Hello guys, I’m new here so let me introduce myself a bit;

I’m a Dutch student working on arc flash mitigation for ships (and offshore units). I would like to apologise in advance for my English since this is not my first language and I sometimes have difficulty with this.

I have a question about ‘standard practises’ when working on de-energised systems.

When working on a (690V AC) compartment the steps you take are;
Whilst wearing the appropriate PPE and using the appropriate tools you de-energise the circuit and rack out the breaker.
Take measurements to check if the compartment is really de-energised.
Lock the breaker in ranked out position so that there is no way someone can switch the power back on in the compartment.
Check if the compartment is still de-energised.

Now assuming that the steps I describe above are correct, my question is: can the engineer take off the PPE while doing his work (because it is de-energised and there is no way the power can come back on) or must he keep the PPE on during his work?

Hello and welcome to the Forum. You will find there are many knowledgeable people here and I hope you will continue to visit and participate in the forum.

Typically, once you have verified that the hazard is no longer present (locked out/tagged out, tested de-energized) there is no hazard, therefore PPE not needed. Well, the PPE for the arc flash hazard but if you are say drilling in the cabinet, then eye protection is needed or other PPE appropriate for the work.

Once the flash and electrical hazards have been removed, locked out and verified de-energized the arc flash and electrical safety can be removed.

Both shock and arc flash hazards require PPE when it is not possible or practical to eliminate the hazard, BUT this is meant in a limited sense, too. For instance with shock the definition is that contact will not occur inadvertently. Recessed, "touch safe" contacts do not prevent someone from using a screwdriver or meter probes to make contact, but accidental contact is prevented. Similarly, just walking by equipment or operating equipment which is operating normally, although there is very small risk of an arcing fault (around 1 in a million per year), should not be a concern. As rack mounted equipment is 5 times more likely to develop an arcing fault and all of the additional information faults involve the racking mechanism, clearly the act of racking the breaker definitely requires arc flash PPE. This is the primary reason that if a disconnect is available or it is possible to lock the breaker, that is preferred. Some people say that breakers themselves are not allowed as legal lockout points but I can point to lots of counterexamples including wording in regulations. It comes down to arguments about design. If we ever go down that road, we will have to explain the difference.

I was primarily concerned about the fact of when an electrical circuit is ‘officially’ de-energized and ‘safe’ to work on from an arc flash and shock hazard point of view. The advantage I have in my studies is that it applies to a vessel that requires high levels off redundancy. Basically there is ‘no excuse’ to work on live part besides taking measurements and such.

PaulEngr wrote:

Similarly, just walking by equipment or operating equipment which is operating normally, although there is very small risk of an arcing fault (around 1 in a million per year), should not be a concern.

Hmm I thought that switching loads on and off locally presented a risk as well, that’s why it is preferred to do remote switching (from a safety point of view). Or is local switching not categorised operating equipment which is operating normally?

PaulEngr wrote:

As rack mounted equipment is 5 times more likely to develop an arcing fault and all of the additional information faults involve the racking mechanism, clearly the act of racking the breaker definitely requires arc flash PPE. This is the primary reason that if a disconnect is available or it is possible to lock the breaker, that is preferred

Again interesting, I thought you should always rack out a breaker if you have the possibility. For example;
Switch off load (remotely).
Go to compartment and visually inspect the situation (is it switched off).
Rack out the breaker yourself or by using a remote racking device when potential energy is high.
Lock the breaker in racked out position.

And for all the steps above use the appropriate PPE (except for switching the load off remotely).

EDIT:

How does one actually verify if a voltmeter is working correctly?

In the NFPA 70E 2012 110.4A (5) it states that at voltages at 50V or above you must check if the voltage meter is working correctly before and after testing if the de-energized circuit is really de-energized.

I found a site with the following description:


What catches my eye is the word ‘similar’, because this means opening another compartment to take measurements on live parts with risk off creating an arc flash, right? Because I would assume that you should write another risk assessment for the ‘work’ you do in the live compartment. I used the word should on purpose because in my mind this would be something that in the field gets skipped (and I base that on no experience what so ever).

As of the 2012 edition of 70E, this is one of the more confusing issues. The standard talks about "interacting with equipment in such a way as to cause an arc flash" and mentions in another note that 600 V class equipment that is "operating normally" is not likely to pose an arc flash hazard. Similarly the "hazard/risk asseessment tables" have "0's" conspiciously placed for operating breakers which would otherwise be rated with much higher arc flash hazards but offers no explanation as to how these hazard/risk categories were derived. If you do your own arc flash study then, you are on your own to determine when a particular task does not require PPE and when it does.

