The K-Zone: An overview of domestic electrical installations

This article outlines the circuit configurations normally used in domestic electrical systems, including the use of ring and radial supplies, kitchen and bathroom requirements, and outdoor installations. There is a also a brief discussion of non-mains installations like telephone and network wiring. This article is written for competent, sensible DIY enthusiasts who may have some experience of electrical work; it is not intended to encourage people to carry out work for which they lack the practical or theoretical background.

Note on the text

Principles

It is possible, in theory, to run one electrical power cable around an entire property, and use it to connect all lights and appliances. The cable would have to be very large, to carry the high current, and would therefore be heavy and expensive. In practise, we normally divide the installation into circuits, each covering a specific part of the property, or serving a specific function. As well as being common sense, this is stipulated by [IEE 314-01-01]. A basic rule is to use the smallest cable that can safely supply a given circuit. This will save money, and will reduce the time taken for installation because smaller cables are much easier to work with.

In a domestic installation, mains circuits fall into three basic types: power outlet circuits, lighting circuits, and single-appliance circuits. Power outlet circuits supply power to electrical appliances via plug-and-socket connections. Lighting circuits operate lights, and single-appliance circuits provide power to appliances that need their own circuits, either because they draw heavy currents, or because it is more convenient or safer to do so.

Of these circuit types, lighting circuits are the most complicated to wire, because they have switches in the circuits. Power outlet circuits normally have switches in the outlets, so no extra switching needs to be provided by the installer.

All mains circuits are required to have an earth conductor, even if it is not used. The requirement to do this in lighting circuits is emphasized in [IEE 471-09-02] because it was commonly thought unnecessary until recently. Bear in mind that even if all your light fittings are plastic, and therefore cannot become live in the event of a fault, someone, someday, may replace them with metal fittings. If no earth conductor is available, this could result in a dangerous installation.

Power outlet circuits

Ring and radial circuits

Power outlet circuits are, in principle, simple to wire up. In concept, the wiring consists of connecting all the `live' terminals together, all the `neutral' terminals together, and all the `earth' terminals together. In practise, there are two main configurations: rings circuits, and radial circuits. In the radial configuration (figure 1) the outlets form a chain, running continuously from the supply point to the furthest outlet in the system. Note that this arrangement must be strictly followed if the circuit is to be described as radial. Separate branches taken from the basic chain are called `spurs', as described below.

**Figure 1:** connection of a radial power circuit
$\begin{figure}\epsffile{radial.eps}\end{figure}$

In the ring configuration (figure 2), the outlets have two complete paths to the supply. In a sense, a ring circuit is like a radial circuit with the final outlet connected back to the supply.

Note that in both these configurations, no connections are necessary except at the supply terminals (about which more later) and at the outlets themselves. Outlets like 3-pin sockets are designed with terminals large enough to take two connections on each pin (or three, at a pinch). It is also possible to make connections in a junction box, if necessary, but this is best avoided.

**Figure 2:** connection of a ring power circuit
$\begin{figure}\epsffile{ring.eps}\end{figure}$

Contrary to the advice usually found in DIY books, there is no regulatory requirement to supply socket outlets from a ring circuit. Neither the Electrical Supply Regulations nor the IEE Wiring Regulations stipulate this. The IEE On-Site Guide does recommend, however, the maximum floor areas that may be served by the various kinds of circuit, and it is clear that the ring configuration offers important benefits. These guidelines are summarized in table 1.

Table 1: Floor areas for various power outlet wiring configurations. Source: IEE On-Site Guide

Conductor size (mm)	Circuit type	Nominal fuse/MCB rating (A)	Floor area (m)

2.5	radial	20	20
4.0	radial	30 or 32	50
2.5	ring	30 or 32	100

Note that the IEE On-Site Guide does not attempt to place a limit on the number of appliances or outlets that can be connected, but only on the floor area to be served. You may legitimately and safely connect a hundred mains outlets on one circuit if you wish¹. It is clear from table 1 that we can serve a greater floor area with a ring circuit made from 2.5 mm

cable than we can with a radial circuit in 4 mm

cable, and clearly this is the way to go. Note that if you are planning to install a new ring circuit, or replace an old one, you need to be careful that the length of the ring meets the voltage drop and disconnection time regulations, as discussed in my more technical article on cable selection. A total cable length of about 30 metres will be safe in most circumstances.

