Deep geological disposal with multiple barriers composed of engineered and natural barriers has been considered as the most suitable disposal method for high level nuclear wastes. In terms of the geological evaluation factors, brittle structures such as fractures and faults should be characterized around the repository site, because radionuclides transfer mainly with groundwater in the subsurface and groundwater flows through discontinuous brittle structures. The geological survey for the characterization of deep geological repository sites is widely conducted by narrowing the survey area from regional scale down to local scale, which could be divided into three steps: 1) using remote sense or geophysical survey, 2) trench and drill core logging including field survey based on the first step, 3) detailed geological survey in the tunnel. In this study, we analyzed the distribution of geological structures to derive the history of brittle deformation in and around the KURT (KAERI Underground Research Tunnel) site located in the KAERI (Korea Atomic Energy Research Institute). The bedrock of the KURT site is mainly consist of Jurassic two-mica granite, which is extensively intruded by andesitic dikes of Cretaceous with N-S to NE-SW strikes. The two-mica granite in the study area was deformed in a ductile deformation environment and has been overprinted by major geological structures such as faults, dikes, veins, and joints. From this study, we identified 8 brittle deformation events based on the cross-cutting relationship among the geological structures, which are obtained from the analyses in and around the KURT. In order to evaluate the reactivation and fluid flow potential of brittle structures, it is essential to determine the characteristics and ages of the brittle structures and the composed rocks around the site.

A methodology is under development to restore and predict the long-term evolution of the natural barrier comprise the site of radioactive waste disposal for surface geological outcrop, tunnel face and drill core. Considering the condition that the radioactive waste repository should be located in the deep part, the drill core is an important subject that can identify deep geological properties that could not be confirmed near the surface. In this study, we investigated proper age dating methods to construct lithological model of the disposal site with regard to the long-term safety. Also, preliminary age dating locations were selected using the lithological distribution results by depth through geochemical and micro-structural analysis for the deep drill cores excavated around KURT. In the study area, the dikes presumed the Cretaceous were intruded by Jurassic granites. As for the granotoids, U-Pb age dating for zircon, which is resistant to deformation or metamorphism and has loss, is often used. In the case of the dikes, K-Ar and 40Ar/39Ar age dating for the argon captured in the rocks after magmatism is often used. Through U-Pb zircon ages of KURT site granotoids, we expect to solve the clustering problem (granite and granodiorite), which is different from precious chemical analysis (XRF) results and TAS-diagrams. 40Ar/39Ar age dating to be used for the dikes is suitable for the perspective of lithological model of the disposal site. Because, it can compensate for accuracy problems such as sample heterogeneity in K-Ar age dating and is used for volcanic rocks. In the further study, we plan to determine the appropriate sampling locations by the selected age dating methods from the perspective of disposal in this study.

The radioactive waste repository consists of an engineered barrier and a natural barrier and must be managed safely after isolation. We classify the geological events of natural barriers for the evaluation of their present and future disposal stability assessment, they can be divided into regional and regional evolutions according to their scale. Regional evolution can be quantitatively explained by plate tectonics and regional rock distribution, and local evolution can be explained by petrological, mineralogical evidence and ductile, brittle deformation. Plate tectonics can explain the change quantitatively by restoring the direction of the Earth’s magnetic field recorded when rocks were formed. The time units for these changes are tens of millions of years to hundreds of millions of years, but plate tectonic is a way to estimate geological history. It can be assessed by extrapolating past knowledge considering the known geological events of radioactive waste repository. It is possible to derive a conservative value of the change of the geological environment in the time unit of disposal stability. The Korean Peninsula belongs to the edge of the Eurasian plate and is divided into Gyeonggi, Yeongnam Massif, Okcheon orogeny belt, and Gyeongsang Basin. To quantitatively determine their geological history, we collected paleomagnetic data using rocks from the Korea Peninsula (paleomagnetic database and papers). We attempted to carry out the apparent polar wander paths (APWPs) on the Korean Peninsula by collecting and sorting data. Since the Korean Peninsula is composed of multiple massifs, this APWP is expected to serve as a basis for explaining the local crustal rotation or brittle ductile deformation. Furthermore, by extrapolating the change pattern from the past to the present, it can contribute to the estimation of the future geological evolution.

