ECN Forum Forums General Topics of the Trade Non-US Electrical Systems & Trades Ring circuits again!

C-H, Thanks for your explanation.

>but what do you use in Belgium? Please tell me you use type B for sockets! Do you use 20A or 16A for them, BTW?

In belgium we have a TT system (in residential installations) and therefore use type C curve. We use the radial system (as you understood) and we use 2.5mm2 wire/cables with a max value if breaker being 20A. All sockets have to be with ground.

Even if you use a TT system, you must have something to protect the .75 mm2 lamp cords from short-circuit? (A 20A type C breaker will allow 100A current for several seconds. There won't be a trace left of the lamp cord by then...)

Some people might be wondering what TT and TN has to do with tripping curves of breakers. I'll attempt to explain it. In a TT system, there is no connection between earth and neutral. The Protective Earth is only connected to a ground rod. The resistance of this is too large to allow a short-circuit to ground. (With a 25 ohm ground rod, the current will only be 230V/25ohm = 10A.) Therefore, these systems always have an RCD (GFI), in the 100-500 mA range which cuts the power in case of a ground fault. Even if a time delay RCD is used for this, the response time is short enough to prevent the dangerously high voltage on metal objects connected to the protective earth present during a fault condition from killing someone.

In a TN system, there is a connection between neutral and earth. If there is short-circuit between live and earth, the current will be high enough to trip a breaker. For this reason, these systems don't need to be equipped with an over-all RCD. However, care must be taken that the fault current is high enough to trip the breaker within a short period of time (no more than 5 seconds). To check this, you do an earth loop resistance calculation and sometimes also test it once you have made the actual wiring. A fast breaker (type B) will allow a higher resistance than a slow-blow (type C), and hence longer runs.

If you can't get a low enough resistance, you need to install an RCD, just like in the TT system.


[This message has been edited by C-H (edited 10-14-2002).]

>Even if you use a TT system, you must have something to protect the .75 mm2 lamp cords from short-circuit? (A 20A type C breaker will allow 100A current for several seconds. There won't be a trace left of the lamp cord by then...)

I presume that by lamp cord you mean the cord which goes from the cieling to the lamp.
Are you so sure that a few seconds of 100A on a 0.75mm2 will make that there is no trace left of the lamp cord? Remember that the lamp cord is relatively short.
If the lamp cord does melt, then it will be at the connecting points (b/c of the higher resistance at the connections). Therefore, The idea of the breaker is to protect the wiring in the walls (b/c of it's unaccessibility and fire hazard), not the lamps and appliances, which can be easily accessed.

What curve do you use in Sweden? (It is Sweden, right?) What's your solution to protecting the lamp cords?

I think David has provided a good outlined of the ring circuit.

Yes, the ground wire run in a ring gives redundancy in that a single break will not leave any outlet without a protective ground connection.

The same ring configuration on the phase and neutral means that a single break on either of these also results in no outlet being left without power. That's when problems with overloaded cables can start. I encounter such breaks in ring continuity on a fairly regular basis.

In cases where I find that both phase and neutral have no ring continuity, it almost always turns out to be somebody's DIY attempt at adding outlets which has left the ring broken.

I should also add to David's explanation the fact that our Regs. allow a ring to have "spurs." A spur is a single cable feeding a single or duplex receptacle which is connected to the ring at an outlet or junction box. The spur cable is also 2.5 sq. mm.

Although the IEE considers it unlikely to happen, it would be possible to connect two 3kW loads to a double socket and overload the spur cable.

Some months ago, we also discussed the exact distribution of current around a ring. The fact is that the ring is not completely foolproof. Given heavy loads connected close to one end of a long ring, it is still possible to exceed the cable rating on the short leg.

For some circuits BS7671 requires a disconnect time of under 0.4 sec., although 5 seconds is acceptable for lighting circuits, fixed appliances, etc.

Re the breakers, we generally use type B in all residential applications.

We used to use metal cable clips. These were nailed to the surface first, then the cable laid in and the strip bent around the cable and back on itself to secure it.

These days, the plastic saddle with a single nail driven on one side of it is standard.

Paul,
I'm in total agreement with what you say regarding ring continuity.
Breaks are usually as a result of botched DIY, growl! (I'm on the soap box again!)

