Lightning protection Protection of structures and open areas against lightning using early streamer emission air terminals.Part 3


Once the E.S.E. lightning conductor installation is completed, it should be

inspected to make sure that it complies with the provisions of this standard.

The purpose of this inspection is to make sure that:

– the E.S.E. lightning conductor is 2 metres or more above the entire protected


– the materials and the gauges used for the down-conductors are suitable,

C 17-102 -July 1S95 – 36·

– the down-conductors are routed, focated and electrically bonded as required,

– all the installation components are firmly secured:

– the safety distance(s) is(are) respected and/or equipotential bondings are


– the earth termination system resistance values are correct,

– the earth termination systems are interconnected.

This inspection should be performed visually under the conditions stated in part 6

of standard NF C 15-100.

However, where a conductor is entirely or totally hic;1den,its electrical continuity

should be tested. Such a test should conform to part 6 of standard NF C 15-100.


The inspection frequency is determined by the protection level. The following

inspection intervals are recommended:





Note: The intensified interval is recommended in a corrosive atmosphere.

An LPS should also be inspected whenever the protected structure is modified,

repaired or when the structure has been struck by lightning.

Note : Lightning flashes can be.recorded by a lightning flash counter installed on

one of the down-conductors.

7.2.1 Inspection procedure

A visual inspection should be performed to make sure that:

– no extension or modification of the protected structure calls for the installation of

additional lightning protective measures,

– the electrical continuity of visible conductors is correct, .

– all component fasteners and mechanical protectors are in good condition,

– no parts have been weakened by corrosion,

– the safety distance is respected and there are enough equipotential bondings

and their condition is correct.

·37 – C 17-102· July 1995

Measurements should be taken to verify :

– the electrical continuity of hidden conductors,

– the earth termination system resistance values (any variation should be


7.2.2 Inspection report

Each scheduled inspection should form the subject of a detailed report containing

all the findings of the inspection and the corrective measures to be taken.


Any faults found in the LPS during a scheduled inspection should be corrected as

soon as possible in order to maintain its optimal effectiveness.

C 17·102 -July 1935. -38 –



A 1.1


“~ ‘

…… :~:- …





Striking point determination

The formation or arrival of a stormy cloud .creates an electrical field (ambient)

between the cloud and the ground. This electrical field may exceed 5 kV/m on the

ground, thereby initiating corona discharges from ground reliefs or metal parts.

The lightning stroke begins with the formation of a downward leader within the

stormy cloud which propagates in steps towards the ground. The downward leader

conveys electric charges which causes the ground field to”build up.

An upward leader develops from a structure or an object linked to the ground. The

upward leader propi;lgates until it joins the downward leader and the lightning

current flows through the resulting channel. Other upward leaders can be emitted

~y several ground structures. The first one which joins the’ downward leader

determines the lightning striking point (Fig. A1).

Figure Ai

Note : This description only concerns the negative downward lightning stroke,

which is the only application case of the electro-geometrical model. This type of

lightning stroke is by far the most frequent.

A 1.2 Leader propagation velocity

Recent experimental data obtained from the nature shows that the average

velocities of the upward and downward lead~rs are comparable during the

attachment phase and the velocity ratio vuplvdown is close to 1

-39 – C 17-102-July 1995

Assuming that v = vup = vdown = 1 m/lJs (average measured leader velocities),


vup is the upward leader velocity,

vdown is the downward leader velocity,

v is the common velocity.



A 2.1 Triggering advance

An E.S.E. lightning conductor is built to reduce the average statistical time related

to the upward leader initiation. An E.S.E. lightning conductor features an triggering

advance as compared with a simple rod lightning conductor installed under the

same conditions. This gain in time is assessed in a l:Iigh-voltage test laboratory as

recommended in paragraph and Appendix C to this standard.

A 2.2 Gain in length of the upward leader

The gain in upward leader length L\L is given by 6L(m) = V(m/lJs) .6T(lJs).

The protected volume is determined from the protection model described above

on the basis of the electra-geometrical model.



Protection radius of a S.R. lightning conductor

In the caze of a simple rod, according to the electra-geometrical model, the

lightning striking point is determined by the ground object which is the first one to

be located at a distance D from the downward leader even though this object is

the flat ground itself. The distance 0 between the strike point and the upward and

downward leader joining point is known as the “striking distance” : this is also the

development length of the upward leader.

