Heavy starting induction motor

[polvalrot]

It was the 3rd failure in 900kw 6kV 4 pole induction motor, turning a fan in a steel plant. Originaly the closed damper starting dictates that the load starting characteristic is all below the 0.3 of nominal torque at nominal speed. Neverthless, the measured current at closed damper condition and nominal speed was at 84Amps. THe starting acc to manufacturers data sould be 24 secs at 0,85Un, while for all 3 failed motors was 48 secs. The damper is claimed to be realy closed. Motor inertia is 33 and load inertia 413kgm2. Acc to my calculations, with 84Amps at closed damper (considered as no-load(!)) is really 47secs. What's wrong?

 

[polvalrot]

Dear Petronila
if you check in the early potings, you will find the results of your formulas. For me there is only one question. We are all engineers in plants or engineering firms. Do we or do not rely on manufacturer's data? what else do we have? If i was not lucky enough to take the starting curves before failure, how could i stand against the "inappropriate" motor case? Almost all 6kV 4pole motors 1000kw , should be able to rotate at no load (closed damper)the fan. The fan at low speed is just like an inertial (whorl?) load.
Thank you all for your advice
chronis (

http://www.rotechen.gr)

 

[LionelHutz]

I did some quick simulations.

I assumed 1800rpm motor.
I assumed 85% voltage.


I first assumed the fan provided 90% load at full speed.

If I assume a typical poor starting torque motor with ~75% torque from 0 - 60% speed to a breakdown peak of 180% then I get 49 sec. as the start time.

If I assume a better motor with ~110% torque from 0 to ~60% speed with a breakdown peak reaching 220% then I get 28 sec. as the start time.


I then assumed the fan provided 40% load at full speed.

With the same motor as the first example I get 36 sec. as the start time.

With the same motor as the second example I get 24 sec. as the start time.

So, it seems to me that your numbers are not correct. The fan starting time should not double when the fan changes from closed damper to open damper.

As another note, those formula's posted will give an approximation of the accel time but I have seen that answer off by as much as 10 sec. compared to an iterative calculation that goes in small speed steps using the motor and load torque at each speed step. Those formula's don't even account for the required load torque.

 

[polvalrot]

Dear Lionel

motor and fan manufacturers claim 25secs starting with closed damper.
I measured : 84amps with "closed damper" at 1500rpm (50hz)and 46secs starting time.
What i say is that what they call "closed damper" is not "no load" condition.
The motor is 22Amps at no load (uncoupled) condition.
Real numbers: Nominal torqe 5770Nm, Rel.starting torque=0,6Tn,
84Amps at 1500rpm are 4600Nm. THerefore the motor starts with 80+% of nominal load.Please check inertia figures in early postings. My calculations just indicate to me that the "closed damper" condition is the 80+% of full load, under start up. Therefore, the net accelerating torque is very small and the consequent starting time rises.

 

[LionelHutz]

I mean the fan inertia you provided is the dominating factor in the acceleration and changing the fan curve from 40% load at full-speed to 90% load at full-speed does not seem like it should double the acceleration time with everything else being equal.

How are you converting amps into kW without the power factor?

What is the actual voltage drop on the system during the start? A larger voltage drop will cause a longer acceleration time but this typically has little effect on the heating of the motor.

 

[polvalrot]

Dear Lionel

i disagree. The cange of inertia changes a little the schenery. (it is just a multiplier). On contrary, the load curve that is 80 and not 40% at nominal speed, is the dominant parameter, since the startind time is proportional to the sum(d(speed)/d(Net torque)).
steady state dumpe closed: 84amps with pf 0,88 (measured), 6kV
starting voltage drop (measured)=1000V = 15% of 6,1kV nominal measured.

The point is that we find out that the load is not the intended for these motors.Do we agree on this?

 

[LionelHutz]

Well, by destroying 3 motors it seems you have already proved that the motor is not suitable for starting the present load.

We will have to be content disagreeing about which factor is the dominant one with respect to the starting time.

 

[polvalrot]

We are not the manufacturer, we are not the owners, nor the system designers. The manufacturers of the motor and the fan are well known. We only participate to the point we can advice regarding the nature of the failures.I strongly believe that we have the case of a large motor in respect to the load.
As regarding to the starting time is fairly easy using an excel sheet to decide whether the damper or the inertia have more effect to the motor starting severity.
Now, isn't it correct that 84amps at 6kV with a pf 0,88 at 1500rpm is a 4,7kNm of torque? That is reality.
Isn't there a diference between the claimed nominal figure of 0,5Tn=2800Nm for the closed damper?

