Just wondering how often people run into really old panels, breakers etc. that have you questioning whether they would work and should you use the trip time as part of the arc flash study? If not, do you use a device upstream? It would be easy to say "replace it" but in the real world that is difficult to achieve. Thanks!

My job now involves going through and doing maintenance on exactly that kind of equipment. Generally speaking what happens is that especially electro-mechanical relays drift out of calibration very slowly for the most part but then rapidly drift out of calibration as wear becomes a significant factor. I don't know how else to put this into words. With electronic relaying it basically either works or it doesn't. And due to the nature of this, using equipment that has drifted out of calibration is taking on a lot of risk...it is only going to get worse over time and very rapidly at that. So it is almost better to just model it as if the device isn't even there if you insist on not repairing/replacing it.

Also consider the consequences...a bad trip unit is one thing, but a bad mechanical assembly can be downright dangerous if it opens so slowly that it arcs severely and blows apart, which often happens when the bearings are shot or the grease has hardened. Better to just get the whole thing rebuilt if it's that bad.

You can replace trip units on switchgear fairly easily, even upgrading from old electro-mechanical relays to newer electronic microprocessor ones. And it's even easily possible to buy a carcass or a fully refurbished one to do this. Doing so keeps your downtime down significantly since it becomes just a breaker swap at that point. Doing entire switchgear lineups takes about as much time as pulling breakers, reworking them, and sticking them back in. It is POSSIBLE to go so far as to keep the frames for say an old drawout unit and replace the breakers with new modern equivalents mounted on the frame but the cost and engineering gets to be pretty prohibitive so I don't recommend going down that road.

So...everyone does this all the time. No excuse not to do it.

I agree with Paul - in simple terms all protective devices should be tested! This does not happen all the time. Any study must be based actual conditions but to know the actual conditions would be cost prohibitive but proper maintenance is the cheaper route. I generally make a statement that manufactures recommended maintenance is assumed.

I disagree slightly with the statement that "all protective devices should be tested". That's the ideal situation and as you said...cost prohibitive. Fuses in particular are an interesting case in that there is essentially almost no maintenance beyond basic periodic visual inspection. Circuit breakers though because they are "active" devices are a pretty big issue.

As another example NETA MTS actually recommends testing all wiring insulation routinely. A megger test really only tells us about the contamination (moisture and dirt) of the insulation except when it is severely degraded and with the failure rates of most wiring (which is not exposed to mechanical damage) being out in the neighborhood of over a million times less than any other device, the act of disconnecting and reconnecting the wiring and doing the testing is actually incurring more risk of failure compared to not testing at all.

Fuse elements may not be 'testable', however switches 600V 800A and above, and enclosed MV ones, are usually of some type of bolted or high pressure contact design. These switch mechanisms need routine maintenance similar to that of power breakers.

The simple phrase 'circuit breaker' should not be consider as being all inclusive. For example, 600V maximum Molded case circuit breakers, Insulated case circuit breakers, and Power circuit breakers have similar parts and construction but vastly different maintenance requirements.