To translate CFU, MPN, NTU and a water laboratory test report correctly, preserve the organism or parameter measured, the laboratory method, the sample volume and the reporting limit. A result of 20 CFU/100 mL does not mean the same thing as 20 MPN/100 mL merely because both refer to microbial contamination. Likewise, a turbidity result of 2 NTU is an optical water measurement, not an estimate that two bacteria were found. Accurate water-analysis translation carries the evidence, not just the number.
This practical guide answers searches about water quality report translation, CFU versus MPN meaning, colony-forming units per 100 mL, most probable number testing, nephelometric turbidity units, total coliform and E. coli result interpretation, water sample chain of custody, non-detect and limit-of-detection wording. It is intended for translators, educators, laboratories, water-system communicators and international readers who need to understand technical reports without rewriting their scientific claims.
Our specific search intent is how to translate the results and qualifications in a microbiological or turbidity water-test report without changing the sample matrix, counting method or reported uncertainty. The overarching technique remains with Master Art of Translation; the protected Vocabulary Learning Hub and How English Works cover general language development. This is a specialist document-reading problem, not an alternative translation or vocabulary hub.
The 60-second answer
- CFU stands for colony-forming units and is associated with colony growth under a specified culture method.
- MPN means most probable number, a statistical estimate derived from a specified test design; it is not a direct colony count.
- NTU means nephelometric turbidity units and describes an optical characteristic of a water sample, not bacterial abundance.
- Always keep the denominator, such as per 100 mL or per mL, and the precise sampling location and date.
- Not detected, below reporting limit and zero organisms in the entire water system express different claims.
- Drinking water, recreational water, raw intake water and treated effluent can involve different test methods and jurisdiction-specific criteria; do not transplant a limit from one context into another.
Why water-test translation begins with the sample rather than the result
The first danger in translating a water laboratory report is treating the laboratory output as a statement about every litre of water in the environment. A report describes a defined sample, taken from a named location at a recorded time and tested using a selected method. The sample may have been drawn from a tap, a storage tank, a river, a cooling tower or a wastewater discharge. These are not interchangeable matrices. Even accurate laboratory results can be misread when the location, sample type or timing is lost.
Suppose a report states ‘raw intake, grab sample, collected 09:40, E. coli 18 MPN/100 mL’. If the target text calls this ‘treated drinking water, 18 bacteria per glass’, it has changed both the sampled medium and the reporting method. The error is not merely terminological. It may create unwarranted concern or false reassurance. Preserve the collection point, whether the material was treated, and the original count or estimate format. A clear translation can be simple while still giving readers the right evidence boundary.
A disciplined workflow constructs a sample identity record before translating the table. Record sample ID, location code, collection date and time, collector, material, preservation, receipt date, method, units and analytical qualifiers. These are often scattered across the first page, an appendix and footnotes. Moving them into a target-language summary is acceptable only if their relationships remain accurate and no fields are invented.
CFU is about colonies under a method, not a direct census of cells
Colony-forming units arise from microbial culture methods in which organisms capable of growing under defined conditions produce visible colonies or another countable growth endpoint. One colony does not prove that exactly one single bacterium was in the original sample; it may arise from a cluster, and organisms that do not grow under the chosen conditions may not be captured. That is why ‘colony-forming units’ is a meaningful technical expression and ‘exact number of bacteria’ is often an overstatement.
The denominator matters. A value of 15 CFU/mL is not numerically identical to 15 CFU/100 mL. A direct unit conversion of 15 CFU/mL to a per-100-mL concentration would give 1,500 CFU/100 mL, assuming the reported concentration is correctly normalized and suitable for conversion. A translator usually should preserve the laboratory’s reporting unit rather than calculating an alternative. If a target audience requires another unit, the conversion needs a clearly marked explanatory note and method review.
Do not strip the organism label. A heterotrophic plate count, total coliform count and Legionella count may all be reported using CFU, but they concern different analytical targets and often different culture procedures. Keeping CFU while replacing the target with the generic word ‘germs’ makes the result impossible to interpret. If the laboratory reports ‘presumptive’ or ‘confirmed’ colonies, retain that status too.
