Electric shock from power supply (and how to fix it?)

[PRR]

> what the average (US) man in the street thinks mains voltage is; I've seen 110V, 115V, 117V, 120V.

Voltage is invisible. The average man has NO idea. Not counting pro electricians, I can only think of one guy I know who could figure it out, and he doesn't care.

As I said, the historic trend runs to 1/4V rise a year. In the 1930s it was around 110V, and the number lingers. 1960s saw more like 117V, another lingery number. Nearly all recent incandescent lamps (before they went out of favor) were rated for 120V. Filament lamps are by far the fussiest loads. 5V low and they run dim, 5V hot and they burn out too fast. I assume nearly all US homes had 120V (117-123) in the 2000-2010 period. Recently 123V-125V (no load) seems common. Sag is actually specified (if not enforced) and may be 3%-6% in most houses at 1/2 to 3/4 load. (My 15% sag at half rated load or max ordinary load is unusual.)

> the drop is in the house wires.

Should be 2% in the transformer if you and all your neighbors on the transformer are pulling heavy load. Drop-wire loss depends a lot on set-back: 500 feet (me) is too far, 20 feet (cluster housing) opens problems with short-circuit fault current magnitude.

> so many vendors.   We have large distances

I thot the UK grid was mostly nationalized? In the US, very little; but the various players mostly cooperate on technical workings.

Glasgow to London is 340 miles. St Louis to Boston is 1,230 miles, but 1,500 miles the way the large wires run (through Canada!).

> reluctance to bond Neutral to Earth at the property

As long as it IS bonded, somewhat nearby, I call it equal. In my cellar, at the meter, out on the street (though at 500 feet back from street, I already have significant difference due to earth current and line-drop).

Dirt-bonding may be a lot to do with protecting the street gear. Lightning happens. Good dirt-rods localize the damage. In the US, the utility's problems and the user's problems are two very different authorities; the UK may take the tack that if the utility is good, the users should be OK. And while I am not hip on utilities finding all problems in their dirt-rods, they gotta be better than the average user who does not know about a dirt-rod.

I too have pondered. One BIG fault in the US electric implementation is that devices (outlets, switches) are not required to be High Temperature material. In very old devices it was not an issue: many devices were ceramic and some paper-insulated lamp sockets were allowed to burn-up. But the rise of cheap plastic is now a plague upon our system. It is very common for overloaded outlets or multi-way strips to arc and burn with enough energy to ignite the house. In fact early "Arc Fault" breakers intended to mitigate this apparently really sensed the increased leakage of half-melted plastic. It also means our "#12" wire, which would take 25 Amps easy and is used at higher current in non-house wiring, can only be used as 20 Amp stuff because the devices can't take the temperature. We are stuck with the billions of installed devices for at least the next 50 years.... but why aren't we forcing better devices now, instead of the dubious Arc-Fault breakers?

[anotherjim]

Uk grid is privately run with multiple  private "suppliers" who are the ones who take your money.  A government regulator is supposed to keep them in check.

For all of the regulations that installers are meant to work to, does nothing for the quality of the equipment on the market. We have pendant lamp sockets that can't stand the heat of anything hotter than a 40w filament lamp (filament lamps are supposed to be banned here anyway). Socket strips with spot welded busbars that fall apart. Screw terminals that can't reliably clamp more than one wire and held in plastic walls that crack if you tighten the screws a bit too hard in an attempt to actually grip all of the 3 wires you have to fit in it.

[R.G.]

> what the average (US) man in the street thinks mains voltage is; I've seen 110V, 115V, 117V, 120V.

s I said, the historic trend runs to 1/4V rise a year. In the 1930s it was around 110V, and the number lingers. 1960s saw more like 117V, another lingery number. Nearly all recent incandescent lamps (before they went out of favor) were rated for 120V. Filament lamps are by far the fussiest loads. 5V low and they run dim, 5V hot and they burn out too fast. I assume nearly all US homes had 120V (117-123) in the 2000-2010 period. Recently 123V-125V (no load) seems common. Sag is actually specified (if not enforced) and may be 3%-6% in most houses at 1/2 to 3/4 load. (My 15% sag at half rated load or max ordinary load is unusual.)

