Dave, thanks for that!
Do you have any general thoughts on when to use a line reactor, and how to size it properly for the application? And what about the other forms of industrial power filters, and how can we judge what type is most appropriate?
[Without getting an EE degree, that is]
Evidence is all around in layman's eyeshot.
The USEFUL cousin to what Dave mentioned came ages ago as a free bonus to railways and power company grid operators and even natural gas and petroluem pipeline operators.
A carrier signal could carry telemetry data to monitor and operate the <whichever> system over existing POWER wire as a piggyback. It wasn't long before voice and data for other purposes followed, then even leasing-out bandwidth to other users. All without having to buy more wire or such than was already run over massive networks spanning extreme distances.
BUT.. transformers meant to operate at 60 Hz blocked some of those useful signals and needed a bit of bypass gadgetry.
Fast forward to already a rather long while ago, and in-home LAN gadgets ran Ethernet over the house power wires, relying on the powerco transformer to keep it "local" to just the few homes served off the same transformer. Those are still hanging on the display wall at Best Buy and such, and may be among the several reasons you'd like to keep VFD or DC Drive noise to a minimum, your own shop and office or home where 'puters are in use.
Where I am heading with this is that "generally" those Corcom-style EMI/RFI filters can do rather a lot of good towards taming a VFD as well. Especially newer VFD of major makers that are already better behaved than cheaper or older tech ones.
The line reactors, then, are for the tougher challenges where there is more "power under the curve" in the problem zone, and/or the gear is installed in more sensitive environments.
You'd not be allowed NEAR a major telco Data Center with that noise. You be hunted down like a terrorist.
Too many billions of dollars being moved every few seconds by the world's major banks, for one thing, to not react to the detection of anomalies as a possible threat. Too many other folks with comparable interest in not being messed with.
On a more personal scale, I just like my WiFi and TV to work well, the microwave and fridge to not go bug***k.
Your service entrance panel maker - or maker of any OTHER loadcenter - list filtering/ protective devices with the general goal of reducing entrance of EXTERNAL nastiness or reducing the spread of INTERNAL nastiness.
They are not a perfect fit to knocking-down what a DC Drive or a VFD or even a compromised Phase-Perfect may get up to, but they have two MAJOR advantages: That of being Code-Compliant as "listed" or approved goods, and of being more affordable because of their (relatively) higher sales volume.
Start there. It's easiest. It even may be all you need to cover a whole facility.
Then go and ID any truant units and make the spend to put what THEY need right up close and personalized as Dave suggested. Eg: BEFORE their s**t has had time to reach the proverbial fan, broadcasting like a beacon all along the wire on the route as if it were an antenna before it reaches a filter.
"Test gear" to find the noisy ones? AM radio. Older and more cheaply made the better.
Tune it of off the bottom of its band to where there are no stations. Go listen to heterodynes and "stuff.
It's kinda neat even around 'puters. Back when they had only 600 Khz CPU clocks, yes "KILO Herz", not MHz or GHz, we'd set an AM radio next to one chunking away, do something useful. Knew by the sound when it had got hung in a loop or had finished its peas and taters and was ready for the next meal.
20CW