Hi All

Just looking for some advice here. In my sessions up here in Ontario, I will often pose the question to the group- "What is an arc flash" or "What causes an arc flash"? Considering many of these folks have a limited knowledge of electrical systems, when I finally get the answer "the result of a fault, or short circuit condition" I proceed with an example such as the incorrect connections in a 120 V receptacle or a ceiling light. Obviously on a 120 Volt residential circuit- a loud pop, perhaps a very small amount of shrapnel and bright light is all one will get. How do we define or refer to this? We cannot really refer to this as an "arc flash" as we know it. Do we just simply call it a spark?

Any suggestions would be greatly appreciated

Thanks
Len

You are mixing up fault types. A bolted fault is one in which there are solid connections such as your mis-wiring example. This causes the highest fault current and the greatest mechanical stress from magnetic fields proportional to the square of the current. Very little heat is generated. The fault is contained in the enclosures (if AIC is sufficient) and over current device interrupting capacity is also critical. It is possible to calculate bolted faults conservatively using minimal circuit information. Conversely an arcing fault is one in which electricity passes through air. It makes a lot of heat and equipment ruptures. Arc resistant equipment simply ruptures in a controlled way. The amount of damage is extensive.

However arc flash usually is most severe for 250-1000 V. At each half cycle when the current passes through zero, the arc extinguishes. If the conductors are close enough and the air is hot enough, the arc restrikes as voltage increases during the next half cycle beyond a threshold value generally around 150 V. From a DC perspective below a threshold (voltage available, current, temperature, and gap dependent) at low level conditions the arc (actually a spark) is weak and almost harmless. Beyond a threshold though it becomes a power arc and very destructive. In AC this just happens in a cycle 120 times per second. This results in a "square wave" pattern that an AFCI recognizes in 120 V circuits. Due to three phase interactions development of a similar 400-600 V device has not occurred. Below 250 V the time interval between the AC voltage getting over the threshold to ignite the arc and current passing through zero becomes fairly short. Without much of an arc to hear air to the point where it becomes conductive, most arcs below 250 V self extinguish harmlessly, but not all. In 2009 an electrician in Georgia was killed from an arc flash in a temporary construction panel according to OSHA injury data.

Hi Paul

Thanks allot for the reply here. Yes of course a bolted faul such as improper connection regardless of voltage system versus an arcing fault

Makes sense and thanks for the explanation here

All the best

Several years ago Lloyd Gordon (PhD and Chief Electrical Safety Officer from the Los Alamos National Laboratory) answered a similar question for Department of Energy folks. I can summarize the anwer here.

Not all arcs will result in an arc flash. Some arcs are controllable and desirable (i.e. arc welder or arc furnace).
Arc hazards include heat, molten metal, sparks, UV rays (think of the PPE a welder wears).
An arc flash occurs when an arc has sufficient voltage and current to allow the arc and resultant plasma ball to self-sustain and grow.
Arc flash hazards include heat, molten metal, sparks, UV rays, shrapnel, sound waves, pressure waves, and shock.

Now to sum up the questions with answers in my own words:

What is an arc flash? It is a self-sustaining, expanding plasma ball.
What causes an arc flash? An arc flash is the result of an arcing fault occuring in an electrical system that has sufficient voltage and energy to sustain and expand the plasma cloud resulting from the initial arc.
What causes an arcing fault? Two conductors of different electrical potential getting close enough to each other to allow current flow through the air gap between them. This could be the result of defective equipment or could be caused by human error.
Will all arcs cause an arc flash? No.
Are all arcing faults bad? No, some are desirable and controllable. Welders and arc furnaces are two examples.
Are all arc flashes bad? I can't think of a good one.

I'm open to comments on these.

NFPA 70E-2015 gives the definition of an Arc Flash Hazard as "A dangerous condition associated with the possible release of energy caused by an electric arc." That is pretty much all they have as far as official definitions for arc flash. It really doesn't differentiate between the hazard posed by arcs and arc flashes.

As you said, CONTROLLED arcs are fine and in fact in many cases can be very useful. Not only for arc welding but also by nature, all electrical disconnecting devices (relays, switches, circuit breakers) work by forming and breaking a controlled arc. And all discharge lighting (neon, metal halide, mercury vapor, sodium, fluorescent) works using a controlled arc. All types of fuses also operate by forming/breaking an arc under controlled conditions.

Where I deviate from the above is that I would agree that there are multiple physical manifestations of an uncontrolled arc, many of which can cause serious bodily harm. However to focus on two in general, the radiant heat generated by an arc is called an arc flash and the incident energy is the calculation associated with the heat. Plasma effects are currently not really directly taken into account. The effect of plasma is being studied because it can actually exceed the injury due to the radiant heat itself in some cases. But as of today, the focus is exclusively on radiant heat energy. One of the striking issues is that the most common PPE made from porous cloth is less effective against plasma compared to radiant heat but arc resistant rain wear is much more effective than predicted.