The equipment is not inherently dangerous if it is in good working condition. Lots of actual operating data collected by IEEE (standard 493), EPRI, some military databases, and others have continued to conclude that properly maintained equipment has a fairly low failure rate. Circuit breaker arcing faults are more common than other types of equipment, and yet still have arcing failure rates that are less than the national average for all arc flash related fatalities.

As of right now you are on your own though to determine when a particular task is considered "interacting in such a way as to cause an arc flash" unless you simply follow the tables, which again are not much help. However the 2015 1st and 2nd drafts contain a table which definitely indicates that operating breakers and disconnects locally which are operating normally would not require PPE. Hopefully it will survive to the final draft at which point we can use the table to apply to simplify the risk assessment.

Quote:

Again interesting, I thought you should always rack out a breaker if you have the possibility. For example;

Quote:

Quote:

Racking breakers has a 5 fold increase in causing an arc flash. This is what the data tells us. Visual inspection is required as per 70E (but not OSHA) if it is possible to do so but not required. OSHA Subchapters R and S are very vague on what is an acceptable lockout mechanism in general other than saying that it can't be a control device and that plug-and-cord devices are acceptable, and that anything that can be locked out under OSHA 1910.147 is acceptable. This is where there is a much better definition of an acceptable lockout device:

"Energy isolating device. A mechanical device that physically prevents the transmission or release of energy, including but not limited to the following: A manually operated electrical circuit breaker; a disconnect switch; a manually operated switch by which the conductors of a circuit can be disconnected from all ungrounded supply conductors, and, in addition, no pole can be operated independently; a line valve; a block; and any similar device used to block or isolate energy. Push buttons, selector switches and other control circuit type devices are not energy isolating devices."

Clearly both racking out AND operating breakers are acceptable under 1910.147, and these are also acceptable by reference in Subchapters R and S. Although Subchapter J (1910.147) is the general or "mechanical" standard for lockout, this is the best, most comprehensive, definition I can find anywhere in an OSHA regulation.

Quote:

How does one actually verify if a voltmeter is working correctly?

The issue is that you want to be sure that when you do the test that the meter did not fail immediately before doing the test and thus provide invalid results. Ideally the test should be with similar test conditions and voltages because most multimeters have a fixed set of circuits that works only within a certain voltage range, and then a set of voltage dividers to sense voltage at ranges other than the internal voltage range. The voltage dividers can be manually or automatically selected but the key point is that somehow, there are physically different circuits used for different voltages, and just because the meter works at 12 volts does not mean it will work correctly at 600 volts.

Quote:

This is mostly correct, and the concern is what I just outlined. First, at some point there is going to be some minimal exposure of some kind. The actual test itself is going to be done on a presumed energized bus until the bus is tested to prove that it is dead. The same amount of shock protection and arc flash protection is required on a live as a dead bus. Personally I prefer to do the first test on the actual bus in the actual locations where the test for absence of voltage is going to take place for a couple reasons. First is because it is the best test for the meter. Second, frequently corrosion or other issues may result in false readings and thus it is best to test for these ahead of time. More than once I've had to scrape clean an area to provide a good, provable ground.

The final, "after" test is somewhat more difficult to do. If you are exposing the actual switching mechanism and have access to the "line" side of the disconnecting equipment, then this provides an avenue which is just as good as the test point for "dead". This would be more difficult if there are shutters in the way. Second choice would be an adjacent piece of equipment which is still energized but this brings in additional risks as mentioned, and gets even more difficult to do for instance if the act of powering off the equipment truly cuts off the only available source of power. Third choice is to use whatever you've got and it might just be a 120 V receptacle or the battery in the truck. These are terrible because all they accomplish is to prove that the meter will twitch when it sees "something" but not necessarily a "similar" source of power.

No matter what you do, testing for absence of voltage is a task which requires exposure to energized components. The gloves, arc flash suits, etc., are already out when doing the test even on equipment which is supposed to be "dead" so the risk is already minimal. By itself, voltage testing is fairly low risk as the tester is merely making light contact with possibly energized equipment.

