The thousands of abandoned, inactive, and operating mines present serious potential risks to human health and the environment. The extraction and processing of metals, non-metal minerals and coal, result in the generation of large quantities of contaminated solid wastes, wastewater and air emissions which cause significant environmental impacts, including groundwater and surface water contamination, soil erosion and soil contamination.

The legacies of past uranium mining and milling activities continue to be a cause of concern and require assessment and remedial action. Over the next several decades, a large investment will have to be committed to clean up sites contaminated with hazardous waste and petroleum products. This commitment will result in a continuing demand for site remediation services and technologies, which implies more research needs.

The special risks associated with uranium mining in the past were mostly addressed to workers mines. Over the past decades, these concerns for worker health and safety have been the driving force for gradually tighter controls and improved work and safety practices. Only in relatively recent times, however, impacts on public health and on the wider natural environment, related to uranium mining, have been received the deserved attention. In the case of the natural environment, these concerns include the risk of environmental degradation, contamination and ecosystem biodiversity, aesthetics, access to land and future land beneficial use.

The radionuclides released into the environment can give rise to human exposure by the transport through the atmosphere, aquatic systems or through soil sub-compartments. A multi-compartment model has been developed to predict activity concentrations in predefined end-points. The research work has been mainly studying and modelling the transport and fate of the radionuclides in the environment for the assessment of the exposure to the uranium tailings.

The important routes of contamination have been identified as a basis for determining the modelling approach and the radionuclides of major concern were identified in each one of the compartments based on their chemical, physical and radiological properties. Each compartment represents an environmental media: air, soil, water, vegetation and biota. A radionuclide transfer model through the food chain is also incorporated.

When modelling contaminant release and transport mechanisms for each environmental compartment, the major output chosen is often the contaminant concentration in each exposition point selected (for instance, breathing air zone, superficial soil and well water). This will allow the assessment of doses, if an additional exposure scenery is created. The objective is to quantify the potential exposure levels of the hazard at the receptor location answering to these main questions: what is the contaminant concentration in the exposition point? How much of a contaminant do people inhale or ingest during a specific period of time? Is the situation acceptable?

In the first model the radionuclide activity in the tailings was estimated assuming initial equilibrium in the uranium series ²³⁸U and ²³⁵U for two different situations: the tailings resulting from the treatment of an ore with an average grade of 1 kg/ton, being leached at 90% and the tailings resulting from the treatment of an ore with an average grade of 0,2 kg/ton, being leached at 100%. This will gives us the radium content in the tailings necessary for simulating the radon release from the tailings to the breathing air zone, which is an input for the atmospheric transport.

The undisturbed uranium ores contain low concentrations of the decay daughters of uranium and thorium radionuclides in equilibrium with the parental materials. In the uranium mineral processing some of the radionuclides are not dissolved during the leaching process, which breaks down the equilibrium chains of the uranium decay, remaining in the tailings at their original activities. The expressions used to estimate the activity for each one of the radionuclides in the tailings results from the successive radioactive decays for the two uranium decay series, the uranium series ²³⁸U (4n + 2) and the actinium series ²³⁵U (4n + 3).

The principal radon isotope, ²²²Rn, formed from the ²²⁶Ra radioactive decay, is an inert gas, which emanates from the solid tailings particles and is free to diffuse to the surface of the pile, escaping to the atmosphere where it may be transported by the wind into the surrounding area, dispersing the potential damages.

The basic equations of diffusion were used for estimating the theoretical values of the radon flux from the ²²⁶Ra content in the waste material. The fundamentals of the conceptual model are based in the principles of diffusion across a porous medium, which allows the mathematical description of the radon transport through the waste and the cover.

Radon migration to the surface is a complex process controlled mainly by porosity, and moisture, leading the cover efficiency in attenuating the radon flux. This efficiency depends on the capacity of the cover material for keeping the diffusion so slow that radon decays to another non-gaseous nuclide, becoming trapped by the cover system. The algorithm incorporates the radon attenuation originated by an arbitrary cover system placed over the radioactive waste disposal. As an alternative, it can be estimated the thickness of the cover that allows a radon flux inferior to the acceptable one.

The contaminant concentration released is estimated by a box model formulation which has implicit a mass balance formulation. The box volume is defined by its length, width and the mixing height. As a consequence of a steady state assumption, the pollutant concentration is constant in time and the mass flow rate entering into the box is equal to the flow rate leaving the box.

In the atmospheric transport a two-dimensional model is used for calculating the flux diffusion from the radioactive waste disposal, having as result the hazard concentration at a defined distance from the soil (the breathing or mixing height) which will be the starting point for the dispersion which can be considered either simultaneously in each wind direction or only in the prevailing wind direction. The atmospheric dispersion is modelled by a modified Gaussian plume equation which estimates the average dispersion of the contaminants released from the source in each wind direction. The plume dispersion model accounts for the gaseous contaminant transport from the source area to a downwind receptor. The atmospheric transport is done at wind-speed to a sampling position located at surface elevation and transverse horizontal distance from the plume.

For the hydrologic transport a two-direction model is proposed for simulating the contaminants release from the waste disposal and its migration process through the soil to the groundwater. The final result is the contaminant concentration in the groundwater as function of the elapsed time, at a defined distance from the waste disposal, generally the location where the exposition point is considered being represented by an hypothetical well.

