I was working on a compressed air energy storage system for a marine application many moons ago. It was a rubber bladder reinforce and shap-constrained externally by Nylon fabric that would assume the load under tension. I can't recall the operating pressure, but what's vividly burned into my memory was the large fear the fabrication shop had of testing the setup (that if a unit ruptured, it would send large/micro projectiles hurtling through the walls of the facility). The solution was to test under water. We never ended up building the design, but storage under water is plausible safety mitigation I think, though that seems a bit unwieldy.

Yah, but heat pumps have compressors too. I'm not sure what failure modes are the most concerning. The pressure vessels probably have some max number of load/unload cycles in them for instance.

Most refrigerators have compressor (exists mostly mobile schemes, powered by heat, but anyway they have pressure inside).

From conditioners installers, I've hear, they mostly have issues with external connections, but near none with what made on big enterprise.

So window conditioners, or mobile conditioners, or heavy duty industrial conditioners with salt water as working liquid, are extremely reliable.

Same thing with heat pumps - if it is possible, to make all pipes with high pressure on enterprise, there will be no problems (reliability 3 Sigmas usually). Only potentially problematic, what done at fields.