I have to label some 208/120 circuit breaker panelboards.

First, is the limit for the transformer capacity below which you don't have to perform analysis 125kva? All my xfrms are less than that.

This size limitation is in the NFPA code, did it come out of the IEEE code?

Is there a PPE level that goes with this guideline.

For something a little more indepth, the consensus seems to be that when the main breaker in part of the panelboard, that we can not use the downstream clearing time, as an arcflash on the bus could ionized over to become a arc on the line side of the breaker. Let me ask if this is the consensus? If it is, is the correct method then to assume a 3 phase fault on the line side of the main breaker, and caculate the arc current, then using the transformer ratio to calculate primary xfrm current, and then find primary overcurrent protection device time to clear?

First, is the limit for the transformer capacity below which you don't have to perform analysis 125kva?

Is there a PPE level that goes with this guideline.

Haze10,

I remember seeing that you do not have to consider arc flash for a source smaller than 125 kVA. I would think that no PPE level would be required. However I would always recommend that an electrician wear FR clothing and safety glasses for any energized work.

As far as an arc ionizing to the main bus, I think we could do "what ifs" forever. With the main an integral part of the panel, my concern would not be the ionized arc, but since it is in tight quarters, suppose the employee actually causes an arc on the line side. I would think that you would have to use the level and clearing time of the "box" being worked, and in this case, it is possible to actually have the fault occur on the line side of the main (assuming a large enough source to be concerned about to start with).

Alan

Hi Everyone,

Here is the exact IEEE 1584 wording about the transformer size. It is from Section 4.2 paragraph 5 about the model.

".... Equipment below 240V need not be considered unless it involves at least one 125 kVA or larger low impedance transformer in its immediate power supply."

I believe this refers to the actual calculations. The thought was the fault current would be low downstream of the small transformer. I have conducted short circuit studies the same way for years cutting them off at small transformers that can not produce currents greater than 10 kA which is the lowest rated device. I think a similar philosophy was used here.

There is not a "default" PPE in this case so it would still be a good idea to address safety / PPE. You would likely find that the low fault currents yield a very low PPE category. However the haunting question of a fault current too low making the O/C device operation too long is always lurking out there.

Jim,

Do we agree that the energy level should be calculated worst case for the box being worked given the integral main and of course assuming that the source is large enough?

My case is that there is a 208v panel on the secondary of 500kVA transformer without a main breaker in the panel. I have got dangerous hazard on this panel. What is the best way(cost effective) to mitigate these kind of hazard? I think it is very common in industrial plants. Any suggestion?

Noah,

That's a tough one.
I have a link to a video below where they captured an arc flash on the secondary of pole mounted transformers. I don't know the size of the transformers involved.
[url="http://www.brainfiller.com/documents/FromKGWNews.mpeg"]http://www.brainfiller.com/documents/FromKGWNews.mpeg[/url]

500 kVA is a good size at 208V and can allow quite a bit of short circuit current = 25 to 30 kA for an impedance around 5%.

The correct answer to your situation is don't work it live. However, that is not always possible or practical as it sounds like your case. Some have been using current transformers on the secondary to sense and trip a primary device if one is available. Short of that, adding a remote secondary main such as current limiting fuses is an option although $$$$. The IEEE 1584 Standard suggests you might be able to cut off the clearing time at 2 seconds assuming a person can jump away and not be trapped. Unfortunately there aren't too many options. Maybe some in the forum has a more creative solution??

I believe Alan's posting is correct in that it should be labeled for worst case.

Thanks for all the responses. I've been wanting to buy the IEEE guideline but even with my member discount its over $500. That's somewhat unreasonable according to my manager.

I'll probably calculate the EI for the panelboards, and then use a PPE level of 1 as a minimum. All our electricians are wearing Level 1 shirts and Level 2 pants anyway as their normal uniform. If I assume the line side of the main is the root of the arc, then I am curious to see what the tripping time will be on the primary Over current device.

Noah,

I would concur with the above recommendations. Installing current limiting fuses on feeder to the secondary panelboard would have the greatest reduction of EI. You can also look at installing CL fuses on the xfrm primary or setting the primary breaker to provide the fastest possible response. Then assume the panelboard is in fault, calculate arcing current, and transpose this to primary current, and see how fast the overcurrent device will trip. Some transformer primary devices can be set to follow an Extremely Inverse Time Current Curve which could react quite fast to a secondary fault.

Other option is to work on that panel de-energized.

Thanks for the opinions! Most of time the arc fault is not in the current limiting range of Fuse. So the current limiting fuse still acts as time-delay function. In that case, it still can't mitigate incident energy. Can any one provide more information on current transformer sensing in order to trip the primary breaker faster?

