In order to maintain confidence, it is important that the coating will provide adequate protection under the conditions for which it was designed. If it is vital that the coating has to be continuous i.e. free from cracks and pinholes, then test methods are available which can be used to prove the coating.
When to test is determined by the way in which the coating is applied. If the coating is thin and applied in a number of coats, it may be necessary to test between each coat. A fault in the first coat often causes a fault in the subsequent coats. Frequently, the initial fault may be more easily repaired after all of the coats have been applied. However, in other cases, the fault may prove to be more easily repaired after each coat.
The type of test may also effect the stage at which testing is carried out. If a thin paint coating is to be tested with a Wet Sponge Tester, could the application of water with sodium chloride plus the wetting agent, affect the way in which the next coat adheres to the last one?
The time taken doing the test and the time needed to prepare the surface for the next coat also have to be taken into consideration. It may be better to leave the testing until the last coat has been applied. This is but one example, the point of which is to show that the type and manner in which a coating is applied can have a great effect on the final test result.
Types of High Voltage Testers
There are four different types of High Voltage Tester currently in use:
- Spark Tester, High Frequency A.C.
- Holiday Detector, Low Frequency A.C.
- Holiday Detector, Pulsed D.C.
- Holiday Detector, Continuous D.C.
Spark Tester (High Frequency A.C.)
This is one of the oldest types of tester available, and utilises a Tesla coil to produce the high voltage. These units are normally mains powered and voltages in excess of 50,000 volts are easily attained. When in use, testing a surface, a blue brush discharge will be observed around the electrode.
If a fault is located, the discharge will reduce to a single white spark, pinpointing the problem. Since the output voltage of this equipment is pulsed high frequency, and the waveform contains many high voltage spikes, it cannot be measured by a conventional meter. One way to measure the voltage is to use the Sphere-Gap method as outlined in B.S. 358.
Another problem encountered with this type of equipment is the Capacitive Loss Effect. Since the output voltage is high frequency A.C., the size of the electrode will have a marked effect on the output voltage, in some cases reducing it so much that the test becomes ineffective. Also, the material will absorb some of the power, this is because the material is not perfect as far as the dielectric properties are concerned. Owing to this fact, the electrode should always be kept in motion and never be allowed to remain stationary for any length of time. Inactivity of the electrode may cause localised heating and eventual breakdown of the material under test.
Holiday Detector (Low frequency A.C.)
These testers are normally mains powered and use transformers to step the mains voltage up to the required level. They can be rather large in size and capable of delivering excessive amounts of power. Some means of limiting the output current to a safe level is also required. Although the output is low frequency A.C. (50 or 60 Hz) these units still suffer from the same problems as the high frequency A.C. types, but to a lesser extent.
Another problem with the simpler types of tester is that the output is dependant on the mains voltage, and as this changes, so will the output voltage. One main advantage over the high frequency A.C. testers is that an alarm circuit can be incorporated to detect the fault. The circuit can be designed to actuate an audible or visual warning device.
Holiday Detector (Pulsed D.C.)
This was the first of the truly portable testers, being battery powered and light in weight. They work on the same principle as a capacitor discharge ignition system used in many car ignition circuits. The principle of operation is that a capacitor is charged up to a voltage of up to 400 volts, which is then discharged into the primary of a high voltage coil and produces a high voltage pulse of short duration.
There are two methods used to control the output voltage.
- The first is to control the voltage to which the capacitor is charged and in so doing control the output voltage. The problem with this type of control is that the test voltage will be pulled down depending on the load on the output coil, size of the electrode, thickness of the material etc.
- The second way is to run the generator at full power and then use a pre-set spark gap across the output to limit the maximum voltage to the required level.
This is an improved method of controlling the output, for as long as the spark is jumping the gap, the output is controlled regardless of the load. These testers also incorporate an alarm circuit which gives a positive fault indication, so it is not necessary to actually observe the spark.
A serious objection to the use of pulse testers is the difficulty of accurately setting and measuring the peak voltage (0.0002 second duration at 30 to 300 pulses per second). As a result, tests are all too often made using an unknown and excessively high voltage, with resultant damage to a otherwise good coating. Another objection on safety grounds is that as this type of tester can produce high energy pulses, and if the operator accidentally makes contact with the test electrode, he or she will receive a shock, and will keep receiving a shock because the tester cannot tell that the operator is connected across the output. Also, if the pulse frequency is low (30 to 100) the operator may find difficulty in releasing their grip owing to the fact that the pulse frequency is close to the Faradic Frequency of the motive muscles of the human body.
Holiday Detector (Continuous D.C.)
These are also portable battery powered testers, the output of which can be as low as a few hundred volts to over 40,000 volts. The principle of operation is that the battery drives a high frequency oscillator which converts the battery voltage from 12 volts D.C. to a high A.C. voltage. The increased voltage is then fed into a Cockcroft Ladder which steps up the voltage and changes it from A.C. to D.C.
