From a previous post it might not be obvious just what is going on in an arc flash. There are lots of flashy presentations out there meant to scare someone but they don't really do a good job of explaining what's going on. I hope this quick little description helps make sense out of what happens during an arc flash and what the calculations are meant to accomplish.
1. From Wikipedia: melting point of copper is 1984 F, boiling point is 4643 F.
2. Same source, when it vaporizes, it expands to over 1,000 times it's original size as a solid or liquid at 1 atmosphere (normal pressures; don't feel like doing the partial pressure math for this example).
3. Same source, copper is 8.94 g/cm^3 at room temperature.
4. It takes 300 kj / 63.546 g = 4727 joules (aka watt-seconds) to melt 1 gram of copper.
Most importantly, and not in the wikipedia though:
5. Copper is just as conductive as a gas as it is as a liquid or solid.
6. Electric arcs can reach roughly 10,000 F (plenty of heat to vaporize a little copper).
So...imagine a 1000 A fault at 480 Volts. These numbers are actually a little on the low side for most 480 V distribution gear but serves the purpose. Using simple D.C. math (volts X amps), it takes about 4.5 milliseconds to vaporize a half gram of copper with the energy in this fault. The copper vaporizes and the volume expands from 65 microliters to 65 milliliters, or just shy of a quarter cup of volume for those who are metric-challenged. By this point depending on how the equipment is built, all 3 phases are likely to be covered in a mix of vaporized copper and plasma, and these gases are just as conductive as the original solid material. Effectively all 3 phases are shorted together. All this is happening in 4.5 ms, or 1/4 cycle, the time it takes for a fuse to "instantaneously" trip.
This is assuming that the fuse "instantaneously" trips. If it doesn't open soon enough, the heat and energy continue to vaporize more copper and continue to expand out over all the electrical gear, vaporizing everything in the area until it the gap gets so large that there's not enough energy to continue to sustain the arc and it goes out. Older circuit breakers only open in 1-2 cycles for smaller circuit breakers, up to about 12 cycles for the very large medium or high voltage units. In the mean time, that arc just keeps tearing up everything in the area. I'm sure almost all the electrical equipment of concern has a lot more copper in it than just 65 microliters.
The gases are white hot, so it is also radiating lots of heat outwards towards any available surface such as exposed skin, clothing, flammable material, insulation, etc. Not a pretty sight for anything weaker than fire brick, and even refractories and ceramics are only good for a short period of time at those temperatures. The best that one can hope for is that it takes a certain amount of time to penetrate the material, and that it will not sustain a flame. The thicker the material is, the longer it will hold out before failing. This is all that "arc resistant" clothing does...it is designed to be sacrificial.
All the air and remaining solid components aren't just going to sit there idly as something nearby expands to several times it's original size. Nope, they are going to come flying out of there at a pretty high rate of speed. You can't bottle this one up...it is going to find a way to relieve the pressure whether the equipment was designed for it or not, and most switch gear is not designed to handle this much pressure. The doors usually come flying off and all the internals are warped and bent out of shape. This is arc blast, and right now, not enough research has been done to create safety standards to deal with it.
So....what can we do about this? Plenty!
Increase the impedance in the circuit, which reduces the available fault current. Usually there's not much that can be done about this except if you use current limiting fuses or circuit breakers, or switch to a different grounding system (only applies to line-ground faults, not line-line faults).
Decrease the voltage...if it's under 50 volts, generally you can't get an arc to ignite without direct contact. This is one very good reason for switching from 120 VAC to 24 VDC controls.
Increase the spacing...wider is better. It is harder to light or sustain an arc.
Get away...if someone across the room using a remote control to open/close breakers or rack them in/out, the heat and other stuff spreads out more before reaching the victim.
Open the circuit faster. At 65 mL worth of material, that's not much. It might make a big bang but it probably won't cause too much damage. Move the trip settings down as low as possible, change to faster/better protective equipment, or even use a fiber optic arc detection system that trips in a fraction of a millisecond.
All of these things will reduce the amount of damage done to both people and equipment if an arcing fault happens.
Arc flash calculations simply determine the amount of damage that a person in the "wrong place at the wrong time" might experience. Most of the folks reading the Mr. PLC forums are of course troubleshooters at one point or another. The reason that they are troubleshooting is because something is already wrong. So it just takes that last ingredient (wrong time) to make them have a very, very bad day.
Full Version: Arc flash anecdotal info
QUOTE (paulengr @ Sep 30 2009, 06:27 PM) 
From a previous post it might not be obvious just what is going on in an arc flash. There are lots of flashy presentations out there meant to scare someone but they don't really do a good job of explaining what's going on. I hope this quick little description helps make sense out of what happens during an arc flash and what the calculations are meant to accomplish.
