Electrical Services: Installing Lighting & L1/L2 Feeds

[sophiecentaur]

Yes. That's true. What bothers me is the appliance lead is not protected for a fault on the appliance because it is upstream of the appliance and can be supplied with far too much current from upstream. A fused plug (yes, only with the right fuse) gives the highest protection. I already made the point that moulded plugs with the original fuse constitute the majority of loads these days. Few people will "tinker" with a plug - what reason would they have when there is nothing 'mendable' about any modern appliance?

 

[krater]

What bothers me is the appliance lead is not protected for a fault on the appliance because it is upstream of the appliance and can be supplied with far too much current from upstream.

Under the right conditions an overcurrent device is totally capable of protecting conductors upstream of the device.

Let that sink in for a second. Yes I realize it goes completely against the basics of circuit design. Nevertheless it is true.

Short circuits, ground faults, and overloads are all types of overcurrent condition. Should a short circuit or ground fault occur ahead of, or on the line side of an overcurrent device, the device will not actuate. Why? Because the circuit current never flows as far as the OCP, it takes a path back to the source without ever reaching it.

With me so far? Good. Now an overload condition is a special case. The circuit is operating pretty much as it is designed to, except the current draw is in excess of its design. So let's say an appliance is designed to run at 7 amps. It is connected to its supply branch circuit through a flexible cord, let's say it is 16 AWG which will supply 7 amps quite comfortably under normal conditons and in fact is only at about two thirds its rated capacity per NEC. But let's say this is some sort of motor load, maybe a fish tank pump or something, anyway it has a low volt control circuit supplied through a small power limited transformer. The control circuit passes through a thermal overload device set at 9 amps.

Now, somehow, your favorite goldfish manages to get the filter screen on the pump loose and swims into the impeller. Poor goldie, but you've got a bigger problem now as he has jammed your pump and intends to burn the whole residence down around you. The no longer turning motor is acting as a dead short, drawing 20 amps as it sits in a locked rotor condition. The branch circuit protection, back in the panel where the circuit originates, does not see a problem as the 12AWG wire is sized properly to carry this current safely for hours. But the cord supplying the pump, with only half this safe capacity, is on its way to temperatures well in excess of its insulation rating. This is not good, and is not acceptable by code for obvious reasons.

Lucky for you the designers of the pump system were smarter than your goldfish. They knew that even though the pump would almost certainly be supplied by a circuit capable of delivering current in excess of the cord's rating, should anything catastrophic happen to the pump which would leave it in a locked rotor condition, the heat created would trip the thermal overload and open the circuit. With the hot conductor to the motor broken, the current flow stops and the cord cools down; even though the cord was upstream of the overload relay, by virtue of breaking the circuit the overload has protected the cord.

As stated in my previous post, branch circuit devices at the panel would under most short circuit and ground fault conditions clear the circuit of power before any major damage could be done to the conductors. Yes, this sort of protection needs to be upstream, or ahead of, the potential fault. But overload protection is special, just as overloads are a special sort of overcurrent condition. I once saw a room full of 60-80 electricians, many of whom had more years in the field than myself, have this concept spelled out to them as a number of them sat in disbelief, recalling what they had been taught years ago about how OCP had to be located upstream of a fault. The instructor, who had 40+ years in just about every respect of the field as well as a master's degree in teaching, did a good job of illuminating his point to many who were resistant to the idea. This is a situation seen very often not just in appliances but also larger motor loads as well as things like welders, elevators, and many other applications.

The code is your friend, it has all the answers, if you can understand what it is telling you the power you wield is almost limitless.

 

[krater]

One thing to note here is this: A purely inductive load that draws 15 amps on one side and a purely capacitive load that draws 15 amps on the other side (can't imagine this in the real world) will put 30 amps on the neutral. This was discussed a while back here on PF. Can't recall what thread. A search will show it up I am sure. Jim Hardy was involved in the discussion as well as myself. It had never occurred to me up to that point that is was indeed possible to overload a neutral. But, it most certainly is.

Two words: Hair salon.

Blowdryers and curling irons plugged in at every chair. If you're going to network your branch circuits it is a good idea to pull that 10AWG neutral to go with your 12AWG hots.

 

[Averagesupernova]

Both blowdryers and curling irons are resistive so my example does not apply here. Not sure what you are getting at.

