Modelling the behaviour of compacted bentonite for nuclear waste disposal

压实膨润土被设想为地质处置设施（GDF）中工程屏障系统（EBS）的一部分。作为核废料罐和主地层之间的缓冲区，它们将在与主地层的界面处经受水合作用，并在与主地层的界面处经受高温。EBS设计的目的是通过水合作用促进膨润土缓冲层的膨胀，从而增加其体积，以密封罐和主体地层之间的施工空隙。与膨润土的建模相关的是其在压实状态下的双重孔隙结构的识别，包括粘土聚集体内的微观孔隙度和粘土聚集体之间的宏观孔隙度。这种结构随着水合作用而减少，导致完全水合（饱和）时的单孔隙材料。伦敦帝国理工学院（ICL）最近的一项博士研究开发了一种新的压实膨润土双结构本构模型（IC DSM; Ghiadistri等人，2018; Ghiadistri，2019）。该模型是巴塞罗那扩展框架（Gens & Alonso，1992）的扩展和通用版本，并且在定制的计算平台ICFEP（帝国理工学院有限元程序; Potts & Zdravkovic，1999，2001）中实现，其操作控制有限元方程的完全热-水-机械（THM）耦合公式（Cui等人，2018年）。该模型已成功地应用于两个实验室规模的膨胀压力实验的模拟（Ghiadistri等人，2019 a）在压实膨润土上（例如Dueck等人，2014）和大规模田间实验（Ghiadistri等人，2019 b），如FEBEX实验（ENRESA，2000），暴露于100 ° C以下的温度。自2017年以来，这些建模工具也已用于BEACON Euratom项目（批准号745942），对压实的膨润土块和颗粒进行模拟实验室膨胀压力测试，以及其他大规模的现场实验。（实验，现场，数值）对压实膨润土的行为，与核废料处理，曾考虑将其暴露在高达100摄氏度的温度下。拟议研究的目的是探索膨润土缓冲剂在100 ° C以上温度下的行为，方法是使用ICFEP软件对缓冲剂和宿主岩石的热、水力和机械演化进行预测建模，与瑞士Grimsel试验场的HotBENT实验有关。该研究旨在帮助优化GDF设计，包括GDF内地下拱顶和沉积孔网络的覆盖范围。 该研究将首先使用已发表的文献对高温下膨润土行为的现有实验证据进行审查。与背景研究中描述的方法类似，将首先模拟小规模实验室实验，以验证建模工具在100 ° C以上温度下的性能。然后，将数值工具应用于大规模HotBENT实验的模拟，在该实验中，膨润土缓冲层暴露在超过250 ° C的温度下。膨润土的THM演变的数值预测将进行比较，从这个实验中收集的现场测量。现场数据将由英国放射性废物管理局（RWM）提供，该机构还将担任该项目的工业监督员。 数值模拟将进一步研究近场效应的主机形成。特别是，这将涉及量化的可能变化，由于温度，在渗透性和孔隙水压力制度在地面周围的工程屏障，以及这些变化的程度与一个单一的沉积孔/拱顶。

Compacted bentonite clays are envisaged as part of engineered barrier systems (EBS) in geological disposal facilities (GDF). Placed as a buffer between the nuclear waste canister and the host formation, they will be subjected to hydration, at their interface with the latter, and to high temperatures, at their interface with the former. The objective of the EBS design is for hydration to promote the swelling of the bentonite buffer and hence increase its volume to seal construction voids between the canister and the host formation. Pertinent to the modelling of bentonite is the recognition of its double-porosity structure in the as-compacted state, comprising the micro-porosity within the clay aggregates and macro-porosity between the clay aggregates. This structure diminishes with hydration, leading to a single-porosity material at full hydration (saturation). A recent PhD research at Imperial College London (ICL) developed a new double-structure constitutive model for compacted bentonite clays (IC DSM; Ghiadistri et al., 2018; Ghiadistri, 2019). The model is an extended and generalised version of the Barcelona Expansive framework (Gens & Alonso, 1992) and is implemented in the bespoke computational platform ICFEP (Imperial College Finite Element Program; Potts & Zdravkovic, 1999, 2001), which operates a fully thermo-hydro-mechanically (THM) coupled formulation of the governing finite element equations (Cui et al., 2018). The model has been successfully applied to simulations of both laboratory-scale swelling pressure experiments (Ghiadistri et al., 2019a) on compacted bentonite (e.g. Dueck et al., 2014) and large-scale field experiments (Ghiadistri et al., 2019b) such as the FEBEX experiment (ENRESA, 2000), exposed to temperatures under 100^o C. Since 2017 these modelling tools have also been used in the BEACON Euratom project (grant no. 745942), simulating laboratory swelling pressure tests on compacted bentonite blocks and pellets, as well as other large-scale field experiments.Most of the existing research (experimental, field, numerical) on the behaviour of compacted bentonite, in relation to nuclear waste disposal, has considered its exposure to temperatures of up to 100^o C. The objective of the proposed research is to explore the behaviour of bentonite buffers at temperatures above 100^o C, by conducting predictive modelling with the software ICFEP, of the thermal, hydraulic and mechanical evolution of the buffer and host rock, associated with the HotBENT experiment at Grimsel Test Site in Switzerland. The research is aimed at helping optimisation of GDF design in terms of the footprint of the network of underground vaults and deposition holes within a GDF. The research will first conduct a review of existing experimental evidence on bentonite behaviour under high temperatures, using published literature. Similar to the methodology described in the Background research, small-scale laboratory experiments will be simulated first to verify the performance of the modelling tools at temperatures over 100^o C. The numerical tools will then be applied to simulations of the large-scale HotBENT experiment in which bentonite buffers are exposed to over 250^o C temperatures. The numerical predictions of the bentonite's THM evolution will be compared to field measurements collected from this experiment. The field data will be provided by the Radioactive Waste Management (RWM), UK, who will also act as industrial supervisor for the project. The numerical modelling will further investigate the near-field effects in the host formation. In particular, this will involve quantification of the likely changes, due to temperature, in the permeability and the pore water pressure regime in the ground around the engineered barrier, as well as the extent of these changes in relation to a single deposition hole / vault.