Note: Descriptions are shown in the official language in which they were submitted.
2 ~ 7 3
The invention relates to a method and an ap-
paratus for checking the state of electrical insulation
of an electrically condl~cting work placed in an electro-
lytic medium and connected by a network of electrical
conductors, on the one hand to a counter-electrode placed
in the abovementioned electrolytic medium and, on the
other hand, to a reference electrode which is in contact
with this same electrolytic medium, a voltage existing
na~urally between this electrode and the said work.
During trials undertaken within the context of
the invention, the electrically conducting work consisted
of a buried gas conduit made of a metallic material. The
prime usefulness of the method which is the s~bject of
the invention consisted in checking the quality of the
placement of this pipe which was in this particular case
covered with an electrically in~ulating coating forming
a passive protection intended to remove the pipe from the
electrolytic environment of the ground.
Experience having shown that placements in
particular of such buried pipes are not always perfect
and that furthermore the coatings for passive protection
are not always entirely effective, thought was con-
sequently given to the problem of checking the state of
electrical insulation of these works, once the latter are
in place in their medium.
Now, ~uch checking in practice encounters a
certain number of difficultie~ relating in particular to
the variable nature over time of the parameters to be
meaRured, these variations possibly in particular being
induced by poorly insulated DC in~tallations (railway
traction, electrolysis plant, etc).
Problems of measurement error~ also arose due to
the phenomena of polarisation of the works, these polari-
sations changing over time and as a function of the
electxical change in particular of the worX, of the
counter-electrode and of the disturbances which may,
nearby, cross the electrolytic medium.
After lengthy studies, the Applicant succeeded in
developing a truly reliable and effective method making
2~8~73
it possible to check the state of electrical insulation
of a conducting work, this metho~ being characterised in
particular in that:
a) the voltage between the work and the reference
S electrode is sensed at least once and recorded,
b) ah impos0d circulation of current
who5~ intensity which is sensed and recorded
,is established in the conducting
network, between the work and the counter-electrode,
c! the voltage between the work and the reference
electrode, current circulating, is sensed at least once
and recorded anew,
d) the difference ~U between the voltages recor-
ded in a) and c) is calculated and recorded without and
with imposed current circulation respectively,
e) the said imposed current circulation is inter-
rupted and steps a), b), c) and d) are repeated a given
number of times whilst varying, during step b), the
intensity of the imposed current,
f) the change is next deduced therefrom in the
said voltage difference QU as a function of the intensity
I o~ current established in the network, which change is
linear o~ the form ~U = AI (A = constant) from I = 0 up
to a specified value of intensity of this current,
25 g) the slope A is calculated,
h) and the state of eleotrical insulation of the
woxk is deduced therefrom by identifying the said slope
with the electrical re~istance of this work.
In this way, through rapid capture of the mea-
sured or sensed potentials, it will be possible to obtain
quantities approximating the actual values existing on
site.
Moreover, by limiting oYer t.ime, through rigorous
monitorinq, the settling periods of the imposed currents,
it will be possible to undertake series of trials of
r~asonable duration by repeating the measurements a large
number of times and by statistically processing them so
as to detect and negate in particular the disturbing
presence of the stray curren~s.
2~8~73
-- 3 --
With this aim, a complementary characteristic of
this invention moreover provides that, the voltages
between the work and the counter-electrode being
disturbed by electrical sources generating frequencies
S which modify over time the voltages to be measured, these
said frequencies are eliminated by filtering ~uring the
abovementioned steps a) and/or c) when the said
frequencies are beyond a predetermined threshold and/or
correspond to known frequencies identified as disturbing.
As will be seen below in the description which
will follow, and still in order to meet this problem of
electrical disturbances, provision has also been made in
the invention to preferably statistically process the
measurements so as to take into consideration only values
judged to be consistent and reasonable, as explained in
particular in Claims 3, 4 and 7 attached.
Apart from such a checking method, the subject of
the invention is also an apparatus intended to check the
state of electrical insulation of the chosen work.
