Wye-delta start û peak current and instantaneous setting 6

[electricpete]

We have one MCC-fed compressor motor application where motor uses an open-transition wye-delta start. It is the only application in the plant with wye-delta start. Don’t ask me why we have it... I know of no good reason.

We measured 40 millisecond dead time between open of the wye and close of the delta. And we saw a current spike that occurs upon transition from wye to delta as result of out-of-phase condition of motor residual voltage and incoming voltage as expected. Even if motor speed didn’t change, there is phase difference between delta and wye connections.

We have experienced intermittent tripping of the molded case circuit breaker. I don’t know details of the settings. I was asked to provide general input/reference about the setting of molded case circuit breaker instantaneous for this type of application.

My questions (in order of importance):
1. What guidance is available for selecting molded case circuit breaker instantaneous setpoint (other than obvious coordination requirements with upstream devices and protection of cables).

2. NEC gives suggested mcb instantaneous settings as multiple of FLA, are these applicable to wye-delta start?

3. Any rough idea of how much variability in spike upon starting to delta from one start to the next (we have measured the current spike once, not sure how representative it is)

4. Is there a means to predict the worst-case current spike?

5. Is there any logical reason to have a wye-delta start in a power plant ?

Thanks in advance for any insight.


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

 

[rbulsara]

My two cents. I had seen and operated many Y-D starters up to 250 HP in industrial plants in 80's. None of them had detrimental issues. The current spikes during transition is normal. 40ms sounds very reasonable for contactor operation.

1. I would start with 10 to 12X.
2. Yes.- Use one those for molded case switches (700%-800%).
3. No, but it is more like LRA but only momentary.
4. May be, I do not know that math. But it is akin to energizing transformers or even plugging the motor, the worst case. Wait for more theoretical people on this one or if someone currently in a plant.
5. Cost. It is less of an incentive now a days but if they worked in 80's, there is no reason as to why they would not work now.

Key is to make sure that contractors are properly sized and installed.







Rafiq Bulsara

http://www.srengineersct.com

 

[waross]

Under the Canadian Electrical code, your motor may be protected by a thermal/magnetic breaker rated at 200% of FLC. The instantaneous trips would be capable of setting at 10 times that value. That equates to 20 times FLC. When we had instantaneous trip issues, we generally set the instantaneous a notch higher. It is providing short circuit protection, not overload protection.


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

 

[jraef]

A "trick of the trade" regarding open transition Y-D starting (assuming you can't afford to trash it in favor of a solid state starter).

"Roll" the conductors on the S contactor that go to the motor. Rolling is a process by which you shift the phases but maintain the same relative rotation. It makes a difference sometimes because of changing that phase shift.

By the way, if you search long enough on this site (I'm too lazy), you will find a similar thread where we discussed this years ago, and you yourself did some impressive math proving that the transition spike can theoretically exceed 2200%!


"If I had eight hours to chop down a tree, I'd spend six sharpening my axe." -- Abraham Lincoln
For the best use of Eng-Tips, please click here -> faq731-376

 

[electricpete]

Great comments.

It makes sense that rolling leads could help. Perhaps our goal is to get the delta lagging the wye by 30 degrees we might minimize the spike ASSUMING very minimal deceleration? Or is it more of a trial and error thing?

I searched but couldn't find discussion of 2200%. It must be a sign of my age that I didn't remember that thread. I can vaguely recall some logic: If the motor drifts to 180 degrees out of phase, then we have twice the voltage applied accross one times the locked rotor impedance, resulting in twice the current as a DOL start. If we add in the factor of 2*sqrt2 multiplier for peak instantaneous (including dc) above rms (not inclding dc), then the instantaneous peak could be 4*sqrt(2) above rms locked rotor current (very much theoretical worst case... actual expected much lower). Maybe I was thinking something like 4* 550% = 2200%.

My search was not wasted because I did find many of the good comments from jeff, Mark, Keith, Bill, Gunnar etc. It does seem we can and do routinely get spikes with wye-delta higher than we would see with simple DOL, so using DOL rules for instantaneous setting doesn't seem a good approach to me.

