unexplained 10% high amps 2

[Tmoose]

- A customer has 4 of our hammer mills. All verified to the best of their ability as having identical feed rates, internal condition, etc, etc.
- 3 draw 60 amps
- 1 draws about 68 amps, (not over nameplate) and has ever since being installed 5 or 6 years ago.
Recently the motor on #4 was overhauled just to remove motor condition as a factor for the high amps, since they want to make changes to the line that will work the mills harder, and #4 has no margin presently.

The "report" came back fine from the motor shop, but it still draws 68-69 amps in service.

1 - Is there a motor condition would make a "good" motor draw almost 10% more amps than an identical motor?
2 - Could that condition be detectable with vibration analysis, or some other on line test?
3 - How much extra air gap (turned undersized rotor OD?) would it take to cause a motor to draw 10% higher amps?

thanks,

Dan T

 

[electricpete]

Some questions for Tmoose:

Are you interested to discuss what you have done to rule out a difference in loading from the mill? Others may have suggestions on that. My suggestions included measurement of slip which is very easy to find from vibration, along with other listed electrical measurements... I don't know anything about mills.

What voltage and horsepower is the motor (just trying to get a mental picture).

Is there a drive involved? I'm assuming no, but wanted to check.

Was there in fact some machining done on the rotor that you heard about, or just came up with that as a possible cause on your own?

One other tool in the toolbox during a site visit might be infrared camera. It MIGHT point out something of interest on the motor or the mill, but that is somewhat a shot in the dark.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[edison123]

You could decouple one of the good motors and check its no-load current and compare it with the suspect motor's. That will tell you whether the high current is due to the turned rotor. Assuming all the motors have the same originally designed stator windings (with the same no. of turns per phase, coil pitch etc.), any increase in no-load current will be due to the turned rotor.

Muthu

http://www.edison.co.in

 

[Tmoose]

Team, thanks for the inputs.

Please don't be impatient with me. None of "our" requested tests have been done because I'm still assembling one (hopefully) comprehensive and thoughtful (with all your help) checklist of non-invasive real time online tests.

First on my list is, and was, Voltage and amperage balance of all 3 phases on at least one "good" and the "bad" motor. If it is determined that the bad mill's motor is being underfed, then I expect that will be addressed first.
Not long ago, On another thread, about another project, more than one responder suggested that my wish to measure voltage and current balance was the sign of a reckless unsafe cowboy.

As EPete alluded, shutting down mills and swappin' motors is NOT trivial. By most any measure big bucks are involved. Big outages are planned a year or more in advance. People have lost their jobs when poor or unlucky decisions have resulted in delayed completion of outages. Not much harm in getting as much info as possible while they remain online. Call it stacking the deck.

Perhaps the test report for the recently overhauled-out-of-desperation motor will include details of measurements of rotor gap and eccentricity. I can only hope. For me, measuring the gap is the only way I'll be sure it is the not the one undersized rotor made by Westinghouse ( or whoever) that year.

The professionals at a power plant have spent several years conscientously working to understand why 1 of 4 mills continues to consistently draw about 10% more amps than the other 3. The list of what they have tried is reportedly long and includes Mechanical inspections, calibration of feeders, and lots else. I do not (yet) have details of which, if any, components they have swapped already. I intend to ask.

Among the checks WE are going to do (sitting a thousand miles away at a computer) is Using coal characteristics, fuel flows, coal fineness, tons of steam produced, and other stuff to predict the power we think the mills >>should<< be using. Maybe 3 mills have problems that are causing them to draw 10% too few amps? (just kidding)

The physicist Richard Feynman in one of his books tells a story that as a young man amazed an older relative. The relative had a broken watch or clock or something. Young Feynman checked the watch for symptoms, then paced back and forth for a while, head down, looking at the floor, just thinking. Then he did something simple to the watch, which immediately started working. Thereafter the relative would often proudly tell the tale that Feynman fixed the watch "just by walking around."

 

[Skogsgurra]

Now, I am really curious! What did he do with the watch? Wind it?

Sometimes, it is as easy as that.

