What the ammonia number is actually saying
A farm report lists ammonia as 0.6 mg/L. Before interpreting that figure, ask what chemical quantity was measured. Does it mean total ammonia nitrogen, abbreviated TAN, reported as nitrogen? Does the result represent unionised ammonia, NH3, expressed as the whole molecule? Is it expressed as NH3-N, meaning only the nitrogen mass within the unionised form? Or was it a field-test kit with a different convention? These quantities are related but not interchangeable. The distinction can reverse the meaning of an apparently simple comparison.
In aquaculture, ammonia nitrogen arises from animal excretion, feed metabolism, uneaten food and decomposition. In water, it exists principally as unionised ammonia, NH3, and ammonium ion, NH4+, in a pH- and temperature-dependent equilibrium. The total often combines both forms, while the biologically important un-ionised fraction can be much smaller than the total at low pH and larger at high pH. A reported concentration is therefore inseparable from unit, chemical basis, temperature and pH when used for risk interpretation.
This article translates ammonia reports into defensible measurements and conditional calculations. Examples are hypothetical learning cases, not field observations, and the guide does not prescribe one safe level for all species or a chemical treatment. Genuine animal-welfare decisions require validated methods, species-specific evidence and the responsible facility’s operating procedures.
The acid–base equilibrium from first principles
The chemical relationship may be written NH4+ ⇌ NH3 + H+. When water becomes more alkaline, meaning a higher pH and lower hydrogen-ion activity, equilibrium tends to favour a larger share of NH3. When pH falls, a greater share is ammonium NH4+. Temperature also shifts the equilibrium; under common freshwater conditions, warmer water generally increases the unionised fraction at a fixed pH. Salinity and ionic composition affect activity coefficients and therefore the appropriate equilibrium relationship. A freshwater calculation cannot silently stand in for seawater.
An equilibrium does not mean two equally sized populations. At pH well below the relevant pKa, NH4+ predominates. Near pKa, the two forms are present in comparable amounts. Above pKa, NH3 becomes increasingly important. The word “total” means NH3 plus NH4+ on a defined basis, not that a testing instrument measured two independent numbers that should be added again. TAN is a combined pool from which form-specific estimates may be calculated.
The reason this matters in communication is biological, not semantic tidiness. Unionised NH3 more readily crosses biological membranes and is generally of greater toxicological concern than NH4+ in many aquatic organisms. Toxic response varies with species, age, exposure time, physiological condition, water chemistry and other stressors. The same TAN reported on two mornings may correspond to different unionised NH3 fractions because the pH and temperature have changed.
TAN versus NH3 versus NH3-N versus ammonium
A laboratory may report TAN as mg N/L, often written TAN-N or mg/L as N. That label expresses the mass of nitrogen contained within the sum of the ammonia and ammonium forms. Another laboratory may report the mass of the whole NH3 molecule, mg NH3/L. The molecular mass ratio is approximately 17.031/14.007, or about 1.216, because ammonia molecules include three hydrogen atoms as well as nitrogen. Therefore 1.00 mg NH3-N/L corresponds to about 1.216 mg NH3/L, not 1.00 mg NH3/L.
Ammonium, NH4+, has its own molecular mass and charges; a quantity labelled “ammonium” is not automatically TAN. Analytical methods may first measure total ammonia and then report it under a local convention. Some handheld test kits express their display as NH3/NH4+, yet their accompanying manual defines the numeric output as a nitrogen equivalent. Preserving only the symbol without the stated basis is inadequate.
When transferring a result between languages, write the form and reporting basis explicitly. A useful example reads “total ammonia nitrogen, 0.50 mg N/L; estimated unionised ammonia nitrogen, 0.027 mg N/L, calculated from measured pH and temperature.” Another reads “unionised ammonia concentration, 0.033 mg NH3/L.” Those two latter numbers can describe the same sample on different molecular-mass bases. Treating them as disagreement would manufacture a false laboratory incident.
