Anyone have comments on the cost / benefit between the different types protection? I'm working with someone that is insisting on Arc Rated switchgear and (to me) your money is much better spent installing optic sensing technology that would reduce the amount of time and energy of a fault versus merely directing a blast outside.

My thoughts are
1. arc rated gear is much more expensive vs an optic solution
2. a fault inside arc rated gear = destruction of that section and possible replacement of ducting.
3. a fault inside optically protected gear = much quicker clearing time and likely much less damage to the gear section.
4. Arc rated gear does nothing to reduce amount of energy released and therefore can have a very high incident energy
5. Optically protected switchgear would reduce the amount of time regardless of the amount of fault current
6. Can I even put IR ports in Arc rated gear or did I just make a new exhaust port?

To me the best solution (provided nuisance tripping is not an issue) is to have remote racking, use optic sensing and install IR ports. They can do their IR scanning without opening anything, keeps them well away during racking / unracking and provides protection 24/7....

First off, I really haven't found a legitimate use for the optical relay stuff yet either but I'll back into that.

What does arc resistant switcgear do for you? It keeps the incident energy down to around 4 cal/cm^2 for the guy doing switching while standing in front of closed gear. The 4 cal/cm^2 number is based on research and testing from CIGRE which has been trying to correlate the arc resistant gear test with incident energy. It is NOT contrary to popular belief 1.2 cal/cm^2, so be aware of that little gotcha. Second, it only protects during switching operations. It does NOTHING for maintenance of the equipment. Guess when the vast majority of arc flash injuries occur? I'll give you a hint...arc resistant gear does nothing for those conditions. Arc resistant gear is a great marketing ploy, but doesn't help for the cases that protection is needed in the real world.

Frankly that's my issue with arc resistant gear. It's really expensive or it used to be. The price is coming down. You have to also make sure the room can handle it especially if it vents into the room. If it vents outside, it has to be secured so that varmints don't move in. And at the end of the day, the workers doing switching, the ones that are least likely to ever be injured, are the ones being protected. Maintenance workers are still in harms way and the arc resistant gear is harder to work on and takes longer to take apart and put together, never mind all the potential gotchas if you ever have to modify it to add another relay, switch, repair it, or do pretty much anything but operate it.

Second issue: optical ("arc flash") relays. If I do my coordination properly most of the time, the incident energy is down below 1.2 cal/cm^2 so I don't care. But when it is elevated, there are a number of relaying techniques that can be brought to bear. For instance if protective relays are already in use (a prerequisite for arc flash relays), then "maintenance switches" which temporarily enable instantaneous tripping can be added sometimes for no cost. Second with modern relays it is frequently easy to add zone selective interlocking which allows all breakers to use instantaneous tripping or something equivalent. So you get the speed of the optical relay again "for free" but also improved protection for bolted faults and other types as well, again "for free". There is also differential relaying which is frequently inexpensive to add and again offers massively improved protection for busses among other things and incidentally gives "instantaneous" tripping functions. I would consider optical relays to be just one option but when you look at the bigger picture, they are often really only useful for retrofits.


As to IR windows, did you know that the IR seen through one at any angle other than direct, straight ahead, is decreased so that the readings are not directly comparable? This is a property of ALL windows no matter what the material is. So it can't be avoided and those little tiny ports just don't give you enough of a view to be worth very much.

IR window manufacturers, switchgear manufacturers, and relay manufacturers all have some very interesting devices that can help when applied in the correct situation. But this ignores traditional protection technologie that are mature and have been around for decades and offer vastly improved protection for all faults, not just arcing faults. IR windows are available as an option on SOME arc rated gear. If it doesn't come with it though, the window manufacturer or switchgear manufacturer would have to test/bless it, and we know that will never happen.

Also when it comes to remote racking, again, it sounds nice. Have you actually used the stuff? How many breakers did it destroy, fail to work with, or jam or turn itself into a pretzel before you decided it was not such a great idea?


Finally I'll leave you with this thought. Modern switchgear is quite literally just a breaker mounted on a frame with a stab mechanism. The same breaker is usually available in bolted form. In fact Siemens specifically makes them as bolted breakers and ads the frame if you want "switchgear". The panelboard version just uses the bolted version. The breaker maintenance cycles are now at 8 years and sometimes longer. In fact at least some switchgear made by S&C, Elatimold, Tavrida, and G&W among others is sealed for life and can't be maintained. Maintenance means disposal. You can test it but that's all you can do. 80% of arcing faults occur in the stab mechanism. Said another way, we are decreasing the reliability of the equipment by a factor of 5 by making it in switchgear form. The tradeoff for this massive decrease in reliability is that some repairs (but definitely not all) can be made while energized. And when the breaker itself requires testing on an 8+ year cycle of some gear and has a 20-30 year life that means it gets tested 2-3 times in its entire lifecycle. Some other gear may be both tested and lubricated but that's about it. Now some of the gear intended for the underground market (Elastimold, S&C) is completely sealed and water tight. It is connected to other equipment and/or breakers via elbow connectors which are effectively mini-disconnects. So you could cable everything together with shielded bus bar or cables and disconnect it with a real disconnect (the elbow connector) and avoid the problem with the stabs altogether. This is exactly how the underground group at Duke (largest utility in the U.S.) does it.