Therefore, it appears as if a fictitious sphere of radius 0 was centred on and

moving rigidly with the downward 1eader head.

Considering a simple rod of height “h” relative to the reference surface (building

roof, ground, etc.), there are three possibilities (see figure A 2) :

Figure A 2 • Fictitious sphere method

– if the sphere comes into contact with the vertical rod (A’) only, the vertical rod will

be the strike point,

– if the sphere comes into contact with the reference surface and not with the

vertical rod, the strike point will be on the ground at S only,

– if the sphere comes into contact with both the simple rod and the reference

surface at the same time, there are two possible strike points: A’ and C’, but the

lightning discharge will never strike the hatched area (see figure A 3).


Figure A 3

The striking distance 0 is generally given by the ‘following equation:

D(m) = 10.1213,where

I is the peak current ?f the first return stroke in kiloAmperes (kA).

A 3.2 Protection radius of an E.S.E. lightning conductor

In the case of an E.S.E. lightning conductor of triggering advance ~T, and with jL

=: v.,;lT. the possible strike points are A and C (Figure A 4) with a protection radius

Rp. such that: .

·41 – C 17-102 – July 1995

Rp = J h (20 – h) + tlL (20 +llL)


o is the striking distance

I1L is the upward leader length gain defined by ~L = v.~T

h is the E.S.E. lightning conductor tip height above the surface to be protected.

Rp is the E.S.E. lightning conductor protection radius

11T is the triggering advance ofthe E.S.E. lightning conductor. ,

Figure A4

C 17-102 -July 1995 . – 42 –






The lightning risk assessment guide is intended to assist the design manager in

the analysis of all the criteria used to assess the risk of damage due to lightning

and to determine the need for protection and the required protection level. Only

the damage caused by a direct lightning stroke on the structure to be protected

and the lightning current flow through the LPS is covered. .

In many cases, the need for protection is obvious. EXamples are:

– large crowd,

– service continuity, –

– very high lightning stroke frequency,

– tall or isolated structures,

– building containing explosive or flammable materials, or irreplaceable cultural


Some typical consequences of a lightning stroke on several types of common

structures are listed in table B1 for information.

Structure classification Structure tLyipgehtning consequences


to the lightning

Risk of fire and dangerous sparks.

tRRhieisskkIQocSofSnsostefepqvuevenontlitltaagtitoeon. p·ocwoenrtroflaialunrde fo:ocdattdleistrdibyuintigona. s a result of

ThReisaktreo,f spcahnoicol,and fire alarm failure resulting in delayed fire

hypermfigahrktientgs., sports

Bank, insurance

ab~ve plus problems related to loss of information

and computers malfunction.



Same as above plus problems related to patients in intensive

care units and evacuation of handicapped persons.

Additional effects depending on the factory contents, ranging

lforosms. minor damage to unacceptable damage and production

Irreplaceable losses in cultural heritage.

-43 – C 17-102 – July 1995

Table 81

Note : Sensitive electronic equipment may be installed in any type of structures

and can be. $asily damaged by voltage surges due to lightning.

A riskasses’Sment method is proposed in this guide, it takes into account the

lightning risk and the following factors:

.. ,

,1. Building environmer;1t,’: .

.’····2. Type of construction’, :….

. .’..- 3. Structure contents· •… ··

‘ ..’ 4. Structure occupancy,

, ..·5~ Ligjltning st~oke consequences.

The building’ location in the environment, and the euilding height are taken into

consideration for the computation of the’exposure risk ..

In some cases ho~ever, certain criteria specific to ~ ~iven structure cannot be

assessed and may prevail over any other consid.eratjon. Protective measures can

then be applied which are more .stringent than’th.ose resulting from the application

of the guide. ….’: .’ .’

The selection of the suitable protection level for the ELPI to be installed is based

on the expected direct lightning stroke frequency on the structure or the area to

be protected and on the accepted yearly lightning stroke frequency Nc.

B 2 Determination of Nd and Nc

8 2.1 Lightningflash densityNg

The lightning flash density is expressed as the yearly number of lightning flashes

per km2 and can be determined by :

– ~ theslrokedensitymap Na i1 FJgUre 84. rnthiscase,Ng= Na/2.2

– oonsuIlinga [ghtningkx:ationnebNork, .