Ah, the invisible hand of Georg Ohm. What most people don't consider is that they and their neighbors form a distributed and highly variable load. This load drags down the voltage on the power lines leading to them. It's less easily observable in metropolitan areas, but out here in the sticks, if everyone's AC is humming HARD and Tom is out in the barn welding while Betty Lou is inside cooking a roast in the electric oven while drying clothes in the electric dryer, things sag a bit near them. Especially if Fred is welding and Linda Lou is cooking and drying in the AC'd house just next door. I happens to be September or October and although it hits 80-90 in the afternoon when this is happening, but about 6:00, the laundry is done, the roast is done, and the welding is over and it drops back to the mid 70s and the ACs start cycling off. The family's TVs go on and have an AC line voltage that has bounced to 135-140 because the power grid sees a massive drop in overall power. The grid has to react and takes some time to get regulated back down. In the morning, the same happens, but in reverse. The utilities actively manage for sag and bounce, but can't fix it everywhere instantly. Long wiring lines make this even worse. And in the country, lines are looong.

There is some benefit that might happen in an isolated house to buying a whole-house ferroresonant transformer. They're big, loud, and heavy, but if you do that, you get tightly regulated power over a wide range of input voltages.

> reluctance to bond Neutral to Earth at the property

As long as it IS bonded, somewhat nearby, I call it equal. In my cellar, at the meter, out on the street (though at 500 feet back from street, I already have significant difference due to earth current and line-drop).

Dirt-bonding may be a lot to do with protecting the street gear. Lightning happens. Good dirt-rods localize the damage. In the US, the utility's problems and the user's problems are two very different authorities; the UK may take the tack that if the utility is good, the users should be OK. And while I am not hip on utilities finding all problems in their dirt-rods, they gotta be better than the average user who does not know about a dirt-rod.

I have a dirt rod at the pole, about 50 feet from the garage; one at the garage; one at the east end and one at the west end of the house. I also have massive transient clamps at the garage, and each end of the house. When we moved into the house, we had only the dirt rods and lost a $200 smart thermostat about once a year. As the saying goes, the first four times that happened, we thought it was a random accident  then I got serious and started spending money on transient protectors and not thermostats. Haven't lost a thermostat since. Also since then, the local electrical utility company has updated their lines to thicker, higher current and higher voltage transmission lines which in turn reduce sag and bounce.

It is quite difficult for the average Joe, Fred, Jane or Frieda to know enough about this to make any kind of decent decisions about it.

[PRR]

> they and their neighbors form a distributed and highly variable load.

Thanks for detailing your situation.

"All": somewhere I had a hint where R.G. lives, and peeked in Google Maps. Except for the trees (many here, few there), we live in "similar" places. The low-tax district outside a minor town. Ex-agricultural land (logging here, grazing and farming there) re-purposed for low-density houses when house-tracts became more profitable than logs/cows.

Yes, the wires are l-o-o-o-o-ng.

Longer than they "need" to be. Here, and I think near R.G., a big transmission line runs high above woods/fields, going somewhere else, not tapped to local loads.

Nevertheless, details (like Ohm) matter. My street seems to be 125V SOLID. I have both Volt and Amp meters in my cellar, so I know when I am low-load and can estimate my hi-load drop. It always works out to 125V at the street- no, one hot day last summer it was 123V all afternoon.

This despite the fact that our neighborhood lines are not built-out. There's 3-phase around the peninsula, but only 2 transformers each end of the loop, on a platform sized for 3 cans. My crossroad is one phase. They could add the missing transformers and run 2 more lines over my road. All this was done 35 years ago. They musta been told there was going to be a huge housing boom out here; didn't happen, because even the minimal network appears to give low-low losses.

Three near neighbors and I have arc welders, though we have switched from the big tombstones to modulated switchers, and don't do 48" pipe.