MPN is a statistical estimate, not a colony count in disguise
Most probable number methods use a pattern of positive and negative test results, sometimes across multiple volumes or dilutions, to estimate the concentration of an organism. The reported MPN depends on the protocol and its statistical model. Unlike direct colony enumeration on a filter or plate, the MPN value is inferred from the response pattern. Translating 23 MPN/100 mL as ’23 colonies in 100 mL’ implies an observation that was not made.
An MPN result may have a confidence interval or a method-defined range. If a table gives 18 MPN/100 mL with an interval, retain both the point estimate and interval. Omitting the interval may suggest more precision than the test supports. Avoid rewriting the statistical estimate as a whole-number count of organisms, even if the number looks like an integer. The term ‘most probable’ is part of the measurement method and should be given an understandable target-language equivalent.
Comparisons between CFU and MPN are possible in some study designs, but they are not automatically numerically interchangeable. Different recovery mechanisms, sample treatment, growth conditions and statistical procedures can lead to different results. A translator should not harmonize two columns because they appear to address the same organism. Respect the laboratory’s chosen test identifier, and if an external interpretation is needed, ask the relevant analyst to approve it.
NTU measures turbidity: do not turn cloudiness into a microbe count
Turbidity describes optical scattering and related behaviour of light passing through a water sample. The U.S. Geological Survey explains that materials including fine particles, algae and other suspended or dissolved components may affect turbidity. It is commonly reported in nephelometric turbidity units, NTU, under an applicable method. A result in NTU is not a bacterial concentration, a total dissolved solids reading or a direct percentage of particles present.
High turbidity can matter to treatment processes and environmental interpretation, but a turbidity result alone cannot establish the identity or concentration of a particular pathogen. A translation saying ‘NTU 3: three organisms detected’ is a category mistake. Similarly, lower NTU indicates less optical turbidity under a comparable method; it does not by itself guarantee microbiological safety. The correct translated noun is about turbidity or water cloudiness, with NTU retained as the measurement unit.
The measurement method may distinguish NTU, FNU and other optical unit variants. The USGS describes different optical arrangements and illuminations. Do not change FNU to NTU simply to make the target-language table uniform, and do not claim equivalence without the method basis. In a bilingual lab report, preserve the symbol printed by the laboratory and give a full explanation once.
Per millilitre, per 100 millilitres and unit conversion
Analytical results frequently report CFU/mL, CFU/100 mL, MPN/100 mL or other normalized concentrations. The slash means ‘per’; the denominator is not a footnote. Changing it changes the reported magnitude. A translation system that drops the digits after the slash can produce a hundredfold error even while spelling the organism correctly. For that reason, keep units in a protected data field in the translation workflow and perform a separate numeric quality-assurance pass.
Suppose the source gives 0.8 CFU/mL. Multiplying by 100 converts the numeric concentration to 80 CFU/100 mL if a conventional proportional unit conversion is appropriate. That is a mathematical illustration; it does not mean the laboratory observed exactly 80 colonies on a 100 mL plate. For MPN, a change in reporting volume may be mathematically straightforward for a concentration but should not be presented as a new independent test result. Preserve the original value where possible.
Make notation locale-aware but numerically stable. Some countries use commas as decimal separators and periods as thousands separators. A result of 1,250.5 must not turn into 1.2505 or 12505. Check nonbreaking spaces, superscript formatting and whether the report uses ‘per litre’ in descriptive text while the table uses ‘per 100 mL’. The source and target should agree dimensionally.
Not detected is not the same as proven absent
A laboratory’s ‘not detected’ statement means the target was not detected under the specified method and sample circumstances. It does not prove that the target could not exist anywhere in a water system, nor necessarily that concentration is precisely zero. The method may have a finite detection threshold, limited sample volume, interferences or other constraints. Translation should preserve this evidentiary limit even when the target language prefers short and categorical sentences.