1. From Wikipedia: melting point of copper is 1984 F, boiling point is 4643 F.
2. Same source, when it vaporizes, it expands to over 1,000 times it's original size as a solid or liquid at 1 atmosphere (normal pressures; don't feel like doing the partial pressure math for this example).
3. Same source, copper is 8.94 g/cm^3 at room temperature.
4. It takes 300 kj / 63.546 g = 4727 joules (aka watt-seconds) to melt 1 gram of copper.
Most importantly, and not in the wikipedia though:
5. Copper is just as conductive as a gas as it is as a liquid or solid.
6. Electric arcs can reach roughly 10,000 F (plenty of heat to vaporize a little copper).
So...imagine a 1000 A fault at 480 Volts. These numbers are actually a little on the low side for most 480 V distribution gear but serves the purpose. Using simple D.C. math (volts X amps), it takes about 4.5 milliseconds to vaporize a half gram of copper with the energy in this fault. The copper vaporizes and the volume expands from 65 microliters to 65 milliliters, or just shy of a quarter cup of volume for those who are metric-challenged. By this point depending on how the equipment is built, all 3 phases are likely to be covered in a mix of vaporized copper and plasma, and these gases are just as conductive as the original solid material. Effectively all 3 phases are shorted together. All this is happening in 4.5 ms, or 1/4 cycle, the time it takes for a fuse to "instantaneously" trip.
This is assuming that the fuse "instantaneously" trips. If it doesn't open soon enough, the heat and energy continue to vaporize more copper and continue to expand out over all the electrical gear, vaporizing everything in the area until it the gap gets so large that there's not enough energy to continue to sustain the arc and it goes out. Older circuit breakers only open in 1-2 cycles for smaller circuit breakers, up to about 12 cycles for the very large medium or high voltage units. In the mean time, that arc just keeps tearing up everything in the area. I'm sure almost all the electrical equipment of concern has a lot more copper in it than just 65 microliters.
The gases are white hot, so it is also radiating lots of heat outwards towards any available surface such as exposed skin, clothing, flammable material, insulation, etc. Not a pretty sight for anything weaker than fire brick, and even refractories and ceramics are only good for a short period of time at those temperatures. The best that one can hope for is that it takes a certain amount of time to penetrate the material, and that it will not sustain a flame. The thicker the material is, the longer it will hold out before failing. This is all that "arc resistant" clothing does...it is designed to be sacrificial.
All the air and remaining solid components aren't just going to sit there idly as something nearby expands to several times it's original size. Nope, they are going to come flying out of there at a pretty high rate of speed. You can't bottle this one up...it is going to find a way to relieve the pressure whether the equipment was designed for it or not, and most switch gear is not designed to handle this much pressure. The doors usually come flying off and all the internals are warped and bent out of shape. This is arc blast, and right now, not enough research has been done to create safety standards to deal with it.
So....what can we do about this? Plenty!
Increase the impedance in the circuit, which reduces the available fault current. Usually there's not much that can be done about this except if you use current limiting fuses or circuit breakers, or switch to a different grounding system (only applies to line-ground faults, not line-line faults).
Decrease the voltage...if it's under 50 volts, generally you can't get an arc to ignite without direct contact. This is one very good reason for switching from 120 VAC to 24 VDC controls.
Increase the spacing...wider is better. It is harder to light or sustain an arc.
Get away...if someone across the room using a remote control to open/close breakers or rack them in/out, the heat and other stuff spreads out more before reaching the victim.
Open the circuit faster. At 65 mL worth of material, that's not much. It might make a big bang but it probably won't cause too much damage. Move the trip settings down as low as possible, change to faster/better protective equipment, or even use a fiber optic arc detection system that trips in a fraction of a millisecond.
All of these things will reduce the amount of damage done to both people and equipment if an arcing fault happens.
Arc flash calculations simply determine the amount of damage that a person in the "wrong place at the wrong time" might experience. Most of the folks reading the Mr. PLC forums are of course troubleshooters at one point or another. The reason that they are troubleshooting is because something is already wrong. So it just takes that last ingredient (wrong time) to make them have a very, very bad day.
1. From Wikipedia: melting point of copper is 1984 F, boiling point is 4643 F.
2. Same source, when it vaporizes, it expands to over 1,000 times it's original size as a solid or liquid at 1 atmosphere (normal pressures; don't feel like doing the partial pressure math for this example).