How to calculate the unionised fraction
For a simplified freshwater calculation at a specified temperature and suitable ionic conditions, the fraction f of total ammonia present as unionised ammonia is f = 1 / [1 + 10^(pKa − pH)]. Here pKa is the temperature-dependent equilibrium parameter chosen for the water chemistry, and pH is measured under a compatible scale. The formula follows from the acid–base equilibrium and is not a universal constant. The output is a fraction between zero and one, which can be multiplied by 100 to express percent of total ammonia.
Take an illustrative freshwater pKa of 9.24 at approximately 25°C, consistent with historical FAO tabulations. At pH 8.00, the denominator is 1 + 10^1.24, approximately 18.38, giving f around 0.0544 or 5.44%. At pH 7.50, it is about 1.79%. At pH 8.50 it is about 15.4%. The movement is nonlinear: a pH difference of half a unit can multiply the unionised share rather than adding a fixed number of percentage points.
When translating a formula, preserve its exponent and brackets. Writing f = 1/1 + 10^(pKa-pH) changes the mathematical order of operations. Writing pH − pKa in the exponent without changing the algebra also changes the answer. Most importantly, the fraction must be multiplied by TAN expressed on a matching nitrogen basis to obtain NH3-N, and only then converted to molecular NH3 mass if desired. The distinction should be shown next to the calculation.
Worked example: 0.50 mg N/L at 25°C and pH 8.00
Suppose a hypothetical freshwater tank records TAN-N = 0.50 mg N/L, pH = 8.00 and temperature = 25°C. Choose the illustrative pKa 9.24 and the simplified equilibrium formula. First compute 10^(9.24 − 8.00) = 10^1.24, approximately 17.38. The fraction NH3 is 1/(1+17.38), or approximately 0.0544. Multiplying 0.50 mg N/L by 0.0544 gives about 0.0272 mg NH3-N/L.
If a report instead requires molecular NH3 concentration, multiply 0.0272 by the molecular mass ratio 17.031/14.007. The result is approximately 0.0331 mg NH3/L. The remaining TAN-N is mostly in the NH4+ form; under the simplified two-form assumption, about 0.4728 mg N/L corresponds to ionised ammonium nitrogen. Do not describe 0.4728 mg N/L as molecular NH4+ mass without another conversion.
A unit-safe sentence is: “At pH 8.00 and 25°C, approximately 5.4% of the measured TAN-N is estimated to be unionised, yielding 0.027 mg NH3-N/L, equivalent to about 0.033 mg NH3/L.” It also needs the assumptions of freshwater chemistry and the chosen equilibrium source. Neither this fraction nor the result is a complete toxicological diagnosis. Biological evaluation needs exposure duration, species, independent water-quality data and appropriate decision standards.
Worked example: the same TAN with a higher pH
Keep the hypothetical TAN-N at 0.50 mg N/L and temperature near 25°C, but increase pH from 8.00 to 8.50. With the same illustrative pKa 9.24, compute f = 1/(1+10^0.74). Since 10^0.74 is about 5.50, f is approximately 0.154, or 15.4%. Estimated NH3-N is therefore 0.50 × 0.154 = 0.077 mg N/L. As the whole NH3 molecule, this is about 0.094 mg NH3/L.
The TAN concentration did not change, yet the estimated unionised ammonia nitrogen increased from about 0.027 to 0.077 mg N/L, roughly 2.8 times. An editor saying the ammonia “almost tripled” would be ambiguous and potentially false: total ammonia did not increase; only its estimated distribution between chemical forms changed. The careful wording identifies the affected fraction and the mathematical assumptions.
This example also explains why a TAN graph without pH or temperature may be insufficient for comparing biological exposure across days. An afternoon rise in pond pH from photosynthesis can change the unionised fraction even if total ammonia is comparatively stable. Sampling schedules should preserve simultaneous pH, temperature and TAN observations rather than pairing measurements collected at unrelated moments. Interpolating a missing pH record can be useful for exploratory modelling, but it should never be represented as a directly measured value.
