A containment dome? Nice concept but sounds expensive. Any thoughts?

http://www.youtube.com/gearcflashinnovation#p/u

Interesting approach! Guess it is better than replacing the gear, but probably not cheap.

I can't get to the link from here, but if this is the new GE arc flash addition, it is an interesting approach. They presented this at the PCIC in September in Anaheim, and had a demo unit on display.

The unit, as I understand it, uses a light detection system with an overcurrent relay to initiate the operation of their device. Instead of a crowbar switch type device, they instigate another arcing fault within the containment. This, coupled with tripping the upstream device through auxillary relays can dramatically reduce the energy seen in the first arc flash event. The unit can be added onto existing gear because it's about the size of a standard MCC segment.

It is a nice concept, that takes a lot of the "heebies" out of a crowbar switch.

i'm having a hard time figuring how starting another arc will reduce the primary arc

the optical detection will shorten time ok for that part
but the second arc ??

The optical detection can be tied to the trip mechanism on an upstream breaker, but this is not necessary for this concept.
If it is tied to an upstream breaker, you must still wait for the breaker to operate. This could be several cycles.

By creating a second arc, they are reducing the system voltage available to drive the primary arc. The current demanded by the second arc, because (they say) it is lower impedence than the first arc, will draw current and energy from the first.

To see the arc vault sales video visit this site:[url="http://www.geindustrial.com/products/static/switchgear/arc_vault.htm"]http://www.geindustrial.com/products/static/switchgear/arc_vault.htm[/url]
Interesting, how the plasma gun redirects the flash to the containment dome. Thats like watching Scotty put up the shields on the Enterprise to protect the ship, just to cool.

Equipment to Reduce Arc Flash Dangers

Sq D has also the ARC FLASH Terminator
http://ecatalog.squared.com/pubs/Electrical%20Distribution/0613BR0302.pdf

More info on SQ D ARC FLASH Terminator

See attached PDF file

WDeanN is correct

Hello, in the interest of disclosure, I work for GE ! Below is a very simplified explanation of how the arc vault works.

The system is comprised of three major subsystems:
1 Arc flash detector such as an arc flash relay
2 Arc Vault device to absorb system energy, similar in purpose to a crowbar
3 Low or medium voltage circuit breaker to stop current flow

The arc flash detection system which could be an arc flash relay, or could be a simple current detection system that compares current measured to a threshold, is used to sense for an arc flash event. It controls the Arc Vault triggering mechanism and a line side circuit breaker.

The Arc Vault simply provides an alternate current path for current via an arc it creates inside of itself. That arc has lower impedance and is more stable than an arc on an equipment bus system or other conductors inside an enclosure. The arc in the Arc Vault will become the only arc that conducts current as the system voltage will drop enough that the other arc will extinguish. Inside the enclosure the arc is kept controlled and separated from the immediate environment.

The circuit breaker will start operating at the same time as the Arc Vault, however it is much slower. The Arc Vault once commanded should extinguish the fault arc within a millisecond or so. The circuit breaker may take 3-5 cycles. However, in the interim the arc is in the Arc Vault not where the fault arc originally started or in the open environment.

The intent of the device is to provide an alternative to a crowbar with the benefits of a crowbar and arc resistant equipment with regard to hazard reduction, but without the concerns created by having maximum bolted fault current flowing through the system. In addition allowing testing of the system in the field. All testing up to now indicates this is, indeed, the case. A system such as this can also be installed in existing gear.

Regarding how the device works; It is as WdeanN stated- The arc in the vault becomes a preferred conduction path and the system cannot sustain two arcs simultaneously. The one with the highest impedance will extinguish. Not sure it is technically correct to state the arc in the vault takes energy or current from the first arc. But it does take it from the system causing available voltage to lower and as the first (fault) arc's current lowers, its impedance increases, quickly extinguishing the original fault arc.

This is a brief summary. Hopefully it helps.

Marcelo,
Glad you made it by. I'm sure Jim would consider you a great addition to the arc flash forum.
I hope you stick around and contribute!

Glad you found the forum Marcelo!
FYI everyone, Marcelo is also one of the IEEE 1584 committee members.

At one of our plants there was once a city ordinance requiring service entrances to be rated at 208V. Needless to say, when a transformer (even as small as 750kVA) brings 13.09kV down to .208kV there is an insanely high arc flash potential. Easy Power calculated between 600 and 1300 cal/cm2. This is both extremely dangerous for maintenance, but also preventing me from continuing the study of the plant. I am looking for a solution for this problem and this one seems to be the best I have found. This leaves me with two questions.

1) Does anyone have any other possible solutions? I have considered attempting to fuse the secondary of the transformer before it enters the switchboard. The fused disconnect would obviously still be dangerous, but it should mitigate the danger in the switchboard.

2) How would I add this vault into Easy Power or even calculate it's reaction myself?