In accordance with tha invention, this apparatus
is characterised in particular in that it comprises:
- means of measuring the voltage between the
reference electrode and the work,
- a generator for generating a current
Z5 between the work and the counter-electrode,
- means of measuring thi~ current,
- a clock for interrupting and establishing
sequentially the circulation of the current Lmposed by
the generator between the work and the counter-electrode,
and triggering measuremants of the said current and of
the said voltage,
- means of processing the measure-
ments made in order to calculate, for each intensity of
current, the difference ~U between the voltages
measured without and with circulation o~ said current,
and in order to calculate also the ratio A between the
difference in the abovementioned vol~ages (~U) and the
said corresponding intensity of current,
- and means of displaying the calculated ratio A.
20~8~3
-- 4
Other characteristics and advantages of the
invention will emerge further from the following descrip-
tion made with reference to the attached drawings in
which:
- Figure 1 represents a general configurational
diagram of a circui-t permitting implementation of the
me~hod of the invention,
- Figure 2 is a block diagram representing in
more detail the processing and dis~lay unit of
the apparatus of the invention,
- Figure 3 presents an illustrative curve of a
possible change, with respect to a work, in the potential
difference U as a function of the Lmposed current I,
- Figures 4a, 4b present two curves which can be
obtained with the voltage-measuring means used by the
apparatus of the invention, in the absence of a filter
permitting elimination of the interference frequencies
(Figure 4b) and in the case where such a filter is used
(Figure 4a).
- And Figures 5a and 5b present curves of the
change over time of current and voltage signals
respectively.
If referencs is made to the drawings, there is
seen firstly in Figure 1 a metal pipe 1 constituting in
this example the electrically conducting work whose state
of insulation it is desired to check. This pipe 1 is
buried withln an elec~rolytic medium constituted here by
the ground 3. In this ground i~ also buried a counter-
electrode 5 connected to the pipe by an electrical
network in which a current will be made to circulate.
Under the action of such a current, the piece 5 generally
behaves like an anode serving to inject the current into
the ground. In practice, the piece 5 may take the form of
a crosspiece or even a metal stake made of for example
steel. Hence it will be po~sible to promote a cathodic
reaction in the region of the work, whilst transferring
the natural oxidation reaction to the anode 5, de~enera-
tion of which is accepted a priori (it will however be
noted that sometLmes for the protection of certain
2~847~
-- 5 --
metals, such as certain stainless s~eels or aluminium
alloys, which can be passivated in the electrolyte, it
may be appropriate to make the ~counter-electrode~ play
the role of cathode, then making an "anodic protective~
current circulate).
To ensure the intended circulation of the current
in the network, use is made of a current generator 7
mounted in series in the network and able to deliver a ~c~shn~cr
direct current to the piece 5~ ad~nt~g~o~si~
For the purposes of monitoring, with this genera-
tor 7 have been associated in series a clock 9 making it
possible to interrupt and reestablish sequentially the
circulation of the current imposed by the generator 7 and
to control the triggering of the measuring means 15 and
17 presented below. The clock 9 may consist for example
of a generator of low-frequency signals forming an
interrupter with adjustable pulse times, for example from
l ms to 1 s.
In the field of the protection of metal works
placed in an electrolytic environment, it is known that
the potential difference between the work and the elec-
trolyte is representative of the electrical state of this
work.
In actual fact, thi~ potential difference is
usually measured between the metal of ~he work and a
reference electrode in contact with the electrolyte but
situated some distance from the metal surface of the
work. This is why there have been represented, at 11,
such an electrode placed on the ground and connected up
to the abovementioned electrical network, at 13, a
conductor connecting the negative terminal 7a of the
generator 7 and, at la, the metal structure of the pipe
1.
As known per se, a metal electrode immersed in an
electrolyte can be regarded a~ a half~cell. Now, only the
potential difference of a cell can be measured. In the
present case, this cell will therefore consist of the
electrochemical chain comprising the metal work to be
studied and the reference electrode which, in regard to
2 ~ 7 ~
practice, is usually of the Cu/CuSO4,Ag/AgCl type or else
in calomel (Hg/HgCl-KCl), whose respective potentials are
known.
