Science of Arc Flash & Engineering

tersh

Senior Member
I'd like to know the following information about arc flash which I didn't learn much in engineering in college.

In youtube video about it which I watched a lot. I learn it has 4 times the temperature of the sun. It can explode with energy of grenade. What started as a small millimeter arc can cause chain reaction that can blow the electrician meters away.

<font color="#222222"><span style="font-family: Verdana">

1. Does all arcing do this? what is the threshold voltage and current before it can reach more than the temperature of the sun?

2. Should there be certain humidity or air particles before it can cause continuous chain reaction? What must be present together for something like in the video above to happen? Does it happen in typical home panels?

3. The energy can vaporize metal and surrounding material. Have they integrated this into some kind of weapon? I mean. I just want to know the principles of it and by understanding how they can engineer it at will.. understand more about arc flash.

4. Who has experienced arc flash? How big or small is it?

5. Generally. I'd like to understand the exact process or science whereby the surrounding air becomes ionized enough for millimeter arc (between two conductors very close by) to become meters wide arc flash (like what is the fuel for it). What would stop it from chain reacting into the entire room or bigger arc flash and blast?

6. To what extend can AFCI breakers prevent it or limit the intensity? Like make it 20% only or half the intensity or totally prevent it?

Thank you.
 
Last edited:

cpickett

Member
I can't answer most of your questions, but I am interested in hearing other responses.

I do believe AFCI breakers are designed to protect against arc faults in wiring, not arc flash. Not sure of their effectiveness in an arc flash event.

Arc flash protection is given by enclosures that have thicker steel than normal, and vents and baffles designed to direct the blast up and away from an operator, without blowing apart under the pressure. There are also breakers that have maintenance modes which temporarily decrease the trip settings in an effort to reduce incident energy. I think they also make breakers that have a photocell-like device that will trip the breaker when it detects a flash.
 

gar

Senior Member
190204-0806 EST

tershL

See https://en.wikipedia.org/wiki/Electric_arc

Negative resistance.

Over a wide range an arc's current increases as voltage goes down across the arc. The external circuit determines current because voltage remains somewhat constant but goes down. Power = V*I. As current increases the voltage drops, but not as fast as the current increases. Thus, power increases. Temperature increases with increasing power. If the region in which power is dissipated is relatively contained, then very high temperature occurs, and can produce explosive effects.

.
 

gar

Senior Member
190204-0906 EST

See https://commons.wikimedia.org/wiki/File:Glow_discharge_current-voltage_curve_English.svg
See https://quod.lib.umich.edu/b/bhlead/umich-bhl-87210?view=text one of my professors.
His book on Fundamentals Of Engineering Electronics has a discussion on arcs and a curve I don't find on the Internet. I only see the 1937 edition listed. I studied from his second edition of 1952. When Dow was close to 100 he still walked in our neighborhood. His mind was still sharp at that age. He lived about 5 blocks from me and there is quite an elevation change that he had to traverse.

.
 

tersh

Senior Member
190204-0806 EST

tershL

See https://en.wikipedia.org/wiki/Electric_arc

Negative resistance.

Over a wide range an arc's current increases as voltage goes down across the arc. The external circuit determines current because voltage remains somewhat constant but goes down. Power = V*I. As current increases the voltage drops, but not as fast as the current increases. Thus, power increases. Temperature increases with increasing power. If the region in which power is dissipated is relatively contained, then very high temperature occurs, and can produce explosive effects.

.
Why is it not common? The process seems to be:

1. An initial short
2. The shorting material destroyed as plasma forms
3. Plasma is self sustaining and draws more and more current (in a short current is huge almost unlimited)
4. It explodes in a pressure arc blast melting metal and surrounding and injuring or killing people.

But why don't I hear it reported more often of houses getting blown from arc flash and blast. I heard of LGP exploding in houses but not these.

What conditions can exactly produce them?
 

gar

Senior Member
190204-1016 EST

tersh:

Impedance external to the arc is what determines the arc current. The source voltage, transformer, and wires, or anything else in series with the arc determines the current. The arc itself may not be a major factor.

All devices such as arc lamps, LEDs, fluorescent bulbs, Zener diodes, etc., require some form of external current limiting.

.
 

kwired

Electron manager
Why is it not common? The process seems to be:

1. An initial short
2. The shorting material destroyed as plasma forms
3. Plasma is self sustaining and draws more and more current (in a short current is huge almost unlimited)
4. It explodes in a pressure arc blast melting metal and surrounding and injuring or killing people.

But why don't I hear it reported more often of houses getting blown from arc flash and blast. I heard of LGP exploding in houses but not these.

What conditions can exactly produce them?
Most of the time dwellings don't have large enough supply to deliver extreme destructive level arc flash. Plus 120 to ground won't self sustain arcing as easily as say 277 volts can. Total incident energy is also based not only on how much power is given up in the incident, but also takes duration of the event in consideration. A high current but short duration event can have less incident energy than a lower current but more time event.

The higher the incident energy the more potential destructiveness the event can be. At very least you are typically exposed to higher burn risk, but more energy also means more blast pressure and more debris possibly being thrown as well.