There are times when I wish we could go back to radial circuits, but these are only when I am pulling my hair out attempting to locate a break in continuity, normally caused by the above.

I have to say that generally I like ring circuits for a number of reasons:
1) Permits relatively long cable runs, with limited voltage drop & low earth loop impedance (TN-S & TNC-S).
2) High integrity earthing, required to comply with section 607 of BS 7671.
3) Providing it is correctly designed, it allows a lot of sockets to be installed on a circuit with little chance of over-loading.(Kitchens excepted.)
4) In the event of a cable fault, you can disconnect the faulty section & temporarily restore power until a permanent repair / replacement can be made. Obviously precautions would have to be taken to prevent over-loading.

I was grateful of #4 above when I was called out this morning.
A ring main serving bedroom sockets in a nursing home went faulty, tripping the 32A MCB in the sub panel & also the 63A sub main MCB at the main panel.
I traced the fault to a live (phase)/ earth cable fault between the sub panel & 1st socket on the ring. Disconnecting this cable at the sub panel & 1st socket enabled me to re-energize the circuit, using a 16A MCB to prevent over-loading.
Next week I will go back & replace the faulty cable (after the joiner lifts chipboard flooring) & install a new split load sub panel incorperating RCD for the socket circuits.

I don't think unfused spurs should be allowed on a ring, they tend to be extended indiscriminately by those who know no better.

Breakers for ring circuit are normally "B curve" for residential, "C curve" for commercial / industrial. Nothing to prevent you installing "C curve" for residential, providing you can achieve the lower loop impedance values.
Generally socket circuits on TN systems have to disconnect within 0.4 second.
Socket circuits on TT systems must be protected by a 30mA RCD, also sockets which may reasonably be expected to supply equipment used outdoors on TN systems require 30mA RCD protection.

I have a policy of installing 30mA RCD protection to all residential socket circuits, in new work & when replacing an existing board (panel).

BTW what's "THHN in PVC flex conduit"?


[This message has been edited by David UK (edited 10-14-2002).]

>2) High integrity earthing, required to >comply with section 607 of BS 7671

How high is high ?
One break is relatively OK, you have a reduced capacity fault circuit, two breaks amd who knows ? (been there been very scared)
It's about time we had self diagnostic outlets - RCD outlets do it !

What price an LED

Testing is too much like an MOT - only valid on the day it's done.

Chris

Ah! You mean putting a green LED between live and earth to tell if the earth is OK? It sounds like a great idea. There really are some very simple things the lawmakers could do to improve safety.

Or from a socket to a lamp or any other small appliance (TV, radio, tape recorder)
Couldn't find a good word.


No, it was taking it too far. Evaporating the copper will take a huge amount of energy. (Copper melts at 1100°C, I think)

It is however easy to destroy the PVC insulation.

The maxium allowed operating temperatur for PVC is 70°C.

The maximum allowed short-circuit temperature of PVC is 160°C.

Thus there is 160-70 = 90°C of cable heat up available to handle a short-circuit.

How hot will a 0.75mm2 wire be if it is exposed to 100A for one second?

For copper, the specific heat is 0.39 kJ/(kg*K)

The density of copper is 8.9 x 10^3 kg/m3

The resistivity of copper is 1.7*10^-8 ohm*m2/m

I assume there is no change of state in the PVC and therefore neglect the heat absorbed by the insulation. (This can be justifed by doing the full calculations but I'll jump this part.)

I also assume the heat up is adiabatic, i.e. without any heat exchange to the surroundings.

energy = power x time
power = voltage x current
voltage = resistance x current
resistance = resistivity x length of wire/cross sectional area

Thus:
energy = resistivity x (length of wire/cross sectional area) x current^2 x time { * }

energy = increase in temperature x mass x specific heat

mass = density x length x cross sectional area

Thus:

increase in temperature = energy/(density x length x cross sectional area x specific heat) = {inserting * } = (resistivity x (length of wire/cross sectional area) x current^2 x time)/(density x length x cross sectional area x specific heat) = (resistivity x time x current^2)/(density x cross sectional area^2 x specific heat)

With numbers:

Temperature increase = 1.7x10^-8 x 1 x 100^2/(8.9x10^3 x (0.75x10^-6)^2 x 0.39x10^3) = 87 °C

We get an increase in temperature of 87°C, which is just within the 90°C limit. Hence, the cord survived. But had the breaker taken two seconds to trip it wouldn't.

