On a whim, I wandered down the road. It had stopped raining. About 50m down the road there was a dry patch of pavement and a smell of burning plastic. Another 100m down the road was the electricity substation, with a guy from the electric company who said he was trying yet another 300A fuse. He said that often the cable can be shorted out by ground water, but connecting it up again may heat it enough to drive the water out and make it fine. I asked him after breaking three fuses already, how he thought another one would solve the problem, and would he like to know where the short was? Took a bit of persuading to get him to walk 100m up the road to where the dry patch and smell of burning plastic was, and of course by then the smell had worn off a bit. Another electric company guy arrived with a smelling device, which they bashed into the tarmac to try to identify the location of the break, eventually settling with exactly where I told them it was.

Within a couple of hours after that, a team had ripped up the pavement, dug down to the cable, spliced it, wrapped it up in waterproofing, refilled the hole, and relaid the pavement.

But I think your story wins.

This has 3 important implications:

1) The return conductor (e.g. coaxial cable) may not be adequately-sized to carry the current you're asking it to, but your circuit breakers are rated in relation to the fixed internal wiring of your property, so they won't trip. This risks melting and subsequently breaking the conductor, leading to:

2) The exposed metalwork in your house becomes live. Screws at lightswitches, metal ceiling lighting pendants, copper water and gas lines, the works. All of it. This leads to:

3) An RCD (GFCI) or a device that contains an RCD (like an RCBO) will not save you when you come into contact with any of it, because the current is first passing back through that RCD on your internal neutral wiring, so there is no current imbalance from its perspective.

So yes, very dangerous. In the UK, we use concentric cabling for our street supplies, so it is very hard to accidentally sever the outer neutral conductor without also severing the inner line conductor.

This arrangement remains safe regardless of any single fault in any of the 4 wires. If the PEN conductor breaks but the local earthing still has decently low resistance, the internal PE protection that connects to the case of electronic appliances is still connected to the ground. If strong currents are drawn the voltage of the PE in relation to the real ground will increase, but the mains voltage will also fluctuate with the difference; so you might get a 20-30V drop in the mains voltage and the same voltage appears on the PE, not usually dangerous and you can see something is wrong with the power. Differential and overload protections still work since they have a safe return via the grounding rod.

If, on the other hand, the local grounding malfunctions, the PE is still secured to the PEN, which should be grounded at the nearest pole or local transformer. This is a harder defect to detect and periodic checks are recommended on the local ground, but at least you have single fault protection, and two things need to happen before a dangerous condition can occur (both local grounding and the PEN conductor must have a fault).

An adjacent benefit of having good local grounding in this configuration is cleaner and more stable voltage, since all groundings of all houses in an area are connected in parallel to the same PEN line, providing a path to the ground that might have a lower impedance than the PEN line feeding it.

Usually, our TN-C-S supplies are just augmented with PME, where the PEN conductor is earthed at multiple points along its run. This is almost equivalent to your description of a separate earth rod in every property, but it's on the main run instead of each branch, and it's not done for every point (e.g. house), nor even every other point or perhaps not even every 5 points.

This usually saves you from the consequences of a PEN conductor fault if it's broken along its run somewhere (because there's still an earth rod after the break), but it can't save you if the neutral is broken after it splits off to feed your property, and it can't save you if your property is the last on the line.

Where we have separate property earthing rods (either as the sole earthing arrangement, as in a TT earthing arrangement, or instead, in addition to a TN-S or TN-C-S supply, for example in an outbuilding containing extraneous conductive parts like copper water pipes), these are not permitted to be connected to a TN-C-S supply earth because practically no-one else in the street will have such a setup and then a broken PEN conductor sees your earth rod carry, perhaps, half of that street's supply current, which it could obviously never safely do.

I imagine we don't mandate supplementary earth rods in TN-C-S installations here because:

1) PME is common, reducing the impact of a PEN conductor fault, and

2) As I said, TN-C-S installations here use concentric cable, which is very hard to break in such a manner as to cause a PEN conductor fault without also killing the line conductor, rendering the fault moot.