 

[krater]

Blowdryers have a motor. It may not be as much of a factor as the heating element but it contributes to the reactance of the circuit.

 

[MrSparkle]

the blowdryer's motor has a practically inconsequential contribution.

 

[sophiecentaur]

Under the right conditions an overcurrent device is totally capable of protecting conductors upstream of the device.

Let that sink in for a second. Yes I realize it goes completely against the basics of circuit design. Nevertheless it is true.
Etc.
.

Not "totally" but very often. I am not at all surprised at your statements about the protection provided by a downstream device. I am more surprised that those 80 electricians were surprised. But that protection can only sense and react to what's going on downstream of itself but how can it deal with the following? A common scenario is a damaged power cord - perhaps with a heavy box resting on it for days and days or maybe jammed under a closed door. Any intelligent circuit protection inside the appliance can hardly know about that. Eventually, the wear or pressure can cause a partial short across the conductors and let 20A flow through conductors rated at 5A. There is no protection without an upstream 5A fuse. However smart the protection device happens to be, it cannot disconnect at the socket and that's the only way of eliminating the effects of the fault.
You seem to have taken against the idea of fused plugs. I wonder why? At the very worst, a UK plug can have a 13A fuse - which is significantly less than the protection (32A) back at the board. In practice, these days, they are (a point I have already made) unlikely to be fused wrongly.
You seem to be claiming that every mains-powered appliance (including non electronic) on the US market has a sophisticated protection circuit in it. How true is that? Can you rely on all imports to be compliant? Would it also apply retrospectively to fifty year old appliances like lamps and small power tools? Even fifty years ago in the UK, there were fused plugs, giving protection to all power cords. I have not heard of a single imported UK style plug that was not fused.
One of the reasons that appliances in the US are required to have smart protection may well be that there is no plug fusing. Of course, a lot of UK equipment also has overheat and overload protection in it - it's a very sensible feature. But it can only do so much.
A point of information: what is the current available at a typical outlet in a living room? How many outlets on a typical circuit? Another question: what is the practice with multi socket extension blocks? Would you expect a fuse on its input? UK versions would normally have one, in my experience (in addition to the plug fuse, of course)

Things are changing. Almost the majority of outlets these days are to supply low voltage equipment and the Wall Wart is very common. That is a good solution UK or US to the protection problem - as long as the low cost imports can be relied on to follow the rules. How long before we can expect a 12V distribution system in all homes?

 

[psparky]

Not "totally

Things are changing. Almost the majority of outlets these days are to supply low voltage equipment and the Wall Wart is very common. That is a good solution UK or US to the protection problem - as long as the low cost imports can be relied on to follow the rules. How long before we can expect a 12V distribution system in all homes?

Probably not soon? You would need 10 times the copper in each house.

At 12 volts, a #12 (4mm2) wire would now need to be #2/0 (70mm2) to power a good microwave!

That would drive the cost of copper up as well.

But who knows...

 

[Averagesupernova]

A point of information: what is the current available at a typical outlet in a living room? How many outlets on a typical circuit?

15 amps. Code requires 2 20 amp counter-top circuits in the kitchen. A 20 amp circuit is required for a bathroom outlet also. Washing machine will require a dedicated 20 amp circuit. Code does not typically dictate how many outlets are on one circuit but there are some obvious exceptions as the above implies. Every 120 volt circuit in the house now requires an arc fault breaker excluding kitchen, bathroom, garage, dedicated circuits like washing machine and furnace. Any living room, bedroom, den/parlor and I believe hallway will require arc faults. Ground fault protection is required in any unfinished area such as garages and basements. Bathrooms also require it. Kitchens counter-tops within a given distance of the sink require it too. 32 amps at 240 volts equals an available 7680 watts at each outlet in the UK. In the US typically 1800 watts available. Is it any wonder the UK has a fuse in each plug?

 

[sophiecentaur]

Probably not soon? You would need 10 times the copper in each house.

At 12 volts, a #12 (4mm2) wire would now need to be #2/0 (70mm2) to power a good microwave!

That would drive the cost of copper up as well.

But who knows...