Comprehensive identification and systematic classification of all features, events and processes (FEP) that influence on the performance of a high-level radioactive waste disposal system is essential for safety assessment. Nuclear energy agency (NEA) has been developing and updating the standardized generic FEP list, so-called NEA international FEP list, which may be used as the basis to develop project-specific FEP lists to reflect diverse system and site characteristics in different countries. On the basis, Finland and Sweden have recently got licenses to construct spent nuclear fuel deep disposal facilities. Also in Korea, timely construction of a high-level radioactive waste disposal facility is an urgent issue for stable operation of nuclear power plants. For this end, a FEP list that properly considers for system and site characteristics of Korean high-level radioactive waste disposal facility needs to be developed. In this study, the most recent NEA international FEP list published in 2019 was comprehensively reviewed with focus on the structure of the classification system and the physicochemical mechanisms associated with the key elements. The obtained results will be used for the comparative analysis of domestic and oversea project-specific FEP lists and for the development of a generic FEP list relevant to Korean high-level radioactive waste disposal system.

Since July 2021, the Korea Radioactive Waste Agency has been conducting a safety case development study for the Korean deep geological repository program. The safety case includes generating scenarios in which radioactive materials from spent nuclear fuel repository reach the human biosphere by combining selective FEPs (Features, Events, and Processes). This safety case should be able to transparently explain the process in which conclusions have been drawn not only to stakeholders but also to the public by presenting safety arguments. The scenario development stage consisting of FEP screening, scenario generation, and uncertainty analysis procedures should have a database management system. Database management system was performed in countries such as Sweden, which obtained approval for the construction of spent nuclear fuel repositories, and the United States, where various preliminary research was carried out. Korea Atomic Energy Research Institute also has experience in designing and operating its own database, which has conducted preliminary research on disposal of the spent nuclear fuel. Currently, the safety assessment of the Korean spent nuclear fuel repository is in the early stages of research, but it is necessary to set up a basic framework for database design while the collection of FEP data from domestic and international preliminary studies is under development, and it is advantageous for efficient database construction and operation. Therefore, this paper presents the current status of database design considering completeness and transparency from the FEP screening stage to the scenario development stage in the safety assessment process of the Korean spent nuclear fuel repository. In this process, the functional requirements that the database should provide, the database schema capable of implementing them, and simple examples are presented together. The objectives of this database design are flexible FEPs management, high integrity and consistency, and expandability for linking with the safety case database. The FEP data to be inputted into the database includes a list of major opened FEPs, including International FEPs from Nuclear Energy Agency, which were referred for PFEPs (Project-specific FEPs), and PFEPs applied to POSIVA's Olkiluoto repository. As an additional function, queries from the database are used to visually express the process of deriving scenarios through Rock Engineering System, a widely known scenario generation methodology.

Deep geologic disposal of high-level nuclear wastes (HLW) requires intensive monitoring instrumentations to ensure long-term security. Acoustic emission (AE) method is considered as an effective method to monitor the mechanical degradation of natural rock and man-made concrete structures. The objectives of this study are (a) to identify the AE characteristics emitted from concretes as concrete materials under different types of loading, (b) to suggest AE parametric criteria to determine loading types and estimate the failure stage, and finally (c) to examine the feasibility of using AE method for real-time monitoring of geologic disposal system of HLW. This study performs a series of the mechanical experiments on concrete samples simultaneously with AE monitoring, including the uniaxial compression test (UCT), Brazilian tensile test (BTT) and punch through shear test (PTST). These mechanical tests are chosen to explore the effect of loading types on the resulting AE characteristics. This study selects important AE parameters which includes the AE count, average frequency (AF) and RA value in the time domain, and the peak frequency (PF) and centroid frequency in the frequency domain. The result reveals that the cumulative AE counts, the maximum RA value and the moving average PF show their potentials as indicators to damage progress for a certain loading type. The observed trends in the cumulative AE counts and the maximum RA value show three unique stages with an increase in applied stress: the steady state stage (or crack initiation stage; < 70% of yield stress), the transition stage (or damage progression stage; 70–90% of yield stress) and the rising stage (or failure stage; > 90% of yield stress). In addition, the moving average PF of PTST in the early damage stage appears to be particularly lower than that of UCT and BTT. The loading in BTT renders distinctive responses in the slope of the maximum RA–cumulative AE count (or tan ). The slope value shows less than 0.25 when the stress is close to 30% of BTT, 60% of UCT and 75% of PTST and mostly after 90% of yield stress, the slope mostly decreases than 0.25 in all tests. This study advances our understanding on AE responses of concrete materials with well-controlled laboratoryscale experimental AE data, and provides insights into further development of AE-base real-time diagnostic monitoring of structures made of rocks and concretes.

Uranium isotopes (238U, 235U, and 234U) found in natural environments and their activity ratios (235U/238U and 234U/238U) have been used as an important tool in investigating various geological processes, especially in natural analogue studies. Occurrence and fractionation of uranium isotopes in nature between 238U, 235U, and 234U were investigated. Various measurement methods have been used for the determination of isotopic ratios and geochronology. Thus, we reviewed and summarized the measurement methods such as alpha spectrometry, gamma spectrometry, thermal ionization mass spectrometry (TIMS), secondary ion mass spectrometry (SIMS) with sensitive high resolution ion microprobe (SHRIMP), and multiple-collector inductively coupled plasma mass spectrometry (MCICPMS). Research status of natural analogue studies carried out using uranium isotopes and their isotopic ratios were also reviewed and summarized in terms of long-term behaviors of radionuclides in various foreign uranium deposits as analogues of high-level radioactive waste repositories. Research results for mineralogical, geochemical, and biogeochemical behaviors of uranium in various natural analogue sites were collected and analyzed to investigate the migration and retardation processes of uranium through geological media. These long-term behaviors of uranium and uranium isotopes include dissolution/precipitation of uranium minerals, interactions of uranium with various fracture minerals including sorption and incorporation, redox reactions by minerals and microbes, and hydrological groundwater flow thorough rock fracture systems including identification of flow paths and groundwater circulation. The results of this study will contribute to performing future natural analogue studies in domestic uranium deposits and provide basic information and knowledge for understanding long-term geochemical and geochronological behaviors of radionuclides in a high-level radioactive repository.

A deep geological disposal system, which consists of the engineered and natural barrier components, is the most proven and widely adopted concept for a permanent disposal of the high level radioactive waste (HLW) thus far. The clay-based engineered barrier is designed to not only absorb mechanical stress caused by the geological activities, but also prevent inflow of groundwater to canister and outflow of radionuclides by providing abundant sorption sites. The principal mineralogical constituent of the clay material is montmorillonite, which is a 2:1 phyllosilicate having two tetrahedral sheets of SiO2 sandwiching an octahedral sheet of Al2O3. The stacking of SiO2 and Al2O3 sheets form the layered structures, and ion-exchange and water uptake reactions occur in the interlayer space. In order to reliably assess the radionuclide retention capacity of engineered barrier under wide geochemical conditions relevant to the geological disposal environments, sorption mechanisms between montmorillonite and radionuclides should be explicitly investigated in advance. Thus far, sorption behavior of mineral adsorbents with radionuclides has been quantified by the sorption-desorption distribution coefficient (Kd), which is simply defined as the ratio of radionuclide concentration in the solid phase to that in the equilibrium solution; the Kd value is conditional, and there have been scientific efforts to develop geochemically robust bases for parameterizing the sorption phenomena more reliably. In this framework, application of thermodynamic sorption model (TSM), which is theoretically based on the concept of widely accepted equilibrium models for aquatic chemistry, offers the potential to improve confidence in demonstration of radionuclide sorption reactions on the mineral adsorbents. Specifically, it is generally regarded in the TSM that coordination of radionuclides on montmorillonite takes place at the surficial aluminol and silanol groups while their ion-exchange reactions occur in the interlayer space also. The effects of electrical charge on the surface reactions are additionally corrected in accordance with the numerous theories of electrochemical interface. The present work provides an overview of the current status of application of TSM for quantifying sorption behaviors of radionuclides on montmorillonite and experimental results for physical separation and characterization of Ca-montmorillonite from the newly adopted reference bentonite (Bentonil- WRK) by means of XRD, BET, FTIR, CEC measurement, and acid-base titration. The determined mineralogical and chemical properties of the montmorillonite obtained will be used as input parameters for further sorption studies of radionuclides with the Bentonil-WRK montmorillonite.

