VFDs - Trouble restarting loaded conveyors

[dgauthier]

Hello, all

I'm a process control guy and I'm having to work with VFD-controlled motors more frequently lately. As an old boss used to say, "I'm not sure I understand all I know" about motors.

I just built a PLC controlled panel for coordinated control of a conveyor line. VFDs were selected according to the "Heavy-duty" ratings and all motors (common 460V 3ph induction, 3 to 20hp)run well below FLA at frequencies ranging from 65 to 110Hz, even heavily loaded. Still, many current-limit and stall on loaded restarts and have had to be geared down. All are configured for sensorless vector and have been autotuned. According to specs, these drives can only supply 150% of rated current for 3 seconds, have a hardware limit of "200%, typical" and an instantaneous trip of 300%. All of these numbers are well below the typical 6x FLA permitted by a standard starter. Is this the reason for the restart problems? I see in the archives that torque is related to applied voltage. Are these current limits causing the drives to scale back the applied voltage? Is more torque what I'm even looking for?

V/hz is an available control option. Can I use this method to get better starting torque? If so, what are some guidelines for properly setting the associated parameters?

Thanks in advance
DG

 

[dpc]

Standard off-the-shelf drives are generally designed for use on variable torque loads such as centrifugal pumps and fans. The conveyor is a constant-torque load which requires a high starting torque. It appears that your drives cannot supply adequate current to break the conveyor loose when loaded.

Loaded conveyors can be challenging to start even with a constant-speed motor. Often, conveyors require reduced-voltage starting to limit the maximum torque during starting since excessive torque can damage the conveyor belt.

I suspect that a slightly larger drive would be able to start the conveyor. It is possible to purchase "constant torque" drives that provide higher current output capability.

If you don't want to buy a new drive, you might be able to install bypass contactors to allow you to start the motor across-the-line or through a soft starter when trying to start a full conveyor.

For a normal shutdown, it is common to put in a time delay to allow the belt to purge before shutting down.

 

[misom]

Well, here are some thoughts and questions to ponder.

1. Is the motor big enough and geared properly for the application?
2. How much torque do you need to start the loaded conveyor?
3. How fast do you need to get to set speed - usually process dependent?
4. If you know speed, torque to accelerate the conveyor loaded (worst case), how fast you need to get to set point, than a motor can be selected.
5. Drive selection - you should consider Flux vector drive that is a closed loop (no feed back required)speed control that will generate full toque at stall with 200% overload. 6. V/Hz drive topology will not help you.
7. Motors have peak torque overlaod limits also.

 

[tmahan]

Try setting your voltage boost to the max value suppoted by the drive. That will by incressing the voltage for a given frequency the output torque is increased. Then reduce the acceleration on the VFD from zero to the base speed that you plan to operate at. The drive may have programable acceleration ramps. If so design one that takes a while to go from 0 to 20 hertz or so an then increases the rate of change from 20 on up. That will give the motor time to take advantage of the voltage boost it is being provided.

As DPC mentioned an across the line start for a loaded conveyor may be an option in your case if all else fails. Frequently though across the line is not a valid option for other reason such as product damamge or mechanical damage to the conveyour itself. I do disagree at the idea of using a softstarter.

If a VFD is unable to start the load then neither will a softstarter. Softstarters reduce voltage without reducing frequency so they will produce less torque than even a variable torue drive does.

 

[tmahan]

Misom,

I think I have to disagree (or at least question) with a couple of points. Someone including you may show me to be wrong later but here goes

Point 5. The original post indicate a sensorless vector drive. That is a "Flux vector drive". I think that if no feedback is used/required then it is not a closed loop system but rather an open loop system which uses mathmatical approximations to substitute for the feedback sensor.

Point 6. I think Volts per herts adjustments may very well provide a great deal of help on this application. Using a higher ratio (V/f) will result in greater torque output. There limits to this but all the torque boost features on VFD's do is provide an increase to this ratio.

7. Can you please provide some further explanation

Thanks

 

[electricuwe]

dpc,

I have to disagree to most of your statements.

1.There are some low cost drives, providing only V/f-control, on the market especially designed for pumps and fans, but the majotity of drives available should be able to work on a constant torque load.

2...even with a constant speed motor.
With this statement you state the implication that this would be the best option, which is definitly wrong.

3.I agree to your statement that a larger drive would solve the problem.

4. With a bypass contactor you would lead the VFD-idea ad absurdum.

5. Maybe true

 

[electricuwe]

Just a hint to dgauthier:

since you have drives (motors and inverters) of different ratings it might be sufficient to buy one larger unit and replacing each smaller one with the one becoming redundant.

 

[dpc]

To electricuwe,

Thanks for your comments.