I didn't mean 12V for everything. I meant an extra circuit for all those little things that work off wall warts and USB outlets. They are pretty much in the the majority - not to mention the 12V lighting rail that's so popular. I wouldn't suggest a 12V, 2.5kW kettle! We had an LV supply on all the Physics Lab benches when I was at School. Massive great black box on the wall with a Captain Nemo style wheel and switches.

15 amps. Code requires 2 20 amp counter-top circuits in the kitchen. A 20 amp circuit is required for a bathroom outlet also. Washing machine will require a dedicated 20 amp circuit. Code does not typically dictate how many outlets are on one circuit but there are some obvious exceptions as the above implies. Every 120 volt circuit in the house now requires an arc fault breaker excluding kitchen, bathroom, garage, dedicated circuits like washing machine and furnace. Any living room, bedroom, den/parlor and I believe hallway will require arc faults. Ground fault protection is required in any unfinished area such as garages and basements. Bathrooms also require it. Kitchens counter-tops within a given distance of the sink require it too. 32 amps at 240 volts equals an available 7680 watts at each outlet in the UK. In the US typically 1800 watts available. Is it any wonder the UK has a fuse in each plug?

Reading all that, I can see the advantages of the ring system. A single 32A ring is easy to instal and feeds all of one floor - or even two floors. It's also very easy to modify with extra sockets when needed. Of course, it's swings and roundabouts but the higher voltage gives
you twice the power for a given weight of copper. Shock risk vs fire risk always calls for a compromise choice.
Protection in the UK consists of MCBs, rather than fuses in most places and an RCD covering the whole supply (and one for any shower unit, too).

The incoming mains fuse for an average house is up to 100A. A typical installation would have one or two lighting circuits, one or two rings, a cooker circuit, water heater and possibly a shower.

 

[Averagesupernova]

In the ring system how are the conductors sized? If it is protected at 32 amps at the service panel are the conductors able to handle 32 amps if the ring should become broken? If so, then it seems to me that copper is saved due to wire size only because of the higher voltage. In the US wires are fused at 15, 20, and 30 amps for #14, #12, and #10 wire respectively.

 

[sophiecentaur]

In the ring system how are the conductors sized? If it is protected at 32 amps at the service panel are the conductors able to handle 32 amps if the ring should become broken? If so, then it seems to me that copper is saved due to wire size only because of the higher voltage. In the US wires are fused at 15, 20, and 30 amps for #14, #12, and #10 wire respectively.

2.5 mm, I think.

 

[Averagesupernova]

Ok. So I would say that is a #10. I don't see how that is saving copper.

 

[sophiecentaur]

2.5mm copper is rated at 18.5A in the UK.
Ah, I may have spotted the misunderstanding - that's 2.5mm²
I was being sloppy and using the vernacular.

 

[Averagesupernova]

Ok. That's different. So if part of the ring is broken, which could go for years unknown, the wire can be regularly overloaded.

 

[sophiecentaur]

Yes. So it would; you certainly have a point there. But testing the system involves testing continuity of the ring.

 

[Averagesupernova]

Ok so in the UK are there regular 'tests'? Here in the USA there are inspections at the time of install and that is it. I can see some advantage of a ring. I also see the potential of disadvantages. I would think that there would be more reason to use it on a low voltage system. I look at it this way: Break the ring at mid-point, fuse each half at 15 amps. 15 amps on a 240 volt circuit is double the power available compared to the USA. The one main disadvantage of the system in the USA is losing the neutral before the service. This gives potential of putting 240 volts on some 120 volt appliances. In the USA typically our transformer is located closer to the residences. As I understand in the UK they will run secondary wire up to half a mile. Is this the case? I doubt secondary wire would ever be run that far in the USA.

 

[psparky]

2.5mm copper is rated at 18.5A in the UK.
Ah, I may have spotted the misunderstanding - that's 2.5mm²
I was being sloppy and using the vernacular.

Amazing how the same size copper wire safely conducts different amps in different locations.

In Chile Factories, 2.5 mm² copper wire is rated at 20 amps.

If I take this same wire and install in USA, it only safely conducts 15 amps...and apparently 18 .5amps in UK.

Perhaps different types of copper are used? Perhaps copper behaves differently in different geological locations? Perhaps they are using different temperature rated insulators? Perhaps some are in conduit...perhaps some in open air tray? Perhaps the definition of "safely conducts" or "rated" differs from country to country?

