When putting a patient on CRRT, the choice of buffer these days is largely dependent on whether or not you want to use citrate for anticoagulation. Most of our patients get either citrate or bicarbonate. However, not so long ago, the main buffer was lactate. When bicarbonate was first being used as a buffer for CRRT, it had to be added separately to each bag and, in our institution at least, it came in a glass bottle next to the CVVH fluid.
More recently, the bicarbonate comes in a separate compartment of the same replacement fluid bag and just prior to use, a valve is broken and the bicarbonate-rich fluid is mixed with the rest. I had previously assumed that this was because you do not want to mix bicarbonate with calcium because of the risk of precipitation. However, in our institution, we currently use calcium-free replacement fluid and so there is no risk of precipitation (the calcium is given intravenously to the patient based on a sliding scale).
It turns out that the reason for the separate bicarbonate bags is much more interesting. Most i.v. fluid bags are gas permeable. Therefore, if you leave bicarbonate in the bag for a prolonged period of time, CO2 will leach out of the bags. By a passive process, the bicarbonate in the fluid will then be converted to CO2 which will come out of solution and will in turn leach. You will, in the end be left with very little bicarbonate in the bags. To get around this, the manufacturers of dialysate fluids used put the bicarbonate in a separate glass bottle, This is expensive and cumbersome and is prone to errors if somebody forgets to add it to the solution. Instead, now the bicarbonate-containing dialysate fluids are double-bagged. The inner bag contains the solution and is gas-permeable as before. The outer bag is constructed of a thicker, non-permeable plastic that keeps the CO2 inside. Also, to reduce diffusion further, the air between the two bags has a relatively high CO2 concentration.