3. Same source, copper is 8.94 g/cm^3 at room temperature.
4. It takes 300 kj / 63.546 g = 4727 joules (aka watt-seconds) to melt 1 gram of copper.
Most importantly, and not in the wikipedia though:
5. Copper is just as conductive as a gas as it is as a liquid or solid.
6. Electric arcs can reach roughly 10,000 F (plenty of heat to vaporize a little copper).
So...imagine a 1000 A fault at 480 Volts. These numbers are actually a little on the low side for most 480 V distribution gear but serves the purpose. Using simple D.C. math (volts X amps), it takes about 4.5 milliseconds to vaporize a half gram of copper with the energy in this fault. The copper vaporizes and the volume expands from 65 microliters to 65 milliliters, or just shy of a quarter cup of volume for those who are metric-challenged. By this point depending on how the equipment is built, all 3 phases are likely to be covered in a mix of vaporized copper and plasma, and these gases are just as conductive as the original solid material. Effectively all 3 phases are shorted together. All this is happening in 4.5 ms, or 1/4 cycle, the time it takes for a fuse to "instantaneously" trip.
This is assuming that the fuse "instantaneously" trips. If it doesn't open soon enough, the heat and energy continue to vaporize more copper and continue to expand out over all the electrical gear, vaporizing everything in the area until it the gap gets so large that there's not enough energy to continue to sustain the arc and it goes out. Older circuit breakers only open in 1-2 cycles for smaller circuit breakers, up to about 12 cycles for the very large medium or high voltage units. In the mean time, that arc just keeps tearing up everything in the area. I'm sure almost all the electrical equipment of concern has a lot more copper in it than just 65 microliters.
The gases are white hot, so it is also radiating lots of heat outwards towards any available surface such as exposed skin, clothing, flammable material, insulation, etc. Not a pretty sight for anything weaker than fire brick, and even refractories and ceramics are only good for a short period of time at those temperatures. The best that one can hope for is that it takes a certain amount of time to penetrate the material, and that it will not sustain a flame. The thicker the material is, the longer it will hold out before failing. This is all that "arc resistant" clothing does...it is designed to be sacrificial.
All the air and remaining solid components aren't just going to sit there idly as something nearby expands to several times it's original size. Nope, they are going to come flying out of there at a pretty high rate of speed. You can't bottle this one up...it is going to find a way to relieve the pressure whether the equipment was designed for it or not, and most switch gear is not designed to handle this much pressure. The doors usually come flying off and all the internals are warped and bent out of shape. This is arc blast, and right now, not enough research has been done to create safety standards to deal with it.
So....what can we do about this? Plenty!
Increase the impedance in the circuit, which reduces the available fault current. Usually there's not much that can be done about this except if you use current limiting fuses or circuit breakers, or switch to a different grounding system (only applies to line-ground faults, not line-line faults).
Decrease the voltage...if it's under 50 volts, generally you can't get an arc to ignite without direct contact. This is one very good reason for switching from 120 VAC to 24 VDC controls.
Increase the spacing...wider is better. It is harder to light or sustain an arc.
Get away...if someone across the room using a remote control to open/close breakers or rack them in/out, the heat and other stuff spreads out more before reaching the victim.
Open the circuit faster. At 65 mL worth of material, that's not much. It might make a big bang but it probably won't cause too much damage. Move the trip settings down as low as possible, change to faster/better protective equipment, or even use a fiber optic arc detection system that trips in a fraction of a millisecond.
All of these things will reduce the amount of damage done to both people and equipment if an arcing fault happens.
Arc flash calculations simply determine the amount of damage that a person in the "wrong place at the wrong time" might experience. Most of the folks reading the Mr. PLC forums are of course troubleshooters at one point or another. The reason that they are troubleshooting is because something is already wrong. So it just takes that last ingredient (wrong time) to make them have a very, very bad day.
Just to get things in perspective
I did a master thesis on arc welding and eye / skin injury ie weldors' flash it does not take that long even if you are 5 to 10 feet away - and that is done with only 28 volt or so and 90 to 300 amps with the electrode size 3/32 to 1/4.
I have repaired switch gear that blew and was quite impressed with melted bus bar and teh ash inside. worst was at Trojan where a MCC had a two fool dia hole blown in it.
Saw a case at Boeing where electricians forgot to take the breaker handle off the stabs in back on a 480 V 1200 A (??) tie breaker. When they racked it in they both got sunburns.
Have seen my share of electricians with scars from burns. I am surprised there are not more burned scarred electricians and how well equipment contains the faults.
Dan Bentler
QUOTE (paulengr @ Sep 30 2009, 07:27 PM)
5. Copper is just as conductive as a gas as it is as a liquid or solid.
Interesting, I did not realize this
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.