Setting up a trial NFT system

I am in the process of setting up a small NFT (Nutrient Film Technique) system to start a trial commercial basil farm. I am planning on using a header tank system to supply the nutrients to the plants. I’m choosing this system because I feel at this time it provides me with more safeguards. The flow rate of the nutrient solution to the plants from the header tank would be controlled by a manual inline valve. I will have a primary pump to transfer solution from the catchment tank to the header tank. I also plan on using a secondary pump as a backup in the event of primary pump failure. When I change the solution I will swap the pumps so the primary becomes the back up and vice versa. Both pumps would be activated either by their own float or perhaps one float would control both for simplicity, but I need to figure out which would be more practical. Answer by RICK DONNAN

Currently, I’m trying to find a submerged float valve or some other solution to get my filtered tap water to the header tank in the event of total power failure (I don’t have a back-up generator at this time) and I realise that my nutrient solution will become diluted and the water won’t be pH balanced at first (automation is in the future), but at least it will keep the roots wet.

Since I am planning on having about 288 plants and based on your recommendations of 0.5 litre per small plant from your article entitled, ‘What are the fundamentals of setting up an NFT system?’, I found out I will need to quadruple the initial calculations for my reservoir. This is not a problem, but I was wondering, do the header tank and the catchment tank have to be close to equal in volume (minus whatever was in the tubes supplying the channels, the channels themselves and the drainage system to the catchment tank), or if one tank can be greater than the other? I have four 10 gallon (38 litre) tanks. I was planning on connecting three for the catchment tank and one mounted higher for the header tank.

The float system I have in mind would be a ball float with a vertical rod in the header tank. The rod would be made of all thread so I could use some nuts and washers placed on the rod to activate a switch to turn on a pump when the nutrient solution dropped to a certain level in the header tank, much like how a pedestal sump pump works. The pumps would be in the third catchment tank, as far away from the drain inlet from the channels as possible. The first pump might be set to turn on when two gallons had left the header tank, and if the pump failed the backup pump would turn on when the tank had lost two and a half gallons. In theory, with the right sized adjustable pump, I could create a state of equilibrium and have a constant flow of nutrient solution just like a direct pumping setup.

My final question is about aerating nutrient solution. In my research on the internet about NFT (I have the book The ABC of NFT by Dr Allen Cooper), many people say to aerate the nutrient solution, and many say don’t. After all, isn’t that what NFT is all about, the best mix of nutrients and oxygen since the roots that take up the nutrients are in the film and the ones that take up the oxygen are on top of those and out of the film. When the pump pumps nutrient solution into the header tank I can set it up so the nutrient solution falls into the header tank reservoir aerating it a bit, or I could stick some poly tubing all the way to the bottom of the tank so it doesn’t. Which is better? The channels I’m planning are four inches (100mm) wide and only six feet (1.8 metres) long.


Let’s start with your last question first. Those who suggest that you should deliberately not aerate the recirculating nutrient solution are completely wrong.

Please read my answer on respiration in my previous column (September 2015).

Photosynthesis is the process where the plant leaves take up energy from the sun to produce sugars (plus take in carbon dioxide and emit oxygen). Respiration uses this energy to convert those sugars into the huge number of compounds found within plants: that is, proteins, celluloses, starches, vitamins, etc. In this process, oxygen is taken into the cell and carbon dioxide released.

There are two aspects of respiration that are critically important. These are that: (a) respiration is an ongoing 24-hour process, and (b) respiration occurs in all cells in the plant.

Where this is especially important for NFT is that all the cells in the submerged roots must be able to respire. That is, there must be enough dissolved oxygen in the flowing solution to enable those root cells to take it in for respiration. If this doesn’t happen, those roots will die and hence the whole plant dies. Failure to have adequate oxygen in the solution is a disaster for NFT (and for all hydroponic systems). Also note that many commercial NFT systems successfully operate with all the plant roots permanently submerged.

The basic premise of Nutrient ‘Film’ Technique is that by having a thin film there is a large surface area of solution exposed to the air per volume of solution. Hence, as the oxygen dissolved in the solution is used by the plant roots for respiration, it can be replaced by absorption of oxygen from the air at the film surface. To deliberately avoid aerating the solution before feeding it into the channels is counterproductive and certainly living dangerously (sorry, but basically just plain stupid).

Channel design and geometry
Crop failure in NFT systems can usually be put down to poor design. As just explained, failing to have adequate aeration of the recycled solution before feeding it into the tops of the channels is asking for trouble.

The other aspect where things can go wrong is in channel design and geometry. The basic requirement of NFT is that the solution flowing down the channel still contains adequate dissolved oxygen for the plant roots at the bottom end of the channel. Channel characteristics that increase the risk of failure are: width too narrow, slope too flat, length too long.

Trial system design
It is critically important that, when setting up a commercial NFT trial, you use the same channel geometry as you plan for your scaled up version. That is, the same channel profile, length and slope. Change any of these and the trial is meaningless.

I’ve seen cases where trials have performed well using short sloping channels, but on the commercial scale, they were made much longer and flatter. The end result was crop failure when the plant and its roots grew and made the solution deeper in the channel. Consequently, there was insufficient oxygen left in the solution at the lower ends of the channels, so the plants there died, and eventually every plant in the system.

Your channel set-up
In Australia, the normal minimum length of commercial NFT channel is 6m (about 20 ft). I would recommend to normally not go beyond a maximum of 12m channel length. The longer the channels the fewer inlets and outlets are needed, but going too far can lead to the problems mentioned earlier. For a level site, the most common arrangement is to have two 6m channels sloped from each end and flowing into a central collection pipe. This halves the number of collection pipes needed.

If you had originally planned to cut down longer lengths of commercial channel, then don’t do this and keep the original length, say 24 ft. For the same number of plants, this only requires you to install one-quarter the number of channels, This has some other advantages, such as only needing one-quarter the number of inlets and outlets, and one-quarter the total flow into the system.

Read the full answer in October 2015 / Issue 160. Ω