Spatio-temporal Indoor Human Exposures in Homes Affected by Chemical-contaminated Soil and Groundwater

Abstract

INTRODUCTION: Public health problem statement: Vapour intrusion is a process that involves the migration of volatile chemicals from contaminated soil and/or groundwater into dwellings or other confined structures where inhalational exposure may occur. The process is exemplified by naturally occurring soil radon, which is considered a major risk factor for lung cancer in a number of countries. However, the extensive use of industrial chemicals and fuels over many years has left a legacy of soil and groundwater contamination potentially posing an even wider variety of disease endpoints. Volatile substances such as benzene and trichloroethylene (TCE) have also been shown to intrude into buildings from single or multiple sources off site. Limited environmental epidemiological studies have reported increased risks of cancer and non-cancer effects. Owing to the number of people potentially exposed including vulnerable subpopulations, regulatory agencies have developed frameworks for risk assessment based on environmental sampling and predictive models. These approaches, however, are based on idealized contaminant source and migration characteristics, and have been found to have limited predictive value for health risk assessment. Further research is required to provide more confidence in risk assessment outcomes. Initial literature review: In Australia, the federal enHealth’s human health risk assessment framework and a National Environment Protection Measure provide the basis for site contamination assessment. These documents, although recognizing the complexity of vapour intrusion, provide limited guidance on exposure assessment and have a focus on petroleum hydrocarbons. The human health risk assessment of vapour intrusion can be structured into three key areas: • Sub-surface fate and transport models and vapour measurement that establishes the vapour concentration at the building boundary. • Ventilation models and measurement which consider indoor air concentrations in space and over time within the building; and • Human inhalation dosimetry which considers absorbed doses over time. The peer-reviewed scientific literature on vapour intrusion over the past thirty years combined with the international regulatory documentation is extensive. However, the majority of this literature is oriented towards the initial phase of sub-surface transport to the building boundary. There has been limited focus on ventilation dynamics and less so on inhalation dosimetry. In the past five years, however, increasing attention has focused on spatial and temporal indoor air contaminant changes with one public health publication on residential indoor air spatio-temporal variability and another on linking indoor air contaminant concentrations to biological markers. In terms of considering inhalation dosimetry, however, there is as yet, no discourse on how these indoor environments may result in differing inhalation doses. This may be particularly important where high level peak doses due to environmental effects on the distribution of an indoor volatile, result in adverse pathologies. Indeed, there is some evidence, for example, in the case of TCE, for peak tissue concentrations precipitating acute neurotoxic effects. Gap in knowledge: The gaps in knowledge in vapour intrusion exposure assessment include the following: • Models of dynamic and time-dependent (non-steady state) vapour migration processes and their validation. • An understanding of spatio-temporal variability in indoor concentrations and the correlates of the variability that might lead to an evidenced-based indoor air sampling protocol. • The time dependence of absorbed dose, especially tissue concentrations, that results from time-dependent inhaled air concentrations. PURPOSE STATEMENT: Through a critical review of the literature and a series of empirical case studies, this research seeks to: • Elucidate the nature of spatial-temporal changes in indoor contaminant concentrations within houses affected by vapour intrusion and the factors that may influence those changes. • Provide an evidence base for a time-dependent vapour intrusion model with empirical evaluation, applicable to Australian conditions. • Explore the utility of biological monitoring for risk assessment in a common vapour intrusion scenario. GENERAL RESEARCH QUESTIONS: • What is the short- and long-term spatio-temporal variability of indoor air contaminants arising from vapour intrusion? • Which factors are significantly associated with indoor air concentration variability? • What is the relationship between biological monitoring data and indoor TCE concentrations? METHODS: A critical literature review and experimental case study approach were used. The experimental case studies were opportunistic and reflected real-life conditions. Case Study 1 was a termiticide treatment (including xylene) in a suspended floor home and Case Study 2 was a slab-on-ground house in a TCE-affected area. Critical literature review: Computerised searches of the published literature were conducted using the Web of Science, Scopus and PubMed. The logic grid included “vapour intrusion”; “ventilation”; “inhalation dosimetry” and “exposure”. The yields were complemented with author searching and forwards and backwards searching. The literature on vapour intrusion was critically reviewed in terms of its utility for human health risk assessment. Case Study 1 – Suspended timber floor home construction - Indoor air concentrations: The upper portion of the soil in the subfloor of a 1950’s home was treated with technical grade xylene containing m-, p- and o-xylenes as part of a termiticide treatment. Analyses were conducted of soil xylene and moisture concentrations; subfloor and indoor air xylene concentrations; and air exchange rates in the subfloor space and occupied space. Concurrent meteorological data were collected from a weather station. A published Australian non-steady state model, developed in previous national guidance, was used to estimate (and compare with) indoor air concentrations based on the empirical measurements. Case Study 2a – Concrete slab on ground home construction - Indoor air concentrations: A four-bedroom public housing property in a residential area impacted by chlorinated hydrocarbon contaminated groundwater was used over a period of 14 months to assess indoor air TCE levels. Passive TCE sampling occurred at five indoor and two outdoor locations over various time intervals. Air exchange rates were calculated at front and rear indoor sampling locations. Detailed local meteorological data were gathered from a weather station. Indoor temperature and indoor relative humidity were measured at 30 minute intervals over a 3-month period at each of five indoor air sampling locations. Soil vapour, sub-slab vapour and flux chamber measurements were carried out during one week concurrent with 6 hour passive sampling. Case Study 2b - Slab on ground home construction - Human exposure experiments: A biological monitoring pilot study was conducted with 5 volunteer adults who occupied the TCE-contaminated house for 12 hours. End-exhaled breath samples and blood samples were collected. Participants were also asked to provide urine samples at baseline, at the end of the exposure period and on three subsequent occasions. Passive indoor air sampling and surface flux testing was undertaken. Sub-slab TCE samples were also collected inside and outside the house. RESULTS: Critical literature review: Papers on vapour intrusion mainly focused on issues associated with the sub surface. These included areas such as development of one- and three-dimensional steady-state models; estimation of attenuation factors; lateral exclusion distances; factors affecting subsurface migration such as moisture levels and oxygen concentrations and reconsideration of the United States (US) Environment Protection Agency vapour intrusion database. There have, however, been some new areas of focus in the last five years, which have included the use of new real-time measurement techniques; an increased focus of the role of pressure differences on indoor air contaminant concentrations; seasonal and diurnal differences and spatio-temporal variability in homes across an affected community. One recent study examined indoor air TCE concentrations and blood TCE levels. The recent literature has increasingly examined the above-ground and indoor environment but has not further considered within building spatial differences nor a more detailed examination of short-term indoor air average concentration changes and associated influencing variables. In addition, the literature is silent on the issue of inhalation dosimetry in vapour intrusion and the potential for non-invasive methods of biological monitoring. Case Study 1– Suspended floor home construction - Indoor air concentrations: Xylene air concentrations decayed to non-detectable levels within two weeks. Subfloor xylene air concentrations were greater than living space xylene air concentrations, and the decay of the concentrations following a generally consistent pattern. Air exchange rates between the sub-floor and living space differed by up to an order of magnitude and demonstrated the influence of subfloor ventilation on vapour intrusion. Statistically significant associations were found for air exchange in the sub-floor space and locally measured minimum and average wind speed. Site-specific variables in a non-steady state model showed general consistency with measured data, but the modelling estimated a greater shorter-term initial peak with more rapid decay of xylene concentrations than those measured. Case Study 2a - Slab on ground home construction - Indoor air concentrations: Air sampling data revealed spatial and seasonal indoor TCE variations. Winter month results were up to an order of magnitude greater than summer months. Monitoring over 6-hour (h) periods demonstrated the occurrence of diurnal peaks that were not evident with a 24-h sampling regime. Moreover, the use of a continuous data logging instrument showed occasional spikes over rapid time intervals which were an order of magnitude or greater compared to the common baseline value. Air exchange measurements revealed consistent early morning declines in ventilation. Correspondingly, the highest surface TCE flux was noted during the day with the lowest occurring during the evening. Soil vapour measurements at progressive depths at the rear of the property showed high source concentrations of TCE with lower concentrations progressively up to the subslab. Using the 6-h average TCE concentration as the outcome variable, it was found that ventilation, internal temperature, barometric pressure and wind direction were significant predictor variables in a multivariate model. Ventilation had the greatest impact in the best fit model with one air change per hour predicting a 4.4 μg m⁻³ decline in the indoor TCE concentration. Assessment of model predictions showed close agreement with the dataset. Case Study 2b - Slab on ground home construction - Human exposure experiments: The pilot biological monitoring exercise yielded mixed results with most biomonitoring data below the limit of reporting (LOR) which was which was <5μg m⁻³ for breath and <0.01 μg L⁻¹ for blood. End-exhaled breath TCE concentrations were generally below 5μg m⁻³ with two results above the LOR. Composite end-exhaled breath samples for baseline and at 02:30 and 08:30 were 2.0, 1.5 and 1.2μg m⁻³ respectively. Blood concentrations were all below the level of reporting of 0.01 μg L⁻¹. While blood TCE concentrations could not be quantified in accordance with standard protocols, discrete peaks were observed on the chromatograms. CONCLUSIONS AND RECOMMENDATIONS: The recent literature has increasingly examined the above-ground and indoor environment but has not further considered within-building spatial differences nor shorttermindoor air average concentration changes and their influencing variables. In addition, the literature is silent on the issue of inhalation dosimetry in vapour intrusion and the potential for non-invasive methods of biological monitoring. Case Study 1 confirmed the influence of dwelling features and that of ventilation and meteorological variables such as wind speed for a suspended-floor dwelling. Case Study 2 captured greater resolution across all measurements and although the extent of the variables measured varied, sufficient data were captured to provide a more detailed examination of time-dependent change. Statistically significant spatial differences were observed suggesting the need to account for prevailing wind direction in worst case indoor sampling strategies. Mixed-effects regression models were consistent with the observed seasonal and diurnal differences. The two case studies provide evidence for a worst-case sampling strategy, that is, sampling in winter and during the evening and accounting for spatial variance. Overall, the results demonstrate the complexity of indoor ventilation dynamics and that spatial and temporal influences are important to understand for exposure assessment purposes. Short term, peak TCE exposure periods were observed and may be of toxicological significance based on information suggesting TCE exhibits a nonmonotonic dose-response relationship for foetal malformations. On the basis of the research the following recommendations are made: • More detailed and extensive (>1 year) longitudinal studies capturing time dependent changes in indoor air concentrations and all influencing variables including air pressure changes, should be undertaken. • A human volunteer biological monitoring study using end-exhaled breath and blood TCE analyses should be undertaken, using sensitive analytical techniques such as Selected Ion Flow Tube Mass Spectrometry. • A retrospective epidemiological study in TCE-affected areas should be conducted in Adelaide.