Asbestos in soil and quantitative risk assessment – discussion on a way forward?

- by
Tags: Featured

Article contributed by:

Jo Wilding, Associate Technical Director, RSK (SoBRA Executive Committee)

Simon Cole,  Technical Practice Lead, AECOM (SoBRA Chair)

Feedback to

Laboratory reporting of asbestos in soil is a key data component of quantitative risk assessment for human health, and for the risk-based management of asbestos in soils. Asbestos in soil poses a hazard to human health when it becomes airborne, and the inhaled fibres can result in diseases including mesothelioma and lung cancer. Therefore the relationship between asbestos in soil and the resulting fibre concentration in air is a fundamental, but complex, element of human health risk assessment. In the UK, there is no current consensus on which air quality guidelines should be used to assess potential risks to human health from exposure to asbestos in soils once airborne. This in turn means that there is no UK regulatory or industry-agreed good practice for the assessment of risks from asbestos in soils, which are being – or could be – released to air and subsequently inhaled. Key elements to an asbestos risk assessment are (i) asbestos identification and characterisation, (ii) receptor exposure, (iii) an understanding of background exposure and (iv) assessment criteria at an appropriate risk level that distinguish between relative risks from different types and forms of asbestos.

The Society of Brownfield Risk Assessment (SoBRA) has recently published two discussion papers relating to asbestos in soil laboratory analysis and reporting. These were published alongside an update to the discussion paper on guidelines for airborne concentrations of asbestos fibres in ambient air: implications for risk assessment.  All three papers build on the series of SoBRA papers published by the asbestos sub-group in 2015 that looked at sampling protocols, activity-based sampling, conceptual models, and example decision making under Part 2A of the Environmental Protection Act 1990.

Data mining of laboratory data

SoBRA issued a request to major UK soil laboratories in 2016 to provide SoBRA with anonymised asbestos in soil laboratory data. The distribution of asbestos in soil discussion paper provides a factual presentation of the data shared by five UK laboratories (ALS, DETS, Envirolab, i2, and REC). The data presented in the paper is, naturally, a reflection of the soil samples sent to the testing laboratories rather than a reflection of background concentrations of asbestos in soils in the UK.  The datasets do not distinguish between sample origin – be that a greenfield site or a brownfield site nor do they distinguish between large datasets from one or a small number of sites and small datasets from a larger number of sites.

Typically laboratory testing follows a sequential three stage process akin to the one presented in the Standing Committee of Analysts ‘Blue Book’ method (withdrawn in October 2020; SCA, 2017). However, the data presented in the paper is a result of different laboratory methods or sub-sets of methods and will include analysis undertaken by different methods over time by the same laboratory.  The three stages are:

Stage 1: The determination and identification of presence or absence of asbestos using stereomicroscopy, plus higher magnification polarised light microscopy (PLM) analysis for fine fibres (see HSG 248, HSE, 2005).

Stage 2: The removal of asbestos containing material (ACM) and fibre bundles with identification and gravimetric analysis to determine percentage by weight.

Stage 3: The dispersion and collection of free fibres followed by fibre identification, counting and measurement of fibres to determine percentage by weight.

From a dataset of approximately 175,000 samples, during Stage 1, asbestos was not detected in the majority of samples submitted to the five laboratories. When positively identified the majority of asbestos detected was chrysotile. The majority of reported concentrations of free fibres detected in soils that have undergone Stage 3 analysis following a positive identification at Stage 1 were below the method reporting limit of 0.001% wt/wt. (Note that these samples are typically, but not always, those that have had a positive identification at Stage 1; it is rare for samples with a negative ID at Stage 1 to progress to Stages 2 or 3).

Variability in UK laboratory methods and reporting

There is, however, significant variability in UK laboratory methods for the identification and quantification of asbestos in soil. A SoBRA survey of 10 UK laboratories in 2018 raises questions about the data that is reported and its subsequent applicability for use in human health risk assessment.  The survey was designed to complement a similar survey of UK laboratories undertaken by the AGS and reported in February 2019 (Mitcheson, 2019).  The survey highlighted that laboratories do not follow the same sample preparation and analytical procedures, nor do they report in the same way.  Only 70% of laboratories follow the ‘Blue Book’ method; the method used by the remaining 30% of laboratories was not explored.  Even those who follow the ’Blue Book’ implement that method in different ways that could significantly influence the results reported.

When considering accreditation and proficiency, which is a key element for data quality, all laboratories surveyed participated in the HSL Proficiency Testing Scheme Asbestos in Soil Scheme (AISS) but only 70% of the laboratories responded that they held UKAS accreditation for Stage 1 identification of asbestos.

Total sample size (weight) requested, sample size (weight) used in Stage 1, and sample preparation for Stage 1 and Stage 3, varied considerably between laboratories. No laboratory routinely reported at what sub-stage within Stage 1 asbestos was detected (i.e. during visual inspection using stereomicroscopy or during higher magnification PLM microscopy) and therefore when the sample inspection stopped. All such variables could have implications for the determination and identification of presence / absence of asbestos in the soil samples, and the type of asbestos reported to be present in Stage 1.