So before you automatically worry about arc resistant anything, you may want to ask yourself if switchgear is even the right answer. Can you stand an 8 year maintenance cycle vs. 2-3 for traditional gear? Many times the answer is yes. And the sealed underground stuff doesn't even require cleaning and repairs...it can't even be repaired because in the case of S&C it's a stainless steel box that is filled with SF6 and plug welded shut.

Thank you for your response.

By no means am I an expert of the subject but perhaps the following may clarify my thoughts:

I have to say that I am not a huge fan of maintenance switches as it requires human interaction but they are quite inexpensive and certainly worthwhile if looking for a simple low cost solution. Putting myself in the owner's shoes, I would not want someone in a hurry or maybe just having a bad day and forget and skip turning on the maintenance switch. Or to the reverse, leaving the switch on and ruining any coordination downstream. That said, it is a perfectly legitimate way to mitigate arc flash hazards. The other issue associated with a maintenance switch is having two tags on the equipment - one for the switch being on and one for the switch off. Have heard of complaints that it's confusing to maintenance personnel.

I would agree that through ZSI and differential relaying are good choices but perhaps not viable where ZSI is not used anywhere else (no other zones installed anywhere else for which to interlock with) or the instantaneous settings are not used or are set too high in order to coordinate and the owner is insisting on getting to HRC 2 or less. The optical relaying is not predicated upon reaching a certain fault level but rather sensing a visual aspect of a fault, confirming a concurrent rise in current flow and reacting (very quickly). I am bypassing the TCC of the relay and regardless of what someone does to that relay trip curve or replace it with something else in the future, the optical sensing will still be there. With the confirmation of rise in current flow or (sometimes) sound levels and self test features, one would think false tripping would be to a far lesser extent than when this technology first came out. The only thing I am not clear on is the amount of maintenance needed over time and initial cost for installing optical sensing.

IR windows have (some) sizing options and the viewing angle of your IR camera and distance to target should be considered so that you do not have to angle the camera to shoot the target. Transmittance of the window should be considered so that you understand what you are looking at etc... (good summary here: http://www.datacentir.com/downloads/fil ... indows.pdf). Yes, I would agree that many in house maintenance departments do not get the right training and don't read the literature (not a dig on them, they just have too much to do). They end up just shooting the picture. If nothing else, you would have a baseline from which to compare from measurement to measurement of the exact same IR port to identify a trend over time. Worst case, hire someone that has trained personnel to scan but I would say that they bring great value to the table for those companies that have good preventative or predictive maintenance programs.

From what I see in the literature, remote racking systems have clutches in them to help prevent breaker damage while racking. I cannot find instances mentioned anywhere online about damage done while racking using a remote racking system. Would be interested to know if this is a problem that existed prior to having clutch mechanism or if this still is an issue with modern remote racking equipment.

Warm regards and many thanks for responses.

From a purely academic point of view, IR windows make sense. In actual implementation unless it began to look like Virgin Galactic's spacecraft with bubbles everywhere you can't see into enough places in practice to make IR windows work. Remember that the goal is to find defective connections between conductors, not to simply look at disconnect blades for intance.

From a purely academic point of view, remote racking devices work. However in practice when we tried them (more than one), they either jammed equipment or came off the equipment or damaged equipment, or simply wouldn't turn anything. There doesn't seem to be an ideal solution here yet. In practice the devices do not live up to the sales literature.

From a purely academic point of view, optical relays work. They are fooled by camera flashes which is the purpose of the di/dt CT. In fact this is also how you test them to be sure that they operate. The big issue though is that if I go to a switchgear manufacturer and ask for either an optical sensor install or to validate someone else's install on old equipment, they won't do that. The optical relay vendor is also happy to sell me equipment but won't give me any guarantees on any kind of installation. So the best you can do is install and hope for the best. At least with conventional relaying techniques, I don't have this uphill battle.

ZSI (whether you make your own in a multifunction relay or buy a product called that), as well as differential relaying and all kinds of other relaying methods all work. But they can't be used every time, everywhere. For instance if you go from brand A to brand B, the suppression signal for ZSI may not be compatible. And you have to run extra wiring between breakers in the same switchgear (minor issue) and from the main breaker relay to the next upstream breaker (major installation issue), depending on how big you want a "zone" to be.

As to electricians not using maintenance switches or messing it up, most of the "generic" label formats out there especially the ones that are defaults in the various label making software are terrible and I would never want to use them. They are confusing to the end users because they contain a lot of data but little in the way of explanation. Putting on multiple stickers is even worse. But NEC has gone crazy on requiring signage and labelling to the point where it is probably a violation of the ANSI label standard (no more than 3 warnings per task). But one thing that the current NEC has in it that is very relevant is the rrequirement to label not only equipment over 600 V but pretty much any equipment over 50 V with a warning label that there are exposed conductors present behind a door. This makes it easy for instance to determine whether or not the act of opening a door could in fact initiate an arc flash with a likelihood more likely than just the act of walking by.