I don't think a fuse needs an electronic trip unit with a sampling algorithm to detect an overload by harmonic components of the load. It just melts. Which handles with the overload simple, elegant, but not unintelligent way.

 

Worst case pull a separate full size neutral for each phase. You know why BMWs are junk as in money pit? Because the engineers over engineered everything with electronics.
An engineer once told me the more complicated you make something, the easier it is to break.

 

Back in the 80's when AC drives were just coming out we thought they were the best thing ever. We installed a crap load of Volkman AD-1, AB 1333 and ETrac Drives in the plant I worked in. We had 5 substations and checked current regularly to make sure they were not overloaded. One day lights and machines started flickering so we went to the substation. I start flipping the voltage and current meter switches to check loads and turn to my boss and say I'm cutting the main the Voltage is all over the place.

I push the open main button and look over and my boss is laughing and pointing up. I look and smoke is pouring out of the xformer. I was so interested in checking loads I never looked up, he saw it as soon as we walked in.

After opening the covers we found a puddle of melted aluminum on the floor, we told plant manager that it was a uncommon thing to loose a xformer like that not even 75% loaded. Week later same thing to another substation. Corporate sent out and engineer and they thought it was due to the heat from harmonics, not showing up on the load current.

 

After opening the covers we found a puddle of melted aluminum on the floor, we told plant manager that it was a uncommon thing to loose a xformer like that not even 75% loaded. Week later same thing to another substation. Corporate sent out and engineer and they thought it was due to the heat from harmonics, not showing up on the load current.

And in the early days of drives, this did happen, this was figured out and standards were written for drives and transformers.

 

Bi-metal trip elements CAN be designed to withstand the possible nuisance tripping doe to high harmonics, but most are not. There is evidence of this happening, but because it tends to make the devices trip just a little EARLIER rather than later, it isn't a thing that people tend to worry about on non-adjustable devices. That's especially true when looking at line side devices, such as circuit breakers FEEDING VFDs, because the earlier tripping tends to still fall within the tolerance band of the trip curve of the breaker.



Where this has reared its head is in things like adjustable thermal OL relays used DOWNSTREAM of a VFD where the human response to a nuisance tripping is to "crank 'er up" until it stops bothering them, so the result can be lost motors. But most VFDs now have built-in OL protection, so it's no longer necessary to have a thermal OL down stream of a VFD. The only time it happens now is if you use one VFD to run multiple motors at the same time. In that case, you need a compensated OL protective device or one that is insensitive, such as eutectic melting alloy style OLs.