>Remember that the lamp cord is
>relatively short.

I know you know the answer to this yourself. Why is a short cord better in a short-circuit situation? Is it:

1.) Because the cord will be cooler despite the same current going through it

or

2.) Because it has a lower resistance, increasing the current and thereby triggering a faster response from the breaker?

>If the lamp cord does melt, then it will be
>at the connecting points (b/c of the higher
>resistance at the connections).

Ah! I've never thought of this. It sounds reasonable.

>What curve do you use in Sweden?

Well - since 10A circuits are the norm (or at least used to be) type C breakers are common to prevent nuisance trips when you start motors (like the vaccum cleaner). D-I-Y:ers always use type C for everything. They would use type D if they could buy them... Electricans use what is deemed suitable for the circuit, hence often type B.

Most homes have cartridge fuses, and these are now slow-blow. You could buy both fast and slow fuses a few years ago, but since nobody bought the fast ones, they have now disappeared. (Ranger - what is the situation concerning Diazed fuses in Austria?)

Many fuses also feature the "infinte" trip curve. (Copper wire or similar wrapped around the fuse. Or a nail through the fuse) These are so very convenient - never any blown fuses. *sarcasm*

What kind of fuses are used in Belgium?

Sorry Paul - I'm ruining your thread!

C-H, Let me add another point which came to me overnight. About the lamp cord (or whatever appliance).

As aelectrician it's his duty to protect HIS wiring, not the wirings of the radio/lamp manufacturers. That is their duty.
On their behalf I can presume that they think this way:
Why would there be a 100A flowing through a 0.75mm2?
If there is a ground fault, then the RGI will trip in 0.6 seconds. If it's a short, then the breaker will trip even faster.

Thank you for all those calculations!

As to what we use in Belgium. Well we don't use fuses at all anymore. We use only breakers C curve (residential). Even the main one is a breaker. We do find still older installations with fuses, including the "infinite" type.

Paul, I'm sorry, too for ruining it.
I've had time to think abbout your ring circuits. My personal opinion is that it's a bad joke. What's the idea behind it? To save a bit of work or copper? Then why don't you branch it off a few more places (in midlle of the ring) to the 32A breaker? That way you could use 0.75mm2!
That is, of course, untill that one of them has a loose connection and the whole house is on fire. It's tippicaly British to think of something like that! This would never be allowed in France or Belgium.

[This message has been edited by Belgian (edited 10-15-2002).]

C-H & Belgian,

No problem on the diversion to cords and proper protection. One thing leads to another in these discussions, and they're often related somehow anyway.

I'm often amused by the instructions that come with appliances sold in the U.K. along the lines that the plug "MUST be fitted with a 3A fuse" (their emphasis). Yet the same appliance, fitted with the same 0.75 sq. mm cord is now likely to be sold right across Europe, where in most cases it will be connected to a circuit fused at 16 or even 20A.

By the way, although 0.75 sq. mm is the smallest size generally used on Continental cords, some small appliances sold specifically for the U.K. are fitted with 0.5 sq. mm (e.g. table lamps). That's when the protection of a 3A fuse in the plug is really needed.

I agree with David about the extensions to spurs on our ring circuits. (Another sidenote: Most laymen here call them a "ring main," although the correct term is "ring final circuit.")

I've seen extra spurs daisy-chained from an original spur far too often, sometimes three or four. They're often feeding outlets in adjacent rooms, and thus it's quite likely that somebody could plug two or more high-power devices in simultaneously.

I seem to be inserting "by the ways" a lot here, but another point is that the original ring specification set out by the IEE actually allowed a spur to feed two separate outlets, not the one which is permitted now. (I haven't got my old Regs. handy, but I have a feeling it may have changed only in the 15th Edition, 1981.)

David,
I carried out similar hacks before to get power restored. In fact in one house a while back tearing into the tiled walls to find the problem was going to cause so much damage that I converted a 30A kitchen ring into two 20A radial circuits permanently.(fortunately, the break came at a suitable place and there was a spare way in the panel!).