whether problem lies in the motor or the power system. But I'm having a hard time seeing that it would reduce the imbalance. For only a problem with a power system (symmetric motor), I would think the currents should look the same except shifted one lead.... for only a problem with the motor (balanced power supply voltages excluding imbalance induced by the motor... ie balanced voltage with motor secured), then I'd expect no change at all when swapping leads.

The only time you could actually reduce the imbalance is if you had unbalanced system voltages (even with motor secured) AND unbalanced-impedance motor. Is that the situation you're referring to?

 

[Run1on]

I have some more information to pass along on this subject. We received downloads from the chiller unit and compared time frames on the voltage and current levels. When the chiller is not running, the voltage readings are fairly constant; not more than 2-3V apart. Once the chiller is called, the voltage takes a tremendous drop on one leg;
i.e. 1- 245, 2-275, 3- 278. The corresponding current readings for this time frame are:
1- 2 , 2- 66, 3-510. We see this throughout the entire data record which is two days long sampled at one minute intervals at every startup occurance. We are in the process of taking readings on the ground to see if a partial fault or phase to ground event is occuring.

 

[Run1on]

It seems that the problem lies potentially in the electric system. Diagnosing problems is undercover work. We have found out that the chiller technicians "rolled" the leads to the chiller at the breaker located at the MDP and the imbalance stayed with the phase not the motor. We are going to trace this phase throughout the system and see what, if any, item may be the cause. If we find nothing it would have to be in the utility supply. Agree or disagree?

Thanks

 

[electricpete]

That's a lot to chew on. Please excuse my thinking out loud....

Some more discussion on dividing the problem into motor or supply (and by supply I mean in-plant plus utility):

Normally the observation that the location of high current stayed with supply phase vs motor phase during lead swap would lead you to suspect the supply, not the motor.

Normally the observation that the voltage was balanced without the chiller running would lead you to suspect the motor, not the supply.

You mentioned current imbalances on other motors also, tends to point toward the supply since it's not likely all motors have a similar problem.

Put the above together and you start to suspect that it is a supply-based problem which only shows up when you put a motor load on it. perhaps a high impedance on one phase of the supply circuit which doesn't show up until you put a load on it.

But there's another piece - the currents you reported
"1- 2 , 2- 66, 3-510". Obvsiously a HUGE imbalance. (When did it changve from 1- 200 2 - 215 3 - 252 as originally reported?). In fact there is not enough current within phases 1 and 2 to provide return path for current from phase 3. So you have to suspect an alternate return path through a ground.

But if the ground were within the motor, then your swapping lead trick would have left the high current at the same motor lead. UNLESS - the location of the ground was upstream of where you swapped the leads but downstream of where the motor is normally switched on/off. Assuming you swapped leads at motor terminal box, this could be anywhere between the load side of your starter and your terminal box.

Getting a little far into the speculation stage. If there is a ground it shouldn't be hard to find. Start by meggering from motor with all leads attached. If ground detected, then disconnect leads and see which side it's on.

Let us know if you get more info.

 

[peterb]

Run1on - The readings that you report don't represent any kind of 3-phase load, so I suspect that there may be something amiss with your chiller circuit or with the metering. Can you get independent readings on the supply to the chiller when it is running? Does the chiller function correctly??

 

[busbar]

You’re better off building an air-tight case before extending any blame. Careful measurements and documentation are vital for resolving the problem.

From a purely mechanical aspect, did the chiller seem to be operating acceptably? With the listed current readings, the rotor would likely be stalled. Have [chiller] readings been verified as correct at the motor terminals with known accurate instruments? The wide range of current readings are suspect, but you know the history of the system better than others here. Remember that, for utility equipment to be involved, the readings must be reflected at the service entrance; i.e., where the utility and customer’s facilities meet.

(For the sake of other readers, phase rolling to match service conditions is discussed in §11 at:

http://www.pge.com/002_biz_svc/pdf/058087g.pdf

but my understanding is that the process is intended for a one-motor / one-transformer arrangement.)

 

[jbartos]

Suggestions:
1. Voltage imbalance is often related to load imbalance on multi-phase supply systems.
2. Traditionally, first attempt to rectify the problem was to use clamp-on ammeter and measure currents at panel's feeders to find out the load imbalance. If the load imbalance is spotted some one or two-phase circuits, causing the load imbalance and potentially the voltage imbalance, have to be switched to different phases. This may need several tries since the certain loads may or may not be on, but it is worth it. The three phase loads are usually fairly balanced. If they are not, they are experiencing some internal malfunction, either open circuit nature or short circuit nature. This procedure is widely used and yields reasonable results. It is a good idea to record reading and perhaps input it in a good spreadsheet without bugs for quick updates.

 

[Run1on]

Thanks for all the input. To answer some of the useful questions posed:

1) The chiller runs reasonably well after startup; it has a problem with cycling that needs to be addressed by the manufacturer.

2) We just finished checking other feeders from the MDP and found current imbalance on almost every breaker in the panel.

3) We turned the chiller off at the MDP and the current imbalance was still present in other feeders

4) There was only 2-4 amps to ground and various neutrals showed minor amperage.

5) the feeders from the utility transformer showed only a 1 volt deviation among the phases buring both chiller operation and stoppage. They have said they will help set up monitoring of the transformer and are providing paperwork from tests done previously. They state they are in ANSI compliance.

6) We have decided to take the approach noted by jbartos and examine every panel for current imbalance and adjust as necessary. It is possible since there was a major renovation at this school, the engineering firm may not have done an adequate job of load balancing between new and old circuitry. Will find out.

Thanks everybody