Personal tools

Protection against indirect contact ( full page )

From Electrical Installation Guide

Jump to: navigation , search
General rules of electrical installation design
Connection to the MV utility distribution network
Connection to the LV utility distribution network
MV and LV architecture selection guide for buildings
LV Distribution
Protection against electric shocks and electric fires
Sizing and protection of conductors
LV switchgear: functions and selection
Overvoltage protection
Energy Efficiency in electrical distribution
Power Factor Correction
Power harmonics management
Characteristics of particular sources and loads
PhotoVoltaic (PV) installation
Residential and other special locations
ElectroMagnetic Compatibility (EMC)

Exposed-conductive-parts used in the manufacturing process of an electrical equipment is separated from the live parts of the equipment by the “basic insulation”. Failure of the basic insulation will result in the exposed-conductive-parts being alive.
Touching a normally dead part of an electrical equipment which has become live due to the failure of its insulation, is referred to as an indirect contact.


Measures of protection: two levels

Protection against indirect contact hazards can be achieved by automatic disconnection of the supply if the exposed-conductive-parts of equipment are properly earthed

Two levels of protective measures exist:

  • 1st level: The earthing of all exposed-conductive-parts of electrical equipment in the installation and the constitution of an equipotential bonding network (see chapter G section 6).
  • 2nd level: Automatic disconnection of the supply of the section of the installation concerned, in such a way that the touch-voltage/time safety requirements are respected for any level of touch voltage Uc(1) (see Fig. F7).

(1) Touch voltage Uc is the voltage existing (as the result of insulation failure) between an exposed-conductive-part and any conductive element within reach which is at a different (generally earth) potential.


Fig. F7:  Illustration of the dangerous touch voltage Uc

The greater the value of Uc, the greater the rapidity of supply disconnection required to provide protection (see Fig. F8). The highest value of Uc that can be tolerated indefinitely without danger to human beings is 50 V CA.

Reminder of the theoretical disconnecting-time limits

Uo (V) 50 < Uo ≤ 120 120 < Uo ≤ 230 230 < Uo ≤ 400 Uo > 400
System TN or IT 0.8 0.4 0.2 0.1
TT 0.3 0.2 0.07 0.04

Fig. F8: Maximum safe duration of the assumed values of AC touch voltage (in seconds)

Automatic disconnection for TT system


Automatic disconnection for TT system is achieved by RCD having a sensitivity of I_{\vartriangle n}\le\frac{50}{R_A} where RA is the resistance of the
installation earth electrode

In this system all exposed-conductive-parts and extraneous-conductive-parts of the installation must be connected to a common earth electrode. The neutral point of the supply system is normally earthed at a pint outside the influence area of the installation earth electrode, but need not be so. The impedance of the earth-fault loop therefore consists mainly in the two earth electrodes (i.e. the source and installation electrodes) in series, so that the magnitude of the earth fault current is generally too small to operate overcurrent relay or fuses, and the use of a residual current operated device is essential.
This principle of protection is also valid if one common earth electrode only is used, notably in the case of a consumer-type substation within the installation area, where space limitation may impose the adoption of a TN system earthing, but where all other conditions required by the TN system cannot be fulfilled.
Protection by automatic disconnection of the supply used in TT system is by RCD of sensitivity: I_{\vartriangle n}\le\frac{50}{R_A}
RA is the resistance of the earth electrode for the installation
IΔn is the rated residual operating current of the RCD
For temporary supplies (to work sites, …) and agricultural and horticultural premises, the value of 50 V is replaced by 25 V.
Example (see Fig. F9)


 Fig. F9: Automatic disconnection of supply for TT system
  • The resistance of the earth electrode of substation neutral Rn is 10 Ω.
  • The resistance of the earth electrode of the installation RA is 20 Ω.
  • The earth-fault loop current Id = 7.7 A.
  • The fault voltage Uf = Id x RA = 154 V and therefore dangerous, but

IΔn = 50/20 = 2.5 A so that a standard 300 mA RCD will operate in about 30 ms without intentional time delay and will clear the fault where a fault voltage exceeding appears on an exposed-conductive-part. 

