Aquaculture Ozone: ORP, Dose, Residual Oxidants and Bromate

Why an ozone reading needs interpretation

An aquaculture water-treatment report may describe an ozone generator, an oxidation-reduction potential reading, a contact vessel, an oxidant residual and a claim about cleaner water. These are related but different quantities. An ozone machine’s rated output is not the mass that actually dissolves; an electrical ORP measurement is not a direct ozone concentration; and disappearance of a visible colour does not prove pathogen inactivation. To translate such a report into meaningful language, identify what was measured, where the sample came from, how long water contacted the oxidant, and whether the water ultimately reached fish or people. Losing any of those distinctions can turn a cautious observation into an unsafe instruction.

This article explains ozone treatment in recirculating aquaculture systems from first principles, with particular attention to dose, demand, ORP, contact time, residual oxidants and the risk of bromate formation in bromide-containing water. The numerical examples are constructed teaching exercises, not operating data, regulatory limits or prescriptions for any farm. An actual installation requires species-specific welfare assessment, water-chemistry measurement, appropriate engineering, occupational safety provisions and site validation. Readers should not infer a safe dose or ORP target from a hypothetical calculation. The goal is to read an operational claim accurately and recognize what extra evidence would make it defensible.

The recurring question is deceptively practical: what exactly does the reported number justify? A label saying 30 grams per hour describes a generator under specified conditions; it may not describe absorbed ozone. A plot reaching 350 millivolts describes a redox electrode response with calibration and location qualifications; it does not by itself quantify disinfection. A laboratory report of bromate has its own sampling, detection and concentration basis. Clear translation keeps these separately attributed so managers do not confuse electrical indicators, chemical exposures and biological outcomes.

1. Start with the chemistry and the boundary of the system

Ozone, O₃, is a reactive form of oxygen that can oxidize selected substances in water. In a recirculating aquaculture system, that reactivity can help address dissolved or suspended organic matter and some microorganisms, but the same chemical property creates hazards for gills, biofilters and operators. Ozone is introduced through gas handling and a contact device, commonly separate from the fish-rearing tank. Its effects depend on concentration, reaction kinetics, water temperature, dissolved substances, pH and mixing. The correct conceptual boundary is the treatment loop, not merely the number on the ozone-generator display.

A practical diagram should trace a feed water line into the contactor, the ozone gas supply, vented off-gas, downstream residual removal, and the return line to cultured animals. Label sample points before and after each stage. If a report measured ORP upstream but residual oxidants downstream, those readings are not interchangeable observations of the same water parcel. Water continually moves while reactions proceed, so measurements need time stamps and hydraulic context. A translator should identify whether flow rates refer to the contactor sidestream, total system recirculation, or water supply replacement; otherwise an apparently precise dose calculation uses the wrong denominator.

2. Gas output, injected mass and transferred mass

A generator output such as 24 grams of ozone per hour is a mass rate. Suppose the generator operates at its documented rating for 30 minutes. The nominal generated mass is 24 g/h × 0.5 h = 12 g. This does not mean twelve grams entered the water or remained biologically active. Ozone can be lost during generation variability, gas transfer, off-gas ventilation and reactions in the contactor. If a separate validated test estimated 60 percent net transfer during that half hour, the illustrative transferred mass would be 12 × 0.60 = 7.2 g. That figure still is not a measurement of residual dissolved ozone.

Engineering reports should distinguish generated ozone, injected ozone, absorbed ozone and measured residual. A mass balance can use inlet and outlet gas ozone concentrations and gas flow when those are measured correctly; liquid samples may reveal the remaining oxidants after a defined residence time. Saying an installation uses twelve grams is ambiguous unless the sentence names which mass and time basis. When comparing equipment quotations, read the specified feed-gas purity, temperature, pressure, nominal oxygen supply, operating point and any manufacturer derating. Two machines with the same nameplate output can deliver different dissolved exposure in water with different organic loads.

3. Distinguish concentration from dosing rate

An ozone dose reported as milligrams per litre requires a volume basis. As a toy example, a measured transferred mass of 6,000 mg over a well-defined treatment interval and a corresponding treated water volume of 15,000 L yields an applied mass-per-volume of 0.40 mg/L. This is an average transfer-based dose, not necessarily the concentration found at the outlet. Divide by the volume treated, not by the entire facility volume unless that whole volume passed through the contact stage during the interval. Reporting only 0.40 without mg/L or the treated-volume definition makes the result uninterpretable.

Concentration is a snapshot or time-resolved value at a location, whereas a mass flow rate is a quantity per unit time. In a continuous system, a mass-loading estimate can be written as water flow multiplied by concentration difference when the analyte basis is valid and reactions are understood. An ozone consumption estimate may differ from pathogen exposure because demand can be taken up by organic matter and other reducing substances before all targets are reached. A good explanation therefore names whether the reported number is nominal feed, transferred dose, instantaneous measured dissolved concentration, time-integrated exposure or calculated oxidant demand.

4. Demand makes identical doses behave differently

Water is not an inert container. Particles, dissolved organic compounds, nitrite and other substances may react with added oxidants. If a test liquid consumes most of a nominal ozone dose very rapidly, the remaining chemical exposure available further down the contactor may be small. Another water sample with lower oxidant demand may retain a residual longer under a similar applied dose. This is why dose-response claims need a defined water matrix. Reporting a successful treatment in clear laboratory water does not automatically demonstrate the same result in heavily fed, turbid, biofilter-associated aquaculture water.

A sensible interpretation asks how ozone demand was determined. Was it measured with an appropriate method at several contact times? Was it inferred from changes in ORP or organic carbon? Were water temperature, salinity, pH, solids, nitrite and treatment history recorded? An apparent fall in turbidity may have more than one explanation, and a measured change in nitrite cannot be equated with a uniform disinfection outcome. A quality-controlled experiment normally includes untreated controls, replicated measurements, specified sample handling and explicit uncertainty. Without them, the report may support a useful observation but not a universal mechanistic conclusion.

5. ORP is a mixed redox signal

Oxidation-reduction potential, or ORP, is typically displayed in millivolts. It reflects the electrochemical condition at an electrode in a solution containing multiple oxidizing and reducing couples. It is not a species-specific ozone sensor. The reading responds to water chemistry, electrode condition, pH, reference electrode, temperature and equilibration time. Consequently, the sentence ‘ORP was 350 mV’ should be interpreted as one instrument’s reading at one place and time, not automatically as an ozone concentration or proof that a particular organism was inactivated.

This distinction matters during translation because the words redox potential, oxidation potential, ozone residual and disinfection potential can sound similar in everyday language. They have different operational meanings. If an operator uses ORP as part of a site-specific control strategy, the relationship must be established under relevant conditions and backed by independent water-quality observations. A single threshold borrowed from another system is not transferable as a safety limit. The New South Wales Department of Primary Industries specifically cautions that ORP measurements in ozone treatment are approximate and that electrode response can lag changing conditions.

6. Electrode position and response time

Imagine a probe installed immediately after an ozone injection point. Water may not be fully mixed there, and bubbles or local concentration gradients can affect what the probe experiences. A second probe after a contact vessel samples a different stage of treatment, while a third in the culture tank tells a different story again. Calling all three ‘the system ORP’ erases the positions that make the data interpretable. A robust equipment log includes the sampling port, direction of flow, sensor maintenance and the elapsed time since flow, injection rate or water chemistry changed.

ORP is also not instantly responsive to every disturbance. Following a feed event, chemical addition or temporary generator interruption, an electrode may take time to settle. Comparing a delayed signal with an instantaneous setting can create a false impression that control is tight when the process has already moved. In a worked reading exercise, suppose a change occurs at 10:00, the trace begins shifting at 10:03 and settles by 10:10. The log should report a response with lag rather than claiming the water was stable at 10:01. Sensor dynamics belong in the interpretation.

7. Contact time is hydraulic, not merely a timer label

Contact time describes how long water spends under relevant treatment conditions. For an idealized, completely mixed volume, dividing 1,200 L by a throughflow of 400 L/min gives a nominal hydraulic residence time of 3 minutes. This is an arithmetic residence-time estimate, not a guarantee that each water parcel receives exactly three minutes of exposure. Real contact vessels can have short-circuiting, dead zones, imperfect mixing and variable flows. A design report may therefore use tracer-derived residence-time distributions or a conservative effective contact-time measure rather than a single volume-over-flow number.

If an ozone treatment report combines residual concentration and time into a CT expression, inspect how that quantity was derived. A measured concentration at the final outlet multiplied by nominal vessel residence time is not generally equivalent to the integral of concentration along the exposure history. Ozone can decay quickly and different organisms have different susceptibility. Readers should ask for the organism, validation method, measured oxidant species, exposure calculation, temperature and water chemistry. Avoid substituting an attractive CT number for a demonstrated microbial outcome.

8. Residual oxidants must be named

The phrase ‘residual ozone’ can hide different analytical entities. In freshwater, dissolved ozone may react or decompose relatively rapidly. In saline or bromide-containing water, ozonation can produce other oxidants that persist and remain hazardous. A laboratory or on-site method may report total residual oxidants rather than dissolved ozone specifically. It may use an equivalent concentration convention that needs to be preserved when translating a table. Converting a total residual oxidant result directly into molecular ozone without knowing the method changes the underlying claim.

A good report states the analytical method, calibration or reporting basis, sample point, sample age, storage conditions and detection or quantification limit. The return line to live animals must be evaluated separately from the contactor because treatment effectiveness and animal exposure are different endpoints. No chemical-indicator reading, however reassuring, replaces a design that prevents hazardous residuals from reaching stock or exposing workers. University of Florida aquaculture guidance emphasizes removing ozone before reintroducing treated water and evaluating seawater by-products carefully.

9. Bromide, bromate and marine systems

Bromide is naturally present in seawater and can participate in ozonation chemistry. Under some conditions, ozonation can generate bromate and other bromine-derived oxidants or by-products. Bromate concentration must be treated as a laboratory analyte in its own right, with an identified sampling point and reporting limit. A low or absent ozone reading does not prove that bromate or other products are absent. Nor does a particular example of bromate formation in one system establish that another system will form an identical concentration.

The risk depends on treatment chemistry, contact conditions, matrix and sequence. In a marine recirculating system, a translator should preserve the difference between a dissolved oxidant present immediately after contact and a stable chemical by-product measured later. Mixing those terms could cause someone to believe a short wait erases every risk, when the analytical answer is compound-specific. Studies indexed by the US Environmental Protection Agency describe bromate formation in marine ozonation and emphasize that the chemistry must be examined rather than assumed. No universal safe setting follows from those studies alone.

10. Nitrite reduction is not the same as biofilter capacity

Ozone can oxidize nitrite to nitrate, so a lower nitrite reading downstream of treatment may reflect chemical oxidation as well as microbial processing. This matters because a nitrifying biofilter’s capacity is biological and responds to its loading history. If chemical treatment handles much of the nitrite for a prolonged period, a subsequent ozone interruption can reveal that the biofilter does not immediately compensate. Calling ozone ‘a stronger biofilter’ would be a poor translation; the mechanisms and failure behaviours differ. A well-worded report separates the chemical removal contribution from biological conversion.

Suppose an influent contains 0.60 mg/L nitrite reported as nitrogen and an effluent contains 0.20 mg/L on the same nitrogen basis. The numerical difference is 0.40 mg/L as N. At a properly paired flow of 10,000 L/h, the illustrative net change is 4,000 mg N/h, or 4 g N/h. It is not 4 g of ozone per hour, and it cannot establish how much was converted by biofilm versus oxidation unless the experiment separates pathways. Note whether the reported concentrations are NO₂⁻ mass or NO₂–N because the conversion factor changes the arithmetic.

11. Understand nitrogen reporting bases

Chemical reports frequently use ‘as N’ to express the mass of nitrogen contained in a nitrogen species. A value of 1.00 mg/L NO₂–N means 1.00 milligram of nitrogen per litre, not one milligram of the complete nitrite ion. The conversion from NO₂–N to NO₂⁻ is about 46.01/14.01, or 3.28. Thus 0.50 mg/L NO₂–N corresponds to approximately 1.64 mg/L NO₂⁻. A table mixing these bases can create an apparent threefold change that is only a unit error. Retain the original label next to every imported value.

The same care is needed with ozone mass, oxidant-equivalent units and micrograms per litre of bromate. Never normalize results across laboratories by silently deleting the ‘as’ suffix. A technician translating a cross-language certificate can use a separate normalized comparison column, but the original measurement basis and factor should remain traceable. This principle joins aquaculture chemistry with good scientific communication: an analyte identity, quantity and denominator form one inseparable claim. Without all three, even accurate numerical transcription can mislead downstream decisions.

12. Read percent removal with the right denominator

A treatment report may claim ’80 percent removal’ after comparing two water samples. The straightforward concentration formula is 100 × (inlet concentration − outlet concentration)/inlet concentration. In a hypothetical matched pair of 10 mg/L and 2 mg/L, the reduction is 80 percent. That arithmetic assumes the two concentrations describe comparable water and measurement bases. It does not show an 80 percent reduction in total mass if flow changed or sidestreams joined between sample points. It also says nothing about whether the remaining 2 mg/L is acceptable for a particular species or use.

An alternate calculation uses mass loading: inlet flow multiplied by inlet concentration minus outlet flow multiplied by outlet concentration, with consistent units. For example, 6 m³/h × 10 mg/L equals 60 g/h entering a defined component; 5 m³/h × 2 mg/L equals 10 g/h leaving in the measured outlet, but the missing 1 m³/h must be located before calling 50 g/h ‘destroyed.’ Backwash, drainage or sideflow could carry some constituent away. Methodological clarity prevents a concentration improvement from being oversold as complete chemical elimination.

13. Oxidation is not guaranteed sterilization

Microbial disinfection claims require tests specific to the target organisms and conditions. A decrease in colour, dissolved organic carbon or turbidity can improve treatment but is not the same variable as a reduction in viable pathogen concentration. Depending on the study, the biological endpoint may be culture counts, infectivity, molecular detection or a validated surrogate. Molecular detection of genetic material after treatment can remain positive even when infectivity changes, while a lack of detection may have sampling-limit explanations. That is why translating a paper’s word ‘inactivation’ into ‘all germs destroyed’ is unjustified.

A log reduction is logarithmic, not an ordinary percentage. A 1-log₁₀ reduction is a tenfold decrease in measured count under its stated method; 2-log₁₀ is a hundredfold decrease. An initial 100,000 units per millilitre and an outlet 1,000 on the same valid basis give log₁₀(100,000/1,000)=2. If the outlet is below detection, report it as censored within the assay’s limits rather than replacing it with zero and announcing infinite reduction. Method choice, replicates, uncertainty and controls matter more than an impressive log number alone.

14. The return-to-stock checkpoint

The contactor may accomplish a useful chemical purpose while still making its effluent temporarily unsafe for fish. A responsible flowsheet therefore has a deliberate separation between treatment exposure and culture exposure, with suitable gas handling, residual removal and verified return-water controls. The operational report should make the return-to-stock checkpoint explicit: what is measured immediately before the animals receive water, who validates the limit for the species and life stage, and what action occurs when monitoring fails? Translating ‘treated’ as ‘ready for fish’ skips this essential safety boundary.

Some equipment claims imply that mixing treated water with a larger system volume automatically resolves exposure. Dilution is not equivalent to removing reactive oxidants or eliminating persistent by-products. A concentration can be diluted, but dose, timing, sensitive life stages and localized mixing zones still matter. An evaluation also needs contingencies for probe drift, intermittent flow, generator shutoff, air-handling failure and carbon or other residual-treatment exhaustion. A safe report states verified performance rather than relying on a single nominal design feature.

15. Workers and off-gas

An ozone installation has air-side hazards as well as water-side hazards. Gas lines, seals, pressure conditions, off-gas destruction and ventilation belong in the equipment description. The fact that operators sometimes notice a pungent smell does not make smell a reliable exposure instrument. A translated operations manual should not substitute casual sensory impressions for appropriate monitoring and relevant workplace standards. Staff training and exposure-response procedures are separate from the aquaculture water-quality analysis, but they intersect at the contactor and its vent lines.

Consider a log that reports that off-gas destruction was ‘on’ throughout a shift. This confirms only a command or status indicator unless the equipment’s actual function was monitored. A cautious statement would be ‘the controller recorded the off-gas unit as enabled,’ followed by the inspection or measurement that would substantiate operational performance. Language precision matters because a status light, a verified operating condition and an occupationally safe environment are three different levels of evidence. Electrical alarms and chemical exposure monitoring should not be conflated.

16. Worked example: generator and water-flow reconciliation

Take a teaching scenario with an ozone generator nominally supplying 18 g/h into a contactor sidestream of 30,000 L/h. If we assume, solely for arithmetic, that 50 percent of generated mass transfers to water during this interval, the transferred rate is 9 g/h or 9,000 mg/h. Divide by 30,000 L/h to obtain 0.30 mg/L as a calculated applied mass per volume. Check the units: mg/h divided by L/h is mg/L. Neither an outlet ozone concentration nor pathogen inactivation has been measured in this example.

Suppose another analyst uses the whole recirculation flow of 90,000 L/h as denominator and reports 0.10 mg/L. That may describe a different system-level averaging assumption, not the same concentration at the treatment point. Both calculations could be arithmetically correct while answering different questions. The repair is to label the 30,000 L/h sidestream and the 90,000 L/h system total, then state the location to which each derived value applies. If downstream mixing is incomplete, even the system-average figure is not a measured local exposure.

17. Worked example: an ORP chart and a delayed alarm

Imagine an ORP line that reads 290 mV before a feed event, falls to 255 mV, and returns to 300 mV after operators adjust treatment. It would be wrong to say that dissolved ozone first decreased by 35 mg/L and later increased by 45 mg/L. The chart is in millivolts, not milligrams per litre, and is not ozone-specific. The change may reflect additional reducing load, instrument dynamics or multiple chemical processes. A defensible description says the electrode’s redox potential declined and later recovered while relevant operating conditions changed.

Now add a three-minute response lag to the probe, a five-minute logging interval and a brief 90-second excursion. The recorded minimum may not capture the true timing or amplitude of the disturbance. It would be safer to flag a potential monitoring-resolution limit than to interpolate a precise event. This exercise teaches why charts need the instrument response time, sampling interval, location, maintenance records and linked chemistry. Good language does not turn a convenient process indicator into a unique diagnosis.

18. Worked example: detection and uncertainty

A bromate laboratory certificate may report ‘less than 5 micrograms per litre.’ This means the result was below a stated reporting threshold for the method and sample; it does not mean bromate is chemically absent. A second sample reported as 7 ± 2 micrograms per litre may not represent a clear increase when uncertainty and detection context are considered. The exact interpretation depends on whether the ± term is standard uncertainty, expanded uncertainty, a confidence interval or something else. Preserve the laboratory’s definition.

Converting units is straightforward: 5 µg/L is 0.005 mg/L, because 1 mg is 1,000 µg. The translation must carry the less-than qualifier across the conversion, yielding <0.005 mg/L, not 0.005 mg/L as an exact measurement. If a graph cannot display a censored value conveniently, mark it appropriately rather than silently placing the point at zero. A carefully reported limit can be more honest and useful than a fabricated decimal.

19. Read a mass-balance worksheet without losing the basis

A treatment ledger may contain three columns: water flow, pollutant concentration and calculated mass per hour. Before interpreting trends, verify whether every concentration is measured on the same analyte basis, whether sampling times match flow measurements, and whether the flows were steady. For a hypothetical inlet of 12 m³/h at 4 mg/L, the input mass rate is 48 g/h. If an outlet of 12 m³/h measures 1 mg/L, the output mass rate is 12 g/h and apparent net disappearance is 36 g/h. This calculation does not identify the reaction products.

Chemical transformation often changes the species carrying an element without removing that element from water. Oxidation of nitrite to nitrate, for instance, shifts nitrogen speciation. A report claiming ’36 g of nitrogen removed’ would need evidence that nitrogen left the defined system boundary, not merely that a named intermediate decreased. The stronger wording is ‘a net reduction of 36 g/h in the measured analyte across this stage, subject to matched flows and analytical uncertainty.’ Precise wording preserves useful progress without claiming more than measurements support.

20. Compare ozone with other treatment components fairly

Mechanical filtration, biological nitrification, UV treatment, degassing and ozonation work through distinct mechanisms. A drum filter physically captures certain suspended solids; a biofilter transforms nitrogen species; a UV device delivers radiation exposure to water passing through its reactor; a degasser changes dissolved gas equilibrium; ozone reacts chemically. A site may need several stages. Ranking them as interchangeable ‘disinfection power’ products hides what each does and what each fails to do. Compare under the same explicit endpoint: removal of a defined solid fraction, reduction of a measured organism, oxygen delivery or control of a chemical by-product.

Energy consumption, maintenance, labour and failure behaviour must be included alongside apparent water quality. A treatment stage that improves one indicator while adding another risk is not unambiguously superior. For example, higher ozonation could improve water clarity but alter residual oxidants, nitrogen chemistry or bromate formation depending on conditions. Read comparative claims with controls, units and uncertainty intact. That is especially important when a report is translated between laboratories, suppliers and farm operators with different assumptions.

21. Troubleshooting misleading success indicators

When ORP rises but livestock behaviour worsens, do not assume higher is better. When colour clears but microbial tests remain positive, do not assume the tests contradict the visual observation. When nitrite falls during ozonation but rebounds on interruption, do not assume the biofilter was permanently repaired. These are separate measurements with different causal pathways. The troubleshooting sequence is to document the actual symptom, identify the corresponding measurable endpoint, trace the water path and check sensor calibration and safety conditions before forming a mechanistic hypothesis.

A useful incident narrative uses four sentences rather than a vague claim: the observed change; the time and place of observation; the checks completed; and the remaining uncertainties. An example might say that outlet clarity increased after a contactor change, while no controlled pathogen reduction test was performed and residual oxidant verification was incomplete. This does not hide the improvement. It makes the improvement actionable without turning it into unsupported assurance.

22. Cross-language terminology that changes the decision

The word ‘dose’ often points to mass delivered, but in some supplier documents it may denote generator setting. ‘Residual’ can mean ozone, total oxidants or an equivalent measurement. ‘Disinfection’ can imply different test endpoints from ‘sterilization.’ ‘Contact time’ can mean nominal vessel residence time or verified effective exposure. When interpreting a multilingual specification, keep the original term next to a controlled English definition and its measurement basis. An ambiguity note is better than confidently choosing a stronger meaning.

Use a two-column glossary that protects the distinction between equipment setpoint and measured outcome. Record ozone generation in g/h; transferred mass in g/h; application normalized to treated water in mg/L; ORP in mV; concentration of a defined residual in appropriate mass per litre; bromate in µg/L or mg/L; and biological reduction in properly specified log units. The translator’s task is not to make the document more impressive. It is to make each claim testable.

23. Source hierarchy and what a reader can verify

For practical interpretation, start with first-party guidance from competent aquaculture authorities and universities, then examine original methodological studies for the exact question. University of Florida extension explains ozone’s aquaculture benefits and risks, including the need to eliminate residual ozone before return to fish. The New South Wales Department of Primary Industries describes why ORP control is approximate. Peer-reviewed work on seawater ozonation examines residual toxicity and bromate by-products under its own conditions. Each source addresses a different piece of the system; none supplies a universal operating target for every farm.

A strong literature note should identify whether a source is extension guidance, a controlled experiment, a review or a product claim. Record species, salinity, equipment, water chemistry, dosing basis, measured endpoints and limitations. A translated sentence that strips those conditions away may sound clearer but becomes less accurate. Reproducibility is the best test of translation quality: could another qualified reader reconstruct what was done, at what location, with what units and how much confidence?