RESPONSE:
Beginning with some assumptions, it is presumed that you do NOT have access to the primary of the transformer - Correct /Not?

If the answer to the above is "Correct" then the Arc Vault is a good approach presuming you have a secondary main breaker, probably rated for 2400A and to interrupt at least 42kA (presuming a 5%Z transformer).

A secondary fuse sould not be of much help because the secondary fault would probably be just less than 42kA and the arcing fault at 208V would be about 10-12kA. The secondary main fuse would probably be 2500A Class L and would take about 6 to 8 sec to react to the arcing fault resulting in a very high incident energy. (Not a Good Idea!)

If the answer to the primary question is that you do have access then other possibilities could exist.

In answer to the question about fixing the high current scenario, on the secondary side of the transformer, there is always a zone on the transformer leads leading to your chosen interrupting device which will always have a high incident energy issue if you intend on tripping on the secondary side. This brings up immediately two options.

Option 1: Trip on the primary side. Since there are no "branches" then tripping on the primary side is just as effective as tripping on the secondary side. You can put bushing mounted CT's on the secondary side of the transformer and then trip on the primary side with a breaker. Possible options include 50/51 protection (instantaneous and time-overcurrent) as well as 87 relaying (instantaneous differential) which also can encompass the breaker itself. Note that using AC for your relaying circuit will limit you to 16 millisecond tripping times. Generally DC powered systems operate in 3-4 milliseconds and at that point you are limited to the interrupting capability of your breaker, which is usually 3 cycles these days for VCB's and even a lot of SF6 CB's. This method is also often called a "virtual breaker" since you have sensing on the secondary side but no actual physical breaker.

Option 2: Cable limiters on the transformer secondary lugs. Its a fuse that is small enough to mount in the transformer enclosure. The characteristics of the fuse are not really that great but it definitely gets you down to more reasonable incident energy ratings. Be aware that these devices are rated to match cable thermal limits so they have ratings like "4/0". You have to ignore this and get the actual TCC's for the devices and design according to the fuse characteristics, not the cable size.

Option 3: Series two secondary devices. An example of this is that I have a particular piece of equipment with a 10 MVA transformer at 7200 V feeding a 1 MVA transformer with a 480 V secondary. The incident energy at the breakers that thankfully in some ways are mounted on the transformer termination cabinet itself are ridiculously high. Downstream of those breakers the incident energy is <4 cal/cm^2. So when I replaced the downstream MCC rather than installing main lugs as in the original design, I added a main breaker unit. This second breaker is set lower yet and achieves <1.2 cal/cm^2 on the MCC bus. Now in practice we ONLY use the breaker in the MCC when doing any sort of work whatsoever. The only time that we'd ever work on the transformer or the wiring going to the MCC's is if we power down the 7200 V side of things. In this case you are again adding a second layer of devices just like option 2 but the device capabilites are considerably better.

In terms of the "arc termination" type idea there two fundamental designs out there. Design one (Square D "arc terminator" style) uses a very high speed contactor to cause a bolted fault as close to the main breaker as possible. This works but puts a lot of stress on the main breaker since it will be a bolted fault every time. The second design that I've seen has resistors placed in series on the main bus with a high speed contactor in parallel with the fuses. Under normal operation the contactor is closed. When the device trips, the contactor opens and increases the system impedance to the point where the available fault current is considerably reduced though it doesn't actually go to zero. This puts ZERO additional stress on the main breaker. There is a third option but I've never seen it actually exercised anywhere. S&C at one time (still?) advertised a true solid state "breaker". Think IGBT's that normally operate in full conduction 24/7. The device can interrupt current in about 1 millisecond or less. By itself, this is very expensive to do and not very efficient. However if you merge it with a requirement to do high speed power factor control (a STATCOM) then it is not that expensive at all. Who would need such a beast? Large mining excavators operate on DC. Current power factor regulation is provided by synchronous motors operating as synchronous condensors that are limited by the electrical speed of the motor itself which has a resonance at around 2 Hz. Using a DSTATCOM would eliminate that resonance frequency and allow for true "1.0" power factor operation rather than having a variable power factor in the dig cycle of the machine. Unfortunately the cost was around $5 MM so this idea got shelved pretty quickly...we'll deal with the performance issue.

Edwin: It is the power company's transformer so primary access will be extremely limited if even possible. As far as the time delay in the fuses being more dangerous, the fuse would be current limiting so it would begin choking the current immediately... meaning 6 seconds would not be as big of a deal, as far as arc flash is concerned. At least this is how I understand fuses, it is very possible I have a misunderstanding.

Paul: Therefore primary breakers and cable limiters are not an option which is what brought me to fusing the secondary side before it enters the Switchboard. Thus leaving the fused disconnect extremely dangerous, but greatly reducing the available incident energy at the switchboard