OTOH, I do know places the farmer put up a wire, tapped it when he sold 5-acre sub-plots, and when that burned down the local utility bought it out but did minimal upgrading. Everybody on the line knows when your well or dryer starts.

> It is quite difficult for the average Joe, Fred, Jane or Frieda to know enough about this

Indeed.

> lost a $200 smart thermostat about once a year.

For a while I was hoarding mercury thermostats--- never fail.

The $69 Honeywell serves fine for 3-4 years-- we are not thermo-geeks.

In another place I was losing a modem or two every summer. The place was once overhead power and phone, they put in underground wires, but only the power was moved-over. Lightning tickled the overhead phone, passed out the underground power..... THROUGH my modem. For a while I had a knife-switch to break the phone line when not using the modem. Finally a BIG hit took out the antique telco protector block, the telco guy did a "WTF?", and switched the phones over to the underground line that had lain idle. No more trouble after that.

[amptramp]

One more issue with measuring voltage: the voltage is rated as the RMS (root mean square) voltage which is the equivalent DC voltage that would cause the same heating when driving a resistor.  But people tend to use peak-reading meters calibrated in RMS values.  All is well until the waveform changes from pure sine to something a bit more ragged like the glitch that occurs with some fluorescent lights above and below the switching point and various loads like switching power supplies.  Then the meter needs to be a true RMS rather than a peak-reading meter calibrated in RMS.  I get 121 VAC in my area in Mississauga but the peak-reading meter may be responding to changes in waveform as much as voltage.  So I really don't know what I have, nor do most people even if they have a typical DVM.

[PRR]

Older DVMs tended to read the Average.

Nowadays "True RMS" is fashionable.

I assume you have the bench chops to make a few wave-forms to see what your meter calls it. (We are of course getting very far afield of that Average Consumer.) Aside from peak-catchers and spike wave forms, quite large differences in shape make smaller difference in measured Avg/RMS voltage.

It has been so very long since I had a 'scope on a power line (through say 120:6V transformer!) that I do not know if they can have significant spikes routinely. (Unless there is a tombstone welder on the line.) It takes large energy to spike-up a utility supply.

[thermionix]

One more issue with measuring voltage: the voltage is rated as the RMS (root mean square) voltage which is the equivalent DC voltage that would cause the same heating when driving a resistor.  But people tend to use peak-reading meters calibrated in RMS values.  All is well until the waveform changes from pure sine to something a bit more ragged like the glitch that occurs with some fluorescent lights above and below the switching point and various loads like switching power supplies.  Then the meter needs to be a true RMS rather than a peak-reading meter calibrated in RMS.  I get 121 VAC in my area in Mississauga but the peak-reading meter may be responding to changes in waveform as much as voltage.  So I really don't know what I have, nor do most people even if they have a typical DVM.

I'm glad this was brought up, I almost asked about it the other day.  My meter is a Fluke 73, it's not true RMS, so I always figured the AC readings were probably incorrect.  I don't have a scope or signal generator to conduct my own tests.

FWIW, I just read 123.8 at my wall socket.  It's nice and cool outside, so nobody is running AC, and probably most are asleep.

[R.G.]

For those lacking a true rms meter, but lacking the funds, Analog Devices AD8436 is availble in a SOP package for $11 from Mouser; the LTC1966 is available for $15 from Digikey, and uses an internal A-D conversion and computation of the RMS value, so it has low and constant errors over its whole range. You'll need a suitable attenuator/divider as they only take in aboutu 3Vrms without damage, and some damage control parts for when you set the divider wrong, but you could conceivably make one of these as an add-on to your multimeter for $30-$40.

However, the AC mains line is never far enough from a sine wave to make all that much difference. If you're measuring square waves or sawtooths, yeah, you need true rms. Doesn't happen all that often.

[Rob Strand]

Voltage is invisible. The average man has NO idea. Not counting pro electricians, I can only think of one guy I know who could figure it out, and he doesn't care.

Thanks, I've learnt something here. That's an eye opener for me.

I thot the UK grid was mostly nationalized? In the US, very little; but the various players mostly cooperate on technical workings.

I'm in Australia.