Fascinating, indeed.

 

[sophiecentaur]

Amazing how the same size copper wire safely conducts different amps in different locations.

In Chile Factories, 2.5 mm² copper wire is rated at 20 amps.

If I take this same wire and install in USA, it only safely conducts 15 amps...and apparently 18 .5amps in UK.

Perhaps different types of copper are used? Perhaps copper behaves differently in different geological locations? Perhaps they are using different temperature rated insulators? Perhaps some are in conduit...perhaps some in open air tray? Perhaps the definition of "safely conducts" or "rated" differs from country to country?

Fascinating, indeed.

This is true because the environment of the cabling makes a difference to its operating temperature. It should be rated according to the routing - for instance, cable wound on a drum can be an embarrassment and they have thermal cutouts to protect it. I do not know the regs in detail but you should not swathe cable in fibreglass insulation but give it an air space. There is chapter and verse on all of this stuff. Many houses in the US are insulated much better than UK houses - which may account for the different rule-of-thumb figures.

@Averagesupernova: I take it that you really don't like the ring main system. Fair enough. It is bound to have its disadvantages - along with the two / split phase system.

The basic US and UK distribution systems are radically different and this, I am sure, is due to the different spacing between your average houses in the two countries. The UK does not give everyone their own transformer. The standard urban layout is for a 'large' 3 phase transformer and the three phases are taken along a road with houses fed 1,2,3,1,2,3,1,2,3 along the road. (Essentially a LV distribution) Only under special circs would a house be fed with three phases. One transformer would feed dozens of surrounding houses. There can be problems with progressive voltage sag along the path. I have experienced over-volts, right next to the transformer. I created such a fuss because lights kept blowing that they reduced the volts (I had a chart recorder installed for a week and they told me there was nothing wrong - but I read the chart and they did reduce the volts). I have also experienced low volts and lights going dim when the cooker was turned on. Clearly, one big transformer is cheaper but individual feeds from a long Intermediate Voltage line is necessary when the spaces involved are great.

The two voltage system in the US means that you can use many European white (power) goods with not much modification, I presume. But what about your 230V fusing? Are devices double pole fused? I remember getting US equipment that was double pole fused in the 60s and we had to modify it all to single fuses in the L feed. I can see that neither system is perfect.
We are, at least, not coming to blows about this. (Except I know my system is better than yours! )

 

[krater]

Aha, so your 2.5 mm wire size refers to area, not diamater? I was also under the conception that we were talking about a 10AWG, not a 14AWG.

So you fuse a 14 at 32 amps? Jesus! I realize that it is ring connected and ideally should not be a problem but my concern would be exactly the problem described above, if one side of the ring goes open the circuit may continue to work just fine, however it is drawing the entirety of the load down one leg which is regularly overloaded.

Splices go bad more often than one might like to think. Over here a bad splice, even when it happens on a neutral which leads to stuff taking both lines and burning up, tends to have the result of someone getting a call to service it. But if everything keeps right along humming who's to be the wiser until insulation starts to melt? I'm reminded of one old electrician who once told me that it takes a good 50 or 60 amp load before a 14AWG copper actually vaporizes. I didnt ask him how he knew that.

psparky, you certainly don't have to travel the world to see a 14AWG wire have multiple different current ratings. In fact, you can concievably see several different ratings even in the same structure. The conductor run in MC feeding an outlet next to the panel has to be on a 15A breaker. The other cable which feeds the air conditioning unit outside can be on a 20A breaker. The 10AWG in the inch and a quarter conduit which runs out of the panel along with nineteen other hots and neutrals might only be allowed on a 20A breaker as well, while the one feeding a motor might be connected to a 70A breaker and were it be used in a motor control circuit could legitimately be protected by a 150A fuse.

Sophie, overcurrent devices here are required to be installed in any ungrounded conductor. If the circuit has two then yes, it has to be a double pole OCP. I'm not sure how things work in other parts of the world, but one important point when it comes to electrical codes here is that they are NOT retroactive. You can install something legally by code that will be illegal to do the same way tomorrow, nobody can make you go back and change it and in fact even when new code is in force the standard a project is inspected under is usually the one the permit was drawn under. The NEC by its own definition is a set of minimum requirements, it is not a design manual, and honestly it is not of the code's concern if you want to keep using that 50 year old fridge, vacuum, or table lamp. The lamp is going to have a lousy cord, the vacuum will probably trip your AFCI breakers in your new house, and the fridge full of beer in your garage is going to trip the GFCI outlet it's plugged into. And if you burn your house down with something electrical it's between you, your homeowners insurance, and the manufacturer of the offending product which may or may not still be in business.

