Np(V) migration in a single fracture from Äspö, Sweden: Experiments and reactive transport modeling

Florian Huber1, Paolo Trinchero2, Jorge Molinero2, Thorsten Schäfer1
2 Amphos21 Consulting S.L., Spain

O 1.8 in Numerische Simulation von Strömungs- und Transportprozessen in Grundwasserleitern und angrenzenden Kompartimenten

30.05.2014, 14:20-14:40, H17, NW II

In the context of nuclear waste disposal reactive transport modelling represents an important tool for long-term predictions of radionuclide migration. In this work, the Lagrangian-based framework FASTREACT (FrAmework for STochastic REACive Transport) (Trinchero et al., 2013) coupled to the geochemical code PhreeqC is applied. This newly developed approach is tested to model experimental data on HTO and 237Np(V) migration through a natural single fracture in Äspö diorite. Migration experiments have been conducted on a single fractured drill core (length 13.5cm; diameter 5.05cm) from Äspö, Sweden using a radionuclide cocktail on basis of natural Grimsel ground water to mimic the influence of glacial melt water intrusion into a repository. To model the Np reactive transport we tested different approaches with increasing complexity ranging from simplified Kd-type sorption models (1st order reversible sorption/desorption kinetics) to more complex, mechanistically based surface complexation models for sorption of NpO2+ onto one (Fe2O3x2H2O (HFO)) and two (HFO and biotite (Bt)) mineral phases.

The experimental Np breakthrough curve (BTC) shows a retardation (retardation factor Rf = 1.8) compared to the conservative tracer (HTO) BTC and a pronounced tailing indicative of an interaction with the fracture surfaces. Np recovery is ~76% at the end of the experiment, but the tailing seems to be still on-going thus a quantitative Np recovery is reasonable to expect. The high Np recovery is in line with measured Eh values (~ +200 mV) and thermodynamic modelling yielding no Np reduction and NpO2+ as dominating species, which is known to sorb weakly.

The experimental HTO BTC is captured well by FASTREACT verifying the correctness of the coupling of PhreeqC and FASTREACT. Using a single Kd derived from the migration experiments (Kd = 0.18) in the different SCM models, it is not at all possible to adequately describe the Np BTC. While the peak position is captured relatively well, the model completely fails both in describing the Np BTC peak concentration and tailing. In case of the 1st order kinetic model, the model fit describes the ascending part of the curve quite well, and also shows a pronounced tailing, but the overall fit is not satisfactorily. Adjusting the (unknown total reactive) mineral surface area using the parameter estimation code PEST improved all model fits, but still deviations to the experimental BTC are observed. The combined study leads to the conclusion that fracture geometry induced flow heterogeneities are only important for the conservative tracer (HTO) while the Np migration behaviour is governed by the chemical interaction with the fracture surface showing pronounced sorption kinetics within the duration of the experiments. To sum up, the FASTREACT approach coupled to PhreeqC has been shown to represent an effective computational framework for the interpretation of laboratory experiments of radionuclide transport.

Trinchero, P., J. Molinero, and G. Roman-Ross, FASTREACT - Final Report -  Streamline Approach for the Solution of Multicomponent Reactive Transport Problems. R-10-45, 2011, Svensk Kärnbränslehantering AB (SKB).

Export as iCal: Export iCal

Letzte Änderung 31.10.2013