[Rob Strand]

However, one of the worst fault conditions is the 3-ph transformer Neutral feed breaking open somewhere downstream, although ground at the transformer can be expected to remain good. That allows interaction between all 3 phases according to load and the Live-Neutral voltage at a property can be anything from 0 to 415v. I suppose this "floating neutral" condition is mitigated if your Neutral is individually grounded? A danger of this is, as the Earth bond is only in a little part made of wire, large currents passing via Earth can cause a voltage lift at your Earth rod, and it isn't a safe 0v any more.

I guess a break like that isn't good in any system.

As a kid I was fortunate enough to see the insides and workings of many factories.  Neutral currents are considered bad.   Most factories use three phase equipment and any imbalance of the phases is a sign of a fault.   There's been a trend here over the last 25 years to so for houses to be single phase.  This is done for costs but also so the authorities can juggle the loads on the phases balance the phases.  Statistically if you juggle enough houses you can ensure the phases balance despite individual residence doing different things.

As for safety the different systems present different issues under different faults.   All houses here have a neutral connected to an earth stake at the house.  There is no ground wire along the street.  The wires are three-phases and a neutral.  The next layer up in the system drops the the neutral.

Having the earth connection at the house means you are likely to get the *highest* shock current if you touch the active because the impedance from the ground where you stand to the ground stake is relatively low, unless you live in apartments.   There's a good chance it will pop the RCD every time however we have had this system long before RCDs were on the radar.

I'm not sure if the street wires have both neutral and earth in the UK?    If you touch active in the UK I suspect you will get less of a shock than here because the ground impedance from the point where you stand to the earth stake is more distant.  However, it is variable depending on where you live.

A lot of discussions here about our earthing system often indicates is creates a shock hazard, and it does, however it also prevents hazards in other cases.

Here's a paper with some arguments.  Unfortunately is has a bit of bias to hospital systems.
http://www.sciencedirect.com/science/article/pii/S0004951414607601

[PRR]

> *highest* shock current if you touch the active because the impedance from the ground where you stand to the ground stake is relatively low

Insignificant. A "poor" ground is 100 Ohms. Your body is 100K down to 1K when wet (such as after a tingle begins to break-down cell walls). 1K in you through 100r in grounding, you still get 90% of full voltage.

To be safe under this theory you want ground resistance much higher than body resistance, say 1Meg. That does not happen in real life even if you try. (Yes, in a lab, maybe a hospital, you can stand/lie on a glass sheet. But in the real world everything leaks.)

[anotherjim]

We don't distribute a Neutral or an Earth until after the "street" transformer. 3 supply voltage phases are rotated along a route so adjacent properties single phase is different from their neighbours. Single phase feeder cables are LN + Steel Armour. The armour should only be connected to Earth from the supply end.

...But... Multiple Earth-Neutral bonding is coming in here too. And we have Gutshots!

This is what a UK utility may now install at a single phase property...

The incoming black (lower left) has had its armour cut back below in the gland. It looks a bit thin but the Brown & Blue "Meter tail" wires are a standard one-size-fits-all. Anyway, zoom in to read all the labels. Over the Neutral at the fuse/link is a cover with a "PME" label. That's Protective Multiple Earthing. It would seem that it's up the domestic installer to make any earthing connections to that.

Here's a 3 phase PME with the domestic earth installed...

There are a mess of standards here. Incoming fuses have the old UK Red-Yellow-Blue phase colour spots. Standard Brown meter tails used have had to be labelled to current standard as L1, L2 & L3. Outgoing phases from the MCB are the new (EU) colours. Brown=L1, Black=L2 & Grey=L3. I think the odd white fuse/link is a Neutral test point.
Note that the Yellow/green Earth wire is not from the incoming supply cable.

[Rob Strand]

Insignificant. A "poor" ground is 100 Ohms.

I suspect that is the case.  In anotherjims case the earth point could be 500m to 1km away but it probably doesn't affect the ground resistance.

Over the Neutral at the fuse/link is a cover with a "PME" label. That's Protective Multiple Earthing.