If no one on the street has an earthing rod and the PEN breaks, then all the PE internal wiring of all houses will be connected, via transformers, light-bulbs etc. to the live wire. So every grounded appliance in the area is live. This is such a dangerous scenario that a grounding rod carrying half the current of the street is the least of your worries. All differential protections are also disabled.

Additionally, houses will typically be spread around on a three phase supply with the PEN carrying the disequilibrium current. If PEN breaks uphill from such a distribution cell then you end up with two houses connected in series to two phases, giving a voltage √3 = 1.732... higher than the mains (208 vs 120V in US or 400V vs 230V in Europe). The voltage will spread depending on the resistance of the two houses. So if a house has large consumers it will see a very low voltage, while it's series pair might see a huge voltage spyke. This is an incredibly dangerous scenario from a fire hazard perspective. A similar thing can happen in a 120/240V single phase system.

So, if you want good, solid, safe voltage in TN-C-S systems, a good grounding (preferably less than 1 Ohm) should be connected at the separation point between the TN-Combined supply and your internal TN-Separated wiring. This guarantees your differential protections will work regardless of what happens uphill.

Of course, if this is an uncommon practice in the area and not allowed by local code then that's that. But I wouldn't feel very safe without, at a minimum, a ground fault detection relay (which has it's own mini-earthing rod and detects when PEN floats at a too high voltage), and also a over-voltage protection.

PEN faults are very common, it's not a question of if, it's a question of when. Coaxial supplies are used around here too, with recommended local earthing and mandatory ground fault detection relay if local earthing does not exist. Still, it happens that lighting destroys both the central conductor and the electronic ground relay, yet line voltage still finds a way to shock people (you don't need a very good connection for that). Other times, the aluminum-steel core of the coax breaks down. Other times, idiot electricians work on the network and create PEN faults, exploding all the devices of a certain home by exposing them to three phase voltage.

I agree, I even cringed a little bit when I wrote it. But I didn't want to fully give everything away so it was just a bad edit :) Your suggestion would work a lot better.

Here in Australia the coax connection from the street typically goes through a galvanic isolator to avoid this situation, rather than being bonded to the earth.

http://www.yandina.com/galvanicIsolator.htm

I had a vanilla problem of power from neighbor's faulty set top box that was backfeeding into the cable network and killed my modem (and I think some other people's) one day.

I was told by the ISP's technician what had happened, I assumed the modem died by itself (i.e. bad OS) since it wouldn't fully turn on, only shows the power LED. In reality the tuner and some electronics probably died.

I tried calling electricians but couldn't get anybody out quickly enough so I ended up calling the power company. They dispatched a guy with a small truck with a short cherry-picker.

He was able to diagnose the problem as branches rubbing on the power line. He trimmed around it so I could turn the power back on. Our house was built in 1914 and the tree wasn't there. Somewhere over 100 years the tree in our neighbor's yard was planted and grew large enough to interfere with the power line.

While we were gone the power company replaced the connection with a shorter, more direct line that avoided the tree entirely.

As another poster said, I think your story is better because of your diagnosing the problem yourself.

A few months ago, my house started having intermittent failures.

It turns out that some road works caused one of the phases in my house to leak into the other phase, which caused a whole array of problems, but only in specific areas of the house.

We discovered this by piping 40kV into the wiring and listening to crackling in the ground, where we found the faulty cabling.

Here's a good video about it: https://www.youtube.com/watch?v=CsXUuuV4_sw

I had an electrician come and perform this step. The house was disconnected from the source, and then a "thumper" was used: https://megger.com/portable-fault-location-system-stx40- 2000

Is there a good explanation for how this can be the case in a developed country?

My city also has regular brown outs, power outages, etc. that have greatly inconvenienced my remote work (I have done a lot of work recently over a mobile hotspot). And it's not a HUGE city, but it is sizable (600k metro, ~800,000 in the metro + surrounding areas).

But for how this can be the case: It's as another commenter says. The provider is a duopoly in my area so I have no choice. Even worse the provider doesn't have many employees who do this work, they mostly hire 3rd party consultants to do repairs/installations of infrastructure.

