Biofiltration

A biofilter is a reactor that provides surface area and conditions for nitrifying bacteria to grow as a biofilm. Those bacteria oxidise total ammonia nitrogen to nitrite, and nitrite to nitrate, converting the most acutely toxic nitrogen waste in the system into the least.

Unlike every other item of equipment in the loop, its capacity is biological. It grows into its duty over weeks, responds slowly to increased load, and can be lost far faster than it can be rebuilt.

Without nitrification, ammonia accumulates in a closed loop until it reaches concentrations the fish cannot tolerate. The only alternatives are water exchange at rates that defeat the purpose of recirculation, or feeding less. The biofilter is what makes closed production possible at all.

• It sets the standing biomass ceiling, jointly with the oxygen system.
• It is the slowest process in the loop to respond to change, so it constrains how quickly stocking or feeding can be increased.
• It consumes alkalinity and produces carbon dioxide, which makes pH control a downstream consequence of its operation.
• It is vulnerable to disinfectants, medications, chlorine in makeup water and sudden temperature or pH change.

Nitrification proceeds in two microbially distinct steps. Ammonia-oxidising organisms convert total ammonia nitrogen to nitrite. A separate group of nitrite-oxidising organisms converts nitrite to nitrate. Both are autotrophic, both require dissolved oxygen, and both consume alkalinity. Because the two steps are carried out by different populations with different growth characteristics, they can be out of balance — which is why nitrite accumulates during startup and after a disturbance.

Biofilter sizing starts from the ammonia the system will produce at peak feeding, and converts it into a required area of active biofilm using an areal removal rate. That rate is the whole difficulty: it is not a property of the media, but of the media operating in a specific system.

• Nitrification is oxygen-consuming. The biofilter's own oxygen demand must be supplied, or it limits itself.
• Removal rate falls as bulk ammonia concentration falls, so a biofilter is least efficient exactly when water quality is best.
• Organic loading favours faster-growing heterotrophic bacteria that physically displace nitrifiers. Mechanical filtration upstream is therefore part of biofilter design.
• Temperature strongly affects nitrification rate, so a system designed at one temperature does not carry the same capacity at another.
• Nominal media surface area is a geometric figure. Effective area depends on colonisation and on whether flow actually reaches it.

• Size for peak feed load at the lowest operating temperature and least favourable pH the system will run at.
• Place the biofilter after solids removal, always.
• Provide alkalinity dosing as designed infrastructure, with control and measurement, rather than as a manual task.
• Ensure oxygen is available within the reactor, whether by aeration in a moving bed or by the dissolved oxygen carried in the incoming flow.
• Design for even flow distribution. Channelling reduces effective area far more than any media specification improves it.
• Provide sampling points before and after the reactor so removal can be measured rather than inferred.
• Allow a route to add, remove and inspect media without dismantling the vessel.

• Increase feeding in steps and allow the population to catch up. The biofilter grows into load; it does not step with it.
• Monitor ammonia and nitrite together. Ammonia controlled with nitrite rising is a specific signal about the second population.
• Monitor alkalinity and pH as biofilter parameters, and dose to hold the buffer rather than reacting to a pH reading.
• Never allow chlorinated makeup water to reach the reactor untreated.
• Treat any medication or disinfection decision as a biofilter decision, and plan for reduced capacity afterwards.
• Clean media conservatively. Aggressive cleaning removes the biofilm that is doing the work.

• Inspect media distribution and movement regularly; static pockets in a moving bed indicate failing aeration or mixing.
• Verify aeration or mixing equipment output rather than assuming it from motor run hours.
• Confirm alkalinity dosing equipment is delivering the dose it is set to deliver.
• Measure ammonia and nitrite across the reactor periodically to quantify removal, so decline is detected as a trend rather than as an excursion.
• Keep a record of every event that could have affected the population — cleaning, dosing, temperature change, power interruption — alongside the water quality log.

• Sizing from a published removal rate treated as a media property rather than a system-specific measurement.
• Stocking before the biofilter is cycled, and treating the resulting excursion as bad luck.
• Increasing feeding in a single step and expecting the biofilter to follow.
• Letting alkalinity fall, then diagnosing an ammonia problem as a biofilter capacity problem.
• Allowing solids into the reactor and then blaming the media for fouling.
• Cleaning media thoroughly, and losing weeks of nitrification in an afternoon.
• Adding a medication to the loop without planning for the biofilter's response.

Denitrification is the natural extension. In an anoxic reactor with an available carbon source, bacteria reduce nitrate to nitrogen gas, allowing much lower water exchange. It introduces a process that must be controlled carefully: insufficient carbon leaves nitrite accumulating, excess carbon leads to sulphide production, and both are consequential in a closed loop.

Alternative nitrogen pathways, including anaerobic ammonium oxidation, are studied for their lower oxygen and carbon requirements. Their application in commercial recirculating aquaculture remains an area of research rather than established practice.

The biofilter is a cultivated population, not a piece of equipment. It removes ammonia in two microbial steps, consumes oxygen and alkalinity, is suppressed by organic loading and disinfectants, and grows into load over weeks. Protect it with mechanical filtration upstream, feed it alkalinity and oxygen, load it gradually, and measure across it — because everything the biofilter does badly appears elsewhere as a water quality problem.