RAS Aquaculture

A flow-through farm borrows water from a river or a well, uses it once and returns it. A recirculating system does the same work internally: it removes the solids the animals produce, oxidises their ammonia, strips the carbon dioxide they exhale, replaces the oxygen they consume, and returns the water to them, typically many times per hour. Make-up water is added only to replace what leaves with sludge and backwash and to control the accumulation of nitrate and other dissolved constituents.

That substitution is what makes recirculation an engineering subject in a way that other production approaches are not. In a pond, natural processes provide aeration, nitrification and sedimentation without being designed. In a recirculating system each of those is a named piece of equipment with a capacity, a power draw, a failure mode and a maintenance requirement. The system delivers control precisely because nothing in it is left to happen on its own.

Recirculation is chosen when the culture environment must be independent of the surroundings. That includes producing a species outside its natural climate, producing near a market rather than near suitable water, operating where water abstraction or discharge is constrained, achieving a biosecurity standard that open systems cannot, and delivering a predictable harvest date irrespective of season. Each of these is a reason that no amount of skill in pond management can supply.

It matters equally as a cautionary subject. Recirculating projects fail more often through under-appreciated operating cost, insufficient redundancy and inadequate operator capability than through any deficiency in the treatment concept. The technology is well established; the discipline it requires is where projects come apart.

The treatment loop

Water leaves the culture tank carrying solids, dissolved nitrogen and carbon dioxide, and depleted of oxygen. It passes through a sequence of processes that address these in an order determined by their interactions, and returns. The order is not arbitrary: solids are removed first because they otherwise foul and organically overload the biofilter, degassing follows the biofilter because nitrification itself produces carbon dioxide, and oxygenation is placed last so that the water reaching the animals is at its highest oxygen concentration.

Recirculation rate and exchange rate

These two rates are frequently confused and describe different things. The recirculation rate is how often the system volume passes through treatment, which determines how quickly a metabolite is removed after it is produced. The exchange rate is how much new water enters the system, which determines the steady-state concentration of constituents no process removes. A system can recirculate rapidly and still accumulate nitrate, because rapid recirculation does not remove anything the loop does not treat.

Hydraulic residence and tank hydraulics

Tank hydraulics determine whether the treatment loop actually receives the water it is meant to treat. A circular tank with an appropriate inlet arrangement moves settleable solids to a central drain quickly, which is why the design of the inlet and the outlet does more for water quality than the size of the tank does. Poor hydraulics create dead zones where solids accumulate and decompose inside the culture unit.

Nitrification as the slow process

The autotrophic bacteria that oxidise ammonia grow slowly, are sensitive to temperature and pH, and compete for surface area with faster-growing heterotrophs when organic load is high. The biofilter therefore has a response time measured in days to weeks, while every other process in the loop responds in seconds to minutes. Almost every operational rule about ramping feed follows from this single asymmetry.

Instrumentation and control

Continuous measurement of dissolved oxygen and temperature is the minimum, with water level, flow and pump status typically included. Control ranges from simple interlocks to full supervisory systems, and the important design question is not how much is automated but what happens when the automation stops. An alarm chain that reaches a person capable of acting, and that fails safe when its own power or network is lost, is worth more than additional measurement points.

Waste streams

Recirculation concentrates waste rather than eliminating it. Sludge from solids removal, backwash water and any denitrification residue all have to be managed, and the discharge consent that governs them frequently determines the site's permitted production. Treating waste handling as an afterthought is one of the more common and more expensive design omissions.

Commissioning is where the operating standard of a recirculating facility is set, and it cannot be compressed because the biological part of it proceeds at its own pace.

