Power spikes 2

[ChillerGuy]

We have a customer who over the years have lost several compressor motors (35 HP) on their air cooled chillers. Teardown reports gave no real help except confirming burnt stators.
Phase voltage monitors were installed about six months ago and have tripped but no motor failures. This makes us suspect that the power supply(460/3/60) may have been the culprit.
The customer is a TV station with lots of electronics and UPS systems which have not been affected.
I once heard that power spikes show up the the largest load in the building and in this case the compressors are the biggest motors.

Any truth to this and why?

 

[electricpete]

Motors are susceptible to damage from voltage spikes.

Electronics may also be susceptible to damage from voltage spikes, but they often have built-in surge suppression which protects them.

 

[busbar]

The largest or highest electric-power using components are not automatically the weakest part of an electrical system.

Has the voltage at the compressor terminals been verified to be balanced and consistently within the range specified by the manufacturer? Regardless of advertising promises, voltage phase-imbalance devices are not a reliable means of protection where there are multiple motors at the same voltage as the chiller, or for phase loss while the protected motor is running because this it tends to mask the loss by generating the “missing phase.” They are best intended to inhibit starting where the pre-start voltages are not balanced. Ice-cube/socketed/potted phase-loss relays should be avoided.

Given the cost of compressor replacement and the outage complications, current-imbalance protection may be in order. Because of increased current density in semi-hermetic motor laminations, they are more susceptible to damage than plain-vanilla ‘open-drip’ motors.

 

[rodar]

Check to see if there is any significant Radio Frequency
voltage on the power mains referenced to ground.
This in my opinion could cause insulation breakdown.
If so a common mode Inductor in the form of a large
ring could be used with all three phases fed througt it.

 

[electricpete]

interesting comments.

busbar - can you provide any more details of problems that are more likely in hermetic motors and why.

rodar - I have heard of ferrite beads used for noise supression in a signal circuit. Also a choke is used with electronic drive to smooth input and output power circuits. But I have never heard of any device of the type you mention being used. Can you elaborate any more on whether your suggestion has been practically applied for motor surge protection?

The traditional method for motor surge protection can include a capacitor and possibly an arrestor....but not generally applied at 480v to my knowledge.

 

[rodar]

I have seen servo manufacturers sell a
large ferrite ring to supress the emissions
of their servos by passing the power cables
through this ring. They stated that in operation
the ring will get very hot from dissipating
switching noise. I believe the company was
Emerson Electric.
I thought it not surges damaging the motors but
continual presence of rf to ground which might
heat and degrade the insulation of the motors.
A simple test with an O Scope and High Voltage
probe should suffice to determine this.

 

[buzzp]

I do not agree that voltage protection devices are not effective at detected loss of a phase while the motor is running, at least with my experiences. I would be interested in hearing why busbar feels this way and talking about it. See new post.
Current monitoring can be purchased in the same package as the voltage monitor for a small additional cost for higher cost equipment.

 

[busbar]

buzzp-- I acknowledge that there are newer, more sophisticated protective devices for plant-floor motors and even refrigeration systems, but my comments are based on the "installed base" and smaller, generally more numerous sizes (e.g., ‘35 hp’) and quantities of refrigeration equipment.

e-pete— Semi-hermetic and hermetic refrigeration compressors are, on average, made with less iron than T-frame motors. Although there are some newer chillers that have more advanced protection, there are a great number in service that have fairly unsophisticated compressor protection, and low thermal mass; so a fair portion of hermetic stators now in service have reduced headroom to survive mechanical overloading and electrical unbalance/undervoltage problems. They rely on specific saturated-vapor refrigerant flows for heat dissipation. Stator thermal failures are pricey, for a couple of reasons. Field repairs are rarely possible short of replacement of the entire motor-compressor assembly, and have a fairly high price tag on the replacement new or rebuilt compressor compared to an equally sized T-frame motor. More so than conventional motors, they are somewhat sole sourced, so may be less competitively priced than NEMA-standard motors. Preventive measures like insulation-resistance tracking are rare, and in newer chiller systems, correct operation of electrical protective/display devices are sometimes misunderstood and overridden.

