Groundwater PFAS laboratory results are easier to read when three questions are kept separate: which chemicals were specifically measured, how confidently the laboratory recovered and identified them, and what the numbers can actually tell us about the well. A report expressed in ng/L is not a photograph of every fluorinated substance in an aquifer. It is a set of measurements for named target analytes produced by a stated procedure, with reporting limits and quality-control evidence that shape every interpretation.
The practical starting point is the method line. EPA Method 1633A addresses a defined group of forty per- and polyfluoroalkyl substances across certain aqueous and other environmental matrices using liquid chromatography with tandem mass spectrometry, often abbreviated LC–MS/MS. Drinking-water methods such as EPA 533 and 537.1 have different lists and sample-preparation requirements. Check the actual method revision, sample matrix, analyte list and project requirements rather than assuming every PFAS report is comparable.
A defensible reading joins the result table to the field log, chain of custody, instrument and extraction quality controls, laboratory narrative and any applicable site-specific decision criterion. A numerical result without those companions is a useful clue, not an entire conclusion. The aim of this guide is to translate unfamiliar laboratory terminology into an audit-ready sequence that a student, consultant, landowner or technically interested reader can reason through independently.
Begin with the question the sample was designed to answer
Suppose a groundwater monitoring well is sampled downstream of a former industrial storage area. The reader may ask whether named PFAS were detected, whether concentrations changed from the previous quarter, whether different wells display the same mixture, or whether a particular regulatory condition was exceeded. These sound related, but they require different evidence. Presence needs credible identification and blank assessment. A trend needs comparable methods and uncertainty through time. A source inference requires hydrogeology, spatial context and alternatives. Compliance needs the correct rule, jurisdiction, matrix and averaging protocol.
Before opening the concentration columns, write one sentence describing the decision. For example: “Determine whether measured PFOA and PFOS in monitoring well MW-4 changed materially between two sampling events using comparable procedures.” That sentence excludes many seductive distractions. It reminds you that a combined PFAS total is not automatically interchangeable with individual compounds and that a result from a river or household tap has a different sampling purpose than raw aquifer water.
Understand what targeted LC–MS/MS can and cannot see
Liquid chromatography separates chemicals by how they interact with a stationary phase and flowing solvent. Tandem mass spectrometry selects ions or ion transitions that help identify and quantify selected compounds. In targeted analysis the instrument is instructed to follow known analytes with standards, transition ratios and retention-time expectations. Its selectivity can be impressive, but it does not certify that all organic fluorine in the sample has been inventoried.
EPA explicitly distinguishes targeted analysis from non-targeted approaches. A targeted method answers whether compounds on its list can be quantified under its validation conditions. A result marked “not detected” for one analyte cannot mean that the sample contains no other PFAS, precursor, transformation product or unmeasured fluorinated material. It also cannot guarantee absence below the attainable detection or reporting limit. When an investigation seeks a broader inventory, the analytical plan—not wishful interpretation—must change.
Locate the matrix, method revision and laboratory scope
The expression “EPA method” is only the beginning of a defensible analytical description. Ask whether the submitted material was raw groundwater, finished drinking water, porewater, leachate or soil extract. Matrix effects differ, and the chosen method must be validated for the relevant application. A laboratory’s internal standard operating procedure can be based on a published method yet contain documented adaptations; readers should look for those deviations and what validation supports them.
For non-potable groundwater, a laboratory may cite Method 1633A or another fit-for-purpose method. For drinking water, EPA 533 and 537.1 are different established methods with different analyte coverage. It would be a mistake to say that the larger number in a method name proves superior performance. The decision is about permitted use, the selected compounds, the concentration range, sample handling and the evidence available. Write the method revision beside every dataset before combining results.
Why nanograms per litre feel small but still need exact units
A nanogram is one billionth of a gram. One ng/L is one microgram per cubic metre, but not one milligram per litre. Confusing prefixes changes the reported concentration by enormous factors. For a hypothetical measurement of 8.0 ng/L in a well, the concentration is 0.008 micrograms per litre and 0.000008 milligrams per litre. These are mathematical conversions, not independent analytical findings.