Specified maximum disconnection time

The tripping times of RCDs are generally lower than those required in the majority of national standards; this feature facilitates their use and allows the adoption of an effective discriminative protection.
The IEC 60364-4-41 specifies the maximum operating time of protective devices used in TT system for the protection against indirect contact: For all final circuits with a rated current not exceeding 32 A, the maximum disconnecting time will not exceed the values indicated in Figure F10

Uo(1) (V) T (s)
50 < Uo ≤ 120 0.3
120 < Uo ≤ 230 0.2
230 < Uo ≤ 400 0.07
Uo > 400 0.04

(1) Uo is the nominal phase to earth voltage

Fig. F10: Maximum disconnecting time for AC final circuits not exceeding 32 A

  • For all other circuits, the maximum disconnecting time is fixed to 1s. This limit enables discrimination between RCDs when installed on distribution circuits.

RCD is a general term for all devices operating on the residual-current principle. RCCB (Residual Current Circuit-Breaker) as defined in IEC 61008 series is a specific class of RCD.
Type G (general) and type S (Selective) of IEC 61008 have a tripping time/current characteristics as shown in Figure F11. These characteristics allow a certain degree of selective tripping between the several combination of ratings and types, as shown later in sub-clause 4.3. Industrial type RCD according to IEC 60947-2 provide more possibilities of discrimination due to their flexibility of time-delaying.

X IΔn 1 2 5 > 5
Domestic Instantaneous 0.3 0.15 0.04 0.04
Type S 0.5 0.2 0.15 0.15
Industrial Instantaneous
Time-delay (0.06)
Time-delay (other) According to manufacturer

Fig. F11: Maximum operating time of RCD’s (in seconds)

Automatic disconnection for TN systems


The automatic disconnection for TN system is achieved by overcurrent protective devices or RCD’s

In this system all exposed and extraneous-conductive-parts of the installation are connected directly to the earthed point of the power supply by protective conductors.
As noted in Chapter E Sub-clause 1.2, the way in which this direct connection is carried out depends on whether the TN-C, TN-S, or TN-C-S method of implementing the TN principle is used. In figure F12 the method TN-C is shown, in which the neutral conductor acts as both the Protective-Earth and Neutral (PEN) conductor. In all TN systems, any insulation fault to earth results in a phase to neutral short-circuit. High fault current levels allow to use overcurrent protection but can give rise to touch voltages exceeding 50% of the phase to neutral voltage at the fault position during the short disconnection time.
In practice for utility distribution network, earth electrodes are normally installed at regular intervals along the protective conductor (PE or PEN) of the network, while the consumer is often required to install an earth electrode at the service entrance.
On large installations additional earth electrodes dispersed around the premises are often provided, in order to reduce the touch voltage as much as possible. In high-rise apartment blocks, all extraneous conductive parts are connected to the protective conductor at each level.
In order to ensure adequate protection, the earth-fault current {Id}=\frac{Uo}{Zs} or 0.8\frac{Uo}{Zc} must be higher or equal to Ia, where:

  • Uo = nominal phase to neutral voltage
  • Id = the fault current
  • Ia = current equal to the value required to operate the protective device in the time specified
  • Zs = earth-fault current loop impedance, equal to the sum of the impedances of the source, the live phase conductors to the fault position, the protective conductors from the fault position back to the source
  • Zc = the faulty-circuit loop impedance (see “conventional method” Sub-clause 6.2)

Note: The path through earth electrodes back to the source will have (generally) much higher impedance values than those listed above, and need not be considered.
Example (see Fig. F12)


 Fig. F12: Automatic disconnection in TN system

The fault voltage Uf=\frac{230}{2}=115\ V and is hazardous;
The fault loop impedance


If ZBC and ZDE are predominant, then: Zs=2\rho\frac{L}{S}=64.3\ m\Omega, so that
Id=\frac{230}{64.3\times{10^{-3}}}=3,576 A (≈ 22 In based on a NS X 160 circuit-breaker).

