VSD Starting and Line Transfer 2

[Petrochem111]

We have implemented ATV 630 with by pass contactor scheme for starting 03 numbers Multistage Plunger pump system. ATC 630 donot have line synchronization provisions, so we have opted for open transition.

Motor is 400 V 50Hz 200 KW 1480 rpm and pump is 208 rpm with gear system in between. Pump starts smoothly on VFD but trips on high current once transfer to line. From our earlier experience of similar scheme in grinding mill we have included a delay of 200 milisec in PLC controlling all contactors transfer to minimize motor magnetizing current effect. Just want to clarify my doubt of over speeding limit of the motor before the transfer. What will happen if we increase motor speed to 60-65 Hz before transfer and at time of re-engaging it is more than 1480rpm as it will give us optimum time to complete transfer or it strictly has to be below 1480 rpm . I think due to 208 rpm pump speed , motor speed is decreasing very rapidly during power shifting.

Any suggestion will be highly appreciated. Understand that vfds will synchronization will be useful but they are too expensive, and most of their other high end functions are not utilized in this scenario. Reduced voltage Soft starters usually donot assist with such PD loads as we already had to replace one for our earlier grinding case due to available genset size.

 

[waross]

Motor current is largely dependent on the rotor slip frequency.
1500 RPM at 50 Hz is 30 RPM per Hz.
1480 @ 50 Hz is 20 RPM or 0.67 Hz slip.
The expected current will decrease with increased speed until minimum current will be seen at 1500 RPM. The current will then increase (as an induction generator) as the speed is increased above 1500 RPM.
At 1520 RPM full load current is expected.
At 60 Hz, the expected current will be based on a slip frequency of 10 Hz or 300 RPM slip.
Look at your motor speed/current graph to estimate the current at 300 RPM slip or 1500 RPM - 300 RPM = 1200 RPM.
You may be in the range of 200% rated full load current,depending on the motor characteristics.
That is for an in-phase bypass connection.
You should connect within about plus or minus 10 electrical degrees.
So, out of 360 degrees, you have a safe window of about 20 degrees and an unsafe window of about 340 degrees.
That is a ratio of about 340:20 or 17:1 that the connection will be out of phase.

You have options:
Both options depend on closing the bypass within the 40 RPM window between 1480 RPM and 1520 RPM.
Overspeed and a slow open transition. The timing may be difficult as you must allow enough time for the residual magnetism and the induction generator effect to decay before closing.
Using a sync check relay and a very fast in phase open transition.

Challenges.
Slow transfer; The speed may decay below the lower limit of the safe window before the residual has decayed to a safe level.
Fast transfer. It may be difficult to find contactors fast enough for a very fast transfer.
Transfer switches that depend on a very fast transfer use a rotating shaft and cam operated contacts for very fast operation.
One solution may be to use a very fast transfer transfer switch to switch between FVD and line operation.

CONTROL FEATURES (continued)

INPHASE MONITOR
FOR MOTOR LOAD TRANSFER
Inphase monitoring logic controls transfer and
retransfer of motor loads, so that inrush currents do not
exceed normal starting currents. It avoids nuisance
tripping of circuit breakers and mechanical damage to
motor couplings.

Link to Asco manual

Hint:
If you mirror the speed/current graph at the 1500 RPM point, you may use it to estimate the normal running current at higher RPMs.
This is also useful when investigating the effects of over-driving a motor above synchronous speed.
Warning, this will not indicate the excess currents that may occur if the supply voltage is not in phase with the residual voltage.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[waross]

I realize that it is not feasible to change the motor at this time, but in future, a 1440 RPM motor has three times the slip and three times as wide a safe zone for transfer.
You mention a grinding mill.
I would not expect a grinding mill to have a low slip motor.
Beware of transferring experience from a grinding mill to a low slippump motor.
But I will never see everything.
It may be a 1480 RPM motor on the mill, but it may be well to check the RPM on the grinding mill.
If the mill is subject to surge loading expect a high slip motor, but as I said, check the actual rated RPM.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[Petrochem111]

Thanks a lot for taking out time to clarify my problem. From your response, it seems that only slow open transfer option is available for me with my setup. From previous threads, I learned that Jref(sorry for mentioning the name) has mentioned that it has been done quite often in irrigation pumps. One thing need to be clarified for slow transfer, if we succeed in setting the delay appropriately, why we have to be concerned about phase angle as residual magnetism and the induction generator effect will already decay, and with 60 Hz (300rpm slip) even 200% current is acceptable to us as we will be able to controlled the starting current from stated 9xIn to something in region of 2xIn which will give much more liberation in time setting. Will the motor still be feeding any reverse current if it is rotating at 1800 rpm when comes in connection with 50 Hz line voltage.