Noah,

If the primary breaker has overcurrent relays installed for control you should be able to adjust your pickup level, and unless it is electromechanical, the TCC as well. If the relay has instantaneous attachments, you can also take a look at them. It may also have multi-ratio current transformers to give you more flexibility. If you cannot limit the current, time is your only other controllable variable for IE (aside from distance).

You asked about sensing.....can you be more specific with your question?

And I meant to add source size. You say that it is 208 v and the source is less than 125 kVA. I would refer back to IEEE 1584.

Hope it is some help.
Alan

You'd have to give more specifics on the primary device. I don't know what you have, but if you have primary fuses then you should be able to size them to get them to respond to an Int Fault. For a fuse or breaker relay, you should be able to size the fuse or set the INT of the relay to clear in two cycles on the primary. This is especially true if you don't have super large motors trying to start across the line. If you can get even a reasonable clearing time on the primary, you'd be surprised how low the EI can be.

Hi, please refer to Brainfiller's response for this thread. I think his suggestion is to add current transformer on the secondary and send trip signal to the primary protective device to clear the fault very quickly. However, I want to know more details about that if possible.

For CTs on the secondary to work, you need several things in place:
1) This will only work on a circuit breaker that can accept an external trip. Any drawout breaker should be able to handle this. Molded case breakers generally have to have this feature built in when purchased. Fuses have no means to trip remotely and won't work at all.
2) CT's are an analog device. They put out a 0-5Amp AC current proportional to their primary ratio. A 500/5 CT would be used on up to a 500A feeder, and will output a calibrated 0-5A output, 200A in feeder is 2.0 amp output of CT, 356A in feeder is 3.56A output CT, etc. This analog signal is not able to trip the primary breaker. You need a logic device, often called a 'relay' or 'controller' that will allow you to adjust how you want the output from the CT manipulated (what fuse curve you want it to follow) and then to output a discrete contact from this relay to the primary breaker trip input.

This is the identical way that a switchgear breaker on the primary works, only the CT ratio is probably lower as primary current is lower than secondary. The switchgear breaker has the relay either built in or mounted in its cabinet. What you would be doing is just building a parallel tripping device on the secondary, and tying it into the primary breaker.

However, I still say that you would have to have a very unusual installation if you can not set the existing primary breaker to something that would provide reasonable secondary IE. If you have a draw out breaker, you can change the relay on it, to one that gives you INT settings and even a 'maint switch' that changes to superfact response just to limit IE. Changing the relay on the existing switchgear breaker could be substantially cheaper than adding secondary parallel trip, especially if there is any distance between the transformer and its primary breaker.

Thanks haze10!Just one more case want to discuss with you, I have a shawmut A4BQ-750A fuse feeds to a 500kVA (480V/208V, 6.0%) transformer. The arcing fault current on the secondary is only 6.5kA. I got dangerous level on the secondary bus (there is no main on the secondary). How to mitigate the IE level? I really have no solution here. Thanks in advance!

Anyone tried using semiconductor fuse to mitigate the arc fault level? Any advantages?

Noah,
You are stuck on this one. I checked the shawmut website, but they list a A4BY 750 not Q. Trip time on primary for your secondary arc current is 40 seconds. You could change tot he 600A unit and get that to 20 secs. But this still doesn't get you low enough. Changing to the A4J-600 gets you under 2 secs, and based on a quick check would result in IE levels of 38 and 23 at the 85% and 100% arc current. I came up with a arcing current of 7.2KA and 6.2KA on secondary at 100 and 85%. This would be Level 4 but at least you could work on it.

A 600A primary fuse would probably be fine if you are not starting large motors.

You could also look at changing the primary fuses to a programmable molded case breaker, Square D makes a nice one, but I would have to check if it is any faster. A breaker like that won't be cheap.

It might be interesting to see what a larger transformer or one with a low impedance would do. Having higher fault current might actually help. Other option would be secondary fuse/breaker, or primary breaker with optional secondary trip relay and ct's.

Thank you Haze10, really appreciate it!

Noah,

This is a link to Square D Micrologic relay.

http://www.squared.com/us/products/circ ... ameset.htm

According to the literature the units equipped with Int setting can be set as low as 1.5x the longtime setting. So if you set the longtime at 750 amps, the int could be set to trip at 1075 Amp. It can also be set higher, that's the lowest. This a lot less than your arcing current and would provide a trip in 2 cycles. I'm guessing that the IE would be less than 1.

The only thing you got going for you to put the breaker on the primary rather than the secondary is that it would be a little cheaper buying an 800 amp frame rather than a 1600 amp frame.

Thanks Haze10, I agree that this is the best solution for the case!

Easy solution.... simply label this as Cat 2. End of business! Don't get lost in numbers or rules. Bottom line is make it safe. Cat 2 for this application maybe slightly over kill but always engineer back from the witness stand.