This was sketched for a discussion on orbital invasions and occupation, but I think it can apply pretty generally to managing equipment failures, epidemics/toxins and general robustness in habitat design. Mostly this was just meant to serve as a counterpoint against the hard-SFnal notion that space station habitats are places where "the station owners can just cut off the air and space anyone they don't like" garbage, because space machismo or whatever