-l.ISirg thelocalisoI<eraunK:: level~ : Ng l1BX =0.04 ~ 125 == ~/10

The value Ng max takes into account the maximum lightning density and the

Note: The map of figure 84 shows the stroke density. The constant 2.2 is the

average ratio of the number of strokes to the number of flashes.

C 17-102 -July 1995 -44 _.

B 2.2 Expected frequency Nd of direct lightning to a structure

The yearly average frequency Nd of direct lightning to a structure is assessed

using the following equation:

Nd:: Ng max .. Ae . C1 10-6/year, where: (Equation 6)

Ng is the yearJy average lightning flash density in the region where the structure is

located (number of lightning flashes/year/km2);

Ae is the equivalent collection area of the isolated struct~re (m2);

C1 is the environmental coefficient.

The equivalent collection area is defined as the ground area having the same

yearly direct lightning flash probability as the structure.

According to table 82, the equivalent collection area Ae for isolated structures is

defined as an area of ground surface which has the same annual frequency of

direct lightning as the structure. It is the area between the lines obtained by the

intersection of the ground surface and 1 :3 slope line passing through the top of

the structure and revolving around the structure (see figure 83).

For a rectangular structure with length L, width W an~ height H, the collection area

is then equal to :

Ae:: LW + 6H (L + W) + 9;rH2 (Equation 7)

The topography of the site and the objects located within the distance 3H from the

stru(;ture significantly affect the col!ection area. This effect is taken into account by

applying environmental coefficient C1 (table 82.).

Table 82 – Determination of environmental coefficient C1

– When the equivalent collection area of a structure entirelY.covers that of another

structure, the latter is disregarded.

– When the collection areas of several structures are overlapped, the

corresponding common collection area is considered as a single collection area.

Note: Other more sophisticated methods may be used to. assess the equivalent

collection area with greater accuracy.

C 17-102 -July 1995

Figures 83 – Typical computations

Note: Specific regulations may impose other values for Ne in some cases.

C 17-102 – July 1995 -48 –


The tolerable lightning frequency Nc is compared with the expected lightning

frequency Nd.

The result of this comparison is used to decide whether an LPS is required and, if

so, the protection level to be used:

– If Nd ~ Nc, the LPSis not required systematically.

– If Nd > Nc• an LPS of effectiveness E ~ 1 – Nc/Nd should be installed and the

associated protection level selected in table 8 10.

The LPS design shall meet the specifications given in the standard for the selected

protection levels.·

When an LPS with an effectiveness factor E’ smaller than the computed factor Eis

installed, additional protective measures should be taken. Typical additional

protective measures are:

– measures limiting the step or contact voltage,

– measures restricting fire .propagation.

– measures reducing the effects of voltage surges induced by lightning on

sensitive equipment.

A practical method for selecting the protection level is given in fig!Jre 89.

Table 810 gives the critical effectiveness values Ec corresponding to the limits

between the protection levels and the protection levels corresponding to computed

effectiveness E.

– 49- I 17-102 ·July 1995

Table 89 – Determination of protection requirement and protection level

Result Data input Computation


Figure 84 – Map of lightning stroke density Na in France

This map is based on statistical data coming from measures collected since 1987 by the

national network of lightning detection.

– 51 – C 17-102 • July 1995





The effectiveness of an E.S.E. lightning conductor is assessed by comparing the

upward leader triggering time emitted by the E.S.E. lightning conductor against the

upward leader triggering time emitted by an S.R. lightning conductor.

For this purpose, the SR lightning conductor and E.S.E. lightning conductor are

assessed one after the other under the same electrical and geometrical conditions

during laboratory tests simulating the natural conditions of the upward leader

initiation (positive upward leader). .•

C 1.1 Ground field simulation

The natural ground field existing before a lightning stroke affects the conditions of

corona formation and of existing space charges. The natural ground field should

therefore be simulated: its value ranges from 10 kV/m to 25 kV/m.

C 1.2 Impulse field simulation

To reproduce the natural phenomenon as closely as possible, the ground field

bUild-up is simulated by a waveform the rise time of which ranges from 100 ~sec

. tsoho1u0ld00be~bseetcw.eeTnhe2.w10aaeafonrdm2.1s0l0g>Ve/mw/isthoin the. upward leader initiation region


C 2.1 Positions of lightning protection systems to ~e compared

The upper plate/air-termination distance should be sufficient for the upward leader

to propagate in free space and, in any case, over a length greater than 1 m (d ~

1m). The objects to be compared should be placed in the same electrical

environment which is independent of their locations : they should be tested one

after the other and centred on ground above the· plate and their height should be

the same.

