Repository design |
123 |
programme-specific requirements. Examples of further potentially relevant criteria include waste retrievability and repository ‘‘footprint’’. When using MAA methods, care should be taken not to mix ‘‘exclusion’’ criteria (i.e., those that result in on/off decisions) with criteria comparing the relative value of various options.
In the case where the number of sites is very small, it may be decided to include a step of further evaluation of design variants to determine if other suitable options exist – particularly if some or all of the sites are marginal.
The variety of requirements and constraints inevitably requires a structured approach for developing repository concepts. The repository design process in Japan, e.g., is structured to respond to the boundary conditions set by law and a decision to proceed with site selection via a volunteer process. A wide range of options is being considered, which maximises flexibility. Nevertheless, as volunteer sites are investigated, the range of options should be narrowed down rapidly to allow focusing of limited R&D resources. At present, the focus of work is on the clarification of design options with a high emplacement density and preparation of the information needed to identify relevant repository concepts during the process of literature study of volunteer sites.
5.4. Detailed design/specifications of subsystems
The repository, including the various EBS designs, can be controlled and optimised at all stages up to operation and closure (IAEA, 1999). There is a continual need to evaluate whether the level of understanding and extant modelling capabilities are fit for their intended purpose. Models used to support a decision-in-principle on the potential feasibility of a particular concept, e.g., may not need to be as fully developed as models used to support a licensing decision or final safety case.
Making design decisions that help to ensure that the EBS components will operate under conditions in which process interactions are relatively well understood could help to circumvent problems relating to limitations in scientific understanding. An example might be a design decision to separate the locations for ILW and HLW/SF disposal (see Figure 5.2b) and, thereby, circumvent the need to fully understand the potentially complex interactions that might occur in a co-disposal system as an alkaline plume migrating from a cementitious ILW vault begins to interact with bentonite-based buffer and waste in a HLW/SF vault. In addition to limitations in scientific understanding, a range of other uncertain factors affects programmatic decisions, including cost, time schedule and the capabilities of existing models.
Detailed repository design is specific to waste type and to geological environment. Regardless of the relative weightings of geosphere and engineered barriers in different disposal concepts, detailed characterisation of the geological environment and its evolution with time will certainly be required. The data requirements and the effort required to produce such data may vary considerably between different host rocks and hence ‘‘explorability’’ may be an important criterion for comparison of different options.
The dimensions of the various EBS components are established by an iterative analysis procedure. Key constraints for the HLW/SF case include:
the waste heat output;
the presence of void space in the fabrication waste package (from operational procedures);
124 |
H. Umeki |
the desired overpack lifetime, which will avoid requirements to quantify nuclide transport in the presence of significant thermal and water content gradients;
a desire to keep the bentonite at temperatures below design criteria, e.g., 100LC (to avoid possible loss of swelling properties, chemical alteration, etc.);
specified rock thermal and mechanical properties.
For L/ILW except hull and ends, heat generation is low and thus temperature effects on physical and chemical processes can be assumed to be negligible. The dimensions of the caverns are therefore constrained mainly by rock mechanical properties and requirements for operation (e.g., transport and emplacement of the container). Special attention should, however, be paid to material compatibility in repository design taking a range of waste types into consideration (see, e.g., comments in Chapter 8 on the hyperalkaline plume).
Geological repositories presently being considered have underground dimensions varying from a few square kilometres to as much as about 20 km2, depending on the types and inventories of waste, on its thermal output and on the repository design.
5.4.1. Near-field processes and design issues
The performance and design of an EBS are not dependent solely on the undisturbed ambient conditions and the initial properties of the EBS materials and repository host rock, but also on how those conditions and properties change as a result of repository construction, operation and waste emplacement. These repository-induced influences evolve in a complicated manner that depends on the type of waste, host rock formation and EBS design.
Host rock excavation, construction of an EBS and operation of a repository cause changes in the stress regime and the hydraulic conditions in the surrounding rock. The closure and sealing of the repository also affect the hydro-mechanical evolution of the system. Furthermore, the emplacement of heat-generating wastes modifies the temperature of the disposal system and this affects the THMC evolution of the system through a range of coupled processes. Medium-term (centuries) and long-term (millennia) changes in environmental conditions and alterations in the properties of the EBS materials and the surrounding rock may also influence the performance of the disposal system. The near-field environment is usually considered to include the EBS and that portion of the rock mass in which there would be significant changes (temporary or permanent) in THMC properties.
The manner in which the THMC conditions in the EBS and host rock are affected is further dependent on factors such as the rock excavation methods used, the duration of operations prior to closure, the heterogeneities and discontinuities in the host rock, the engineering measures taken to maintain excavation stability and reduce water inflow, the chemical and mechanical properties of the materials used for backfilling and sealing and the thermal load to which the rock is subjected.
To characterise the near-field environment, it may be necessary to consider a range of coupled processes, including the couplings between THMC processes, as well as possible associated impacts from biological and radiological processes. For example, heat from the waste can cause thermal gradients, which induce mechanical stresses and deformations, which can mobilise water, which can subsequently dissolve minerals.
Repository design |
125 |
Gas generation, due to e.g., biodegradation of organic materials in the waste, can lead to two-phase flow conditions in the buffer and induce additional water movement, mechanical stresses and chemical processes.
