In order to appreciate the limitations of traditional techniques of using a voltmeter and earth contact in cathodic protection and corrosion control monitoring, it is necessary to examine the basic axioms upon which established theory is based.
Cathodic protection specialists would have us believe that there is something special about cathodic protection that sets it aside from normal electrical measurement. However the instruments that are used are electrical meters which are only capable of making the same measurements as any other electrical meters.
Whatever system of metering is used, it will ultimately only indicate the voltage between the two terminals on the meter.
Null-balancing, high internal resistance, digital/analogue electronics and all other refinements of present-day technology do not make the meter capable of doing anything more than measuring the voltage between the potentials at the meter end of the conductive test leads.
The voltage on the meter is therefore the Potential Difference (PD) at the test lead contact points.
A voltage is the difference between two potentials, and a potential
is an electric property. All materials are electrically conductive, but
the degree of conductivity or resistance has an effect on the passage of
electrical current. Some materials have such a high resistance as to be
disregarded as conductors in the circuits we shall be discussing, while
others have such a low resistance that they can be regarded as
electrically homogeneous. The materials in between these extremes can be
studied to reveal the behaviour of the voltmeter under different
electrical conditions at the contact points.
Here's an experiment that shows one such use of the voltmeter
Take an insulated trough of damp sand and a 9-v dry cell, with the
terminals connected to either end of the damp sand. Place the voltmenter
conductors at various locations along the sand trough and observe the
readings. It will be found that a variety of readings between 0 and 9v
will be displayed on the meter, dependent on the position of the voltmeter
probes.
Observation
The sand can be divided into zones of electrical potentials and it
is the differences between these zones which results in the voltage
displayed
The concept of potential zones can be considered in two dimensions by
repeating the experiment using an insulating tray, instead of a trough,
full of damp sand, as shown in the drawing. With one voltmeter probe placed in a fixed position, it
is possible to scribe in the sand 'isobars' of constant voltage round each
of the battery contact points. These lines will define the boundaries of
the zones of similar potential. No voltage will be shown on the meter when
both of the probes are placed on one of these lines even though the probes
are in different positions along the line. The lines form 'contours' of
the potential shape within the tray.
It will be noted that the shape of these contours can be altered by
altering the conductivity of the sand by adding salt solution and the shapes
are only predictable when the sand is of uniform conductivity.
It is possible to make an electrical model of the electric
components further to demonstrate the complexities involved in DC
electrical fields of this nature. If two chains of resistors link two
different potentials and the value of each resistor is different, then we
can see the variations in voltage between the two nodes is equal to the
square of the number of resistors. If a grid of resistors is constructed
with random values and the probes of the voltmeter placed at various
locations, it would require a computer to predict the results.
Put a third dimension onto the grid, and the computer would
have to be fed with every resistance value in order to predict the voltage
resulting from the probes placed at surface nodes. Two or more resistances
of unspecified value included in the system would make the result
impossible to predict with any computer. (It is,of course, true that given
the same source, the voltage between any two nodes would be
constant.)The illustration below shows the results of an actual field survey and demonstrates the complexitites of normal field conditions.
We can use the following experiment to test the integrity of imagining an 'electrical picture' caused by varying ground potentials.
Choose a buried cathodicaly-protected pipeline running though areas
of soil of different resistances.
Place a metal coupon in a high-resistance soil and another in a low-resistance soil, both connected to the same pipeline.
Using a high-resistance voltmeter and two probes, plot the potentials surrounding the coupons in the same way as was done in the sand tray.
In the case of the high-resistance soil, the potential gradient will spread much further than that of the coupon in the low-resistance soil.
Observation
The area of influence of the coupon can vary from a few millimeters
in low-resistance soils to many meters in extremely-high resistant
circumstances such as fresh water, sand and gravel.
The recognition of this fact is important in the understanding of such survey techniques as the two half cell survey (which is now known as DCVG).
The above illustration shows the effect of various depths of cover on the
voltages between the meter probes in the circumstances where corrosion on
a pipeline at a coating fault location is causing a potential difference to
exist between a cothodic and an anodic pipe-to-soil interface. It can be
seen that the exact depth of cover and all the soil resistances
between the probe and the pipe/soil interface must be known to calculate
the interface potential.
This is readily demonstrated where an excavation is dug over a coating fault on a cathodically-protected pipeline. Using 2 half-cells across a voltmeter(as probes) it is possible to plot the potential zones as the work progresses towards the coating fault. It is found that these zones vary in shape according to the conductivity of the backfill.
In high-resistant soils, the zones break the surface but in some low-resistance soils they do not. It is for this reason that the 'close- interval potential survey' is not reliable for finding coating defects.
It is now appropriate to consider the half-cell in the measuring
circuit. Similar illustrations to the one below can be found in various
publications, thus acknowledging the fact that it is possible to measure a
different voltages between the same half-cells when they are placed in
different contact positions with the ground or water in which the pipeline is submerged. The author has developed
several successful field survey techniques based on this fact.
