Speaker
Description
Migration of radionuclides (RN) in the underground environment is strongly dependent on geochemical conditions that influence aqueous speciation and sorption processes. Current studies of RN transport with groundwater from the engineered near-surface disposal facility (ENSDF) for radioactive waste (RW) located in Chornobyl exclusion zone (ChEZ) use sorption models based on constant values of distribution coefficients (Kd) that do not take into account changes in geochemical parameters (e.g. pH, Eh, concentrations of dissolved ions). This assigns large uncertainties to measured Kd values and thus contributes to unnecessary conservatism of safety assessments. Earlier developed mechanistic smart-Kd values approach, implemented in this study, provides an advanced alternative to the conventional Kd approach by describing the variable geochemistry more realistically. Improved understanding of Kd-driven parameters will allow to derive less conservative RW inventory limits of ENSDF under regulatory established dose limits for post-closure safety assessment.
To implement this global research objective, in this study an overall model of RN migration from ENSDF with groundwater was constructed and parameterized with two categories of data: site-specific (geophysical characteristics of sediments; representative minerals of ChEZ sediments: quartz, orthoclase, kaolinite, montmorillonite, illite; chemical composition of groundwater; dose relevant RN: ¹³⁷Cs, ⁹⁰Sr, ²³⁹⁺²⁴⁰Pu, ²⁴¹Am and ²³⁷Np, and invariant (thermodynamic) data for aqueous speciation, mineral solubility and surface species for relevant element/mineral pairs.
To describe sorption by mechanistic, so-called smart-Kd values, thermodynamic datasets were developed and used for calculations:
1) surface complexation model (SCM) parameters:
- for quartz, orthoclase and kaolinite: diffuse double layer model;
- for montmorillonite and illite: non-electrostatic model;
2) ion exchange parameters: for montmorillonite and illite.
As SCM data for some element/mineral pairs were missing, chemical and mineralogical analogies were applied (e.g. U(VI) used as chemical analogue for Np(VI) and Pu(VI); montmorillonite used as min-eralogical analogue for kaolinite).
The sets of smart-Kd values (smart-Kd matrices) were obtained as a function of the following variable geochemical parameters: pH, Eh, SLR, [RN] (4095 combinations of parameters).
To quantify the contribution of each above-mentioned modelled geochemical parameter to Kd variability, a comprehensive variance-based global sensitivity analysis (SA) was performed. Based on the analysis of calculated values of sensitivity indices of the parameters, ranking of the significant parameters was carried out. The SA results demonstrate that smart-Kd values of Cs⁺ are predominantly driven by pH; Kd’s of Sr²⁺ are mainly influenced by [Sr²⁺]; Kd’s of Am³⁺ are the most sensitive to [Am³⁺] changes while pH and Eh have less impact; Eh is the most meaningful parameter defining sorption of Np and Pu whereas pH changes have less influence on their Kd values.
The calculated smart-Kd values will be further utilized in a transport simulation software tool to re-estimate maximum allowed specific activities of RN based on inverse modelling approach.
The authors gratefully acknowledge the funding provided by the Siebold Sasse Foundation (Leibniz University Hannover, Germany).