It has been said that a professional is someone who has specialist knowledge, skills and expertise that has been independently validated and who operates under a code of practice.  However, another take on this is a professional is someone wise – and brave – enough to say that something is outside their area of competence – and then works with someone who has that missing competence.

This article is written as a conversation between an analytical chemist with deep expertise in how accredited laboratories operate and a risk assessor who advises on the need for remediation under specific legal contexts.

Modern analytical chemists are operating at the critical interface between complex environmental matrices and the generation of legally and scientifically defensible data. Within laboratories accredited to standards such as BS EN ISO/IEC 17025:2017, their role extends beyond the execution of methods to the critical evaluation of method applicability to the substances and media of interest, measurement uncertainty, detection capability, and data fitness for purpose. This includes ensuring appropriate sub-sampling strategies, preservation techniques, extraction efficiencies, calibration models, and quality control systems are in place and functioning. For emerging contaminants such as PFAS, the analytical chemist must also recognise the limitations of current methodologies, including compound coverage, matrix effects, and potential artefacts, and communicate these transparently to data users – sadly this is not usually the case.

The GB-based risk assessor works usually under the context of the planning regime or Part 2A of the Environmental Protection Act 1990 (as amended). Both are chemical agnostic. In the context of protecting human health, remediation is needed to ensure land is safe and suitable for its intended use or to remove a significant possibility of significant harm.

Turning to that large and very diverse family of synthetic chemicals, PFAS pose a specific challenge to both risk assessor and analytical chemist. Never mind that they may be persistent and widespread, the real challenges arise from what Greenleaves III calls epistemic uncertainty – the lack of relevant knowledge about the behaviour of PFAS in soil, water and within our bodies.

Analytical chemistry’s epistemic uncertainty stems from the inherent limitations in our ability to comprehensively measure the large, structurally diverse of PFAS family using finite, targeted methods. Current PFAS analyses typically quantify a relatively small subset of known substances, while thousands of other PFAS, particularly precursors and transformation products, remain undetected or only partially characterised. This gives rise to uncertainty associated with:

• Scope of analysis: Targeted methods exclude unknown, novel, or poorly characterised PFAS, leading to systematic underestimation of total burden.
• Transformation and precursors: Many PFAS exist as precursors that can degrade into terminal compounds over time; standard analyses may not capture this latent mass without specialised approaches (e.g. oxidative conversion techniques).
• Matrix effects and recovery: Complex environmental media (soil, sediments, biota) can influence extraction efficiency, affecting quantification accuracy and comparability.
• Analytical variability and uncertainty: Measurement uncertainty, particularly at low concentrations near reporting limits, can be significant and is often underappreciated in downstream decision-making.
• Standard availability and calibration: Quantification relies on the availability of authentic standards, which do not exist for many PFAS, constraining both identification and accuracy.
• Cross-laboratory comparability (PT Scheme): Differences in methods, reporting limits, and compound lists can lead to variability between laboratories, complicating the interpretation of datasets.

Taken together, these factors mean that PFAS analytical data are not absolute representations of environmental concentration, but conditional estimates bounded by method capability. The “numbers alone” therefore risk misinterpretation unless accompanied by a clear understanding of what has, and has not, been measured, and with what degree of confidence.

The risk assessor has to inform decisions on the need for remediation that accounts for epistemic uncertainty in, for example, the dose-response relationship, effects of mixtures, rate of transformation of precursor to terminal substances in the environment, behaviour of different species (neutral or ionised).

A better understanding of these two sources of epistemic uncertainty will help refine the decisions on whether or not to remediate even as science develops a better understanding to reduce those uncertainties.

On analytical uncertainty

What does the method detection limit (MDL) mean?
The MDL represents the lowest concentration of a substance that can be distinguished from analytical noise with a defined level of statistical confidence, rather than a threshold of quantification or risk relevance. It is method, matrix, and laboratory-specific, and is influenced by sample preparation, instrument sensitivity, and background contamination. For PFAS, MDL can vary significantly across compounds and matrices and may fall above or below levels of toxicological concern. Consequently, a “non-detect” does not equate to absence, but rather indicates that any presence lies below the method’s capability to reliably quantify it.

How does inter-laboratory comparability help?
Inter-laboratory comparability, typically assessed through proficiency testing (PT) schemes and inter-laboratory studies, provides an external benchmark of analytical performance. It allows laboratories to evaluate their bias, precision, and consistency relative to peers using similar or different methodologies. For PFAS, where method variability and compound coverage differ from laboratory to laboratory, such comparisons help identify systematic discrepancies. However, comparability does not guarantee accuracy; it demonstrates consistency within the bounds of current methodological limitations.

Are Harmonised analytical protocols possible?
Harmonised protocols are both desirable and partially achievable, particularly for well-characterised PFAS in relatively simple matrices such as drinking water. Standardised methods (e.g. based on LC-MS/MS) can improve comparability and regulatory confidence. However, full harmonisation remains constrained by the diversity of PFAS chemistries, the absence of standards for many PFAS, and the variability of environmental matrices. As such, harmonisation is likely to remain partial and evolving, with a combination of standard methods and fit-for-purpose adaptations required.

Are quality assurance/quality control (QA/QC) procedures at the same level of stringency in different laboratories?
While accredited laboratories operate under common frameworks (e.g. BS EN ISO/IEC 17025:2017), the implementation and rigor of QA/QC procedures can vary in practice. Differences may arise in areas such as blank control, use of isotopically labelled standards, calibration strategies, acceptance criteria, and treatment of data near detection limits. For PFAS, where ultra-trace analysis and contamination control are critical, even small differences in QA/QC practice can materially affect results. Accreditation provides a baseline of competence, but not complete uniformity.