I only know of two cases where a voltage test of any kind turned into an arc flash. The first case was around 2001 in North Carolina where contractors had accidentally wired a 2300:480 V transformer "backwards". This was also prior to requiring "category" rated meters. On contact with the presumed energized 480 V "secondaries", the meter exploded resulting in an injury. The second case occurred in New Jersey about the same time and involved an electrician connecting a "Wiggie" to a 2300 V line rather than the intended 480 V wire, resulting in the wiggie "exploding" and an injury which is very typical for these devices (and a reason the original design is no longer sold). In both cases the total work force for each of the companies was around 10,000 and records at each site were known to be fairly good for about 10 years, so the sum total information available to me at this point is perhaps 20,000 man-years. The injuries in both cases were relatively minor (not life threatening). Both cases would have been eliminated through the use of a properly rated meter, and both companies today have policies in place which specify Category III-1000 V rated meters which would have resulted in no injuries.

Testing injuries do occur. In another case around 1998 in McIntyre, Georgia, an electrician went into an enclosed substation powering several subconducting magnets to take current readings. What happened next is not clear except that the meter was destroyed beyond recognition and the electrician spent 6 months in an induced coma in a burn unit. The electrician has no recollection to this day of any of the events of the day or anything else until waking up months later at the burn unit. This was at a 1500 person site in a specialty chemical company with around 10,000 employees with a very proactive safety organization.

I can give you a third case. In the mid-1990's I was doing some work on a motor test stand. The documentation I had for the piece of equipment was a bit vague in a few areas, so I called the manufacturer to make certain of what voltage was where. The test stand consisted of seferal transformers and one large 3-phase variac. It was fed with 6600V and had a selector to connect several taps to select the test voltage for the motor (from 6600 down to 600). Based on information from the manufacturer, I was led to believe that the 3 variac phase voltage was set by the taps, so if i set the tap to 600, the maximum voltage there would not exceed 600.

That was not the case. The voltage was 6600V and my Fluke 77 meter exploded and shut down the entire facility when I attempted to measure the voltage across the phases. Luckily there were no injuries.

I saw quite a few;

IEEE1584 Annex C Example 12, 17, 18, 20, 31, 33 and 39, most are pretty vague about what really happened but this is what I based my knowledge on ‘the dangers of using meters improperly’ on.

And then there is one of my ‘favorites’ (because when I read it I get the feeling I’m reading the plot of an exciting fictional novel); [url='http://ecmweb.com/arc-flash/case-deadly-arc-flash']http://ecmweb.com/arc-flash/case-deadly-arc-flash[/url]

That is a very good article

Dear PaulEngr and other forum members,

I have a follow up question about NFPA70E’s definition of an arc flash hazard and interacting with equipment.

The NFPA70E states:

I had a discussion with my father on this subject, he is an retired electrician (they let him go because off cutbacks in the workforce). He always says ‘an electrician can only make 1 mistake in his life (because that mistake kills him)’, it is a bit exaggerated but I get the message.

He stated that there is no way you can know for sure that equipment has been properly installed and maintained so you should always assume that it is NOT properly installed and maintained.

My question is if the NFPA states that equipment that has been properly installed and maintained is not a hazard, how do people verify this?

For example if I know a 690V compartment that can create a 35 cal/cm2 flash but has been visually inspected 5 weeks ago, is it okay to turn loads on locally while wearing 1.2 cal protection (normal work clothes)?

What I am getting at is when does equipment fall under the section of ‘safe’, and should one rely on this because there is still risk involved?

NFPA70E 130.6J states:


But how this evidence is found I cannot find anywhere.

Sounds like two questions. As to properly installed, this would be inspecting to NEC requirements. I forgot the number but there is an NFPA inspection standard, NETA ATS (Acceptance Testing Standard), and the new NFPA standards for inspections for on site Labelling inspections.

As to properly maintained, this is covered in Chapter 2 of 70E and generally references 70B or NETA MTS.

AAsAAs ttoto aanantantianticanticianticippapatpatipatinpatingg ffafaifaillulurluree, wwewe ddididd tththiiss iinn hhohouhoussese aanandand ddedeffifinfinefinedfined 'aababnnonornormnormanormalnormal oopopeopeoperoperaatatiatiatioationation'ation'.

As to immediate inspections, there are obvious signs. Black/char marks, right after a trip, failed/damaged enclosures, water/moisture on the surface or visually running down in it. That sort of thing. OSHA requires a full diagnostic after every trip except overloads. All molded case breakers require a visual inspection per NEMA AB4 after every trip. Other than trips, this is a pure visual 2 second inspection over and above 70B style PM's.