A certain level of fault may not be as serious in some aspects in open overhead lines as same energy in a panelboard with just one opening facing the technician working on it. The open line energy can travel all directions, in the panelboard situation the back and sides of the enclosure will reflect energy toward the technician that otherwise would have been traveling away from the technician.
 

cpickett

Member
Total incident energy is also based not only on how much power is given up in the incident, but also takes duration of the event in consideration. A high current but short duration event can have less incident energy than a lower current but more time event.
I saw a presentation on Arc Flash that showed several common plant electrical systems and the corresponding arc flash analysis. It showed a feeder with a lower short circuit current due to a small transformer that actually had more incident energy than a feeder with a larger short circuit current. The explanation was that the higher short circuit current resulted in a faster fault clearing time at the upstream breaker.

It highlighted the need for a full study to understand the incident energy available at each location, and that you can't really rely on any rules of thumb to know what the hazard is.
 

hbiss

EC, Westchester, New York NEC: 2014
6. To what extend can AFCI breakers prevent it or limit the intensity? Like make it 20% only or half the intensity or totally prevent it?
AFCI breakers have NOTHING to do with Arc Flash. Once again you are asking questions and making statements based on your imagination.

-Hal
 

tersh

Senior Member
Most of the time dwellings don't have large enough supply to deliver extreme destructive level arc flash. Plus 120 to ground won't self sustain arcing as easily as say 277 volts can. Total incident energy is also based not only on how much power is given up in the incident, but also takes duration of the event in consideration. A high current but short duration event can have less incident energy than a lower current but more time event.

The higher the incident energy the more potential destructiveness the event can be. At very least you are typically exposed to higher burn risk, but more energy also means more blast pressure and more debris possibly being thrown as well.

A certain level of fault may not be as serious in some aspects in open overhead lines as same energy in a panelboard with just one opening facing the technician working on it. The open line energy can travel all directions, in the panelboard situation the back and sides of the enclosure will reflect energy toward the technician that otherwise would have been traveling away from the technician.
Let's say your home has the typical black and red 240v main feeder, and the white neutral and green ground wires. And let's say the main breaker is 125A. If the electrician accidentally drop a wrench that shorts the bus bar of the 240v black and red lines. Can it produce arc flash enough to blow the electrician away?

How big is the arc flash (if it exist) that would be produced in this. Let's actually do calculations now.

Let's also compute what would happen if the load side of 2-pole 30A breaker (240volts) get shorted by another tool by the electrician (separate scenario). Would this produce arc flash?
 

Jraef

Moderator
Staff member
The blast pressure wave stems from the fact that metals, like copper, expand by 67,000 times their volume in the matter of milliseconds. For a perspective, the metal in a penny will expand to a plasma-induced vapor ball larger than a 5 gallon bucket in a fraction of a second. How MUCH material gets vaporized and expanded is a function of the amount of available fault current in the system and the time it takes for something to clear the fault. So if it's a fault in a 120V wire touching your diagonal cutters (as I just did a few weeks ago) where the breaker trips in a second, there is not a lot of energy available in that circuit, compared to a 480V switchboard fed by a 1MVA transformer in which a wrench drops across the bus bars, even though the breaker senses and trips in under a second.
 

jim dungar

Moderator
Staff member
How big is the arc flash (if it exist) that would be produced in this. Let's actually do calculations now.
First, you have not provided anywhere information in order to make a reasonable calculation. We need to know the available fault current, the impedance of the conductors from point the fault current was calculated at to the point of the fault, and the time-current characteristics of the upstream protective device. If you are going to just assume this data, you might as well just assume the results.

Second, what you are asking for cannot be calculated directly. Arc Flash incident energy (AFIE) for three-phase circuits is fairly easy to calculate using formulas like those found in IEEE1584, however there are no standard industry accepted methods for calculation AFIE for single-phase circuits such as those found in a residential load center.

Third, a person would be 'thrown back' by the pressure wave. At present there is no generally applicable methodology for determining the blast effect from an arcing fault.
 

tersh

Senior Member
First, you have not provided anywhere information in order to make a reasonable calculation. We need to know the available fault current, the impedance of the conductors from point the fault current was calculated at to the point of the fault, and the time-current characteristics of the upstream protective device. If you are going to just assume this data, you might as well just assume the results.

In an average house in the United States. What is usually the average fault current when there is busbar short of red and black (phase to phase), what is usually the impedance of the conductors from point the fault current was calculated to the point of the fault, and time-current characteristics of say 125A General Electric
TQD22150 breaker.

I'm just asking the average above. And yes if we assume the data, and we can assume the results. How big would the arc flash be created for the above scenario. What are the electricians experience in the entire USA for such home arc flash? What are the general reports about it?

Second, what you are asking for cannot be calculated directly. Arc Flash incident energy (AFIE) for three-phase circuits is fairly easy to calculate using formulas like those found in IEEE1584, however there are no standard industry accepted methods for calculation AFIE for single-phase circuits such as those found in a residential load center.

