Surge and lightning arresters

By Ahmed El-Rasheed

Introduction

Though often seen as minor components in electrical power systems, surge and lightning arresters perform the vital function of protecting assets against overvoltage transients in the supply system. These transients are produced by a wide range of issues including system faults, disconnections, reconnections, resonance, ferroresonance and lightning strikes. In this article, we will take a brief look at the functions and types of arrester, and also at recommended methods for testing them.

What do arresters do?

This article focuses on the use of arresters in power transmission and distribution systems, but similar devices are used in almost every area of electrical engineering. Whatever the application, the key function of an arrester is to prevent damage caused by voltage surges and lightning strikes. Arresters for use in power systems are available from a variety of manufacturers including ABB, Honeywell, Toshiba, Eaton, and Siemens. The characteristics and performance of arresters are covered by many national and international standards, but two of the most frequently referenced are IEEE C62.11 and IEC 60099-4. Many different types of surge and lightning arresters are available as can be seen in Figure 1. Arresters are usually categorised by voltage rating, and the most frequently encountered are:

• Low-voltage surge protection devices: Used in distribution systems operating at less than 3 kV and in consumer applications, these devices protect electrical appliances and lowvoltage distribution transformer windings.
• Medium voltage distribution arresters: These are also called intermediate class distribution arresters and typically have ratings between 3 kV and 30 kV. They are used in power distribution systems to protect distribution transformers, cables and substation equipment.
• High voltage surge arresters: These are also called ‘station-class’ surge arresters and are used to protect transformers operating at 30 kV or higher and associated substation equipment.
• Magnetic blow valve station arresters: Used to protect communication systems, transformers and other equipment with ratings from 30 kV to 700 kV or even higher.

Figure 1: Typical surge arresters

Whatever their application, surge arresters must be able to withstand the large amount of energy released by the current that flows in them when they are subjected to a surge voltage. Lightning strikes in particular can generate very high levels of energy in an arrester, and devices that are designed and rated to provide protection in such cases are usually referred to as lightning arresters. The need for lightning protection is graphically illustrated in Figure 2 which shows a voltage transformer before and after a lightning strike. Lightning arresters, provided that they had been properly selected and installed, would have diverted the energy from the lightning strike safely to earth, but in this case, the poorly protected transformer was destroyed.

Figure 2: Voltage transformer before and after a lightning strike

Arresters are installed in parallel with the equipment they are protecting. At normal operating voltages, the arresters have high resistance and little to no current flows through them. When they are subjected to a surge at a higher than normal voltage, the arresters change to a low resistance state and conduct heavily. As a result, the energy in the surge is diverted away from the equipment under protection and is channelled safely to earth. When the voltage across the arrester returns to normal, it reverts to its high resistance state.

To ensure that they provide the best possible protection and to guard against damage to adjacent equipment during a surge event, arresters must be carefully installed and solidly bonded to ground. Figure 3 shows examples of lightning arresters installed on transformer bushings.

Figure 3: Arresters on transformer bushings 

What are arresters made of?

Most modern arresters are made of metal-oxide varistors (MOVs), which are semiconductor devices that are sensitive to voltage. At its rated operating voltage, an MOV behaves as an insulator, but at higher voltages it behaves as a conductor. Figure 4 shows the voltage/ current curve for a typical MOV at constant temperature. Note that the current scale is logarithmic so when the voltage across the MOV reaches the point at which it starts to conduct, the current increases very rapidly for only a small change in voltage.

Figure 4: V-1 curve for a typical metal-oxide varistor (From ABB Application Guideline - Overvoltage Protection in Medium Voltage Systems) 

Referring to this curve, voltage Uc is the continuous operating voltage at which there is very little current flow in the arrester. Uref is the knee point voltage where the MOV starts to change to the low resistance state. Beyond this point, in Zone B, the resistance of the MOV falls rapidly. Upl is the peak permissible voltage, and the MOV may be damaged if it is subjected to higher voltages.

In practical devices, multiple MOVs are contained in a rigid insulating housing that is provided with heat dissipation vents or fins, as shown in Figure 5.

Testing and verifying arresters

Since a defective arrester will, in many cases, produce no warning symptoms under normal service conditions, regular checking is necessary to ensure that installed devices are still capable of providing the required protection against damage and potentially catastrophic failures. There are three main methods of carrying out these checks: visual inspection, passive infrared inspection and electrical testing.

