lnUoduefion The main objectives of the ANDRA/CCE study presented here is understanding how water in contact with a clay formation acquires its chemistry. The knowledge of the different processes involved should help determine the major parameters that are needed to achieve geochemical modelling. Several organisations (BGS, BRGM, CEA, CEN/SCK, CNRS, ERM and GRAM) are taking part in this project which is funded jointly by ANDRA and CEC. BRGM is more particularly in charge of modelling claywater

more »

... uilibria from the study of the mechanisms of solid-and fluid-phase interactions. The first results of this study are presented here. Method and main results The clay formation studied is the Boom Clay Formation, about 35 Ma old and about 100 m thick which is sited between two sandy aquifers (Rupelian and Anversian sands) at about 225 m depth. Clay samples have been taken out from the Underground Research Facility in Mol, Belgium, from horizontal cores drilled under anaerobic conditions in order to minimize oxidation of the clay samples by air. Chemical analyses carried out on the clay samples have shown that they are remarkably homogeneous. Chemical data along with mineralogical determinations using X-Ray diffractometry, scanning and transmission analytical electron microscopy and electron microprobe analysis have allowed us back calculating the mean composition of the studied clay: quartz: 18.9 %, albite: 2.8 %, K-feldspar: 6.0 %, calcite: 1.0 %, dolomite: 0.9%, siderite: 0.4 0/0, pyrite: 4.2 %, iron oxide: 2.7 0/0, titanium, manganese and phosphorus oxides making up respectively 1.0 %, 0.03 % and 0.1%. The clay fraction amounts to 57.4 % by weight from which kaolinite makes up 10.1%, illite 16.7 %, illite-smectite mixed layers 28.5 % and chlorite 21%. Water content for these clays ranges between 18.0 and 22.0 % by weight of dry sample. Cation exchange capacities (CEC) have been determined following three methods cobaltohexamine chloride, cesium chloride and ammonia acetate, which were applied on two kind of samples: oxidised and non-oxidised samples. The results obtained show that: 1. The most appropriate method for our study is the one using cobaltohexamine chloride, as, even if it is not perfect, it allows us to determine exchanged cations in addition to the bulk CEC. Cesium chloride exchange tends to under estimate the real CEC, as part of cesium remains trapped in clay. Ammonia acetate yielded over estimation of the bulk CEC along with an over estimation of exchangeable calcium and magnesium by formation of calcium and magnesium acetate complexes; 2. The values obtained by cobaltohexamine exchange are fairly homogeneous. For oxidised samples (either air dried or dried at 105~ in a drying cabinet) the CEC obtained amounts to 24 meq/100g dry rock and is very close to the sum of exchanged cations. For non-oxidised samples, a large discrepancy exists hetween the bulk CEC measured (about 30 meq/100g dry rock) and the sum of exchanged cations (18 meq/100g only, what is different from oxidised clay). This discrepancy can be attributed partly to an over estimation of the bulk CEC because cobaltohexamine is readily complexed by organic matter and in particular humic acid present in non-oxidised samples, and partly to the under estimation of the sum of exchanged cations (occurrence of about 5 meq/100g dry sediment of exchangeable protons). From the whole set of experiments we conducted, we can consider that the average CEC for this clay is 24.0 meq/100g dry sediment. For non-oxidised sediments, concentration of exchangeable sodium, potassium, calcium, magnesium and protons averages respectively 8.7, 2.3, 3.8, 3.7 and 5.5 meq/100g dry rock.