The “instantaneous” magnetic trip unit adjustment of the circuit-breaker is many time less than this short-circuit value, so that positive operation in the shortest possible time is assured.
Note: Some authorities base such calculations on the assumption that a voltage drop of 20% occurs in the part of the impedance loop BANE.
This method, which is recommended, is explained in chapter F “conventional method” and in this example will give an estimated fault current of  \frac{230\times{0.8}\times{10^3}}{64.3}= 2,816 (≈ 18 In)

Specified maximum disconnection time

The IEC 60364-4-41 specifies the maximum operating time of protective devices used in TN system for the protection against indirect contact:

  • For all final circuits with a rated current not exceeding 32 A, the maximum disconnecting time will not exceed the values indicated in Figure F13
  • For all other circuits, the maximum disconnecting time is fixed to 5s. This limit enables discrimination between protective devices installed on distribution circuits

Note: The use of RCDs may be necessary on TN-earthed systems. Use of RCDs on TN-C-S systems means that the protective conductor and the neutral conductor must (evidently) be separated upstream of the RCD. This separation is commonly made at the service entrance.

Uo(1) (V) T (s)
50 < Uo ≤ 120 0.8
120 < Uo ≤ 230 0.4
230 < Uo ≤ 400 0.2
Uo > 400 0.1

(1) Uo is the nominal phase to earth voltage

Fig. F13: Maximum disconnecting time for AC final circuits not exceeding 32 A

Protection by means of circuit-breaker

(see Fig. F14)

If the protection is to be provided by a circuit-breaker, it is sufficient to verify that the fault current will always exceed the current-setting level of the instantaneous or short-time delay tripping unit (Im)

The instantaneous trip unit of a circuit-breaker will eliminate a short-circuit to earth in less than 0.1 second.
In consequence, automatic disconnection within the maximum allowable time will always be assured, since all types of trip unit, magnetic or electronic, instantaneous or slightly retarded, are suitable: Ia = Im. The maximum tolerance authorised by the relevant standard, however, must always be taken into consideration. It is sufficient therefore that the fault current   \frac{Uo}{Zs}   or   0.8\frac{Uo}{Zc}determined by calculation (or estimated on site) be greater than the instantaneous trip-setting current, or than the very short-time tripping threshold level, to be sure of tripping within the permitted time limit.


Fig. F14: Disconnection by circuit-breaker for a TN system

Protection by means of fuses

(see Fig. F15)

Ia can be determined from the fuse performance curve. In any case, protection cannot be achieved if the loop impedance Zs or Zc exceeds a certain value

The value of current which assures the correct operation of a fuse can be ascertained from a current/time performance graph for the fuse concerned.
The fault current \frac{Uo}{Zs}   or   0.8\frac{Uo}{Zc} as determined above, must largely exceed that necessary to ensure positive operation of the fuse. The condition to observe therefore is that  Ia<\frac{Uo}{Zs}   or   0.8\frac{Uo}{Zc} as indicated in Figure F15.


Fig. F15: Disconnection by fuses for a TN system

Example: The nominal phase to neutral voltage of the network is 230 V and the maximum disconnection time given by the graph in Figure F15 is 0.4 s. The corresponding value of Ia can be read from the graph. Using the voltage (230 V) and the current Ia, the complete loop impedance or the circuit loop impedance can be calculated from  Zs=\frac{230}{Ia} or Zc=0.8 \frac{230}{Ia}.This impedance value must never be exceeded and should preferably be substantially less to ensure satisfactory fuse operation.

Protection by means of Residual Current Devices for TN-S circuits

Residual Current Devices must be used where:

  • The loop impedance cannot be determined precisely (lengths difficult to estimate, presence of metallic material close to the wiring)
  • The fault current is so low that the disconnecting time cannot be met by using overcurrent protective devices

The rated tripping current of RCDs being in the order of a few amps, it is well below the fault current level. RCDs are consequently well adapted to this situation.
In practice, they are often installed in the LV sub distribution and in many countries, the automatic disconnection of final circuits shall be achieved by Residual Current Devices.

Automatic disconnection on a second fault in an IT system

In this type of system:

  • The installation is isolated from earth, or the neutral point of its power-supply source is connected to earth through a high impedance
  • All exposed and extraneous-conductive-parts are earthed via an installation earth electrode.

