Motors and Starters Burning 3

[hbendillo]

Here is the set up:

Medical duplex vacuum pump (alternting 10 HP motors) fed at 480-volt three-phase. Circuit parameters are 3-#8 & #10 G., 30-amp molded case circuit breaker. The pump is located in a boiler room which of course gets hot but we have not determined how hot. They only measured the temperature in their once and it was 104 F. The distance of the vacuum pump from the electrical source in terms of wiring length is no more than 320' but probably only about 300'. This pump is being fed from a panelboard that is on a separate service than is serving this part of the hospital. The service in question was for an addition to the existing building.

Motors and starters are burning up. Have replaced both motors and both starters. It appears the terminals on the starters are getting very hot. The circuit breaker has never tripped as far as we know and apparently the overloads in the starters are not taking the motors out. I don't see an severe arching within the control panel but when I looked at it the old starters had already been removed. A little plastic wiring tray for control wires that run in the control panel has a part that is melted and this part is right under one of the starters. Sounds like some kind of ground fault condition. The voltage on the circuit has been measured and is reading about 485 I think.

Any thoughts about what is happening. When the owner gets the unit up and running I am going to get him to install a meter to mearure and record voltage and amperage. But I don't want to burn up another motor or starters if I can help it. Thanks for the input.

 

[itsmoked]

Yes the voltage drop on that long wire run is causing the starters to chatter... This causes immense arcing on the starters and thrashes the motors. It burns out one pair of contacts which promptly leads to a loss of phase to the motor which immediately fries it.

Keith Cress
Flamin Systems, Inc.- <

http://www.flaminsystems.com>

 

[rmiell]

I don't think that voltage drop is a factor here. The motor rating for a 10hp motor at 480 volts is 14 amps. Using a voltage drop calculator I come up with a required size of #10 for 3% drop. A #8 is even better. Installation of a phase loss/reversal, as well as an undervoltage/overvoltage unit will help, if phase loss does happen alot. Correct fusing will also protect the motor.

Only further investigation will show the true cause.

Rick Miell

 

[waross]

Hello itsmoked and rmiel;
I'm about in the middle in this. I did some iterations on a spread sheet with different voltage drops. I was doing a lot of estimating and extrapolating so corrections are welcome and even expected.
My methodology was to start with a voltage drop and estimate the starting current at the reduced voltage. Then take this current to the voltage drop tables to determine maximum allowable footage for #8 AWG copper.
My estimates showed that about 15% voltage drop under starting conditions would result in about 60 amps starting current. This equates to about 310 feet of #8AWG. Starting torque will be about 72%.
Is this a new installation? Is this a type of vacuum pump that requires a lot of torque to start? Workmanship may be a factor if the motors failed soon after installation.
I have found that 7 strand #8, #6, and #4 are particularly prone to loose connections. I have demonstrated this numerous times. Put a #6, seven strand cable into a connector and tighten the connector. Twist the cable about 15 or 20 degrees so that the strands untwist and watch the cable fall out of the connection. If the installation was done by the same person there is a possibility of multiple poor connections. These connections will start to heat cycle and corrode and get worse. I saw one instance where a switch was connected to a bus gutter with a 6" nipple. Metalic bushings were used. (Legal at that time and place). The loose connection at the connector on the switch heated the cable. The copper conducted the heat and the cable insulation at the bus gutter end of softened enough to let the copper push through and short out to the metalic bushing. It was a long time ago but as I remember the cable was a #4 or #2.
I saw another installation with multiple dry-type lighting transformer failures due to poor connections.
I suggest that you inspect the contactors if possible. If the problem is as itsmoked suggests, then the contacts will be badly burned and melted. Some contacts may be burned off.
The connection terminals will be in fairly good shape as the heat has been originating at the contacts and will have been conducted to the wire terminals.
If the problem is poor connections the heat will have been originating at the wire connection terminals. One may even be burned off. The contacts will be in quite good shape because the heat will have originated at the wire connections and been conducted to the contacts.
Excessive heat damage of the cables and melted insulation on the cables is also a sign of poor connections.
Either excessive voltage drop under starting conditions or poor connections can lead to motor burnout.
Let us know what you find after an examination. As far as recordind instruments, I would suggest measuring the voltage and current at the time of start-up of the replacement motors and starters. Contact chatter (We called it "Machine gunning", same thing) will be readily evident. Recording instruments will tell you what happened after the next motor burns up.
respectfully

 

[hbendillo]

The contacts were badly burned and melted. I also did a voltage drop calculation and thought that would not be a problem. But as suggested voltage drop on starting may be a factor. Maybe this pump is cycling a lot therefore lots of starts. Guess that could be a factor. The damage does seem to originate at the contacts of the starters. I will furnish additional info when I can. Thanks for the input.

