the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 18 Jun 2024
Numerical uncertainty identification, classification and quantification in radioactive waste management
Abstract. The work package “Uncertainty Management multi-Actor Network – UMAN” within EURAD European Joint Programme on Radioactive Waste Management was dedicated to the management of uncertainties potentially relevant to the safety of different radioactive waste management stages and programs. One important goal there was to compile, review, compare and refine strategies, approaches and tools for the management of uncertainties in the safety analysis and the safety case that are being used, planned to be used or being developed in different countries. This paper presents major findings from the UMAN deliverable D10.3 "Uncertainty identification, classification and quantification" that addresses approaches to identify and classify uncertainties that might be of relevance in the various stages of radioactive waste management as well as on the quantification of numerical uncertainties. The section on methodology compares Bottom-up and Top-down strategies, describes which sources were used for the report as input: expert elicitation (here primarily based on a respective questionnaire send out to UMAN participants) and literature survey. It then advices on how uncertainties can be structured to pave the way to a comprehensive assessment of numerical uncertainties: fishbone diagrams and tables for uncertainty characteristics. Results support the identification of uncertainties with high relevance for RWM. Nine suitable categories are identified; the uncertainties are then grouped (including representative examples utilizing fishbone diagrams and tables) according to the occurrence by system phenomena, following the themes and subthemes of the EURAD Roadmap. The last part is treating with the evaluation as well as quantification of uncertainties. The paper closes with recommendations aimed at future research directions for parameter uncertainties. Finally, it provides definitions for some terms frequently used (uncertainty in general, parameter uncertainty, uncertainty models, and aleatory vs. epistemic uncertainties) in a glossary.
- Preprint
(1316 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
CC1: 'Comment on sand-2024-1', Fabien Magri, 03 Jul 2024
Methodology:
The method is based on bottom-up and top-down strategies and literature review. It would be useful to shortly state how these methods were specifically applied to ensure a comprehensive identification and classification of the uncertainties. How was the filtering and prioritising the expert responses and literature sources done?
Numerical uncertainties:
The paper states that it deals only with numerical uncertainties. It explicitly excludes scenario uncertainties due to future state changes. It is known that that future state changes can significantly affect numerical parameters. So how is the exclusion of these scenario uncertainties justified and what are the risks / limitations? Please add a comment in the MS.
Quantification techniques:
The discussion of quantifying uncertainties mentions various approaches such as fishbone diagrams and tables. A few information on the uncertainties quantification would be useful, e.g. techniques, statistical, computational methods and most importantly validation to ensure their reliability and accuracy in the context of radioactive waste management.
Citation: https://doi.org/10.5194/sand-2024-1-CC1 -
AC1: 'Reply to revieweres comments on sand-2024-1', Vinzenz Brendler, 07 Nov 2024
Dear editors,
This is our reply to the review of the article sand-2024-1 entitled “Numerical uncertainty identification, classification and quantification in radioactive waste management” by Vinzenz Brendler and Solveig Pospiech (both Helmholtz-Zentrum Dresden-Rossendorf e.V., Bautzner Landstrasse 400, 01328 Dresden). We have copied all remarks from the three reviewers into this document, together with our responses (all in GREEN). The changes / additions discussed have been done in the manuscript accordingly.
We are thankful for the comments and are confident that the manuscript has now reached a higher quality.
