10.12.2012 Views

Euradwaste '08 - EU Bookshop - Europa

Euradwaste '08 - EU Bookshop - Europa

Euradwaste '08 - EU Bookshop - Europa

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

ehaviour expected based on measurements on rock samples, data modelling also shows that for<br />

strongly sorbing tracers, complexities in mass transfer (borehole mixing, diffusion in filters, etc.)<br />

strongly influence experimental data, making clear-cut estimations of values for diffusion-retention<br />

parameters in the undisturbed clayrock difficult. Some of these ambiguities will certainly be removed,<br />

at least for the more mobile tracers, when information on tracer distribution in the surrounding<br />

rock becomes available (post-FUNMIG) for providing further constraints on model interpretations.<br />

Finally, Pu(V) diffusion into Opalinus clayrock samples kept under in-situ confining pressures was<br />

measured by FZK-INE, with results from both batch and in-diffusion experiments showing that Pu<br />

is reduced to the Pu(IV), probably by Fe(II) contained in rock minerals (probably chlorite), and is<br />

retained (sorption or other process?) on preferential sites.<br />

Callovo-Oxfordian<br />

The Callovo-Oxfordian research program is, by design, quite similar to that described for the<br />

Opalinus clayrock since one of the objectives of RTDC3 was to generate comparable data sets for<br />

different clayrocks in order to identify common characteristics and eventual significant differences<br />

in RN migration properties. The principal difference between the two is in the working hypothesis,<br />

and consequent experimental approach, taken by a consortium of French partners (CEA, ERM,<br />

Hydr’asa, Andra) for quantifying and modelling migration of highly-sorbing RN in clayrock. The<br />

guiding assumption here was that, given the very small spatial scales covered by RN migration during<br />

the time frames of in-diffusion experiments owing to the preponderant effect of sorption, it<br />

would be important to be able to quantify how tracer mass present in the diffusion profile was distributed<br />

relative to rock mineral constituents and associated porosity. The approach taken involved<br />

carrying out, on a single, oriented cm-scale volume of clayrock:<br />

Characterization of the mineral-pore space organization and construction of the corresponding<br />

2-3D numerical models, including analysis of grain organization (cf. §2.2), followed by TDD<br />

modelling of diffusion of non sorbing tracers (cf. §3.2);<br />

In-diffusing a highly-sorbing tracer (Eu, Cu) into the rock (gradient perpendicular to sedimentation<br />

plane) for a given period, then sectioning the rock perpendicular to bedding;<br />

Simultaneous mapping of tracer mass and mineral grain 2D spatial distributions, relative to the<br />

in-diffusion surface, using LIBS and Electron Probe Micro Analyser (calculation of the average<br />

diffusion profile), followed by construction of numerical models of the in-diffusion zone;<br />

Inverse modelling of tracer diffusion (source term, spatial distribution) to determine Da values<br />

for the porous mineral zones (clay matrix + disseminated pyrite), followed by comparison with<br />

batch Kd.<br />

The results for Cu(II) (the most complete dataset available) are rich in information concerning both<br />

tracer migration phenomena and potential experimental artefacts which can make measurements on<br />

such small rock volumes difficult to interpret. Regarding the former, it was found (i) that Cu does<br />

not penetrate into, nor sorb significantly onto, the carbonate and quartz grains and associates preferentially<br />

with the pyrite minerals dispersed in the clay matrix, (ii) Cu diffusion profiles (clay and<br />

pyrite) developed over a distance of roughly 2 mm into the rock and (iii) that evaluation of the entire<br />

dataset (reservoir + profile) yields values for De (2.5 * 10 -10 m 2 .s -1 ) and Rd (3500 mg/L) which<br />

are consistent both with results showing that De for cations in the COx are generally significantly<br />

higher than the value for HTO (~ 2.5*10 -11 m 2 s -1 ) and measurements of Kd(Cu) which give values<br />

in the 3000 to 9000 range (note that Rd could be controlled by the redox reactivity of this tracer).<br />

On the other hand, detailed analysis also shows that the roughly 50 m layer of rock in contact<br />

with the source solution is significantly perturbed, both in terms of its mineral-porosity composi-<br />

323

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!