Hello AF followers,

We have a 480v/4000 amp, 5 section switchboard with an incoming line section, a main fused disconnect switch section and three distribution sections. The gear is has a open concept ( no substantial barriers). The fused DS does not provide any AF mitigation and the AF hazard is 116 Calories/Dangerous. There is no other protective device upstream as the gear is fed directly from 2500kVA xformer.

The entire gear is labeled Dangerous, how can the client confirm it is de-energized if he needed to install a new breaker?

Our recommendations are:
1. to ask utility to open fuselinks on pole and confirm is off (utility), then perform the work
2. in order to mitigate the AF install a separate device upstream that can limit the AF

Do you agree with our recommendations? What else would you recommend?

Other Questions:

If they were to OPEN the main fused disconnect in the gear, gear rated dangerous, what do they need to wear to do voltage testing confirming that the rest of the gear is off? Can they do it at all safely?

The gear is an open concept with line side (assuming breaker open) is still 116 cal, would you consider it safe working on a gear two sections over from the main disconnect? Could installing a new distribution breaker create and arc on the line side of the open breaker (vibration of buses when bolting the breaker or something) that could injure the worker two section over?

Sorry for too many questions, but I did not find a good thread that covers similar situation yet.

Thank you advance for your input.

Marek

My first thought is to suggest analyzing the incident energy at some non-standard working distances. You may be surprised how much the energy drops at 24" or 36".

I did that and unfortunatelly the room is very limited in space, there is only 48" space between the front of the gear and the wall, plus some parts of the wall have equipment installed on it. I'm not sure a person can get 36" working distance if wall is 48" from the gear, plus what task can you perform at that distance? I'm a tall person but I can't reach stuff 3 feet away,maybe voltage testing with a device that has long probes? But then again the wall is right there and I don't see it. I believe with 36 inch distance I was still in Dangerous hazard level.

Thanks

The arc flash hazard is spherical in nature. Don't just think straight out. If an arcing fault occurs within the gear but upstream of the fuse but you are a few sections "over", the working distance increases substantially over the "standard" working distance which assumes looking at bus bars at the back of the cabinet out to the position of the worker's face/chest area.

Second, along the same lines, what if you are checking for absence of voltage with a meter on a stick? It's not impossible. Look at the Bierer PD-25.

Third, if your meter is insulated properly (insulated against both user-to-phase contact and phase-to-phase contact), can you describe a condition under which measuring voltage can induce an arcing fault? If so then an arc flash hazard exists...and you should consider the likelihood that it can occur.

Paul,

I get you the AF hazard nd distance, but the fuses don't mitigate the AF hazard (if I was to have two labels line and load) and even with 36 inches I was over 40 Cal.

I also get the stick, but the reduced AF hazard label would (still over 40 cal) would only apply for voltage testing.

Third, sorry not sure I follow you here. Are you saying there is not AF hazard when confirming the gear is off? No PPE needed based on the label? Where is that stated?

thanks

This depends on methodology. Check carefully because many software programs give invalid results when you use current limiting fuses. IEEE 1584 has some additional formulas for the specific case of current limiting fuses. It does matter greatly whether the arc initiates on the line or load side of the fuse. On the load side a current limiting fuse may or may not reduce arc flash hazards. Sometimes current limiting fuses increase the arc flash hazard by increasing opening time. Other times they reduce it by limiting the arcing fault current. Note that current limiting circuit breakers also do this and can avoid the slow opening time issue that can occur with fuses, although a fuse is faster than any except a few exotic or very small circuit breakers.



The label is to warn you of a potential hazard. You must mitigate the hazard. In this case because the hazard is so great, you are responsible for creating procedures and methods for mitigating the hazard and in this case, PPE is not an option. The label is there for informational and warning purposes. Somehow anyone performing the work needs to understand how to respond appropriately. A label that simply says "DANGER! KEEP OUT! DO NOT OPERATE" is confusing at best because then it is not clear where and how you can safely perform any required tasks, even if it just refers to an incident energy level with no guidance.



