VFD & Overload
- Thread starter sii
- Start date
- Status
kwired
Electron manager
- Location
- NE Nebraska
- Location
- San Francisco Bay Area, CA, USA
- Occupation
- Electrical Engineer
A couple of minor warnings:
IEC bimetal overload relays are known to have issues with heating up disproportionately to the current going through them at lower frequencies and higher harmonics in the output of a VFD. So if your load is anything other than a centrifugal pump or fan*, that might result in nuisance tripping. Nuisance tripping is not inherently bad, but the human RESPONSE to a nuisance is usually to "make it go away", which on those IEC OL relays, means tweaking the dial up. That then can result in the OL relay NOT responding properly to a REAL overload condition and you can lose the motor. Eutectic melting alloy overloads are insensitive to frequency or harmonic heating issues and are the preferred OL for using down stream of VFDs. If you do want to use IEC OLs because they are cheap, make sure to use the little accessories that lock out or seal the dial adjustments after you get them set correctly.
Also, most solid state OL relays will not work down stream of a VFD.
*Centrifugal loads, aka "Variable Torque" loads, usually don't become a problem because at the same time that you lower the frequency, you also lower the load. So even if the OL sensor would be prone to trip too soon, the actual load has decreased at the cube of the speed drop too and that more than makes up for the added heating effects. Since the vast majority of VFDs are used on VT loads, most people never see the problem.
IEC bimetal overload relays are known to have issues with heating up disproportionately to the current going through them at lower frequencies and higher harmonics in the output of a VFD. So if your load is anything other than a centrifugal pump or fan*, that might result in nuisance tripping. Nuisance tripping is not inherently bad, but the human RESPONSE to a nuisance is usually to "make it go away", which on those IEC OL relays, means tweaking the dial up. That then can result in the OL relay NOT responding properly to a REAL overload condition and you can lose the motor. Eutectic melting alloy overloads are insensitive to frequency or harmonic heating issues and are the preferred OL for using down stream of VFDs. If you do want to use IEC OLs because they are cheap, make sure to use the little accessories that lock out or seal the dial adjustments after you get them set correctly.
Also, most solid state OL relays will not work down stream of a VFD.
*Centrifugal loads, aka "Variable Torque" loads, usually don't become a problem because at the same time that you lower the frequency, you also lower the load. So even if the OL sensor would be prone to trip too soon, the actual load has decreased at the cube of the speed drop too and that more than makes up for the added heating effects. Since the vast majority of VFDs are used on VT loads, most people never see the problem.
The bimetal OL relay responds to the heating effect of total load current whether it is linear or not. So I expect it to respond to motor overload conditions properly.
- Location
- Placerville, CA, USA
- Occupation
- Retired PV System Designer
The most likely explanation is that, for whatever reason, the resistance (not impedance) of the elements is frequency dependent.
Skin effect? Magnetic heating?
- Location
- San Francisco Bay Area, CA, USA
- Occupation
- Electrical Engineer
Well, it's a real problem, so just saying that it does not happen is more like religion than engineering.
My suspicion is a skin effect issue because of the fact that a bimetal OL is, by definition, two dissimilar metals with inherently dissimilar properties relating to heating, so undoubtedly different skin effect responses in those metals that become more pronounced under low frequency. I don't know if anyone has investigated it too much further than to understand that it IS happening, because there is a simple fix.
The bi-metal relay responds to the RMS value of the motor load current and so harmonics is no issue for it.
kwired
Electron manager
- Location
- NE Nebraska
Do you have a response to the last portion of the post you didn't quote?
Does seem to have at least some conflict with what you mentioned above. I know it was mostly speculation at this point but was also said that it is fairly recognized that something is happening in the bi-meta relays and this seems to be a logical thing that may be happening.
I have none because it seems gibberish to me.
- Location
- San Francisco Bay Area, CA, USA
- Occupation
- Electrical Engineer
As does blindly accepting a precept in spite of evidence to the contrary. Sure, bi-metal overloads are SUPPOSED to react only to RMS current, but look at the specs: at 50-60Hz. No evidence that they have been tested, or approved, for use at any other frequency. Here in the U.S., we still have a few places where 25Hz is still in use (many of the big pumps in New Orleans if anyone was wondering). Bimetallic OL relays are not used there as well, only melting alloy or very high-end solid state MPRs. I did a project for one of those after Katrina and that's what raised my curiosity. The specs explicitly restricted any OL protection from being anything other than melting alloy or solid state Motor Protection Relays that explicitly show approval at 25Hz (Multilin dies). A little more delving on my part after reading that dug up warnings from some bi-metal OL relay mfrs about using them behind VFDs or at frequencies lower than 50Hz. Nobody offered an answer as to why, only that it can cause nuisance tripping, leading to inappropriate adjustments.
Sure, that spec for linear loads. But the industry has found that bi-metal O/L relay works with non-linear loads as well.
Thermal overload relays work both with alternating and direct currents. US community took to melting alloy type O/L relay and European community to bi-metal relays.
Because then there would be a competition between thermal O/L relay and VFD for tripping on overload.
petersonra
Senior Member
- Location
- Northern illinois
- Occupation
- engineer
Modern VFDs provide overload protection already so why would anyone need to add an external overload (multiple motor situations excepted)?
- Location
- San Francisco Bay Area, CA, USA
- Occupation
- Electrical Engineer
Yes, multi-motor would be the primary reason, but also if the VFD is much larger than the motor. Most have some lower limit on sizing because of the accuracy of the OL sensing that's built-in. Typically if the motor is less than 1/2 of the amp rating of the drive, you'll want to add an external OL.
Last edited:
- Location
- Wisconsin
- Occupation
- PE (Retired) - Power Systems
Many, if not most, US manufacturers, like Allen-Bradley, standardized on bi-metallic overload relays as far back as 50yrs ago. From what I recall there were only a handful of manufacturers, like Square D, that had melting alloy OLRs as standard. Of course everyone offered both styles as options.
Probably the biggest difference is that US bi-metallic relays were typically sold with replaceable 'heaters' sized to protect the motor, while European designs used relays with fixed heaters with adjustable setpoints over limited ranges.
- Status