Visual inspection

Visual inspection is the first and simplest step, but it must be carried out by trained engineers and technicians who carefully examine each arrester looking for cracks which might lead to malfunctioning under surge conditions. The inspectors also look for corrosion, frayed wiring connections, and signs of water intrusion. High-voltage arresters are sometimes fitted with surge counters and, if these are present, the count should be checked. Some manufacturers recommend a maximum number of surges that their product can withstand and, if this number has been reached, replacement should be scheduled.

The final part of the visual inspection is to compare the records of current coming into the facility and the amount of power being used. A trend of increased power consumption in a transmission or distribution facility is often an indication of developing faults and should be investigated further, bearing in mind that an efficiency drop is not always caused by a faulty arrester, but a faulty arrester can cause efficiency drops.

Thermal inspection

Thermal inspection of arresters is carried out using either permanently installed sensors or a thermal camera to detect hot spots resulting from power loss in the arrester. Ideally, the thermal sensors should be installed near power lines, transformers and similar locations where the arresters are used. With a thermal camera, inspectors can detect specific problems such as breakdowns in the wiring or cracks in the cooling vents, as shown in Figure 6. It is important to use the camera to view the arrester from all sides, as some faults affect only one side. It is also useful to look at thermal trends over long periods as these can reveal developing problems before they lead to outages or major faults.

Electrical tests

Electrical tests are carried out by certified inspector whose first task is to identify the main function of the device. The recommended procedure is for the inspector to record this along with the operating voltage, the range of voltage, power factor, frequency, range of frequency, power supply type, load limits, and more. The inspection report should also include details of the equipment maintenance record and the number of times the main circuit breaker has tripped. The tripping data is useful because if a surge arrester is not operating correctly, a common consequence is frequent operation of circuit breakers as overvoltages are not clamped effectively. After this background information is gathered, there are three main types of electrical test that can be carried out: 

• Watts (loss) test using a power factor test set
• A standard test voltage of 2.5 kV or 10 kV is applied and the resulting current through the arrester is measured. The higher test voltage is preferred, provided that it does not exceed the maximum continuous operating voltage for the arrester. In addition to the measured current, the test set will automatically provide a loss value in watts, power factor and capacitance values. Note, however, that power factor is calculated by dividing loss (in watts) by the product of the test voltage and measured current. As the measured current is likely to be very small, a small change in the measured current will lead to a large change in the calculated power factor. For this reason, power factor is often disregarded as a reliable diagnostic indicator if the measured current is less than 250 µA. In such cases, the arrester is assessed on the basis of watts loss alone as this can be accurately measured. However, care must be taken to ensure that surface leakage is not included in the measured current.
• Watts tests are often performed on station­class arresters at the same time as the assets they protect are being power factor tested.
• Higher than normal losses may be the result of dirt or moisture on the inside and outside surface of the arrester housing, cracked or broken housing, salt deposits or contamination in general. Lower than expected test results could be the result of open shunt resistors or defective elements. Note that loss values will differ between manufacturers and between different styles or classes of arrester.
• V-1 response test using hi-pot test set
• The knee on the V-1 curve is the point of interest. To investigate this, the test voltage ramps up from zero to twice the rated voltage of the arrester (sometimes more) and then returns to zero. The current through the arrester is monitored throughout the test, and a plot of current against voltage will yield a V-1 curve that allows the knee point to be determined.
• For high voltage arresters this method is impractical because of the very high test voltages needed.
• Failure of the arrester during a V-I response test may cause a build-up of gases inside the arrester and lead to an explosion. Because of this, appropriate safety measures must always be put in place before testing starts.
• Insulation resistance test using a specialist megohmmeter
• This test is often used on low and
  medium voltage arresters and is the simplest and quickest verification test.
• When arresters fail, they often do so in a short circuit (or low resistance) mode, and this can be detected by a simple megohm test using an insulation resistance test set. The test set must apply at least 60 % of the rated voltage of the arrester, and the resistance measured should be compared with that of similar arresters. A lower than expected resistance value indicates failure of the arrester.
• It is essential that the test set guard terminal is used when carrying out an insulation test, otherwise surface leakage may result in inaccurate results (the use of a guard terminal is discussed in more detail later in this article).
• Third harmonic analysis
• This test is often carried out by the manufacturers of arresters as part
  of their quality control procedures
  but is generally considered too time consuming for work in the field.
• A 10 kV test voltage at power frequency
  (50 or 60 Hz) is normally used and the third harmonic of the resistive current is measured. For accurate results, a current of around 1 to 3 mA is needed.
• The test typically takes between 30 minutes and 1 hour for each arrester, which limits its usefulness on installed equipment.