Consider ‘E. coli not detected in 100 mL analysed’. A careful translation keeps the organism, negative observation, method or sample size and collection point. It should not say ‘the whole building is completely free of E. coli’. That turns one result into a universal claim. It also should not silently change ‘not detected’ to ‘less than one colony’ unless the laboratory’s formal reporting convention supports that equivalence.
Similarly, ‘below limit of detection’, ‘below limit of quantification’ and ‘below reporting limit’ can mean different things. One refers to detectable signal, one to reliable quantification and another to the laboratory’s reporting rule. In a chemical parameter table, the differences are especially common, but microbiology uses its own method-specific language. Translate the actual qualifier and retain the named limit rather than flattening every symbol into ‘zero’.
A less-than sign can carry an entire analytical conclusion
Values such as
If a report reads ‘less than 1 MPN/100 mL’, translate the relational phrase without changing the statistic into CFU. Do not write ‘no bacteria’ unless the actual method statement explicitly supports the wording in the correct sample context. Where a laboratory uses symbols such as ND, BDL, LOD or LOQ, consult its legend. Abbreviations are not globally standardized merely because several laboratories use familiar letters.
A data quality pass should inspect these symbols one by one. Operators are especially vulnerable when spreadsheets are exported to text, PDF extraction strips glyphs, or a target language is typeset right-to-left. A reviewer must compare target and source visually, including every less-than and greater-than sign, units and footnotes. A fluent sentence is not proof that the analytical value survived.
Which organism was tested: total coliforms, E. coli and other indicators
Total coliforms describe a group used in water-quality monitoring, while E. coli is a more specific organism that is often used as an indicator of faecal contamination under applicable testing programmes. Enterococci, heterotrophic bacteria and Legionella are other distinct targets. A report may show several of them, each with its own detection method and unit. A translation must not merge these targets into one generic ‘bacteria’ result, because readers would no longer know which parameter the laboratory measured.
An organism name can have a scientific Latin form, an accepted common form and abbreviations. Many professional documents retain the Latin name, such as Escherichia coli, at first use and E. coli thereafter. Keep capitalization and italicization where the publication style requires them. Avoid substituting local colloquial disease names for organisms; not every detection is a diagnosis, and an environmental test is not the same as clinical infection evidence.
Also respect specimen scope. A Legionella result from a cooling tower belongs to that sampled water system and method, not automatically to the drinking-water supply. CDC materials emphasize that sampling and processing choices influence reported values. Translate the system label and sample location consistently so readers do not mix risks from unlike water uses.
Sampling time, chain of custody and holding conditions
Microbiological results can be affected by the period between sampling and analysis, transport temperature, sample volume, preservation and whether the container was appropriate. The translator is not a laboratory scientist, but these fields are critical parts of the record. If a chain-of-custody form lists ‘collected’, ‘received’ and ‘analysed’, use different labels for those events. A report received on Tuesday may describe water collected on Monday and analysed later.
Time zones and local date formats need scrutiny in international reporting. The string 04/05/2026 can mean 4 May or 5 April depending on the convention. Do not choose by intuition. Spell out months or retain an ISO-style date where publication guidelines allow, after confirming the source. Preserve the sampling time if it matters to the report, and do not shift it to the target reader’s time zone without showing what changed.
Field blanks, duplicate samples and control results may appear in the same packet. Their codes can resemble actual customer sample identifiers. Label them accurately rather than counting every row as an independent environmental observation. A chain-of-custody discrepancy may require clarification, not narrative repair. Translators make the record legible; they do not alter the record.
Accreditation, scope and analytical method references
Water-test reports may identify accreditation, laboratory number, analytical method, method edition and any deviations. A general statement that a laboratory is accredited does not necessarily mean every possible analyte or method is inside its accredited scope. The translated report should retain the laboratory’s actual scope language and any explicit exception or subcontracting note. Avoid strengthening ‘method performed under quality system’ into ‘officially certified safe’.