C 2.2 Dimensions of experimental set-up

The upper plate/ground distance (H) should be greater than 2 m. The ratio h/H of

the air-termination height to the plate height above ground level should range from

0.25 to 0.5. The smaller horizontal dimension of the upper plate is the upper

plate/ground H distance.

C 17-102 -July 1995 ·52·



.PTS h

Configuration 1

Figure C1

Configuration 2


C 3.1 Electrical parameters

– Applied voltage waveforms and amplitudes (ambient field calibration, pulsed

voltage wave, associated current, etc.);

– Continuous polarisation setting;

– Initiation setting on the reference equipment (simple rod lightning conductor) :

initiation probability equal to 1.

C 3.2 Geometrical conditions

The distance d should be strictly the same in each configuration : it should be

checked before each test.

C 3.3 Climatic parameters

The climatic conditions should be recorded before and after testing in each

configuration (pressure, temperature, absolute humidity).

C 3.4 Number of lightning strokes in each configuration

The number of lightning strokes should be statistically adequate in each

configuration, e.g. about one hundred lightning strokes in each configuration.

C 3.5 Triggering time

The criterion adopted for assessing the effectiveness of an E.S.E.lightning

conductor is its capacity to initiate an upward leader before an SR lightning

conductor under the same conditions. The average upward leader triggering time

T is measured for each usable lightning stroke on the SR lightning conductor and

then on the E.S.E. Iight~ing conductor.

– 53- C 17:’102-July 1995

· C4

C 4.1


Experimental Assessment of the average triggering times

The upward leader triggering times measured during usable shocks on an SR

lightning conductor and an E.S.E. lightning conductor are used to compute the

average triggering times T’SRLC and T’ESELC in compliance with the selected

experimental curve parameters.

C 4.2 Reference waveform

The reference waveform is defined by a rise time TR of 650 I-Isecand a shape as

shown in the graph of Figure C2.

C 17-102 – July 1995 – 54-

Reference waveform

C 17-102· july 1995

C 4.2 Determination of the triggering advance of the E.S.E. lightning conductor

The experimental curve is plotted on the same graph as the reference

waveform to which is assigned the same field value EM as the experimental

field EMexp.

Lines are dropped from T’SRlC and T’ESElC onto the rererence curve and the

ordinates of.the intersection points give the E field values. The triggering times

are obtained by projecting lines from the E values to the points where they

intersect the reference curve; the associated values on the x-axis gives the

triggering advance ~T (JJsec)= T’SRlC and T’ESElC’

Note : The method proposed above can be used to determine a AT value in a

laboratory. Using the upward leader initiation fields which only-depend on airtermination

height h, a tlT value independent of d can-be _determined. This

transposition is accomplished using the continuous leader starting threshold

field model developed by Rizk & Berger.

C 17-102 – July 1995 ·56·





D 1.1 . Impulse component waveforms (discharges) of a lightning stroke

Figure 01 shows a few lightning current waveforms. Such lightning currents

have been recorded at the San Salvatore Mount research station in

Switzerland. Tables 03 to 015 show the cumulated frequency distributions of

the lightning characteristics. .

Negative and positive lightning currents measured on San Salvatore

lugano (Switzerland)

Figure D1 – Examples of lightning currents

D 1.2. Distribution of the different lightning parameters

A considerable number of parameters are used to describe the lightning

impulse (or impulses in case of negative lightning), including in particular: .

current amplitude, rise time, decay’time, charge and specific energy.

These parameters refer to the actual lightning stroke waveforms as measured

to compute the distribution statistics. Initially, the amplifude, decay time and rise

time may be considered as defined as in a laboratory. The charge corresponds

to f idt and the specific energy to I j2dt. The usefulness of these parameters is

explained below.

The steepness (steepest current slope in kNlJsec) is also sometimes a useful

data for characterising an impulse though it is related to other parameters

already defined: rise time and amplitude. .

C 17-102 – July 1995

The total lightning flash, including the impulse(s) and the following current

flowing in the interval between two impulses is essentially characterised by its

total duration.