Spurs

There is often a requirement to add a socket outlet to an existing circuit. Ideally we should do this by extending the ring, so that the new outlet becomes part of the ring itself. Often it is impractical to do this, and we will want to take a branch from the ring to supply a socket. This is called a spur. Again, there is no regulatory limit on the number of spurs that can be taken from a ring, but the IEE On-Site Guide recommends no more than one spur for each outlet on the ring. What is forbidden, because it is dangerous, it to take a spur from an existing spur. The problem is simply that the spur does not benefit from the doubling of the cables that serves a ring, and cannot carry the same current. However, if it has electrical outlets fitted, the user has no way of knowing this.

A particular important type of spur is the fused spur. In this arrangement the spur is connected to the ring by a fuse and, optionally, a switch. A fused spur becomes, in effect, a separate circuit, because its fuse limits the load it places on the main ring. Fused spurs are very important where we need to connect appliances designed for a lower-current circuit into a higher-current one. For example, suppose we have a wall-mounted lamp designed to be fitted on a 5-amp lighting circuit. The lamp will be constructed in such a way that it can stand a current of 5 amps (in a fault), but will expect never to see more than 5 amps, because somewhere a fuse or MCB will trip. Suppose we want to connect two such lamps to a 30-amp power circuit. We can't take a simple spur from the power ring, because it has a 30-amp fuse or MCB protecting it (and we need to restrict current to 5 amps). So what we do is fit a fused spur unit and connect the lamps to that. The fused spur unit will be fitted with a 5-amp fuse. The spur unit will cost about three pounds, and fit the same mounting as a standard lightswitch. For an extra few pounds you can get one that contains a switch as well, so it could be the lightswitch.

You may be wondering why we would want to connect lamps to a power circuit, rather than a lighting circuit. First, it is sometimes simply more convenient to do this, especially with wall lamps. Second, some lamps - especially security floodlamps - may take as much as 4 amps. Although one of them will run on a 5 amp lighting circuit, it doesn't leave a lot of current over for the rest of the lamps. And, clearly, two 4-amp security lamps will overload a 5-amp lighting circuit.

Sockets and flex outlets

Most electrical appliances will be connected to the power system by plug-and-socket arrangements. In the UK, the 13-amp, 3-pin plug is usually chosen, although smaller, 6-amp or 3-amp connectors may be appropriate in some installations (e.g., for a hi-fi system with 5 components all together).

Despite its convenience, plug-and-socket connection is not always appropriate. For example, in the UK plugs have to contain fuses; if an appliance is large and heavy, you may not want to have a fuse in a plug hidden behind it (because the fuse may blow, and then you'll have to find a way to get at it). You may also want to discourage, or even prevent, people from switching an appliance off (e.g., an aquarium pump or freezer). If there is a plug and socket, it is always possible to switch off by pulling the plug out. In these circumstances, a flex outlet may be more appropriate.

A flex outlet connects an appliance's flexible cable directly into the power system, without a plug and socket. The cable emerges from a small hole in the unit. In other respects the outlet is like a socket outlet: it is wired into the power system in the same way, and fits the same mounting. The outlet will have a small hole at the front or to the side of the casing, from which the flex emerges. Inside the outlet is a clamp to prevent the flex being pulled out.

Flex outlets are very appropriate for fridges and freezers, washing machines and dishwashers, pumps, heating equipment, security systems, and alarm systems. They are available with or without switches, and with or without indicator lamps (that light up when the appliance has power). They may also have built-in fuses (for reasons that will be explained below).

As well as for convenience, another reason for using a flex outlet is to prevent overheating of the plug and socket with an appliance that is running very close to the maximum current allowed. For example, a 13-amp plug running at 13-amps can get quite warm. This is a particular problem in the vicinity of heating equipment or hot-water tanks. The 13-amp limit of the 3-pin plug applies at room temperature, not in a hot closet. An electric immersion heater, although it may require less than 13 amps, is better connected to a flex outlet than a three-pin plug.