Volume matters equally when explaining extraction. If a 250 mL groundwater aliquot genuinely contains 4.0 ng/L of a compound, the aliquot contains approximately 1.0 ng before extraction: 4.0 ng/L multiplied by 0.250 L. That illustration tells you how little analyte a method may need to track. It does not establish the minimum mass any particular instrument can identify, since extraction concentration factors, recoveries, sample matrices and calibrations determine the workable range.
Sampling is part of the analytical result
The water in a bottle should plausibly represent the water at the sampling point, not material introduced by equipment, dust, hands, containers or an adjacent workflow. PFAS projects often adopt carefully specified field materials and contamination-control plans. The relevant test is not whether a material is fashionable or universally forbidden; it is whether the project has evidence that the sampling system can add the target compounds and what controls were used to expose that possibility.
Record well identification, screen interval, sampling date and time, purge or low-flow approach where applicable, field parameters, containers, cooler condition and laboratory receipt. Compare the same sampling depth and hydrologic setting between rounds. A PFAS sample collected after a drilling or maintenance intervention may tell a different story than a stable low-disturbance sample. Documentation of the disturbance enables careful inference; lack of documentation does not convert the disturbance into proof of pollution.
Make field blanks answer a precise question
A field blank helps test whether contamination was introduced during a specified part of sampling and handling. An equipment blank may focus on sampling apparatus, while a trip blank or other blank has a different design purpose depending on the project and method. Their names are not interchangeable substitutes. The blank’s relevance depends on how and when it was prepared, the contact surfaces and which analytical compounds it covers.
Imagine that well MW-4 has a reported PFOA concentration of 2.1 ng/L and an associated field blank has 0.7 ng/L. A tempting calculation is to subtract 0.7 and announce a true groundwater concentration of 1.4 ng/L. That would overstate what the evidence permits. The blank does not demonstrate that a fixed fraction of the blank contamination entered the field sample. Instead, apply the project’s data-validation rules, qualifiers and follow-up sampling logic. Blank contamination may weaken a low-level result without yielding a reliable arithmetic correction.
Understand the laboratory method blank
A method blank travels through the laboratory preparation and analysis workflow without intentionally receiving field contamination. It helps identify contamination originating in solvents, equipment, reagents, glassware, extraction materials or instrument carryover. A clean field blank with a contaminated method blank can signal a laboratory-side problem; the reverse pattern raises different questions. Neither pattern automatically identifies a single culprit.
For a low-concentration sample, compare the blank result against the sample’s concentration and the method’s established qualification criteria. If a blank has a signal close to a measured sample, the sample finding may need qualification or reanalysis. If the sample is far above a trace background, the significance can differ. Do not invent a universal blank-to-sample ratio: laboratory protocols and data-quality objectives decide how potential blank contamination is assessed, qualified and reported.
Isotope-labelled compounds are process witnesses, not magical erasers
In isotope dilution, a known amount of an isotopically labelled analogue is introduced so its behaviour can help account for extraction losses and signal response. Labelled compounds are chemically similar to their native target but distinguishable by mass. By following the labelled material through the analytical process, the laboratory gains evidence about recovery and instrument behaviour that a native-result number alone cannot supply.
This technique does not erase every source of error. Matrix suppression, poor calibration, contaminated extraction materials, an incorrectly selected analogue or handling anomalies still matter. Some analytes are quantified using their own labelled analogue; others use a designated extracted internal standard. When you review quality control, ask which compound was associated with which labelled standard and which recovery acceptance interval applies. A footnote stating “isotope dilution used” does not substitute for the actual recovery table.
Work through a recovery calculation without confusing it with the report
In a simplified teaching example, a sample receives 5.0 ng of labelled reference material and the analytical accounting identifies 3.7 ng of that material through the process. Apparent recovery is 3.7 divided by 5.0, or 74%. This is a numerical illustration; actual laboratories infer recovered amounts from calibrated response ratios and apply method-specific corrections. A value of 74% might be acceptable for one analyte under one method criterion and unacceptable under another.