One good thread in particular was:

http://www.eng-tips.com/viewthread.cfm?qid=29369

Unfortunately the linked Nailen article on wye-delta transition spikes is no longer available at that address. Does anyone know how to find that article?

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

 

[electricpete]

Hmmm. Actual if the wye residual voltage was at 100%, shifted phase and did not decay, then I think it would contribute 173% voltage when reconnected in delta. So worst case would start with 100% incoming + 172$ residual = 273% instead of 200% voltage for out of phase connection. At any rate I don't think that worst case calculation has much practical meaning.

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

 

[jraef]

In that old thread, I referred to an article by Richard Nailen that has unfortunately been removed. But that company, Master Controls, also still published this one that has similar corroborating information. They also had a device that can detect what they call the "leading phase" in an open transition starter, called a "Leading Phase Indicator". Its just a little box with 3 lights on it that tells you if the phase shift is leading or lagging, so you can roll the conductors and diminish the effect. Last time I spoke with Bill Stelter a few years ago, the inventor and owner of Master Controls, he had all but forgotten his invention but was considering marketing it again (at my prodding). I don't know if he did. But here is a copy of his patent application and in it he describes the issue in detail, using references from that Richard Nailen article. Even this patent application is an interesting read on the subject.

http://www.patentstorm.us/patents/5142213/description.html

"If I had eight hours to chop down a tree, I'd spend six sharpening my axe." -- Abraham Lincoln
For the best use of Eng-Tips, please click here -> faq731-376

 

[electricpete]

Very good information. Thanks Jeff!

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

 

[waross]

Thanks Jeff.
Yours
Bill

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

 

[Skogsgurra]

I have been looking for some recordings from starting an old rotating welding converter that caused us big problems. Not found so far, will look more.

Anyhow. Rolling phases did not help and the spike we saw was high and short. As I remember it, it was like 300 A (instantaneous) and FLA was around 15 A (at 380 V 3 phase).

We finally solved that particular problem by increasing the delay between Wye and Delta to 500 milliseconds. Then, the EMF from the rotor field had decayed enough not to cause a high current spike.

The motor speed doesn't drop much in 500 milliseconds thanks to the high inertia. And since all loads that need a soft start or Wye/Delta usualy have a high inertia, I think that the long delay can be used as a universal solution to most situations where the second peak causes problem.

Gunnar Englund

http://www.gke.org

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

 

[rbulsara]

To backup what waross said, all of the the Y-D starters I talked about had fuses sized 2-3 times the FLA of the motors. That would be akin to using thermal-magnetic breakers with INST at 10 or so, which is like 20-25X FLA of the motors.

Rafiq Bulsara

http://www.srengineersct.com

 

[waross]

Gunnar's example is also 20 times FLA.

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

 

[Marke]

The magnitude of the transient is a function of the residual voltage generated by the rotating motor and the phase angle between the generated voltage and the applied voltage.
The longer the OFF time before the reclose, the lower the voltage that is generated and the lower the transient, but during this OFF time, the motor will slow and that can mean that in addition to the transient(reduced) the steady state current immediately after the reclose may be close to LRC.

The first thing I would look at, is what speed the motor is rotating at when the transition occurs.
If the transition is occurring at less than full load speed, the steady state current will follow the DOL curve until the motor reaches full speed.
I find that in the majority of installations, the transition from star to delta occurs well below full speed. Often, this is due to the torque available in star being insufficient to accelerate the motor to full speed, so transitions around 40 - 50% are not uncommon.

If the transition occurs at significantly less than full speed, there is still the reclose transient as well as the DOL steady state current.
Typically, it is the steady state current that causes the breaker to trip.

If the motor is operating at full speed when the transition occurs, and if the transition timing is very consistent, and the motor deceleration during the transition is also very consistent, then the phase shift at the reclose will also theoretically be consistent.
The delta connection can be configured to either the leading phase, or the trailing phase. One configuration will minimize the phase difference at reclose and minimize the transient. This can be proven in a laboratory with an open shaft motor, but in real applications, it is less relevant.

So, I would check that the motor is always at full speed before the transition is initiated.
If the motor is at part speed and lacks start torque, then you would be better to revert to DOL.

Best regards,
Mark.

Mark Empson
L M Photonics Ltd