All of the key processes described above need to be considered to establish appropriate design requirements for the EBS, and to evaluate the performance of the EBS and the repository in a safety case. Modelling of coupled processes, experiments and field measurements in, e.g., underground research laboratories (URLs) are essential for the development of a better understanding of the key processes and for devising sensible and technically defensible design criteria based on scientific understanding rather than on arbitrary assumptions (see also comments in Chapter 8).
Radionuclide retardation in the buffer is important to safety in many repository concepts. The relative importance depends on, e.g., the thickness of the buffer and the expected lifetime of the container. In most HLW/SF disposal concepts, the buffer is designed to play a significant role in preventing the release of radionuclides in colloidal form by filtration. In contrast, the buffer has to allow transport of gas at a rate high enough to prevent ‘‘explosive’’ release, but not so high as to disrupt the other barrier roles, such as colloid filtration.
As gases such as H2, CH4 and CO2 will be produced by anaerobic corrosion of metals and by microbial degradation of organic matter in the ILW containers, the gas permeability of the buffer is more important for the cementitious L/ILW system design. The L/ILW tunnels contain waste solidified mainly with cement grout, surrounded by a relatively high-permeability cement-based mortar and repository porewater would be equilibrated to hyperalkaline pH levels. Such chemical evolution should be also taken into account, in particular for the co-disposal of L/ILW and HLW/SF, as noted above. L/ILW containing significant quantities of organic material (e.g., bitumen) and nitrate (from SF reprocessing1) may influence the chemical conditions in the repository and should be located to limit possible detrimental effects on the chemical conditions in other L/ILW tunnels and those for HLW/SF, for the case of co-disposal.
A geological repository is expected to remain open for many years, owing to the duration of disposal operations, and, in some cases, for an undefined period even after waste emplacement has been completed. National programmes may require that the wastes remain retrievable, perhaps also readily accessible, until decision-makers are comfortable to proceed to closure. An alternative retrievability option could be the partial or total closure of repository openings through the emplacement of the EBS in a reversible way. Depending on the disposal concept and on the closure strategy, at some time sections of the repository may be constructed and completely backfilled, while others would not be backfilled until final closure. Thus, over a period that might potentially last for many decades, it could be required that the repository remain stable and capable of maintenance and monitoring. This would require that the repository openings are excavated in stable blocks of rock and that any necessary support systems are designed to last for the necessary length of time (although with the possibility of remedial maintenance). For example, rooms filled with ILW might not be backfilled until just before closure, while some concepts envisage large volumes of such openings never being completely backfilled.
1 In most L/ILW repository concepts, the nitrate levels are relatively low but, in the Japanese design, a total of 10,000 tonnes of waste are expected by mid century (FEPC & JAEA, in preparation).
126 |
H. Umeki |
In many geological environments in which there is significant groundwater movement, the repository will need to be pumped and ventilated to keep it dry right up to the time of closure, although some sections that have been completely backfilled may start to resaturate as hydraulic gradients begin to re-establish themselves and groundwater moves into regions that had previously been drained. In all types of geological environment, during the open period exposed rock surfaces will interact with ventilation air passing through the facilities. Rock may dry out or be oxidised and some unlined excavations in sediments may crack and require support. If ventilation air were to flow from warmer to cooler sections of the repository, moisture would condense. Microbial activity will develop and flourish in regions where water carrying nutrients flows into excavations. Steel support systems will corrode and require maintenance, while cement surfaces may be partially carbonated by interaction with atmospheric carbon dioxide. All these processes will need to be monitored and their effects accounted for during the entire pre-closure period.
Following closure of a repository located below the water table, the groundwater regime will be progressively re-established and the whole system will resaturate. Any remaining oxygen in trapped air will react with the rock and EBS materials and the whole system will become chemically reducing. Microbes may play an important role in consuming the trapped oxygen. Rock stresses will re-equilibrate and lithostatic loads will be transferred onto parts of the EBS, particularly in weak host rocks that experience creep. The main determinants of the performance of the near-field in the majority of disposal environments will, however, be the content, movement and composition of groundwater in the rock immediately surrounding the EBS.
From the discussion above, many FEPs can influence the long-term performance of the EBS, depending on the particular waste types and site characteristics. Potentially important FEPs which should be taken into account in repository design can be summarised as:
Waste form dissolution/waste leaching rates.
Container corrosion rates/container leaching rate/container defects.
Buffer resaturation, swelling and long-term alteration.
EBS material degradation and chemical evolution in repository porewater.
Radionuclide transport in the buffer.
Gas generation in containers and transport through the buffer/backfill.
Slow changes in the geosphere (e.g., due to climate change, tectonic uplift, etc.).
It is difficult to conceive of a geological repository being constructed without some use of concrete2 and other common construction materials. In addition to making underground construction possible, these are needed to ensure a safe working environment for long periods of time. Concrete will be used for lining shafts and drifts or as shotcrete sprayed onto tunnel walls. In host rocks with significant groundwater flow, it is necessary to limit the water inflow into underground openings by sealing the fractures, especially fast-flowing features often related to ‘‘channelling’’, using cement-based grouts3. Cement is also needed to attach the rock bolts which might be necessary to provide additional stability to repository rooms.
2Alternatives, such as low-pH cements, geotextiles, etc. are currently under active consideration in many national programmes (see Chapter 10).
3Again, alternatives are actively under consideration, including bentonite, silica gel, resins, etc.