In this context, the United Kingdom Accreditation Service (UKAS) is actively working towards greater standardisation of PFAS methodologies. However, the process of extending a laboratory’s accredited scope to include PFAS analysis requires a high level of technical rigor, including extensive method validation, demonstration of measurement uncertainty, contamination control, and ongoing performance verification. This reflects both the analytical complexity of PFAS and the need to ensure that reported data are robust, reproducible, and fit for regulatory purposes.

How do the above influence whether nor not reported concentrations are a reliable and fit basis for regulatory decision-making?
These factors collectively determine the confidence that can be placed in reported concentrations. Data reliability is not solely a function of the numerical concentration reported on the certificate of analysis,  but of the supporting analytical context. Laboratories should therefore be prepared to report, on request, key quality indicators alongside results, including measurement uncertainty, recovery of surrogate standards, system suitability performance, blank contamination status, and other relevant QA/QC criteria.

For regulatory decision-making, particularly where threshold values are comparable to the MDL and consequences significant, transparency in these parameters is essential. Reported concentrations should be interpreted in light of method scope, detection capability, and demonstrated analytical performance. Decisions should therefore be based on a weight of evidence, incorporating these quality metrics and, where appropriate, complementary analytical lines of evidence, rather than relying uncritically on single numerical values.

On risk assessment

Should risk assessors consider the presence and potential transformation of precursor substances?
Yes, risk assessors should explicitly consider the presence and potential transformation of precursor substances when evaluating PFAS contamination, because failure to do so can lead to a systematic underestimation of both current and future risk. Many PFAS precursors are not captured by standard targeted analytical methods, yet they can undergo abiotic and biotic transformation in soils, groundwater, and biota to form persistent terminal compounds such as perfluoroalkyl acids, which are often subject to regulatory criteria. This means that a site with relatively low concentrations of measured PFAS may still represent a significant long-term source of contamination as precursors tranform over time. Incorporating this dynamic requires moving beyond static concentration data toward a more process-informed conceptual site model that accounts for transformation pathways, rates, and environmental conditions. In practice, this may involve the use of complementary analytical techniques such as the TOP assay to estimate precursor potential, alongside cautious interpretation of targeted data, and the application of total organic fluorine (TOF) measurements to provide an estimate of the overall fluorinated organic burden, including unidentified or unquantified PFAS. Where TOF significantly exceeds the sum of organic fluorine in targeted analytes, it provides a clear indication of unaccounted for missing mass and highlights the potential for both unknown compounds and precursor-derived contributions to risk. However, both the TOP assay and TOF approaches are subject to important limitations: the TOP assay may not fully oxidise all precursors, while TOF lacks compound specificity and may include non-PFAS organofluorine contributions, with results also influenced by extraction efficiency. TOP Assay is also affected by the intensity of the oxidation step.  Recognising precursor presence and the broader organic fluorine mass balance, while understanding the constraints of these techniques, aligns with a weight-of-evidence approach and is essential for ensuring that risk characterisation remains protective under conditions of analytical uncertainty.

What role do complementary analytical approaches (e.g. TOP assay, total organic fluorine)?
Analytical approaches such as the Total Oxidisable Precursor (TOP) assay and total organic fluorine (e.g. EOF/AOF) could help contextualise concentrations of targeted PFAS by addressing key sources of epistemic uncertainty inherent in targeted analyses. While targeted LC-MS/MS analysis provides robust, compound-specific quantification for a limited suite of well-characterised PFAS, it systematically underrepresents the total burden due to the exclusion of precursors and other unknown compounds. The TOP assay partially resolves this by forcibly converting oxidisable precursors into terminal perfluoroalkyl acids that can be measured, thereby revealing the latent PFAS mass that may tranform to terminal PFAS over time in environmental systems. In parallel, total organic fluorine approaches quantify the aggregate fluorinated organic content irrespective of molecular identity, enabling a mass balance

How could mass balance evaluations improve understanding of the total PFAS burden and long-term leaching risks?
Atoms can neither be created nor destroyed in a chemical reaction – the total number of fluorine atoms remains the same. Mass balance involves counting the number of F atoms in a sample, discerning which ones are bonded to carbon – organic fluorine – and accounting for those that are part of targeted analysis. The remainder are part of molecular structures that are not able to be analysed using targeted methods.  An understanding of how much of the organic fluorine load is present in identified molecular structures can be an important line of evidence in risk assessment and even more so in risk management.

Conclusions
Regulatory decisions about the significance of PFAS in water, soil and other media should be based on a broader appreciation of analytical results than simply the concentrations reported on Certificates of Analysis.

The need to consider PFAS beyond those that can be subject to targeted analysis requires methods such as TOP assay and organic fluorine analysis to establish the potential for precursor transformation into terminal PFAS.

Increasingly, high resolution mass spectrometry could bring more PFAS into the analytical spotlight and reduce the need for untargeted TOP assay and organic fluorine analyses – but that is a subject for a future article.

References
BS EN ISO/IEC 17025:2017 General requirements for the competence of testing and calibration laboratories (Currently Under Review)

NATHANAIL, C P, WILLIAMS, G and NATHANAIL, J F (2024) Good practice guidance: some per- and polyfluoroalkyl substances (PFAS) in soil and the water environment, C819, CIRIA, London, UK (ISBN: 978-0-86017-965-8)

Article by Ken Scally (Normec and Calibre Scientific) and Paul Nathanail (LQM)