Naturally occurring radionuclides (NOR) such as 238U, 232Th and their decay-products are abundant in the environment. Contamination with NOR can arise from various anthropogenic activities that exploit raw materials for commercial purposes (e.g. metal mining and smelting, the phosphate industry and coal mining). For instance, it is well established that long-term use of phosphate fertilizers increase the soil U concentrations whereas the evidence for its effect on Th accumulation is weak.
The ability to estimate the activity concentration of NOR in biosphere components is a key step in evaluating the long-term impacts of these contaminants on human health and the environment. Radiological assessment models have integrated modules to predict NOR accumulation in soil and vegetation. These modules are often formulated in terms of simple parameters such as the equilibrium solid-liquid distribution coefficient (Kd) and the soil-to-plant transfer factor (TF). Although simple to measure and to use in radiological impact assessments, the Kd and TF are highly variable. This variability can be attributed to several sources including soil and plant properties, radionuclide speciation, the time after contamination and land management practices. Several approaches have been proposed to reduce the variability in the Kd and TF data with varying degree of success. One approach is to group available data by soil texture (i.e. sand, silt and clay), organic matter content and crop type. Another approach is to relate the Kd to physicochemical parameters that can be used as surrogates for the underlying mechanisms that govern radionuclide retention in soils (e.g. pH, cation exchange capacity, organic matter content). This cofactors approach has shown a greater potential to reduce the variability in the Kd data for some radionuclides. Determination of the appropriate set of cofactors however requires understanding of the underlying processes that control radionuclide biogeochemical behaviour in the soil-plant system. Kd can also be estimated with an established geochemical sorption model and some soil-specific information. However, applicability and accuracy of such approach requires further evaluation.
This PhD project will focus on one or more of the following radionuclides: 238U, 232Th, 226Ra, 210Pb and 210Po.
The overall objective of this research is gaining a better understanding of the geochemical fate of NORs in a wide range of soils (collected from Belgium and across Europe) and their uptake by typical crops in the temperate environment. This understanding will be translated to simple but realistic tools to predict the mobility and plant uptake of NORs under a wide range of environmental conditions. Ultimately, these predictive tools should contribute to the reduction of overall uncertainty associated with radiological impact assessments.
More specifically, this project aims to link solid-liquid distribution coeffients (Kd) and soil-to-plant tranfer factors (TF) with soil characteristics, the NOR labile fraction or site history/mineralogy. The feasibility of linking a geochemical sorption model with soil characteristics to estimate Kd is assessed.
The general hypothesis is that soil testing for labile (adsorbed) NOR, in combination with factors affecting the sorption (soil properties) will improve prediction of Kd and TF values as compared to the current approach in which soils are categorized in texture and organic matter classes. By replacing these parameters with functions of specific environmental properties, uncertainties associated with environmental impact assessment models will partially decrease.
It is envisaged to use a collection of uncontaminated soils representative of different land uses that will be spiked with specific NOR. In addition, there is the possibility to work with a collection of European NOR-contaminated soils, representative of some NOR situations. A large range of soil characteristics, identified as being potentially important in determining NOR mobility, availability and uptake, will be determined such as texture, organic matter, cation exchange capacity, pH, Fe, P, soil solution composition, bulk density, field capacity, etc.
Through laboratory experiments Kd values will be determined. Subsequently, it is aimed to determine the labile NOR in contaminated soil for which we propose to extract soils with a range of selective standard soil extractants. Next, one (or more) native plant(s) will be selected to perform transfer studies in the greenhouse to determine soil-to-plant TF for the characterized soils.
Subsequently, the experimentally obtained Kd and TF values will be linked to extractable fractions of NOR and to soil parameters via a parametric model. In addition, the parametric Kd values will be used in a simple multi box model (soil, soil solution, plant) to predict TF for different soil-plant combinations and compare it against the measured TF values. The selective extractions will serve to establish a soil test for bioavailability. A part of the collected and characterized soils will be used to validate the Kd and TF equations. Predicted Kd and TF values will be compared with the values obtained under lab conditions. In addition, a comparison will be made between soil-to-plant transfer studies in the lab and soil-to-plant transfer studies in the field to investigate if field data can be predicted from lab data.