Temperature changes the result too
At a fixed TAN-N and pH, warmer water commonly favours a larger NH3 fraction in freshwater. For illustration, FAO’s historical table gives a pKa near 9.24 at 25°C and near 9.09 at 30°C. At pH 8.00, the simplified fraction is about 5.4% at 25°C but approximately 7.5% at 30°C. With TAN-N at 0.50 mg N/L, that difference implies about 0.027 mg NH3-N/L versus 0.038 mg NH3-N/L, before considering other chemical or physiological effects.
A report that says “TAN remained constant, so ammonia exposure was unchanged” ignores this temperature dependence. Yet another report that attributes a fish response solely to temperature is also overconfident: warm water changes oxygen availability, metabolism and other processes. Chemical speciation is one piece of a broader water-quality picture. The observation should remain distinct from the inferred mechanism.
The precision of a calculation is constrained by its inputs. If pH is known only to a tenth of a unit and temperature only to a degree, reporting an NH3 fraction to four decimal places implies more certainty than the data support. If the pKa expression is valid only for freshwater within a temperature range, do not extrapolate it to a salty tank or extreme temperature without validation. A good translation includes the domain of the formula, not just the answer.
pH measurement is a moving target
pH in an algae-rich pond often changes over a day as photosynthesis consumes dissolved inorganic carbon in light and respiration releases carbon dioxide in darkness. Some systems have more stable pH because of buffering or active control; other systems experience abrupt operational changes. A sample taken at dawn and a TAN sample collected after midday may not represent a simultaneous chemical state. Combining them as though they were co-located and co-timed can generate a misleading unionised-ammonia estimate.
pH electrodes require suitable calibration or verification, temperature compensation and maintenance. High ionic strength, fouling, old buffers, incorrect calibration sequences and measurement in a sample that has exchanged gas with room air can change the result. A pH of 8.20 recorded from an old bottle is not necessarily the pH of the pond at the instant ammonia was sampled. State whether pH was measured in situ, immediately after collection or much later under laboratory conditions.
If a dashboard gives one pH value for an entire pond complex, ask where it was measured. Different tanks can experience different feeding loads, aeration and carbonate chemistry. Do not silently use one pond’s pH to calculate another’s NH3 fraction. If a proxy must be used for a scenario analysis, label the assumption and calculate a plausible range instead of a point estimate disguised as direct observation.
Sampling, preservation and laboratory methods
Ammonia can change between collection and analysis through biological activity, volatilisation under some conditions, contamination, sample handling and storage. Laboratories use method-specific preservation and holding-time requirements. A sample that was drawn from the surface, carried in warm sunlight and analysed late may not be equivalent to a promptly handled sample collected under a validated procedure. The correct remedy is adherence to the selected method, not a universal preservation recipe from an internet summary.
Colorimetric test kits and laboratory methods may have detection limits, interferences and uncertainty that vary with the water matrix. Salinity, colour, turbidity, disinfectants and reagents can affect particular methods. A result reported as “below detection” is not zero. It means the method did not detect the analyte above its stated detection capability under the conditions assessed. If the laboratory reports a value beneath its quantitation limit, it may be a detected but less reliably quantified result.
A translated laboratory table should retain method identifier, sample identifier, units, detection or quantitation limit, dilution, collection date and time, preservation notes, quality-control flags, and whether the reported number was corrected for a blank. A unit-free “ammonia level” embedded in a progress report may be accessible prose, but it cannot serve as a fully reproducible analytical record.
The nitrogen cycle is more than an ammonia number
Ammonia can be transformed by microbial processes including nitrification, in which microbes oxidise ammonia to nitrite and then nitrate under appropriate conditions. The first step often increases nitrite transiently, and the process consumes oxygen and alkalinity. A decline in TAN may therefore coincide with a rise in nitrite rather than complete removal of nitrogen from the system. It is incorrect to call a falling TAN concentration proof that the nitrogen has disappeared. Nitrogen can also enter biomass, leave with harvested or wasted water, be transformed through other pathways, or be affected by dilution.