Do not multiply the laboratory’s already corrected native PFAS concentration by 100 divided by 74 as a second “fix.” That can double-correct the measurement. Instead, establish whether the reported sample value already incorporates isotope dilution, whether the labelled recovery passed the chosen acceptance criteria and whether the result carries a qualifier. The meaningful question is whether the method’s prescribed quantification and quality checks were followed, not whether a reader can improve the number with an improvised factor.
Read calibration as evidence, not decoration
A standard curve relates known concentrations or amounts to the instrument response under specified analytical conditions. Quality evaluation includes the calibration range, initial standards, continuing checks, retention times and the expected mass-spectrometric signals. A beautiful line on a graph is insufficient when the sample signal lies outside the validated range or matrix interference disrupts identification. Some laboratories dilute or re-extract samples, and such changes should be recorded.
Suppose a calibrated extract range is illustrated as 0.5–20 arbitrary concentration units while a sample’s response corresponds to 31 units. Simply accepting 31 as a precise result extrapolates beyond the demonstrated range. Appropriate validated dilution, reanalysis or qualification is needed. Conversely, a signal below the lowest verified reporting concentration does not become quantitative merely because software prints three decimal places. Precision in typography is not evidence of measurement precision.
Learn the difference between detection and quantification
Detection limits, reporting limits, quantitation limits and laboratory practical limits can refer to distinct concepts. A detection decision addresses whether a signal differs sufficiently from noise or blank response; a reporting or quantitation level addresses the confidence to provide a concentration under a method’s conditions. The exact terms must be read from the report legend and laboratory accreditation or project protocol. Never assume that “ND,” “U” and a less-than sign all represent identical definitions.
If one well reports “<2.0 ng/L” and another reports 1.4 ng/L from a method whose reporting level is 1.0 ng/L, a statement that the first well is cleaner is not justified. The first value is left-censored at its relevant limit; its underlying concentration might be zero or near 2.0 ng/L. A defensible comparison notes differing censoring thresholds and, where necessary, uses a statistical approach designed for censored environmental measurements.
Qualifiers belong to the result
Letters such as J, U, B or E can mean estimated result, not detected, possible blank association or calibration exceedance in many reporting conventions, but meanings vary across laboratories and projects. Read the actual legend. A result of “3.2 J ng/L” should not be reproduced as “3.2 ng/L” without retaining the estimation qualifier. The qualifier is part of the data, not a visual inconvenience.
One useful interpretation table contains sample ID, date, method, analyte, reported value, unit, reporting limit, qualifier, relevant blank finding and data validator’s disposition. That table prevents a polished narrative from accidentally stripping uncertainty away. When comparing wells, separate robust detections, qualified detections and nondetects. A map that uses the same colour for all three can imply a spatial pattern the underlying analytical evidence does not support.
Linear and branched isomers complicate ordinary-looking sums
Certain PFAS, including PFOS, can appear as different structural isomers. Chromatographic behaviour can separate or partially resolve their signals. A laboratory may report a target analyte using a defined integration protocol for linear and branched components, depending on the method. Therefore a number labelled PFOS total, a linear-only result and an individual branched-isomer peak are not automatically exchangeable without checking the method’s naming and integration rules.
For a purely illustrative example, suppose the method requires adding two appropriately identified components of 6.2 and 1.8 ng/L. The combined value is 8.0 ng/L. The arithmetic is simple, but the acceptance of both peaks as the correct analyte is a chemical and validation decision. Avoid adding a published total to its components a second time. Ask whether the lab report already includes isomer contributions and whether the relevant comparison criterion specifies the same definition.
A sum of forty analytes is not a sum of all PFAS
A targeted forty-compound list measures a selected subset of a very broad chemical class. A report’s “sum of PFAS” might mean the sum of detected substances on that list, the sum of a smaller regulatory group, or a project-specific combination. Two reports can print the same phrase while measuring different sets. Always write down the names included before comparing a sum over time or across laboratories.
Consider a hypothetical table showing PFOA 4 ng/L, PFOS 3 ng/L and PFHxS 2 ng/L, with all other listed targets absent or reported below their respective limits. The sum of these three quantified detections is 9 ng/L. That does not demonstrate a total PFAS concentration of 9 ng/L in the water. Nor does it tell you how to replace nondetects when constructing a broader sum; substituting zero, half the reporting limit or the limit itself involves assumptions that must be declared.