The recognition of the variablility of the voltages obtainable between
two half-cells should be sufficient in itself to indicate that neither
half-cell can be regarded as a reference potential which could be used as
a fixed electrical value against which to plot a survey graph of pipe-to
-soil potential. However, the half-cell, used as an earth contact probe, does
overcome the problems of introducing a metal/environment reaction into the
circuit and is therefore a useful tool.
The extent to which half-cell to half-cell voltages can vary can be
dramically demonstrated by plotting the potential gradients around a
groundbed of an impressed-current cathodic protection system.
With one
half-cell placed over the pipeline but at least 100m remote (from the
groundbed ), the other half-cell is stepped over the groundbed.
A typical
plot, like the one below, is shown, where a difference of over 8v is
measured between one 'reference' electrode and the other. T
he polarity of
the connections to the meter does not alter the value of the voltage shown
(provided the meter is digital or centre-null) and the movement of either
half-cell produces a changing reading when it is placed in a zone of
different potential.
This raises the question How can the half-cell be regarded as a reference when the voltage shown between the two half-cells is not consistent when used
in this way?
It is common practice to connect one side of the voltmeter to the
pipeline, or other structure to be surveyed, using a metal conductor and
the other side to the copper in a CuSO4 half-cell. The
half-cell is then put in contact with the ground and a voltage shows on
the meter. It is shown below, that this voltage that the voltage will
increase as the half-cell is moved towards the center of a CP groundbed
area.
The connection to the pipe remains unaltered and the only change in
the circuit is the location of the half-cell. This will increase
the voltage shown to almost the transformer-rectifier voltage when the
half-cell is close to the anode. It is the potential on the half-cell side
of the meter which has increased, for the other side of the meter is
untouched.
If a coating fault exists, then the voltage will decrease as
the half-cell is moved towards the fault, as explained above in relation
to the use of two half-cells. The pipe is the 'reference' against which the
ground potentials are measured in this type of survey.
During a voltage survey between a pipe and a 'roving' half-cell, the
ground potentials can be altered by the presenceof a bi-metallic coupling
buried under the half-contact point.
As the half-cell is moved over the
anodic part of this coupling it is in contact with a zone of higher
potential and the voltage on the meter will increase.
Similarly, when it
moves over the cathodic part of the coupling the voltage will decrease.
It
does not matter that the potential zones are caused by an electromotive
force which is indepentent of the pipeline being monitored.
Using an insulated tray of damp sand and a short length of steel
pipe, illustrated below, it is possible to make a model which will clearly
demonstrate this misleading effect.
A bi-metallic coupling is set up
between a piece of magnesium and a copper rod buried in the sand, and the
pipe-soil potentials are taken in the traditional way using a
high-resistance voltmeter between the pipe and the half-cell.
By moving
the half-cell contact point to different points in the tray it is possible
to obtain readings between 0.005v and 1.3v.
The pipe is not in metallic contact with the magnesium or copper, but must not be completely insulated in order that it should make contact with the damp sand. (This is to allow the measuring circuit to be completed, as it would be through imperfections in the pipe coating in field conditions.)
Using this model it
is seen that a voltage can be measured between two potential fields
even though they are not interfering with one another.
This experiment can be repeated using an actual pipeline in the field and
placing a bi-metallic coupling in the ground.
Pipe-to-soil readings can be altered by as much as 0.8v by placing the half-cell in various positions in the electrical field of the coupling.
In one over-the-line potential survey a potential depression was
excavated at a fault location signal identified originally on a
Pearson-type survey. The cause was found to be a steel wellpoint pipe
which has been abandoned after construction of the pipeline. The wellpoint
pipe had a galvanized fitting which caused the electrical potential ground
distrubance, but no contact with the steel pipe existed. The pipe coating
and metal condition were perfect at this location.
On a cathodically-protected pipeline a coating fault will provide an
area of low-resistance whereby the TR(Transformer-Rectifier) current
can pass onto the metal pipe. This creates a field of varying
potential zones around the coating fault, and the placing of the half-cell
within these zones will affect the voltage shown on the meter. If the
bi-metallic coupling is placed with its anodic interface over the coating
fault, the effect would be to 'mask' the meter from the pipeline's influence, and the fault would go
undetected. This can be demonstrated in the field.
At a cathodic protection test connection, we can create an
artificial coating fault by placing a steel coupon in the ground and
conncecting it to the pipeline. The potential zones around the coupon can
be plotted in the
ground. If the soil resistance at the test site is sufficient, the zones
of influence will extend over 1m from the coupon. We can then draw a
straight line through the area, the same distance from the coupon as the
depth of the cover to the pipe. By moving the half cell along this line we
can acheive a plot readings similar to that which would be obtainable in
conducting an over-the-line type pipe-to-soil potential survey. We have in
fact created a 'section' of such a survey flat on the ground. Using a
bi-metallic coupling we can produce an independent electrical field which
mask the voltage shift caused on the meter by the presence of the coupon.
It can be easily arranged that the current flowing into the coupon is not
distrubed, thus demostrating that traditional voltage readings can be misleading.