To measure the potential gradient between the
pipe 1 and the electrode 11, provision has of course been
made for a Yoltage-measuring sensor schematised at 15 in
Figure 1 and conn~cted up between the electrode and the
point 13 of the network. This first measuring means could
consist of an oscilloscope with memory, with an associa-
ted printer.
Given that in the invention the recording of theintensity of the currents which may circulate in the
network is to be taken into consideration, a second
measuring means 17 has been provided. To this end, the
use could be envisaged of an oscilloscope connected up in
parallel across the terminals of a resistor 19 placed in
series between the positive output 9a of the clock 9 and
the piece or outlet 5. Hence, by correctly choosing the
value of the resistor 19, a direct reading of the current
circulating in the network could be made on the measuring
apparatus 17.
In order to process these voltage and current
measurements, the apparatu~es 15 and 17 have been con-
nected to a capturing, processing and viewing unit 21, a
more detailed illustration o~ which is presented in
Figure 2.
In this figure is ~ound firstly the terminal la
of the pipe, the connection terminal of the outlet 5, as
well as the connec~ion ter~inal of the electrode 11. The
current qenerator 7, the clock 9 and the resistor 19 have
also been represented schematically. Finally, there are
the two means 15 and 17 for measuring respectively the
~oltag~ U between the pipe and the electrode and, on the
other hand, the current I between this same pipe and the
piece 5.
During the trials carried ou~, it wa apparent
that the voltages read between the work and the anode
piece were sometimes greatly disturbed by electrical
sources generating disturbing frequencies in the
2~8~7~
-- 7
electrolytic medium 3. Provision was therefore made for
the possibility of adding at the output of the measuring
apparatus 1~ a filter 23 of the low-pass filter type
possibly with a rejector making it possible thereby to
eliminate by filtering the sensed ~requencies having
values situated beyond a predetermined threshold and/or
corresponding to frequencies identified as particularly
disturbing. During the tests carried out, it had for
example been chosen to eliminate frequencies above 5 ~z.
For the sake of clarity, in Figures 4a and 4b
have been illustrated the shape of the signals obtained
without filter (Figure 4b) and with filter (Figure 4a),
from the voltage-capturing apparatus 15.
In Figure 4a, one can immediately observe the
beneficial effect of the filter which, without especially
deforming the signals, makes possible a smoothing benefi-
cial to the quality of the measurements.
As has certainly been understood from the above,
an Lmportant aspect of the invention lies in fact in the
manner in which the readings sensed and captured by the
two, voltage and current, measuring means 15 and 17 are
processed.
In Figure 2 ha~ been schematised with the block
25 the processing unit used and which contains essen-
tially a microprocessor 27, programmable memories 29,proce~sing programs 31 and a calculator 32.
O~ course, a viewing unit 33 has been added to
the proces~ing unit. This viewing unit could take the
form of for example a printer or a computer screen linked
with a keyboard 35 itself connected to ~he processing
unit 25 for programming and control, via the control unit
24, of the operations which will now be presented.
As has been stated, the aLm of the apparatus just
presented is to permit automatic checking of the state of
electrical insulation of a conducting work.
For this purpose, it will therefore be sought to
establish rigorously the relationship existing between
the pipe l/electrode 11 voltage U and the current I
circulating in the installed electrical network.
20~8d7~
-- 8
Choosing to take such a relationship into con-
sideration is explained by the fact that, whatever the
work and its environment, the characteristics obtained
have the particular feature of being linear and of the
type ~U = AI with (A = constant), from the origin I = O
up to a certain value of current (see Figure 3). It is in
fact the slope A (that is to say, as understood, the
resistancq of the work) which will enable the operator in
possession of this information to analyse the state of
electrical insulation of the work.
To accomplish this analysis, the rate imposed by
the clock 9 will firstly be suitably adjusted, so that
for example the generator 7 establishes an imposed
current I in the network for about 0.5 to 1 second, with
exposure times to zero imposed current of the order of
for example 8 to 10 seconds.
In Figures 5a and 5b respectively have been
represented two examples of signals recorded, for current
and voltage. It will be noted that the duration of the
in;ections of current made by the generator 7 is repre-
sented by the interval T3 - T1, whereas the interval
between two injections is seen to correspond to T5 - T3.
