6

I know that electric arcs are dangerous as they can start fires, cause explosions, melted insulation, etc. and they are generally undesirable.

However, I am a bit confused as to why. Arcs generate heat, but shouldn't normal current flow also generate heat? Is the specific heat of plasma air, is it low enough to get extremely hot under normal current flow, which then quickly heats up the surroundings?

I am referring to simple arcs that are a few amps, not arc faults that go to hundreds (probably more) of amps.

Voltage Spike
  • 75,799
  • 36
  • 80
  • 208
Agniva Chaudhuri
  • 109
  • 7
  • 6
    *Arcs generate heat, but shouldn't normal current flow also generate heat?* Yes but if the **current** is the same, the **voltage** across the wire will much less than the voltage across the arc making the amount of energy loss **much less** in the wire. – Bimpelrekkie Jul 15 '21 at 17:08
  • 1
    @Bimpelrekkie I see, so arcs would mainly be a problem when current is high? So if the circuit has sufficient resistance (let's say resulting in a 50 mA current) it should not be as big an issue, since that would result in only 12 watts (in 240V), likely 10 watts/10 joules per second across the arc. Obviously most arcs are not like that, hence the danger. – Agniva Chaudhuri Jul 15 '21 at 18:27
  • 2
    They also generate UV which is bad for your eyes if the arc is sustained and large enough. – qrk Jul 15 '21 at 18:36
  • 1
    @qrk True, but I'm talking more about the fire hazard. UV light will definitely be very bad since it ionizes things and breaks chemical bonds, in addition to the eye damage. – Agniva Chaudhuri Jul 15 '21 at 18:46
  • 1
    *so arcs would mainly be a problem when current is high?* No, I never wrote that. For a small current we would generally use thinner wires and smaller contacts. I mean, why spend money on a switch which can handle 100 A when you're only swicthing 50 mA? For this small switch the arcing at 50 mA is just as bad as 100 A arcing would be to a large switch. The arc will still cause a plasma that will "eat" into the metal contacts. There is no bad/worse. Arcing should be avoided as much as possible. Also arcs cause radio transmissions which disturb wireless communications. – Bimpelrekkie Jul 15 '21 at 18:49
  • 1
    @Bimpelrekkie Well obviously using smaller contacts and thinner wires will worsen the effect of an arc, and small arcs would have more effect on smaller objects. I meant that it's not as big of a problem in the same scenario; the same way a handgun bullet instead of a railgun slug hitting a ship is not as bad, but handguns are pointed at people which are less durable. – Agniva Chaudhuri Jul 15 '21 at 20:10

5 Answers5

14

Arcs generate heat, but shouldn't normal current flow also generate heat?

The whole point about normal current flow is that it is through a medium called a conductor i.e. a copper wire and, the resistance of the copper wire is extremely small. This means that the heat generated (\$I^2\cdot R\$) is also quite small because the resistance is low.

Arcs usually and commonly occur through air and generate a lot of heat because the current x voltage of an air arc (also equal to power) is very high. Alternatively, you can say the equivalent resistance of air is very, very high compared to copper.

Andy aka
  • 434,556
  • 28
  • 351
  • 777
  • 1
    If the resistance is very low, then current should also go up, and cause more heat. However I think I'm getting it - the current in a standard electrical setting is limited, but then there's an area with particularly high resistance, so the heat is concentrated in that area. Is that correct? – Agniva Chaudhuri Jul 15 '21 at 18:31
  • 4
    For a copper wire (where current is dictated by the load attached to the wire and the voltage applied), the currently will barely increase at all because the load resistance dominates the situation @AgnivaChaudhuri – Andy aka Jul 15 '21 at 18:34
  • 2
    @AgnivaChaudhuri: We're not talking about a copper wire shorting out a power socket, that's also *extremely* dangerous if no circuit breaker interrupts that massive current! We're talking about a circuit that includes a toaster or hair dryer for example, i.e. a 10 ohm resistor (1500W = 120V ^ 2 / 9.6 Ohms). That sets the upper bound on current with perfect conductors elsewhere in the loop. With the wire resistance being much less than the load resistance, almost all the power is dissipated in the load, not the wire. (This is the opposite of impedance matching.) – Peter Cordes Jul 16 '21 at 06:39
  • 4
    Arcs are actually a negative incremental resistance (avalanche mode) – Tony Stewart EE75 Jul 16 '21 at 07:34
  • @PeterCordes Of course. I was referring to that only; e.g. a current of 10 Amps. Not shorting the power outlet :P – Agniva Chaudhuri Jul 17 '21 at 14:53
  • 1
    @AgnivaChaudhuri: ok right, the total current can't go up a lot. Think about impedance matching: when the resistance of an element (like the wire) is lower than the resistance of something else in a series circuit (like the real load), lowering the wire resistance reduces the voltage drop across it by more than the total increase in current. So the total power increases slightly, but the power dropped across the wire reduces much more. The maximum wire heating point is when the wire resistance matches the load resistance (impedance matching), for a given supply voltage. – Peter Cordes Jul 17 '21 at 17:00
9