There has been a lot of focus on air gaps in this thread. I have a problem accepting that someone started the commissioning with taking a rotor out of a motor and turning it. The commissioning was only five or six years ago. Surely such an uncommon action must be remebered by someone? Perhaps also noted in the commissioning report?

Checking slip is what I would do first thing. But I would not calculate slip on an assumed 60 (or 50) Hz, but on actual grid frequency when doing the (also accurate) speed measurement. Small deviations in grid frequency mean a lot to calculated slip.

If there is a key or something else that could be measured with a proximity switch and a recorder, then I would do that. And record grid frequency (cycles) simultaneously. Then evaluate as many cycles as I think needed to find out real slip. You can actually see the pulses from the key 'slipping' behind the grid cycles.

If possible, I would also include power, using current clamps, VTs and a power transducer with analogue output. And I would do that measurement on all four motors.

Such measurements are extremely valuable when one sits down to discuss what is going on in the plant.




Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[waross]

How are the rotors constructed?
I had four identical pumps on a cooling skid. One ran hot. The rotors had cast aluminum squirrel cages. The hot motor had an internal void in the squirrel cage casting. A replacement rotor fixed the problem.


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

 

[Skogsgurra]

Such a defect shows up clearly when you do an exact slip measurement. More slip if one or more bars defect. Not sure if it actually increases current drawn from grid. Does it?

I think of it as having a slip ring rotor. More resistance increases slip, but not current drawn.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[desertfox]

hi electrical guru's

I am intrigued as to why 10% more current is being investigated so intensely, is it really that significant?
what would be the normal tolerance on no load current for two identical motors?
I can see that variations in what the motor drives would give rise to additional current, ie friction, inertia, changes in mechanical load etc and these in my opinion can never be identical across four different mills.
All the mechanical components will have been manufactured to a tolerance and therefore each build will inherit different masses, which in turn will result in different friction, inertia etc.
Just curious as to why the additional current is so critical.
Although I do understand if they want to work the mills harder, then they can't overload the motors, but I would of thought there was more than 10% margin in the motors to start with.

desertfox

 

[Tmoose]

why 10% may not be enough - I don't have the details (yet) as to what this customer wants to do, but when changing from that fine bituminous coal ( heating value around 12,000 BTUs) to the cheaper, lower sulfur PRB coal ( heating value 8-9k btu ) a LOT more poundage must be processed to make the same amount of steam ( and ultimately electrical power, if that's the product). Even if the extra volume of coal can be delivered to (rev up the gravimetric feeders), and forced thru (rev up the PA fans) a mill, almost certainly there will be a reduction in fineness, which can bring on various combustion and emission changes/problems. Power plant boiler design is based on meeting the contract requirements that must define steam requirement and "design" coal, but they were and are so danged expensive most, maybe all, customers don't (didn't) want to pay for a whiff of extra capacity. 10 or 30 years ago assumptions about the future were only a little better than guesses we might make today about life law and liberty in 2040.

 

[Tmoose]

Hi rbulsara ,

yesterday I think you said -
"OP has to get past what dpc suggested in the very first response to the original question. Combined with the QUESTION ABOUT THE VOLTAGE and is the motor is fed from the same source?"

I have not asked this customer to check their (4000) volts/amps yet, but you may recall last month, a customer declined to perform those tests on a 460 V motor.

http://eng-tips.com/viewthread.cfm?qid=263195&page=4

The voltage test request will be the first item on my list, but frankly I am anticipating encountering a certain amount of (non-electrical) reluctance

 

[rbulsara]

Tmoose:

Ah, that is interesting.

This tells me that even you do not have the first hand information. Not only that, there is not enough second hand information available to do any meaningful analysis. On top of that the customer appears non-cooperative.

Per your statement, the motor is not even overloaded. The 10% high current is relative to other supposedly identical machines means nothing. A motor draws what it needs to draw. You first need to eliminate (or confirm)the load as the cause. Decoupled motor test, as Edison123 suggested, would be the least you need to do.

