Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations IFRs on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Trouble switching motor from VFD to line power

Status
Not open for further replies.

eeprom

Electrical
May 16, 2007
482
Hello,
I am trying to drive a 50 HP induction-motor pump alternately with a Yaskawa VFD, and line power. What I am trying to do is to use the VFD to control the pump while the synchronous speed is less than 60 Hz. When the synchronous speed reached 60 Hz, I use a set of reversing contactors which disconnect the pump from the VFD, and then connect it directly to the grid. Sometimes this works fine. The problem is that occasionally the connection to the line is very rough, and it results in popping a breaker. It seems that the transition should be smooth. I can start the motor directly from the line with no problem. But when the VFD drops out and I try to connect from the line, I get the bump. Any ideas why this would be happening? And any ideas on how to fix it?

thanks,
Andrew
 
Replies continue below

Recommended for you

Andrew,
It is highly likely you are out of phase with the motor. You could be 180deg out and this will lead to your breakers popping out. It is the same with a Star-Delta starter, the transition from Y-d can lead to the same result. Once you switch off to the motor and then back on, you are in no-man's land as far as the phase direction.
When you go from line power to your Yaskawa VFD, this will probably have a function called "flying Start" that will measure and 'catch' the motor without any of the problems you enounter going DOL. You can control the output from your VFD but DOL will simply provide full voiltage at 60Hz.

DOL=Direct on Line
 
When the motor is disconnected, the magnetic field takes a little time to decay. If the motor is reconnected before the field decays, the field can be work to make the inrush current higher than it would be with a normal line start. The problem can be prevented by delaying closing one contactor by up to several seconds after opening the other. At 50 Hp, something like 1.5 seconds may be enough. Someone else here might have a better estimate. It is possible to synchronize the two power sources and use a quick or even closed transition, but that is difficult to do. I believe that even some who have done it successfully are not inclined to do it again.
 
Sed2developer,
I have considered this situation. But the motor is an induction motor, and so as soon as power is removed, there is no more field, and therefore there is no phase. Am I incorrect about this? I would assume that the disconnected rotor has no remaining induced current, and so it is simply free spinning. Can you please explain why there would be a phasing to considered before connecting to the line?

thanks,
Andrew
 
eeprom--

When you disconnect the motor from the VFD, it's still spinning AND the magnetic field has not yet decayed. For a few cycles the motor is generating 3-phase voltage. If you close the contactor with the motor's decaying three-phase voltage out of phase with the line supply, the end result is a very high current. This is why your breaker trips.

The fact that it doesn't happen every time is due to the randomness of the situation's several factors.

old field guy
 
Thanks for the help. This makes sense.
 
Andrew
I should have added the part on that CJCPE and oldfieldguy have added.
There will be motor EMF and possibly opposite phase that will lead to a high short-circuit surge current. Pop.
The 1.5secs is about right. There will still be a surge as the motor gets back to rated speed but nothing like going DOL or trying to reconnect too quickly.
 
The motor field will take some time to decay. I would expect a 50hp to require around 0.5 seconds for the terminal voltage to decay. I have seen larger motors (100's of HP) require 1.5 to 2 seconds.

A wye-delta starter can be built with resistors. The resistors are connected first during the transition to "pull" the motor into phase with the power system. Then, the main contactor is closed. I know at least one chiller manufacturer that requires their wye-delta starters to be closed transition because an open transition damages the chiller.

Maybe you could do something similar here?

 
eeprom; What you're doing is guaranteeing that you will be randomly popping breakers or something worse eventually.

Why are you using a VFD? If you don't need the 'V' you should probably just change to a soft-starter as they usually deal with the contactor transition automatically.

Generally pumps are nice things to start DOL. Do you have voltage sag problems if you go that way?

