The water inventory of both N2 and H2O on Earth (0.2 cubic km of water per inhabitant) is extravagant. Ceres is roughly 50% water (like the generic outer solar system or interstellar object) and we could use all of it but it adds up to about the mass of our oceans, and that is most of the water inventory the asteroid belt.
I’ve looked at various large rotating structures up to the size of the Banks Orbital and concluded you could solve the structural problems if you could make a reliable journal bearing (structural parts don’t need to be spun) that could handle the velocity difference but no matter what you cannot afford tall mountains or deep oceans.
Very tempting to swap nitrogen for helium, but AIUI the main downsides are that the speed of sound is higher so air will leak out from a breach faster, and there may be negative effects on flammability due to helium's high thermal conductivity and low specific heat.
There have been proposals to use scoopers with structurally integrated electrodynamic propulsion ("tether without the tether"),[0] which would use zero propellant.
[0] https://forum.nasaspaceflight.com/index.php?topic=44542.msg1...
Nitrogen needs a high-pressure bottle which is wasted mass, whereas water you can send to orbit in thin plastic and fabric bags like they do on ISS.
Often on ISS they're actually shipping oxygen as water, but still the wasted mass of water being 11% hydrogen plus the wasted mass of the bag is more efficient than a high pressure gas cylinder.