5000HP Compression Motor Starting Methods. Design Phase 2

[gokulkrish2]

Hello All,

We are in the process of designing an air compressor which uses a 5000HP Induction motor. The data about the motor is attached.

This motor is proposed to be fed from a 15MVA - 69KV/4160V transformer which is inturn supplied from a 125MVA - 230KV/69KV transformer from the UTILITY end. I am sure the 125MVA system is quiet strong for this load but i am concerned if the 15MVA transformer would be able to start it.

The electrical head from the contractor is claiming that the motor is started in the unloaded condition and that this motor can start Direct online without any problems.

Is that possible. Would this 15MVA transformer be able to handle the initial inrush of the 5000HP motor? Isnt that sized small? The contractor dont want to spend the money on a VFD and the motor which is already procurred is not insulated enough to be able to run on a soft starter.
What would a solution here.

Please comment

Thanks

gokul

 

[electricpete]

Jeff - I'm sure you know much more about soft starting than me. (We have only a handful at our plant and only because they were supplied by a vendor... not needed at a power plant). But the discussion does not sound quite balanced to me. Please consider my comments below.
Respectfully

I know of no reason why any particular motor design is not suitable for a soft start. The time / damage curve issue is often misunderstood. The energy it takes to start a motor from a dead stop is always the same regardless of the method. Soft starting puts no more stress on the motor that DOL. Where people get confused is ....

In the context of this thread (unloaded start) it is true. However this logic would be incorrect if applied to loaded start. The heat energy dissipated in the motor while going from rest to full speed increases as voltage decreases if load torque is applied during the start. I can post the math proof (which is also a claculation recipe) if anyone is interested.

But think it through a little more; if the reduced starting torque that results from using a soft starter is insufficient to accelerate the load, then the starter application was not valid in the first place!

It sounds like circular logic to me. How do you know what is insufficient (*) until you or the OEM has done the calcs. NEMA MG-1 says starting the motor must safely start the specified load at 90% terminal voltage. Most people specify 80% terminal voltage (we do). If it's not specified, you shouldn't bet on it imo. (* checking for safe start involves not just checking load torque remains below reduced motor torque, but also plotting simluated starting time/current curve against motor thermal limit curve and verifying enough room to fit protection curve in between).

Maybe it is ok for most applications and I gather it is usually that way in your experience. The troublesome ones would by large motors with heavy load during start. At our plant we have 8000 hp 1200rpm motors driving single stage radial flow pumps located in a fluid loop that has no valves... must be started under equivalent of "valve wide open" conditions. It also has flywheel and takes 12 seconds to start at 100% voltage and 21 seconds to start at 80% voltage. The total heating during that 20-second 80% voltage start is 140% of the total heating added during the 12 second 100% voltage start. We also have 3500hp 324rpm motors driving single stage axial flow pumps that must be started against closed valve for water hammer prevention... Closed valve means maximum starting load torque for these pumps. The time current curves on log/log scale come much closer to the protection for 80% start for these curves. Of course "close" is a little ambigous on a time current curve especially log scale.

Another question: why would NEMA even bother to specify a minimum voltage for starting if it is somehow irrelevant to motor starting performance ?

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

 

[burnt2x]

I think all these contradicting comments all boils down to the differences in motor application and types of motors! jraef maybe talking about a different scenario, and maybe epete's case is another application/ scenario.
Heat developed in the motor does not always result to high temp (which is a problem). The ability of a certain motor to cool-off is one factor. Motor speed affects the temp rise as most motors have vent fans driven by motor shaft in the same way as the method of starting employed. The point I see is that the posters ahead of me explained their cases very well, and these are very helpful in finding solutions to design problems. Kudos to all.

 

[electricpete]

Thanks Keith. I agree 100% it depends on application. That was in fact my point. I interpreted (misinterpretted?) some previous comments to be offered as general rules... whereas I thought clarification of the assumption/application (no-load start) would be helpful.

At the risk of beating a dead horse, I think Richard Nailen has expressed similar viewpoint in Electrical Apparatus magazine:

http://findarticles.com/p/articles/mi_qa3726/is_200603/ai_n17176681/

One might suppose that a reduced-voltage or "soft" start would ease the thermal stress on an accelerating motor. After all, heating within the motor can only result from current flow. That's the only energy source during acceleration. Lower current should mean less heat-doesn't that follow from the I^2*R relationship?