 

[polvalrot]

We are not the manufacturer, we are not the owners, nor the system designers. The manufacturers of the motor and the fan are well known. We only participate to the point we can advice regarding the nature of the failures.I strongly believe that we have the case of a large motor in respect to the load.
As regarding to the starting time is fairly easy using an excel sheet to decide whether the damper or the inertia have more effect to the motor starting severity.
Now, isn't it correct that 84amps at 6kV with a pf 0,88 at 1500rpm is a 4,7kNm of torque? That is reality.
Isn't there a diference between the claimed nominal figure of 0,5Tn=2800Nm for the closed damper?

 

[polvalrot]

sorry: small motor for the application (not large)

 

[LionelHutz]

Sorry, by destroying 3 motors it seems your customer has already proved that the motor is not suitable for the applied load.

I used some custom software written to calculate the acceleration in 0.01 second time steps. With the limited data you have given, my opinion is that it is very unlikely that the accel time will change from 48 sec to 24 sec by unloading the fan further. I've literally seen 100's of motor curves so by creating some "typical" motor curves based on the data you provided as well as using a fan load and the inertia you gave I just can't make it happen. The closest I could come is a time change of 1.5 times.


So, to answer you initial question - What's wrong?

#1 - There is an issue with the closed damper fan load.

#2 - I'm suspect of the 24 sec vs 48 sec accel times.

 

[petronila]

Hi Polvalrot,

Are this motors new , surplace or remanufactured ??
Taking account the manufacturer data ¿where the 25 seconds is coming from? From the manufacturers calculation only for your application, or is the maximum allowable for their motor during a DOL start?
the difference could be there , the manufacturer could be citing the maximum for the motor, rather than what the load does.

Regards
Petronila

 

[waross]

I think that it is time to re-evaluate all parameters, don't assume anything.
Most likely the dampers are not closed. However, the wk^2 of the fan may be inaccurate or it may have been calculated for the wrong RPM. How is the voltage drop, and the voltage balance? Somewhere in the data is a parameter that is being accepted as valid that is not valid.
However, given the excess running current with the dampers supposedly closed, and the high ratio of "Damper closed" current to "Damper open" current, check the dampers, the air temp and, by the way, the voltage drop and balance. You may also want to check for a phase shift on the supply power. That will lead to high motor current.
If this is a retrofit of an existing fan, the customer may have replaced a double squirrel cage, or other high slip motor with a low slip motor in the pursuit of energy savings.
Still, the 84 amps points to either a partly open damper, voltage or phase shift problems or possibly an air temperature problem.
respectfully

 

[rmw]

I have one other possibility for consideration. What is the integrity of the ductwork? I immediately jumped to the dampers and made posts early in this thread, but the symptoms would lend themselves to any source of cold air inleakage. So, check the duct work, I've seen some so mighty awful looking holes in duct work. Check the expansion joints. There are probably some either side of the fan. If metal, the convolutions crack, rot (rust-corrode), and if fabric, they could be ripped or rotten.

Point being, the dampers are getting a lot of attention and to me they are highly suspect, but I believe the topic needs to be broadened to any significant (or lots of insigificant) source of air leakage in the suction side of the fan.

rmw

 

[polvalrot]

maybe this will help our condsiderations
A slip ring motor was employed, 950kw, 1500rpm, with an oil cooled starter, is operating excellent for more than 3 weeks now, with a very smooth starting 30 secs(acc to the owner). As i see it, seems to me like a starting with " not closed" damper
regards, chronis

 

[edison123]

chronis

What is the no-load current of this slip ring motor under "damper closed" condition ?

 

[polvalrot]

Owner claims 75 Amps, "closed damper"

 

[Skogsgurra]

Thanks for the feed-back, polvalrot.

Good ol' WRIM to the rescue again! There's a reason for their existence. Low starting torque in squirrel cage motors, some voltage drop in supply and cables, limited rotor heat capacity and a high load inertia are some of the reasons why they are needed.

A soft starter could have helped, but it produces just as much rotor heat. Reduced current drain and a "well behaved" source and cable voltage drop may have helped. But a WRIM with external starter always does it. Mainly because you can have a high torque during all of the acceleration and not a very weak starting torque like most squirrel cage motors have.

Your 30 seconds acceleration time with external starting resistors vs calculated 24 seconds at .85xUn makes me wonder if your inertia value is correct and if that wasn't the "root problem"? Using rated torque and 413 kgm2 results in 11.3 seconds. Adding motor inertia and some windage makes 15 seconds a probable acceleration time with WRIM and starting at rated torque. So. I wonder if there isn't a problem with the old GD2 and inertia getting mixed up?




Gunnar Englund

http://www.gke.org

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