I'm still unconvinced that this is the right solution, but it has a number of advantages:
- no return plumbing - the water is evaporated and thus we don't need to return the water to the fish tank. This in turn implies that the plants can be below the tank level.
- lower power consumption - less moving water means less power
- more reliable - with suitable topology the device can be passive
- prevent cross contamination from diseased plants
- low flow means simpler irrigation for living walls.
On the other hand, it has a number of disadvantages:
- complex feedback control to prevent wasting water - autopot sell the smart valve, which would solve this, but they are $30 a pop, and we'll need hundreds.
- salt buildup on the plant substrates (unconfirmed)
- no oxygenation
- no water purification other than the water removed each step. This would result in a rather poor dilution rate unless we have lots of plants.
A lettuce requires about 5L of water over its 70 day lifetime (70mL/day). Let's say I'm adding 4.5g daily of nitrate equivalent in the form of fish food, insects or nitrogen fixing plants to a 4.5tonne tank. That's 1ppm NO3 increase each day. DC claims that silver perch can live comfortably at 60ppm (Bob claims 5ppm for aquaria) so with no nitrate removal I should have 60 days per refill. That means I need to remove 75L of tank water a day to maintain steady state. This means I would need 75L/(70mL/lettuce) = a thousand lettuces to use up the water.
So a closed loop or hybrid system seems preferable. My current preference is to run a closed loop inside and near the greenhouse, with some open loop stuff further away. Another possibility is to have a separate algae tank and add some nutrient water back to this. Growing pure algae is beneficial for the omega fatty acids (DC claims that only algae can produce them).
An interesting idea is to use the algae as an integrating photoactive window for the greenhouse - algae growth is predicated on the number of photons, and if we are removing some proportion of the total algae continuously we will have a classic exponentially damped FIR filter So the transparency of the window will be the averaging integral of the total light received - so over a period of bright days the window becomes opaque, and in a cloudy week it returns to clear. This would possibly be useful for regulating plant growth or temperature in a greenhouse.
A simple arrangement would be to use welded PE film to make a relatively thin layer that sat on the GH roof.
I have a large fish tank from avril which I shall set up tomorrow and see if I can start an algal culture.