Without going back through the whole message thread to see if I've missed something, I
think I've got to take issue with what you say in the paragraph below. Yes it is true that as
long as liquid remains, the equilibrium pressure of the gas phase will be maintained, and a
difference in isotopic composition between liquid and gas, if one exists, will be maintained.
However, if the gas is isotopically different than the liquid, material balance requires that
removing gas has to change the composition of the total remaining liquid plus gas in the
tank. The contents of the tank will drift over time, although much more slowly than what
happens in a Rayleigh process because of the buffering effect of the liquid.
>Remember, compressed CO2 is a liquid. As long as liquid CO2 remains in the tank, there is an equilibrium between liquid CO2 and gaseous CO2 sitting at an equilibrium pressure of about 55 bar at room temperature. As long as there is liquid neither the cylinder pressure reading nor the isotopic composition of the CO2 taken off will change. The moment the liquid reservoir is exhausted, the pressure will start to drop and the isotopic composition of the CO2 taken will continuously change. Withdrawing amounts from a limited reservoir is a Rayleigh process (in the same way d2H and d18O values of meteoric water change over subsequent precipitation events from the same but perpetually decreasing cloud).