Best regards from Rossendorf,
Vinzenz Brendler and Solveig Pospiech
-
AC1: 'Reply to revieweres comments on sand-2024-1', Vinzenz Brendler, 07 Nov 2024
-
RC1: 'Comment on sand-2024-1', Anonymous Referee #1, 07 Jul 2024
The paper provides a summary of some results of the UMAN project. This is in principle fine, but the contribution of the paper should be stated more clearly. In particular, large parts of the text in most sections are just a copy from the UMAN report without marking them as a quote: For example, L293 - L355 (two pages of text), L196-L246, L262ff, L166-185, large parts (in fact most of the text) of the Outlook (section 5) are also just copied from the UMAN report. This has to be changed! Either mark them as direct quotes, or put them in boxes as extracts from the report which you discuss in the manuscript.L59: The term "numerical dispersion" has a very specific meaning, which is not the one intended here. Better call it "effects of computational imprecision". In the same paragraph, it is said that such imprecision, e.g. due to convergence criteria of iterative algorithms, are dealt with validation, qualification and verification. Validation and verification address two very distinct aspects in modeling (physical adequacy of the model and mathematical correctness of the implementation). Please reformulate to remain precise.The definition of epistemic uncertainty is here put into the context solely of numerical models. In the beginning of the paper, it is said that the study is limited to "numerical uncertainty", where the term numerical is used in the sense of data with numerical values. This is a broader category than numerical (in the sense of computational) modeling. The overall wording in this sense might need a bit of sharpening to avoid confusion. The definition of epistemic uncertainty can be broadened accordingly. One could also discuss that the practical distinction into aleatory and epistemic (with a view on reducibility) can be context-specific (see, e.g., "Special Workshop on Risk Acceptance and Risk Communication Aleatory or epistemic? Does it matter?" by Der Kiureghian and Ditlevsen). In a RWM setting, generic studies are very different from site-specific characterization, for example, or URLs can be explored differently than a proper repository.Something seems to be wrong in Tab. 2: there's a column of large amount of knowledge and one for high amount of knowledge. Probably the latter should be low amount of knowledge.L237: How would such an internal consistency be ensured? What is best practice in this context?The authors discuss fishbone diagrams and show two examples. Explain, how they are linked in a technically (if they are). In the present example, sorption and RN migration are two linked diagrams. How does one move from one to the other? Is this system set up as a dynamic data base or, at the moment, a collection of examples that just shows that in principle this is a practical way of ordering uncertainty information.The authors have a background in sorption, reactive transport, radiochemistry, etc. and show examples of the BU approach in the form of fishbone diagrams. Can you add a similar example for the TD approach coming from RN assessment? That would add valuable context to the general concepts discussed in the paper.Typos:L19: "treating with" -> "treating" or "dealing with"L71: "known knows" -> "known knowns"L104: pdfs -> probability density functionsL 191: effect -> affectCitation: https://doi.org/
10.5194/sand-2024-1-RC1 -
AC1: 'Reply to revieweres comments on sand-2024-1', Vinzenz Brendler, 07 Nov 2024
Dear editors,
This is our reply to the review of the article sand-2024-1 entitled “Numerical uncertainty identification, classification and quantification in radioactive waste management” by Vinzenz Brendler and Solveig Pospiech (both Helmholtz-Zentrum Dresden-Rossendorf e.V., Bautzner Landstrasse 400, 01328 Dresden). We have copied all remarks from the three reviewers into this document, together with our responses (all in GREEN). The changes / additions discussed have been done in the manuscript accordingly.
We are thankful for the comments and are confident that the manuscript has now reached a higher quality.
Best regards from Rossendorf,
Vinzenz Brendler and Solveig Pospiech
-
AC1: 'Reply to revieweres comments on sand-2024-1', Vinzenz Brendler, 07 Nov 2024
-
RC2: 'Comment on sand-2024-1', Anonymous Referee #2, 12 Jul 2024
According to the abstract, the paper presents major findings from the UMAN deliverable "Uncertainty identification, classification and quantification". As these results have already been published, the authors' own further reflections are of particular importance for the scientific value of the publication. Where results from UMAN are presented and where the authors' own reflections come into play should therefore be more clearly labelled.
The glossary represents a significant added value of the paper. In addition, the glossary should explain what understanding of risk the publication is based on: Is risk only associated with negative effects that need to be assessed or also with positive effects? What role does the probability component play? Against this background table 1 should be reconsidered.