Theoretical nitrification stoichiometry offers a useful scale: oxidising one gram of ammonia nitrogen fully to nitrate requires about 4.57 grams of oxygen and consumes about 7.14 grams of alkalinity expressed as CaCO3, under the conventional overall reaction. Real systems include microbial growth, partial conversion, exchange and other reactions; the ideal coefficients are not a direct field oxygen-demand measurement. A translation should identify them as theoretical mass ratios rather than operational predictions.
If a filtration unit claims to “remove ammonia,” investigate whether it converts, assimilates, adsorbs or physically exports nitrogen. Each process has different subsequent consequences. An apparently good TAN result should not distract from nitrite, nitrate accumulation, dissolved oxygen, pH and alkalinity. The system’s material balance connects all these measurements. Good explanatory writing makes those connections without treating one species as a proxy for the whole nitrogen cycle.
Worked example: the nitrification oxygen budget
Suppose a hypothetical system converts an estimated 120 g of ammonia nitrogen completely to nitrate over a period. Multiplying by the theoretical 4.57 g oxygen per gram N gives about 548 g oxygen demand, or 0.548 kg. Multiplying by 7.14 g CaCO3 equivalent per gram N gives about 857 g CaCO3-equivalent alkalinity consumption. These results are idealised reaction-scale calculations, not observed oxygen removal or a recommendation to add 857 g of a particular chemical.
Why is the last distinction important? Different buffer products have different chemistry and purity, biological assimilation may divert nitrogen from nitrification, and water exchange can carry nitrogen or alkalinity into and out of the system. The theoretical oxygen demand should also not be confused with a measured dissolved-oxygen deficit in one tank. Simultaneous aeration may replace part of the consumed oxygen. Converting the result to mg/L would require an identified effective water volume and a model of concurrent transport.
The corrected explanatory sentence is: “Complete nitrification of a hypothetical 120 g N would require approximately 0.55 kg oxygen and consume roughly 0.86 kg CaCO3-equivalent alkalinity, under the ideal stoichiometric assumption.” The sentence names input basis, chemical transformation, output units and limitation. It is a much stronger translation than “120 g ammonia uses half a kilogram of oxygen,” which leaves uncertainty about whether the 120 g means nitrogen mass or whole-molecule ammonia.
Nitrite is not NH3 and nitrate is not TAN
Nitrite (NO2−) and nitrate (NO3−) are separate inorganic nitrogen species with distinct physiological and environmental significance. A report may show NO2-N or NO3-N, which expresses nitrogen mass in each species, or may show NO2− and NO3− as whole-ion mass. Conversions use different molecular mass ratios. For example, converting nitrite nitrogen to nitrite ion mass uses approximately 46/14, whereas converting nitrate nitrogen to nitrate ion mass uses approximately 62/14. Numbers that look equal on the page may describe very different molar amounts and risks.
A field strip that says “nitrite 1.0” without a unit should not be assumed to equal 1.0 mg/L as nitrogen. Read the test instructions and reporting convention. Nitrite stress depends on water chemistry, including chloride, and on species. The presence of nitrate can indicate nitrification but does not prove that all ammonia has been processed or that the system is safe. Nitrate can also accumulate and require its own management.
For a nitrogen budget, convert all forms to the same nitrogen basis before adding them. Do not add 1 mg/L NO3− to 1 mg/L TAN-N and report 2 mg N/L. That adds whole ion mass to nitrogen mass. Keeping the “as N” label throughout the dataset is a simple way to prevent a major interpretation error, particularly when data have passed through several laboratories or software systems.
