Hierarchical Holographic Modeling for Risk Identification in Complex Radioactive Waste Disposal Systems, 20-R9567

Printer Friendly Version

Principal Investigators
Olufemi Osidele
Osvaldo Pensado
Jude McMurry
Sitakanta Mohanty

Inclusive Dates:  10/03/05 – Current

Background - Risk analysis models for radioactive waste disposal systems generally offer limited flexibility to accommodate changes in regulations, design, and scientific knowledge. One reason for this limitation is the prior screening of risk factors whereby plausible risk scenarios are excluded because their likelihood of occurrence or perceived consequences are low. In addition, several alternative conceptual models are often considered in integrating the complex physical and chemical processes associated with radioactive waste disposal systems. Thus, whenever regulatory or technical criteria change, model revision takes considerable effort in reassessing the risk scenarios and conceptual models. Moreover, the utility of the revised risk analysis models — especially for prioritizing subsequent regulatory and technical activities — may not be determined until revisions are complete and extensive analyses are conducted.

Approach - The goal of this project is to develop a capability for identifying the important technical aspects of risk analysis models at the conceptual stages. This would potentially relieve efforts expended in revising risk analysis models while providing information to support early decision-making and prioritization of regulatory and technical program activities. The project draws on ideas from decision theory and analysis, including techniques such as Hierarchical Holographic Modeling and the Analytic Hierarchy Process. The overall approach is to conduct a comparative assessment of hypothetical case studies of geologic radioactive waste repository systems. Risk factors obtained from expert judgment are organized and depicted as a hierarchical model. Based on available technical information, a related conceptual model of risk is identified. Matrix data structures and algorithms are used to map the risk factors to the conceptual model. This mapping is then applied in ranking the importance of the technical elements of the conceptual model. A comparison of the rankings obtained from the case studies is used to infer the flexibility of the conceptual model to implement changes in regulatory and technical criteria.

Accomplishments - A case study of a Northern Temperate Repository (NTR) has been developed. The NTR concept describes a geologic repository in saturated rock located in a climate representative of northern America and Europe and includes features, events, and processes being considered in radioactive waste management programs worldwide. Also, a decision-analysis procedure has been developed and tested on the NTR concept. This new procedure applies a modified principal components analysis to numeric scores computed by mapping the risk factors to the conceptual model. The eigenvalues and eigenvectors resulting from this analysis provide a set of metrics for ranking the technical topics considered in the conceptual model. Numerical experiments have demonstrated that the procedure effectively integrates risk perception (via the risk factors) with technical evidence (via the conceptual model) in judging the importance of technical issues that are considered in risk assessments for radioactive waste disposal systems.

2006 Program Home