Aquaculture CO2 Stripping: Packed Columns and Mass Balances

Why dissolved carbon dioxide needs a separate report

A fish-rearing tank can have enough dissolved oxygen yet still accumulate carbon dioxide from fish and microbial respiration. This is why the words aeration, oxygenation and carbon-dioxide stripping should not be used as interchangeable descriptions of one machine. In a recirculating aquaculture system, a packed-column degasser can bring water into contact with air so that CO₂ transfers out of the water. But its effectiveness depends on the gas composition, water chemistry, packing, flow, temperature and measurement method. A credible interpretation starts by identifying which carbon-dioxide quantity the instrument actually reports.

This guide explains how to read a CO₂ stripping report through first principles, engineering dimensions, transparent examples, troubleshooting and independent practice. All numerical scenarios are constructed learning exercises, not real installation measurements or safe operating instructions. Actual animal-health decisions require species-specific information, qualified design, validated sampling and attention to oxygen, alkalinity, pH, salinity and total dissolved gas pressure. Good translation keeps equipment performance distinct from animal exposure and does not turn a test result into a universal promise.

The most common source of confusion is a percentage called “carbon-dioxide removal” without a description of the measurement. Was the change in free dissolved CO₂ or total inorganic carbon? Was the outlet sample tested immediately or after carbonate chemistry had equilibrated? Was water flow steady, and was air flow measured under actual pressure conditions? Each answer changes what a number means. These questions are not academic complications; they determine whether a technician can reproduce the calculation and whether another reader can trust it.

1. Trace the source of CO₂ before treating it

Fish breathe, microorganisms respire, and both processes can introduce carbon dioxide into recirculating water. Biofiltration may consume oxygen and produce CO₂ even as it removes other water-quality problems. A degasser cannot prevent all further carbon dioxide generation downstream; it removes a portion at a defined location and time. If a report compares gas levels in a tank with the outlet of a degasser, these are not interchangeable water samples. They represent different stages of the treatment loop and may be separated by additional respiration, mixing or oxygenation.

Draw the loop before reading the figures: culture tanks, solids separation, biofiltration, degasser, other gas treatment, and return to the animals. Mark the sampling positions and direction of water flow. The word inlet must specify whether it means inlet to the entire farm, to a mechanical filter or to the carbon-dioxide column. A technician who labels the ports can often identify an apparent disagreement between two reports without changing any number. Measurement location is part of the evidence, not decorative detail.

2. The mechanism is gas exchange

When water containing excess dissolved CO₂ contacts a gas stream with relatively low CO₂ partial pressure, transfer can occur from water to gas. A packed column exposes water as films, droplets and streams over structured surfaces while air moves through the available voids. Greater effective interface and good mixing can improve transfer under suitable conditions. Yet the process slows as the driving force decreases; a column cannot be assigned a universal removal percentage independent of incoming concentration and gas composition. A useful model therefore relates actual gas and water conditions to a defined mass-transfer expression.

As air contacts water, the air can become enriched with CO₂. This means the gas at the outlet is not necessarily equivalent to the fresh air at the inlet. Low ventilation, bypass, fouling or recirculated exhaust can change the removal performance even if the water pump stays at the same setting. A report that assumes constant fresh-air composition throughout the bed needs to explain why that simplification is justified. Translating this correctly is a way of protecting the causal mechanism rather than merely replacing technical words with familiar ones.

3. Carbonate chemistry changes the apparent answer

Dissolved CO₂ is part of a carbonate system that also includes carbonic acid, bicarbonate and carbonate. The distribution among these forms depends on pH, temperature, salinity and other properties. Removing molecular CO₂ can shift the equilibrium and may lead to further CO₂ formation from related species. Total inorganic carbon and free dissolved carbon dioxide therefore describe different quantities. A report that says free CO₂ decreased by half does not prove total inorganic carbon fell by the same fraction. Accurate translation preserves the analyte and definition in every table.

USGS researchers Colt, Watten and Pfeiffer explain that apparent carbon-dioxide removal measured after chemical equilibration may differ from true transfer during stripping, particularly in high-alkalinity or marine waters. That does not make either quantity inherently wrong; it means each represents a different measurement or calculation. If a sample was allowed to stand before analysis, record the holding conditions and whether it exchanged gas with the surrounding air. The difference between an immediate operational reading and an equilibrium-based result must be carried into conclusions rather than buried in methods.

