2 METHODS FOR THE STRUCTURED ASSESSMENT OF MODEL UNCERTAINTY BY EXPERTS IN PERFORMANCE ASSESSMENTS OF RADIOACTIVE-WASTE REPOSITORIES

Citation
E. Zio et Ge. Apostolakis, 2 METHODS FOR THE STRUCTURED ASSESSMENT OF MODEL UNCERTAINTY BY EXPERTS IN PERFORMANCE ASSESSMENTS OF RADIOACTIVE-WASTE REPOSITORIES, Reliability engineering & systems safety, 54(2-3), 1996, pp. 225-241
Citations number
43
Categorie Soggetti
Operatione Research & Management Science","Engineering, Industrial
ISSN journal
09518320
Volume
54
Issue
2-3
Year of publication
1996
Pages
225 - 241
Database
ISI
SICI code
0951-8320(1996)54:2-3<225:2MFTSA>2.0.ZU;2-D
Abstract
The assessment of the performance of high-level radioactive waste repo sitories is based on the use of models for predicting system behaviour . The complexity of the system together with the large spatial and tem poral scales imposed by the regulations introduce large uncertainties in the analysis. The difficulty of validating the relevant models crea tes the need of assessing their validity by means of expert judgments. This paper addresses the problem of model uncertainty both from a the oretical and a practical point of view and presents two mathematical a pproaches to the treatment of model uncertainty that can assist the ex perts in the formulation of their judgments. The formal elicitation of expert judgments is investigated within the Technical Facilitator/lnt egrator (TFI) framework that has been proposed by the Senior Seismic H azard Analysis Committee,Within this framework, the mathematical formu lations for the treatment of model uncertainty are regarded as tools f or sensitivity analyses that give insights into the model characterist ics and are helpful in structuring the expert opinion elicitation proc ess itself. The first approach, referred to as the alternate-hypothese s formulation, amounts to constructing a suitable set of plausible hyp otheses and evaluating their validity. The second approach to model un certainty is referred to as the adjustment-factor formulation and it r equires that a reference model be identified and its predictions be di rectly modified through an adjustment factor that accounts for the unc ertainty in the models. Furthermore, both approaches require a clear u nderstanding of the distinction between aleatory and epistemic uncerta inties. The implications that these two formulations have on, and the issues that they raise in, the elicitation of expert opinions are expl ored. A case study of model uncertainty regarding alternative models f or the description of groundwater flow and contaminant transport in un saturated, fractured tuff is presented. (C) 1996 Elsevier Science Limi ted.