First fault situation

In IT system the first fault to earth should not cause any disconnection

On the occurrence of a true fault to earth, referred to as a “first fault”, the fault current is very low, such that the rule Id x RA ≤ 50 V (see F3.2) is fulfilled and no dangerous fault voltages can occur.
In practice the current Id is low, a condition that is neither dangerous to personnel, nor harmful to the installation.
However, in this system:

  • A permanent monitoring of the insulation to earth must be provided, coupled with an alarm signal (audio and/or flashing lights, etc.) operating in the event of a first earth fault (see Fig. F16)


Fig. F16: Phases to earth insulation monitoring device obligatory in IT system

  • The rapid location and repair of a first fault is imperative if the full benefits of the IT system are to be realised. Continuity of service is the great advantage afforded by the system.

For a network formed from 1 km of new conductors, the leakage (capacitive) impedance to earth Zf is of the order of 3,500 Ω per phase.
In normal operation, the capacitive current(1) to earth is therefore:
 \frac{Uo}{Zf}= \frac{230}{3,500}=66\ mA  per phase.
During a phase to earth fault, as indicated in Figure F17,the current passing through the electrode resistance RnA is the vector sum of the capacitive currents in the two healthy phases. The voltages of the healthy phases have (because of the fault) increased to \sqrt3 the normal phase voltage, so that the capacitive currents increase by the same amount. These currents are displaced, one from the other by 60°, so that when added vectorially, this amounts to 3 x 66 mA = 198 mA, in the present example.
The fault voltage Uf is therefore equal to 198 x 5 x 10-3 = 0.99 V, which is obviously harmless.
The current through the short-circuit to earth is given by the vector sum of the neutral-resistor current Id1 (=153 mA) and the capacitive current Id2 (198 mA).


Fig. F17: Fault current path for a first fault in IT system

Since the exposed-conductive-parts of the installation are connected directly to earth, the neutral impedance Zct plays practically no part in the production of touch voltages to earth.

(1) Resistive leakage current to earth through the insulation is assumed to be negligibly small in the example.

Second fault situation

The simultaneous existence of two earth faults (if not both on the same phase) is dangerous, and rapid clearance by fuses or automatic circuit-breaker tripping depends on the type of earth-bonding scheme, and whether separate earthing electrodes are used or not, in the installation concerned

On the appearance of a second fault, on a different phase, or on a neutral conductor, a rapid disconnection becomes imperative. Fault clearance is carried out differently in each of the following cases:
1st case
It concerns an installation in which all exposed conductive parts are bonded to a common PE conductor, as shown in Figure F18.


Fig. F18: Circuit-breaker tripping on double fault situation when exposed-conductive-parts are connected to a common protective conductor

In this case no earth electrodes are included in the fault current path, so that a high level of fault current is assured, and conventional overcurrent protective devices are used, i.e. circuit-breakers and fuses.
The first fault could occur at the end of a circuit in a remote part of the installation, while the second fault could feasibly be located at the opposite end of the installation.
For this reason, it is conventional to double the loop impedance of a circuit, when calculating the anticipated fault setting level for its overcurrent protective device(s).
Where the system includes a neutral conductor in addition to the 3 phase conductors, the lowest short-circuit fault currents will occur if one of the (two) faults is from the neutral conductor to earth (all four conductors are insulated from earth in an IT scheme). In four-wire IT installations, therefore, the phase-to-neutral voltage must be used to calculate short-circuit protective levels i.e  0.8\frac{Uo}{2Zc}\ge Ia^{(1)}  
Uo = phase to neutral voltage
Zc = impedance of the circuit fault-current loop (see F3.3)
Ia = current level for trip setting

If no neutral conductor is distributed, then the voltage to use for the fault-current calculation is the phase-to-phase value, i.e.   0.8\frac{\sqrt{3}Uo}{2Zc}\ge Ia^{(1)} 

(1) Based on the “conventional method” noted in the first example of Sub-clause 3.3.
  •  Maximum tripping times

Disconnecting times for IT system depends on how the different installation and substation earth electrodes are interconnected.

For final circuits supplying electrical equipment with a rated current not exceeding 32 A and having their exposed-conductive-parts bonded with the substation earth electrode, the maximum tripping time is given in table F8. For the other circuits within the same group of interconnected exposed-conductive-parts, the maximum disconnecting time is 5s. This is due to the fact that any double fault situation within this group will result in a short-circuit current as in TN system.
For final circuits supplying electrical equipment with a rated current not exceeding 32 A and having their exposed-conductive-parts connected to an independent earth electrode electrically separated from the substation earth electrode, the maximum tripping time is given in Figure F13. For the other circuits within the same group of non interconnected exposed-conductive-parts, the maximum disconnecting time is 1s. This is due to the fact that any double fault situation resulting from one insulation fault within this group and another insulation fault from another group will generate a fault current limited by the different earth electrode resistances as in TT system.