Arc Flash Hazard definition, informational note #1: "An arc flash hazard may exist...provided a person is interacting with the equipment in such a manner that could cause an electric arc." Hence the reason for looking at the task. If the interaction cannot in and of itself cause an electric arc, then the equipment is not likely to pose an arc flash hazard.

Unforunately, IEEE 1584 is no help when it comes to determining this last part. The tables in 70E have included a risk assessment that was performed by the 70E Technical Committee. In some cases they determined that the hazard was minimal (H/RC=0). In others, they have reduced the PPE requirement but do not give explanations as to why this was done.

If you do your own calculations (or you have to because the tables don't apply), then you have 3 choices:
A. Assume that an arc flash hazard is likely irrespective of the task. Essentially, ignore the definition of an arc flash hazard in 70E.
B. Create a list of tasks where the hazard is unlikely based on either a qualitative risk assessment process or by looking at the tasks where the 70E Technical Committee assigned an H/RC of 0.
C. Perform a quantitative risk assessment to determine the likelihood using reliability statistics to determine when an arc flash hazard is acceptably unlikely. I've put an article in the articles section discussing an approach to doing this.

A procedure should be prepared for how to do this by a knowledgable trained Liscenced Professional Engineer, or Certified Electrical Inspector, or knowledgable Electrical Worker. Otherwise , farm this work out to either a testing laboratory or liscenced contactor eaperienced in this work. Here is a possible outline of what can be done.


1- Have the Utility lock and tag out the sevice and install grounds on incoming service transformer either the primary or secondary or both.
2- Add a your own lock and tag on both the grounding device, and on the power companys disconnect, if possible, along with a big wrning sign notice to clearly indicate that the service was disconnected for customer work and must not be reenergized untill cleared by customer representive.
2- It would not be overly cautionous to have the utility disconnect the service temporily and connect all service lines directly to ground.
3- Check for voltage at the terminals of the equipment. This needs to be done with protective gear. Since a 116 cal suit is needed for close approach this is the time to use the lineman type voltage tester with a 10 foot hot stike or what ever is needed to allow using a 40 Callory suit.
4- it is advisable to install grounds on the incoming line for assurance.

Warning: The life you save will be your own, If it is n't your own, regrets can last a lifttime!!!!!!!!!!!!!!!

Installing a large Circuit Breaker or disconnect along with current limiting fuses is advisable.

I would not considered working any where in the described gear until I was asured it was totally deergized.

Have You considered replacing this equipment with a mordern draw-out type switchgear; one accident could cost more then the cost of replacement

Have you considered installing panel mounted instrumentation (volt-ammeter)? Analog would work but if digital is installed, ensure it has battery operation.

Doesn't help. 70E requires you to be able to test it before and after. I've found more nonworking panel meters over time than working ones, both reading "zero" and reading something other than zero.



OK. And that's what the OP asked for help on. The response was to get an outside firm to do it, presumably to transfer the liability to them.



Ever try to get the utility to do anything? It works out great when you have a major customer service representative and you have a very close relationship. Otherwise, they are a monopoly and they typically act like it.



First off, based on current recommendations in 70E, a "116 cal suit" does NOT protect sufficiently. An arc blast is fatal irrespective of the suit. Second, OP also explained why you can't get away from it so the hot stick is out. Third, noncontact voltage testers are not certified as valid for doing testing for absence of voltage below 1 kV. So this whole response is incorrect.



Have you considered the fact that the reliability of draw out gear is much less than bolted gear? This can be proven by the simple fact that in a lot of modern gear, the breaker itself is IDENTICAL to it's bolted cousin. The major difference is in the draw out hardware which adds lots of extra joints, mechanical latches, etc., all prone to additional failure modes. 80% of failures in drawout gear have occurred in the drawout mechanism according to a CIGRE report. And most of the failures that occur in the drawout mechanism result in an arcing fault. If you are going to push for replacement then the obvious choices are either to insert another breaker or fuse in series or to recommend going to something with very low maintenance or no maintenance design. It's hard to beat the current design with a bolted, molded case breaker. Even old data from IEEE 493 from the 1970's confirms this. Powell and ABB for instance are already advertising 8 year maintenance cycles on VCB's. Tavrida, Elastimold, and S&C Vista gear is sealed at the factory and has ZERO maintenance requirements except to periodically exercise and test.