I haven't specified a drive in three or four years, but in the US, every drive manufacturer would very clearly point out that the horsepower rating of a given drive was based on a variable-torque load unless a constant-torque drive was explictly specified. And I was involved in specifying a lot of drives. In some cases, we have variable torque loads (sewage pumps) that would stall using standard drive due to ragging, etc. To eliminate the problem, we had to use constant torque drives. I'm not making this up. Maybe things are different in IEC land. The problem comes from the use of horsepower to specify drives, rather than output current. Most U.S manufacturers would give two different prices and models - one for variable torque and one for constant torque. This was true up through 250 hp or so.

In many instances this was necessary to be cost competitive. Since most drives were being used for variable-torque applications, the manufacturer could under-size the drive, and sell it cheaper. Any supplier who quoted a fully-rated constant-torque drive for the same application, would be at a cost disadvantage.

I certainly was not trying to imply that a constant-speed motor would be a better solution - although I have seen these used many times for conveyor belts. In the old days, we used wound-rotor motors in order to limit the starting torque to avoid damaging the belts. I was simply trying to point out that starting a loaded belt conveyor is problematic.

Bypass contactor would allow the conveyor to be started, and would be a less expensive solution (perhaps) than replacing the drive. Excessive torque issue would have to be addressed. I agree that this is not the best solution, but it might be an improvement over shoveling off the conveyor whenever it tripped loaded.

It is easy to give a solution of replacing the drive, but it is harder to find someone to pay for it.

 

[electricuwe]

dpc,

from my own experience in designing inverters for VFDs I know that for a voltage source VFDs, which is the most common type in that power range, you safe nothing by specifying it just for variable torque loads because the main issue in inverter design is output current which directly relates to torque.

Limiting output current at very low speed is sometimes done to reduce the stress related to junction temperature swing, but this is only an important issue in larger drives and for longer periods of time.

For variable torque loads there are only some savings in the control section because V/f-control is usually sufficient for that application. Thats the main reason why thes type of drive is still on the market.

For the motor part of the drive its diffent:

constant torque application requires oversizing or an independent fan for cooling

 

[electricuwe]

If you operate a 460 V motor on 460 V line using an inverter and you operate with frequencies in excess of the line frequency you operate the motor wuth reduced flux leading to significantly reduced torque!

Torque is proportional to flux² !

 

[gigahertz]

Try setting the acceleration time in the drive to maximum ...

 

[electricuwe]

A drive set up correctly should reduce acceleration rate if the current limit is reached.

 

[gigahertz]

That's correct. But if the drive goes to current limit and holds at 150%, it will trip after 3 seconds if protected according to the ratings given. If the accel time is set long enough to get the load going and not stay in current limit for extended periods, trips will be avoided. That's assuming current limit is at 150%. If current limit is lowered to about 125%, the load can probably be started more easily, but there may be problems maintaining speed if load is applied suddenly. Either will work depending on the process.

 

[TheDOG]

My $0.02
I have heard of applications that could not be started across-the-line, being started by a SLV (sensorless vector) VSD. A VSD attempts to use only the "steep" part of the motor torque Vs speed curve ie. near full speed.
If the acceleration rate on the VSD is set to short, slip increases and it is possible that the drive is using the curve before (@ a lower speed) breakdown torque.
dgauthier;
Have you measured the starting current? What is the acceleration time set to?
I would sugggest decreasing the accel rate to see if this made a difference.

 

[dgauthier]

Hello, all.
Thanks for the response to my original post. You're a goldmine of information.

FYI, the drives in question are A-B PowerFlex 70s.

Reducing the accel rates per advice from A-B and others did help. Also, (according to AB) virtually any properly-configured drive should outperform across-the-line starters with regards to available torque both starting and running. Reason cited was that because of inefficiencies related to slip at startup, A-T-L starters with NEMA B motors can provide at most 170% to 200% rated torque. Because VFDs control slip they produce much more torque with significantly less current. I'll leave further discussion on the matter to you.

Tuning the drives, gearing down two motors and making minor control changes to better manage loading has helped a lot and the system is working fine.

 

[dpc]

electricuwe,

I'm not an expert on drive design, so I'm not disputing what you say.

Maybe the "variable" and "constant" torque drives marketed in US are defined a little differently. My recollection is that the so-called constant torque drives had a higher output current rating than the variable torque drives, so that agrees with what you are saying, I think.

Thanks for the viewpoint of the drive designer.

 

[jbartos]

Suggestion: Visit

http://www.oit.doe.gov/bestpractices/energymatters/wint2002_ask.shtml

http://www.facilities.upenn.edu/uop/3_electrical/Section_16483_doc.pdf

(This is university stuff!!!)

http://www.esmagazine.com/CDA/ArticleInformation/features/BNP__Features__

(With a couple of graphs.)

http://www.vfds.com/benefits.htm

(Good benefits of variable speed drives.)

http://www.vfds.com/specify.htm

(How to specify VFDs.)

http://www.caconline.com/Solutions/EP/ePAC_PCP.htm

(Vector Flux Drives for pumps.)
etc. for more info.