One of the major issues in the safety of a deep geological repository (DGR) for radioactive waste is the migration of hydrogen (H₂) resulting mainly from anaerobic metal corrosion and from water and waste radiolysis. The migration of hydrogen may have adverse effects on the mechanics of the host rock and of the engineered barrier systems within a DGR. Some studies pointed out involvement of hydrogen “piston effect” in the transport of radionuclides. In this mechanism, the gas phase displaces the contaminated water, along backfilled but still relatively permeable drifts, towards the main shafts. The precise quantification of such a displacement is still a challenging task because it involves simulations at the scale of a DGR and only few studies were dedicated to this subject. In this work an attempt has been made to verify, through a simplified DGR model, if the usual parametrizations could give rise to a substantial and continuous water movement during the first 100,000 years after the closure of a DGR, when a significant H₂-flux reaches its main drift. The assessment of the piston effect importance, related to the H₂ production and release, is evaluated in terms of cumulative liquid-phase travel distance (CLTD) within the main drift and its persistence in time. Scenarios simulated by our model show that, in the investigated conditions, piston effect is not negligible (i.e., CLTD-values are greater than 100 m) and that water would be displaced towards the main shafts. However, this work is focused on the mechnisms involved, and the results obtained cannot therefore be genralized to any disposal concept. Furthermore, additional studies are necessary for improving this model by analyzing uncertainty propagation in its parameters, and by considering, e.g., gas-entry pressure and hysteresis phenomena usually neglected in the simulation models.

放射性废物的深层地质处置库（DGR）安全性的主要问题之一是氢（H ₂）主要是由于厌氧金属腐蚀以及水和废物的辐射分解造成的。氢的迁移可能会对DGR中的基质岩石和工程屏障系统的力学产生不利影响。一些研究指出，氢“活塞效应”参与了放射性核素的运输。在这种机制下，气相将污染的水沿着回填但仍相对可渗透的漂流移向主轴。这种位移的精确量化仍然是一项艰巨的任务，因为它涉及DGR规模的模拟，并且只有很少的研究专门针对该主题。在这项工作中，我们尝试通过简化的DGR模型来验证，如果通常的参数设置会在前100个水位期间引起大量连续的水运动，₂通量达到其主要漂移。与H ₂相关的活塞效应重要性评估根据主要漂移内的累计液相行进距离（CLTD）及其持续时间来评估生产和释放。我们的模型模拟的情况表明，在研究的条件下，活塞效应不可忽略（即CLTD值大于100 m），并且水将流向主轴。但是，这项工作集中在所涉及的机械上，因此获得的结果不能归纳为任何处置概念。此外，需要通过分析其参数中的不确定性传播并考虑例如模拟模型中通常忽略的进气压力和滞后现象来进行改进，以改进该模型。