Matrix suppression changes how a number should be trusted
Groundwater can contain salts, dissolved organic matter, iron, particles and other constituents that accompany extraction and influence ion formation in electrospray mass spectrometry. A weakened signal may be a true low concentration, a recovery issue or ion suppression. Labelled standards and quality-control samples help distinguish these possibilities. Matrix spikes and duplicates are particularly informative when the actual water composition differs from clean laboratory reagent water.
If a laboratory reports a matrix-spike recovery outside acceptance limits, the correct next question is which analytes are affected and what the data qualifier means. It does not justify declaring every analyte present or absent incorrectly. A high-salinity well may require a validated dilution or different preparation. Report any modified reporting limits, because a diluted sample might be less sensitive to trace contamination even when the instrument response becomes easier to quantify.
Investigate carryover and analytical sequence
High-concentration samples can leave residues within preparation equipment or the instrument path. If a low-level sample follows a strongly contaminated sample, analysts use blanks and sequence controls to examine carryover. The sample list and analytical sequence can therefore matter as much as a single result page. However, temporal adjacency alone is not proof that carryover occurred; the appropriate blank and instrument checks must support that interpretation.
Suppose a run processes a high sample, then a method blank, then a low sample. A contaminated blank suggests the low sample warrants scrutiny. A clean blank and other satisfactory checks increase confidence, though they do not eliminate every possible source of error. When a low-level trend turns on one isolated result, ask whether the laboratory reran the sample, how carryover was investigated and whether the stored extract remained within appropriate holding and stability constraints.
A duplicate has more than one possible meaning
Field duplicates assess combined variation introduced by sampling, handling and laboratory analysis, subject to how independently the two samples were obtained. Laboratory duplicates assess a narrower preparation or instrument component. Neither should be treated as an identical experiment. For two positive measurements, relative percent difference is often computed as absolute difference divided by their mean, times 100. A project may have separate rules near reporting limits.
Take duplicate results of 7 and 9 ng/L. Their mean is 8 ng/L, the absolute difference is 2 ng/L and the relative percent difference is 25%. The arithmetic does not tell you whether the pair passes quality control, because that decision requires the project’s criterion and concentration context. Near trace limits, relative statistics can look unstable even for small absolute differences. Preserve both values, the calculation and the method-specific decision rather than reducing them to an unsupported “good” or “bad.”
Do not infer a chemical source from a single concentration
PFAS are associated with diverse historical uses and transport pathways. One measured compound or ratio rarely identifies a source uniquely. Even a recognizable mixture may evolve through selective movement, precursor transformation, varying recharge and changes in the sampled interval. Hydrogeology supplies the missing spatial structure: upgradient and downgradient positions, hydraulic gradients, screened depths, recharge boundaries and plausible alternative activities.
A convincing source interpretation compares multiple lines of evidence. Are concentrations coherent with groundwater flow? Do repeated well samples show a persistent pattern? Do background wells and relevant source-area records support the hypothesis? Could sampling equipment have contributed a low-level blank signal? An impressive-looking PFAS fingerprint without these checks should be called a hypothesis. Labelling it as established attribution risks confusing chemical resemblance with causation.
Trend analysis requires comparable measurement conditions
Imagine quarterly PFOA results of 5.0, 4.4, 3.9 and 3.5 ng/L. The apparent sequence is declining. Before making that conclusion, compare the reporting limits, laboratory methods, qualifiers, blank findings, well pumping conditions and sampling season. The arithmetic decline from first to last is 1.5 ng/L, or 30% of the first value, but its environmental significance depends on uncertainty and whether samples represent comparable water.
Short sequences do not establish a long-term exponential attenuation process. Seasonal recharge can dilute or shift source contributions, and changes in well chemistry can affect recoveries. If a site decision turns on a trend, consider replicate sampling, data validation and a time-series method appropriate to censoring and autocorrelation. An honest chart displays both concentrations and limits, and marks changes in methods rather than smoothing them out of the story.
