Over the years, I have been asked this many times. The reply is not always straightforward since a number of competing considerations need to be taken into account. Here, I would like to share the logic and rationale used to formulate an answer in each case:
Functionality vs. Age
Deciding on the ‘end of life’ of an arrester in a power system can prove highly subjective because this component can survive for a long time without obvious loss of functionality. Of course, in those cases where a dramatic electrical arcing event occurs, the decision is made for us. Since the purpose of the arrester is to protect other equipment, still being able to meet this expected function should be an integral part of a decision to remove it from service. Another way to state this is that arrester lifetime should be considered more from a functional than an age perspective. Most power utilities expect their substation equipment to have a service life of between 20 and 50 years. Assessing the continued functionality of an arrester in the absence of an end of life event can be done in service with routine thermal imaging as well as with temporary or permanent watt meters. If off line, arresters can also be checked with high-pot or Doble testers. Therefore, before a removal decision is made, the arrester’s ongoing functionality should be assessed, assuming all other criteria are met.
Role of Sealing
Provided that the varistor elements are not overstressed, from a technical perspective long service life is achievable. However, it is the condition of the enclosure and seals that will most likely become the determining factor. For all surge arrester designs (i.e. porcelain as well as polymeric housed), once the sealing system has deteriorated it can no longer protect the active elements from external climate. Moisture ingress will then progressively weaken the dielectric strength of the insulation within the assembly.
The reliability of seals is subject to many factors, not the least of which is the quality of manufacturing. Even if the materials used have a long functional life, their reliability is ultimately subject to the influence of production defects. Estimating a likely life span for any sealing system can prove difficult for other reasons as well. For example, operating environment will play a decisive role in whether or not the arrester will achieve its expected service life. Some polymeric-housed arresters have now been in service for about 30 years and, even with them, moisture ingress and sealing problems remain an issue. Keeping the active elements free from moisture is the key to longevity.
When the first polymeric distribution arrester was introduced in 1987, I’m sure it surprised many in this industry at how fast this design became the standard. Since then, it has become evident that even economics cannot drive change in the power sector as fast as does safety. For utilities that had been dealing with exploding porcelain arresters for decades, it was an easy decision to replace them with polymeric units in order to avoid all risk of injury. Even today, the potential for an exploding end-of-life event should be seriously considered whenever raising the question of when an arrester should be removed from service.
In the case of substation arresters, the conversion to polymeric housings has been far slower since the safety risk to the public is not perceived to be as high. Moreover, since these arresters are generally equipped with venting ports for pressure relief, there is less risk of explosion. Still, this risk for a porcelain-housed arrester increases if it is the practice of a utility to either automatically or manually restore power at a substation without first examining its condition. Those utilities that follow this practice should ideally replace porcelain arresters with polymeric-housed units that have less explosive end-of-life events, even under severe conditions. To add to the risk, there are times when available system fault current may have been increased over time with no consideration given to the capability of existing arresters. If they were installed e.g. 30 years ago, it is quite probable that they do not have the short circuit withstand needed. Indeed, the risk of fragmenting porcelain should always be seriously considered whenever deciding when to replace an arrester.
Margin of Protection
Many times the removal question is answered by the fact that the arrester comes from the bygone era of silicon carbide technology. Arresters of this vintage simply cannot protect equipment as well as newer MOV technology. The chart below demonstrate how protective characteristics have changed over the past 50 years. If old arresters are protecting even older and probably lower BIL equipment, then the margin of protection calculation should be critical.
Once a utility has made a decision to remove old arresters, they are still faced with the task of identifying them. In this regard, I recommend you visit www.NemaArresters.com for more information on this topic. Another source is “An assessment of the reliability of in-service gapped silicon-carbide distribution surge arresters“ by Darveniza, Mercer, and Watson, and available at www.IeeeXplore.com.