The "arc" itself is a very thin region of air, millimeters thick at most. Plasma is the "4th phase" of matter where temperatures are so hot that the electrons just sort of float around the nucleus of atoms freely without being tied to any particular molecule. We also say that the air "dissociates" which means that chemical bonds such as the bond holding two oxygen atoms together does not really exist anymore...everything just floats around in a single gaseous mass. Because of this, it is a highly conductive state of matter. And with conductivity very high, heating is limited. Thus its not really the core where the heating occurs but the surrounding air. The temperature in the arc is thus in many ways self-limiting. The plasma core can't really grow substantially without exponentially more energy involved, and the temperature rise tapers off once plasma forms because resistance is extremely limited. Air itself becomes luminous at a temperature below the temperature necessary to turn it into plasma to what we physically see as an arc is not the plasma but the surrounding glowing air that indicates the presence of the arc.

It's also a misnomer to suggest that it's just a matter of coming into close contact. Although the critical flashover distance and several related electrical calculations are an important design consideration, the actual distance is somewhat ill defined depending on temperature, humidity, altitude, contamination, and shape and material of the components. Thus most equipment is vastly overbuilt, well beyond any critical values. With rare exceptions, even fairly large deviations from proper design practice rarely result in an arc. At medium voltage levels (above 2 kV-3 kV) at best it is not uncommon to see corona/tracking/partial discharge that results in equipment breakdown and damage over time. These can but rarely do result in an arc flash, and even if they did, the odds are very much against someone injured while just walking by.

What is more common is if for instance one wire or bus bar falls into another such as from external damage. This is usually a bolted fault. An arcing fault happens if it is magnetically flung or simply vaporizes at the point of contact so that an arc gap is formed in the process. In actual lab testing the most common method is to tie a piece of bare #14 copper wire between two bus bars. So the vast majority of arc flashes are created or set off by either defective equipment or human error. Electrical equipment that is installed and maintained in good working order rarely, if ever, spontaneously arcs. It is energized work or disturbing damaged/faulty equipment that is generally the cause of an arc flash.

UV is taken into account indirectly. Face shields are designed specifically to block a certain amount of UV but any other effects are not taken into account.

Arc blast is the pressure wave caused by heating the air around the arc. It is taken into account in some "arc resistant" (actually arc blast resistant) equipment designs, and by a requirement to wear ear plugs. There is a theoretical calculation due to Lee but it has been largely discredited because it produces unrealistic results with regards to the pressure developed by an arc. Actual test work on arc blasts has shown that the arc blast is over with in 1-2 cycles (16-32 milliseconds), and that it is essentially independent of the arc power. More powerful arcs cause the arc blast to occur sooner but the arc blast itself is dictated by the structural strength of the enclosure. Once the enclosure ruptures, all of the pressure developed is released. Larger enclosures can result in a longer onset to an arc blast but with more surface area to contain the pressure, generally result in lower arc blast pressures. These results contrast with the theoretical equation developed by Lee that suggest that arc blast is essentially proportional to arc power and that there is no theoretical limit on the arc blast.

Shrapnel is only indirectly accounted for by procedures which recommend standing to one side of an enclosure while operating a disconnect as an example but no specific engineering guidelines, rules, regulations, etc.

So although we are talking about uncontrolled power arcs and they do generate a large number of physical phenomena, it would be unfair/incorrect to characterize 70E and IEEE 1584 as addressing all of them when in fact they don't address ALL physical phenomena so far.

The reason that I can say that plasma is rarely an issue is that if plasma were the issue, we'd see a lot of residual vaporized copper coating all the surfaces around an arcing fault. Molten balls of metal are pretty common but not a coating that looks like vapor deposition. There is definitely plasma in an arc but it is very little and confined only to a very small core region millimeters thick or less. The amount of energy necessary to generate more than this becomes essentially impossible and as I said earlier, as the resistance within the plasma is extremely low, electrical current conducts freely and thus you really can't have an "expanding sphere of plasma" at all. It is confined to a small area.

Hi Folks

Excellent and thanks allot for the information. This is extremely helpful and much appreciated and will certainly lend a hand when explaining this to the non- electrical and even the electrical worker

Enjoy your day

Len

Len,

To summarize the answers to your quesitons:

What is an arc flash? The thermal energy given off by an uncontrolled arc through open air.

What causes an arc flash? Energized work or disturbing damaged/faulty equipment that is generally the cause of an arc flash.

How do we refer to the pop, light, and material thrown off of a 120V residential circuit when short circuited? Do we simply call it a spark? Yes.

Great explanations here and most beneficial when speaking to the non electrical workers

Thanks once again

All the best

Len

I kinda figured things hadn't changed to the point where every time someone shorts out a 120 volt receptacle it's an arc flash event... Now to get that to our Safety gurus...