31
\$\begingroup\$

What's the point below which electrical current is generally considered safe for "casual" human contact?

Is either voltage or amperage more "dangerous" (e.g. high voltage / low amperage vs. low voltage / high amperage), or is the only consideration the total current?

\$\endgroup\$
  • 1
    \$\begingroup\$ Here's a start. It's a huge filed and there are countless books written about electrical safety and human contact. Basically, there is no "casual" contact. In short: current kills, voltage hurts. Frequency is also very important! \$\endgroup\$ – AndrejaKo Jan 19 '11 at 23:41
  • 3
    \$\begingroup\$ It also varies from person to person. Someone with cardiac arrhythmia is much more susceptible to being driven into fibrillation then someone with a health heart. \$\endgroup\$ – Connor Wolf Jan 20 '11 at 1:06
  • \$\begingroup\$ Related video: youtube.com/watch?v=8xONZcBJh5A \$\endgroup\$ – JYelton Feb 1 '14 at 8:56
  • \$\begingroup\$ This is very complicated. It depends how much current flows near heart etc. en.wikipedia.org/wiki/Electric_shock \$\endgroup\$ – Kamil May 15 '14 at 22:51
21
\$\begingroup\$

As rules of thumb, I've generally thought of myself as a 70 kΩ resistor to ground that feels pain at around 1 mA, which can be driven by 70 V or so. In my experience, the pain threshold is slightly above 48 V.

I can't say that I have any good medical science to back this up, but there are a few empirically obtained data points in that I'm not dead yet.

\$\endgroup\$
  • 6
    \$\begingroup\$ Thanks, and don't feel obliged to gather any more data points on my account! \$\endgroup\$ – Mark Harrison Feb 25 '11 at 23:24
  • 6
    \$\begingroup\$ That 70kOhm is very optimistic scenario (small surface contact). \$\endgroup\$ – Kamil May 15 '14 at 22:47
  • 1
    \$\begingroup\$ If your hands are soaked in salt water 70V will be rather dangerous \$\endgroup\$ – Dirk Bruere Apr 19 at 22:38
30
\$\begingroup\$

Here is an article titled "A review of hazards associated with exposure to low voltages" which I used as a reference when answering a medical safety question (I design embedded hardware and firmware for medical devices which go through FDA approval).

Because the body has a minimum resistance of around 550 ohms, to get enough current to do damage a minimum theoretical voltage level of around 16.5 Vpeak is required (corresponding to a current of 30 mApeak, which can induce respiratory paralysis if conducted across the chest for several minutes of contact at this low voltage). Based on the cases studied by the author, the single lowest voltage reported to cause transdermal electrocution in an adult is 25 volts.

For less than one minute of contact, currents >40 mA are required to cause ventricular fibrillation, corresponding to a theoretical voltage of 27.5 Vpeak. For less than one second exposure, >100 mApeak and 55 Vpeak are required. The author states that in all the cases he studied, there was no accidental electrocution from short-term exposures to voltages below 50 Vpeak.

\$\endgroup\$
  • 1
    \$\begingroup\$ Good article. I would add that the discussed in your answer is a very much worst case analysis. Very large contact areas, thin/wet skin, long duration exposure, abrasions, etc. Skin resistance can change considerably in response to such situations. For instance the article lists dry unbroken to have resistance of 100k-2M ohms and that under 12V little to no skin resistance breakdown occurs. This meshes with my experience doing electrical work when younger, 50-75V was not noticeable work work on live (hand to hand contact) unless your hands were sweating a lot. 120V hurts though. \$\endgroup\$ – Mark Jan 20 '11 at 22:07
  • 1
    \$\begingroup\$ What's the ring voltage on phone lines? 48v? Or is it 96v? Either way, it gives a nice sharp poke if a phone call happens to come in while you're working on live wiring... :-) \$\endgroup\$ – Brian Knoblauch Jan 21 '11 at 21:20
  • \$\begingroup\$ @Brian -- ringing voltage is around 90v AC at 20 Hz \$\endgroup\$ – tcrosley Jan 21 '11 at 21:47
  • 3
    \$\begingroup\$ In the US the ring is theoretically 90VAC at 20Hz, but i've seen it vary quite a lot, especially in rural areas I'd see 100-110V routinely. When ideal there is supposed to be -48VDC on the line, but again it can vary especially in rural areas where i worked. The ring voltage(90VAC) is imposed on the supervisory voltage (-48VDC) as well. When current is drawn the Tip goes down from ground and Ring rises, the current limit is roughly 20mA as a result of the system resistance. I've been zapped by ring pulses a couple times, it hurts but the current drops the voltage quickly. \$\endgroup\$ – Mark Jan 21 '11 at 21:53
  • 3
    \$\begingroup\$ But 12 volts HAS killed volunteers when applied across the chest, despite there being medical expertise on hand in case problems occurred. Once conduction occurs the effective impedance can be very low. \$\endgroup\$ – Russell McMahon Sep 6 '11 at 6:37
9
\$\begingroup\$

Voltage doesn't really matter, it's a requirement to get a certain voltage to pass through the skin, but voltage doesn't have any impact on "damage".