Methods can involve membrane filtration, plate culture, defined substrate tests, multiple-tube fermentation, molecular detection or optical analysis. Different techniques answer related but not identical questions. If the source method is stated by code, copy that code exactly and translate its description. Version numbers, revisions and issue dates may determine which procedure was used. A translation that drops the version can frustrate subsequent review.
Country-specific regulatory limits and accreditation rules require jurisdiction-specific professional interpretation. A translator should not replace a source-country standard with a target-country requirement because the target audience is located elsewhere. The translated document can provide a clearly separated explanatory note about a relevant local framework if commissioned, but should never imply that the original laboratory tested under a standard it did not use.
Water classification: why drinking, recreational and industrial are not synonyms
Water testing takes place in many settings. Potable drinking water is evaluated for human consumption, recreational water may be monitored for bathing exposure, industrial process water may be evaluated for equipment operation, and treated wastewater may be monitored for discharge compliance. A result meaningful in one context does not carry the same regulatory interpretation in another. A translated table should preserve the source’s classification and the sampling design that produced it.
Suppose a river sample reports E. coli MPN/100 mL alongside temperature and rain conditions. A translation describing it as a household tap-water result would be wrong, even if every number were copied accurately. Similarly, a cooling-tower Legionella control report should not be promoted to a certificate for drinking safety. The organism names may overlap, but the sampled systems, relevant methods and intended decisions differ.
Readers sometimes ask whether a particular number means water is ‘safe’. That answer depends on the water’s use, applicable regulation, sampling representativeness, timing, laboratory method and the rest of the evidence. The translation can accurately communicate the finding and the source’s own compliance statement; it cannot make an unsupported new safety determination. When health decisions are at stake, refer the report to relevant local water authorities or qualified professionals.
Chemical test results that accompany microbial data
A single water report may also include pH, conductivity, total dissolved solids, free residual chlorine, metals and nutrients. These fields should not be folded into CFU or MPN because they describe different physical or chemical characteristics. pH is a logarithmic measure of acidity; conductivity relates to electrical conductance; turbidity concerns optical behaviour; disinfectant residual is a measured concentration under a specific chemical method. A clear translation keeps units and analytes distinct even when the original table is cramped.
TDS may be reported in mg/L or estimated by an instrument using conductivity and a conversion factor. That estimate should not automatically be treated as a direct gravimetric measurement. Similarly, total chlorine and free chlorine are not necessarily identical. The translator can preserve source values and explain the difference, but should not reconcile them by changing a column heading. If a laboratory defines ‘residual’ in a note, reproduce that note.
Uncertainty also varies by analyte and method. A quantified metal concentration might carry an uncertainty interval, while an organism result may use a qualitative positive/negative report or MPN confidence interval. Resist formatting every cell as though it represents the same type of evidence. Good bilingual presentation makes differences legible instead of hiding them.
What an NTU value does and does not establish
A higher NTU result generally indicates more light scatter under a comparable turbidity measurement arrangement. It may reflect particles such as silt or algae, but the instrument is not identifying each particle or a particular microbe. This is why translating a turbidity figure as ‘amount of dirt’ can be too narrow. It is an optical indicator influenced by sample and instrumentation conditions. Retain the measured unit and, if given, the instrument method and measurement conditions.
The same water body may change after rainfall, reservoir mixing, dredging or other events. A temporal rise in turbidity may be relevant to environmental monitoring, but a translation should not assign a specific cause that the report has not established. Where a report says ‘elevated NTU after rainfall’, preserve the temporal observation and source’s level of causal certainty. Do not rewrite correlation as proof of contamination.
Water quality decisions usually combine turbidity with other indicators and operating information. A field report might use NTU as one warning signal; a treatment plant might use it to assess filtration conditions. Neither document is interchangeable with a microbiological culture report. The target language should retain this division so readers do not infer an organism count from optical scattering.
