Repository systems

There are inherent scale limitations in the time and space dimensions covered by laboratory studies. The applicability of the near field geochemical models derived from laboratory observations have to be applied to long-term, large-scale situations like the ones involved in the safety assessment of nuclear waste repositories. Hence, there is a need to test the models developed from laboratory investigations in field situations that are related to the ones to be encountered in repository systems. 

There are three issues often raised with regard to the mechanical behavior of the canister-salt-repository system when HLW is the emplaced waste form. 

Another alternative is to prevent modifications to the records at the data collection and acquisition domains. The permanent repository system may be located at the business logistic systems level.5 At this level, the electronic records are either maintained while in active use, or are maintained off-line when their use is less frequent. This setting must be evaluated and, preferably, documented as part of a company wide strategy. The selected settings require procedural and technological controls. 

Given the current desire of management to be able to mine data and see trends across studies and the availability of PK/PD repository systems (which we discuss more fully later), the first and foremost operating procedure requirement in pharmacometrics is the definition of key metadata that describes the process flow. Metadata, from the information science perspective, is simply information that describes data that is, where the data goes, what the data is, and what possible... 

For systems such as PK/PD repositories, a broader view is needed. Such systems by definition are intended to exchange data with other systems and integrate with analytical tools such as those described earlier. In this case, the pharmacometrician needs to have a well established process in place and be able to document how such a repository system will be implemented to either augment or replace current manual processes. 

We also note that the approach of quantity over quality of documentation is preferred by many organizations. This course of action will lead to a general disillusionment with the validation process and should be avoided at all costs. A good installation qualification should fit on one or two sides of a page (three with boilerplate, if that is unavoidable). A user requirements specification (URS) may be no more than a paragraph. While it is plausible that a URS may turn out to be several hundred pages for an internally developed repository system, that would be an exception rather than the rule. If an FDA inspector arrives with a method to determine the mass of your documentation, rather than with a desire to view the processes that such paperwork documents, it will be time to find other sources of advice on validation. 

The important point to recognize here is that software systems such as repository systems are considered enterprise software. The implementation is not customizable to an individual s requirements or a department s needs. The architecture of the software is the vendor s impression of how a clinical pharmacology effort... 

Skagius K, Strom A,Wiborgh M (1995) The use of interaction matrices for identification, structuring and ranking of FEPs in a repository system. SKB Technical report 95-22, SKB, Stockholm, Sweden... 

Geological repository systems for safe disposal of spent nuclear fuels and radioactive waste... 

DEVILLERS, C., Repository System Integration and Overall Safety, this Seminar. McCOMBIE, C., Safety Assessment Methodology and Confidence Building Measures, this Seminar. 

DEVILLERS, C., "Repository System Integration and Overall Safety", these Proceedings. 

The overall safety of a deep repository relies on a multi-barrier system and a limited contribution from institutional control. The question of safety indicators is discussed, including the need for societal risk indicators. The process of determining the robustness of a repository system is presented with the selection of adverse situations and the provision for adequate lines of defence, starting at the site selection stage. Performance assessment principles, methodology and sensitivities are addressed and examples of the usefulness of sensitivity studies are shown. 

While the question of low level waste disposal has found practical answers in several countries, there is to-day no concrete experience available on deep geological disposal of high level or long lived waste. To compensate for die lack of eiqierience, a specially careful safety approach is deemed necessary, in particular at the stages of site selection and deep repository design studies, a major present-day issue. Deep geological repositories will be addressed in this presentation from the point of view of repository system integration and overall safety. 

In the following, we will discuss die issues diat we to-day consider as important for the safety authority to form its opinion on die overall safety of a deep geological repository system. This will introduce the next paper devoted to confidence-building measures in the quantification of safety. The next sections deal respectively with ... 

Within this scope, the fundamental safety objective of a deep repository system for the post-closure period is to protect future societies and the environment from the risk associated with the dispersion of and the eiqnisure to die radioactive material contained in the repository. 

IMPORTANT ISSUES FOR JUDGING THE POST-CLOSURE SAFETY OF A DEEP REPOSITORY SYSTEM... 

Note that the effort put into the quality of activities involved in working out its protection adds to the robustness of the repository system. This is particularly true ... 

Before licensing a repository system, the safety authority, on the basis of the safety demonstration provided by the applicant, will first of all carry out a qualitative analysis of the set of lines of defence proposed by the latter against a set of selected normal and abnormal situations. This analysis will relate to two aspects the comprehensiveness of the set of selected situations and the qualities of the lines of defence. 

The next step in the safety assessments consists in performance assessments relyii on modelling the repository and its boundary conditions. Performance assessments are the essential tool to understand, quantify and optimize the respective roles of the different lines of defence of a given repository system. The roles of the different barriers in that repository system can then be explained from the results of performance assessments. Indeed, the roles of the different barriers dq>end greatly on which radionuclide is considered. Therefore, results of performance assessments obtained with a specific radionuclide spectrum cannot be generalized to any other spectrum. 

On the other hand, specific scenarios corresponding to events of unpredictable probabilities can be developed to check the robusmess of the repository system to such events ("what if approach). Such specific scenarios may be introduced to represent extreme realizations of situations that generally fall into the "normal scenario. For example the ageing and progressive loss of effectiveness of shaft and access tuimel seals is assumed, due to lack of qualification, to prematurely allow for a short-circuit pathway in these seals this case is dealt with as an hypothetical scenario. Another example of a hypothetical scenario comes from the difficulty in very heterogeneous rock formations to correctly characterize the medium from a limited set of measurements this leads to introducing a hypothetical scenario corresponding to an undetected defect in the host formation. 

The overall safety of a deep repository system relies on a multi-barrier concept including the geological barrier. 

Preselection of sites on the basis of their lines of defence against the most adverse simations (fast underground water circulation, high seismotectonic instability) increases the probability of obtaining a safe repository system. However, engineered safety features could, to a certain extent, compensate for weaknesses in the site. 

Quality assurance and peer reviews are inq>oitant aspects in assessing the robustness of a repository system. 

Performance assessments are, through scenario development and modelling, necessary to understand, quantify, optimize and assess a repository system both as concerns its intrinsic robusmess and from the point of view of the dependence of the safety demonstration to uncertainties. Performance assessments also provide input for ... 

As pointed out earlier, models are needed in performance assessment to simulate the behaviour of repository systems because of the long time periods and large spatial scales that are of interest. The modelling process itself consists of a series of basic steps (see Fig. 1), each of which requires input information or data ... 

---