Crack on Surge Arresters 2

[charz]

Almost all the surge arresters in all the feeders of MV switchgear (6kV) got cracks as shown in the picture.

What are the reasons for it apart from dirt, moisture and inadequate capacity?. The said surge arresters were in operation for just more than a year. We have planned to replace it with another surge arrester with the same dimension with increased MCOV? will that help? In general, when the surge arrester starts conducting due to over voltage, why dont the ground fault protection activate?

 

[RRaghunath]

"11kV ...required 18kV to stop LA failures" - I guess this solution to stop LA failures would have initiated cable / equipment failures in 11kV system.
I don't see it appropriate to install LA with MCOV of >12kV considering that the system (i.e. all the equipment installed in 11kV system) is designed with a max system voltage of 12kV.
It is the earth fault detection and removal process that needs to be improved and instead increasing the MCOV of LA could be counter-productive (by shifting the failures elsewhere).
I haven't worked in an ungrounded system but I read why the concept of ungrounded system has gone out of popularity and is replaced by high resistance or medium resistance grounding concepts. It is the difficulty in identifying the faulty feeder and consequent delay in detecting and removal of earth faults that give rise to 'arcing grounds' which can lead to overvoltages much higher than the equipment rating.
I thought the above will be helpful to OP as he is also planning to increase the MCOV of the LAs.

Rompicherla Raghunath

 

[krisys]

In my view it is better to review the system grounding and provide at least high resistance grounding if not low resistance grounding. Now a days in the medium voltage you may not find many installations with ungrounded neutral.

In the low voltage the unrounded systems are still used, for the specific (critical???) applications. But in MV and HV I don't envisage any justification to have an ungrounded system.