Uncertainty, confidence intervals and statistical limits
MPN techniques may report confidence intervals because concentration is estimated from a response pattern. A value of 12 MPN/100 mL with a broad interval is not a declaration that there were exactly twelve organisms in the vessel. The interval helps communicate plausible concentrations under the statistical method. Translate both the point estimate and its uncertainty language, while retaining the stated confidence level if provided.
CFU counts also have variability and practical bounds. A plate may be labelled ‘too numerous to count’, reported as estimated count or excluded because of interference. These labels are part of the analytical result. Do not turn them into a made-up numerical count or replace them with the convenient word ‘high’. If the laboratory provides a reporting convention or quality code, attach it to the same row in the target table.
Trend comparisons require care because differences may reflect sampling variation or analytical limits. A report with one low result and one higher result does not automatically prove a permanent deterioration across a system. A translator should communicate the report’s statements about variability, repeat sampling and method comparability without adding a scientific conclusion not made by the source.
Geometric means, percentiles and averages in regulatory summaries
Water datasets are sometimes summarized by geometric means, arithmetic means, medians or statistical thresholds. These statistics give different weight to observations. A geometric mean commonly appears in microbial water-quality assessment because concentrations are often skewed, but it is not a simple total of all organisms. An average drawn from several sampling dates is not the same as a single measured value. Keep the method of aggregation named in the translated table.
A document may state ‘geometric mean of five samples’ or ’90th percentile of observations’. A translation that says ‘the fifth sample was 90 percent’ loses the statistical structure. Preserve number of samples, observation window and calculation basis. If the source uses a standard or legal criterion, it may specify which statistic counts for compliance. Do not silently replace that criterion with one familiar in another country.
In a bilingual report, include the aggregation word in the column header as well as the explanatory footnote. If a software system strips footnotes during export, a source reviewer may misinterpret the numbers. This is a strong reason to perform a cross-table integrity review instead of only checking paragraphs.
The difference between an observation and a compliance declaration
Laboratories frequently provide analytical results without making a regulatory decision. An operator or authority may compare those results with an applicable standard and issue a compliance finding. Those are separate functions. The phrase ‘result below reporting limit’ is not equivalent to ‘legally compliant’; neither is ‘complies with one listed parameter’ equivalent to ‘water is safe for all purposes’. Translate exactly who made the statement and what standard or criteria were used.
Some reports display a limit alongside each result. The limit might be a regulatory maximum, a customer-specified action level, an internal guideline or an analytical reporting limit. They can all look like numbers in a column titled ‘limit’. The translator must distinguish their nature from the document’s legend or context. If the source is ambiguous, query the author before replacing ‘limit’ with ‘legal limit’.
Different jurisdictions use different criteria and assessment periods. A multilingual report for international readers should preserve the named governing document and the date or version cited. It may add an explanatory note stating that criteria can differ by use and location, but should not transplant one jurisdiction’s threshold to another laboratory result without a separate authorized review.
Charts, control dashboards and time-series captions
Water dashboards may display turbidity in NTU, disinfectant residual in mg/L, pH, temperature, organism results and operational flow. An alarm threshold can be shown beside a measurement without being a legally binding limit. Translate ‘alarm’, ‘action’, ‘target’, ‘maximum’ and ‘measured’ distinctly. A threshold line on a graph may be a configurable plant setting rather than an external standard. A wrong adjective can change the interpretation of a system’s operation.
Time-series plots can mix instantaneous readings with laboratory results from grab samples. Turbidity sensors may report every few minutes while microbiological samples are processed over longer intervals. A translator must preserve the sampling frequency and delay. A single plate result plotted at a collection timestamp should not be represented as a continuously measured bacterial concentration.
Check missing values, outage gaps and substituted readings. A zero in a sensor series may mean ‘zero measured’, ‘instrument offline’ or ‘missing data encoded as zero’, depending on system design. Do not guess. Read legends and export conventions, then mirror the source’s distinction. Faithful translation of a chart is also translation of its data provenance.