  • Protection by circuit-breaker

In the case shown in Figure F18, the adjustments of instantaneous and short-time delay overcurrent trip unit must be decided. The times recommended here above can be readily complied with. The short-circuit protection provided by the NSX160 circuit-breaker is suitable to clear a phase to phase short-circuit occurring at the load ends of the circuits concerned.
Reminder: In an IT system, the two circuits involved in a phase to phase short-circuit are assumed to be of equal length, with the same cross sectional area conductors, the PE conductors being the same cross sectional area as the phase conductors. In such a case, the impedance of the circuit loop when using the “conventional method” (sub clause 6.2) will be twice that calculated for one of the circuits in the TN case, shown in Chapter F
The resistance of circuit loop FGHJ = 2R_{JH}= 2\rho\frac{L}{a}  where:
ρ = resistance of copper rod 1 meter long of cross sectional area 1 mm2, in mΩ
L = length of the circuit in meters
a = cross sectional area of the conductor in mm2

FGHJ = 2 x 22.5 x 50/35 = 64.3 mΩ
and the loop resistance B, C, D, E, F, G, H, J will be 2 x 64.3 = 129 mΩ.
The fault current will therefore be 0.8\times \sqrt 3 \times 230 \times 10^3/129 = 2,470 A.

  • Protection by fuses

The current Ia for which fuse operation must be assured in a time specified according to here above can be found from fuse operating curves, as described in figure F15.

The current indicated should be significantly lower than the fault currents calculated for the circuit concerned.

  • Protection by Residual current circuit-breakers (RCCBs)

For low values of short-circuit current, RCCBs are necessary. Protection against indirect contact hazards can be achieved then by using one RCCB for each circuit.

2nd case

  • It concerns exposed conductive parts which are earthed either individually (each part having its own earth electrode) or in separate groups (one electrode for each group).

If all exposed conductive parts are not bonded to a common electrode system, then it is possible for the second earth fault to occur in a different group or in a separately earthed individual apparatus. Additional protection to that described above for case 1, is required, and consists of a RCD placed at the circuit-breaker controlling each group and each individually-earthed apparatus.
The reason for this requirement is that the separate-group electrodes are “bonded” through the earth so that the phase to phase short-circuit current will generally be limited when passing through the earth bond by the electrode contact resistances with the earth, thereby making protection by overcurrent devices unreliable. The more sensitive RCDs are therefore necessary, but the operating current of the RCDs must evidently exceed that which occurs for a first fault (see Fig. F19).

Leakage capacitance (µF) First fault current (A)
1 0.07
5 0.36
30 2.17

Note: 1 µF is the 1 km typical leakage capacitance for 4-conductor cable.

Fig. F19: Correspondence between the earth leakage capacitance and the first fault current

For a second fault occurring within a group having a common earth-electrode system, the overcurrent protection operates, as described above for case 1.
Note 1: See also Chapter G Sub-clause 7.2, protection of the neutral conductor.
Note 2: In 3-phase 4-wire installations, protection against overcurrent in the neutral conductor is sometimes more conveniently achieved by using a ring-type current transformer over the single-core neutral conductor (see Fig. F20).


Fig. F20: Application of RCDs when exposed-conductive-parts are earthed individually or by group on IT system

Measures of protection against direct or indirect contact without automatic disconnection of supply

Extra-low voltage is used where the risks are great: swimming pools, wandering-lead hand lamps, and other portable appliances for outdoor use, etc.