The parameters mentioned in the foregoing do not generally have the same

effects or failure modes as regards the different types of equipment.

The current amplitude is useful for addressing the voltage surge problems and

mechanical load problems generated by lightning.

The rise time is only used to address the voltage surge problems.

The decay time is related to mechanical loads in that it is used to determine the

electromagnetic force application time; it is mainly representative of the

lightning stroke energy in connection with the amplitude. To represent this

energy, the amplitude/decay time binomial can be replaced by :

– Specific energy I j2dt (amplitude and decay time) when the LPS component

dimensions are considered (connectors, conductors, etc.);

– Charge I idt (amplitude and decay time) in the case of the characteristics of

surge protective devices connected to lightning protection systems (E.S.E.

lightning conductor + earth-termination system) -or metal melting at the

lightning strike point.

o 2.1 Thermal effects related to charge quantity Q

Thermal effects are observed in lightning protection inst::lll:Jtion:; especially

when the air-termination systems have sharp tips on· which melting is

sometimes observed over a maximum of a few millimetres. In the case of flat

surfaces (sheet-metal plates), evidence of melting is found which may result in

complete piercing.

An exceptional lightning stroke (300 C) is capable of piercing sheet-metal plates

of up to 2-3mm thick.

This accounts for the minimum thickness requirement when a metal plate is

used or likely to serve as a lightning collector (e.g., 4 mm for iron, 5 mm for

copper) ..

Low-intensity discharges with a long duration may readily ·cause ignition. As

lightning discharges are usually accompanied by a continuing current. lightning

strokes are seldom cold. Even dry wood can be ignited by this kind of lightning

with long-lasting continuing currel}ts.

Poor contacts are particularly dangerous points along the lightning current path.

Contact resistance values of a few thousandths of an ohm already generate

enough heat to melt appreciable quantities of metal producing sparks. When a

readily flammable material is located near such poor contact points, indirection

ignition may result. This kind of sparking is particularly dangerous in premises .

exposed to a risk of explosions and in explosive manufacturing plants.

C 17-102 – July 1995 – 58 –

“D 2.2

D 2.3

Thermal effects related to current integral J j2dt

When the lightning current enters a metal conductor in which it can propagate,

the resulting heat dissipation obeys the Joule’s law which involves the square of

the current i2, current flowing time t and ohmic resistance R.

Significant thermal effects are therefore encountered especially at highresistance


The direct-current resistance measured on a conductor should not however be

taken as the resistance value R. Lightning currents are short shock waves

which produce a skin effect as in the case of high-frequency currents, Le., the

current flow is confined to a thin conductor surface layer a few tenths of a

millimetre thick, as measured in direct current, which corresponds to the total

cross-sectional area.

There are no visible consequences of this heating, in spite of the skin effect,

when the conductor gauge is large enough. Temperature rises up to the melting

point .temperature only occur in conductors having a small gauge or high

resistivity. Melting effects are often observed, for instance, in antenna cables

and wires. On the other hand, cases of melting are seldom observed on larger

gauge wires of a few millimetres in diameter (such a barbed wires). Melting has

never been seen in lightning conductors having the gauge recommended in this


On the other hand, the current flow in poor conductors releases a large amount

of energy in the form of heat. This is why the water contained in wood, concrete

and similar materials is heated up and vaporised. The entire phenomenon lasts

a very short time and, as a consequence of the subsequent pressure rise,

trees, wooden masts, beams and walls burst. Explosive effects of this kind

more particularly occur in places where mpisture has accumulated (slits,

vessels full of sap) or the current density has risen significantly, i.e., at the

points of current entry or exit between a material having a poor conductivity

(cement) and a material having a high conductivity (attaching clamps of a

damaged lightning down-conductor, electrical conduit cramps, water and gas

pipe steel clamps).

Electrodynamic effects

Significant mechanical loads may occur only when sections of the lightning

current path are laid out one relative to the other in such a way that one of them

is located within the magnetic field generated by the other. In this case, the load

increase is inversely proportional to the distance between these sections. Small

turns are subjected to considerable enlarging loads. Considering a 10 cm

diameter ring made of 8 mm diameter wire, a very heavy lightning current of

100 kA will apply a force of 1200N to each centimetre of the periphery. With a

2m” diameter, the for.ce would drop to 140N. Due to the reciprocal interaction

between the lightning current in a conductor and the Earth’s magnetic field,

mechanical effects of only about 10N per metre of conductor can be produced;

such effects are trivial.

