We recently got arc flash calculations on and industrial facility and the numbers are quite high.

The switchgears are 440V AC and most of them built in 1975-1985. All metal clad with doors in the front.

The highest IE numbers are from 60 to 100 cal/cm², and that is with calculations using 2s as maximum due to the IEEE thing. The actual trip time is between 2-8 seconds for the minimum short circuits and the high IE is due to long trip time. Which means that the actual IE could be a lot higher. The switchgears only has overload protection, not short circuit.

There are plans to install arc flash guards, but there will be some time until it is ready. Until then we have to live with these high values. The old, analogue overload protection has limited adjustment options, so those who did the calculations strongly recommended installing arc flash guards. Adjusting the current relays wouldn't help a lot.

We got ATPV for the entire body with ratings a little over 40 cal/cm². On switchgear where the IE is over 25, we are instructed to try to disconnect before work is done when doors are open. Over 40 and we are instructed to reduce IE before opening doors, which means disconnection. Not always easy to achieve, but at least we got something to argue with.

The main question is how we treat work on these switchgears with all doors closed. I think I saw "Arc flash tested" mentioned in the documentation, but not how it was done. It also mentioned a pretty high short circuit current, which could mean they only tested at currents high enough to make the overload protection relay trip fairly quick. As mentioned, it's the lower short circuit currents which lead to the highest IE.

We sometimes have to switch off equipment for maintenance or pull out main incomers by turning screws. These actions are done with all doors closed. Let's say we were using the handle to screw the main incomer back in to the cubicle (on this old switchgear), wearing APTV of some rating and an arc flash with 80-100 cal/cm² IE happened inside the cubicle. The door hinges are spot welded and no where near as well built as our newer switchgear, where every part of the door is bolted.

What could happen if we used PPE2? PPE4? How do we evaluate the risk of the operation and establish sufficient protective measures to secure the health of the operators? I struggle with making a good risk assessment and finding good enough measures without having to disconnect the supply. Some in our company say that the switchgear can take 25 cal/cm² and operating any board, arc flash tested or not, up to 25 cal can theoretically be done without needing any ATPV (not that we do this). For the 60-100 cal cases, we are told to wear PPE4 when operating, but I don't feel comfortable with it. The chances of a 60-100+ cal incident happening are of course slim, but I don't want to find out what happens if one day we are very unfortunate.

Just had a study done? Our old switchgear was swapped out and our newer and has energy reduction switches on the main and also all of our breakers over 1200 amps which is code now. You might be able to use a protective relay such as https://selinc.com/solutions/arc-flash-solutions/. I don't know the make or type of main breakers you have, but it you can shut trip them quicker with a protective relay, then that would be a solution. Most switchgear never get down to the level of live work, so disconnection is always your first choice. Hospital or other safety for the reason you cannot disconnect?

A few questions. Is the voltage 440 or 480? Just wondered. Also, have you gotten good information on the upstream transformer? Impedance values as well as primary fusing. Consider looking at arc flash reduction devices by Littelfuse as well as other manufactures. This might be a better option. Also, look at remote operators for removing and connecting LVPCB.

I don't believe you can distinguish between the doors being opened and the doors being closed unless the switchgear were rated as arc-flash resistant. Doors on switchgear do not provide any level of protection against arc flash unless it's proven that they can contain the blast.
I think your best bet is some sort of remote operator. I'm sure there's a lot of companies out there, but one I've used before is CBS ArcSafe. They make remote operators for all kinds of vintage of switchgear, panelboards, etc. Note I'm not affiliated with them in anyway, so hopefully this isn't viewed as an advertisement..

There is specific arc flash PPE all the way to 140

Only if you want to look like Ralphie's little brother (a Christmas Story).

Mike

This appears to be the real problem which is using the 2 sec maximum allowed cutoff time. Why was this used? Was it because to actual available fault current on the primary side, the primary side protective device and utility transformer info was not obtained?

It seems that an infinite bus was used for the fault current and then 2 sec used for the time factor for this equipment which seems be the service entrance equipment. This will usually always result in unrealistic incident energy values.

Do a proper study and you will find the IE values are probably much lower.

We have also done a study on older (1994 ish) unit substations, right on the tail end of 2500kVA 13.8-480xfmrs. There is a fused main breaker, followed by about six lower breaker cubicles with top protection/metering/control cubicles. The gear is not arc rated, so (per IEEE guidelines) I am marking all of the breaker cubicles and backside connection doors with the really high (~65cal/cm2) ratings. Now - I DID deviate from the recommended 24" working distance as this gear is really deep and the identified tasks (racking breakers, opening doors for thermography, etc...) would not bring you within 36". I am recommending they put in arc sensing relays, and consider a buss-differential scheme (but it has to trip the 13.8 breaker to the unit sub, not the 480 main).
My point is, review the tasks and establish the working distance, then train on it. Consider using signage to "Refer To Report" for other working distances (if using a multimeter on the 480V bus is allowed, for instance) then include multiple working distances in the table in the report - or posted in the switchgear room.
The working distance is very important when calculating incident energy, and it'd be advantageous to the worker and the electrical safety program management to understand the relationship and consider it as part of the risks for the identified tasks.

First of all, thank you for all the replies!

We are working on ways to reduce IE or operate remotely, but everything won't be implemented for quite some time and we need to make risk assessments on short term operations. If that is to stop all operations until it's done, that's fine.


Correct. We've been asking for it for years, but studies have now been done by a reputable company. Our company has also done their own calculations and the results are somewhat the same. We have installed arc sensing relays in some of the switchgear, but all of them won't be ready for some time.

Petroleum fascility and loss of production is the reason they don't want to disconnect.


440V AC 60 Hz. The company that did the calculations has been given all data on the facility.


Thanks. Yeah, with higher levels I don't see how we can risk assess any work on these old switchgear. Remote operation is also something we have discussed and may be introduced as a temporary measurement.


The IEC standard says it uses 2 seconds since that is a reasonable time in which they expect people to get away from the situation. The incident energy is also at a distance of 61cm since that is where one typically works. When the IE is 100 61cm from the short circuit, I don't want to know what it is in the middle of it. Nor what could happen to doors and the entire switchgear. Turning away from the flash doesn't help if projectiles come flying after you.

This is a proper study done by a reputable company.


Distance is certainly an important factor. When the energy can be over 100 cal 61cm from the flash, it's a lot higher up close. We rarely get that close to the bus, but projectiles is also a concern.

One of the previous responders mentioned that they do not use the typical working distances in the IEEE 1584 standard. If you know the planned work tasks keep the workers at specific distances from the arc then using those distances would be appropriate. There is no mandate to use the distances shown as typical in the IEEE guide. The proper distance to use is what represents how the task is carried out and how far the worker is from the arc.

If you are adding AF relays or any other relay to control the CB via its shunt trip make sure you include the extra time the shunt trip probably adds. Generally an integral trip unit with instantaneous, specially if the trip unit has a protective algorithm as is typically used for the RMS function, will be faster than if the CB is controlled from an external relay. If you CB still have analog trips I would recommend updating them to modern electronic trips.

Though it may seem common sense that sheet metal doors offer some degree of protection, the fact is they are not tested for that unless it is arc resistant switchgear tested per ANSI C37.20.7 In fact if there is considerable pressure build up in untested equipment parts like doors may be ejected and increase the hazard.