In the light of this result, we should re-consider our concept of
'interference' from other CP systems. The field caused by a groundbed or a
coating fault on a 'protected' foreign service need not cause any damage
to our own plant, even if massive voltage changes are shown during
'interference testing', the ground potential swings of several volts need
not result in current flowing onto or from the pipes. In other more
dangerous circumstances, however, current can be flowing from a pipe
without the effects being detectable at the ground surface.
Further alterations to the voltages on the meter can be caused by the
resistance of the layers in the environment between the pipeline metal and
the point of contact of the probe or half-cell, in the diagram below.
These varying resistances alter the shape of the zones of potential and
the position of the half-cell can alter the reading obtained. The feature
had been overcome in recent years by the introduction of digital meters
which draw very little current, but it is still very noticeable in arid
climates and over surfaces such as tarmac and concrete.
It can be seen that the potential on the half-cell side, of the meter can
be shifted by electrical fields in the ground at the point of contact of
the half-cell. Over the past ten years, the author has witnessed the
voltage shifts shown in the following table:
Pipeline Diameter And Location(ins) | Coating Fault Shift(v) | Undefined Shift(v) | Total Known Error(v) |
---|---|---|---|
30 - rural | -0.200 | -0.600 | -0.800 |
24 - rural | -0.100 | -0.400 | -0.500 |
24 - rural | 0 | -0.150 | -0.150 |
24 - gas works | 0 | -0.400 | -0.400 |
During investigations at other sites, scrap batteries, buried tools,
welding rods, farm implements and parts, abandoned 'foreign' services,
abandoned concrete bases, disused tips both officical and unofficial, and
natural unidentified features have all been found to affect ground
potentials by shifts of up to 1.0v.
To complete the half-cell side near the meter, it can be demonstrated
that the use of a vehicle's mobile radio transmitter can affect the ground
potenial by up to .5v at up to 100m distance, and poorly-earthed
AC generators can cause ground potential fluctuations of 0.200v or more.
Other sources of fluctuations on this side of the meter can be identified
giving totals of over 1v in many locations. The pipeline metal potenial
itself may be caused to fluctuate by outside influences causing further
confusion to the interpretation of results. Indeed, there is no way of
telling which side of the meter is fluctuating using the
pipe-meter-half-cell measuring technique.
Connection Of The Meter To The Pipeline
We must now consider the side of the meter which is connected to the
metal of the pipeline or structure. It is often conveniently ignored that
we make a connection to a single location and that this point is in
low-resistance contact with all the interfaces between the environment and
the structure/pipeline under test. If we examine a very small pipeline
with a very good coating, and there is only one fault/contact, then the
potential on this side of the meter is the pipe/soil potential at that
fault. If we have more than one fault in the coating it is unlikely that
the conditions at the metal-to-environment interfaces will be identical.
The potential at each interface will be different and a current will flow
between them. This side of the meter is connected to the metal of all the
faults, and will be at an average potential resulting from the activity of
them all.
Where there is corrosive activity or a difference in metals, we can
regard this side of the meter as being connected to many potentials in
parrallel. It is impossible to identify the value of any one of these
potentials from the single reading on the meter.
An experiment is not necessary to demostrate this fact, as all
field-experienced personnel who have worked with sacrificial anode systems
can verify the readings obtained. A voltage can be obtained between the
pipe-to-soil potential and the half-cell side of the meter. This will be
less than the voltage read between the anode and the half-cell side of the
meter. When the anode is connected to the pipe and the pipe-to-soil
potential is compared to the half-cell side of the meter, the reading will
be an average of the two previous readings. The potentials of the pipe
metal-to-soil interface and the anode-to-soil interface combine to give a
joint potential on this side of the meter.
In a normal field situation, there would be many hundreds of
interfaces of differing potentials combining to produce a joint 'natural'
potential to the metallic unit which is connected to one side of the
meter. It is impossible to identify the value of any one of these without
disconnecting it from the others (as we can do with the sacrificial
anode). An experiment can be set up to demostrate that corrosion can occur
on a steel surface even though the average potential of the metallic body
has been reduced to -1.00v with reference to a Cu/CuSo4
half-cell. This experiment can only be set up using the Alexander Cell,
and will be described below.
In summary, it can now be stated that the potentials on each side of
the meter can vary by ±0.500v, and that most of the causes of these
variations have no effect on the corrosive activity on the surface of the
pipeline or structure being examined. Conditions have been experienced
where the current from a CP system was flowing onto a magnesium anode
(thus protecting it) when the pipe-meter-half-cell voltage was less than
-0.650v. When the anode was disconnected, the reading decreased to -0.575v
and the steel pipeline was in perfect condition despite very poor coating
and an active local environment.
When attempting to interpret readings obtained by the pipe-to-soil
potential method using a half-cell, we must therefore know :
Only by using these steps can a criterion be fixed to ensure that corrosionis stopped by cathodic protection, and it can be seen that most of the steps are commonly ignored and others of them are impossible to apply in the field practice of monitoring corrosion control. We can thus see that using a half-cell is similar to using a balance scales(the voltmeter) to assess the weight on one side by placing a known and an unknown weight on the other. Indeed, we even try to do this when the whole assembly is in motion(fluctuation).