Does anyone know how to properly calculate the amount of molecular sieves needed for a pressure swing adsorption system? I have searched application notes, books, and of course google, and while I find the basics of how they are constructed I cannot find any information on the quantity required. My theory is that I will need enough surface area to hold the output flow rate worth of nitrogen with a minimum of 2x "safety factor". For example I want to build a 5LPM flow oxygen concentrator. For 5LPM O2 output I would have 23.8LPM of air of which 18.8L would be N2, and 5L would be O2. Assuming I have a 10 second cycle time I would need to hold 3.13L of N2, and with a 2x "safety factor" I would need say enough sieves to hold 6L of N2 before exhausting. I would then double that amount to get a continuous flow switching between 2 beds at approximately 50PSI input pressure. This is however just a guess. Any input is appreciated.