A leaching model based on a sorption-desorption process is used for describing the contaminant release from the waste disposal. The leachate concentration is determined by a distribution or a partition coefficient which describes the relative transport speed of the contaminant to the water existing in the pores; soil properties such as bulk density, and water content; the extent of contamination, described by the contaminated zone thickness, area, and the amount of contaminant in the source.

The transport for the dissolved contaminants is considered to occur either in the vertical direction through the unsaturated zone until an aquifer is reached either in the horizontal direction, through the saturated zone flowing to an hypothetical well, where the contaminants become accessible to humans or other forms of life. The vertical flow is considered to be one-dimensional. It is assumed that there is retardation during the vertical transport that is estimated assuming that the adsorption-desorption process can be represented by a linear isotherm.

Movement and fate of radionuclides in groundwater follow the transport components represented by the basic diffusion/dispersion–advection equation with radioactive decay and retardation for the radionuclide transport in the groundwater.

The radionuclides initially considered in the model simulation were uranium, thorium, radium, polonium and lead but only the results obtained for uranium and radium were included in the model exploration. For the other radionuclides considered, the results were in accordance with the expected based on the assumption that because of the slow rate of contamination migration only the radionuclides with relatively long half-life are of importance in the transport process. In addition, data on uranium and radium activity in groundwater and superficial waters from the Urgeiriça site were used in the model simulation comparing the model results with the measured data.

The hydrologic model also quantifies the superficial runoff and its transport to the superficial waters. The activity present in this environmental compartment, superficial water, may be a potential pathway to food chain contamination, either by the direct use of streams waters or indirectly through the contaminated aquatic biota.

As a result, the hydrologic model estimates the radionuclide activity concentration in each exposition point, namely in the underground waters (represented by a well) and in the superficial waters (represented by a streamlet).

Modelling the radionuclide transfer to vegetation has implicit several processes which describes and quantify mathematically the radionuclides transfer mechanisms, transport, absorption and translocation to vegetation. The main goal is to develop a radionuclide transfer model through food chain by the ingestion of contaminated vegetation. This pathway can be quite significant because of the biological concentration in the foodstuffs.

Vegetation may be contaminated through direct deposition, root uptake or irrigation with contaminated water. The material resuspension from superficial soil may occur due to the wind, rain or mechanical factors, with later deposition onto vegetation surface.

A model was developed to describe each one of these contamination pathways: root uptake, deposition and resuspension, either from deposition either from irrigation with contaminated water from a well. Radionuclides evolved in the transfer processes depends on the contamination route, deposition or root uptake.

The different contamination processes were combined in a global model for simulate the radionuclide transfer and estimate the vegetation activity resulting from each one of the contamination processes. The final output is the total concentration in the vegetation combining the internal contamination with the external contamination.

The model is rather complex as it is necessary to understand and transcribe to the conceptual model the interactions between the contaminants (radionuclides or heavy metals) and the soil components, vegetation, as well as the interactions between the contaminants itself.

The conceptual model is based on the assumption that each one of the transfer processes may have either origin in soil either in the air. In the first case, the processes involved are deposition (characterized by the deposition velocity), interception (described by the interception factor) and retention (described by the weathering half-live). In the second case, the radionuclide behaviour in soil and its mobilization reflects the radionuclides physics and chemical properties, soil properties, the type of vegetation and local hydrology and geology characteristics.

The next model will complete the previous one by simulating the radionuclide transfer through the food chain resulting from the cattle ingestion of contaminated vegetation.

Contamination of the trophic chain by radionuclides released into the environment will be a component of human exposure by transferring the radionuclides into animal products that are part of the human diet. This can occur by first ingestion of contaminated pasture by animals and then by ingestion of animal products contaminated (dairy or meat). The relevant incorporation of the radionuclides into cow’s milk, for instance, is usually due to the ingestion of contaminated pasture. This transfer process is often called the pasture-cow-milk exposure route.

A compartment dynamic model was developed to describe mathematically the radionuclide behaviour in the pasture-cow-milk exposure route and predict the activity concentration in each sub-compartment. The dynamic model is defined by a system of linear differential equations with constant coefficients describing the mass balance in different compartments taking in account the fluxes in and out of the compartment and the radionuclides decay. For each compartment, a transient mass balance equation defines the relations between the inner transformations and the input and output fluxes. The fluxes between the compartments are estimated with a transfer rate proportional to the amount of the radionuclide in the compartment. The concentration within each compartment is then transcribed to doses values based on a simplified exposure pathway and a pre-defined critical group.

The first model considered for the transfer through the food chain is relatively simple and classic and considers as initial state a contaminated pasture that is consumed by a cow that produces a certain quantity of milk. The transfer coefficients for soil and pasture compartments are expressed as function of soil characteristics and ecological parameters. A more sophisticated model is also described taking into account the spread of the radionuclide within the cow by including the sub-compartments involved: the gastrointestinal system (GIT), the plasma and the bones, in the case of radium simulation.

The transfer coefficients for the sub-compartments within the cow are combined with biological half-lives which is the time taken for the radionuclide activity concentration in tissues or milk be reduced by one half of its initial value.

A complete model simulation was already done for radium and the model can also be applied to uranium, thorium, lead and polonium depending on the availability of data.

For all sub-models at least one simulation was done. The necessary parameters were adopted from different sources: some parameters were adopted from measurements referring to a particular contaminated site, the Urgeiriça Uranium tailings piles, and others were adopted from published data. The unknown parameters were estimated from available data or from literature references in cases where on-site data were not available.