Heater failures

[MTUgEEk]

Recently I have a couple of 480V 3ph hot oil immersion heater fail. The configuration is a bunch of 240Vac cartiridge heaters series, paralled together to get to the desired kW. The failure is a group of cartridge heaters in a small space all have both leads burn off. This seems like to me a small arc flash type failure, as i can't explain why the both leads would burn off if it were an overload situation. If one wire opens up the current stops flowing and the overload/overheat of the wire should stop. The problem that I am having with the arc flash scenario is that the heater connection head is purged with N2, which i would think would help negate an arc flash. Inspection is not possible as itis located at a facility in another country.

Additionally These heaters are very old and do not have the "stand off" cold area. Is purge gas the way, one properly disipates heat in the connection head if convection cooling by the surrounding metal is inadequate


Any thoughts on this would be welcome.

Thanks

 

[iceworm]

Possibly this applies.
This heater had maybe 15 elements connected parallel delta, 480V, ~70kw. Connection box had studs for each end of each heater - all crammed into a 10" dia circle. Heater terminals were strapped with small CU busses (maybe 1/16 x 1/4) to make up the parallel combinations. In some places the phase to phase clearances were under 1/2 inch.

The heater connection box showed evidence of phase to phase arcing between the heater studs/busses. Nothing showing heater stud/ground arcing - which isn't suprising, we are 480Y HRG.

The conclusion:
The stud/buss connections were loosening from temp cycling. The heater studs are small - difficult to get a good torque. The small arc plasma that developed from the arc between the stud and CU bus, instigated a phase to phase fault to close clearance opposite phases.

The fix:
Cleaned all studs, CU busses. Chased the threads on the studs. Assembled with external star washers, high temp anti-oxidant grease. Purchased an excellent torque wrench suitable for the small studs.

PM twice a year to check torque - being carefull to not increase torque. We are finding a few (very few) mildly loose connections. Disamble, clean grease, torque. It's been a couple of years - so far no repeats of the arcing failure.

Just some thoughts on what may be a similar failure mode.

ice

Harmless flakes working together can unleash an avalanche of destruction

 

[iceworm]

Inspection is not possible as it is located at a facility in another country.

Trouble shooting by telephone over a few time zones - and maybe even add some language barrier, is tough on a good day.

Good luck

ice

Harmless flakes working together can unleash an avalanche of destruction

 

[Compositepro]

Nitrogen will have no effect on arcing. 480v creates arcs are self-sustaining over large gaps, and having two heaters in series means that you have a ballast resistor to limit the arc current so that the breaker will not trip quickly. The arc can be initiated by any contamination that eventually causes tracking along the surfaces between the two leads, like oil residue. The fact that you have nitrogen purge indicates that there is a problem with contamination in the box.