Number of consecutive starts on MV motor

[rockman7892]

I have been doing some investigating as of lately on the number of starts that are permissible on an MV motor. From what I see this is dictated by the NEMA MG-1 standard which says the motor should be capable of 2 cold starts and 1 hot start. I understand however that this is usually a minimum requirment and is one that is not closely regulated.

For a given motor datasheet which has a listing of 2 cold starts and 1 hot start is this a minimum that the manufacturer is stating the motor is capable of meeting or is this usually a maximum? Or this somewhat of a gray area?

I also understand that the thermal capacity used during a start depends on the inertia of the load or how much of the load is present. So lets say that the motor datasheets lists the 2cold 1hot rating for a given inertia. So esentially when starting this inertia you would expect to use about 50% thermal capacity on each start thus allowing only 2 consecutive starts from cold before 100% thermal capacity was used.

Lets say however that the load being started in much less and has a much smaller inertia than is listed on the motor datasheet (maybe 1/4 of that listed.) With this case when starting the motor we are only using about 10% of the motors thermal capacity. If only using this 10% at a reduced load, would this allow for more than 2 cold starts on a motor if required. For instance if we for some reason needed 4 starts on the motor would this be allowable since with each start at 10% we would only use 40% of the motors thermal capacity?

If this is the case then I would think the start inhibit function of a protecive relay would be the best option for determining how much thermal capacity it takes to start the motor compared to how much thermal capacity is avaliable in order to determine how may starts are allowed as opposed to limiting it to some hard number.

I wanted to hear what others thoughts and experiences are.

Thanks

 

[electricpete]

You are considering only the load inertia in your ration 1/4... you need the ratio of the total inertia (motor plus load). For a typical centrifugal pump load, the motor inertia is much larger and there is not much difference in heat used up if you change the pump inertia by factor of 4.

Also of course load torque plays some role in the picture and I think you are safe with your approach (modified to consider total inertia) as long as your load torque-speed curve does not exceed torque~T100*(speed/speed100)^2 where T100 and speed100 are rated values at full speed.

=====================================
(2B)+(2B)' ?

 

[dpc]

You need to obtain this data from the motor supplier for any MV motor. There is no standard. The 2 cold, 1 hot is a reasonable assumption, but far from a standard that you can rely upon.

A good motor relay can calculated thermal capacity used during a start (an approximation at least). This is a good approach if you have such a relay and it is properly configured.


David Castor

http://www.cvoes.com

 

[electricpete]

Also I'm sure you know voltage deviation from nameplate plays some role in the thought process. The NEMA requirement one start hot and two cold actually requires motor nameplate voltage above 90% which may not be realistic. Your specification or data sheet may have other more lenietn requirements. We normally specify our motors to meet the same requirements (one start hot and two starts cold for the specified load) with terminal voltage at 80% during the start.

=====================================
(2B)+(2B)' ?

 

[electricpete]

If you want to do more detailed evaluation there was some good discussion by Lionel and others here along with some examples.

http://www.eng-tips.com/viewthread.cfm?qid=277213&page=1

A brief summary:
With torque vs speed curves for motor and load, you can generate curve of speed vs time.
If you have current vs speed curve you can convert that to current vs time.
With current vs time, you can numerically compute total I^2*t
As a conservative approach, compare your I^2*T to LRC^2*T where T is 100% voltage locked rotor thermal time limit. If you are splitting among two starts, compare the sum of the I^2*t.

Now I was just thinking... what if you had locked rotor on 2nd of two consecutive starts (or more realistically was locked both times...didn't realize it the first time and just hit the start button again).... would the protective relaying take out the motor in time to prevent thermal damage. Off the top of my head I don't think that would be guaranteed using standard approach to protective relaying using electromechanical induction disk relays which really only looks at one start at a time (not smart digital relays). I'll think some more, but I'm interested if others have comments on that aspect.

=====================================
(2B)+(2B)' ?

 

[electricpete]

I understand however that this is usually a minimum requirment and is one that is not closely regulated.

For a given motor datasheet which has a listing of 2 cold starts and 1 hot start is this a minimum that the manufacturer is stating the motor is capable of meeting or is this usually a maximum? Or this somewhat of a gray area?

The only reasonable approach imo is to limit yourself to the less restrictive of what is stated in NEMA or the manufacturer data sheet. If you don't have manufacturer's data, then NEMA is all you have (unless you feel lucky).

=====================================
(2B)+(2B)' ?

 

[rockman7892]

dpc

I am using a motor relay to calculate the starting thermal capacity. All data has been entered based on the motor datasheet and thermal damage curves of the motor.

When a particular 2500hp motor of ours starts it gets up to full speed in a little over a second and the relay shows a thermal capacity of only 19% being used during this starting. Does this mean if necessary we could start this motor 5 times consecutivly and still stay within the thermal limits of the motor based on the TC used during starting that we are seeing? (This question is a theoretical question, for we have no current plans of starting motor 5x consecutively.)

 

[dpc]

Theoretically yes. But the relay calculation is a fairly crude approximation, so I would not want to bet a 2500 hp motor on it. If you have RTD biasing in the motor model, that improves the accuracy a little, since it knows where the starting point of temperature is. Relays that run both a stator and a rotor model are probably a little more accurate, also, at least in my opinion.



David Castor

http://www.cvoes.com

 

[rockman7892]

The motor relay does have stator RTD's that are used for biasing in the motor model.

Are these RTD's quick enough to respond to temperature changes during the starting of a motor to where they would be able to accuratley bias the thermal capacity used during a start?

 

[electricpete]

For a large motor, the rotor heats faster than the stator... I don't think there is a lot of credit that can be taken from looking at the stator RTD's.

=====================================
(2B)+(2B)' ?

 

[electricpete]

I also agree with previous comments - in theory it should all come down to I^2*t and if you obey the curves (as interpreted by the relay), you should be ok.

Somehow if the motor data sheet says 1 start hot and 2 starts cold on a 2500hp motor, I personally would be reluctant to allow more than that (for example during balance) even if the curve suggested it was ok. A few reasons:
1 – Selfish reasons. No-one would question me for enforcing a limit from an OEM data sheet. On the other hand if I go beyond that, the risk lies on my shoulders. More margin makes me more comfortable.
2 – These limits are not exact. We have had a rotor bar failure on (by coincidence) a 2500 hp 4-pole motor. Now we have another one that seems to be going through the same thing. They don’t have frequency starts in their routine operation. Exact cause of the failure is unknown… but for all I know we could have had some unusual repeat starting sequences.

That’s just one way of looking at it fwiw.

=====================================
(2B)+(2B)' ?

 

[dpc]

The RTD biasing is only useful in establishing the starting temperature. The RTDs are generally too slow to respond the rapid stator heating during starting and are of no use in sensing rotor heating during starting.

One of the difficult values to estimate in a motor model is the rate of cooling after a start. The RTDs help with that.

I agree with electricpete - the rotor is usually the limiting factor during starting.

David Castor

http://www.cvoes.com