The use of SELV (Safety Extra-Low Voltage)

Safety by extra low voltage SELV is used in situations where the operation of electrical equipment presents a serious hazard (swimming pools, amusement parks, etc.). This measure depends on supplying power at extra-low voltage from the secondary windings of isolating transformers especially designed according to national or to international (IEC 60742) standard. The impulse withstand level of insulation between the primary and secondary windings is very high, and/or an earthed metal screen is sometimes incorporated between the windings. The secondary voltage never exceeds 50 V rms.
Three conditions of exploitation must be respected in order to provide satisfactory protection against indirect contact:

  • No live conductor at SELV must be connected to earth
  • Exposed-conductive-parts of SELV supplied equipment must not be connected to earth, to other exposed conductive parts, or to extraneous-conductive-parts
  • All live parts of SELV circuits and of other circuits of higher voltage must be separated by a distance at least equal to that between the primary and secondary windings of a safety isolating transformer.

These measures require that:

  • SELV circuits must use conduits exclusively provided for them, unless cables which are insulated for the highest voltage of the other circuits are used for the SELV circuits
  • Socket outlets for the SELV system must not have an earth-pin contact. The SELV circuit plugs and sockets must be special, so that inadvertent connection to a different voltage level is not possible.

Note: In normal conditions, when the SELV voltage is less than 25 V, there is no need to provide protection against direct contact hazards. Particular requirements are indicated in Chapter P, Clause 3: “special locations”.

The use of PELV (Protection by Extra Low Voltage)

(see Fig. F21) This system is for general use where low voltage is required, or preferred for safety reasons, other than in the high-risk locations noted above. The conception is similar to that of the SELV system, but the secondary circuit is earthed at one point.
IEC 60364-4-41 defines precisely the significance of the reference PELV. Protection against direct contact hazards is generally necessary, except when the equipment is in the zone of equipotential bonding, and the nominal voltage does not exceed 25 V rms, and the equipment is used in normally dry locations only, and large-area contact with the human body is not expected. In all other cases, 6 V rms is the maximum permitted voltage, where no direct contact protection is provided.


Fig. F21: Low-voltage supplies from a safety isolating transformer

FELV system (Functional Extra-Low Voltage)

Where, for functional reasons, a voltage of 50 V or less is used, but not all of the requirements relating to SELV or PELV are fulfilled, appropriate measures described in IEC 60364-4-41 must be taken to ensure protection against both direct and indirect contact hazards, according to the location and use of these circuits.
Note: Such conditions may, for example, be encountered when the circuit contains equipment (such as transformers, relays, remote-control switches, contactors) insufficiently insulated with respect to circuits at higher voltages.

The electrical separation of circuits

(see Fig. F22)

The electrical separation of circuits is suitable for relatively short cable lengths and high levels of insulation resistance. It is preferably used for an individual appliance


Fig. F22: Safety supply from a class II separation transformer

The principle of the electrical separation of circuits (generally single-phase circuits) for safety purposes is based on the following rationale.
The two conductors from the unearthed single-phase secondary winding of a separation transformer are insulated from earth.
If a direct contact is made with one conductor, a very small current only will flow into the person making contact, through the earth and back to the other conductor, via the inherent capacitance of that conductor with respect to earth. Since the conductor capacitance to earth is very small, the current is generally below the level of perception. As the length of circuit cable increases, the direct contact current will progressively increase to a point where a dangerous electric shock will be experienced.
Even if a short length of cable precludes any danger from capacitive current, a low value of insulation resistance with respect to earth can result in danger, since the current path is then via the person making contact, through the earth and back to the other conductor through the low conductor-to-earth insulation resistance.
For these reasons, relatively short lengths of well insulated cables are essential in separation systems.
Transformers are specially designed for this duty, with a high degree of insulation between primary and secondary windings, or with equivalent protection, such as an earthed metal screen between the windings. Construction of the transformer is to class II insulation standards.
As indicated before, successful exploitation of the principle requires that:

  • No conductor or exposed conductive part of the secondary circuit must be connected to earth,
  • The length of secondary cabling must be limited to avoid large capacitance values(1) ,
  • A high insulation-resistance value must be maintained for the cabling and appliances.

These conditions generally limit the application of this safety measure to an individual appliance.
In the case where several appliances are supplied from a separation transformer, it is necessary to observe the following requirements:

  • The exposed conductive parts of all appliances must be connected together by an insulated protective conductor, but not connected to earth,
  • The socket outlets must be provided with an earth-pin connection. The earth-pin connection is used in this case only to ensure the interconnection (bonding) of all exposed conductive parts.