Considered and as PaulEng mentioned, they don't meet 70 E.

What is the arc flash hazard at the fused disconnect? Is it possible to lock open and verify a no voltage condition at the disconnect? If the gear is only fed from this disconnect switch I believe performing a live-dead-live check at this location would be suitable. PaulEng do you agree? We also have 480V switchgear with a 4000amp main breaker directly coupled to a 2500kVA transformer with fused primary disconnect switch. Unlike you, the disconnect switch is fed from one of our 4160 volt breakers. We have adjusted the setpoints on this relay to achieve something <40cal downstream of the 4000amp main breaker but still have extremely high incident energy upstream. We have begun replacing the 4000 amp breaker protective relays. The new relays have a selector switch for normal and maintenance setpoints. Switching to maintenance gives us a Cat 2 downstream of the breaker.

What is the arc flash hazard at the fused disconnect? Is it possible to lock open and verify a no voltage condition at the disconnect? If the gear is only fed from this disconnect switch I believe performing a live-dead-live check at this location would be suitable. PaulEng do you agree? We also have 480V switchgear with a 4000amp main breaker directly coupled to a 2500kVA transformer with fused primary disconnect switch. Unlike you, the disconnect switch is fed from one of our 4160 volt breakers. We have adjusted the setpoints on this relay to achieve something <40cal downstream of the 4000amp main breaker but still have extremely high incident energy upstream. We have begun replacing the 4000 amp breaker protective relays. The new relays have a selector switch for normal and maintenance setpoints. Switching to maintenance gives us a Cat 2 downstream of the breaker.

Is there a way you can verify the voltage has been isolated downstream of the switchboard? Meaning, the qualified electrician opens the Disconnect, and downstream of the distribution sections can such electrician verify "OFF," at the load side of a possible Circuit Breaker (CB)? For instance is there is a 480 panel or MCC with a main CB, can you verify off load side at such breaker with it closed or even a smaller breaker downstream from that main, which would also be closed (and verified energized with DS closed and de-energized immediately after DS is opened). Or, better yet if there is say a smaller 208/120 transformer hard wired to one of the Switchboard sections, that provides power to a Main Panel and do the same at the load side of the smallest CB knowing all CBs are closed, with only the Disconnect transitioned to the open position. Hence, the qualified electrician may be able to verify off at a lower Arc Flash Incident Energy level. Is everything downstream from the Switchboard hard wired connected? If so, than can the operation of a possibly connected motor that stops after the DS is opened be suffice for verification? My take is to verify farthest downstream of the secondary in your case of a 2500KVA XFMR, wherever possible. Now if the switch board is back-fed from an alternate source - indeed all bets are off. Otherwise I agree with the distance of using a hot stick or long lead probes to get within a reasonable amount of arc flash boundary below 40 cal/cm^2 to perform such verification.

FWIW...

Michael, as mentioned in my description the AF hazard level at the fused disconnect, line or load side is 116 Cal. Opening the fused disconnect should theoretically work, but what if one blade does not open and you're voltage testing a 116 Cal load side of that disconnect?! Plus it seems that even operating the fused disconnect is hazardous. Other problem I see is that the gear is an open concept and assuming the fused disconnect is open you still have the potential of 116 calories on the line side of the open disconnect in the section next to you, the disconnect can be easily reached by hand. I understand that the work performed in different section should not affect line side of the main disconnect, but I'm not 100% convinced and I would really only trust someone that completely understands the risk that this gear has before I told him: "yes, the main is open go work on the load side!"
Would you guys?

thanks, m.

Yes, this switchgear feeds several panelboards, switchboards and other loads, so you could potentially do what you're saying. But then you still have to go work on that high hazard level gear with 116 cal on the line side of the fused disconnect. Please read my response to Michael with my concern with that hazard.