Lighting circuits

Lighting circuits may include fixed lighting units, like ceiling pendants and uplighters, and outlets for flexible lighting systems (like track lighting). These circuits are often more complicated to construct, because they include a switch.

**Figure 3:** a theoretical lighting circuit with switches
$\begin{figure}\epsffile{lightingcircuit1.eps}\end{figure}$

Notice that switch is on the live, not the neutral, part of the circuit. This ensures that the light fitting is safe when the switch is off (so, for example, the bulb can be changed without risk of electrocution).

In practise it would be very difficult to wire a lighting circuit like figure 3, because the live and neutral cables are separate. In addition, we need an earth connection at each switch, and each light fitting, for safety reasons (not shown on diagram). Normally we use two-core-and-earth cable for domestic wiring, so by convention the wiring of a lighting circuit is as shown in figure 4.

**Figure 4:** a more practical lighting circuit
$\begin{figure}\epsffile{lightingcircuit2.eps}\end{figure}$

Apart from only showing two lights rather than three, this configuration is identical to figure 3, although it looks more complex. This additional complexity is to ensure that all connections can be made with two-core-and-earth cable. The central point for each light fitting is a junction box with four terminals. There are two ways to implement this junction box. First, you could use a specific, four-terminal lighting junction box (cost: about 50p in bulk). The junction box will normally be concealed in a ceiling or floor void. Second, you could use an integrated ceiling rose, where the terminal blocks are part of the rose body. The rose connects the pendant lamp holder, and hides all the connections. For simplicity, the rose will often be supplied with exactly the right number of terminals to accommodate all the cable connections. That is, there will be two blocks of three terminals, one block of two, and one block of four (for the earth wires).

One important point to note about the standard lighting circuit is that when the switch is on, both wires (red and black) to the switch are live. If you find this situation (a black live wire) unpalatable, you can buy two-core-and-earth cable with two red conductors. Alternatively - at much lower cost - wrap a small red marker (e.g., red insulating tape) around the black wire wherever it is visible.

When integrated ceiling roses are used throughout a circuit, this is called a `loop-in' configuration. Note, again, that the only connections are inside the fittings; there are no concealed junction boxes.

In practise, strict conformance to the loop-in system is only practical for standard ceiling pendant lamp holders. It is inappropriate for, for example, wall uplighters.

Lighting circuits are usually wired with 1.5 mm

cable, although 1 mm

is not uncommon. In conditions where the cable is entirely enclosed in wall plaster, at 30 degrees celcius even the smaller of these two cables has a current carrying capacity of 11 amps: still well below the likely load imposed by a standard lighting circuit (11 amps will support more than twenty 100-watt lightbulbs). However, if the cable run is long, it is more likely to exceed the maximum voltage drop regulation. In a worst-case configuration (a long cable with maximum allwable current drawn at the far end) it turns out that a 1 mm

cable can be about 35 metres long before this happens. If the cable is longer than that, you'll need a larger cable².

Lighting circuits are not normally wired as a ring system, because the total current requirement rarely even approaches the capacity of the cable.

Single-appliance circuits

It is often more convenient, or safer, to provide certain appliances with their own circuits, wired directly to the consumer unit. If, for example, an electric shower requires 40 amps, it would be dangerous to attempt to power it from a socket outlet circuit with a capacity of 30 amps. We must either run a separate circuit, or up-rate the power outlet circuit to handle at least 40 amps. The latter would not be cost-effective.

A single-appliance circuit that is installed to supply an appliance rated at more than 13 amps cannot terminate in a three-pin plug (13 amp limit). You can buy specific connection panels for these appliances (like a flex outlet, but bigger).

Sometimes even standard 13-amp appliances may require separate circuits. For example, a dishwasher, a washing machine and a kettle in a kitchen may together draw enough current to exceed the 30-amp limit of a standard power ring. This problem can be overcome by providing a separate circuit for one or two of these appliances (usually the washing machine and/or dishwasher, as the tend to stay in the same place).

