Arc Flash Resistant Switchgear and Incident Energy Levels

[stevesummers]

Hi All,

so i am doing an arc flash study to IEEE 1584 using ETAP, and have been working through the various calculations etc.. Most of the answers are fine, but a few results are high and i have various mitigation measures like ZSI and arc flash sensors to consider. I am at early concept stage so we don't have any specific switchegar in mind, and i am also using typical fault clearnace times and switchgear dimensions so i know things are a little imprecise.

But i started to think about the actual switchgear design itself. I know that in addition to the various mitigation measures, we can also specify arc-proof switchgear, which is rated to withstand an arc blast, and direct some/most of the arc energy to a vent duct or similar.

So my question is from a power system studies point of view, does specifying normal or arc-proof switchgear actually make any difference to the study results? I.e. if i am getting 50cal/cm^2 with a generic switchgear would using a similar arc resisitant switchgear actually reduce that incident energy level down to help protect the individual / reduce PPE, or is it more to protect the physical switchgear itself, and doesn't cause any reduction to the energy levels if someone was undertaking mainteance work?

I know there are vairous levels of arc proof design, i was just wondering about the basic concept...

 

[ScottyUK]

Here are a few thoughts from a UK private network owner / operator's perspective. We've installed a fair amount of internally arc classified equipment in the past ten years or so.

The IEC TR61641 technical report - it hasn't quite made it to the status of 'standard' yet - has quite a number of clauses which allow you to specify the magnitude and duration of fault, and which faces of the board are protected - front in all cases, lateral in many, rear in some.

The results of the (non-mandatory in the UK) arc flash study performed for us by a consultant advised us that arc-contained boards present a Cat 0 arc flash hazard within the 'safe' zone, which in our case covers the front, lateral and rear positions with the exception of the incomers which present a rear horizontal blast hazard; all other vents are oriented vertically upward, directing the blast away from personnel.

Note that in some board designs, including ours, arc by-products will still be ejected into the switch house and will be a hot, toxic and very dense cloud which would engulf anyone in the area. It's a far more survivable hazard than a direct arc blast, but you'd still be in a world of trouble breathing that stuff in. We didn't have the option to vent externally in our case because of limitations of the pre-existing building structure, but that should be the preferred solution if the opportunity exists.

You ask about safety or personnel v's safety of the board. An arcing fault will inevitably destroy part of the board: the question to consider is just how much extra you're prepared to pay to limit the propagation of the fault to other areas. Typical options are: no limitation (i.e. free to propagate); limited to an individual pillar; or limited to an individual cell. In the event of a fault then the more localised the fault containment, the greater the chance of restoring at least some of the board to service as qwell as reducing the amount of hardware to replace.

We performed some additional testing on our board with some of the feeder / starter compartment doors open, initiating an arc within a sealed compartment and observing breakthrough from the faulted compartment into the adjacent open compartment below. This test is a lot more onerous than the requirements of the IEC techncial report which 'only' requires arc containment be achieved with all covers closed and bolted up tight for the test.

In my opinion the IEC document doesn't fully assess the real-world hazard to personnel working on a board which will likely have a door open in order to access a fixed-pattern compartment or to insert/withdraw a bucket, and that's what we wanted to evaluate. We didn't get an absolutely perfect result but the results suggest that a worker on a board of thsi design would have a high probability of surviving an major event within an adjacent compartment. I am not convinced that this has been adequately addressed by the major manufacturers in their current designs, nor is it considered in the IEC document so far, and I don't think many of the big-name designs would offer protection from a fault of this nature.

 

[stevesummers]

Hi Scotty - thaks for your thoughts.

yes the IEC world seems to be lagging behind the ANSI world a bit!

It is fairly clear that arc-proof designs will help contain the fault and prevent propogation through the switchboard, and i think this is reasonably established on how it works. I have actually worked /designed a few installations that have needed arc flash ducting to divert it away from the switchgear and had similar problems of how to vent it to safe area. Interestingly i believe that there are some solutions out there that can contain the energy safely without needing external ductwork.

I suppose what i am trying to work out is if there is any way of qualifying how an arc resisitant design would effect the required PPE levels. So in terms of a simple theoeretic example.

1) Client mandates that the maximum incident energy levels shall not exceed 40cal/cm2
2) We do the study, and determine that using standard designs we have an actual incident energy level of 45cal/cm2, which would not be acceptable.
3) We need to find a way to adjust the design to bring the incident energy levels down to <40cal/cm2

So, i know we can use the various mitigation techniques like ZSI, dual setting relays (maintenance settings), optical flash sensors to reduce the Fault Clearing Time and the subsequent energy levels and this would give me a solution I am happy with.

But my thought is more that suppose if the Client has already specified that the switchgear has a high level of arc-proofing. Would the design of the switchgear inherently reduce the incident energy level down to say <40cal/cm2, where we could then mandate PPE level 4 and avoid the need for ZSI or dual setting relays etc..?

With a bit of common sense, it would seem to be the case, as a portion of the blast would be directed elsewhere rather than at the operator, and so the blast energy would be lessened, but i cannot seem to find any documents or standards that discuss this point.

am talking slightly theoretical here, so i know that in practice this might not be the best soluton

 

[davidbeach]

I think you had it more correct in the thread subject - arc resistant - than in the body where you refer to it as arc proof. Perhaps it is just semantics, but none of it is truly arc proof, though I suppose GIS might come close, but arc resistant gear can make it much safer if all of the necessary steps are taken. It seems to me that referring to it as arc proof could lead to complacency and failure to keep it all buttoned up, every bolt always tight.

 

[cranky108]

If you look at the PPE required for 40 cal. you can see why it is desirable to keep it below that level. And some of this might depend if you are under NESC or NEC standards.
We use arc-resistant gear, and we still classify some activities as requiring higher PPE levels.

The other issue we started using is faster clearing protection to reduce potential arc-flash concerns.

Other options are current limiting equipment, like reactors, current limiting fuses, or higher impedance transformers.
Example, we specify that transformer low side let through current to be less than 10,000 amps, which results in higher impedance transformers, and higher losses, but reduces the cost of distribution equipment.

We also specify bus differentials for new switchgear to reduce protection speeds. However, we don't use arc-flash protection yet, because we are still waiting to hear other companies problems before we go there.