4. Free CO₂ is not alkalinity

Alkalinity describes acid-neutralizing capacity on a specified equivalent basis. Free carbon dioxide is a concentration of a particular dissolved species or conventionally grouped species. A sample with 100 mg/L alkalinity as CaCO₃ and 10 mg/L free CO₂ does not contain two directly additive quantities on one chemical scale. The phrase “as CaCO₃” signals a reporting convention rather than an identification of the only molecule present. Do not erase it in translation; it can be essential to the carbonate calculation.

If pH rises across a degasser while alkalinity changes little, that can be consistent with CO₂ removal, but it does not yield an exact gas mass from pH alone. A model may calculate free CO₂ from pH and alkalinity, yet it needs a known temperature, appropriate constants and salinity conditions. An apparently precise decimal is only as reliable as those inputs. When a comparison uses two laboratories or years, first check whether the same carbonate method, temperature correction and sampling protocol were applied.

5. Read the column’s physical description

Packed columns commonly have a water distributor, packing medium, air inlets or outlets, a support grid and a route for drained water. An engineer needs the packing depth, plan area, void structure and ventilation arrangement, not only the total height of the tower enclosure. Water can channel along preferred paths; air can bypass wetted surfaces; deposits can alter the void space. The words “one-metre tower” and “one-metre packed bed” are not synonymous. The original equipment drawing should govern how these dimensions are reported.

A full-scale research study indexed by FAO AGRIS examined different structured packings in coldwater recirculating systems and noted effects of water loading and potential air restriction. Its results belong to the studied configurations, not every commercial tower. A meaningful comparison describes the experimental devices, their operating ranges and what was measured. Copying only an impressive removal percentage from a vendor brochure or article would discard much of the information needed to interpret that percentage.

6. Water loading is a flow divided by area

For a circular degasser with internal diameter 1.0 m, the cross-sectional plan area is π×(0.5 m)², approximately 0.785 m². If 40 m³/h of water enters, the superficial hydraulic loading is 40/0.785 = about 51.0 m³/(m²·h). The units simplify to metres per hour, but that does not mean each droplet falls through the packing at exactly 51 m/h. It is a defined engineering ratio between total volumetric throughput and tower area. The actual path and local contact can be very different.

If diameter falls to 0.8 m while flow remains 40 m³/h, area becomes about 0.503 m² and loading rises to about 79.6 m³/(m²·h). A translator should therefore preserve both flow and area rather than saying the two towers are identical because they process the same volume hourly. Hydraulic loading affects whether distribution and ventilation behave properly. However, neither loading figure alone establishes a safe CO₂ concentration or a specific removal efficiency.

7. Gas-to-liquid ratio needs its volume basis

The gas-to-liquid ratio often appears as G/L. Suppose a blower moves 600 m³/h of air through a system while water flow is 100 m³/h. The nominal volumetric G/L is 600/100 = 6, commonly written 6:1. That calculation assumes the flows refer to corresponding intervals and stated volume conditions. If air flow is in standard cubic metres per hour but water volume is measured at operating temperature, readers may need temperature and pressure conventions before comparing ratios across studies.

More air does not always produce a proportional improvement. As the gas phase becomes less enriched and the process approaches other transfer limits, returns can diminish. Increased ventilation can also cause heat loss or demand more fan energy. A FAO AGRIS-indexed experimental study examined carbon-dioxide removal and heat loss as G/L changed in a forced-ventilated trickling filter. Its performance curve is useful evidence in context, not a universal air ratio to copy into any farm’s controls.

8. Distinguish nameplate from delivered airflow

A blower may be advertised at a flow under a defined static-pressure condition, while the installed system has ducts, dampers, filters and packing resistance. Fouling and partial flooding can increase pressure losses and lower actual airflow. A stated “600 cubic metres per hour blower” may therefore not mean 600 cubic metres of air traversed the wetted bed. A robust report quotes measured delivered airflow or a manufacturer fan curve evaluated at the actual operating point. The distinction can explain an unexpected deterioration in gas-removal performance.

Leakage can also create a false sense of ventilation. A fan might move the advertised amount of air, but some could bypass the packing through unintended openings. Record where the flow sensor sits relative to those possible paths. A translated inspection report should distinguish command state, motor running, measured airflow and verified treatment effectiveness. These are successively stronger claims. No status light alone demonstrates that air contacted the moving water as intended.

9. Calculate concentration reduction carefully

Suppose paired inlet and outlet free-CO₂ results are 18 mg/L and 6 mg/L. The difference is 12 mg/L, and percent concentration decrease is 100×12/18 = 66.7 percent. That describes the measured or derived concentration at the named ports under the test conditions. It does not mean total inorganic carbon fell by 66.7 percent, nor does it mean every future operating hour will have the same result. Percentages should always name the starting concentration as their denominator and the analyte basis used.

If inlet was sampled at 10:00 and outlet at 10:00 while tower residence and recirculation dynamics are non-negligible, the samples may not represent the same water parcel. A well-designed experiment may sample after steady operation or adjust for transit and mixing. If water chemistry or instrumentation changed between samples, the apparent efficiency can be biased. Preserve method qualifications rather than turning a single pair into a permanent performance certificate.

10. Convert a concentration change into a mass rate

At a constant water flow of 50,000 L/h, a paired free-CO₂ difference of 12 mg/L corresponds to 600,000 mg/h, which is 0.60 kg/h of apparent concentration-based mass-rate change. We multiply concentration difference by the relevant treated-water flow. The litre units cancel, leaving milligrams per hour, then kilograms per hour after unit conversion. This is not automatically the complete true carbon dioxide transfer if carbonate interconversion and measurement timing affect the before-and-after value.

A claim of “6 kilograms removed” requires an integration interval. If 0.60 kg/h persisted for exactly ten hours, 6 kg would be a simple estimate, but a one-minute test cannot establish that the condition persisted all day. Transient feeding, animal respiration, temperature and fan performance may shift the rate. A time series and appropriate integration method would provide stronger evidence. State when the value is a snapshot, a representative average or a modelled rate.

11. Carbon mass is not molecular CO₂ mass

One mole of CO₂ weighs about 44.01 grams; the carbon within that molecule weighs about 12.01 grams. Thus 44.01 mg of molecular CO₂ corresponds to 12.01 mg of elemental carbon. A spreadsheet field labelled “dissolved inorganic carbon as C” cannot simply be relabelled “CO₂ mg/L”. If a constructed value is 10 mg/L as C, multiplying by 44.01/12.01 yields approximately 36.6 mg/L as CO₂-equivalent mass, assuming that conversion is the intended basis. This exercise concerns mass units, not carbonate speciation at a particular pH.

The same conversion appears in greenhouse-gas inventories and laboratory datasets, making it tempting to reuse an old formula. Yet aquaculture degassing calculations require the correct source species and measurement convention. Ask whether the reported value is free CO₂, CO₂–C, DIC as C, partial pressure or an equilibrium estimate. Preserving the heading prevents double conversion and avoids a substantial arithmetic error that may otherwise seem credible because every displayed number has two decimal places.

12. Separate oxygenation and stripping

Gas-exchange equipment can have overlapping functions but different design targets. Aeration may increase dissolved oxygen when water is below its air-equilibrium oxygen level while also helping remove carbon dioxide. A pure-oxygen transfer device can supply oxygen efficiently without necessarily providing strong CO₂ stripping. A degasser may prioritize release of carbon dioxide or excess nitrogen. Translating all three as “oxygen machines” hides the endpoint being measured. When evaluating equipment, name the desired gas and the direction of transfer for each stage.

USGS research describes modifications to oxygen-absorption equipment that enable concurrent CO₂ removal. The need for special design highlights why an oxygen delivery specification is not enough to prove carbon-dioxide performance. An accurate report lists oxygen transfer separately from carbon-dioxide removal, records the ventilation and pressure conditions, and checks final water quality where animals receive it. A strong explanation asks not “Does this machine add gas?” but “Which dissolved gas moved, where did it go, and how was that change verified?”

13. A pH rise is evidence of chemistry, not a fixed CO₂ percentage

When CO₂ is released from water, pH may rise as carbonate equilibria respond. But the magnitude depends on alkalinity, temperature, ionic conditions and the original dissolved carbon species. A shift from pH 7.1 to 7.4 is a change of 0.3 pH units, not a decline of 0.3 mg/L CO₂ and certainly not automatically a 30 percent decrease. The pH scale is logarithmic in hydrogen-ion activity. Any carbonate-system interpretation requires the model inputs and their uncertainties, not merely the difference between two pH numbers.

If two reports compare a similar pH rise in different waters, readers should ask whether alkalinity and salinity differed. More buffering may alter the relation between a gas change and a measured pH response. The strongest phrasing is that the observed pH shift is consistent with the described carbonate process under documented conditions; a separate validated calculation or CO₂ measurement is needed to quantify removal. This sentence is longer than “pH proves stripping works”, but it is also much more useful.

14. True versus apparent removal and sampling delay

A fast transfer process may lower the free dissolved CO₂ in the column even as chemical reactions continue restoring equilibrium among dissolved carbon species. A water sample analyzed after equilibration can therefore show a different before-and-after free-CO₂ difference from a rapid, appropriately measured process value. The USGS methodological literature explicitly identifies such apparent versus true removal issues. The discrepancy can matter particularly for higher alkalinity or marine conditions. Treat it as a consequence of the chosen physical and analytical definitions, not an instrument mystery.

Record whether samples were obtained from sealed sampling ports, whether they exchanged gas with ambient air, the interval before analysis, temperature, and whether measurements were direct or calculated. A sample left open can gain or lose CO₂ during handling, so its measured condition may not correspond to the water in a closed pipe. If a translated report suppresses the sampling delay, its result becomes less reproducible. A cautious analyst documents both the observed response and the chemistry that may alter its interpretation.

15. Gas-side balances need gas composition and a molar basis

Carbon dioxide moved into the air can be estimated by the difference between gas-side CO₂ mass flows entering and leaving, but that requires measured gas flow and concentration at defined temperature, pressure and moisture conditions. A reading in parts per million by volume is a mole fraction, not automatically milligrams per litre of water. Wet and dry gas measurements may use different denominators. A translator must not multiply a ppm number by an unrelated air volume and announce a mass rate without the appropriate gas-density or molar-flow conversion.

As a constructed arithmetic exercise, suppose a validated gas-flow calculation yields 8 mol/h additional CO₂ at the outlet relative to the inlet. Multiply by 44.01 g/mol to obtain about 352 g/h. That is gas-side CO₂ mass transfer on the defined basis. A water-side estimate may differ because of sensor uncertainty, sampling timing, carbonate reaction, bypass or other streams. Rather than selecting whichever figure looks more flattering, report both methods and investigate a discrepancy.

16. A complete calculation across a packed column

Take a hypothetical square column of 1.5 m² plan area. Water passes at 90 m³/h and air at 540 m³/h, on compatible volume bases. Hydraulic loading is 90/1.5 = 60 m³/(m²·h), while G/L is 540/90 = 6. Paired measured free CO₂ falls from 16 to 7 mg/L, a reduction of 9 mg/L or 100×9/16 = 56.25 percent. Multiply 9 mg/L by 90,000 L/h to obtain 810,000 mg/h, or 0.81 kg/h apparent water-side CO₂ change.

Each output describes a different quantity. The first measures water throughput per geometric area, the second compares gas and water volume rates, the third is a relative concentration change, and the fourth is a mass-rate change inferred under steady paired conditions. None is a validated safety limit for fish, a complete carbon inventory, a measure of blower efficiency or a guarantee of all-day performance. A report should retain the calculated values in separate columns with labelled denominators rather than compressing them into one generic ‘tower performance’ number.

17. Countercurrent, cocurrent and air path

In countercurrent designs the bulk air and water flows oppose one another, allowing the composition of contacting gas to vary along the bed. Other layouts can use different flow directions. A drawing needs arrows for water inlet and outlet and air inlet and outlet, with sample points clearly distinguished. The word counterflow should not be used when the actual diagram shows only a fan direction but not the gas path through wet packing. Air leakage or recirculation can complicate an idealized schematic.

When translating equipment diagrams from another language, preserve original connection labels and add an English glossary rather than altering the drawing to match assumptions. If two lines are visually ambiguous, annotate the uncertainty. A mistaken air-inlet label can reverse the interpretation of gas-phase carbon-dioxide samples and lead to an incorrect conclusion about net transfer. Good technical language makes the physical system reconstructible by another reader.

18. Water retention time is not the same as transfer effectiveness

A degasser containing 1,000 litres of operating liquid volume at 500 litres per minute has nominal volume-over-flow residence time of two minutes. But the real path of water through a packed tower may include films, droplets, reservoirs and short circuits. The ratio is not proof that all water sees the same contact time. The effective gas-transfer area and driving force may change throughout the column. For this reason, a two-minute nominal residence does not imply a fixed reduction percentage in every system.

The word “retention” can be especially confusing because a diagram may include a separate sump or balance tank. Do not add the whole system volume when calculating the hydraulic residence time of the contactor. If a reservoir holds water after degassing, it may alter the timing of subsequent sampling and gas equilibration. Specify which physical volume is included, whether it is a mixed tank or a flowing film, and whether the purpose of the number is hydraulic delay or measured mass-transfer performance.