Maybe our different views of fusing come from our different experiences with it. For someone who grew up in a house with a fusebox identical to the one Jim posted a picture of a couple pages back, I just assume that anything which can be relatively easily meddled with simply will be. There was a 25A fuse in my panel which fed the 120v well pump, the fridge outlet, and about half of the first floor lighting and power oultets (there were maybe ten in all). Granted the load was not carried all at once on the same (14AWG I think, and old) conductors, but still the circuit was grossly overfused. Why? Because someone plugged something in and the smaller fuse blew. So they put in a bigger one and it didn't happen again.

Is this likely to happen with appliances? I guess not. Unless I suppose you knock your radio or something into the sink, the plug fuse blows, you go to the hardware store and buy the biggest they have and, well, what happened to the protection? I have no idea if this situation has any basis in reality, but an appreciable part of our code deals with situations that have happened maybe only once, someone submitted a proposal to the codemaking panel, and they were convinced it needed to be addressed.

 

[jim hardy]

The potential voltage will be dropped across the load, so in theory the current will run along the neutral with no volts pushing it??

As MrSparkle said

Drop along the neutral wire is very small.

eg #14 is 2.5 milliohms per foot.

You indeed have a voltage divider between the load and the neutral
with the load devouring the tyrannosaurus' share of voltage.

A 50 foot #14 neutral carrying ten amps clear across the household will be elevated at the load end by
2.5 X 50 = 125 milliohms, X 10 amps = 1.25 volts.
About 1% in US, leaving load with 99% .

EDIT: Late Addition
Above was posted too quickly in an attempt to achieve minor clarification ( that's likely unneeded) .
Of course the "Hot" conductor carrying power to the load takes its bite as well, leaving 98% for the load.
old jim

 

[Averagesupernova]

Sophie. In the USA I don't know of any single family residences that are fed with 3-phase. All single phase, and I would imagine all but the oldest of houses in the far reaches from civilization are 240 volt split-phase. Over here multiple houses share the same transformer. Different utility companies have different standards on who owns what. Some places the customer owns everything up until the transformer except the actual meter itself. The customer will own the meter socket. Other utilities the customer only owns up until the meter socket sometimes including the socket, other times not. 200 amp services are common in my location. 50 KVA transformers in my location are the largest you can get for 240 volt single-phase, at least for a single customer. However, it is not uncommon at all to have half a dozen 200 amp services on a single transformer of this size. With smart meters the utility can tell just how a customer loads the transformer. If a customers useage gets too close to capacity the utility will make you put in another transformer and meter and split your operation up.
-
I see some advantage of a ring system. I just don't like the possibility of things going wrong and going unnoticed. I am not a big fan of multiple current paths anyway. I have had experiences on PC boards with a ring type ground. Normally ground currents would go mostly to the left for example from a certain component. Suppose that path breaks for some reason and now the ground currents go to the right. Now these currents are sharing a path that they did not before and if they are a high frequency or have fast switching current they can creep into places that they did not before. But, everything seems to be working. No catastrophic failures, just suddenly a noise that didn't exist before. Although it could happen with a star type ground if the PCB was laid a certain way, I recall hoards of circuit boards wrecked when there would be a ground current from an external source that opened up a ground trace in a ring type ground. The result was the ground pin for a 5 volt regulator floating up and putting 12 volts on everything on the 5 volt supply bus. This was just an example of an engineer who did not have a good grasp on design for mis-use or failure. Best case would have opened a single ground trace instead of frying every 5 volt component on the board. Something similar to this is in my opinion the largest drawback of the split-phase system. Losing a neutral ahead of the service (actually that is considered a shared neutral) has the potential to over voltage one side of the line and undervoltage the other side. However, it will never damage permanent building wires hidden inside walls.