A worked laboratory example from sample to bilingual summary
Imagine a constructed report with sample ID W-204, raw reservoir water, collected at 08:15 on 12 June, analysed by a defined culture method, E. coli result 24 MPN/100 mL, and turbidity 4.2 NTU. A clear translation states the site, time, two analytes, each method if specified, and each unit. It does not say 24 visible colonies were counted if the source used MPN, and it does not identify the material as drinking water after treatment if the source says raw reservoir water.
Now suppose a second sample, labelled ‘post-filtration’, shows turbidity 0.3 NTU but its E. coli analysis is pending. The translation must not use the lower turbidity to announce a negative E. coli result. Those are independent tests. The word ‘pending’ must survive even if a marketing summary would look tidier without it. A lower NTU suggests lower optical turbidity under comparable conditions, not a substitute for the unfinished microbiology result.
The point of this example is not to judge compliance or health risk; there is no jurisdiction or complete sample programme here. It demonstrates the proper boundary of translation: two samples, two distinct locations, separate measurements and an explicitly incomplete result. A target reader should understand exactly what has and has not yet been established.
Preparing a terminology matrix that laboratories can review
Create a terminology matrix with one row per analytical concept: parameter, organism or indicator, method, unit, sample basis, laboratory qualifier, source phrase, target rendering, reviewer and approval date. Mark CFU and MPN separately, even if both appear against the same organism. Add a note for the chosen translation of ‘not detected’, ‘estimated’, ‘presumptive’, ‘confirmed’, ‘too numerous to count’ and ‘below reporting limit’. Their meaning must remain stable across certificates, tables and summaries.
Use the scientific name as an identifier when appropriate. Translate common-language explanations for the intended audience, but do not change the analyte. Keep sample IDs, chain-of-custody codes and method numbers unchanged unless there is a documented reason. A regulator or later investigator may need to match the target report to the original laboratory record. An elegant paraphrase that destroys matching ability is a poor translation.
The laboratory should review domain-specific terms before release. Translators can ask excellent questions about units and logic without becoming responsible for method validation. Record unresolved ambiguities rather than silently filling them. Translation integrity depends as much on knowing what not to infer as on choosing the right words.
A realistic review sequence for a multilingual test report
Start with a table-only extraction of sample IDs, dates, analytes, results, units, methods and qualifiers. Compare it to the source. Next, translate headings, footnotes and the scope statement; these often change how numerical results should be read. Then translate narrative interpretation, explicitly checking whether every statement is an observation, inference, recommendation or compliance declaration. Each claim should be traceable to a source sentence or properly identified expert addition.
Perform a second pass for numerals, micro-symbols, degree symbols, subscripts, less-than signs, date formats and locale-specific decimal separators. Pay attention to column wrapping, because a denominator such as ‘/100 mL’ can be pushed onto the wrong line. Scan for conflicting labels between tables and charts. If one page calls a sample ‘untreated intake’ and another ‘finished water’, the discrepancy requires clarification.
Before publishing, obtain technical review for statements about water safety, legal compliance or corrective action. The translator should not decide whether a result is acceptable by consulting an unrelated threshold from a different water use. The final multilingual document should be useful precisely because it makes the original findings and limits visible.
Twelve mini-cases that reveal the usual errors
CFU/100 mL appears next to a source organism
Translate the expanded organism name or retain its accepted scientific form, preserve ‘colony-forming units’ and copy the per-100-mL denominator. Do not change this into a number of infected people, samples or illnesses. The count describes growth under a specific culture method and sample basis. If the method or sampling time is absent, the translation should not invent them. A source query is better than a confident but ungrounded explanation.
A water sample reads 6 MPN/100 mL
Keep most probable number as a statistical estimate. It is not proof that exactly six visible colonies were found. If the report supplies a confidence interval, carry it into the target language. Check whether the value belongs to E. coli, total coliforms or another target. A sentence saying simply ‘six germs’ loses the organism, method and denominator, which are the very facts a laboratory report is intended to preserve.