Alternatively, an appliance may require a separate circuit for safety reasons unrelated to electricity. This is particularly important for security and fire alarm controllers (because you don't want them switched off if a fuse trips owing to a fault elsewhere).

Finally, some appliances will require a separate circuit because of their location; it may be more inconvenient to extend an existing circuit than to install a new one.

Kitchens

In the UK there are no special regulatory requirements for electrical outlets in kitchens. You are not required to use RCD protection, for example, although many people like to. The use of RCDs for kitchen power outlets is a regulatory requirement in many other countries, but in the UK kitchen installations can be identical to those in any other part of the house. It's generally recommended that you don't put socket outlets where they can be operated by a person with one hand in a sink, but since many people have an arm span of about 2 metres, this would be impossible in many kitchens.

Bathrooms

Bathrooms present special problems for electrical installation, for reasons that should be reasonably obvious. Technically the IEE Wiring Regulations does not mention bathroowms as such, but `locations containing a bath or shower'. Therefore these regulations apply to any room that contains a shower cubicle, as well as a bathroomn. The whole of [IEE 601] should be required reading for anyone contemplating electrical work in such a location.

First, the electrical safety hazard in a bathroom arises in the vicinity of the bath or shower. It is recognized that beyond a certain distance from the wet area, hazards are no worse than anywhere else in the house. In the IEE Wiring Regulations this distance is essentially 3 metres horizontally from any part of the bath or shower basin³. Outside this area, the special bathroom regulations do not apply.

Second, it is recognized that in bathrooms used by disabled people strict conformance with regulations may be impossible, and special precations taken instead.

Third, there are few restrictions on the use of approved 12-volt equipment; this may be useful for bathroom lighting.

The most obvious restriction in the bathroom is the ban on using socket outlets, with the exception of approved shaver sockets. These have isolating transformers to reduce the shock risk.

Electrical switches are also banned within 0.6 m of any part of the bath or shower basin, except for those built into equipment (like electrical showers) and switches with pull-cords. The regulations do not even allow waterproof wall-mounted switches, even though they are suitable for use outdoors. Note that the ban on switches does not apply to the whole of the 3-metre bathroom zone, but only to within 0.6 metres.

Equipment that is likely to be installed within the hazardous area include shower and whirlpool pumps, shower heaters, lights, and fans. According to the IEE Wiring Regulations none of these require RCD protection if they are designed for bathroom use. Oddly enough, anything else must be fitted with RCD protection, even if it is designed for bathroom use.

Non-mains circuits

Telephone system

A modern telephone outlet box has six terminals, but in a domestic system normally only four are wired. Of these four, in fact, one is redundant, and another is only used for the bell signal. The terminals and their wiring colours are shown in table 2.

Table 2: terminal numbers and core colours for telephone circuits

Terminal number	Cable colour	Function

1	Green with white stripe	spare
2	Blue with white stripe	signal
3	Orange with white stripe	bell
4	White with orange stripe	earth recall
5	White with blue stripe	signal
6	White with green stripe	spare

The connection for `earth recall' has nothing to do with earthing for safety, it is a method of signalling used on private exchanges for, for example, transferring calls. If you don't have your own telephone exchange, this line won't do anything.

So normally a domestic telephone system is cabled up using light four-core cable. It doesn't need to be earthed, but it does have to use purpose-made junction boxes and accessories.

Adding a new telephone outlet is logically straightforward: connect terminal 2 on the new outlet to terminal 2 on the old one, connect terminal 3 to terminal 3, and so on, using a run of 4-core cable. Ideally you should use specific telephone cable, because the colour coding on its wires will match the colour coding given on the telephone outlet.

Extra-low-voltage appliances

An increasing number of electrical appliances are designed to be powered at 12 volts, rather than the usual 230 volts. Normally such appliances will be provided with their own transformer to generate the correct voltage from the mains. In general, the transformer needs to be mounted as close as possible to the appliance, because the lower voltage implies a very high current. Therefore, installations of this type are very often no different in installation than mains-powered systems; most of the cabling will be at mains voltage anyway. As an example I will describe the increasingly-common 12-volt halogen track lighting systems.