The main advantage with this type of tester is that the output is D.C., and therefore there is constant pressure of voltage applied to the surface, without the problems of power absorption in the material. The only current drawn from the tester is the polarisation current during testing and a small leakage current through the material. Also, a volt meter can be incorporated, continuously indicating the output voltage. This meter will also indicate any loss of voltage due to moisture causing an alternative return path. The output voltage can be controlled by changing the output from the oscillator, so a range of output voltages are easily and accurately obtainable from a single tester. Because the output is so easily controlled, it is possible to incorporate a circuit that will monitor the output voltage and correct for changes in battery voltage and, to a degree, changes in output load current.
The continuous D.C. type of tester also has an alarm circuit to aid in the detection of faults. As the output is continuous D.C., the alarm circuit can be made sensitive enough to enable tests to be carried out on concrete substrates. Another advantage with this type of tester is that if the oscillator is properly designed it will collapse upon location of a fault of less than a pre-set resistance. Therefore, if the operator accidentally makes contact with the electrode, they only receive an initial shock and then the output current is reduced to a safe level.
How to Choose the Test Voltage
The test voltage needs to be high enough to find the fault but not too high as to create one. With reference to B.S. 358 (Measurement of voltage with Sphere-Gaps) it can be seen that 32,000 volts will jump a gap of 1cm between spheres of 5cm diameter. The same voltage will jump a gap of nearly 3cm between needles. This is because the shape of the electrode effects the point at which corona discharge starts i.e. the sharper the points on the electrode, the lower the voltage necessary for corona discharge to start.
Spark over (complete temporary breakdown of the air between the electrodes) will occur when the voltage is increased to cause localised breakdown. Therefore the effect soon spreads throughout the whole of the inter-electrode space and gives the required number of ions to carry the current (this can be tens of amperes). With the sudden increase in current there is a corresponding fall in the voltage across the electrode to a very low level.
For the thicker types of material in the range 1mm to 30 mm the formula used in the NACE Standard RP-02-74 has been found to work well-in most cases.
Test Voltage formula using thousands of an inch:
or Test Voltage formula using microns:
After the test voltage has been worked out it is necessary to check that the voltage is not so high as to damage the material. The Dielectric Strength is the voltage at which the material starts to break down, this is expressed in volts per mm, normally with D.C.
To take an example, a 2mm thick sheet of P.V.C. would require a voltage of 11,180 volts using the above formula. Referring to the manufacturer’s technical data, the Dielectric strength is 8,400 volts per mm. 2 x 8,400 = 16,800, so the test voltage should not damage the material.
If the Dielectric strength was found to be only 5,000 volts per mm, then the test voltage would be too high. In this case, a high voltage test may still be used if tests are carried out to ensure that the test is valid.
- Make a small hole in a test piece
- With the electrode over the hole, increase the voltage until a spark jumps the gap.
- Make a note of that voltage (approximately 5,000 volts)
- Use a voltage halfway between the two (7,500 volts)
- Now make some more holes in the test piece, this time at an angle, and using a 7,500 volt test voltage, make sure that all the faults are detected.
For more information you should refer to Calculating the Test Voltage and Calculating the Dielectric Stength from the Technical Info. section
Types of Wet Sponge Testers
With this type of tester a sponge is saturated with a conductive liquid and is moved over the surface to be tested. The sponge is connected to one pole of a low voltage source, the other pole being connected to the substrate. If the liquid makes contact with the substrate via a fault in the coating, then a current will flow. The current is used to operate an alarm circuit.
There are a number of variables which affect the test results; the first one is the conducting liquid which is normally tap water, to which has been added some wetting agent and salt. The wetting agent reduces the surface tension, allowing the water to penetrate a fault, and the salt increases conductivity. The resistivity of a 3% salt solution is about 100 ohms per cubic cm, whereas normal water is about 1000 ohms per cubic cm. The next problem is to get the liquid into the faults. The liquid will enter a pinhole when the angle of contact is small, this is so that the air has a means of escape and is not trapped. This situation is best achieved by letting the liquid flow over the surface by itself.
The biggest variable of all are the testers themselves, the output voltages of which can range from 2 volts to as much as 1000 volts. Also the sensitivity of the alarm circuits range from 40,000 ohms to 4,000,000 ohms. These two variables have a great effect on how the testers operate. For example, if we take a look at the two extremes we find that, with a voltage of 1000 volts and an alarm sensitivity of 40,000 ohms, a current of 0.025 amps will flow. Now to take the other extreme, with a voltage of 2 volts and a sensitivity of 4,000,000 ohms, a current of only 0.5 micro amps (0.000,000,5 amps) will flow.
