surge due to "bump" test for motor rotation? 13

[electricpete]

During motor retest following motor removal/replacement, we confirm the direction of rotation by momentarily energizing the motor just long enough to get it rotating.

A fairly knowledgeable engineer recently told me that this can cause excessive surge stresses to the motor. Specifically, the full starting current is still flowing at the time the breaker is re-opened. He said that the interruption of this higher-than-normal current caused excessive voltage surges which could over time damage the insualtion. v=L*di/dt and the i in this case is ~ 6xFLA.

By the way we have mechanical circuit breakers.

I don't know enough about surges and circuit interruption to evaluate this concern. I am skeptical for the following reasons: I was under the impression that this bump test was fairly common practice. Therefore if it were creating severe surges, I would think I would have heard more about it.

What do you guys think?

 

[electricpete]

By the way, I realize it is a fairly simple resolution to just allow the motor enough time to come up to speed... but I still want to know whether our existing practice is really creating a problem.

 

[jOmega]

Pete,

How often is this practice applied to any given motor ...

once an hour ?

once a day ?

once a week ?

once a month ?

once a year ?

etc....etc....etc ..


Consider that the junk yards would be full of failed motors if it was really a problem to infrequently 'bump' a motor. And motor manufacturers, NEMA, EASA etc... would publish a warning about such practice.

Kr... jO

 

[nbucska]

Since you need to do it only once
and since it may happen anytime by accident
-- a motor have to be able to survive it
So yes, you get large voltage spike but I don't think the risk is significant.



<nbucska@pcperipherals.com>

 

[electricpete]

It may be a few times during the life of the motor winding. Definitely after initial installation. Definitely after refurbishment of motor which we sometimes do at 10-20 years. Possibly when motor is removed to access pump. I think at various other times when motor is de-term'd re-term'd in place for test or troubleshooting, we just label the leads carefully to ensure they are reconnected the same way.

 

[rbulsara]

While it is true that any switching on inductive load could creat a voltage 'spike' or surge, things has to be put in perspective.

You should not be worried, imho.
Because:

1. Presumably you have had no problems so far so why try to fix a problem which does not exist.

2. Bumping a motor on no load (even starting), the inrush current spike does not last much beyond first half cycle or even say 2-3 cycels. Even fastest manual breaker closing and opening takes more than 1 second (60 cycles) so you are not interrupting any high currents while opening a unloaded motor after a 'bump'.

3. Since you are not repeating this thing several times on a motor in a short period of time, there definitely is no threat to the motor.

4. If any, the concern shall be with the circut breaker contacts. As long as you use a adequately sized breaker you should not have any promblem. Overtime, I am sure, you can surely afford to replace a breaker or two if any problem occurs.

 

[electricpete]

thanks rb.

You mention that inrush current only lasts a matter of cycles. I agree from the exponentially-decaying dc componentn of the current... but I think the high ac starting current ~ 5*LRC persists until the motor is near full speed, somewhere in the range 0.5 - 2 seconds.

 

[afterhrs]

Electricpete,
I deal with 75 hp motors on a daily basis that we &quot;jog&quot; or &quot;bump&quot; on a several times a day. They are always jogged under load.
If you are going to have a machine with this function I strongly recommend you overate the contactors. A good rule of thumb I have seen is about 200% for NEMA contactors and 300% for IEC contactors.

Regards,
Afterhrs

 

[busbar]

There are those goofy little $100 phase-sequence testers that operate over 20mV-600V, and can be used on an idle motor to indicate motor phasing with a decoupled shaft spin. The line-side {energized} connections should be limited to fused PT-secondary connections for anything over low-voltage/low-current applications. [

http://www.rke.com/mrt.pdf

]

 

[swgrmfg]

electricpete:

The engineer who answered you is absolutely correct. The fact is that, during start, the motor has not yet built any back EMF - which would otherwise limit the voltage surge during disconnection from the power line. Instead, you will have a huge Ldi/dt voltage spike across the motor terminals.

GH

 

[dpc]

That's an interesting concern that I've never heard voiced. My experience is that bumping motors to verify rotation is standard practice every place I've ever been.

 

[edison123]

pete,

isn't twice the rated voltage that is the maximum that could occur in a normal ac motor during switching ? if so, bump test couldn't be any worse than normal switching operations ? what do you think ?

kumar

 

[ScottyUK]

Pete,

My experience of this is that it kills contactors if carried out too frequently. If it is a frequent occurence, rather than a DoR test after maintenance, consider uprating the contactor to a larger type or one rated for a more severe duty.

 

[jbartos]

Suggestion: Normally, motors are designed to survive a start in the wrong direction with a possible motor protector, i.e. circuit breaker, overload relay or fuse, trip, if the motor is excessively loaded in the wrong direction. There is usually a bigger concern about the motor load being started in incorrect direction, whether or not it will not be damaged.

 

[electricpete]

Well, it seems there is some difference of opinion within responses here and I really don't know the answer.

For the future to be on the safe side I will just ask our operators to wait 5 seconds between starting and stopping during the bump test. I really don't know if it's required but certainly is easy to do. I am still curious (and skeptical) though.

edison - The doubling that you refer to occurs during energization. During deenergization it's a little harder for me to picture what's going on.

I guess I can vaguely envision a scenario that if one contact opens before the others, it creates a sudden increase a v=Ldi/dt across that open contact. If the increase in V occurs suddenly it creates a traveling wave?