Keith Cress
Flamin Systems, Inc.-
 
I agree with the opinions above. The induction motor becomes a generator for some cycles after the power supply is disconnected. Out of step voltage reconnection develops high currents and mechanical stress surges.
A time delay will allow dissipating the residual magnetism and generated voltage, and the resultant transient.
To avoid the reconnection current surge due to high slip of the motor speed, increase the VFD frequency a little bit over 60 HZ before the VFD is disconnected, that way the motor speed will be still close to 60 HZ synchronism when the main supply is connected.
In my opinion 0.5 seconds delay (30 cycles) is more than enough to dissipate any residual voltage. Remember that the pump is loaded and the speed will drop sharply when the electric power is removed, since the pump inertia is normally light.
 
aolalde makes a good point. Too long of a delay will allow the motor to slow down enough so that when reconnected X-line, there will be a significant inrush current again, along with the associated torque spike. In my work on DC injection brakes where we had the same issues, we found that the field decay time on motor designs under 100HP was approximately 28ms. But to that you must add the reaction and opening time of the contactors, which can range greatly from brand to brand. For example, ABB contactors that size take up to 130ms to open. Then you add in the closing time for the X-Line contactor of around 80ms and you start to approach 1/2 of your .5 second delay already. So I recommend a delay that is .25 sec., then allow the mechanical issues to do the rest. Any longer and you risk a re-close spike that will damage your equipment.

As an aside, this subject has been dealt with extensively by VFD manufacturers in what are called "synchronous transfer" bypass systems, originally promoted by Robicon drives for years MV systems and taken up by several other manufacturers now on both MV and LV drives. Yaskawa just doesn't happen to be one of them. In a sych. transfer system, the line sine wave is monitored and the VFD output is synchronized to it and both systems are connected in parallel. But a reactor is put into the VFD side during parallel operation to prevent damage to the transistors. Not something I would tackle without experienced help however.
 
I suppose the easy fix is to not bypass. What's gained by bypassing?
 
that last point is valid, I want to know what is the purpose in bypassing the vfd after startup?
 
In my experience it is typically done so that the VFD can be reused for an adjacent pump. The extra contactors and SCPD to do the bypass correctly are less expensive than the additional VFD(s), especially as you get to have more and more pumps. I have seen this done for as many as 5 pumps and one VFD. Theoretically there is no limit to how many you can do but at some point the extra control gear to do it makes it uneconomical compared to buying another VFD. There is also the issue of a lack of redundancy to consider. But I did one job where we put in 3 VFDs for 9 motors, it was less than 1/2 the cost of 9 VFDs. It's a trick often used when budgets are tight, but it is also fraught with pitfalls for the inexperienced.

It's also done sometimes when the full speed run time grossly exceeds the variable speed run time, or the VFD is used just as a soft starter for extremely weak systems where a regular soft starter will not adequately reduce starting current, i.e. 1.0 - 2.0 x FLA situations. To leave the motor on the VFD for long periods of time at full speed means suffering the energy losses in the VFD.
 
Yes, this is correct. The system I am speaking of has three pumps. When the system comes on line, there is only one pump. If that pump reaches full speed and the line pressure is still too low, pump 1 is connected across the line, and pump 2 is then connected to the vfd. Pump 2 then becomes the control pump. This system is less expensive than 3 vfds. However, we are running into problems in trying to get the pumps to switch from vfd power to the grid.
 
One trick that has worked for me in the past is to do what aolalde said in his post; slightly overspeed the pump right before transfer, maybe 5% or there about. Then when you go into open transition and it coasts for that 1/2 second, it is closer to normal speed when reconnected. Tricky to get right however, takes some trial and error which can be hard on your system.

But then again you already have that condition now so it can only get better. (famous last words)
 
Why don't you just run with two fixed speed pumps and one variable? With a bit of simple logic that would give you a controllable output over the full range from zero to three pumps flat out.


----------------------------------
image.php
Sometimes I only open my mouth to swap feet...
 
ScottyUK,
The usual reason is the same for any triplex or duplex pumping station; alternation of the pumps to avoid excessive wear and tear on any one pump, plus complete redundancy. It's the same reasoning behind not just doing it with one big pump.

It's a big deal to the pump people. I'm working on a triplex system design right now (non-VFD) where we are using kWh and cycle counts on each pump to feed back into a PLC. The PLC then determines a more complex alternation scheme based on cumulative motor heating and damage curves from excessive starting during wildly variable flow conditions. It will also allow the system to revert to a full duplex if any one of the 3 pumps goes down, then a full simplex if only one pump is alive. When pumps are then fixed and returned to service, the PLC will then rebalance the service time.
 
Hi Jeff,

Yes, that makes sense.


----------------------------------
image.php
Sometimes I only open my mouth to swap feet...
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor