The Oxitran project - measure and understand the oxygen consumption in argillaceous media

Document available in extended abstract form only. The geological disposal design developed by the French National Agency for Radioactive Waste Management (Andra) for high level waste (HLW) involves stainless and carbon steel components. Up to now, most studies on steel corrosion in clayey media have been focused on reactivity under reducing conditions since, at the timescale of the geological disposal, minerals such as pyrite are supposed to impose a reducing environment. However, soon after closure of HLW cells, the entrapped oxygen introduced in the disposal cell during the operational stage may induce an oxic corrosion of metallic structures until the whole consumption of oxygen. Under humid conditions, oxic corrosion may lead to drastic corrosion rates. Therefore, the assessment of steel corrosion can crucially depend on the oxidizing period and its duration. Corrosion rates on steel discs were measured after 2, 5 and 10 years of contact with an argillite rock in the Tournemire Experimental Station (France). Obtained rates (100 μm/year) underline that the oxidizing period may have lasted between 2 and 3 years. Such a corrosion rate shows the need to identify accurately the oxygen consumption rate with regard to the exposed steel surface, the argillite surface and the volume of the gas phase. Unfortunately, only few reliable in situ experimental data are available concerning the oxygen consumption in media representative of a geological waste disposal. In that purpose, the OXITRAN (Oxidising Transient) project has been developed by the French Institute of Radioprotection and Nuclear Safety (IRSN) to understand the evolution and the duration of the oxidizing transient in the Tournemire argillite, in presence and absence of steel, and in unsaturated in situ conditions. This project aims at answering two questions: - What is the oxygen consumption rate by the argillaceous rock for a given volume of gas and exposed rock surface? - What is the oxygen consumption rate by steel corrosion for a given exposed metal surface, exposed rock surface and gas volume? The main objective of such an experiment is finally to build a sufficient knowledge on oxygen consumption in the conditions prevailing in a HLW disposal cell in order to model accurately the oxidizing transient.. Two tests are carried out in 8-meter long horizontal boreholes (100 mm diameter) within the undamaged zone. To avoid previous pyrite oxidation, each borehole was drilled just prior starting the experiment. Each test consists in isolating a 1-m long measurement chamber at the borehole end. The tightness is ensured by a 1.30-m long packer inflated at 5 bars. Two probes are emplaced at the packer back: an optical O2 probe and a total pressure probe in order to prevent the oxygen content fluctuations from the total gas pressure changes. Firstly, a gas line is used to fill the chamber with pure Argon, and the monitoring of O2 concentration during 24 h showed the Oxygen-tightness of the system (the oxygen content remained nil). Then, the same gas line is used to ensure oxygen circulation until a 19% The first test (OXI 1) was carried out without steel in order to evaluate the oxygen consumption by the rock, which occurs mainly by pyrite oxidation. The chamber volume is equal to 7.9 dm3 and the exposed clayey rock area is equal to 0.31 m2. After one month, a slow and almost constant decrease of the oxygen partial pressure (pO2) is observed: the pO2 decreases by 4.2.10-2 bar. The second test (OXI 2) was carried out in the presence of carbon steel to assess its effect on the oxygen consumption rate. Complementary test series will be started in 2012, in order to highlight the influence of the air relative humidity and the steel reactivity on oxygen consumption. Numerical simulations with the reactive transport code HYTEC are carried out to fit the experimental data obtained on the OXI 1 and 2 tests, using the unsaturated gas phase model implemented recently. This inverse modelling will enable to quantitatively characterize the kinetics of oxygen in situ r eactivity with the argillite, alone and in competition with steel corrosion. (authors)