I wonder why they have decided to change after all these years.  Maybe your old system is prone to a single point of failure and ground conditions.

[PRR]

> In anotherjims case the earth point could be 500m to 1km away

Wow!

OK, that's "only" 1 Ohm or less. Because "earth" is an indefinite thing, I have no objection.

That is IF the power company does NOT have "neutral return leaks". They shouldn't. It is bad practice, and normally poor profit. (I am ignoring an odd Aus system which has no metallic return; I assume it is rare even there.) Every few years in the US, some power company "loses" a Neutral, or has obscene unbalance, and many Amps flows through the dirt. Then the "Earth" in a cable from 1km away could be many Volts different from the dirt at the house. Touching an "Earthed" electric box and the dirt could be a good tingle.

[R.G.]

Out here in rural Texas, often the feeder to one or another section of property (note here that "section" has the technical meaning of "one square mile") will be fed with a single hot line at ?? 13kV?? and feed transformers with a hot and earth ground. Accordingly, it's important that the "ground" at the transformer not be too far away from real, earth ground. This is at least partly accomplished by a thick bare copper wire running down the power pole and stapled in a spiral on the bottom of the pole before the pole is erected. At the low currents enforced by the multi-kV voltages, the volts per foot drop on the copper is negligible.

That is until the copper ground wire opens. So every pole has a copper ground wire. It forms a distributed ground connection that makes the real earth impedance quite low.

That is until another joker decides that the scrap price of copper is so high that they will steal air conditioner coils for their copper coils, copper wiring of any kind, and in particular, several people have been electrocuted, their bodies found next to the copper wire they were attempting to pull off the power poles. Not only are there harsh legal penalties (less harsh than death, anyway) but metal scrap yards now have to have positive identification, perhaps more than one form to accept scrap copper at all. It's so bad that in the Central Valley in California, people have been caught tying the electrical lines running out to the irrigation pumps to their pickup bumpers and driving off the pickup, copper wires trailing. I guess it's good that the people's republic of California fights hard for eliminating guns so guys like that can't get shot while committing their little misdeeds. But I digress.

Fueled by the enormous rise in demand for electronics, copper is rapidly becoming so precious that we'll probably see initiatives to prevent its use for electronics. The motivation is too great.

[PRR]

Here in Maine, someone broke the fence on a *sub-station* and managed to take all the copper without being killed. (Suggests insider training.)

Yes, axing the copper off poles is popular here. When I bought a house this year, I went to both adjacent poles and checked. They looked grounded last week.

Yes, some ID is needed to sell copper. To sell copper pipe scrap I needed my driver license and had to take a check (paper-trail). I think specifically for pole-length wire in a gauge range, the offerer SHOULD be required to show a Work-Order, a Disposal-Order, and payment direct to the utility; but there's a considerable run of private lines which "could" be salvaged by the owner.

One of the wire companies is selling "branded" wire. Stamped in the wire every few feet is Your Company Name and a running number so you can track where it was installed (stolen from). This would go with a stern letter to scrap dealers not to take such wire without ample documentation. I dunno if it has caught on. What I saw was just pole-bond, but I found span-wire of similar idea.

The price of Copper is up. But the surge in theft seems to be driven by drug users. A 2007 study blamed crystal meth, but I am sure it makes very little difference.

Pole bond wire *could* be steel. It would have to be fatter, but space is not critical. Steel is cheaper than copper. Also harder to cut. Corrosion can be controlled with thin copper coat over steel (CopperWeld), though this "looks like" copper enough to fool a druggie.

All of this mostly protects the utility company equipment from lightning. For indoor non-concrete playing, low hum mostly depends on G-N bonding, not bonding to dirt.

[thermionix]

Last apartment, about six years ago, woke up and the TV was out.  Upstairs and downstairs.  What the hell?  Went outside, and somebody had chopped the exposed coax, maybe 15 ft at the most.  What's that worth as copper scrap?  25 cents?  How much meth can you buy for that?

Air conditioner at the guitar store has been hit several times.  Landlord finally built a 2nd story platform for it.

Damn junkies.