 

[itsmoked]

hbendillo; Lots of people seem to do the straight FLA calc and then decide xAWG is just fine. They forget that on starting the current is maybe 6 to 10 times greater, causing a much larger voltage drop then their steady-state calculations showed.

Some motor starters use one of the supply legs for the coil power which means the coil sees precisely the voltage drop that hits on start-up. Sometimes you can fix the problem by changing the setup to where the coil is run entirely by a different source or at least the same source but from far enough up stream to not be effected by the starting voltage loss, hence the contactor doesn't fall out. "Machine gun".

Machine gunning or chattering will kill a contactor after about 4 seconds so all you need is one short bout or a couple of real brief ones.

Keith Cress
Flamin Systems, Inc.- <

http://www.flaminsystems.com>

 

[waross]

hi hbendillo
After your last post, I have to agree with itsmoked.
And I have just come in from looking at a small comercial oven for a friend. The normal 208 volts was 205 volts. When the oven is on the voltage goes down to 185 volts. One day the incoming voltage was low and the contactor chattered/machine-gunned and died. (I think the difference between chattering and machine gunning is the smoke and flames. With marginally low voltage a contactor can chatter but not actually open the circuit. Does this terminology sound okay to you, Keith?)
The same recomendation in both cases.
Bigger supply cables.
respectfully

 

[itsmoked]

I would probably call that a humming/buzzing relay and that will probably fry the relay's coil eventually because it is drawing a larger current than it is supposed to. This is because the armature is not staying sealed and it results in the coil's impedance dropping -> current rising -> smoke leaving.

Chattering/motor boating/machine gunning are all the same to me.

I did a Detroit Diesel governor modification so I could pull a 3-71 to idle. I used a laundry solenoid, which is a class or solenoid. If those suckers didn't pull all the way in they smoked in about 3 seconds. Current would go from about 15A,150Vac to about 0.2A in the last 0.060" of the stroke or 2,250W to 30W sealed. Hence the rapid smoke letting.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[Laplacian]

In your first post, you state that the terminals seem to be getting very hot. Exactly what part of the contactor is failing, the contacts, or the terminals? Fried terminals indicate poor connections which will increase voltage drop to the motor.

Your second post states that the contacts are failing. Fried contacts could be caused by frequent cycling of the motor; is this a NEMA or IEC framed contactor? IEC framed contactors are more duty specific.

 

[hbendillo]

Thanks again for the input. I think ya'll are on the right track. Unfortunately the owner had already changed out the contacts by the time I got to see it. Only going on a description of the damage. So, does your hypothesis also explain why the motors burned up? They replace two motors previously but this time the motors at least did not fail. They brought one to a shop and it checked out OK. Don't know whether it is a duty or IEC framed contactor.

 

[itsmoked]

Yes my hypothesis is correct. I can pretty much guarantee they will burn these contactors and motors up too.

Replacing the motors and contacts without fixing the problem strikes me as pretty dang dumb. Their cost of "vacuum" is going to be pretty high.

As I explained the motors may take the voltage drop in stride as long as the load (vacuum pump) reduces during the voltage sags. But once one contact fails the motors are "single phased" and they WILL die(in seconds). So it is only a matter of does one contact(single phasing/nearly instant motor destruction)[flush] or two contacts(no phasing/no motor damage) occurs first at the tortured contacts.