In the case of a second fault, overcurrent protection must provide automatic disconnection in the same conditions as those required for an IT system of power system earthing.

(1) It is recommended in IEC 364-4-41 that the product of the nominal voltage of the circuit in volts and length in metres of the wiring system should not exceed 100,000, and that the length of the wiring system should not exceed 500 m.

Class II equipment

Class II equipment symbol:  Box.jpg

These appliances are also referred to as having “double insulation” since in class II appliances a supplementary insulation is added to the basic insulation (see Fig.F23).


Fig. F23: Principle of class II insulation level

No conductive parts of a class II appliance must be connected to a protective conductor:

  • Most portable or semi-fixed equipment, certain lamps, and some types of transformer are designed to have double insulation. It is important to take particular care in the exploitation of class II equipment and to verify regularly and often that the class II standard is maintained (no broken outer envelope, etc.). Electronic devices, radio and television sets have safety levels equivalent to class II, but are not formally class II appliances
  • Supplementary insulation in an electrical installation: IEC 60364-4-41 (Sub-clause 413-2) and some national standards such as NF C 15-100 (France) describe in more detail the necessary measures to achieve the supplementary insulation during installation work.

A simple example is that of drawing a cable into a PVC conduit. Methods are also described for distribution switchboards.

  • For distribution switchboards and similar equipment, IEC 60439-1 describes a set of requirements, for what is referred to as “total insulation”, equivalent to class II
  • Some cables are recognised as being equivalent to class II by many national standards

Out-of-arm’s reach or interposition of obstacles

In principle, safety by placing simultaneously-accessible conductive parts out-of-reach, or by interposing obstacles, requires also a non-conducting floor, and so is not an easily applied principle

By these means, the probability of touching a live exposed-conductive-part, while at the same time touching an extraneous-conductive-part at earth potential, is extremely low (see Fig. F24).


Fig. F24: Protection by out-of arm’s reach arrangements and the interposition of non-conducting obstacles

In practice, this measure can only be applied in a dry location, and is implemented according to the following conditions:

  • The floor and the wall of the chamber must be non-conducting, i.e. the resistance to earth at any point must be:

  -  > 50 kΩ (installation voltage ≤ 500 V)
  -  > 100 kΩ (500 V < installation voltage ≤ 1000 V)
Resistance is measured by means of “MEGGER” type instruments (hand-operated generator or battery-operated electronic model) between an electrode placed on the floor or against the wall, and earth (i.e. the nearest protective earth conductor). The electrode contact area pressure must be evidently be the same for all tests.
Different instruments suppliers provide electrodes specific to their own product, so that care should be taken to ensure that the electrodes used are those supplied with the instrument.

  • The placing of equipment and obstacles must be such that simultaneous contact with two exposed-conductive-parts or with an exposed conductive-part and an extraneous-conductive-part by an individual person is not possible.
  • No exposed protective conductor must be introduced into the chamber concerned.
  • Entrances to the chamber must be arranged so that persons entering are not at risk, e.g. a person standing on a conducting floor outside the chamber must not be able to reach through the doorway to touch an exposed-conductive-part, such as a lighting switch mounted in an industrial-type cast-iron conduit box, for example.

Earth-free equipotential chambers

Earth-free equipotential chambers are associated with particular installations (laboratories, etc.) and give rise to a number of practical installation difficulties

In this scheme, all exposed-conductive-parts, including the floor(1) are bonded by suitably large conductors, such that no significant difference of potential can exist between any two points. A failure of insulation between a live conductor and the metal envelope of an appliance will result in the whole “cage” being raised to phase-to-earth voltage, but no fault current will flow. In such conditions, a person entering the chamber would be at risk (since he/she would be stepping on to a live floor).
Suitable precautions must be taken to protect personnel from this danger (e.g. non-conducting floor at entrances, etc.). Special protective devices are also necessary to detect insulation failure, in the absence of significant fault current.

(1) Extraneous conductive parts entering (or leaving) the equipotential space (such as water pipes, etc.) must be encased in suitable insulating material and excluded from the equipotential network, since such parts are likely to be bonded to protective (earthed) conductors elsewhere in the installation.


Fig. F25: Equipotential bonding of all exposed-conductive-parts simultaneously accessible