Arcs generate heat, but shouldn't normal current flow also generate heat?

Arcs are the breakdown of air with the resistance going from GigaOhms to less than Ohms in milliseconds of time. This means a large current is generated which generates heat in the air, but another problem is also the shorting that happens. With amps of current flowing through the air and through copper traces, everything heats up (especially with low impedance sources like AC mains).

With this shorting it can generate currents through components far beyond what they are rated for and destroy them (I've seen some components physically explode because the heating was so fast)

Voltage Spike
  • 75,799
  • 36
  • 80
  • 208
  • Well that would be interesting to see...and with a resistance less than 1 Ohm, that will be very bright and very hot very quickly from ~300 Joules of heat energy. Thanks for the answer. – Agniva Chaudhuri Jul 15 '21 at 18:45
  • https://meta.stackexchange.com/questions/126180/is-it-acceptable-to-write-a-thank-you-in-a-comment – Voltage Spike Jul 15 '21 at 19:03
7

I know that electric arcs are dangerous [etc] However, I am a bit confused why.

An arc creates the fourth state of matter: a plasma. Plasma in air is very hot; it can be over 2500°C in fact. For this reason alone, it is a major fire hazard as almost nothing can survive those temperatures. Even a small arc (such as inside a loose connector) over time, will eat away material and cause malfunction or fire. A non-intended arc will never "get better" and can only worsen the connection.

A conductor such as a wire, as opposed to an arc, is much more conductive, so has much lower loss. It is difficult to get most conductors in air to melt, let alone double that temperature. Once overloading a conductor causes it to get hot, it will melt and break, leading to an arc. The arc has entirely different physical properties from a hot wire.

[the] specific heat of plasma air, is it low enough to get extremely hot under normal current flow, which then quickly heats up the surroundings?

Yes and no, it is a combination of factors. A small arc, with small current, will dissipate a small amount of power, creating a (very) small heat source. While a big arc, with high current, dissipates much more power, creating a much hotter arc. It is easier for the big arc to melt metal, so would seem more dangerous - but the small arc temperature is still very hot; it is just so small that it can't do as much damage as quickly. So yes, the air does get very hot, but air isn't a very good thermal conductor, so the heat is localized to the area inside the arc channel. Transfer mostly occurs at the electrical contact points. For small arcs, not enough heat is transferred to melt the metal (it gets thermally conducted away faster than it can build up.) To produce "big" arcs and melt metal, high voltage is required, which adds another facet of danger.

Current does play a role in the temperature of arcs. While unpleasant, it is possible to touch a small 2MV Tesla coil without so much as (major) discomfort. There, the current is very low and frequency very high, so the current tends to travel over the skin rather than penetrate into it.

The lower the (oscillating AC) frequency, the more dangerous, with DC being the most dangerous to living things. DC arcs are also harder to extinguish and require more exotic methods than AC, since AC naturally reverses current twice every cycle, tending to help extinguish arcs.

Power substation arcs are especially bad, as they employ low frequency (60Hz), high-voltage, and high-current. Near the end, you can see molten metal (>900°C at least) dripping down from whatever the arc touched.

