Arc Flash Curtain around switchgear

I am looking for input for placing arc flash curtains around my 4160 switchgear. The arc flash boundary is upwards of 460' (according to the calcs) in some of my 4160 areas. My thought is placing a curtain around the area would reduce that boundary. Has anyone implemented this type of modification successfully?

There are shock protection blankets but no "arc flash curtains" as such.

There is arc resistant (medium voltage) switch gear that is constructed by having a substantial front face and "blast doors" on the top or back to redieect the arc blast. These work as long as the front doors are closed and all latches attached properly. However, the design for this is a performance test...it reveals nothing about how to do it without lab testing.

In a similar way any sort of substantial wall is going to limit or eliminate an arc flash hazard. The tricky part is that there is no test data or calculation to indicate what would be an acceptable wall. Most folks use a little judgement and assume that a substantial (block/concrete) load bearing wall will do the trick.

By far the best approach is reducing the arc flash itself. This means either modifying the trip units or breakers directly, or moving employees farther away with the use of mimic panels, handle extensions, and the like. Lots of folks like to talk about remote rack out devices but experience has not been so good with a lot of these. It seems that if the switchgear was intentionally designed then the device is more likely to actually work when needed.

Just out of curiosity are you putting the 460’ distance on your equipment labels?

Yes, to protect a large DC heater terminal array we used an arcflash blanket customized by Estex Manufacturing for a unique DC application. However this was AFTER other mitigation techniques were employed first. The dc voltage was reduced physically separating the + and - terminals from each other, then installing a non-conducting barrier between them. This reduced the voltage from 600Vdc + to - to only 300Vdc + to gnd or - to gnd. Cutting the voltage in half reduced the AF by a factor of 4. Next was to keep the protective curtain far enough away that the combined Kevlar/glass layers weren't compromised by the arc heat or blast pressure. The curtain had to be supported top and bottom as well as easily removed, so a mockup was sent to Estex to make sure we got it right the first time. The blanket material had been tested at Kinectrics with actual arcs and test dummies by Hugh Hoagland, but at specific distances and for various exposures expressed in kA and cycles. We had to convert this data to our metrics of cal/cm2 and psf with enough margin to validate the blanket would withstand the energy of not only one, but a cascaded event fed by multiple rectifiers in this installation. The engineers who designed the heater hadn't really thought about the arc hazard before we got involved, and would have been measuring the DC terminations to adjust the rectifiers (really dc drives) in shirt sleeves with multimeters with a (cascaded) exposure of 357cal/cm2 at 18 inches!!!

ASTM F2676 is for arc flash blankets/curtains. These blankets are now mentioned in NFPA 70E (my proposal).

Testing on, and use of the blankets for many years in the utility networks has shown excellent protection for workers.

We do most of the blanket testing and the blanket manufacturers can assist in how to install properly.

I do not know anything about arc-flash curtains but would like to gather more information from you on the arc-flash calcs. What is the fault current, fault-clearing time, and is there an upstream PD? The reason I ask is we had a situation at our station where our FPB was >100' and this was due to a low fault current being seen by the PD. The low current resulted in an extremely long FCT and subsequently a high incident energy. We adjusted the PD's instantaneous setting to fix the issue. Hope this helps some.

Tweason:
Use either 2 or 3 seconds as your manual time and rerun the calcualtions and see what you get.

Elihuiv:
On the topic of blankets, if possible please provide answers the following:

1. Aren't most blankets merely dielectric protection and that is where the testing/retesting is required.

2. Has ASTM actually tested blankets that are protection from incident heat energy.

3. Has anyone tested blankets with the IEEE box as the heat source. The IEEE box subjects material to a different type of heat energy than much of the ASTM arcs produced by two rods.

4. Is there an actual reference in the new 70E that discusses blankets as protection from incident energy?

I am posting the labels with the flash boundary. I am at a nuclear power plant, so there is a nuclear safety aspect that I must adhere by. Changing the setpoints would be my preference and the obvious solution, but changing setpoints or gear is a very long term and expensive solution. I would like to come up with a solution that I can implement in the near term. There are concrete walls on at least 2 sides in the switchgear areas.

Are the HK breakers in question 1E? What types of interactions are you concerned about?

That's what I thought too but after looking, no, there is actually a thing called an arc flash blanket. It's a barrier to an arc flash. They have a rather interesting rating system.

The standard is cited but using them as protection has yet to be defined. We know what they will do but an installation standard has yet to be written and it is up to the user to decide at this point. Some improper installations will provide significant protection but we have yet to reach consensus on how to approach this. NEVER use blankets and put workers in NO protection. They will likely reduce energy reaching the worker and MAY prevent all energy from reaching the worker but the installation and matching to the hazard is the key. The manufacturers who have tested have good guidance. Keep the discussion up. This will be a good viable protection strategy in the coming years. My 2012 IEEE-ESW paper gives a few new ideas if you got it.