Electronic Thesis/Dissertation
 

Analysis of Process Criticality Accident Risk Using a Metamodel-Driven Bayesian Network

Open Access

Since the discovery of fission and subsequent first criticality of Chicago Pile-1,more than 60 criticality accidents have occurred throughout the world. These accidents are divided into two categories: those that occur during critical experiments or operations with research reactors, and those that occur in production facilities, more commonly referred to as process criticality accidents. This dissertation focuses on the development of a methodology that uses a coupled Bayesian network and neural network metamodel to estimate process criticality accident risk. This methodology is essentially a generalized software-based framework that was written in the R programming language and applied to fissile material operations in the Plutonium Facility (Building 332) at Lawrence Livermore National Laboratory. The most significant benefit of using Building 332 as a case study was that it enabled a comprehensive review of the entire process of building a coupled Bayesian network and neural network metamodel and estimating process criticality accident risk. This review also enabled iterative improvements to be made to the methodology as it was being developed. Overall, the coupled Bayesian network and neural network metamodel worked extremely well and has several advantages over existing methodologies.

Author Language Date created Type of Work License
  • All rights reserved
Rights statement GW Unit Degree Advisor Committee Member(s) Persistent URL

Notice to Authors

If you are the author of this work and you have any questions about the information on this page, please use the Contact form to get in touch with us.

Thumbnail Title Date Uploaded Visibility Actions
Preview of Zywiec_gwu_0075A_15503.pdf Zywiec_gwu_0075A_15503.pdf 2021-05-10 Open Access