With regards to the data needed to inform a human health risk assessment, only 20% of laboratory responses indicated that they would routinely provide comment on the condition of the asbestos identified in Stage 1 (i.e. weathered, degraded, non-degraded, disaggregated, not in original form) in addition to identifying the presence or absence of the three principal types of asbestos, and the form of that asbestos. A lines of evidence approach is often needed for asbestos risk assessment. For example, non-degraded asbestos material is that which despite being in or on the ground is in relatively good condition and capable of retaining most of the asbestos fibres. As such, non-degraded asbestos material may pose a lower risk to human health and should be considered within the risk assessment. Asbestos containing materials that have been significantly damaged or degraded will be more likely to release fibres and therefore pose a greater risk to human health.

A small number of laboratories did respond that they could provide asbestos type differentiation at quantification if required during Stage 2 and Stage 3, along with photographic evidence if requested, but this was not routine.

Reporting recommendations

The SoBRA paper on UK laboratory methods for the identification and quantification of asbestos in soil makes a series of recommendations for laboratory reporting of asbestos in soil results based on the requirement for human health risk assessment to be supported by clear, unambiguous laboratory data. One key element is the reporting of types and forms of asbestos, for all each type and form of asbestos identified. Within Stage 1 this should be for each of visual inspection, detailed inspection under x20-x40 stereomicroscope and pinch samples under x80-x500 magnification (PLM). Within Stage 2 individual masses for each type and form of asbestos identified should be reported, with mass reported in mg/kg not %wt/wt.

The SoBRA recommendations (the full detail of which is presented in the SoBRA paper) for laboratory reporting of asbestos in soil results would ensure that the data provided by the laboratory is clear and unambiguous. With the withdrawal of the of the SCA ‘Blue Book’ method for the determination of asbestos in soil, now is an opportune time for industry to adopt a new approach the analysis and reporting of asbestos in soil. Consistency is required in sample preparation, analytical procedures and reporting to ensure the resulting human health risk assessment can be undertaken with confidence.

When considering the data required within a human health risk assessment, there are other laboratory test methods that focus less on the reporting of the mass of asbestos present in the sample and focus more on the potential for airborne asbestos fibres arising from asbestos in soil. Two such methods that have been developed, or are in the process of being developed, by UK laboratories are:

  • Respirable fibre count per unit weight of sample (this can be used to estimate airborne fibre numbers in association with airborne soil particles), and
  • Dustiness tests (modified HSL drum tests designed to estimate the releasibility of asbestos fibres and provide normalised fibre to dust concentrations akin to those reported by Addison et al 1988).

Air quality guidelines for use in human health risk assessment

The SoBRA discussion paper on guidelines for airborne concentrations of asbestos fibres in ambient: implications for risk assessment was initially published in 2017. The paper has been prepared as an evidence base, with the aim of supporting the development of good practice for assessment of potential risks from asbestos at sites affected by land contamination.

The paper sets out a series of issues that need to be resolved before a UK air quality guideline value can be proposed for asbestos, but calls for the Asbestos in Soil Joint Industry Working Group to formulate a position regarding an air quality guideline for asbestos in the UK.

There are a range of existing air quality guidelines for asbestos fibres, provided both by international bodies (e.g. World Health Organisation) and national bodies (e.g. Health Council of the Netherlands). There are also different approaches in literature for calculating air quality guidelines for asbestos fibres, dependent on the exposure scenario under consideration. The SoBRA paper summarises a selection of internationally published air quality guidelines, together with the data upon which these are based, as well as calculating air quality guidelines using two different modelling approaches from literature. The existing air quality guidelines, and SoBRA calculated guidelines, are compared alongside published ambient background airborne concentrations to understand variability in thresholds for asbestos in air and the potential practicalities of those guidelines.

All models are having to extrapolate from relatively high occupational exposures that form the empirical evidence on which the models are based down to much lower environmental exposures.  The authors of these models differ in their interpretation of that empirical data and as a consequence there is variance in the model outputs for a given input exposure.  Risk estimates from these models can vary by an order of magnitude, and published air quality guidelines by more than two orders of magnitude due to differing assumptions on the relative potency of different asbestos types.

The updated paper includes risk estimates calculated using SoBRA’s workbook (beta version freely available at using the linear as well as the non-linear version of the Hodgson & Darnton model (Hodgson & Darnton, 2000) as well as updates on the age adjustment calculations. The paper recommends that the linear version of the Hodgson & Darnton model for pleural mesothelioma is used to estimate risk and calculate air guideline values in conjunction with the non-linear variants for peritoneal mesothelioma and lung cancer.

The tool allows users to select from the algorithm options presented in Hodgson & Darnton (2000), and enables users to perform model choice sensitivity analysis and evaluate the difference the use of alternative algorithms makes to estimated risk.  It is hoped that the tool will provide a consistent basis for the calculation and reporting of risk estimates and feedback on the tool is welcomed at


Addition, J, Davies LST, Robertson, A, Wiley, RJ (1988). The release of dispersed asbestos fibres from soil. IOM Historical Research Report TM/88/14.

HSE (2005). Asbestos: The analysts’ guide for sampling, analysis and clearance procedures.

Mitcheson, B (2019). Variability in asbestos analysis in soil, AGS Magazine, February 2019

SCA (2017). The quantification of asbestos in soil (2017), Methods for the examination of waters and associated materials, Standing Committee of Analysts, April 2017 [Withdrawn in October 2020].


The authors would like to acknowledge the work of the SoBRA asbestos subgroup, and in particular Barry Mitcheson for the development of the workbook tool, and both Barry and Simon Hay as primary authors of the risk calculations presented in the SoBRA air quality guidelines discussion paper.