This week's question has been brought up by several different people so I said I would post it as a Question of the Week and see what the members think.

Here is the scenario:
The main service equipment is a fusible switchboard with three sections
The calculated incident energy is above 40 cal/cm2 due to the next device upstream being a utility primary fuse.

Would you consider replacing fuses on a feeder with only the feeder switch open?

The debate focuses on whether to open the main and take a large outage simply to replace a feeder fuse. Other considerations are that checking for absence of voltage assumes live until tested for dead and live means possibly a very large incident energy.

Opening only the feeder switch to replace the fuse still leaves the line side of the switch energized. If "credit" is taken for the clearing time of the main fuse, the incident energy on the feeder would be below 8 cal/cm2 however, the study models the entire switchboard as one node with one fault clearing time and one short circuit current. One last wrinkle: The maximum available PPE at the site is rated 40 cal/cm2.

Would you consider replacing fuses on a feeder with only the feeder switch open?
Yes, open the feeder switch only
No, open the main and de-energize the entire switchboard
Something else

Your comments are greatly welcome, there has been a lot of discussion about this scenario. What PPE would you consider for this one?

If you have a question you would like to see for the "Question of the Week" feel free to send it to me.

Depends on the construction of the feeder cubicles. If the switchboard feeder fusible switch has a line shield or arc shield over the energized parts while the door is open, then there shouldn't be a safety issue with changing the fuses with only that cubicle switch in the OPEN position. It was quite literally designed for that purpose.

If the switchboard feeder fusible switches didn't have this protection, then you have to apply the full arc flash incident energy of the switchboard to that cubicle and de-energize ahead of the feeder cubicle.

If it were MCCB's and you needed to replace one of those, or check the bolt torque or in some way interact with the energized side of the MCCB, then you'd have to de-energize ahead of the feeder cubicle.

Jim, this a great question - one I have thought about and vacillated on many times.
I recognize there is a potential for many things to go wrong but as Jeff S pointed out, that is what these are made for. But as always, this requires a qualified person that understands the inherent risks of this activity. The next question is - what calorie clothing should be worn? 2 cal would be a minimum since this is a de-energized activity, 4 cal for comfort factor.

Jim, great question. I've had similar questions. The question goes: "if the entire switchboard is labeled danger, how do I verify voltage is off at the switchboard if I have to work on that part of the electrical system?" My suggestion is to check downstream - away from the switchboard. Check for voltage at a point where the hazard is below the 40 calorie level. It's not a perfect answer, but it makes sense to go to a panel or device that's connected directly to the switchboard and check voltage there. Then, once you've verified voltage is off (at this panel or device), go to the switchboard. You're pretty well assured that it's also off. Of course, I still verify the power is off at the switchboard. There are several methods to insure the switchboard is also shut off and vendors do offer remote methods to verify voltage.

On replacing fuses, I'd shut the entire switchboard off. He's why. Inherent in your question is the assumption that the device that's blown the fuse is critical. In fact, it's so critical that if it fails, we can ignore safety and put ourselves and others at risk to get it back up and running. Isn't that really the crux of the decision about replacing the fuses? My question is if this device is so critical, what happens if the utility power goes off? If the device is that important, doesn't the facility have a generator to back up the power and isn't this device on the generator? If they don't, you have your answer to the question of how important it is and if you should shut off the entire switchboard. And if shutting off the switchboard is critical, isn't there a generator that backs that up if power goes off?

We faced ourselves with this question many times in the field whenever we have to replace the Main Disconnect (breaker in my case) for a switchgear/switchboard. We are fully aware the the breaker or fuse have higher interrupting rating (kA) than the utility short circuit. This makes sure that the device will not explode if a fault occurs in its nearest downstream. But unless the utility is de-energized, line side is live and to put a breaker or fuse, we were like mice who got an idea to put a bell around cat's neck but had none to do the job. I called the manufacturer, if we can install a breaker with energized line side (utility) and when it really boils down to arc flash and liability, regardless of the design, even the manufacturers recommend to work in electrically safe condition. So we schedule with utility for short period work and arrange generator (if not available) to run the plant while working on utility main. Once the utility disconnect the fuse on the primary side of their transformer, we check the absence of voltage on line side in PPE & when confirmed, PPE is no longer used. For me, even if someone recommends me that it is safe to work with PPE, I ll not work on energized line side of the main. I would rather arrange and alternate source or other means to avoid it, if I can. The reason is because, upstream of Main breaker or fuse is a blind spot. We have to rely on utility transformer primary fuse to clear the fault, in the event of an arc flash. And those fuses have time delay to avoid nuisance trip. And this could lead to longer arc flash duration. Moreover, utility information comes with one page data and two pages disclaimer. And in the disclaimer, they state that transformer primary fuse is to protect their transformer and customer should not rely on it to protect the service panel or any customer owned equipment.So consider the arc flash hazard associated with the main of switchgear/switchboard and work safe!