That simplistic view fails to take into account another relationship that is more important here: the rise in internal motor temperature is caused by total heat input, which involves time, and is proportional to I^2*t, not I^2*R. When motor voltage is reduced during starting, the associated torque reduction causes acceleration time to increase. The consequent increase in heat developed within the rotor and stator may far outweigh any reduction resulting from reduced accelerating current.

Some argue that reduced voltage spreads out heat generation over a longer time, thereby allowing more of the heat to dissipate without raising motor temperature. That's true-but heat transfer to the surroundings is seldom rapid enough to offset the increased heat production.

I wouldn’t go as far as he does in his last paragraph (“seldom”). I suspect as Jeff suggests most motor applications benefit thermally from soft start vs DOL start. And there is no question for some high inertia, low-torque load such as centrifuge that it is a huge benefit. My spin is just that there are competing effects when load is present (increased total heating disadvantage vs increased dissipation time advantage), so softstart is not universally beneficial in terms of motor thermal performance. Ideally we’d like to apply motors based on specifications, calculations, or thumbrules that demonstrate suitability. As far as I can tell soft start application is beyond specification requirements for most motors. If there is some thumbrule that allows easy evaluation of whether a soft start will adversely affect the thermal performance for loaded start in a given application, I’d be interested to hear it.

Sorry if I am taking the thread a little off-track.


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

 

[gokulkrish2]

edison,

Thanks for the update. I have seen similar kind of inline reactors used with softstarters and bypass switches. But those are for capacitance issues of long cables.

Also,

Is someone here from a motor manufacturing company. Do you guys think it would be worthwhile to call those guys and ask for an explanation of why a soft starter cannot be used for the motor we purchased? would i be welcomed with a question like that?

I have had good and bad times with US motors in different situations. It all depends on who u get when u dial that 800 number

gokul

 

[wolf39]

Gokul:

Whether you are welcomed or not: You as a user/customer are entitled to ask questions regarding the application of products you've purchased. If you are not welcomed, you have the option to change the OEM for your next procurement.

Pete and jraef made valuable contributions. All main aspects were covered and on your behalf I give a star to both gentlemen.

Wolf

http://www.hydropower-consult.com

 

[electricpete]

Thanks Wolf I appreciate it, although I am aware I am not contributing to original poster’s question. It was just an interesting sidetrack that we have discussed before that I wanted to follow up.

As a matter of clarification of my earlier comments: my pumps examples were not good examples since they are centrifugal-like pumps. Even though valve position was not optimal, they still have roughly Torque^speed^2 torque-speed curves. Most of heat is added at lower end of the speed range so centrifugal pump curves are relatively forgiving in this respect since load torque does not slow them much accelerating through lower speed range.

In contrast certain compressor torque speed curves can be more challenging. Example linked at bottom of page 2/47 and top of page 2/48 below shows evaluation of a compressor which turns out not suitable for starting even at 80% voltage (even though motor torque remains above load torque throughout)

http://books.google.com/books?id=V1...suitable for only one cold or one hot&f=false

I’m done now ;-)

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

 

[LionelHutz]

I would say with a full voltage start and a transformer impedance of 5% you could see around a 10% voltage drop. This does depends on the transformer impedance. I would hope for <5% as opposed to an impedance around 7% to 10% I sometimes see.

I would question them about reduced voltage starting if reduced voltage starting is a route you want to persue. An unloaded compressor should be an easy soft-starter application. The only time it could be an issue is if the motor was specifically designed for full-voltage starting with your source and load, but I doubt you have this case. I have seen cases where the motor had atypical torque and current curves which matched very well to it's application.

In theory, the I^2*t heating of the motor is always increased when using a soft-starter or reduced voltage starter. The increase is minimal if the torque is not reduced so much that the motor comes close to stalling. But the heating will become excessive if you lower the voltage so much the motor and load torques almost meet. If the motor actually suffering or not from an increased heating over a longer period of time is the debateable part.

 

[edison123]

Gokul - The inductive reactors in the case I cited are specifically for starting purpose and not for some cable capacitance issues. The system has been working very well since it was installed about 12 years back.

Muthu

http://www.edison.co.in