Sometimes the arc is able to extinguish itself through contaminating the conductors from vaporized materials or removing enough material to create a distance large enough to quench the arc. Sometimes it is not, and results in an oil-filled transformer venting in a fireball. But in all cases, (unintended) arcs are damaging and undesired.

rdtsc
  • 15,913
  • 4
  • 30
  • 67
  • 4
    "The lower the (oscillating AC) frequency, the more dangerous, with DC being the most dangerous to living things." Wait I'm confused. I thought AC around 50-60 Hz is the most dangerous because then spasm/fibrillation occurs? I thought DC isn't as dangerous because there is a single jolt instead of constantly changing direction, and high frequency AC isn't dangerous (ignoring heat) because nerves cannot respond above ~1000 Hz and the skin effect. – Agniva Chaudhuri Jul 15 '21 at 20:11
  • 2
    Never say never. While the erosion caused by the arc usually has the effect to worsen the situation, an unintended arc can get better if the arcing components are spot-welded by the arc. A well-known case where an unintended arc causes a spot weld and "gets better" in the sense of delivering better conductivity without arcing is a stuck contact in a relay. Of course, it isn't really "better" in that application, though. – Michael Karcher Jul 16 '21 at 08:16
  • @AgnivaChaudhuri that's correct, as can be seen from the standards for what's considered extra-low voltage: 50 V AC RMS, or 120 V DC. I don't know if 50-60 Hz is the peak of dangerous frequencies, but it's definitely worse than DC of the same voltage. https://en.wikipedia.org/wiki/Extra-low_voltage – llama Jul 16 '21 at 15:35
  • If a person were shocked by 240VAC, it will hurt but they'll be able to let go (due to the current reversing twice per cycle, giving the nerves/muscles time to respond.) 240VDC causes a "lock-up" of the nervous system, so you *can't* let go and will likely grip it tighter. – rdtsc Jul 16 '21 at 16:19
  • @rdtsc But the locking up also happens for AC. I have read many stories of people spasming and unable to let go of an AC wire. – Agniva Chaudhuri Jul 17 '21 at 18:24
  • True, but there is a *chance* with AC, and *no chance* with DC. – rdtsc Jul 19 '21 at 12:30
  • @rdtsc AC causes fibrillation which the heart can't solve. DC simply stops the heart, and it can restart. 1/50 or 1/60 second is not enough time to let go and pull away from the wire, and even if you start, the current will increase again and grab you. AC causes consistent spasming, whereas DC causes a single jerk which could throw the person away from the shock source. The human body has a capacitive property which somewhat impedes DC but which AC bypasses. – Agniva Chaudhuri Jul 26 '21 at 05:58
  • You're right, I surrender! – rdtsc Jul 26 '21 at 15:09
4

The whole point of arcs is that they are visible light and usually white is 4800’C whereas hot wires might only be 90’C. That temperature detonates the air molecules breaking down the interface with oxides. If there are any combustible gasses within explosive concentration levels, that’s the issue.

Tony Stewart EE75
  • 1
  • 3
  • 54
  • 182
3

Arcs generate heat in areas where you don't want them, like on a side panel near to your skin. The heat from current tends to stay in the vicinity of the conductors. Current through a conductor is a normal, controlled situation. Arcs are not a controlled situation (unless we're talking about high-power switching, in which case some arcing is expected and properly handled or protected against). An arc will produce ionised air which creates paths for electricity where you don't want it.

The heat from a small arc itself isn't the problem. Having an arc where it's not expected is a problem. The existence of the arc itself means something went wrong and the effect of the arc on the air increases the likelihood of more arcing. Copper or aluminium gets damaged, air is ionized and now you have a damaged system with current going anywhere you don't want it.

A small arc is like a little smoke. If smoke comes from an exhaust, it's an inconvenience. If smoke comes from a lightbulb or a discrete component, it's dying. The size has very little to do with it, it's all about where you expect it and where you encounter it. Arcing faults and arc flash effects are dangerous.

The heat produced by normal current is a normal situation to which the rest of the equipment is adapted. There's either passive cooling (radiated heat, possibly assisted by heat spreaders or other surfaces) or active cooling (mechanical ventilation, airco) available to deal with that. The reason that's never a problem (when done correctly) is because it's taken into consideration and calculated. Arcs usually aren't.

This is all assuming we're not talking about devices that arc on purpose, like a spark plug or tesla coil.

Mast
  • 1,263
  • 1
  • 16
  • 27