Multiple sorptive sites on natural illitic clays (e.g., frayed edge [FES], type II [TS], and planar sites [PS]) play an important role to diverse 137Cs immobilization in soil and aquifer environments. This study investigated the Cs sorption capabilities of 10 natural illitic clays at ranged Cs concentrations (1 ×10−7 to 1×10−3 mol·L−1) under various competing potassium concentration (distilled water to 1×10−1 mol·L−1). Additionally, multisite cation exchange model was performed to evaluate the best-fit sorption model and optimize the sorption capacities and affinities of multiple sorptive sites for Cs. Here, the experimental Cs sorption isotherms varied among 10 illtic clays, indicating different sorption capacities of Cs on illitic clays. The best-fit sorption model exhibited that variable Cs sorption of 10 illitic clays was significantly related to the sorption capacities at the FES (1.76 × 10−5 to 1.12×10−4 eq·kg−1), TS (1.59×10−3 to 9.76×10−3 eq·kg−1), and PS (2.14×10−2 to 1.51×10−1 eq·kg−1), respectively. The FES predominantly contributed to Cs sorption at low aqueous concentrations, whereas the TS and PS sorbed Cs at high concentrations. These sorption capabilities of multiple sorptive sites were correlated to illite contents and crystallinity of illitic clays, implicating that such parameters could be key factors to predict the Cs sorption for natural illitic clays in soil and aquifer environments. Finally, 1-D transport simulations represented that the severe Cs retardation occurred at low Cs concentration, implying that the FES predominantly affected to Cs transport in actual radioactive contamination sites (i.e., where low Cs concentration prevails), compared to the TS and/or PS.

Corrosion of copper (Cu) canisters is one of the important factors to ensure the safety of a deep geological repository site. This is because the corrosion of a canister may induce failure of the canister which can lead to a release of radionuclides into the environment. Corrosion of canisters for highlevel wastes is affected by the following multiphysics: thermal-hydraulics, transportation of chemical species, chemical reactions, and interface reactions. This research aimed to develop a multiphysics numerical model for the corrosion of spent nuclear fuel canisters for a deep geological repository in South Korea. The multiphysics model is based on MOOSE (Multiphysics Object-Oriented Simulation Environment) which uses a finite element method. In the multiphysics model, the following multiphysics are coupled and solved together for a deep geological repository design of South Korea: interface redox reactions, porous flow, and heat transport in porous flow. The proposed model was validated with experimental data before being applied to a KAERI reference disposal unit. It was found that the corrosion potential of a Cu canister shows an uneven distribution of corrosion potential along with the surface. In addition, top, bottom, and side surfaces of the canisters show a different lifetime and corrosion potential. Important redox reactions for corrosion are changed along with time from a reduction of O2 and anodic dissolution of Cu by Cl− to sulfidation of Cu and reduction of water. The proposed model will be coupled with some important chemical reactions in engineering buffers and will be the base for the understanding of the behavior of Cu canisters in the KAERI reference disposal unit.

The International Atomic Energy Agency recommends the deep geological disposal system as one of the disposal methods for high-level radioactive waste (HLW), such as spent nuclear fuel. The deep geological disposal system disposes of HLW in a deep and stable geological formation to isolate the HLW from the human biosphere and restrict the inflow of radionuclides into the ecosystem. It mainly consists of an engineered barrier and a natural barrier. Safety evaluation using a numerical model has been performed primarily to evaluate the buffer’s long-term stability. However, although the gas generation rate input for long-term stability evaluation is the critical factor that has the most significant influence on the long-term hydraulic-mechanical behavior of the buffer, in-depth research and experimental data are lacking. In this study, the gas generation rate on the interface between the disposal canister and the buffer material, a component of the engineered barrier, was mainly studied. Gas can be generated between the disposal canister and the buffer material due to various causes such as anaerobic corrosion of the disposal canister metal, organic matter decomposition, radiation decomposition, and steam generation due to high temperature. The generation of gas in such a disposal environment increases the pore gas pressure in the buffer and causes internal cracks. The occurred cracks increase the intrinsic permeability of the buffer, which leads to a decrease in the primary performance of the buffer. For this reason, it is essential to apply the appropriate gas generation rate according to the disposal condition and buffer material for accurate long-term stability analysis. Therefore, the theoretical models regarding the estimation of gas generation were summarized through a literature study. The amount of gas generated was estimated according to the disposal environment and material of the disposal canister. It is expected that estimated values might be used to estimate the long-term stability analysis of buffer performance according to the disposal condition.

