Aquaponics Nitrogen Cycle: Fish, Bacteria and Lettuce

A Blender-built aquaponics loop: fish waste becomes ammonia, nitrite and nitrate, the bed bacteria need weeks to grow, and the lettuce feeds on the result.

Switch to 2D

Presets

Playback

Parameters

Camera

Layers

Quality

Model check

The science

What is modelled

A small aquaponics loop: fish in a tank, a sump and pump, and a raised media bed with lettuce. Fish waste enters the water as ammonia; bacteria living on the bed media turn ammonia into nitrite and then nitrate; the lettuce takes up nitrate; a water change dilutes everything. The model follows the concentrations in the loop day by day, including how long a brand-new bed takes to grow enough bacteria (the “fishless cycle”).

Equations

  • Ammonia produced from the feed: P = feed × Y, with Y = 0.028 g N per g feed (assumed; roughly a 35 % protein feed with about half of its nitrogen excreted as ammonia).
  • Concentrations in a loop of volume V (mg N per litre): dTAN/dt = (P − r1)/V − e TAN, dNO2/dt = (r1 − r2)/V − e NO2, dNO3/dt = (r2 − U)/V − e NO3, with e the daily share of water replaced.
  • Bacterial conversion (Monod kinetics): r1 = A C1 f S1/(K1 + S1) and likewise r2 for nitrite; A is the media surface (200 m² per m³, assumed), C the areal capacity of the bacteria and f = 1.08(T−25) times a pH factor.
  • The bacterial capacity grows only while there is food: dC/dt = μ f C S/(K + S) (1 − C/Cmax) − kd C, which gives the start-up delay.
  • Plant uptake: U = nplants u NO3/(Ku + NO3) with u = 0.025 g N per plant per day; lettuce mass follows from the nitrogen taken up (4 % of dry mass, 5 % dry matter).
  • Toxic fraction of the ammonia: NH3/TAN = 1/(1 + 10pKa − pH) with pKa = 0.09018 + 2729.92/T (Emerson et al. 1975, T in kelvin).

Simplifications, stated honestly

  • Bacterial growth rates, the areal capacities (0.30 and 0.40 g N per m² per day), half-saturation constants, the temperature factor and the pH factor are typical textbook-scale values chosen for illustration, not fitted to a particular system. The simulated start-up takes about three to four weeks, in line with experience of a few weeks, but real timing varies a lot.
  • The fish stock is held constant (no growth or deaths), oxygen, carbon dioxide, alkalinity, pH drift, phosphorus and other nutrients are not modelled, the loop is perfectly mixed and pump flow is assumed sufficient for the bed.
  • The welfare limits shown (ammonia 0.05 and 0.1 mg N/L, nitrite 1 and 3 mg/L) are rough guideline values; tolerance differs between species.

Checks you can see

The five boxes above test the ammonia-fraction formula against tabulated values, nitrogen conservation, the numerical solver against a smaller time step, the steady state against its closed form, and the nitrate build-up rate against the ammonia production rate.

Loading model...
Loading Blender model...
Ammonia, nitrite and nitrate
Toxic un-ionised ammonia
Bacterial capacity of the bed
Toxic share of ammonia against pH
Made with 🟠 Blender · Three.js