Distinguish laboratory results from regulatory conclusions
Regulatory comparisons require a specific governing jurisdiction, medium, contaminant definition and applicable date. A groundwater-monitoring concentration may be relevant to plume assessment but not directly comparable with a limit framed for treated drinking water without specifying why. Legal standards and guidance values can change, and some rules apply to individual PFAS while others address combinations. The responsible conclusion names the exact benchmark and its applicability.
A useful report sentence says: “At sampling event S3, the laboratory reported PFOA at 4.0 ng/L in raw groundwater by the stated method; interpretation against any applicable site criterion should consider this result’s qualifiers and the rule’s scope.” That is much stronger than “the groundwater is safe” or “unsafe” stated from one number. For health decisions, obtain current local authority guidance and accredited analytical advice rather than treating an educational example as a site-specific determination.
Work a complete miniature PFAS data validation case
Construct a teaching dataset with two wells. MW-A has PFOA 8.4 ng/L, PFOS 5.1 ng/L, a method blank reported below 0.5 ng/L, and label recoveries that meet the laboratory’s documented criteria. MW-B has PFOA 1.2 J ng/L and PFOS reported below 2.0 ng/L; its field blank contains PFOA at 0.8 ng/L. The numbers are fictional and chosen to illustrate reasoning, not observations from any real site.
MW-A provides two unqualified measurements for its two target analytes, subject to the complete data package. MW-B offers weaker evidence for low-level PFOA because the result is estimated and a relevant blank is positive. It also has no quantitatively detected PFOS above the stated limit. Comparing the sum of both analytes between the wells would obscure these asymmetries. A better decision is to document MW-B’s blank concern, check validation rules, consider targeted resampling and leave source attribution unresolved.
When a reported concentration is inconsistent with field evidence
Suppose a low-concentration well unexpectedly shows an order-of-magnitude increase immediately after maintenance. One explanation is genuine mobilization; other explanations include a sampling disturbance, a lab artefact, changed sample representativeness or a previously missed event. The proper response is not to delete the result as an outlier. Preserve it, flag the anomaly, inspect field records and QC, and formulate tests that could discriminate between candidate explanations.
A useful resampling design retains comparable methods, captures field blanks, observes well stabilization and records hydraulic conditions. If independent sampling reproduces the increase while QC remains sound, confidence in a true change grows. If the increase disappears and the original blank or carryover record was suspect, the interpretation changes. The lesson is methodological humility: one laboratory number starts a question, whereas corroborated and well-documented evidence can help answer it.
Reporting limits may move between sampling rounds
Method revisions, sample dilution, extraction efficiency or matrix interference can alter quantitation capabilities. The apparent sequence “ND, ND, 1.5 ng/L” cannot be interpreted as emergence when the earlier limits were 5 ng/L and the newer limit was 1 ng/L. A concentration near 1.5 ng/L could have been present throughout but undetectable in the earlier measurements. This is a change in analytical visibility, not necessarily a new contaminant pulse.
When preparing a trend plot, include a second series for each sampling event’s reporting limit or show nondetect arrows at their respective limits. Document any method switch and consider reanalysis of archived material only when preservation and analytical suitability can be demonstrated. Avoid coding all nondetects as zero. That common shortcut creates artificial step changes at exactly the times the laboratory’s sensitivity changes.
Isotope dilution versus a simple internal-standard approach
The conceptual distinction is that isotope dilution follows an isotopically labelled analyte analogue added early enough to encounter the important preparation losses, while a late-added instrumental standard primarily monitors analytical response. Both can be useful; they guard against different parts of the workflow. Exact terminology differs by method, so consult the laboratory’s quantification table and the primary method rather than relying on a casual label.
This matters for interpreting a failed labelled recovery. If the relevant extraction surrogate is anomalously low, uncertainty may involve extraction or matrix effects. A stable late internal standard alone would not rule out those earlier losses. Conversely, a failed instrument check suggests a problem even if preparation behaved consistently. Translate the QC evidence into the stage of measurement it tests. That stage-based diagnosis makes a technical appendix intelligible without replacing its acceptance criteria with guesses.