The article is explicitly focussed on numerical uncertainties, but - as mentioned in chapter 4 - does not deal with the mathematical handling of uncertainties. It would therefore be appropriate to further clarify the question of whether numerical uncertainties can be clearly distinguished from non-numerical uncertainties and what advantages the focus on numerical uncertainties brings for the publication. The authors declare that they want to concentrate on parameter uncertainties. In the text, however, much space is given to other uncertainties, in particular model uncertainties.
While there is general reference to RWM at the beginning of the publication, later statements concentrate mainly on the (long-term) safety of deep geological repositories. A clearer definition of the scope of the paper would be helpful.
The statement in line 64/65 "Uncertainty creates an uncomfortable position for a large part of the public (anxiety)" is not clearly supported by the social science literature. If the topic is not dealt with in more detail in the paper, this sentence should therefore at least be supported by a reference to the literature.
At the beginning of section 2.1, it is not clear what the cost-benefit ratio refers to. "Relevance can be inferred from priority for further investigation" can be read as a circular argument.
Table 2 should be verified. What is the difference between a large and a high amount of knowledge?
In line 418/419 there is the statement "Finally, further research activities should elaborate possible ways to treat uncertainties resulting from human behaviour numerically, taking into account that their impact on the safety could be significant." Was this statement made against the background of the relevant social science literature? A corresponding reference would be valuable.
Citation: https://doi.org/10.5194/sand-2024-1-RC2 -
AC1: 'Reply to revieweres comments on sand-2024-1', Vinzenz Brendler, 07 Nov 2024
Dear editors,
This is our reply to the review of the article sand-2024-1 entitled “Numerical uncertainty identification, classification and quantification in radioactive waste management” by Vinzenz Brendler and Solveig Pospiech (both Helmholtz-Zentrum Dresden-Rossendorf e.V., Bautzner Landstrasse 400, 01328 Dresden). We have copied all remarks from the three reviewers into this document, together with our responses (all in GREEN). The changes / additions discussed have been done in the manuscript accordingly.
We are thankful for the comments and are confident that the manuscript has now reached a higher quality.
Best regards from Rossendorf,
Vinzenz Brendler and Solveig Pospiech
-
AC1: 'Reply to revieweres comments on sand-2024-1', Vinzenz Brendler, 07 Nov 2024
-
RC3: 'Comment on sand-2024-1', Anonymous Referee #3, 13 Aug 2024
General comments
In this paper, the results of the EURAD project UMAN are described with a focus on uncertainties in numerical models. The paper effectively highlights the uncertainties and provides options for managing them, such as using a fishbone structure. However, the paper seems to rely too heavily on a EURAD report, which has essentially been adapted with only minor adjustments to fit the format of a paper. For instance, the outlook states that “A thorough survey in other EURAD deliverables issued so far is recommended for indicators that may help to rank the importance of FEPs and their associated uncertainty.” However, I believe much more literature exists on this topic beyond EURAD deliverables. Additionally, the paper notes that “A revision of ranking and order of various uncertainties might also be necessary for the current EURAD Themes definition.” This again focuses solely on EURAD, while there is likely a wealth of other relevant papers on this subject. Therefore, I believe the paper should reference a broader range of other papers and reports. It is also recommended that the authors review the text, as in some instances, they still refer to “this report” instead of “this paper.”
It is also unclear whether the uncertainties discussed in this paper pertain specifically to the safety assessment or safety analysis of a GDR, or if they are related to numerical modeling in general. I suspect it is the former, but the paper does not make this clear. This distinction should be clarified, and an introduction to the subject might be necessary. Additionally, I believe the paper would benefit from including examples with references. For instance, in section 2.1, the authors refer to textbooks that might provide an in-depth discussion, but they do so without citing specific examples. In other sections, they mention certain methods that can be used. Including a reference to a book or other source in the text would be very beneficial for readers who do not have a background in this area. Lastly, I found that there was little on the different stages of a program and how that would influence the uncertainty.