Third, a person would be 'thrown back' by the pressure wave. At present there is no generally applicable methodology for determining the blast effect from an arcing fault.
 

jim dungar

Moderator
Staff member
In an average house in the United States. What is usually the average fault current when there is busbar short of red and black (phase to phase), what is usually the impedance of the conductors from point the fault current was calculated to the point of the fault, and time-current characteristics of say 125A General Electric
TQD22150 breaker.

I'm just asking the average above. And yes if we assume the data, and we can assume the results. How big would the arc flash be created for the above scenario. What are the electricians experience in the entire USA for such home arc flash? What are the general reports about it?
The breaker in the panel is not important, if the 'plasma' is large enough to throw a person backwards, it is likely large enough to envelope the main breaker in the panelboard. The important device would be anything upstream from the panelboard.

I have never tried to estimate the arc blast effect, but people being 'knocked off balance' seems to be a not uncommon occurrence.
For estimating the results, I have heard that a fatality was recorded for 120V L to ground arc flash, and I know people have not died from one. So, what result do you want to use as the average?


Turn the power off.
Perform a risk analysis.
Wear appropriate PPE.
 

tersh

Senior Member
The breaker in the panel is not important, if the 'plasma' is large enough to throw a person backwards, it is likely large enough to envelope the main breaker in the panelboard. The important device would be anything upstream from the panelboard.

I have never tried to estimate the arc blast effect, but people being 'knocked off balance' seems to be a not uncommon occurrence.
For estimating the results, I have heard that a fatality was recorded for 120V L to ground arc flash, and I know people have not died from one. So, what result do you want to use as the average?


Turn the power off.
Perform a risk analysis.
Wear appropriate PPE.
Anything upstream from the panelboard would be the wiring to the utility pole transformer. So I'm using the United States average split phase utility pole. What size of wiring is usually used to power the 125A main breaker in a US home? AWG 1/0? Then whatever that service entrance wire size is will be the conductor and the impedance in our computational estimate. By these estimate. How big would be the arc flash produced in typical US home phase to phase short from dropped tool?

Just asking and of course no electrician would purposedly drop a tool in live panel.
 

GoldDigger

Moderator
Staff member
A possibly apocryphal story tells of the two tin-knockers who were installing a section of air duct above the + and - busbars of a telco exchange DC distribution system (48 volt). They dropped the duct and only the two ends were left to hit the floor. :happysad:
 

junkhound

Senior Member
Who has experienced arc flash? How big or small is it?

FWIW


Ground burst nuclear EMP testing.
40,000 uF capacitor bank at 10 kV (size = 16 ft high, 24 ft across, 3 ft deep)
Setup in a buck configuration with 2 big mercury arc rectifiers, a 4 ft diameter inductor with 1" square windings held in place by (about IIRC) 16ea 3/4" SS bolts. Supposed to fire the main line ignitron, then after inductor reaches current peak, the other ignitron at input of the inductor fired to create a long tail 100s of kA L/R pulse.

Ignitron Triggers inadvertently laid near each other, both ignitrons triggered at same time producing short circuit.

BIG CT in the circuits saturated at 4 million amps peak. Conductors were 2ea 6" aluminum channels back to back bolted with 1/2" grade 8 bolts every 2 feet.

Current BROKE 250 MCM leads to ignitrons. Alum channel terminal was 2 half inc folts on the 250MCM lugs, broke the bolts.

Test folks in next room

Arc flash put a 1 foot bow into a heavy steel roll up door, moved the opposite steel stud wall with 3/4 fire treated ply on both side 6 inches uniformily into the next room. Did not move the concrete block walls at the ends, no roof damage.

Blew the steel door to the lab opening by shearing the latch .

Great fun.
 

tersh

Senior Member
Or to be realistic. Let's take the issue of loose breakers and bus bar or any loose connections in the main panel in typical US home.

When the breaker or wiring got loose. Arc would be formed. For typical 125A main breaker and 1/0 wire to the service entrance and the utility pole tap in US home. Assume it is the main breaker lugs that is loose. Can this form a self sustaining plasma that can grow in size as it feeds on the current enough to explode the panels and the room it is put on?

I'd like to hear reports of this. Worse-case scenario.
 

hbiss

EC, Westchester, New York NEC: 2014
When the breaker or wiring got loose. Arc would be formed. For typical 125A main breaker and 1/0 wire to the service entrance and the utility pole tap in US home. Assume it is the main breaker lugs that is loose.
125VAC cannot sustain an arc. Arc faults are caused by FAULTS not loose connections. Generally, transformers serving residential services do not provide substantial fault currents.

Do you stay up at night dreaming up things to ask here?

-Hal
 

zbang

Senior Member
What size of wiring is usually used to power the 125A main breaker in a US home? AWG 1/0?
With some PoCo's you're lucky to get 6 or 4g copper for a 100A service, and anywhere from 25' to maybe 500' from the transformer (which could be from 10 to 50KVA, we don't know the loading, and we don't know the current available on the primary size). With those loose parameters, it's kinda hard to give a simple number as the impedance and thus the short-circuit current available at the wrench you drop.
 
Top