I have seen several dangerous suggestions on Physics and Chemistry Stack Exchange, and normally what happens when this is pointed out, is that the suggestor thanks people for their concern about safety.
This question asks how to test a robot under a simulated depth of 40ft of water. Several answers have mentioned pressure vessels, either mentioning caution from the outset or editing it in when concern was raised. However the poster of this answer has told both myself and another commentor that we need to get a "sense of perspective."
He advocates getting a reject propane tank, half filling it with water and pressurizing it with an air compressor. I have pointed out to him that there would be approximately 240kJ of stored energy in the compressed air in the tank, which would be extremely dangerous if the vessel were to fail: I have even pointed him to a video of an exploding beer bottle (admittedly at higher pressure, but much smaller volume and total energy.)
He correctly points out that the design pressure of a propane tank is much higher than the pressure required, but even so, when considering a reject tank, this is playing Russian roulette. It's not the likelihood of something going wrong that concerns me - it's the potential consequences.
I've pointed out in the comments and in my own answer that the safe way to do this is to completely fill the tank with an incompressible fluid such as water before pressurizing in order to eliminate all air and hence the possibility of stored energy (I have professional experience with testing pressure vessels and that is how we prefer to do it: test failures with water are safe events, provided air is fully eliminated from the vessel.) This is completely compatible with what the OP wanted to do but the answer poster continues to argue.
What is the policy on answers suggesting something dangerous? Should they be downvoted? Flagged?
I'm not sure deletion would be helpful, I think the discussion in the comments on the answer is in a way quite useful.