Since his story is worded as if it happened recently, it could be COVID-related.

Back in April I tried to get my cable company to install its home security system where I live. I was told no, because the company wasn't sending techs into anyone's homes until the virus let up.

I tried again in August, and was told it would be a very long while, as the service guys were still catching up on the backlog.

This was certainly an issue with BT-OR in the UK for a short while when lock-down first properly started and there was a combination of OR people not being able to work and home users discovering issues (or reporting those that until now had been to minor to care about) as they tried to work from home. That caused quite a back-log of work in some areas.

A deregulated ISP market would be a huge improvement for this country!

("regulations simply cease to exist" seems like the casual definition for "deregulation," though the actual process of deregulation is for a bunch of people to pass a new set of regulations and if they're pandering to a certain constituency, they will call it "deregulation," regardless of the effect of the regulations they just passed. See: telecom, the airlines, banking, etc.)

[1] https://en.wikipedia.org/wiki/Natural_monopoly

There are some areas that might be a natural monopoly anyway due to low density population, but only way you could know is open things up. That's a tough pill to swallow since the franchise "fees" (taxes) are guaranteed so long as the cable company has a guaranteed monopoly. Yes they could pass a generic tax on cable service, but that would be a new tax.

I focus on cable companies because that seems to be the most common way to get broadband to many parts of the US.

But in 2020, wireless ISP work great wherever they're not regulated away.

One big problem with regulations is that even if they're beneficial when enacted, they will remain in force for many decades after the underlying reality has changed.

Because in CURRENT_YEAR we don't have to worry about rain fade or quirks of local terrain or unintended interference?

That's quite a big statement to make, generalized over the whole world like that.

We tend to avoid principles-based regulation in the US, but it's not destiny that regulations remain the same forever, even with our system. We certainly don't have to choose between that and not having any regulations at all (which isn't even an option when you're talking about radio spectrum, or use of wires at the last mile, or whatever).

For radio spectrum you don't need any regulations or regulators. Just establish that radio frequencies in a certain area can be owned, and use general property right rules.

This is quite different from laws protecting property, where all the state does is settle property disputes.

In slogan form, one is "rule of man" vs "rule of law".

Regulation deals with the here-and-now.

There have recently been well funded pushes into satellite internet such as Starlink and Project Kuiper.

I'll admit I don't know the details well.

I'm assuming a serious plan to run wires above. That is you have found deep packet backers to fund you; you have found qualified people to do the work; you have figured out all the other details to do this. There are good reasons to stop someone running cat-6 wires everywhere, but here I'm assuming you have figured out the right way to do this - whatever it is.

No, agreements between the companies ensure that only one option is available to each region. My city can offer Time Warner the right to run cable here whenever they want, but Time Warner will refuse to do so, because Time Warner has a deal with Comcast that says they will never offer cable in my city. There is no legal barrier to new cable startups like Google Fiber, only the immense cost of it and the threat of retaliatory blacklisting and throttling from the cable monopolies.

That's why the only places these companies compete are ones where the cities have decided to set up a local public broadband to buck the monopoly, or a very well resourced competitor like Google has decided to enter the market and refuses to play their anticompetition game.

As an aside, it's funny, I'm not sure what qualifies for me as "the telephone company" anymore. I guess there probably is one where I live, technically, but it doesn't seem to factor in.

edit: this is evidently going to be one of those posts that fluctuates in "points" over time, bouncing up and down. It'd be fun to have a sparkline implementation that illustrates this change over time

At present the only example of governments running - or at least owning - communications utilities in the United States that I'm aware of are municipal broadband [1] providers, which tend to be much more well-liked by their users than something like Comcast. (and bitterly opposed by The High Priests of The Invisible Hand, who want competition but not that competition)

Of course, other countries have state-owned communications monopolies, which actually run the gamut in terms of quality and cost.

[1] https://en.wikipedia.org/wiki/Municipal_broadband

I think it started with old cable systems where people would connect their cable themselves without paying for service. Then when the techs would go to install service they would see if any of the nearby houses were hooked up without being on the customer list, and cut their cable.