To understand the functioning of the system, we
firstly take up position at the time origin. Up to time
T1, the generator 7 will deliver no current. During this
interval, the voltage between the work 1 and the elec-
trode 11 will then be sensed and recorded, preferably
several times. At time Tl, a constant-value square-wave
current of specified intensity will be established in the
network. The ~ensor 17 will then sense and record the
intensity of this current. Following the settling tLme
for the filter 23 (T2 - T1), the voltage between the work
and the electrode will then be sensed and recorded anew,
preferably several tLmes. This takes place up to time T3
which marks the end of the square-wave current injection.
The processing unit 25 will then calculate and record in
memory the difference ~U between the voltages recorded
previously, without and with circulation of imposed
current I.
2~ 7 ~
Of course, during this time, new work/electrode
voltage measurements have been sensed and recorded by the
oscilloscope 15, current cut, until the clock 9 commands
a new square-wave current for the generator 7 and until
a new measurement of voltage and of current is taken into
consideration for a new calculation of the difference in
the potentials AU.
Thus, as t.he measurements proceed, the processing
unit 25 will be able to supply to the viewing unit 33 the
change, as a function of the intensity I of the imposed
current, in the difference in the pipe/electrode poten-
tials ~U with and without circulation of this imposed
current.
This change being, as indicated above, linear and
of the form ~U = AI (A = constant) from I = 0 up to a
specified value of intensity, the calculator 32 will be
able to calculate the slope A and thus supply the unit 33
with the value of this slope A which can be identified
with the electrical resistance of the work.
Having presented the general principle of the
measurements made, the various improvements which may be
added in order to guarante~ the reliability of the
measurements will now be described briefly.
Firstly, within the current established and
current cut time intervals (T3 - T2 and T5 - T4 in Figure
5b), it will in practice be preferred to sense and record
several voltage values, uniformly staggered over time, of
the work/electrode gradients.
Next, it was seen to be preferable to complement
these staggered measurements with two tests carried out,
advantageously in combin tion, current cut and current
established.
For this purpose, provision has been made to
compare with one another the voltage readings made during
the interval T3 - T2 and then the interval T5 - T4. If
these values lie within a predetermined range, and
evidence a steady potential, the following steps of the
method are undertaken. If on the other hand one at least
of the said values does not lie within this range, no
`` 2~3~73
-- 10 --
voltage value is recorded and a new cycle of measurements
and of generation of new square-wave currents is renawed
a few moments later, by repeating the comparison step
above.
Hence, it will be po~sible to avoid taking into
consideration disturbed readings, during the steady
states of the system with and without imposed current.
However electrical disturbances may also inter-
vene during the transient phases of voltage drop or rise
(T2 - Tl or T4 - T3 in Figure 5b). To remedy this pro-
blem, provision has been made to repea~ several tLmes the
steps of measuring and capturing pipe/electrode voltages
with and without imposed current, for the same intensity
of generated current, so as to thereby engender the
calcula~ion of several potential gradients ~U calculated
from constant square-wave currents. During this calcula-
tion of ~U, the processing unit 25 will then statisti-
cally process these values of potentials and will record
values only if the latter converge to an average value.
It will be observed that these various precau-
tions appertain to the capturing and processing of the
measurements of potential. As f 2r as the checXing of the
measurements of current is concerned, a monitoring loop
has been envisaged7 schematised by the line 37 in Figure
2 and intended to monitor the intensity values actually
supplied by the current generator 7.
More precisely, provision has been made to record
fir~tly the intensity of the square-wave currents con-
trolled by the operator and introduced into the memory
units 29 of the machine. After having ~ensed, with the
measuring means 17, the intensity of current supplied
periodically by the generator 7, the processing unit 25
and more particularly the calculator 32 will compare the
controlled recorded intensity and the sensed intensity.
If the deviation between these two intensities lies
within a predetermined range the sensed intensity will
then be recorded. Otherwise, provision may be made for an
alarm signal such that it may be displayed on the viewing
unit 33, thus enabling the operator to act.