Specific comments
Abstract
R9: Safety analysis is mentioned here but later on in the introduction, it is also referred to as safety assessment. Better use one word and maybe explain what it means.
R14: I am wondering if “for the report” is correct here. In general, I would refer to it as paper.
R17: RWM is not defined.
- Introduction
Overall the introduction looks good, but it can be shortened by focusing on what is done in the paper rather than describing in length was is not done. This will help to make it clearer what the paper is about and it will reduce the length of the introduction and make it clearer.
R25: Safety assessment of the DGR?
R26: Here, radioactive waste management for the first time in the introduction, so maybe introduce the abbreviation here rather than later on.
R29:Is the safety assessment the same as the safety analysis in this context?
R31: I am a bit wondering how various stages of radioactive waste management influence the safety analysis.
R63: Deliverable or paper?
R68: What kind of decisions?
2.0 methodology
I like that the authors start with a list of definitions. I do, however, wonder how their definitions are compared to others.
R103: pdfs?
R151: The abbreviation WMO is already defined somewhere else.
R117: Both strategy and structure are red underlined. Is that for a reason?
R130: It is mentioned that they are used for a multitude of complex applications in science and society, but could some examples be given?
R186: Again report and what is exactly meant by sources being identified and used in this report are listed in the references section.
R205: Sea level changes instead of water transgression?
R244: I don’t think guess is good to say here: it is always based on something.
R251: Design concept of the DGR?
R276: What are first two columns in table 4 in this paper use for? Do they serve any purpose?
3.0 Categorization of uncertainties
R327: The abbreviation FEP is not explained here (or earlier; only later).
R343: EURAD is currently almost finished. So, is there no final version of the EURAD roadmap?
4.0 Categorization of uncertainties
R375: Not the topic of this report? Paper probably.
R376: RW?
5.0 Outlook: What are the gaps in understanding, quantification and processing of uncertainties?
R394: Task 3 deliverables?
R410: Why would it be challenging to convince modellers and programmers? I think most of them want to but cannot do it due to time.
Citation: https://doi.org/10.5194/sand-2024-1-RC3 -
AC1: 'Reply to revieweres comments on sand-2024-1', Vinzenz Brendler, 07 Nov 2024
Dear editors,
This is our reply to the review of the article sand-2024-1 entitled “Numerical uncertainty identification, classification and quantification in radioactive waste management” by Vinzenz Brendler and Solveig Pospiech (both Helmholtz-Zentrum Dresden-Rossendorf e.V., Bautzner Landstrasse 400, 01328 Dresden). We have copied all remarks from the three reviewers into this document, together with our responses (all in GREEN). The changes / additions discussed have been done in the manuscript accordingly.
We are thankful for the comments and are confident that the manuscript has now reached a higher quality.
Best regards from Rossendorf,
Vinzenz Brendler and Solveig Pospiech
-
AC1: 'Reply to revieweres comments on sand-2024-1', Vinzenz Brendler, 07 Nov 2024
Dear editors,
This is our reply to the review of the article sand-2024-1 entitled “Numerical uncertainty identification, classification and quantification in radioactive waste management” by Vinzenz Brendler and Solveig Pospiech (both Helmholtz-Zentrum Dresden-Rossendorf e.V., Bautzner Landstrasse 400, 01328 Dresden). We have copied all remarks from the three reviewers into this document, together with our responses (all in GREEN). The changes / additions discussed have been done in the manuscript accordingly.
We are thankful for the comments and are confident that the manuscript has now reached a higher quality.
Best regards from Rossendorf,
Vinzenz Brendler and Solveig Pospiech
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 380 | 110 | 36 | 526 | 19 | 15 |
- HTML: 380
- PDF: 110
- XML: 36
- Total: 526
- BibTeX: 19
- EndNote: 15
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1