Track lighting systems are appropriate where a large number of small, intense lamps are required, for example to emphasise a number of paintings hung on walls or to illuminate a large work surface. Most systems allow lamps to be clipped into place or removed, and pointed in the appropriate direction. Normally the lamps are 12-volt halogen types, and they produce a very bright, white light. To get the 12 volts requires a transformer; the system should be provided with information about how many lamps can be supplied by one transformer. Tracks can often be connected together to extend the system.

The wiring from the transformer to the tracks, and between tracks, will carry a very large current at 12 volts. Thus these cables will be very large and heavy, and need to be kept as short as possible. In addition, the transformer is ugly, and needs to be mounted out of sight, but not in a confined space (because it will overheat) or concealed under floorboards (unless you can get at it for maintenance).

Figure 5 shows a typical installation strategy, where the tracks are to be mounted on the ceiling. I assume that an ordinary lightswitch is to be used to control the lights; the switch must be between the mains supply and the transformer, not between the transformer and the lights. This will allow the supply to the transformer to be disconnected during repair or maintenance.

In the example, the transformer is mounted on the wall just below the ceiling, and within a few inches of the first track. This keeps the amount of thick, 12-volt cable to a minimum. The transformer could also be mounted in the ceiling void if you have access to it from upstairs. This would put it completely out of sight, but may make it tricky to get to.

The transformer is connected to an existing ceiling junction box; I am assuming we are replacing an existing pendant lampholder here. The lighting cable can be concealed in the wall as normal, and brought to the transformer by a flex outlet. The transformer is probably supplied with a built-in cable, so there has to be some mechanism for connecting it to the installation cable.

Note that while it is possible to conceal the 12-volt cables in plaster as well, we don't have reliable figures for how this affects their current ratings, so this may not be a good idea. In any event, these cables should probably be short enough that you don't need to conceal them.

**Figure 5:** connection of a 12-volt halogen track lighting system into an existing ceiling rose or junction box
$\begin{figure}\epsffile{tracklighting.eps}\end{figure}$

Computer network systems

If you have computers in your house, it may make sense to install a local-area network if you are doing significant re-wiring; installing network cabling is no different to installation of any other kind of cable, and is best done before redecorating, and while you have access to floorboards, etc.

There are at present two network systems broadly suitable for home use; the earlier system based on coaxial cabling, and the more modern `unscreened twisted pair' (UTP) system. The former is essentially obsolete, and it is becoming difficult to get compatible equipment. I would recommend the UTP system for all new installations.

In the UTP system, computers are connected using hubs. For about �20 you can get a hub that will connect up to four computers. It is also possible to connect hubs together to build a larger installation.

Computer networks need much more careful planning than power and lighting systems; this is because the location of the hubs is crucial to a cost-effective and flexible system. For a substantial network, a reasonable solution is to have one hub on each floor of the premises, plus one in any room that has more than one computer. The `backbone' of the system would be a cable that connects only the hubs together; each hub would be cabled to a number of network socket outlets mounted in exactly the same way as a power outlet.

Details of the installation procedure are beyond the scope of this article; for more information see my article on home networking.

Security systems

Commercial security systems are normally based on fixed wiring, just like power systems and computer networks; however, for home use it is increasingly common to use wireless systems. These use radio to communicate between the sensors, alarm sounders and control unit. Such a system is very quick to install.

Wired systems are usually cabled with light, 6-core cable. In a budget system it will be necessary to run a separate cable from each zone (group of sensors) to the main control unit. Using large zones (with many sensors all connected in series or parallel) makes the wiring much easier, but the system cannot distinguish which sensor has been activated. This may be a problem in a large premises.

More expensive systems allow the use of zone repeaters; these collect up signals from a number of sensors, and concentrate them into a single cable for transmission to the control unit. In such a system you could install one repeater on each floor, and perhaps one in each outbuilding, and the bulk of the wiring would be between repeaters and sensors which are close together. This makes the wiring much simpler, without the loss of discrimination between sensors.

The control unit itself is normally powered by a fixed mains outlet (i.e., a flex outlet) from a power ring, or a private mains circuit.

Installation of security systems is a complex and specialized subject, and would need a whole book to describe.

