I am interested in hearing from folks who have experience with lightning arresters (of all types), especially details about how they fail and any thermal signatures that accompany failure.

I downloaded the attached document from the flir site at one time. ("Lightning arresters effect on power line reliability" by Richard Strmiska InfraMation 2003)

It has no copyright, so I assume it's ok to post (anyone please let me know if it's not). It has some pretty detailed info.

When searching for it with google, I saw some other info at flir but I didn't read it:

http://www.google.com/search?hl=en&q=lightning+%28arres...flirthermography.com

We had a failure of a metal oxide lightning arrester a few years back. We hadn't noticed anything on thermography, but when reviewing our trends afterwards we could see a definite increasing trend on the Doble readings (I think it was the watts loss but my memory if foggy on that). Our guess what that the gasket malfunctioned, allowing moisture into the metal oxide unit.

Many years before that we had silicon carbide surge arresters. The reliability on those was poor and we went to metal oxide. I think most plants have done the same.

FlirGoogle_2003_StrmiskaFINAL_LightningArrestersGood.pdf (570 Kb, 42 downloads)