A friend of mine had her cable cut, probably for the same reason. Unfortunately the cable went to her satellite dish, not comcast.

(I guess it could have been competitors too)

When the neutral was cut on the service line, the return path for the mains current became residential ground->cable line->neutral in cable trunk. Since coax drops aren't designed to handle residential mains current, the cable began to heat up and melted its insulation.

Pedantically (but with the justification that this is safety-critical): Electricity always takes every path, just in inverse proportion to the impedance of the paths.

If a fleshy 1000 ohm resistor touched that service ground, you'd get some electrons there too, and possibly enough to kill.

This is one of those things that are bloody obvious in retrospect, but people usually don't think about it at all

I understand how electricity works well enough to handle it safely - and, at least as important, to know when I can't handle it safely and it's time to call in a professional - but that's not the same as trusting it. Honestly, it's a lot more the same as not trusting it. I can and do prototype line-voltage filament LED lamp designs on my bench without a problem, but I think more than anything else it's maintaining a lively awareness of the danger involved that helps there not be a problem.

What's really scary is stuff that's as dangerous as those prototypes and doesn't look it - a condition that I think only ionizing radiation excels electricity in making it possible to create. All anyone has to do is screw up wiring an appliance with exposed metal parts, and the whole thing can be live at potentially lethal mains voltage - and I'll freely admit that having had my batteries very unpleasantly charged that way once, thanks to a miswired ceiling fan, strongly informs my perspective on the subject.

As with fire, so with electricity: an indispensably useful servant, but also an ineluctably dangerous one, to be regarded with respectful mistrust at all times lest a moment's carelessness end in tragedy.

Or for extra fun, the wiring in the house. Lots of people think things like changing a plug are "easy" (they are), but they're also really easy to screw up.

Now you don't need some DIY'd appliance. Your Ikea stuff is now potentially deadly.

The number of DIY jobs I've run across where the hot/neutral are reversed on the plugs... It's insidious because for all intents and purposes it will _work_. Nothing you plug in will have any issues.

But that nice shiny metal lamp you've got? It's got the neutral bonded to the body so that if the hot were to come loose and contact the body it has a path to ground. What happens when the neutral and hot are reversed? The body of your lamp is now live, and when it's off it has no path to ground. Wear rubber shoes when you're turning your lamp on and off.

Your toaster? It only switches the hot. So when your toast is done and you stick your metal knife all up in there to try and dig your toast out, it's all still live with no path to ground. Whoops.

A lot of assumptions that are essential to things being safe all fall apart as soon as someone puts those black wires under the silver screws.

Another commenter pointed this out, but I have to chime in as well: no. Not true. The neutral is NOT bonded to the exterior body of a metal lamp. Not even on old fixtures built in the 1960s.

The old fixtures, if they were metal, were often not grounded, i.e. the lamp might not be connected by a 3-prong plug (hot, neutral, ground). But the neutral was definitely not bonded to the metal body.

If you inspect a new fixture, typically this means: it is either constructed with no accessible conductive parts, OR its metal case is bonded to the ground (not neutral!) and it has a 3-prong plug, OR it has a step-down transformer that converts to lower voltage before it enters the conductive lamp.

*

The thing that does happen when switching the neutral wire on a lamp is that doing so will energize the more-accessible "outer ring" of the Edison female receptacle rather than the less-accessible "center pin". This means that if you change the bulb with the switch on, you'd be more likely to contact the hot "outer ring".

Also, in most cases, the switch is only across (the wire that is supposed to be) the hot lead-- with neutral and hot swapped at the plug, you're not only energizing the outer ring of the light bulb socket, but also leaving it energized when the switch is off.

This leaves a shock hazard exposed even when the switch is off-- the lamp still won't light under normal circumstances because there's no path to ground with the switch off on the neutral line, but if you provide it one (your body) by touching it, you'll still get a shock regardless of the state of the switch.

Yeah, comment kinda got downvoted into oblivion for a slightly off example, but this was kind of my point. There are a lot of appliances that become significantly less safe when you put the power into places it's not supposed to be. And the danger is non-obvious.

oops, yes, good point.