Freshwater and seawater calculations must be distinguished
Ammonia speciation in saline water involves ionic strength and an appropriate equilibrium model. Historical FAO tabulations provide separate values for freshwater and salinity bands. They are not merely alternative formatting of the same curve. A seawater hatchery cannot assume that a pKa approximation chosen for dilute freshwater is accurate across all salinities and temperatures. The chosen equation or table should describe its conditions and the pH scale used.
Suppose a reader sees a marine tank labelled salinity 32 and pH 8.1. The number “32” might mean practical salinity rather than grams of salt per litre, and the pH may be measured on an operational or total scale depending on laboratory practice. If the calculator assumes freshwater inputs, the resulting fraction has a hidden model error. For precision work, use a validated marine-chemistry procedure rather than forcing the freshwater example to fit.
The communication task is to state what was actually used: the salinity measurement convention, temperature, pH method, reference equation and total ammonia basis. When a report compares marine and freshwater units, explain the different speciation assumptions before ranking NH3 exposures. This is especially important in facilities that culture organisms across life stages in different salinities, because a stable TAN trend can conceal a changing calculated fraction.
The combined problem of oxygen, pH and ammonia
Aquaculture water quality rarely deteriorates one parameter at a time. High feed loading may raise TAN and microbial oxygen demand. Nitrification consumes oxygen and alkalinity; a drop in buffering capacity can allow pH to change. Photosynthesis can raise pH and oxygen by day, while overnight respiration can lower oxygen and alter carbon chemistry. A spreadsheet that considers NH3 only during daylight may miss the oxygen burden at night. Conversely, a chart focusing solely on dawn oxygen can overlook a daytime increase in the unionised ammonia fraction.
These processes can create apparently contradictory observations. For example, an afternoon record may contain high dissolved oxygen and a larger NH3 fraction, while a predawn record may contain low oxygen and a lower NH3 fraction. One is not automatically “the safe period” across all hazards. The appropriate response is to describe two distinct exposure pathways, with their own timing and measurement quality, and to examine the culture animals and operating conditions with trained personnel.
The wording “water quality is good” compresses too many different claims. A more useful statement says which indicators were measured, which were missing, when samples were collected and what local criteria were applied. Reports should include a chronology of feeding, aeration, water exchange, mortalities and sampling because operating events help distinguish chemical transformations from sample artefacts. The primary task is to preserve this causal structure rather than to produce a single green or red status.
Estimating uncertainty rather than hiding it
NH3 calculations are sensitive to pH because the equilibrium relation contains a power of ten. A small error in pH can produce a substantial relative change in a small unionised fraction, even though TAN measurement itself has not changed. Temperature uncertainty alters the chosen pKa. Salinity, sample handling and analytical uncertainty introduce additional variation. A final value with six decimal places is therefore not a guarantee of six-decimal accuracy.
A useful sensitivity exercise holds TAN-N at 0.50 mg N/L and freshwater pKa at 9.24, then compares pH 7.9 and 8.1. At pH 7.9, f is approximately 4.37%, giving about 0.0219 mg NH3-N/L. At pH 8.1, f is approximately 6.76%, giving about 0.0338 mg NH3-N/L. This interval is not a formal confidence interval; it is a scenario range showing what a ±0.1 pH difference means under one model. Reporting it as statistical confidence would misstate its origin.
For a true uncertainty estimate, consider calibration evidence, sample variability, laboratory uncertainty and any covariance between inputs. If pH and temperature were measured at different locations or times, the uncertainty may be structural, not just a small standard error. Preserve this limitation rather than assigning precise uncertainty bounds without data. The purpose of the calculation is to help readers identify what to measure next, not to decorate a rough sample with false numerical authority.
Worked example: a mixed reporting basis
A hypothetical laboratory report lists TAN-N at 0.80 mg N/L, while a manager’s summary displays unionised ammonia as 0.050 mg NH3/L. The figures cannot be compared by simple subtraction because their reporting bases differ. Convert the latter to NH3-N by multiplying 0.050 by 14.007/17.031, obtaining approximately 0.0411 mg NH3-N/L. Relative to TAN-N of 0.80, the implied unionised fraction is about 5.14%, if both measurements refer to the same sample and the speciation assumption is consistent.