There are many sizes and brands of semi-hermetic and hermetic refrigeration compressors, so interchangeability is less likely than with conventional motors, with the ready willingness to pay premium interstate shipping fees {and in-house maintenance crew or repair-contractor overtime} to limit the duration of fellow-employee {and hotshot junior vice-president} complaints. The rigging necessary to change many hermetic units is often improvised and only marginally safe, given some of the places they are shoehorned in to. Also, burnt enamel is free to chemically contaminate the refrigerant circuit, requiring some longer term babysitting after compressor replacement and consequently larger repair invoices.

There is a tendency to let them sit in the corner, and not be looked at until there a number of ‘filtered’ complaints—but that is not unique to mechanical refrigeration.

These statements have been empirical and somewhat anecdotal, but also of first-hand observation.

 

[busbar]

Aaaggghh! I got it wrong. It’s a copper and not an iron problem. See IEEE 241-1990 (Grey book) §3.11.3: “Hermitic (sealed) compressor motors that are used in air conditioning seem most susceptible to phase-voltage unbalance. Originally, hermetic compressor motors were limited to small sizes, but they are now being built in sizes up to 1000 hp or more. These motors tend to operate with higher current densities in the windings because of {design relying on} the added cooling effect of the refrigerant. Thus, the same percent in increase of the heat loss due to circulating current caused by phase-voltage unbalance will have a greater effect on the sealed compressor motor than it will on a standard air-cooled motor.”

{There is some other related info in that section, buy I’m real slow at transcribing.]

 

[electricpete]

Thanks busbar. I vote you a star for some good info.

 

[buzzp]

Busbar, I agree that typical monitoring devices (cubed, socketed, etc) will not help with transient protection given the short duration. But as I read your post, it sounded as though you were talking about sustained voltage problems (like greater than 1/8 cycle).
Maybe the typical monitoring device doesn't respond quick enough before damage is done in this case. Is this a possibility given that most standard devices (lower cost)have a response time of 6 cycles or higher (60Hz)? Will damage be done in this short time for something other than transients?

 

[jbartos]

Suggestion: The poor power quality of the power supply should not be ruled out as already pointed out above in terms of Radio Frequency superimposed on the fundamental frequency. The harmonics can be harmful starting from the second harmonic, i.e. 120Hz, often caused by half-way rectifiers.

 

[Phlipper]

ChillerGuy:
Good Morning!

What type off starting does this motor/compressor have? Across-the-line? Wye-Delta Start? Soft Start?

We had experience where a customer had two air compressors, fed from the same 600/3/60 bus, but only one exhibited frequent failures. These were wye-delta start. We O-scoped the motor controls within the compressor package, and noted a faulty/mis-operating contactor which introduced a significant spike to the motor each time it was started (the other compressor did not have such a spike). It was our position that the cumulative effect of this switching was the cause of this failure.

We recommended changing of the suspect component, and also recommended re-wind of the motor with 1600 V insulated magnet wire. No problems since.

Phlipper

 

The electrical spikes can produce short circuits in the winding and burned it. It is not normal that they produce the trigger of the protections.
The UPS, power supplies and other electronic equipment, deform the wave of the network tension to which the motors are very sensible, being able to overload and to burn them.
It would be necessary to take a reading of the harmonics rate of the network tension and in the motor current.

 

[jbartos]

Suggestion: The following motor manufacturers concentrates on the high quality of motor windings and documents it in its literature:

http://www.lincolnmotors.com/literature/index.asp

on "spike-proof" motors.

 

[buzzp]

The motor act as an integrator and it will suppress some of these small transients until it starts breaking down the insulation.

 

[jbartos]

Suggestion: Actually, it is the high frequency of ringing and voltage spikes that are damaging the insulation, if inadequate. Obviously, if the insulation is exaggerated, then the motor will have small common mode currents and voltages, and it will run just fine including its bearings.