I do know that some people consider vacuum breakers create surges which can be damaging to motors. Is it during the closing or interruption? If it is during the interruption, then I can see maybe it is because they interrupt very suddenly, increasing the di/dt. So by same logic, increasing the current would also increase the di/dt and create similar effect.

There can be even higher at the time of current interruption during a phase-to-phase fault. (or phase-to-ground fault for solidly grounded power systems I am not familiar with). In that case, I would expect that if current trips on phase-to-phase fault we might sometimes expect to find shorted line-end-turns during subsequent surge testing. I have not heard of this... has anyone?

As a coincidence, I was just involved in reviewing investigation of a motor which tripped on ground fault in a resistance-grounded power system. In addition to the fault to ground 4 coils in, two of the line end coils failed surge testing at the normal maintenance-level test voltage. But they were not physically near the failed coil... no reason to connect them to the fault. The motor had not been surge tested since manufacture. As I have alluded, the ground fault current was limited by resistance grounding system and there was no phase-to-phase fault possibility based on examination of damage and relays.... so the hi Ldi/dt scenario doesn't seem to apply.

 

[edison123]

pete,

It has been emprically proven all switching surges (in or out) cannot exceed 2 x rated voltage especially with VCB's. This is the reason new or rewound machines are ac hipotted and surge tested at 2 rated voltage + 1 KV. If you talking about one pole of breaker not opening or closing, then it is single phasing with its disastrous consequences.

 

[meenakhi]

-bump test is done for very samll duratuion in comparision to the life of motor.
-the sil/bil is supposed to take care of the surge if any during bump test.
-what if an unintentional bump test occour i.e motor closed /tripped during normal operation.
- a motor has to be designed taking this fact in to account.
-it is correct that there will be surges during bump test , but what about the magnitude and duration.

 

[jOmega]

Edison123:

Consider that the insulation test by means of the highpot method... is typically done with a sine wave source. The NEMA standard for insulation, as you stated, is (2 x rated) + 1000v.

However, in a bump test, you have maximum (6x)....or near maximum..... current flowing into the inductance of the stator... at the just before you interrupt this current, the field around the conductors is very, very strong, and, there is a lot of force or torque stress on the conductors (not to be confused with shaft torque)... So the wires are straining to move... and sometimes are able to cause insulation failure when they encounter a metal edge and there is movement of the conductor...

So maybe it doesn't happen the first time... but every time the motor is DOL started, mechanicl stress occurs in the conductors (winding, etc..) of the stator.

At some point in time... failure will occur... and along the way, insulation withstand is reduced.

Now, to the point... when the energy in the inductance of the stator windings, is suddenly interrupted, the voltage spike consequent to the L di/dt.... is not a gentle sinewave... but rather a steep wavefronted voltage, which can cause the insulation in the motor to fail where under a highpot test...it would not...

An oscilloscope used to capture the voltage transient thusly generated would substantiate that the characteristic is not that of a sinewave.

The insulation test for such occurrences, is not that of a highpot or of a meggar type test.... rather, the windings should be subjected to a pulse injection (surge)type test, whereby the rise time and amplitude of the pulse can be selected.

Typical would be a Baker Instruments Surge & HiPot tester. These Surge and HiPot testers verify insulation integrity by finding weaknesses in Turn-to-Turn, Phase-to-Phase, and Coil-to-Coil insulation systems within the motor.

Many motor manufacturers apply such a test to the stator before the rotor is installed as a quality check. I've seen it done on random wound and form wound stators.

The surge test is also a good means for determining the motor's ability to withstand the stress of dv/dt that is seen when the motor is sourced from a VFD and is suggested for old motors newly being applied to a VFD.

A friend of mine who sits on the NEMA MG-1 committee... tells me that the hipot is not meant to protect against high dv/dt stress, but rather.... to accommodate line transients..... not unlike those caused by the switching of capacitors on the mains.

 

[electricpete]

meenakhi brings up a good point. We have one motor that has on three different occasions tripped upon starting for unknown reasons over a 2 year period. Now has not happened in 8 months. It would be the same transient.

edison - I can easily see where the voltage doubling applies during closing. It is simple reflection of voltage wave (nominal voltage) upon seeing a change of impedance. I don't see where it applies on opening. The scenario that I tried to paint only assumed a slight difference in timing of contacts opening...just because it is easier to imagine where the voltage appears (across the single open contact). It does seem credible that whatever transient exists during opening is proportional to current if we assume simple linear circuit based on dv/dt. I think the nature of transients upon opening are much more complex than I understand. I know there is a lot of literature on circuit interruption but I havent read much of it. Terms like restrike, reverse recovery voltage etc... I have heard them but don't really understand them. Once again... is it not true that vacuum breakers impose a worse voltage transient than than mechanical breakers during interruption? If so that suggests the di/dt term is important and will depend upon magnitude of I. (Just arguing the devil's advocate position here)

jO - Your point out that in the bump scenario there is surge stress at the same time as mechanical stress. Interesting point.

 

[edison123]

pete,

I agree that the inrush current of the cage motor during starting (or bump testing) may be 5 to 6 times, but it does not automatically follow that the surge volatges across windings will also reach to 5 to 6 times the rated voltage (regardless of whether you make or break the circuit). If it were so, I am sure most of the motors will fail during starting since they are not designed to withstand 5 to 6 times rated voltage. If you have any theory about surge voltages exceeding twice rated voltages, I'll be interested to know.