Deep geologic repositories (DGR) are designed to store spent nuclear fuel and to isolate it from the biosphere for an extended period of time as long as millions of years. The long-term performance of the DGR replies on the performance of the natural geologic barriers after the end of the lifetime for the engineered barrier systems. Typically, multiple analytical and numerical models are used to analyze and ensure the safety of the repositories along both engineered and natural barrier systems. Despite the immense advancement in computing power and modeling techniques over the last few decades, a series of models and their linkage often require many simplifying assumptions in this safety assessment. The degree of the reliability and confidence of the safety analysis is thus highly dependent on the validity of those tools used. Considering the significance of the DGR performance and public attention, the highest level of quality control is necessary for the models employed in the assessment. The performance of the ultimate long-term geologic barrier is determined by the expected travel time of the radioactive species of interest, the level of their dilution or radioactivity at compliance areas, and the uncertainty associated with them. As the species of interest can be carried away from the repository location by groundwater flow, the travel time is determined by groundwater velocity along the flow path from source to biosphere while the dilution is a function of the decay and production rates as well as the diffusion and dispersion. Due to the time scale and the complexity of the physicochemical processes and geologic media involved, the models used for safety evaluation will need to become more and more comprehensive, robust, and efficient which is difficult to achieve in principle. They will also need to be transparent and flexible to satisfy the regulatory quality control requirements. This study thus attempts to develop an accessible, transparent, and extensible integrated hydrologic models (IHM) which can be widely accepted by the regulators as well as scientific community and thus suitable for current and future safety assessment of the DGR systems. The IHM can be considered as a tool and a framework at the same time when it is designed to easily accommodate additional processes and requirements for the future as it is necessary. The IHM is capable of handling the atmospheric, land surface, and subsurface processes for simultaneously analyzing the regional groundwater driving force and deep subsurface flow, and repository scale safety features, providing an ultimate basis for seamless safety assessment in the DGR program. The applicability of the IHM to the DGR safety assessment is demonstrated using simple illustrative examples.

To reduce the environmental burden caused by the disposal of spent nuclear fuel and maximize the utilization of the repository facility, waste burden minimization technology is currently being developed at the Korea Atomic Energy Research Institute (KEARI). The technology includes a nuclide management process that can maximize disposal efficiency by selectively separating and collecting major nuclides in spent nuclear fuel. In addition, for efficient storage facility utilization, the short-term decay heat generated by spent nuclear fuel must be removed from the waste stream. To minimize the short-term thermal load on the repository facility, it is necessary to separate heat generating nuclides such as Cs-137 and Sr-90 from the spent fuel. In particular, Sr-90 must be separated because it generates high heat during the decay process. KAERI has developed a technology for separating Sr nuclides from Group II nuclides separated through the nuclide management process. In this study, we prepared Sr ceramic waste form, SrTiO3, by using the solid-state reaction method for long-term storage for the decay of separated Sr nuclides and evaluated the physicochemical properties of the waste form. Also, the radiological and thermal characteristics of the Sr waste form were evaluated by predicting the composition of Sr nuclides separated through the nuclide management process, and the estimation of centerline temperature was carried out using the experimental thermal data and steady state conduction equation in a long and solid cylinder type waste form. These results provided fundamental data for long-term storage and management of Sr waste.

Through constructing statistical fracture network model based on discrete element method, the evolution characteristics of the fracture aperture had been directly simulated and evaluated caused by redistributed stress after the borehole excavation. This study focuses on the size effect of the discrete element method for the analysis of the effective distance of fracture aperture change after the borehole excavation. A two-dimensional trace-type domain with a maximum size of 1.1 m2 was created using a discrete fracture network with stochastic information of KURT. A total of eight domains with different sizes were constructed from the largest domain area to the 0.4 m2 analysis area. The aperture change ratio which can be depending on the domain size was examined. The ratio was investigated by comparing the aperture size before and after the simulation of borehole excavation. In addition, the effective range of aperture changes was analyzed by comparing the re-distribution distance from the center of the borehole. Based on dimensional analysis, input variables (borehole radius, occurrence distance of aperture changes, domain size) were modeled using exponential distribution form. Through the analysis model, two dimensionless variables were derived to investigate the expected distance of the aperture changes and appropriate DFN domain size for simulating bole excavation. As an application example of the 3-inch borehole simulation, the analysis model predicted that the range of aperture changes could occur within a radius of about 0.98 m from the borehole center, and the suitable size of the model had been inferred as about 5 × 5 m for minimizing the domain size effect.