Questions worth asking the laboratory
If a result has a surprising value, seek the exact original chromatographic identification and quantification basis, not merely a rewritten certificate. Ask whether the analyte was within the validated calibration range; which native and labelled ion transitions were used; whether retention time and qualifier checks passed; and whether any laboratory, equipment or field blank contained that analyte. Ask for the extraction batch identifier and a coherent explanation of any qualifiers.
Also ask if the reported value already includes all required correction factors, which isomers were included in an analyte total, whether a dilution raised the reporting limit, and what uncertainty information is available. These are normal scientific questions, not accusations. A well-supported laboratory answer often resolves apparent contradictions that arise when an otherwise careful reader treats a summary results table as an exhaustive analytical history.
Transfer problem: compare a blank to a low result
Practice with a fictional report. A sample has PFOS 1.7 ng/L with an estimated qualifier. A method blank shows PFOS 0.9 ng/L; a different well reports 15.0 ng/L with an otherwise clean QC record. Is the first measurement definitely 0.8 ng/L after blank subtraction? No. The method blank provides evidence of potential contamination but not a sample-specific subtractable amount. Determine the appropriate qualification and whether corrective work or resampling is warranted.
Is the second measurement necessarily free from all uncertainty? No. It appears less vulnerable to a small blank background, but sampling representativeness, extraction recovery and instrumental checks still matter. If a project applies a documented multiple-of-blank threshold, apply that specific criterion, not an invented one. This exercise teaches why two concentrations with the same physical unit can have markedly different levels of evidential support.
Transfer problem: compare sums with non-detects
Two fictional laboratories report the same well. Laboratory A measures three named PFAS at 2, 3 and 4 ng/L, with all other target results below their limits. Laboratory B reports four named PFAS at 2, 3, 4 and 1 ng/L, using a more sensitive method for the fourth analyte. A’s detected three-compound sum is 9 ng/L. B’s detected four-compound sum is 10 ng/L, but this does not prove that total PFAS increased. The analyte definitions and detection capabilities differ.
To compare fairly, calculate the common three-compound subset using compatible procedures: both produce 9 ng/L in this simplified example. The fourth compound is additional information rather than evidence of a temporal increase. Now ask whether matrix and sample date were also comparable. Even matching sums can hide different chemical patterns, so retain compound-level values. The independent transfer lesson is to keep definitions stable while interpreting change.
Transfer problem: convert a small concentration into mass
Assume groundwater contains 12 ng/L of a particular target PFAS and a hypothetical volume of 500 litres is removed. Multiplication gives 6,000 ng, which equals 6 micrograms of that chemical in the removed water. This computation describes a mass in a known volume at an assumed concentration. It is not the total contaminant mass in the aquifer, because the volume of water in the aquifer, spatial concentration distribution and connected phases remain unknown.
Now consider 12 ng/L measured only in a 0.25-litre laboratory aliquot. That aliquot corresponds to 3 ng of the target analyte before extraction. The chemistry of extraction, recovery and analysis still must be handled by the validated method. This comparison teaches the importance of scale: a concentration can describe a tiny aliquot accurately while revealing little about regional contaminant loading without a hydrogeological model.
Reader questions about PFAS laboratory evidence
Does “not detected” prove zero PFAS? No. It means the specified target was not reported as detected under the stated method and limits. Can a method blank be ignored because its concentration is small? Not without comparing the relevant sample concentrations and the validation rules. Does isotope dilution guarantee an exact result? No. It is a powerful compensation and monitoring strategy within a broader quality system.
Can one PFAS ratio identify a specific industrial facility? Usually not by itself: transport, mixtures and alternative sources need examination. Is the sum of forty measured compounds the entire chemical class? No. Can an old drinking-water method be treated as identical to a groundwater method? Not without reviewing the intended matrix and analyte coverage. These answers follow one principle: a defensible result must keep the measurement’s identity, scope and uncertainty attached to the number.
A practical final review before citing the data
Before using PFAS results in a public explanation, check the sampling location and depth, date, chain of custody, matrix, method and revision, compound nomenclature, linear/branched reporting conventions, units, reporting limits, qualifiers, blank findings, isotope-labelled recoveries, calibration and any dilution factors. Then match the result to a specific question: presence, trend, spatial contrast, mechanism or criterion comparison. Do not pretend that all five questions are answered by one analytical certificate.
