Sizing VFD recommendations

I am curious what your take is on sizing VFD's for variable torque applications. In the past I have sized for ND amp ratings paired with the motor (talking PowerFlex) drives but we have started running into limitations on performance so we talked with our VAR and they recommended in the future going to HD rated drives (basically for Rockwell this means the next size up drive).

I spoke with TechConnect the other day about this same issue and their recommendation was to size it so the software current limit could be set at 200% of the motor FLA. On the ND drives I know the current limit seamed to max out at 150% or 165%, I'm not sure about the HD rated drives. But If I use this sizing method the drive is basically 2 sizes larger than the ND rating (amp output). Is that really necessary? It seems excessive, especially if sizing conductors and other components to match the ratings; getting increasingly more expensive on the larger hp drives.

Another thing to take into account is I have been sizing my motors to handle 110% of the load the process guys hand me which has their own safety factor built in.

Thanks,

Eric

I like to put pen to paper and find the peak and average torque with duration for the application. I compare this to the motor current/load curve to get my peak and average torque. I will size the drive to the current with respect to load duration and peak torque with at least 1.25 engineering factor.

I started doing things this way after changing from the pf4 to the pf525. The 525 has a slightly lower peak torque and we were losing drives because the pf4 wasn't sized properly to begin with.

One thing to realize it that a motor started across the line will pull a tremendous amount of current. This can be 8, 10, 12 times the full load current on the nameplate depending on the motor design. This gives your motor a kick start.

However, with a VFD you do not get this kick. You can do some things in the drive with DC boost to help but sometimes it is not enough.

With across the line starters we often size motors to the RMS load of the system which counts on the fact that for short periods of time you can overload the motor if there is enough underloaded time to cool things down. With a drive that is sized close to the motor current limit then you can't really overload the motor.

For almost all hydraulic pump applications we pick a VFD one size larger than the electric motor size because it's much easier to start the motors and we can occasionally overload the motor. It all depends on how the VFD is rated; some include overload capability and some don't.

All too often, in pumping applications, the designer used the motor service factor for the full rating of the pump...

If in doubt of the design specifications, save a headache and up-size.

When sizing VFD's, you must know two things about the load to do it properly. First, continuous running torque, and second, peak short term (1 minute) torque.

You then take these two values and translate them into motor continuous amps and peak motor short term amps.

The drive must be sized to cover both values, continuous and peak.

If you don't know those values or simply don't want to do the work to find them, then you are left with arbitrary oversizing and extra wasted expense.

Congrats to Califflash for getting it just right.

Incidently, if you end up oversizing the motor for the load, you will pay for it over and over again with lower motor efficiency for the life of the motor. It's not so bad today with premium efficient motors as they will unload to about 75% without a large drop in efficiency but the older high efficient and standard efficient motors dropped efficiency rapidly when underloaded. The best motor for the job is generally one sized right to the expected continuous load.

I appreciate the praise DickDV. It does mean a lot. Unfortunately that was a hard won lesson. I came out of that struggle a much better engineer.

One thing that drives me a little crazy on some projects is that issue of the Engineer sizing the motor to use the Service Factor for the pump curve, then using a VFD. Most if not all motor mfrs will tell you that when running a 1.15SF motor from a VFD, it becomes a 1.0 SF motor.

The VFD sizing issue is this: if your load is TRUELY a variable torque centrifugal pump or fan and sized correctly for the load, it should be essentially impossible to cause a torque spike sufficient enough to put the drive into Current Limit. So selecting a ND drive should be absolutely fine if that can never happen.