Keith Cress
Flamin Systems, Inc.- <

http://www.flaminsystems.com>

 

[waross]

Keith is correct in his statements.
When the contactor closes, excessive voltage drop causes the contactor to drop out. It is carrying locked rotor current at that time so the internal arc is much more than normal normal motor stopping duty. As soon as the arc stops the voltage rises and the cycle repeats. This happens several times a second.
Next. The contacts may weld, but more frequently one will burn off.
If the first contact to burn clear is on one of the phases that supply the coil then the voltage drop to the coil will be less than half what it was. The contactor will probably hold in now, and will supply single phase power to the motor. This can burn out the motor before the thermal overload protection operates. If the thermal overload has automatic reset, it will probably automatically burn the motor out on the second or third reset.
If the first contact to burn off is on the phase that does not supply the controls, the cycle will probably continue until a second phase burns clear in the contactor. At this point the current to the motor is interupted. This happens very fast and the motor will often survive.
respectfully

 

[patrick7]

I agree with these previous posts on this problem,
I would install a portable fault recorder and set the trigger levels to capture everything with a slight drop in voltage. Then retrieve the events from the recorder daily and look at the waveforms these will tell you exactly what is happening with this motor. You will be able to see the voltage and current waveforms and analyze there performance throughout any given day. You will also see exactly how many times the motor starts and any effects of multiple starts within a short timeframe. This will allow you to correctly advise the owner of what corrective action to take prior to the motor failing. Also, a thermal imaging devise will show you if your contacts are getting hot.

 

[Laplacian]

I'm having a hard time with why starting a 10HP 480V motor is causing enough voltage drop at the MCC to make a contactor chatter. I've seen this happen for much larger motors with inadequate transformers; 10HP is small by comparison.

hbendillo - what is the size, and the demand load from the transformer that feeds this MCC? Also, can you post the contactor nameplate information?

 

[waross]

Hello Laplacian
Consider the 300 ft cable length, and calculate the voltage drop at locked rotor current. Then allow for low incoming voltage at peak usage times.
respectfully

 

[Laplacian]

Is the starter at the location of the motor or near the source transformer?

My experience is with a large industrial plant with MCC's in one location fed from decently sized oil filled transformers. Voltage drop at the motor would not factor in at the starter unless the entire transformer secondary dropped as well which happens while starting large motors. If the starter is mounted near the motor, then I can see how this would be an issue.

 

[itsmoked]

That's what we are assuming.. Starter near motor at end of long run..

Keith Cress
Flamin Systems, Inc.- <

http://www.flaminsystems.com>

 

[jraef]

Although I stand in itsmoked and waross' camp that this is contactor chattering caused by voltage drop, there is another related possibility here, and having built dozens and dozens of medical vacuum pump controllers I can attest to this being a real issue.

You could also have a control circuit problem. Medical vacuum pumps typically have at least 2 control points; the vacuum switch that calls for the lead pump when vacuum in the receiver is lost (pressure rises), and a receiver high level switch that opens a dump valve to drain the waste material picked up by the vacuum (after shutting down the pump of course). Both of those switches must have a proper hysteresis in them, the difference between changing state and resetting, otherwise they will flutter. For example on the vacuum switch, on at 10in.Hg, off at 30in.hg, so there is a 20in.Hg hysteresis where the pump runs even though you have attained vacuum greater than 10in.Hg. Sometimes the switches are really cheap and have no hysteresis, in which case the controller must have a Minimum Run Timer incorporated that prevents the short cycling of the starter as the switch flutters back and forth. What I have seen happen most often is that they have the cheap switch design, but the timer fails or someone turns the time down to zero because they can't figure out what it's there for. Really common unfortunately. If it flutters fast enough, the alternator relay doesn't pick it up and switch pumps, so the same pump motor is banging on and off really fast, just like what was described for a low voltage condition.

By the way, the other way a motor can be lost to this is if 2 sets of contacts weld, because then you have power flow to 1 phase of the motor windings. Many / most overload relays will not pick that up.

Eng-Tips: Help for your job, not for your homework Read faq731-376

 

[hbendillo]

To answer your questions, yes, the starters are installed at the motors. This is a packaged unit with control panel and motors part of the vacuum pump assembly. Thanks for the info jraef, if I can't solve the problem correcting the voltage drop problem with the starter coil we will look at your possible issue.

 

[imok2]

You may want to check the voltage at the coil of the magnetic starter at the same time the motor is started. In my situation, the starter contacts chattered until the amperage draw of the motor decreased as gained speed. If the voltage on the starter coil is low, consult Table 430-7b of the NEC and calculate the amperes required.