The report shows 1.8 NTU
Use turbidity, optical cloudiness or an approved technical equivalent for the target audience, and preserve the NTU symbol. Do not write 1.8 bacteria or 1.8 mg/L of solids. NTU is not a direct concentration of suspended solids. If a document compares two readings, check that their methods are comparable before treating the ratio as a scientific change in particle mass.
E. coli is marked ‘not detected in 100 mL’
The target should say that the specified method did not detect E. coli in the 100 mL sample. It should not certify that the entire water supply is sterile, nor should it turn the laboratory observation into a diagnosis. Preserve the sample location, method if stated, and reporting convention. An absence of detection has a defined evidentiary scope; it is not an assertion of universal absence.
A cell reads ‘less than 1.1 MPN/100 mL’
Preserve the relational operator. The laboratory is reporting a value below a specified numerical bound, not an exact measured result of 1.1. Do not simplify it to ‘negative’ without confirming that the laboratory uses that shorthand. Maintain the MPN method designation and denominator. Text extraction may drop less-than signs, so numerical proofreaders should compare the original page visually.
An appendix contains a field blank
A blank is usually a quality-control sample or control material, not an environmental water sample from a second public tap. Translate the quality-control designation and preserve its ID. If the blank shows contamination or another issue, do not hide it for visual consistency. The laboratory’s quality discussion may need to explain the effect on results; the translator should not make that assessment independently.
A chart says ‘total coliform positive’
Positive is a qualitative analytical finding for a specified target and method; it is not necessarily a numeric CFU or MPN concentration. Keep the distinction between a binary detection statement and a quantified result. Do not rewrite positive as ‘confirmed pathogenic E. coli’ because the target organism and interpretation are different. If the source has confirmatory testing pending, preserve that status.
A dashboard plots an alarm limit and a measured result
Translate alarm setpoint and observed measurement as separate labels. The setpoint may be operational rather than statutory. If the report does not specify a regulatory limit, do not identify the alarm line as one. The reader needs to know whether a line is a target, an instrument threshold or a legally defined maximum. The difference guides interpretation, even when the numerical values are identical.
Collection and receipt dates disagree by one day
Preserve both event labels and verify local date conventions before assuming an error. Transport, time zones and laboratory shifts may explain the gap. A translated form should not silently change a date to make the timeline appear simpler. If sample holding time affects the test’s validity, the laboratory must make that assessment. The translator can flag the issue without inventing a cause.
The summary claims ‘safe water’ after one NTU reading
If this wording appears only in a translator’s draft and not in the source, remove it. Turbidity does not independently establish all microbiological or chemical safety conditions. If the original makes a safety claim, preserve the claim’s source and context and request professional review if its basis is unclear. The translated output should not strengthen a single measurement into an all-purpose health assurance.
Two tables use CFU and MPN for the same organism
Do not merge the columns into a unified count. Different test designs may produce estimates or counts under different analytical processes. Translate each method label and preserve sampling dates and volumes. If a report compares the two, preserve the author’s comparison and uncertainty notes. A translation is not a statistical harmonization project unless the user explicitly commissions separate scientific analysis.
A result reads ‘TNTC’
Expand the term if the laboratory legend confirms it means ‘too numerous to count’, and retain the abbreviation when needed for auditability. Do not replace it with an arbitrary very large number. If the laboratory reports a qualified estimated concentration alongside the label, preserve both. The distinction between an upper measurement limit and a precise count is essential to the record.
Questions readers ask about translated water tests
Are CFU and MPN the same unit?
No. Colony-forming units describe counted growth events under a particular culture method, while most probable number expresses a statistical concentration estimate derived from test reactions. Both may be reported per 100 mL, but that shared denominator does not make the methods identical. Translate the label and method as well as the value.
Does zero CFU mean every bacterium is absent?
No universal absence follows from one tested sample. A report’s zero or no-colonies observation applies to the analysed sample, method, detection capability and target organisms. Other organisms or other locations may not have been tested. Preserve how the laboratory reported the finding rather than promising that a water system is sterile.