What does sometimes happen however is that people ASSume that a pump is VT when in fact it is not, or something in the load profile can cause rapid increases in torque demand despite that the NORMAL load profile follows a VT curve. A classic example is centrifugal refrigeration compressors. Yes, they CALL them centrifugal, but when used with a slide valve controller, as 99% of them are, that slide valve can open rapidly and send a slug of refrigerant into the compressor that is like having a locked rotor for a fraction of a second. A VT/ND rated drive will trip off to protect itself by using the Hardware Current Trip, which means the demand for current rose faster than the drive can react to limit it. Sizing the drive for Heavy Duty / Constant Torque means the OL rating of the power devices can allow a motor to deliver BDT (Break Down Torque) for a few seconds. BDT is actually HIGHER than Locked Rotor Torque on a standard Design B motor and is almost always sufficient to handle short steep spikes in load demand.

jraef said:

The VFD sizing issue is this: if your load is TRUELY a variable torque centrifugal pump or fan and sized correctly for the load, it should be essentially impossible to cause a torque spike sufficient enough to put the drive into Current Limit. So selecting a ND drive should be absolutely fine if that can never happen.

Click to expand...

Jraef,

I don't totally agree with you here. While you are correct for the case you used I wouldn't use that as a general rule. There are certain applications that require an intermittent overload of the motor, less than 10 seconds for example. This typically does not heat the motor enough to damage it, it depends of course, but can draw enough current to damage the vfd.

The current limit is just another parameter in the drive so is not necessarily set correctly. And there is a response time involved. The faster the processor in the drive the better the protection will be. The pf755 has night and day performance in this regard to a pf4 or pf525, obviously.

I do agree that you should essentially ignore the hd rating, but I ignore the nd rating also. I prefer to size from the rms and peak current of the application and they're duration. Because that's what matters to the drive, not the hp rating of the motor.

Hmmm... I never said to ignore any ratings. Let me try again.

ND/VT ratings of DRIVES has ONLY to do with the overload capabilities of the power devices, actually just the transistors. The reason we CALL the drive ND/VT is because the intended use is for loads that CANNOT experience instantaneous load changes due to the physics of the load. You should NOT use ND/VT rated drives for any load other than true Variable Torque applications, centrifugal pumps and fans. EVERYTHING else needs a HD/CT rated drive.

I'm good with all of jraef's comments with one reservation regarding peak short-term current available from HD/CT rated drives. Since the NEMA B motor curve shows Break Down Torque to be 220% of nameplate torque, that would mean that the current at that point would also be at least 220% of FLA (actually closer to 260% since the current curve and the torque curve begin to divert at that point). While I have seen VFD's rated to 200% short-term, the majority of the ones I have come into contact with are rated 150% for one minute. While it is true that even those rated 150% for one minute can probably do a bit more for only 5 seconds, I would generally expect to upsize the VFD TWO sizes from the ND/VT size to reach break down torque in the motor. This rule holds pretty well for NEMA design A motors as well but is not valid for Design C or D.

Just one additional comment for those not familiar with NEMA motor torque-speed curves. Breakdown torque is the peak torque available magnetically from the motor. That level does not occur at start or stall but rather at the point where a running motor experiences increasing overload to the point where it collapses and starts slowing down to stall. The point just before this collapse occurs is the maximum possible available torque from that particular motor and corresponds to the point where the magnetic circuit in the motor reaches saturation.

Hope this helps clarify some of the terminology used above in this string

DickDV said:

I'm good with all of jraef's comments with one reservation regarding peak short-term current available from HD/CT rated drives. Since the NEMA B motor curve shows Break Down Torque to be 220% of nameplate torque, that would mean that the current at that point would also be at least 220% of FLA (actually closer to 260% since the current curve and the torque curve begin to divert at that point). While I have seen VFD's rated to 200% short-term, the majority of the ones I have come into contact with are rated 150% for one minute. While it is true that even those rated 150% for one minute can probably do a bit more for only 5 seconds, I would generally expect to upsize the VFD TWO sizes from the ND/VT size to reach break down torque in the motor. This rule holds pretty well for NEMA design A motors as well but is not valid for Design C or D.

Click to expand...

Dick, you're right of course, but in my experience most CT rated drives will have a short time rating higher than the 150% OL rating. The 150% for 60 seconds is just an industry "de facto" standard as a point of reference (based on what the transistor mfrs base their specs on). The trick on discovering the extent of that higher peak rating is to look for the "Hardware Over Current Trip" point, or words to that effect.

Thank you for all the great responses.