Electrical tests compared

In the field, a third harmonic analysis test can quantify the condition of the arrester in terms of the number of MOV elements that are still functioning, but it takes somewhat longer to complete than a watts loss test. The V-1 response test is an excellent way to verify the V-1 characteristics and show that the arrester is still functioning, but it can require very high test voltages (440 kV for a 220 kV rated arrester, for example). This might not be practical and there are important safety implications. The DC insulation test is a simple pass/fail test that produces results in just 60 seconds, but it is only suitable for low and medium voltage arresters.

Nevertheless, in applications where there is a large volume of arresters to examine, the DC insulation test saves time and delivers reliable results.

In summary, the simple DC insulation test is usually best for low and medium voltage arresters, while the watts loss test is often the most suitable option for high voltage and station-class arresters.

Use of the guard terminal

To obtain reliable results from DC insulation and watts (loss) tests, it is essential to take into account surface leakage. This is the current that flows through the resistance path provided by contamination (dirt and moisture) on the outer surface of the arrester, which, during testing, will be in a high resistance state – unless it is defective – and will therefore effectively behave as an insulator. Surface leakage is often neglected, but it is very much part of the measurement and it can dramatically affect the results. With a dirty arrester, the surface leakage current during a test can easily be greater than the current flowing through the arrester itself.

By using the guard terminal, the surface leakage current can be automatically measured and removed from the final result. This is important when high values of resistance are expected as they are when testing high voltage components like arresters. Arresters tend to have large surface areas that get exposed to contamination, resulting in high surface leakage.

Figure 7 shows how the parallel current path that flows through the outside surface of an arrester can give a false low resistance reading. In this example, the arrester has an internal resistance of 800 MΩ but there is an external leakage path with a resistance of just 4 MΩ in parallel with this. The measured value is, therefore, 3.98 MΩ, which gives a completely erroneous impression of the condition of the arrester.

Figure 8: Current flow when guard terminal is not used

Figure 9: Current flow when the guard terminal is used

Figures 8 and 9 compare the current flow during testing when the guard terminal is not used against when it is used. It can be seen that the guard terminal effectively diverts the surface leakage current away from the test circuit, thereby ensuring that it does not influence the test result.

Conclusions and recommendations

Surge and lightning arresters play an essential role in electrical power systems. Their inspection and testing are sometimes overlooked in favour of testing more costly assets, but their correct operation is necessary to ensure reliability of supply. Several North American electrical utilities have successfully implemented DC insulation resistance testing programs for arresters. All arresters are tested on delivery from manufacturers, then annually, and additionally after the occurrence of faults. The utilities that have implemented this program are benefitting from improved downtime and power quality metrics – not to mention additional cost savings because arresters that are still fully functional are no longer being condemned. The moral is simple: to save money and enhance reliability of service – inspect and test your arresters regularly!

Bibliography

ABB. (2018). Application Guideline Overvoltage Protection in Medium Voltage Systems

D. Hogue, 2013, Electrical Explosion Donnells Power House Strawberry, CA [Video] https://youtu.be/HICmYZY3prE, Northern California Power Agency (NCPA)

IEEE Standard for Metal-oxide Surge Arresters for AC Power Circuits (>1 kV), IEEE C62.11, 2012

J. Lundquist, L. Stenström, A. Schei, B. Hansen, “New Method for Measurement of the Resistive Leakage Currents of Metal-Oxide Surge Arresters in Service,” IEEE, 1989

J. Woodworth (2011), Arrester Reference Voltage. ArresterWorks. http://www.arresterworks.com/arresterfacts/pdf_files/ArresterFacts_027_ Arrester_Reference_Voltage.pdf

Megger (2006). A Stitch in Time: The complete guide to electrical insulation testing

Megger. (2005). Application guide for power factor testing of surge (lightning) arresters

Megger. (2018). 5 kV, 10 kV and 15 kV Insulation Testing – Use of Guard Terminal

Metal-oxide surge arresters without gaps for AC systems, IEC, 60099-4, 2014