– 59- C 17-102 -July 1995

In addition to these repulsion forces, which may distort conductors in rare

cases, there are also strong attraction forces between parallel lightning current

paths”when they are quite close. In this way, thin tubular antennae are crushed

and parallel conductors knock together.

D 2.4 Potential differences and arcing

The surprising profusion of spark traces observed after a violent lightning

stroke, sometimes even in buildings provided with lightning protectioR systems,

can be explained by two effects well known in electrical engineering : the earthtermination

potential rise, which mainly depends on the peak intensity

(amplitude) of the drained current, and the induction phenomena which mainly

depend on the dildt gradient (leading edge steepness) of this current.

D 2.4.1 Earth-termination potential rise

Due to the earth-termination resistance R, resulting from the resistivity of the

soil itself, there is a potential difference between the LPS down-conductor and

nearby points while the current is flowing. The total potential rise relative to the

unaffected remote “ground’ (therefore remaining at the conventional zero

potential) is expressed by Ohm’s law : U = RI

A 100 kA current flow through a 5-ohm earth-termination system will cause a

potential rise in the lightning current draining system of 500 kV relative to

remote ground points.

Such a potential rise is actually distributed in the ground according to a law

which depends on the type of earth-termination system and the soil


All the conductive parts of the structure which are connected to the earth in any

way (heating systems, pipe lines, electrical systems, cable armours) are also

subjected to a potential rise if they are not interconnected. The only way to

prevent insulation breakdown is to provide an electrical connection through

down-conductors to independently earthed parts. In thi~ way, these become

integral parts of the lightning protection system and can therefore drain part of

the lightning current according to branch circuit laws. Their connection to the

down-conductors make them an integral part of the LPS.

As no conductive connections to live electrical lines can be made, this standard

recommends the installation of voltage surge protective devices known as

lightning arresters (varistors or spark gaps). However, these lightning arresters

should then be sized to withstand a non-negligible portion (from a few per cent

up to 50 per cent, approximately, in the worst case) of the lightning current

striking the LPS.

Note : Given the frequencies involved in lightning phenomena, the earthtermination

system impedance should be taken into account in addition to the

measured earth-termination system direct-current resistance.

C 17-102 -July 1995 – 60-

D 2.4.2 Induction phenomena

Shorter ~istance between down-conductbr and metal building structures

An LPS down-conductor forms open loops with the various metal structures of

a building (water pipes, central heating system, electrical power lines, etc.).

These loops will be subjected to induction phenomena and electromotive ~orces

will appear between their open ends. This standard allows for this phenomena

in article 3.

TABLES 02 TO 014

These’tables are extracted from IEC 1024-1, Part 1, Se.ction 1, Guide A,

“Selection of protection levels for lightning protection systems”

Basic values of lightning current parameters

Cumulative frequency distribution





People standing outdoors run the greatest risk of being struck by lightning,

‘. whether directly or caused by the step voltage. For people inside a structure,

.the hazards are due to :

{a} the abrupt potential rise in items connected to lines leading from the

outside such as power lines, telephone lines, outdoor TV antenna cables;

(b) metal objects within the structure which may” also be brought to high

potentials : contact voltage.

The measures stated in this standard to prevent dangerous sparking are

designed to reduce the risks run by people inside structures.


To protect themselves against lightning, individuals should take the following

minimum precautions : .

(a) look for a shelter in a place covered by an earthed roof is or an all-metal


Note: Conventionally manufactured tents do not provide protection ..

(b) when there is no shelter nearby, reduce one’s height (crouch down) and

surface area on the ground Goin the two feet) and do not touch any earthed

object with the hands, .

(c) do not ride a bicycle or a horse. Do not remain in an open-top car,

(d) do not walk or swim in water,

(e) keep away from high places, or tall or isolated trees. If the vicinity of a tree

cannot be avoided, stand beyond the foliage limits. .

(f) do not touch or stand next to metal structures, metal fences, etc.,

(g) do not carry any object which extends above the head (umbrella, golf club,

tool, etc.),

(h) do not use or minimise the use of cord telephones,

(i) do not touch any metal object, electrical appliances, window frames, radio

sets, TV sets, etc.

C 17-102 – July 1995 – 64-

Deja un comentario