When a current is passed through water, electrolysis takes place whereby the water is split into Hydrogen and Oxygen. This forms gas bubbles at the point where the water and the metal meet, the greater the current, the greater the amount of gas produced. With this phenomena, a fault could be indicated on the first pass of the sponge electrode, but due to the bubble formation the resistance of the fault could have increased to a point where the alarm circuit will ignore the fault.
For this reason, the maximum current available from the tester should be limited to 100 micro amps (0.000,1 amps). The next variable is the resistance at which the alarm circuit operates ( the resistance of the liquid in the fault). If the high resistance is chosen so that as many faults as possible can be found, difficulties will be encountered when locating the fault, because the tester may detect the fault at a distance of 1 metre or more. Therefore, it should be noted that a lower value has to be used, (a good compromise is 200,000 ohms). The voltage need only be 5 volts for the tester to work correctly. See our Wet Sponge Tester section for more information.
Problems in Testing Coatings
The main problems encountered when testing the thinner coatings, are the variations that can occur in the coating thickness. To be able to use a high voltage tester on a thin coating, the minimum thickness of the coating must be capable of withstanding the test voltage, and at the maximum thickness of the coating, the test voltage must be high enough to find any fault. Another problem when testing thin coatings is that because the test voltage is so low and can only jump a small gap, it is important that the electrode is in continuous contact with the coating, so that faults are not missed.
The surface on to which the coating is to be applied should be properly prepared. If a weld has been ground down there may be some small sharp points left behind, which although covered by the coating these points will decrease the dielectric strength of the material. This happens because the points reduce the voltage at which corona discharge starts, in addition to reducing the coating thickness. There is also the problem with sharp edges, where the coating is normally reduced in thickness and easily damaged by high voltage.
There is one problem that may occur when two coatings of differing types are applied, one on top of the other. If the coatings are of the same thickness, but the dielectric constants are different, then the test voltage will not be distributed evenly across the total coating thickness. When dielectrics are in series, the combination can be considered as equivalent to two capacitors in series, with the boundary as the common plate. The voltage across each coating is inversely proportional to their capacities. To calculate the across one coating (v1) the formula:
can be used, where T1 and T2 are the coating thicknesses, and D1 and D2 are the dielectric constants of the respective coatings.
If T1 and T2 = 0.5mm each, and D1 = 3, and D2 = 5, and Test voltage = 11,180 volts, then the voltage across coating 1 (T1 D1) would be 4,192 volts and therefore the voltage across coating 2 (T2 D2) would be 6,987 volts.
Because of the unequal voltages, great care is necessary in selecting the coating materials, otherwise it is possible for one of the dielectrics to break down at a lower voltage than expected. In doing so, the full test voltage will be transferred to the other coating, and it too may break down. See our Frequently Asked Questions.
Problems in Testing Linings
Overlapped joints are difficult to test since the size of the overlap can be ten times the thickness of the lining material. The test voltage has to be up near the maximum that the lining material can withstand, and tests should be carried out to see that the test voltage is high enough to find any long diagonal faults across the joint, but not so high that it blows a hole in the lining.
When re-testing a lined vessel that has contained a corrosive liquid, if the lining has a fault, it may not be detected, because a void may have been formed behind the lining so increasing the distance between the electrode and the substrate. See our Frequently Asked Questions.
How to Choose the Correct Tester
The first thing to decide is the type of tester to use i.e. wet sponge or high voltage. We would recommend Wet Sponge Testers for coatings below 350 microns thick, but it must be remembered that they can only find a fault into which water has penetrated. If it is vital that no faults exist in the coating, then the high voltage technique should be used. When using test voltages of 1,000 to 5,000 volts, it is important that the test voltage is stable, and that the unit has an alarm circuit, which is necessary because the spark is so small and can be missed.
With higher voltages the spark can be clearly seen, so a high frequency A.C. tester could be used (if the limitations of electrode size and unstable output voltage can be tolerated). If the material to be tested is 6.5mm thick or greater, then any of the high voltage testers mentioned could be used, but if a large electrode is to be used then only the pulsed and continuous D.C. testers will be suitable.
Safety
With high voltage testers, the operator should be in good health and not suffer from a cardiac condition. It is recommended that the testing should be conducted well clear of personnel not involved in the testing procedure, or in such a position whereby surprise of receiving a shock, could cause a related accident. For example, tests being conducted close to moving or rotating machinery, or in such an unstable position that the operator could fall.
All hand operated test equipment must not be left switched on without the operator being present.
Danger
Do not use high voltage test equipment in any combustible or flammable atmosphere, as the test voltage will cause a spark, and an explosion could result. Therefore the safety officer should be consulted before proceeding with a test.
Conclusion
Porosity / Holiday detectors serve a useful purpose in the quality control of coatings and linings – if they are used correctly and with forethought.
The tests are important and should be carried out by properly trained, competent individuals. All too often the responsibility for them is left to inexperienced local site personnel.
