Happy New Year!

Here is a question posed by one of our forum members.

Do you feel standalone Adjustable Frequency Drives should have an arc flash label?
Yes
No
It depends (please explain)

If you have a question you would like see asked as “The Question of the Week� please send it in!

A simple wall mounted VFD should be treated the same as any other piece of end use equipment that 'just sits there'. An User Interface, that can be operated externally, does not increase the possibility that an arc flash will occur.

However a VFD that is installed in a 'motor control cabinet' which must be opened for servicing, such as a multi-section drive or a unit with an automatic bypass scheme, probably has a relatively higher chance of an fault occuring.

The question specifically says "Standalone" not "standalone or installed in a MCC or otherwise group mounted" so I voted "No".

I don't consider a standalone drive something that would need to be gotten in to while energized to be maintained.

I used to label all disconnects, drives, cam-lok connection boxes (we have broadcast TV trucks where I work, local PBS, ESPN, ESPN 2 etc.), starters, indoor dry transformers, etc. The models were huge as was the amount of work and greater potential for error to do that. 99% of the time a standalone drive and the rest of the items I mentioned can be shut down upstream from a breaker or switch in a panel or switchboard and not affect anything else. Therefore our policy is if it isn't labeled then shutdown is required.

I've labeled all of standalone drives. The best I can remember, they all had negligible levels of incident energy available.

Per NFPA 70E-2015 Electrical equipment such as switchboards, panelboards, industrial control panels, meter socket enclosures, and motor control centers that are in other than dwelling units and that are likely to require examination, adjustment, servicing, or maintenance while energized shall be field-marked with a label....

We consider a VSD as an industrial control panel. Ask any electrician, there have been many times that they have worked on a VSD with it energized and the door open while troubleshooting or taking voltage/current measurement. We always will label standalone VSD units.

I label them because it is a piece of equipment that is "likely to be serviced while energized", ie there is potential to take the cover off and troubleshoot why it isn't working. Thus it's reasonable to need a label.

As a not-so-simple AC drive, let's consider a locomotive traction drive or a mining excavator drive. I don't know much about the locomotives but I know quite a bit about the excavators. There are a little over a hundred of them in existence world wide right now. The one I'm familiar with uses Semikron IGBT modules. There are 5 drives. Each one is a "common DC bus" type VFD with an active front end (AFE) converting to DC and 3 inverters coverting back into AC driving 2000 HP motors. It runs at 690 VAC on the front and back end. The limit is set because going higher in voltage means that we go from the industry standard 1400 V IGBT's that are used in all the low voltage drives up to 600 V to medium voltage IGBT's or other types of devices, and the price goes up dramatically. The drives are also water cooled due to space limitations.

As with any drive, the power electronics are semiconductors which means that I^2*t has to be respected or the power components can literally explode and fling shrapnel out of the drive. So either fuse (or high speed circuit breaker) the drive or it will "self protect" the downstream loads by becoming sort of an expulsion fuse. Semiconductor fuses are absolutely amazing and trip in 1/4 cycle simply because if you consider the AC sine wave, any given 1/4 cycle segment will have the peak current value in it. That's tripping in just 4-5 ms. Incident energy downstream of this kind of protection is very, very low. In the case of the monster system described above it never got above 1 cal/cm2.

You couldn't open the doors with it energized (door interlock switches) but that's not the case with all of these types of systems. Even if you did...come on, <1 cal/cm2 even with a 6000 HP output drive running as a servo so the instantaneous output is even higher?

I'm not speaking of an exception to the rule either. Every time I run the calculations around a drive, I come up with basically the same result...no significant arc flash once we get away from the enclosure with the fuses/circuit breaker in it. For some of the smaller 100-500 HP discrete style drive packages like this though I can see where you'd basically have to treat it like a feeder breaker enclosure with some extra stuff attached to the downstream side.

Well, if you are absolutely certain you will never need to open the cover to access the configuration or make adjustments, then probably not.

BUT I have been working on variable speed controls for over 30 years. At some point in time the cover needs to come open for one reason or another.

Oh, I calculated the Incident Energy for our facility's repair department, and found that the incident energy was indeed surprisingly low. Especially since I had them change the fuses to something faster-acting.

All of the calculations I have done point to the same results you have. I had found that the highest incident energy is where the 3-phase line voltage is connected. A lot of our drives (especially DC Drives) have internal fuses so the incident energy is higher at the line-side of the fuse, as compared to the drive-side of the fuse.