 

[LionelHutz]

Overspeed it enough that the contactors can switch before the motor speed drops below 1500rpm then use a sync relay to close the line contactor.

Once you implement the sync relay, you can test it with varying over-speeds before the transfer to see what speed yields the lowest inrush surge. Keep increasing the speed until you find the speed where the transfer causes the lowest inrush current.

 

[edison123]

Try increasing the changeover time to 2 seconds for the flux to die, increase OC trip time to 2 to 3 seconds (motor can survive overcurrent for 3 seconds) and increase the speed to slightly over base speed, say 1510 RPM, before changeover thus allowing the decelerated speed to come to nearly rated speed of 1480 RPM.

Muthu

http://www.edison.co.in

 

[waross]

While not exact this is close enough to estimate currents at a given speed;
Look at your slip speed or slip frequency.
The slip speed is the synchronous speed minus the actual speed.
The frequency corresponding to the slip speed is the frequency that the rotor sees.
So, on a 50 Hz system, motor speeds of 1450 RPM and 1550 RPM will both cause a slip speed (and corresponding rotor frequency) of 50 RPM.
Building on Lionel's excellent suggestions, you may have a problem with a sync check relay.
If you allow enough time for the residual magnetism and potential to decay, you may not have enough signal to operate a sync check relay.
I suggest replacing the sync check relay with an accurate speed switch.
Minimum current will be closing at 1500 RPM IF THE RESIDUAL HAS DECAYED.
Glad to help and thanks for your suggestions Lionel.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[Marke]

Open transition switching is always going to result in transients to flux the iron.
If switching is too son after the disconnect, the rotor will still be generating and depending on the phase relationship between the "generated voltage" and the line voltage, and the magnitude of the generated voltage, the refluxing transient can be very severe.
Additionally, if there is an slip between the motor speed and the supply frequency, there will be a major current overload until the motor and supply are synchronised.
If the load inertial is high and the deceleration of the motor and load once disconnected is low, then it is possible to overspeed the motor, disconnect the output and wait until the motor speed is exactly at synchronous speed and then close in.
In the case of pumps, the disconnected deceleration is very high and it is virtually impossible to synchronise.
I use the Emotron Drives to synchronously switch with a closed transition. Works very well on pumps.

Mark Empson
Advanced Motor Control Ltd

 

[LionelHutz]

Pump starts smoothly on VFD but trips on high current once transfer to line.

What exactly is tripping, and does it have to trip?

I think due to 208 rpm pump speed , motor speed is decreasing very rapidly during power shifting.

The motor rapidly decelerating when disconnected from power means it's not a good candidate for an open transition switch. You might just have to overspeed and then "power through" a quick switch to catch the motor before the speed drops too much, transients be damned. I've seen pumps drop most of their speed in 1 second when line power is cut. On a pump like that you don't have time to mess around when making the transition. The contactors have to be switched about as fast as they can mechanically move.

You could try a synchronizing relay to check the VFD output is leading the line power and then transition as fast as the contactors can switch. The amount the VFD is leading will affect how nicely it switches to line power.

Another option could be using resistors. First close a contactor that connects the motor to line power with series resistors and then close the main contactor. It might help to soften the transition inrush. It doesn't need much time, the main contactor can be driven from an auxiliary contact on the resistor contactor. A similar idea is used in closed transition wye-delta starters and when connecting capacitors to line power.

A long open time between switching like 1-3 seconds can just result in across the line starting a stopped or almost stopped pump.

 

[Petrochem111]

With a delay of 300 msec and operating at 53 HZ, our open transfer and starting current go just fine. I think the reduced back pressure assisted us as 300 Amps motor is drawing around 150 Amps due to the reduced available back pressure.

Current during transfer remains around 200 Amps on panel-mounted energy analyzer, however recorded at 600 Amps level on Fluke clamp on meter "Inrush" current selection.

In my understanding, Inrush settings do not give motor starting current true representation, and motor starting current usually stays for some seconds, not like Inrush measurements, which represent only a transitional spike of a few cycles and may be beneficial only for magnetic settings. We took three readings and current on panel analyzer remains around 200 in all three starts but changes significantly on clamp meter Inrush selection.

Once again thanks all for their valuable inputs and guidance, and would be very thankful if you people commented on this Inrush measurement. and the industry standard of measuring starting current when intend to lower it from 9xIn (LRA) as I am sure that on Inrush clamp meter it would be much higher than 9xIn as around 6-9xIn is what usually observed on panel mounted analyzers.

 

[waross]

The first cycle peak current of a transformer may approach over 2.8 times the steady state short circuit current.
The actual value depends o residual magnetism and point on wave of switching.
Your ratio of 600 Amps vs 200 Amps seems reasonable.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!