 

[AlephZero]

I think much of this is extrapolating from the faults in a system you know, to those in a system you don't know. That works both ways. As a Brit, the entire US system seems like it was designed by Rube Goldberg on a bad hair day...

So you fuse a 14 at 32 amps? Jesus! I realize that it is ring connected and ideally should not be a problem but my concern would be exactly the problem described above, if one side of the ring goes open the circuit may continue to work just fine, however it is drawing the entirety of the load down one leg which is regularly overloaded.

It's theoretically possible, but I've never heard of it happening in practice in 50 years.

Splices go bad more often than one might like to think. Over here a bad splice, even when it happens on a neutral which leads to stuff taking both lines and burning up, tends to have the result of someone getting a call to service it.

Again, this probably comes down to the construction system as much as anything else. I don't know exactly what you mean by "splice", but again that type of fault just doesn't happen in the UK. (And it's impossible with the UK house wiring system for a fault to cause "double the voltage and burn up" in any case.)

Remember that we don't live in wood-framed houses either. Faulty wiring doesn't set fire to bricks or concrete wallblocks very easily.

Unless I suppose you knock your radio or something into the sink, the plug fuse blows, you go to the hardware store and buy the biggest they have and, well, what happened to the protection?

"The biggest fuse they have at the store" that will fit a standard UK plug will be 13A, which is the rating of the outlet it would be plugged into. So there is no safety issue with the house wiring. Fuses are to protect the wiring, not the equipment. If a UK appliance needs a low rated fuse for the safety of its internal circuitry, that fuse would be internal to the device and not user-replaceable. For example devices like "wall wart" power supplies typically don't have any user-accessible fuses.

I suspect lower rated plug fuses in the UK are a left-over from the old wiring system which had different sized plugs and sockets rated at 15A, 5A and (occasionally) 2A, each with its corresponding fuse. Even in my current house, I've still got an old 2A plug and socket that powers the water pump for the heating system. It's in such an inaccessible place that replacing it would be a serious building job, not 10 minutes' electrical DIY work.

 

[Averagesupernova]

I don't know exactly what you mean by "splice", but again that type of fault just doesn't happen in the UK.

I suspect it does happen but goes unnoticed. You see that is the beauty of the ring system in that it is redundant and at the same time the drawback in that the failures never show up.

 

[AlephZero]

The redundancy makes it more tolerant of a single "loose connection" fault. A loose connection that carries high current will generate heat and be an immediate fire risk. That hazard is removed by the redundancy.

Since the cable in a 32A ring is rated at 20A not 16A, a poor connection may not actually generate an overload condition in any case, depending how the ring is used - and remember everything is at 230V, so typical currents are half of the US 120V equivalents.

The wall sockets are designed for reliable connection of 3 sets of cabling, without any additional hardware for cable joints - i.e. the two "halves" of the ring, plus an optional spur connection to another socket.

It does introduce the possibility of other wiring errors though - e.g. adding wiring which cross-connects two rings, which means that a fault may leave 20A wiring protected by only a 64A fuse (2 x 32A in parallel).

 

[sophiecentaur]

The redundancy makes it more tolerant of a single "loose connection" fault. A loose connection that carries high current will generate heat and be an immediate fire risk. That hazard is removed by the redundancy.

Since the cable in a 32A ring is rated at 20A not 16A, a poor connection may not actually generate an overload condition in any case, depending how the ring is used - and remember everything is at 230V, so typical currents are half of the US 120V equivalents.

The wall sockets are designed for reliable connection of 3 sets of cabling, without any additional hardware for cable joints - i.e. the two "halves" of the ring, plus an optional spur connection to another socket.

It does introduce the possibility of other wiring errors though - e.g. adding wiring which cross-connects two rings, which means that a fault may leave 20A wiring protected by only a 64A fuse (2 x 32A in parallel).

That's a good example of the high build quality of components used in UK Power distribution. The UK system looks a bit 'quirky' compared with what you find elsewhere but (with a few exceptions - available on some market stalls) it's made very well, with chunky brass components and large contact areas.
Or even cross connecting two rings, brought back to the consumer unit. But all systems are open to abuse by bad installation contractors.

It always amazes me that there is such a contrast between power and lighting circuits. You just cannot buy 'dodgy stuff' for UK power circuits but lighting fitments seem to come in all levels of quality and finish.