Practicalities

Modern wiring accessories are designed to be simple to install, in the standard configurations described above. For example, lighting junction boxes have four terminals (which is exactly what is required by the circuit shown in figure 4) and have four cable entry points (two for the power supply (in and out), one for the switch and one for the bulb).

Types of fitting

Modern (indoor) electrical wiring fittings (outlets, lightswitches, etc.) come in two general type: flush mounting and surface mounting. In a flush-mounted fitting, the wiring and connections are below the surface of the wall, and only a small thickness of the fitting protrudes. This system is ideal when used with fully-concealed wiring. The mounting boxes are usually made of steel, and are very cheap (since no-one will ever see them). These boxes are called patresses.

Surface-mounted fittings are mounted entirely on a wall surface, and may be used with either concealed or surface wiring. Large surface-mount boxes look very ugly, and are best avoided where possible. However, because they don't require any chiselling away of brickwork, they are much easier to fit.

For outdoor systems, fittings are usually designed to join up to standard 20 mm circular truncking. Normally there will be an insulating bush at the junction to keep the system watertight.

Using flush-mounted fittings

In a substantial rewiring exercise, you will probably want to use flush-mounted fittings, and concealed wiring, as much as possible. The problem is that each fitting will require the plaster and brickwork to be cut away. There's no doubt that mounting these fittings is an unpleasant, tricky, messy job; here's two ways to do it.

Traditional method: drill and chisel

This is, of course, easier said than done. You can simplify matters considerably by using a drilling jig. This is a plastic frame with holes in the correct positions for drilling the outline. This should be screwed to the wall before drilling (don't try sticky tape or Blu-Tak: it will fall off).

Modern method: using an electrical box sinker

You can save a lot of time, if you have many fittings to mount, by using a box sinker. This is a tool designed for exactly this one job; it costs about �80, but will save half and hour every time you use it.

However you do the job, fitting the mounting boxes will be noisy and messy. Naturally you don't want to do this unless you plan to redecorate afterwards.

Although you can be lucky, and end up with a perfectly square hole, very often you will find the chunks of plaster and brickwork fall away from the edges of the hole, and you will need to make good with plaster.

This job is relatively straightforward where the plaster is thick, or where the brickwork is aerated blocks (`breeze blocks'). Traditional brick is much more difficult, and decorative brick almost impossible (not because it's hard to drill, but because it shatters and leads to a huge, irregular hole). But by far the worst kind of wall to work on is cement or concrete. It's almost impossible to chisel, and you'll end up with a very messy hole.

Happily the mounting box does not have to be perfectly straight and level; there will be about 5 mm of adjustment in the mounting holes, so the faceplate can be straight even if the box isn't.

Using dry-wall fittings

A dry-wall is a wall made of plasterboard mounted on wooden studs; there will usually be a void behind the plasterboard large enough to conceal electrical wiring and fittings. Note that walls covered in close-mounted wall lining boards need to be treated as an ordinary plastered wall.

The problem with plasterboard is that it isn't usually thick enough, or robust enough, to screw anything into it. This necessesitates the use of specialized dry-wall mounting boxes. These usually push into a hole cut into the plasterboard, and have a pair of plastic clamps that fasten onto the edge of the cut-out. It doesn't take much effect to make the cutouts: a jigsaw is quickest, but a sharp knife will also work in many cases.

Outdoor installations

The best way to provide an outdoor mains supply is to use proper outdoor cabling accessories, with a supply derived from the main distribution board via an RCD. As always, you will need to ensure that the cabling is adequate for the current load you anticipate.

A particular problem is that of meeting the voltage drop regulations when the cable is likely to be quite long. You may find that you have to use a cable that is heavier than you would expect, not to carry the current, but to meet the voltage drop requirements. This is particularly likely if you plan to use the outdoor system to supply power tools.

For example, a 2.5 mm

cable carrying 20 amps can only be 27 metres long before it fails the voltage drop requirements.

You can derive the outdoor supply from a power ring, using a fused spur unit, if the load is expected be less than 13 amps. You may need to include an RCD in the circuit at the point the spur is taken, if the power ring itself is not RCD protected.