This habit of preserving meaning is transferable to any technical vocabulary. A scientific word such as “detected” is not a loose synonym for “present at any concentration”; “estimated” is not another word for “wrong”; and “isotope dilution” names a measurement strategy, not certainty. Readers exploring precise language may find the eduKate guide to translation and meaning across languages useful alongside the guide on why vocabulary changes thinking. For a neighbouring geochemical problem, see the independently published chromium speciation and analytical preservation guide.
Precursors demand an additional qualification. Some PFAS-related compounds can transform into other measurable substances as environmental and chemical conditions change. Therefore, a decline in one named analyte accompanied by a rise in another does not establish degradation of all PFAS mass. Transformation may redistribute which molecules appear on a target list. A defensible investigation compares the same analytical scope over time, considers which precursors can be measured, and resists replacing an inventory of known compounds with an assertion about the whole fluorinated chemical pool. Laboratory reports sometimes contain special analyses or oxidation assays that ask related but different questions; their results should be labelled by the method they actually performed. Even when the numbers look like ordinary ng/L concentrations, the analyte definitions and chemical transformations can make the comparisons substantially different.
Field timelines are another independent line of quality evidence. Record the exact sampling time, preservation and bottle details, transport dispatch, laboratory receipt, extraction date and analytical run date. The applicable method or project contract defines storage and holding requirements. A chromatogram produced after an unexplained delay may require a qualification that a comparable promptly handled sample does not. This does not mean degradation necessarily occurred: stability depends on the compound, container, preservative, temperature and matrix. It means the analyst must establish that the timeline meets the agreed method. A practical way to read a certificate is to place the sample’s chain-of-custody timeline next to the extraction batch and ask whether any stated holding-time exception applies specifically to the compounds being interpreted.
Read geochemistry alongside PFAS only when it answers a mechanism. If two wells have very different dissolved organic carbon, ionic strength or redox conditions, that information might help explain different transport and sorption environments. It cannot be used to predict exact PFAS concentrations from a single indicator. Hydrogeological gradients also matter: the nearest well is not necessarily the well receiving the same groundwater path. A conceptual site model should identify stratigraphy, screens, gradients, recharge, possible preferential pathways and historical releases before drawing arrows between concentrations. The aim is an explanation whose assumptions are exposed, not a colourful map that looks more precise than the available hydraulic data.
For a further worked check, imagine two sampling rounds from the same well. The first reported three PFAS at 4.0, 3.0 and 2.0 ng/L; the second reported 3.0, 2.0 and 1.0 ng/L for the identical three compounds. Their respective detected sums are 9.0 and 6.0 ng/L, a 3.0 ng/L decrease or one third of the original sum. The reader can compute that arithmetic without knowing how the water changed. To claim a real downward environmental trend, however, confirm identical analyte definitions, comparable reporting limits, similar well conditions, method QC and an uncertainty assessment. If one round also measured additional PFAS, keep those new analytes separate when evaluating the common-set trend. State the date and the exact set rather than a vague statement that ‘PFAS fell by 33 percent’.
Finally, communicate the uncertainty without burying the decision. A technically credible one-paragraph finding should name the sample, matrix, date, method, target chemical, numeric result, unit, reporting level if pertinent, any qualifier and the narrow inference supported. Follow it with the outstanding question and what evidence would answer it. For example, a blank-qualified low-level PFOS detection suggests a need to review contamination controls before using it as a plume-boundary anchor. A repeatable, high-confidence concentration pattern across multiple flow-aligned wells motivates a different investigation. The distinction between these interpretations is not caution for its own sake: it is the difference between making a decision from measured evidence and allowing a visually neat number to impersonate certainty.
Primary methodological sources
US EPA: PFAS analytical methods development and sampling research. US EPA: drinking-water research methods, including 533, 537.1 and 1633. US EPA: Method 1633A, December 2024. These references should be consulted for exact method-specific criteria, not replaced by educational examples.