Does NTU tell me the number of microbes in a sample?
No. NTU is an optical turbidity measurement related to light scattering in water. It can indicate changes in clarity, but it cannot directly count a specific organism. Microbial parameters require their own appropriate tests. A translation must keep these categories separate even when both appear on a single water-quality dashboard.
Why do reports use CFU/100 mL rather than CFU/mL?
The chosen reporting volume depends on the analytical method, sample type, programme and expected concentration. A per-100-mL convention is common for various water microbiology indicators. It is a denominator, not a sample-quality rating. Altering it without proper conversion changes the numerical meaning.
Can I change ‘less than’ into ‘not detected’?
Not automatically. The laboratory may be expressing a numerical reporting threshold, while a ‘not detected’ label may reflect a formal method-specific conclusion. Preserve the operator or abbreviation given, consult the legend, and ask the laboratory when wording is ambiguous. Different qualifiers can imply different analytical evidence.
Are total coliform and E. coli interchangeable?
No. Total coliforms refer to a broader indicator group, while E. coli is a more specific organism used in many monitoring programmes. Do not translate one into the other simply because both appear in drinking-water discussions. Preserve the source analyte names and any confirmed or presumptive testing distinction.
Does a result below a limit always mean the water is safe?
No. The reported limit might be a laboratory detection limit, an operational action level or a regulatory criterion for one particular use and parameter. Safety also depends on the water’s intended use, other tested and untested conditions, sample representativeness and applicable rules. A translator should never infer an all-purpose safety declaration from one numerical comparison.
What does MPN confidence interval mean?
MPN is a statistical estimate. A confidence interval expresses uncertainty under the method’s statistical assumptions; it is not an exact range of organisms individually counted by hand. Translate the interval and stated confidence level if present. Omitting them can make the result appear far more precise than it is.
Should I translate the sample ID?
Usually preserve sample IDs and traceability codes as they appear. A translated explanation can describe where the sample came from, but the identifier should still match the laboratory’s chain-of-custody and original analytical certificate. Changing codes for readability can prevent audits and follow-up testing from connecting the correct records.
Why are sampling and analysis dates different?
Collection, transport, laboratory receipt, analysis and report issue are separate events. The source may record several dates because laboratory workflows take time. A target-language report should label them separately, preserve their formats accurately, and avoid merging them into a single ‘test date’ when that would conceal an important distinction.
Can a bilingual report insert a local drinking-water threshold?
Only as a clearly identified, separately validated interpretation requested by the client and checked for the applicable jurisdiction and water use. Do not silently replace the original limit or imply the lab tested under rules it did not use. The translation should remain an accurate record of the original certificate.
What is the most useful final QA test?
Select a result row and ask whether a reader can identify the sample, organism or parameter, test method, measured or estimated value, unit, qualifier, date and source of any compliance claim. Repeat for a microbiology row, an NTU row and a non-detect row. If those relationships survive, the report is much less likely to mislead.
A final evidence check for readers and publishers
Before sign-off, compare all tables against the source with special attention to CFU, MPN, NTU, per-volume denominators, less-than signs and italicized organism names. Verify that ‘negative’, ‘not detected’, ‘below reporting limit’ and ‘pending’ have not been merged. Read the laboratory qualification notes rather than translating only the prominent result box. Keep sample types separate and make clear whether a reported interpretation comes from the laboratory, an operator or an external authority.
Useful educational sources include the U.S. Geological Survey explanation of turbidity, USGS laboratory guidance on turbidity measurement and CDC guidance on Legionella test methods and result interpretation. They clarify technical concepts; they do not supersede a laboratory’s specific method report or the jurisdiction’s governing water standards.
The best translated water report is not the most confident sounding. It is the one that lets readers find the exact parameter, method, sample, unit and limitation behind every conclusion. Preserving uncertainty where the laboratory has uncertainty is a form of precision. It allows the source evidence to remain useful after the language changes.