1. Fill and run the system on water alone, verifying flows, levels, head losses and pump duties against the design before any biology is introduced.
2. Prove every alarm and every standby device by simulating the failure it exists to cover, including a full power interruption.
3. Establish the nitrifying population by dosing an ammonia source and following the ammonia and nitrite response through the full sequence to completion.
4. Confirm that alkalinity and pH can be held under the dosed load, and establish the buffer addition regime before stock arrive.
5. Stock conservatively and increase feed in measured steps, holding each step until the water response has settled.
6. Establish the routine of daily records covering feed, mortality, water chemistry, temperature, alkalinity and equipment status.
7. Set and document the maintenance intervals for screens, media, ultraviolet sleeves, pumps and blowers, and follow them on schedule rather than on symptom.
8. Rehearse the emergency response, including the actions to be taken on total power loss, so that the sequence is known rather than improvised.

The biological consideration that dominates design is the mismatch of time constants. Feed can be increased instantly, oxygen responds within a minute, carbon dioxide within the residence time of the degasser, and nitrification over days to weeks. A system operated as though all four responded together will experience an ammonia and then a nitrite excursion every time feeding is increased sharply, and the animals will pay for it.

The second consideration is energy. Pumping, oxygen transfer, blowers and heating dominate operating cost, and pumping cost is driven by head, which is set by the vertical arrangement of the treatment stages. A layout that minimises the number of times water must be lifted is worth more over the life of a facility than most equipment selections within it.

A third consideration is that the system is a single connected volume. A treatment problem, a chemical error or a pathogen introduction reaches every tank sharing the loop. Splitting a facility into independent loops costs capital and reduces the consequence of any single event, and whether that trade is worth making is a decision that has to be taken at design stage.

Finally, staffing. Recirculating facilities require operators who understand both the biology and the plant, who will follow a maintenance schedule when nothing appears wrong, and who can diagnose under pressure. The availability of such people at a given site is a legitimate feasibility question, not a detail to be resolved after construction.

• Control the feed input, because it is the single lever that sets solids, ammonia and oxygen demand together.
• Ramp any load increase at the pace of the slowest process in the loop, which is the biofilter.
• Remove solids as early and as intact as possible; every particle that disintegrates becomes a harder problem downstream.
• Maintain alkalinity deliberately and continuously, since nitrification consumes it and pH stability depends on it.
• Test standby equipment and alarms on a schedule, because equipment that has never been exercised cannot be assumed to work.
• Keep a full record of flows, pressures and power draw, so that gradual degradation is visible before it becomes a failure.
• Hold critical spares on site, selected by lead time and consequence rather than by cost.

• Size every process against peak feed load, which coincides with maximum biomass immediately before harvest, rather than against an average.
• Design the hydraulic profile to minimise pumping head, since it is a permanent operating cost fixed by layout.
• Commission the biology fully and prove it against a simulated load before stock arrive.
• Provide standby power and independent oxygen supply for the processes whose loss is fatal within minutes.
• Instrument the parameters that move fastest, and route alarms to a person who can reach the site in time to act.
• Plan sludge and effluent handling as part of the design, since the discharge route frequently limits permitted production.
• Write operating and emergency procedures before start-up, and rehearse the emergency ones.
• Recruit and train operators in advance of commissioning, since the plant will be run by whoever is present on the first difficult night.

Advanced practice in recirculation is largely about managing what the standard loop does not remove. Nitrate is the obvious case, addressed either by exchange or by an anoxic denitrification stage that trades water and heat savings for an additional carbon-dosed process requiring its own control. Phosphorus, dissolved organic carbon, coloured compounds and accumulating trace elements are the less obvious cases, and they set the practical floor on exchange rate in genuinely low-exchange designs.

A second theme is the microbial ecology of the system as a whole. A recirculating system holds a large and largely invisible microbial population, in the biofilter, on the pipe walls and in the water column, and the stability of that population is closely connected to the stability of the animals. Operating to maintain a settled community, rather than repeatedly disinfecting the whole loop, is the direction current thinking has taken.

Finally there is the question of what a recirculating system should be optimised for. Higher density raises output per unit of capital and simultaneously reduces the margin for error, shortens the time available to respond to a failure, and raises the consequence of any single event. The right operating point is the one whose worst credible incident the business can absorb, which is a risk decision rather than a biological one, and it deserves to be taken explicitly.