The mechanical, hydraulic, thermal, and chemical properties of the subsurface can have a significant effect on the long-term performance of an underground facility. Therefore, it is important to accurately estimate the aquifer properties in order to predict the groundwater flow and solute transport and thus ensure the stability and safety of a high-level radioactive waste disposal. Using heat as a tracer has become a popular tool for the subsurface characterization. Recent studies have demonstrated that heat tracing is an effective approach to quantify both hydrogeological and thermal subsurface properties. However, most studies in natural conditions assume the local thermal equilibrium (LTE) between the solid and fluid phases, ignoring heat exchange between them. The LTE assumption has not yet been verified by experiments. This work investigates the validity of the LTE assumption by performing the laboratory tracer tests using both solute and heat in a porous medium under natural groundwater flow velocities (Reynolds number, Re < 0.37). The experimental results showed that the LTE assumption can be violated even under natural groundwater flow conditions. The violation of LTE (LTNE) had a significant impact on mechanical dispersion, whereas its effect on velocity was negligible. These results provide the first experimental evidence for LTNE effects in natural conditions. Therefore, it is necessary to consider LTNE effects especially when the mechanical dispersion is evaluated using heat tracing.

Various diffusion experiments using geologic media have been carried out and it is often assumed that aqueous diffusion is the dominant transport mechanism. However, in some cases diffusive migration has been much faster than predicted in the model simulation. To explain such results surface diffusion of sorbing species was invoked. Experimental results were generally open to interpretation but possible existence of surface diffusion, whereby sorbed radionuclides could potentially migrate at much enhanced rates, necessitated investigation. The potential for surface diffusion of some sorbing nuclides on through-diffusion experiments using domestic rocks was examined. The apparent diffusion coefficients for sorbing cations were determined from their steady-state diffusion flux through rocks disks, while effective and pore diffusion coefficients were obtained with non-sorbing tracers through the same rocks. Diffusive transport models through domestic granites and granodiorites based only on pore diffusion did not often described adequately for sorbing cations. Thus, surface diffusion should be considered. Then what was the most important measure to estimate surface diffusion? As far as we examine, the sorption reversibility provides a hint of surface diffusion. The reversible sorbing species, for example, Sr, has a remarkable surface diffusion contribution, whereas surface diffusion has a relatively small contribution for irreversibly sorbing species such as Cs and Am under domestic experimental conditions.

Geologic disposal of high-level radioactive waste is considered the most effective method to isolate high-level radioactive waste from the biosphere. A high-level radioactive waste repository is designed to be placed at a deep depth and generally consists of canisters, buffer material, and host rock. In the disposal system, the heat from the canister occurs for millions of years due to the long half-life of the high-level radioactive waste, and the heat induces vaporization of groundwater in the buffer material. The resaturation process also occurs due to groundwater inflow from the host rock by the hydraulic head and capillarity. The saturation variation leads to the heat transfer and multi-phase flow in the buffer material, and thermal pressurization of groundwater due to the heat affects the effective stress change in the host rock. The stress change can make the porosity and permeability change in the flow system of the host rock, and the flow system affects the nuclide migration to the biosphere. Therefore, it is crucial to understand the complex thermo-hydro-mechanical-chemical (THMC) coupled behavior to secure the repository’s long-term safety. DECOVALEX is an international cooperating project to develop numerical methods and models for predicting the THMC interactions in the disposal systems through validation and comparison with test results. In Task C of DECOVALEX-2023, nine participating groups (BGR, BGE, CAS, ENSI, GRS, KAERI, LBNL, NWMO, Sandia) models the full-scale emplacement (FE) experiments at the Mont Terri underground rock laboratory and focus on understanding pore pressure development, heat transfer, thermal pressurization, vaporization and resaturation process in the disposal system. In the FE experiment, three heaters generated heat with constant power for five years at a 1:1 scale in the emplacement tunnel based on Nagra’s reference repository design. KAERI used OGS-FLAC3D for the numerical simulation, combining OpenGeoSys for TH simulation and FLAC3D for M simulation. We generated a full-scale three-dimensional numerical model with a dimension of 100 by 100 by 60 meters. The pressure and temperature distribution were well simulated with the host rock's anisotropy. Based on the capillarity, we observed vaporization and resaturation in the bentonite under the twophase flow system. We plan to compare the simulation results with the field data and investigate the effect of input parameters, including thermal conductivity and pore compressibility affecting the thermal and flow system.