19. Ventilation can remove heat as well as gas

Air moving over warm water can collect moisture and heat, introducing a thermal penalty. A blower arrangement that increases CO₂ transfer may also increase electricity consumption and evaporative cooling. The importance of this trade-off depends on climate, inlet-air conditions, water temperature, insulation and heat recovery. A comparison that records only outlet CO₂ may miss a significant operating cost. In the FAO AGRIS record on gas-to-liquid ratios and heat loss, the investigators explicitly evaluated the thermal consequences of ventilation.

When reading cost results, keep energy per day, power and energy per removed mass distinct. A plant using 24 kWh over a 24-hour interval averages 1 kW, not 24 kW. If a study reports 30 kJ per gram of CO₂ removed under its experimental conditions, this is an energy-intensity measure, not a universal heat-loss rate. Multiplying by a measured removal mass requires that the operating conditions and interpretation be appropriate. Preserve the energy unit and denominator together.

20. Fouling, biofilm and flooding

Wetted packing can accumulate biofilm or fine solids. This may alter water distribution, pressure drop and effective air passage. A column with the same geometric area and pump flow can therefore perform differently after months of operation. Airflow restriction near the support screen, uneven wetting and local flooding are distinct failure mechanisms that need observation. A higher inlet CO₂ load can also cause worse outlet values without any mechanical defect. Diagnose generation and removal separately instead of assuming every concentration rise means packing has failed.

A useful inspection sequence compares measured airflow, water flow, pressure, packing condition, temperature and paired concentrations before and after maintenance. If cleaning is followed by improvement, record the conditions rather than claiming a universal effect. Sensor location and calibration may also have changed during maintenance. Write “the measured outlet concentration decreased after cleaning under recorded conditions” if that is the evidence; avoid “cleaning permanently fixed the gas exchanger” unless sustained testing supports the stronger conclusion.

21. Where the fish actually experience the water

A degasser outlet concentration is not automatically the concentration that reaches fish. Water may pass through oxygen equipment, mixing tanks, additional biofilters or long pipes before returning to a culture enclosure. Carbon dioxide can be produced or redistributed along that route. Final exposure depends on the whole recirculating system and ongoing respiration, not only one successful unit operation. A defensible report includes representative tank-water measurements and relevant welfare observations in addition to equipment inlet and outlet data.

The same principle applies to total dissolved gas pressure, oxygen and nitrogen. A design that removes CO₂ may change other gas conditions, and a pressurized oxygen stage can create its own considerations. Do not translate an isolated “CO₂ reduced” result into “all dissolved gas hazards removed.” Different endpoints require separate monitoring and interpretation. When people and animals depend on a system, the most important location is often the point of exposure, not the easiest laboratory sampling port.

22. Compare models only within validated conditions

Models of degassing use equations for mass transfer, equilibrium and flow patterns. Some equations assume non-reactive dissolved gases, well-defined gas composition or idealized contact. Those approximations can be useful but may not hold for CO₂ in every carbonate system. USGS authors explain why the physical device, gas enrichment and liquid reactions affect how transfer models should be used. A fitted coefficient for a cold acidic system should not be pasted into a warm marine calculation without checking its domain of validity.

Suppose a model predicts 5.4 mg/L at the outlet while an independent observation is 6.2 mg/L. The absolute discrepancy is 0.8 mg/L. A signed error divided by observed 6.2 differs from a signed error divided by predicted 5.4. The report must state its error definition rather than calling 0.8 an “0.8 percent difference.” Calibration data, independent validation and measurement uncertainty matter. A calculator’s numerical precision should never be mistaken for evidence of model accuracy.

23. A causal troubleshooting map

When a tank’s CO₂ rises, check whether biomass or feeding has increased, water temperature has shifted, biofilter respiration changed, flow to the degasser decreased, or actual airflow is below its designed operating point. Then check column wetting, fouling, gas recirculation and sampling methods. Each hypothesis should have an observation that would distinguish it from alternatives. Vague statements that “the system has too much gas” cannot identify whether the problem lies in generation, transfer, monitoring or an upstream water-chemistry change.

An evidence-matched incident note might read: “Tank free CO₂ increased during the recorded feeding interval; the blower controller remained enabled, but delivered airflow was not independently checked; outlet samples were taken from a different location than earlier comparisons; carbonate calculations will be repeated with current alkalinity.” This wording preserves the observed concern and describes the next diagnosis. It is clearer and safer than declaring the blower broken solely from a changed tank concentration.