24. Independent transfer exercise: define the denominator

Exercise A: a unit reports 15 g/h generated ozone and an estimated 40 percent transfer to 20,000 L/h of contactor water. Compute transferred grams per hour and average mass per volume, then state what is not proven. The answer is 6 g/h transferred and 6,000/20,000=0.30 mg/L application on the defined contactor-flow basis. There is no measured outlet residual, pathogen inactivation or stock safety conclusion. If a colleague divides by 100,000 L/h total recirculation and reports 0.06 mg/L, they have changed the denominator; ask what system question the new figure represents.

Exercise B: nitrite is recorded as 0.30 mg/L NO₂–N in one report and as 0.90 mg/L NO₂⁻ in another. Can you declare the second concentration three times larger? No. Convert 0.30 × 3.28 ≈ 0.98 mg/L NO₂⁻ before comparing, and then consider analytic uncertainty and time. The apparent difference is primarily a reporting-basis issue. This repair also teaches why translated tables should carry a locked unit field rather than placing bare numbers under a general heading ‘nitrite.’

25. Independent transfer exercise: a claim with weak evidence

Exercise C: after adding ozone, an operator records that the water looks clearer, an ORP probe rises from 275 to 325 mV, and a single test detects no bacteria in a small sample. Rewrite the sentence ‘Ozone sterilized the system.’ An evidence-matched version is: ‘Following the treatment change, visible clarity improved, the probe’s ORP reading increased, and the specified small sample yielded no detections in the reported test.’ The next question is whether replicated, validated pathogen measurements and residual assessments support a broader claim. The exercise is about language, not dismissing observations.

Exercise D: a lab writes bromate <2 µg/L and an English summary writes ‘bromate 0.002 mg/L.’ Repair it to ‘<0.002 mg/L under the stated reporting threshold.’ Exercise E: a vessel holds 900 L at 300 L/min. The nominal residence time is 3 minutes, but the report says every parcel spent three minutes in contact. Repair the wording to ‘nominal hydraulic residence time of three minutes; actual exposure distribution requires mixing or tracer evidence.’ These small language corrections preserve chemical and engineering meaning.

26. Frequently asked questions

Does high ORP prove ozone is present? No. ORP is a mixed electrochemical signal and cannot independently identify ozone or quantify a particular disinfectant. Does a nominal dose prove a particular log reduction? No. The microbe, contact conditions, water chemistry and validated biological method must be specified. Does clear water mean microbiological safety? No. Optical clarity and microbial counts are different endpoints. Can a laboratory result below a reporting limit be recorded as zero? No. Preserve the qualifier, units and method.

Can ozone replace a biofilter? Not as a simple one-for-one substitute. Chemical oxidation and biological nitrification respond differently to loading and failures. Are freshwater and seawater ozonation interchangeable? No. Bromide chemistry and residual oxidant products can differ substantially. Is the safest translation always the shortest wording? No. A short sentence that loses flow basis, sampling point or qualifier may be harder to understand correctly than a longer sentence naming what was actually measured.

27. A defensible reporting checklist

Before signing a report, confirm generator output and gas handling, the water volume or flow used in each dose calculation, contactor configuration, sampling-point locations, timing and hydraulic assumptions. Confirm whether ORP values are electrode readings with appropriate calibration; whether oxidant measurements are species-specific or total-equivalent; whether microbiological endpoints are supported by validated tests; and whether bromate and other by-products were considered when relevant. Record what residual-control evidence applies at the actual return-to-stock boundary, not only at a convenient upstream port.

Finally check the language of every conclusion. Remove suggestions that an instrument setting establishes biological safety. Keep limitations beside calculated values. Distinguish sample observations from plant-wide guarantees, derived rates from direct measurements, and operational convenience from validated performance. An excellent translation lets a reader act on reliable evidence while clearly recognizing what remains unknown. The achievement is not a dramatic claim about ozone; it is an accurate, reviewable explanation of what the water-treatment data mean.