The Application of Human Health Risk Assessment to Mustard Contaminated Sites

Principal Investigators
Matthew H Ashmore, C Paul Nathanail

Funding

NERC

Project overview

Risk based land management involves the use of Generic or Site Specific Assessment Criteria to determine the degree of risk contaminants pose and therefore the need or otherwise for remediation. There is a lack of Assessment Criteria for Sulphur Mustard and its environmental breakdown products which can inform risk assessments or remediation targets for contaminated soil. The redevelopment of land formerly used for the production, filling, storage or disposal of chemical weapons containing Mustard is hampered by this. Any Human Health Risk Assessment for contaminated soils requires the four stages of Hazard Identification, Hazard Evaluation, Risk Estimation and Risk Evaluation. A critical review of the literature on the toxicity, fate and transport properties of Mustard Gas, Thiodiglycol, 1,4-Dithiane and 1,4-Thioxane which may be used to produce UK policy compliant Generic or Site-Specific Assessment Criteria has been carried out. Incorporation of Health Criteria Values and Fate and Transport Data into Risk Assessment Models in order to produce Generic Assessment Criteria for contaminated land for Standard Land Uses is demonstrated. Particular problems associated with the selection of Health Criteria Values and Fate and Transport Data are reported, and implications of the Environmental Chemistry of Sulphur Mustard to the Risk Assessment and Remediation of Contaminated Soils is discussed.

Introduction
The UK employs a risk-based approach to the assessment of contaminated land in which the exposure of potential receptors to the contaminant in question, via all relevant pathways, is compared with some target value in order to assess the potential risk. Risk Assessment is a tiered process, whereby the results of a Generic Quantitative Risk Assessment (GQRA) is initially carried out using conservative assumptions about receptor behaviour and site specific properties in order to produce an Generic Acceptance Criteria (GAC), i.e. the concentration of contaminant in the soil which would produce no /minimal ill effects. Should this be exceeded, a Detailed Quantitative Risk Assessment (DQRA) may be carried out using site-specific values or more realistic assessments of receptor behaviour in order to produce a less conservative Site-Specific Acceptance Criteria. Exceeding a SSAC then a Risk Management strategy and Remediation of the contamination may be called for.
The Contaminated Land Exposure Assessment (CLEA UK) model is the Environment Agency’s semi-probabilistic tool for assessing human exposure from soil contaminants and may be used for a number of generic exposure scenarios to produce GACs, or site data can be used to produce SSAC.
The present paper discusses production of GAC values for Mustard Gas and several breakdown products using data from the literature in order to illustrate soil contamination issues at Mustard Gas contaminated sites, including the adequacy and uncertainties presented by existing knowledge and data.

UK Contaminated Land Policy
In the UK, land affected by contaminants is regulated under two separate regimes. Primarily it is considered during the planning process of development, and it is required that the developer of a site show that site is fit the intended purpose. The second regime concerns the management of specific threats posed by historically contaminated land and, under Part IIA of the Environmental Protection Act 1990, the relevant local authority must deem land as ‘contaminated land’ if, by reason of substances in, on or under the land, that, for human receptors, significant harm is being caused or the is a significant possibility of significant harm (commonly termed SPOSH) being caused during its current use, though further threats to e.g. Groundwater would also be relevant.

Health Criteria Values (HCV)
Health Criteria values are toxicological benchmarks against which human exposure to contamination is compared and under the UK contaminated land regime, two types of HCV are defined; TDIs and Index Doses.
Due to its proven genotoxic effects, it is considered that Mustard Gas does not have a threshold of toxicity, i.e. theoretically one molecule could induce a tumour or mutation, and so an Index Dose is derived for chronic effects, representing a minimal level of risk, with the additional requirement to keep any intake as low as reasonably practicable (ALARP). Slope Factors presented in USACHPPM 2000 of 7.7(mg/kg-bw/day)-1 for oral exposure and 300(mg/kg-bw/day)-1 for inhalation exposure mean that the Index Dose for Mustard Gas at the 1x10-5 tumour incidence level (consistent with several other Index Dose assessments) is given in Table 1. No formal health criteria have been produced for TDG or 1,4-Dithiane and so the oral health criteria has been used in its absence.

TDG and 1,4-Dithiane are threshold contaminants, that is there is a level below which no adverse effects are induced, and this leads to the notion that below the threshold the concentration is tolerable and thus a Tolerable Daily Intake (TDI) is considered. Reference Doses (RfD) calculated for these compounds by [Reddy et al (2005)] and [IRIS] are essentially TDIs and have been used here in this context.

Generic Site Scenarios
Two of the CLEA UK generic scenarios used; Residential where no vegetables are grown and Commercial/Industrial have been used to produce GAC values for discussion purposes. These contain general assumptions of the critical receptors such as body weights, inhalation rates and duration of exposure. In addition, particular site characteristics are assumed, such as the type of soil present (sandy) and a low proportion of organic matter in the soil, to give a conservative over-estimate of exposure to the contaminant.
Exposure of the Critical Receptor via different routes is summed where appropriate to give an Average Daily Exposure and the soil concentration where ADE/HCV=1 g the GAC.

Comparison of derived GAC with soil concentrations
The derived GAC for different scenarios are presented in table 1. There are no reliable figures for background exposure for the three breakdown products and so a very high background is assumed (>TDI). Under UK policy, the TDSI under such conditions is considered to be 0.2TDI.

Comparing the GAC with soil analysis results from around a Mustard burning pit gives some initial indications of some possible implications. The sampling was carried out in three phases, an initial surface sampling exercise and two further intrusive sampling rounds which provided samples down to approximately 1.2m and 2.0m respectively. Few samples contained TDG and the concentrations were far lower than the GAC.
The phase 1 analysis results gave the highest concentrations of the three contaminants and also highlighted a common error in the analysis of contaminated land, i.e. the limit of detection of the analysis was greater than the retrospectively calculated GAC values and so all the samples must be assumed to have failed these criteria, though around half were less than limit of detection.
The intrusive phases 2 and 3 had more appropriate detection limits, and the exceedances were fewer and of a lesser magnitude.

Commercial² represents an initial stage of a site-specific assessment, by replacing the generic ‘sandy’ soil with a ‘clay’ soil, and specifying no buildings on-site.
Overall, it is observed that the process of making the assessment more site-specific means that fewer samples exceed the GAC and future work may reduce this further. By far the greatest exceedences occur in the surface samples from phase 1 and, while these obviously present the greatest risk, it also implies a relatively small amount of soil requires initial treatment.
It must be pointed out that compliance with a GAC indicates that the concentration presents an acceptable risk and thus for example, for the commercial land use, it represents an acceptable risk level for an adult spending his entire working life on the site and so a judgement as to whether the site presents a serious possibility of serious harm is needed before remediation would be required under Part IIA due to risk to human health.
In terms of acute toxicity, the UK has no formal mechanism as yet, however on the strength of the highest recorded concentration of mustard (~30mg/kg) and the most conservative soil loading estimate of 1mg soil cm-2 skin area from CLR10, this equates to a concentration of 0.003µg H/cm2, which compares favourably with 0.02mg which OPCW reports as the lowest observed blister forming amount of agent, though sensitization may be an issue in the chronic assessment.