I worked in cement and chemical plants, in my past life and have dealt with many real and perceived problems, like this. (I can almost picture the non-technical reason behind this perceived problem, but that is a different subject.)

You really need to get the first hand information, talk to the customer and involve an electrical engineer to tackle this issue, if it really exists.

I will share one of my experiences, in a similar 'troubleshooting' of a 40HP air compressor motor and starter, some other time.

Rafiq Bulsara

http://www.srengineersct.com

 

[Skogsgurra]

Much ado for nothing...

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[waross]

Hi Gunnar;
In the instance that I mentioned the indication of a problem was a hot running motor. The motor shop diagnosed a faulty squirrel cage. A replacement rotor was supplied under warranty and the problem was solved.
If all the rotor bars have a higher resistance, we can expect a higher slip. If one or two rotor bars have a higher resistance the other bars are working harder (ie: drawing more current).
Consider a low slip motor coupled with a higher slip motor. The lower slip motor will "Hog" the load and run hotter. Similarly, one or more higher resistance squirrel cage bars will be taking less than their share of the load, and the I[sup]2[/sup]R in the other bars will result in more heat in the good bars.


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

 

[rasevskii]

If there is a safety issue in measuring the current and voltage using, say, clamp on meters, which appears to be the case, how do they know what the currents are in the first place? There must be in fact panel meters that can be read somewhere without opening any panels or doors. The same for the busbar voltages. It looks like that the four motors are not in fact on the same busbar...

rasevskii

 

[electricpete]

I have not asked this customer to check their (4000) volts/amps yet, but you may recall last month, a customer declined to perform those tests on a 460 V motor

In general it is easier to gather data on 4kv systems than 460vac systems due to the presence of PTs and CT's on the 4kv system.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[rasevskii]

Tmoose-

It seems that these are 4160V motors, having reread the posts. Obviously then out of the question to do tests with clamp on meters. What sort of metering is actually provided then?

You are not telling us the size of these motors, or what type. They could be synchronous for all we know. Or wound rotor. Therefore other considerations.

Much more info needed by all here, as Rafiq has said.

rasevskii

 

[Skogsgurra]

I will have to digest that, Bill.

I have seen two motors with rotor bar failures in my (working) life. That makes the frequency something like one each 25th year. So, it is not a very common thing.

OTOH, I have seen many motors with a supposed rotor bar failure. But it has almost always been something else.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[electricpete]

Just to join the side discussion between Gunnar and Bill - attached is an article "Broken Bar Detector for Squirrel Cage Induction Motor" by Kliman of GE.

Page 3 of 11 shows figure 3: [Calculated] "Effect of broken bars on steady state performance". The horizonttal axis is broken bars per pole. The vertical axis is full load slip and % change in full load current (two curves). You can see that neither slip nor current is very useful as a predictor until the number of broken bars is huge. Figure 7 on top left of the next page shows that the change in torque characteristic from a broken bar is more evident during start/acceleration than at full load.

Tangent to a tangent:
1 - Apparently, the current actually decreases for small number of broken bars (!?)... I would have liked to see an explanation for that.
2 - By the way, the document is a GE publication "GER".... ot was at one time available for free from GE - no copyright listed.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[rbulsara]

At the least we need to get the nominal voltage right. Some posts here and in the previous related thread mentions, 460V, 4000V and 4160V, sometimes in the same sentence!!!

Rafiq Bulsara

http://www.srengineersct.com

 

[desertfox]

I found this motor specification which details type tests etc for electrical motors:-

http://www.scribd.com/doc/24157736/Motor-Standards

Now another question if the motors had type test certs would that show that it drew more current when compared with the other motor type tests.
Reading the posts now I take it 10% difference is not significant under normal circumstances.

desertfox

 

[electricpete]

It would be a function of measurement and load type among other things. From memory we typically have a 5% variability in panel meter readings among groups of 6-8 pumps carrying similar loads at our plant (I presume mills have more variability than pumps). Frankly I don't think I'd even investigate 10% (we have a lot of other monitoring to rely on... vibration, flow, dp). But if it were threatening to limit production, it can merit further investigation to assign causes to the variability.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.