A translator who writes “the difference is 0.75 mg/L” would subtract incompatible quantities and imply that the remaining amount is ammonium. A corrected calculation uses compatible nitrogen bases: estimated NH4-N is 0.80 minus 0.0411, approximately 0.759 mg N/L. To express NH4+ molecular mass, another conversion is needed. The numbers are illustrative and their precision depends on how each laboratory result was produced.
This example also shows why the label “total ammonia” must not be silently changed to “free ammonia.” If the manager requests a comparison with a regulatory limit, first verify whether that limit is expressed as NH3, NH3-N or TAN-N, and whether it refers to a specific averaging duration or water chemistry. A unit mismatch can be more consequential than a small analytical measurement error.
A zero, a blank and a non-detect are not the same
Laboratory reports sometimes contain “ND”, a less-than sign, an empty cell or a zero. These entries have different meanings. ND generally indicates non-detection under the method’s definition. A result below a reporting limit may still represent a positive but unquantified concentration. An empty cell may mean no sample was taken, the analysis failed or the result has not been received. A zero may be a rounded or software-imputed value, not proof of chemical absence.
Suppose the report states TAN-N <0.05 mg N/L. A graphing program substitutes zero for every such result. The resulting apparent downward trend may be an artefact of censoring rather than a demonstrated improvement in water quality. For a screening summary, show the original censoring notation and the method’s limit. Statistical handling of censored data requires an explicitly selected method, particularly when many values are non-detects.
The same principle applies to calculated NH3. A unionised concentration cannot be declared zero merely because TAN is below detection; it may only be bounded under assumptions. For example, with a TAN upper bound and a known pH fraction, a corresponding estimated upper bound can be calculated, but the calculation must retain the less-than relationship. Translation that preserves uncertainty is more truthful than translation that forces every data field to contain an ordinary number.
Reading trends without confusing concentration and flux
A decline in TAN-N from 0.8 to 0.4 mg N/L represents a concentration change, not necessarily the total nitrogen removed from a system. If the tank volume was halved through water loss or the system exchanged water, the nitrogen mass balance must account for those flows. In a hypothetical well-mixed 20,000-litre tank, 0.8 mg N/L corresponds to 16 g N in dissolved TAN. At 0.4 mg N/L and the same volume, it corresponds to 8 g N. The difference in stored TAN mass is 8 g, not necessarily the mass biologically nitrified.
If water was exchanged between measurements, imported and exported TAN may account for some of the change. If fish biomass and feeding increased, new TAN may have entered as other nitrogen was processed. Concentration is a stock per litre; transformation rate or export requires a time interval and a process balance. A chart that shows “TAN removal efficiency” based solely on before-and-after concentrations may overstate treatment performance without inlet, outlet and flow information.
For comparisons, use matched volumes, time windows and sampling locations. A daily average of TAN at a tank outlet need not equal the daily mass load into downstream water. Estimating load requires flow multiplied by concentration and integrated over time, with units such as grams N per day. Keeping a clear boundary between a measurement and a process claim makes environmental and husbandry reports materially more reliable.
Worked example: dilution versus biological treatment
Imagine a 10,000-litre tank containing TAN-N at 1.0 mg N/L, or 10 g dissolved TAN nitrogen. The operator removes 2,000 litres of well-mixed water and replaces it with 2,000 litres of clean water containing negligible TAN. Ignoring new excretion and reactions during the operation, the remaining old water contains 8 g TAN nitrogen. After full mixing in 10,000 litres, the concentration is 0.8 mg N/L. That 20% decline is a dilution and export effect, not evidence that bacteria nitrified 2 g N.