Thanks, m.

OK, it seems this is going in circles. Read the last sentence of note #1 under the definition of an arc flash hazard in 70E-2012. Also a similar note is placed in Article 130. The salient point is that if it is properly installed and maintained, equipment is not likely to pose an arc flash hazard. IEEE 1584 provides a method for calculating the incident energy if something happens. What it does not provide is a method of calculating the RISK in that activity. There is an article that I put together (with some data) in the articles section that gives one method for doing this. OR you can rely on the collective experience of the 70E Technical Committee in referencing their statements in 70E. You may also want to consider completing an EEWP as an administrative procedure to minimize the number of times that this risk is being taken to a bare minimum.

What I'm saying is that although there is a hazard, you may find that the risk (ie, hazard plus likelihood) is at a tolerable level. However in order to do so you need to carefully document your basis for this decision because clearly you cannot eliminate the hazard.

Hazards by the way that cannot be eliminated are all over the place. You probably drive to and from work. That's a hazard. There is a possibility that an errant meteor could hit the Earth and instantly kill you (among other natural disasters). All of these are hazards that cannot be practically eliminated. The question to be asked is whether or not the issue at hand (arc flash hazard) or any of the other ones I just mentioned carry a tolerable level of risk.

If your equipment is in very poor shape and not maintained properly, then you definitely do not have a tolerable level of risk. Otherwise, the situation is quite a bit different. Yes, the blade may stick in. BUT right away, consider the chance of an arcing fault occurring if a blade stuck in while opening the door and doing a voltage check. I can't think of a case where the blade was ever loose...the reason it is almost always stuck in is that it welded in place the last time the switch was closed. I'd be more concerned about a wire burned off in the panel or the insulation was overheated and cracked, and that the wire is laying in a position where an arcing fault could happen by the action of opening the door. Once the door is open, depending on whether or not you have the probe tip covers on your multimeter probes and thus can cause a line-to-line arcing fault if you slip with the meter probes determines the chances of accidentally causing an arcing fault while taking a voltage reading.

Again...look at the potential hazards here. Just because if an arcing fault happens the result is fatal doesn't mean that it is likely, and this is stated very clear in 70E.

Paul, I understood what you wrote before and I don't disagree, I was just hoping to read Michael's opinion.
Thanks

Marek,

I missed the part about the disconnect at 116 cals, sorry. Does the disconnect switch have an installed window that can be used to verify the position of the blades? Ours do. Like Paul mentioned, you must determine what you consider an acceptable risk. I would look at some type of rope and pulley system to operate the disconnect switch. Very inexpensive and easy. Some examples of this were shown at the IEEE ESW earlier this year in Daytona. If you have an installed window the blade positions can be verified after the switch is operated and you would be able to see if one of the blades is still engaged. I believe a point can be found further downstream where there risk is acceptable to perform the voltage checks.

Have you considered the fact that the reliability of draw out gear is much less than bolted gear? This can be proven by the simple fact that in a lot of modern gear, the breaker itself is IDENTICAL to it's bolted cousin. The major difference is in the draw out hardware which adds lots of extra joints, mechanical latches, etc., all prone to additional failure modes. 80% of failures in drawout gear have occurred in the drawout mechanism according to a CIGRE report. And most of the failures that occur in the drawout mechanism result in an arcing fault. If you are going to push for replacement then the obvious choices are either to insert another breaker or fuse in series or to recommend going to something with very low maintenance or no maintenance design. It's hard to beat the current design with a bolted, molded case breaker. Even old data from IEEE 493 from the 1970's confirms this. Powell and ABB for instance are already advertising 8 year maintenance cycles on VCB's. Tavrida, Elastimold, and S&C Vista gear is sealed at the factory and has ZERO maintenance requirements except to periodically exercise and test.[/quote]

Strongly disagree with the reliability issue. Most metal-enclosed switchgear design is much more robust (rugged) than the panel type described in this post. Well maintained switchgear is extremely reliable.