That would be illegal where I live, and I think in most of the world too. In fact, on a standard residential outlet here, it would immediately bring the house power down.

Either the body is grounded or it's floating. It should never be linked to the neutral.

How?

They wouldn't trip in this case just by plugging a poorly wired appliance in. They detect an imbalance in current flow between the hot and neutral wires. If they detect more current going out than coming back in, they recognize that the current has found another path (which is bad) and trip.

Unless you have a grounded lamp (which is certainly not common here; I've never seen one) there'd be no alternate path until you provided one by being shocked, so there'd be no immediate trip. It would, however, increase safety by tripping once you were being shocked instead of simply allowing you to continue being shocked.

For example, if someone did indeed have a metal-cased grounded desk lamp whose body was also connected to neutral too; it would trip as soon as you plug it in, whether you turn it on or not, and you won't be able to turn the RCD back on until you unplug the lamp [1].

In the US they're called GFCIs as you point out, but this name is bad, since it implies that it only trips on ground faults. In reality they trip on /any/ current imbalance, regardless of how it's happening, over which conductors, or in which direction it is occurring (after all, we are talking about /alternating/ current here). Among other things, this means that if you take a line conductor from a non-RCD-protected circuit, and borrow the neutral from an RCD-protected circuit to power a load, that RCD will still trip, even though the fault isn't on the circuit(s) that the RCD was installed to protect.

As to tripping should the neutral-connected case develop line potential, a circuit breaker will immediately (< 40ms) trip anyway. As to tripping should the grounded-case develop line potential; circuit breakers will also do that if your earth fault loop impedance is low enough (and if it isn't, you'll have an additional single time-delayed master RCD for the entire installation, for fault protection anyway). In a properly designed, installed, and maintained electrical installation, such a fault will simply not exist for long (40-500ms), regardless of whether you have RCDs for additional protection or not. This makes it highly unlikely that you would subsequently come into contact with that mains-potential object; but on the off chance that you do, say, if it only develops a line to earth fault when you're touching it because it wasn't turned on yet; indeed, an RCD for additional protection is a nice thing to have.

[1] Sometimes, depending on design, one can manually hold the tripping latch up to prevent it from disconnecting even when it detects a fault. This is extremely bad design in my opinion, and is not the case with UK RCDs; the tripping mechanism is entirely internal, and even if you epoxy or padlock the latch in the on position, it will still trip correctly. I've seen some circuit breaker designs that suffer this flaw too, but again not in the UK.

Thanks to Big Clive, I have a vague but reasonable sense of what UK electrical systems are like. It's not safe, metaphorically or literally, to assume the US is up to the same standard. As you might find out the hard way if you did make such an assumption, North American electrical systems in a lot of ways are shockingly poor.

You can also get circuit breakers with integrated GFCIs (which is what we call RCBOs); those are usually double-width devices in the US (For example, with the Siemens "QF" product range [1]), however, again, single-width versions are available (For example, with the Square D "QO" product range [2]).

Because a GFCI needs to measure current flowing in both directions, it has an inherent need for a neutral connection too. This is what the flying lead is for. The circuit neutral would be terminated into the GFCI instead of the panel neutral bar; the flying lead is terminated into the panel neutral instead.

I'm also not assuming a ground; a neutral-bonded lamp body sitting on a conductive surface which is in contact with the mass of the earth would still be enough to immediately trip a GFCI under the same circumstances as my previous post, regardless of whether someone is touching it or not, and if it is turned on or not (assuming the switch is in the line conductor).

EDIT: However, this all goes to crap if the lamp only has a 2-prong plug, yes.

I agree that some aspects of US electrical installations are exceedingly bad compared to over here. For example, it is stupidly easy to shock yourself removing a live socket from a wall over there, compared to it being a basic impossibility over here -- yes, I know, safe isolation and all that, but name me a sparky that hasn't done this at least once.

[1] https://cdn11.bigcommerce.com/s-b8b6a/images/stencil/1280x12... [2] https://cdn11.bigcommerce.com/s-b8b6a/images/stencil/1280x12...