Are arc flash studies necessary for 33 kV indoor SF6 insulated switchgear ? I have a client requesting they be done, but I'm of the opinion that as no arc is possible, calculating incident energy etc is unncessary.

Have you never heard of insulator flashover?

If swicthgear is non-drawout gas insulated swicthgear (GIS) then it is inherently arc resistant and no calculations should be necessary. If switchgear is drawout type then arc flash boundry calculations should be done, even if equipment is arc resistant design since working on switchgear with live bus exposes worker to arc flash. With draw out switchgear and with doors closed, it may be arc resistant . But with doors open, say for maintenance, it no longer is arc resistant and therefore calculations performed.

Sure have heard aout it, especially for outdoor insulators, everything from flashovers due to bird guano accumulation during no/low rain periods to nearby lightning arrester failures causing nearby bushing flashover and failure

Should calculate the level of PPE to verify nominal SF-6 pressure (they tend to leak) prior to any further work tasks on such gear as well....

It depends on the GIS construction. IMO, a robust design like the Siemens GIS would have no chance for an arc flash, since all current carrying conductors are encased in a grounded metal housing similar to isolated phase bus duct. However, all GIS is not insulated the same, so pay attention to the construction. You also need to pay attention to cable connection areas, as these aren't encased. You may also have some auxiliary equipment connected to the GIS, such as air insulated fuses or control power transformers. These are susceptible to arc flash events.

Thanks for the feedback everyone. I'm proceeding with the studies as suggested.

Not sure how you are going to do a study. The problem we've had with >15 kV is that IEEE 1584 can't be used. That leaves you with two possible calculation methods. There is Ralph Lee's theoretical calculation which is already predicting 300% higher than experimental evidence at 15 kV, and ArcPro which is designed for outdoor gear but gives some fudge factors for enclosed gear.

The fact that GIS is in isolated phase bus duct doesn't really matter. So is metalclad AIS gear. The big key to looking at the equipment for what it is going to be used for in my opinion is to look at the numerical reliability of the equipment and the likelihood of failure. This has become the norm in nuclear, aviation, and the chemical process industry for years now. You can do an arc flash equation which determines the magnitude of the hazard but if the likelihood of occurrence is well below your company's risk tolerance threshold, then it's much ado about nothing.

See Brown, "Electrical Power Distribution Reliability" for a starting point. I think Richard is far and away too much of a GIS fanatic, but the point is valid.

You may also want to look at vacuum gear with solidly insulated systems such as available from Elastimold. If your idea of switchgear is long hallways with shiny metal surfaces, it's not going to be pretty. But it's a different approach to the alternative (plug welded and sealed-for-life GIS with vacuum interrupters). The prices are vastly better than GIS and the design sizes are similar. A few of our local utilities are adopting these designs even for surface rather than the intended vault-mounted equipment.