24. Exercise: repair a percent-removal statement

Exercise A: paired free-CO₂ measurements are 24 and 9 mg/L at inlet and outlet. Difference = 15 mg/L. Relative change = 100×15/24 = 62.5 percent. At 40 m³/h of properly matched water flow, apparent mass-rate difference = 15 mg/L×40,000 L/h = 0.60 kg/h. The repaired report should say “62.5 percent concentration decrease in paired free-CO₂ samples, equivalent to 0.60 kg/h apparent water-side change at the recorded flow.” It should not say that total inorganic carbon fell by exactly 62.5 percent.

Exercise B: a water loading is given as 70 m³/(m²·h) and a draft English summary calls it 70 litres per second per square metre. The units are wrong. Convert 70,000 litres per hour per square metre to approximately 19.4 litres per second per square metre. If the column footprint is 1.2 m², total water flow at that loading is 84 m³/h. The denominator is not optional: flow per area is different from total pump flow.

25. Exercise: carbon basis, blower ratio and pH

Exercise C: a validated gas balance says 2.0 kg of molecular CO₂ entered the exhaust. Approximately how much carbon is contained in that gas? Multiply 2.0×12.01/44.01 to obtain about 0.546 kg carbon. If the original figure was already reported “as C,” applying this conversion would be wrong. Exercise D: delivered air flow is 420 m³/h and treated water flow is 84 m³/h. G/L is 5 on compatible volumetric bases; if the label gives only standard air volume and an unknown water temperature basis, record that unresolved comparison condition.

Exercise E: water pH rises from 7.0 to 7.3 after a column. Can one calculate exact free-CO₂ removal from the pH change alone? No; valid carbonate chemistry also needs alkalinity and other relevant inputs. Exercise F: a blower is nominally rated at 700 m³/h but actual measurement is 480 m³/h. Which flow enters the operating-point ratio? Use the verified delivered air flow on the stated basis, and retain 700 m³/h only as a rated reference. These exercises require no invented farm data because their assumptions are explicit.

26. Read older and newer reports with consistent definitions

Two studies separated by years may use different CO₂ sensors, computation rules, pH scales or temperature corrections. The word “removal” can refer to apparent free-CO₂ concentration difference, true transfer inferred from chemistry, or gas-side mass flow. Before placing their reported percentages in a single comparison chart, reconcile these definitions. A table should include inlet basis, outlet basis, flow, water chemistry, device class, sampling conditions and the calculation formula. Otherwise an apparently ranked list may be measuring methodological differences rather than technology quality.

Good cross-language work adds a brief methods glossary rather than silently standardizing every term. Keep the original value and method in one column and a clearly marked normalized estimate in another, with factors and limitations. If a report’s conversion cannot be reconstructed, leave it unresolved. Accurate uncertainty is more valuable than a fabricated precise number. This approach lets specialists make defensible comparisons while readers understand why not every study is directly interchangeable.

27. Frequently asked questions

Is CO₂ stripping just aeration? Both involve gas exchange, but the equipment and reported endpoints can be different. Does higher G/L always yield proportional removal? No; gas composition, hydraulics and diminishing returns matter. Can pH alone prove an exact amount of CO₂ removed? No. Does a column’s nameplate flow establish actual ventilation? No. Do two concentrations determine total transfer without flow? No. Does a good outlet value prove the fish tank is safe under every condition? No; exposure and operation across the full loop must be assessed.

Why do sources sometimes report different efficiencies for similar towers? Water chemistry, temperature, packing, airflow, test definitions, sample timing and carbonate equilibration can all differ. Why not call every carbon result a CO₂ measurement? Because free CO₂, DIC as carbon and alkalinity report different things. The translation succeeds when the reader can reconstruct precisely which physical stream, chemical species and calculation the original value represents.

28. How to issue a defensible interpretation

A complete report should show a system flow diagram; positions for paired samples; temperature, pH, salinity and alkalinity; analytical or calculation basis for free CO₂; tower geometry; water and delivered air flows; hydraulic and gas-to-liquid ratios; and the exact formula for the reported concentration or mass reduction. It should also record operating variability, repeatability, uncertainty and any heat or power trade-offs. Separate directly measured quantities from assumptions and derived estimates so a reader knows where confidence is strongest.

The final summary need not be intimidating. It can explain that under defined conditions a tower reduced a specified free-CO₂ measure between named points, with a calculated apparent mass change at measured flow, while the full system and welfare context require separate validation. Such wording is both accessible and scientifically serious. Technical translation is most valuable when it turns numbers into meaning without hiding the conditions that give those numbers their meaning.