With the increase of temporarily-stored spent radioactive fuels, there is an increasing necessity for the safe disposal of high-level radioactive waste (HLW). Among various methods for the disposal of HLW, a deep geological disposal system is adapted as a HLW disposal strategy in many countries. Before the construction of a repository in deep geological condition, a performance assessment, which means the use of numerical models to simulate the long-term behavior of a multi-barrier system in HLW repository, has been widely performed to ensure the isolation of radionuclides from human and related environments for more than a million years. Meanwhile, Korea Atomic Energy Research Institute (KAERI) is developing a process-based total system performance assessment framework for a geological disposal system (APro). To improve the reliability of APro, KAERI is participating in DECOVALEX-2023 Task F, which is the international joint program for the comparison of the models and methods used in deep geological performance assessment. As a final goal of Task F, the reference case for a generic repository in fractured crystalline rock is described. The three-dimensional generic repository is located in a domain of 5 km in length, 2 km in width, and 1 km in depth, and contains an engineering barrier system with 2,500 deposition holes in fractured crystalline rock. In this study, a numerical simulation of the reference case is performed with COMSOL Multiphysics as a part of Task F. The fractured crystalline rock is described with the discrete fracture matrix (DFM) model, which expresses major deterministic fractures explicitly in the domain and minor stochastic fractures implicitly with upscaled quantities. As an output of the numerical simulation, fluid flow at steady-state and radionuclide transport are evaluated for ~106 years. The result shows that fractures dominate the transport of radionuclides due to much higher hydraulic properties than rock matrix. The numerical modeling approaches used in this study are expected to provide a basis for performance assessment of nuclear waste disposal repository located in fractured crystalline rock.

Deep geological repository (DGR) has been considered as a globally accepted strategy to dispose high-level radioactive wastes. During long storage periods of 100,000 years, uranium (U) could be migrated through fractures in deep granite aquifers and interact with indigenous bacteria under anaerobic condition. Anaerobic bacteria can reduce U(VI) and further precipitate in the form of U(IV)-oxide minerals by transferring electrons through c-type cytochrome. In this point of view, a comprehensive understanding of uranium-microorganisms interaction is necessary to guarantee the safety of high-level radioactive waste disposal. Although diverse bacterial communities are present in DGR environment, a number of studies have been focused on some model bacteria, such as Desulfovibrio, Geobacter, and Shewanella spp.. In this study, indigenous bacterial community in deep granitic groundwater at 234–244 m was inoculated to sterile uranium-contaminated granitic groundwater amended with 20 mM of sodium acetate, and then incubated under anaerobic condition for 12 weeks. Bio-reduction of U(VI) to U(IV) by indigenous bacteria in uranium-contaminated groundwater was investigated during whole operation period. Initial U(VI) concentration of 885.4 μg·L−1 gradually decreased to 586.1 μg·L−1, resulting in approximately 33.8% of aqueous U(VI) removal efficiency. Oxidation-reduction potential (ORP) value was gradually decreased from 175.4 mV to –243.0 mV after the incubation of 12 weeks. The decrease in ORP value was attributed to the presence of aerobic bacteria and facultative anaerobic bacteria in indigenous bacterial community. The shift in bacterial community structure was observed by 16S rRNA gene high-throughput sequencing analysis. Proteobacteria (55.6%), Firmicutes (24.1%), Actinobacteria (5.5%), and Bacteroidetes (5.4%) were dominant in initial indigenous bacterial community, while Proteobacteria (94.8%) was found to be the only abundant phylum after the reaction. In addition, great increase in the relative abundance of sulfate-reducing bacteria (SRB) was observed: the relative abundance of SRB increased from 11.4% to 44.3% after the reaction. This result indicates that the SRB played a key role in the removal of aqueous U(VI). This finding shows the potential of aqueous U(VI) removal by indigenous bacteria in DGR environment.

A geological repository system consists of a disposal canister with packed spent fuel, buffer material, backfill material, and intact rock. Among these, the bentonite buffer is one of the most important components to assure the safe disposal of high-level radioactive waste (HLW). As the bentonite buffer is installed as a block type, it is important to fabricate homogeneously. Generally, floating die method and cold isostatic press (CIP) method are used to fabricate bentonite blocks. In this paper, two bentonite blocks were produced using float die method at first, and CIP method was additionally applied to just one block. After that, several samples were cored from two blocks. The dry density and water content of several samples produced from two blocks were measured.