Water Quality

Water quality in a recirculating system is the set of physical and chemical parameters that determine whether fish can grow and whether the treatment processes can function. It covers dissolved gases, nitrogen compounds, the carbonate buffer system, temperature, solids and the ionic composition of the water.

In a closed loop there is no dilution. Every parameter is the accumulated result of what the fish produce, what the treatment train removes and what the exchange water replaces. Water quality is therefore the direct output of the system's engineering, and it is the earliest available evidence that something in that engineering has changed.

• Fish performance responds to chronic marginal conditions long before any acute event occurs.
• Treatment processes have their own water quality requirements, so poor water degrades the equipment meant to fix it.
• Trends carry the information. A single reading tells you a value; a series tells you a direction.

• pH sets ammonia toxicity and nitrification rate in opposite directions, so it cannot be optimised for both.
• Temperature raises metabolic demand while reducing gas solubility.
• Solids left in the loop increase ammonia, oxygen demand and turbidity together.
• Water exchange rate controls nitrate and dissolved organics, and dilutes every other accumulating parameter as a side effect.

• Define target ranges for the species and life stage before commissioning, and record the basis for each.
• Provide sampling points at every meaningful location: tank outlet, before and after the biofilter, after degassing, after oxygenation.
• Instrument continuously where the failure is fast — oxygen and temperature — and by scheduled testing where it is slow.
• Provide alkalinity dosing as controlled infrastructure with measurement, not as a manual addition.
• Set the exchange rate deliberately against nitrate and dissolved organic accumulation.
• Design the makeup water treatment for what the source water contains, including chlorine, metals and gas supersaturation.

• Sample the same points at the same times, at a fixed interval that does not vary with workload.
• Record temperature, pH and feed alongside every reading, since several parameters are uninterpretable without them.
• Plot trends rather than reading tables. Direction and rate of change carry more information than any single value.
• Investigate a change before correcting it, because correcting the observed parameter often moves the problem elsewhere.
• Keep the log where the operators are, and make it the same record that engineering reviews.

• Calibrate every probe on a written schedule and record each calibration.
• Verify routine test kits against a second method or a reference laboratory periodically.
• Check reagent expiry and storage conditions; degraded reagents fail toward plausible values.
• Clean sampling lines and sensor surfaces, which foul and then report the biofilm rather than the water.
• Review the trend record as a scheduled task rather than only during an incident.

• Reading parameters as independent numbers rather than as a connected system.
• Correcting the parameter that moved instead of the cause that moved it.
• Sampling at whichever point is convenient, and comparing readings from different points.
• Adjusting pH during an ammonia excursion, increasing the toxic fraction immediately.
• Trusting uncalibrated instruments, which is worse than having no instrument.
• Recording values without the conditions needed to interpret them later.
• Testing more frequently during an incident and less frequently during normal operation, which is exactly backwards for trend detection.

The carbonate system deserves separate study. pH, alkalinity and dissolved carbon dioxide are three views of one equilibrium, and understanding which of them is driving a change is the difference between dosing chemicals continuously and fixing a degasser once.

Beyond the routine parameters, mature systems accumulate dissolved organic matter, metals and micronutrients that are rarely measured and can affect fish performance and biofilter behaviour. Where a system shows persistent unexplained performance differences, extending the analytical panel is often more productive than re-examining the parameters already being logged.

Water quality is the observable output of the whole system. The parameters interact, most problems are chains rather than points, and interpretation depends on recording conditions alongside values. Measure consistently, trend continuously, walk back along the chain before intervening, and keep instruments calibrated — because every decision downstream rests on their readings.