Current is what does damage.

I've heard tons of claims as to what will kill you. In EE school is was 60mA AC and ~100mA DC across your chest that would send your heart into fibrillation.

I've seen claims that < 10mA directly through your heart could do the same. Honestly both are probably correct. I don't know what a real electrical model of the body looks like, but I don't have a hard time believing that if there were 100mA running through my body from one hand to the other that only 10% would pass through my heart directly.

I've worked on live phone lines before (~58V DC with off hook) and that didn't pass through my skin initially. A half hour of being in the 105°F degree attic and sweaty hands later, it made my finger twitch and didn't feel good. On another occasion I was working on a phone line when someone dialed it... that sucked... the ring pulse is 120V AC (current limit though) and does not feel good at all.

It only takes a couple milliamps to seriously get your attention, 10+mA will lock up muscles, this is highly frequency-dependent though.

To get back to your point... greater than 100-200mA is when you'd expect to start to see flesh burning and things like that. But obviously from the heart discussion above, localized currents that are much smaller can be deadly.

I don't really know if there is a firm rule as to what's "safe". The current debate over the use of tasers, for example, would seem to indicate there isn't much conclusive evidence.

\$\endgroup\$
  • 4
    \$\begingroup\$ It's not current that kills, it's the fibrillation. The path the current takes through your body and the presence of vibrations that interfere with the heart are important. Saying "voltage doesn't matter; it's the current that kills" will only encourage people to look at the current rating of a power supply and think that somehow has something to do with the danger of it. Power supplies are almost always voltage supplies. They don't force a specific current. The amount of current that goes through your heart depends on your resistance and the supply's voltage. \$\endgroup\$ – endolith Jan 20 '11 at 20:52
  • 4
    \$\begingroup\$ @endolith "It's not current that kills, it's the fibrillation." Considering that its current that causes the fibrillation i don't get what your point is. Thats like saying "It's not the lack of breathing that killed you, its the lack of O2 in your blood". I get what your saying on the voltage vs current issue, but honestly if someone is daft enough to look at a supply's current rating and grab on to it there isn't much you could say that would save them. Also, I don't often see a supply thats limited to currents below what i stated are dangerous that outputs enough V to pass through skin. \$\endgroup\$ – Mark Jan 20 '11 at 21:42
  • 1
    \$\begingroup\$ @Mark Meaning that 1 mA of DC has a different effect from 1 mA of 60 Hz AC, which has a different effect from large spikes at 2 Hz that average out to 1 mA. There isn't a magical voltage that kills people, and there's not a magical current, either. \$\endgroup\$ – endolith Jan 20 '11 at 22:23
  • 3
    \$\begingroup\$ Current won't pass through your body without voltage. I've worked with continuity test equipment capable of flowing massive amounts of current and could safely hold the anode and cathode in my hands, due to the low voltage. So, saying it's the current that kills you, is like saying "It's not the lack of breathing that killed you, its the lack of O2 in your blood." Without a difference in potential, you aren't going to get any current. \$\endgroup\$ – bt2 Jan 21 '11 at 0:30
  • 1
    \$\begingroup\$ @bt2 If you wanted to build a safe "shock" device. You don't do so by limiting the voltage because we can't account for wide fluctuations in body resistance. We'd build it with a high voltage and a current limit as limiting the current is how we can guarantee limiting possible damage. As mentioned there may be a frequency component in the mix as well, depending on what you want to achieve. I doubt it matters much in a hand-shake shock device, matters more in something like a tazer if you want to lock muscles. \$\endgroup\$ – Mark Jan 21 '11 at 19:21
4
\$\begingroup\$

The HAM Radio Technician Class examination question provides this answer:

2010 Pool - Question T0A01

Category:   T0A - AC power circuits; hazardous voltages, fuses and
            circuit breakers, grounding, lightning protection,
            battery safety, electrical code compliance

Which is a commonly accepted value for the lowest
voltage that can cause a dangerous electric shock?

A   30 volts
\$\endgroup\$

protected by W5VO Nov 1 '13 at 23:44

Thank you for your interest in this question. Because it has attracted low-quality or spam answers that had to be removed, posting an answer now requires 10 reputation on this site (the association bonus does not count).

Would you like to answer one of these unanswered questions instead?

Not the answer you're looking for? Browse other questions tagged or ask your own question.