NFPA70E requires risk assessment to be a routine activity for a qualified employee.
Some references are:
Article 100 definitions of Qualified Person and Risk Assessment
Article 110.1(G)
Article 110.2(D)(1)(4)

Not necessarily "the employee". Risk assessment standards clearly require a team effort (3 ANSI standards, IEC 61508, 61511, LOPA, etc.). The big thing though with all of these methods is that they are intended for design-time considerations whereas the criteria in the tables in 70E and 1910.269 are a combination of both things that can be analyzed ahead of time such as whether the equipment is designed and installed properly, and other considerations that can only be evaluated at the job site such as whether there are burn marks around the doors. Every risk assessment standard requires evaluation ahead of time by a team of qualified individuals but because of the nature of this case they have to lay out a framework for field evaluation. Similarly, a field service technician cannot properly hope to evaluate for instance if the equipment interrupting rating is adequate. That must be done ahead of time. Thus the risk assessment procedure for electrical work must of necessity be a two step process and requires multiple participants.

I would like to add my two cents to this interesting conversation.

Many engineers are installing two labels in panels with main overcurrent protection devices. One for the line side of the main OCPD and for the load side (bus) of the main OCPD. Normally, the AFIE at the line side of the main OCPD is much higher than at the load side (bus) of the main OCPD.

Since this question is related to Service Equipment (SE), we are going to restrain this discussion to switchboards or switchgears functioning as such.

Many SE are opened frame construction. If you open the adjacent section from the incoming section (where the main OCPD is installed), you can see all the main section because there is no barrier. I think it will be logical to assume that only one label (the line side) for this type of equipment should be affixed. The reason is that nothing will stop the arc flash energies to migrate to the adjacent section. There is no proper segregation.

Other SE's will have barriers between the different sections. I will still place only one label in this situation (the line side label), unless the MANUFACTURER can guarantee in WRITTING that there is PROPER SEGREGATION between the incoming section and the feeder sections.

Before the IEEE 1584.1 (released a couple of years ago) there were no recommendations or suggestions on how to handle this issue.

In this new standard, in section 5.1 Responsibility for Data Verification it is clearly stated that you have to carefully evaluate this situation. It establishes that distribution equipment should be EVALUATED to determine if it is appropriate to include the effect of the main OCPD in the calculations of the AFIE at that panel. Moreover, it encourages the responsible engineer to CONSULT the MANUFACTURER to assure that there is ADEQUATE SEGREGATION of the section housing the main OCPD and the feeder breakers sections.

I know that there is much more to discuss regarding this issue, but eventually it will come to a RISK ANALYSIS performed by the responsible person assigned to electrical safety. This person my determine that, after a documented risk analysis, that it is safe to use the load side (bus) labels in the adjacent sections and the line side label in the incoming section.

It is my opinion that consultant engineers that are performing an arc flash hazard analysis (not an arc flash risk analysis) should not take this additional liability of affixing two labels UNLESS they have a WRITTEN opinion from the equipment manufacturer that there is adequate segregation and that arc flash energies from the incoming section will not migrate to the feeder sections.

I hope that this discussion can discuss since I am sure I have not considered all the info that the engineering community can share.

Have a great day!

There is no specific standard that mentions how many labels to put on each equipment. The only reference I have seen is NFPA 70 (NEC) article 110.16 which mentions that an equipment that are likely to require service in energized condition needs to have an arc flash warning clearly visible to warn qualified persons of electrical arc flash dangers. And ANSI Z535.4 standard for product safety signs and labels mentions the meaning of Caution!, Warning! and Danger!
So using this information, training knowledge, and field experience, I put only one label on each section and the label is based on the line side fault current and incident energy. For incident energy less than 40cal/cm2, warning label is used and for incident energy greater than 40cal/cm2, danger label is used. What is important here is we should not put too many labels that may tire the eye of the qualified workers. Otherwise they won't pay attention and our labels will loose meaning, thus defeating the purpose of study.

ANSI Z535 specifically states no more than 3 warning labels per task.

ANSI Z535.2 and Z525.4 comment on a three panel sign design (signal word / symbol panel / word message). But I don't see where three labels per task is ever mentioned. Can you please quote the chapter/verse of this?

Except for sectionalized switchgear and rare Canadian switchboard designs with isolated main . . . one red Danger arc-flash and shock hazard label per piece of equipment. Label's values are based on worst-case scenario--no confusion added having multiple scenario labels.