Autotransformer start on 4,000 HP 3 stage air compressor, activating at full speed?

[fijensen]

I'm an elecrical specialist. I've been working on MV motors, and controllers for many years. I recently had a call to a facility with mystery trips on a compressor (see title). They had some data recorded by the DCS which showed the current increased from 280 to 500 amps, at which time the machine would trip on high vibration. Sometimes it would run for weeks with no problems. The machine had all internal components replaced (very expensive) 2 weeks ago. I should point out that during the trip events there is no increase in pressure/flow, so guide vanes were ruled out. Last week it tripped again, and I was sent to check it out. The motor tests good with 2.5 KV megger, and DLRO. The only thing that I could think of electrically was; possibly the R contactor might be dropping out (R shunts out autotransformer used for starting). I'm not sure what this would do mechanically if it happened? I brought the controls up in test, and started testing voltages. On the economizer (ATC CEM) that feeds the R coil, I read 8.9 VDC (aha!). The output is spec'd out to be 20 VDC. I ordered a new one and changed it. The new one puts out 20.4 VDC. I checked the old one with a scope. It was putting out 120 VAC. That's a lot of ripple. I think that has to be it! They start the compressor, run it overnight unloaded. the next day they start to load it, and it trips on vibration. I haven't seen the trend on most recent trip, but previously current was first indication followed by vibration. Mechanics are saying case of gear box is distorted? That's hard for me to buy, after seeing what I saw? If it had mechanical issues that bad, how could it run for weeks at a time with no problems? I'm still wondering about that R contactor. Could the AC from the economizer have done something strange to the DC coil, or armature? I'm going to swap it out with the M contactor next week. If M drops out, it stops the machine. R is more problematic to drop out at full speed. Has anyone ever seen anything like this? It's definitely not your average problem. Thanks In Advance, FJ
Attached power diagram of similar installation.

 

[jraef]

Assuming it is a Korndorfer starting scheme, if contactor R dropped out after the motor accelerated (meaning that contactor N is already open), the voltage to your motor would drop precipitously because the AT windings would be in series with the motor. In the Korndorfer method this is always done, but very briefly, as a form of closed transition starting to avoid torque transients. So if that happens again later when under load, your motor torque would severely drop, the slip would increase significantly and the motor would begin to draw high current but would not likely allow the motor to remain accelerating, so it would stall in relatively short order. I'm not a vibration expert, but I suppose it's possible that your vibration detection is somehow interpreting the stall, or a torque transient as the motor slips behind the BDT point in an impending stall, as a chatter in the machine.

I would however think there are at least a couple of more likely issues.

1: One of the contacts in R is failing and intermittently not making good contact. If that happens, you will get a sever current imbalance (because of the added voltage drop of the transformer on that phase akin to above), something that would definitely show up in a vibration monitor. A simple voltage check across the contacts of the vacuum contactor would detect that, although nothing is actually "simple" in MV...

2: You have a broken rotor bar in the motor. I've seen on numerous occasions where this seems to be a "phantom" event, but in reality is is tied to adding the extra stress of loading that was not there during an unloaded acceleration, as is likely the case on a compressor started by an RVAT. There are other ways of detecting a broken rotor bar with a vibration spectrum analysis, that might be relatively simple if your vibration sensing system is capable of capturing data.​

"Dear future generations: Please accept our apologies. We were rolling drunk on petroleum."
— Kilgore Trout (via Kurt Vonnegut)
For the best use of Eng-Tips, please click here -> faq731-376

 

[fijensen]

We aren't gettin an unbalance. We have a Multilin 469 on it. It trips pretty fast on unbalace. We're having PDMA testing done on motor. I was wondering if; when R contact re-introduces the starting transformer at full speed, it might cause axial thrusting in the motor, that could possibly damage compressor housing? I'm going to see if we can get a look inside of motor as well.

 

[waross]

It may be worth a second look at the gear box.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[jraef]

Possible. I've seen it on pumps.

But I'm not tracking your statement about how re-introducing the starting transformer after already running is going to cause axial thrust to begin with. You will get a drop in voltage and thereby torque will drop at the square of that voltage drop. Is it a vertical motor where the loss of your allows it to drop off of the thrust bearings or something? Not sure I understand that.

Another though occurred to me. Are you sure this isn't perhaps the Autotransformer being left into the circuit too long and it is heating up, tripping the thermal cutout, which drops out the R contactor? This can happen when, like in a compressor, there is something like an unload valve that is sticking closed, or someone has messed with the timing (usually in a misguided attempt to solve that problem). Typically the thermal switch on the transformer core is supposed to be in the control circuit for M, not R, but someone may have inadvertently wired it incorrectly.

"Dear future generations: Please accept our apologies. We were rolling drunk on petroleum."
— Kilgore Trout (via Kurt Vonnegut)
For the best use of Eng-Tips, please click here -> faq731-376

 

[fijensen]

It's a horizontal compressor. The mechanics and engineers are thinking it may be a case distortion on the gear box. They changed all internal parts a few weeks ago. The compressor was sent to the manufacturer (Cameron)this week. This has become a high visibility problem. It has cost a lot of money. I'm trying to make sure that if it trips again, it won't be due to electrical problems (heads might roll). I know the economizer on the R coil was bad. (8.9 VDC output, with 120 VAC ripple) It seemed to work alright, but this compressor would sometimes run a month without problems. The trending always showed elevated current as the first indication, followed by vibration trip. My thought was; with the machine running full bore, if the starting transformer was instantaneously introduced to the power circuit, the phase angle change, and lower voltage would try to brake the motor? I've never seen it happen. I was told that a guy with 35 years experience said that he has seen them do that.

 

[jraef]

It would not "brake" the motor, but you would have a rapid loss of torque output from the motor. That's not the same thing, although someone not familiar with the issue might perceive it that way if it were a high friction machine that needed full torque to keep running.

"Dear future generations: Please accept our apologies. We were rolling drunk on petroleum."
— Kilgore Trout (via Kurt Vonnegut)
For the best use of Eng-Tips, please click here -> faq731-376

 

[fijensen]

So, you think it will slow down with no current increase, generator effect, or vibration? I don't think it will like being instantly locked to a different voltage while running at full speed, especially with that momentum. It wouldn't be coasting.

 

[waross]

" I don't think it will like being instantly locked to a different voltage"
Voltage doesn't control the speed, frequency does. A drop in voltage will mean a drop in torque and current. The drop in torque will most likely result in a drop in speed if the motor is loaded. As the speed drops the slip frequency increases and then the current increases. It depends on the load but with full load on the motor and a drop in voltage the current generally exceeds normal full load current.


Bill
--------------------
"Why not the best?"
Jimmy Carter