A second tank shows the same change from 1.0 to 0.8 mg N/L without any water exchange. The causes could include nitrification, assimilation, adsorption, sampling differences or several processes. In the second tank, external evidence is required to attribute the change. Do not claim that identical concentration trends prove identical treatment mechanisms.
A useful report for the first tank says “TAN-N declined by 0.2 mg N/L following an assumed 20% perfectly mixed water replacement with negligible-TAN water.” The qualification matters because a real exchange may be incomplete, and source water may contain TAN. This worked problem transfers a general idea: tracking the amount of a chemical requires a water-volume and flow account, while identifying the mechanism requires more than a concentration curve.
Independent problem: explain a pH-driven change
Two hypothetically simultaneous freshwater samples have the same TAN-N of 0.40 mg N/L and temperature near 25°C, but one has pH 7.5 and the other pH 8.5. With illustrative pKa 9.24, the unionised fractions are about 1.79% and 15.4%. Estimated NH3-N concentrations are about 0.0072 and 0.0616 mg N/L respectively. The second result is roughly 8.6 times the first, even though TAN-N did not change.
Your task is to repair “ammonia nitrogen increased eightfold.” The correct statement is “the estimated unionised share and its NH3-N concentration are much higher at pH 8.5; total ammonia nitrogen remains the same in this comparison.” Next identify the uncertainty: the freshwater pKa approximation, pH reliability, species response and observation duration. The exercise demonstrates why a small change in wording can turn a valid chemical calculation into a false account of how much nitrogen entered a tank.
Independent problem: the percent-to-fraction mistake
A student calculates 5.44% unionised ammonia and multiplies a TAN-N concentration of 0.50 mg N/L by 5.44, reporting 2.72 mg NH3-N/L. Diagnose the error. A percentage must first be divided by 100: 5.44% = 0.0544, so the estimate is 0.50 × 0.0544 = 0.0272 mg NH3-N/L. The incorrect answer is one hundred times too large and even exceeds the measured total ammonia nitrogen, which is an immediate consistency warning.
A useful numerical audit checks that the unionised fraction lies between zero and one, and that NH3-N cannot exceed TAN-N when both use the same nitrogen basis and the chemical pool is correctly defined. The exercise transfers to any speciation calculation: a percentage, a decimal fraction and a concentration have different dimensions and should never be swapped without explicit conversion.
Independent problem: nitrogen-basis conversion
A translated table displays 0.036 mg NH3/L. Express this as NH3-N: 0.036 × 14.007/17.031 ≈ 0.0296 mg N/L. If the corresponding TAN-N is 0.60 mg N/L from the same sample, the implied unionised fraction is 0.0296/0.60 ≈ 4.93%. This arithmetic does not by itself verify the original calculation because pH, temperature and water chemistry are still needed.
The misleading sentence “4.93% of the water is ammonia” must be rewritten as “approximately 4.93% of the total ammonia nitrogen is estimated to be present in the unionised NH3 form.” Percent composition of one nitrogen pool is not a volumetric fraction of the entire water body. This task shows how an apparently tiny vocabulary error can change the scientific denominator of a statement.
Independent problem: nitrogen cycle interpretation
A hypothetical biofilter inlet has TAN-N 1.0 mg N/L and nitrate nitrogen 2.0 mg N/L. Its outlet has TAN-N 0.2 mg N/L and nitrate nitrogen 2.6 mg N/L, measured at the same flow and under a simplified steady-state assumption. The TAN concentration fell by 0.8 mg N/L while nitrate nitrogen rose by 0.6 mg N/L. The missing 0.2 mg N/L cannot automatically be assigned to bacterial biomass; it could reflect other species, measurement uncertainty or unmeasured processes.
A careful report says the observations are consistent with ammonia oxidation but do not close a nitrogen mass balance. A full balance requires flow, nitrite, particulate and dissolved organic nitrogen where relevant, water exchange, sampling quality and storage effects. The transfer lesson is to use “consistent with” where evidence is partial, and reserve “proved” or “fully removed” for measurements that genuinely establish the claim.
