A total closed-loop system is a bit silly to pursue for its own sake. If you can satisfy the other design requirements and achieve it, great, but while you want maximum recycling, you're going to have inputs and outputs at some level.<br /><br />The nutrient input need not be a big deal AFAIK. It seems to me the 'mass multiplier' effect of a barrel of concentrated NPK fertilizer would be very large: A lot of food grown from those nutrients, and nearly all the nutrients would stay in the system.<br /><br />Do y'all agree that the outputs, given incineration (which IMO is a must), could be realistically reduced to just leakage?<br /><br />For aquatic plants, the fertilizer is as simple as bringing (or finding) certain minerals: gypsum, epsom salts, potash, chelated iron, nitrates, maybe a few more. Also a nice concentrated concoction of trace elements. <br /><br />I'm not thinking about harvesting the aquatic plants, but the fish in the same tanks. The plants keep the fish healthy. The efficiency will be lower than growing wheat, but it might be the best - or quickest - way to provide fresh meat.<br /><br />If aquaculture seems to be over-reaching at this stage, then consider an aquatic system to grow the champion nitrogen-fixer: duckweed. Bright lights, basic nutrient input, continuous production; harvest the duckweed, combine it with those symbiotic organisms and you might get a soil-based system to outperform hydroponics on a cost basis.<br /><br />Which is to say that until shown otherwise, hydroponics should be, IMO, considered the baseline space-food production technology. <div class="Discussion_UserSignature"> </div>