Independent problem: time and salinity mismatch
A marine hatchery’s TAN sample is collected at 15:00, its pH is measured at 07:00, and a freshwater calculator is used despite salinity around 30 on the farm’s reported scale. A resulting value is written as “measured unionised ammonia.” List the problems: the NH3 fraction was estimated, not measured directly; pH and TAN were not contemporaneous; the chemistry model may be inappropriate for the saline water; and the salinity convention and pH scale need to be known.
The defensible replacement is “an illustrative NH3 estimate was calculated using mismatched sample times and an unsuitable freshwater assumption; it is not fit for a definitive exposure conclusion.” The remedy is to obtain paired observations and use a validated marine speciation method with documented inputs. The exercise reinforces that calculation software does not repair incompatible source data.
Independent problem: a missing reporting limit
A report contains TAN entries of 0.03, 0.02 and “ND” mg N/L but gives no method or detection limit. A chart draws the last value at zero and announces a 100% reduction from 0.02. The claim is unsupported. Without a detection limit, ND does not define an upper bound, and without a justified censored-data method, the plotted zero falsely creates a precise reduction.
A corrected summary retains “not detected” with the laboratory’s method identifier and reporting limit once obtained. Even then, distinguish the documented concentration from assumptions about chemical form. If the reporting limit were 0.05 mg N/L, an ND could coexist with a concentration above the previous quantified result, depending on the method’s actual reporting conventions. The central discipline is to prevent a missing or censored value from masquerading as proof of elimination.
Independent problem: mass and concentration
A 15,000-litre tank contains TAN-N 0.70 mg N/L. Under a well-mixed approximation, dissolved TAN nitrogen inventory is 15,000 × 0.70 = 10,500 mg N, or 10.5 g N. A report labels it “10.5 kg ammonia.” The unit error is a factor of one thousand, and the chemical form has changed from nitrogen mass to unspecified ammonia mass.
The correction is “approximately 10.5 g of dissolved total ammonia nitrogen in the 15,000-litre tank at the observation time.” It is an inventory estimate, not a treatment rate. This exercise is useful in environmental reporting because regulators and operators frequently move between concentrations, total masses and daily fluxes. Keep litres, milligrams, grams, hours and chemical bases visible on each line.
Frequently asked questions and decisions that remain open
Is a low TAN always evidence that biofiltration works? No. Dilution, assimilation, sample timing, adsorption and other processes can also lower the result. Is pH 8.5 always unsafe? No universal conclusion follows without TAN, temperature, salinity, species, duration and validated thresholds. Does a higher temperature always mean ammonia production increased? No; it can increase the estimated unionised fraction even when TAN is unchanged. Does “as N” indicate a different chemical? It indicates a mass-reporting convention based on nitrogen content, not a new nitrogen species.
Can nitrate be added to TAN to estimate the total without conversion? Only when quantities share a compatible mass basis, and even then the sum covers only the species included. Does a reading below detection mean NH3 is zero? No. Is an equation sufficient for marine water? Only if it is valid for the ionic conditions and pH convention. Should an operator choose a chemical adjustment from this article? No. Interventions require controlled procedures, competent supervision and evaluation of coupled water-quality effects.
The lasting skill is to keep four questions separate: what was measured, what was calculated, which assumptions make the calculation valid, and what the biological or operational evidence actually supports. When that chain is preserved, an ammonia report can inform better questions rather than exaggerating certainty.
Sources and related reading
All worked numbers are declared hypothetical teaching cases. The approximate freshwater pKa values and unit conventions are described and must not be extrapolated beyond their evidence. These references are not a replacement for validated laboratory methods, instrument documentation or local animal-welfare procedures.
FAO, chemical features of water and unionised-ammonia tables
FAO, improving pond water quality and oxygen context
eduKateSG, Dissolved Oxygen, mg/L and Saturation