PSAM 16 - Paper Overview

PSAM 16 Conference Paper Overview

Welcome to the PSAM 16 Conference paper and speaker overview page.

Lead Author: Mostafa Hamza Co-author(s): Alp Tezbasaran - email: alptezbasaran@ncsu.edu Mihai Diaconeasa - email: madiacon@ncsu.edu

Methodology and Demonstration of Git-based Configuration Control in Probabilistic Risk Assessment
Probabilistic Risk Assessment (PRA) has been an integral part of large-scale high-risk industries, like the aerospace, chemical, and nuclear industries. Both fault tree (FT) and event tree (ET) modeling techniques are essential parts of any PRA model. FTs model the possible causes of system failures, whereas ETs are used to track the progression of a certain postulated initiating event and the associated system responses. FTs and ETs can be trivially built, modified, updated, and tracked for simple systems and associated event progressions. However, as the complexity of the system or event progression increases, their associated models become more complex. For models developed, modified, and updated by a singular analyst, increasing complexity does not necessarily present a challenge for model development. However, if multiple analysts contribute to developing the same FT or ET, keeping track of the changes becomes critical to avoid conflicting or replicating work. In addition, rigorous configuration control is imperative to keeping track of PRA model modifications during iterative design stages where system configurations evolve dynamically. Under current configuration control requirements, PRA standards in the nuclear industry, including ASME/ANS Probabilistic Risk Assessment Standard for Advanced Non-Light Water Reactor Nuclear Power Plants, require an explicit process for monitoring design changes, tracking associated model modifications, and tracing model progression. Though some existing PRA tools, such as CAFTA, implement minimal version-control, multi-user collaboration and model-tracking capabilities are limited. To address this unmet need, we present a methodology that utilizes Git to fulfill the configuration control requirements and achieve rudimentary multi-user collaboration. We then demonstrate the presented method on representative models using the MAR-D file format for SAPHIRE and CAFTA. Finally, we offer a brief discussion of the limitations within the proposed methodology and propose improvements for future work.

Methodology and Demonstration of Git-based Configuration Control in Probabilistic Risk Assessment

Probabilistic Risk Assessment (PRA) has been an integral part of large-scale high-risk industries, like the aerospace, chemical, and nuclear industries. Both fault tree (FT) and event tree (ET) modeling techniques are essential parts of any PRA model. FTs model the possible causes of system failures, whereas ETs are used to track the progression of a certain postulated initiating event and the associated system responses. FTs and ETs can be trivially built, modified, updated, and tracked for simple systems and associated event progressions. However, as the complexity of the system or event progression increases, their associated models become more complex. For models developed, modified, and updated by a singular analyst, increasing complexity does not necessarily present a challenge for model development. However, if multiple analysts contribute to developing the same FT or ET, keeping track of the changes becomes critical to avoid conflicting or replicating work. In addition, rigorous configuration control is imperative to keeping track of PRA model modifications during iterative design stages where system configurations evolve dynamically. Under current configuration control requirements, PRA standards in the nuclear industry, including ASME/ANS Probabilistic Risk Assessment Standard for Advanced Non-Light Water Reactor Nuclear Power Plants, require an explicit process for monitoring design changes, tracking associated model modifications, and tracing model progression. Though some existing PRA tools, such as CAFTA, implement minimal version-control, multi-user collaboration and model-tracking capabilities are limited. To address this unmet need, we present a methodology that utilizes Git to fulfill the configuration control requirements and achieve rudimentary multi-user collaboration. We then demonstrate the presented method on representative models using the MAR-D file format for SAPHIRE and CAFTA. Finally, we offer a brief discussion of the limitations within the proposed methodology and propose improvements for future work.

Paper MM182 Preview

Author and Presentation Info

Lead Author Name: Mostafa Hamza (mmhamza@ncsu.edu)

Bio: I am part of the probabilistic risk assessment (PRA) group at NC State University. My current work is supporting the design of X-energy’s high-temperature gas-cooled reactor (HTGR), the Xe-100. As part of the ARPA-e GEMINA program, I am working with their PRA team developing the PRA of Xe-100 to inform the design. Following the Licensing Modernization Project (LMP) methodology, we are developing the technology-inclusive risk-informed performance-based design of the advanced reactor. In addition, I am part of the NC State University initiative developing a dynamic PRA platform. I am responsible for coupling human reliability analysis (HRA) methods with the platform moving towards, building on the capabilities of codes like ADS-IDAC, the development of a complete open-source object-oriented PRA code.

Country: United States of America
Company: North Carolina State University
Job Title: Graduate Research/Teaching Assistant

PSAM 16 Conference Paper Overview

Welcome to the PSAM 16 Conference paper and speaker overview page.

Paper MM182 Preview

Author and Presentation Info

Download paper MM182.

Download the presentation pdf file.

Download the presentation PowerPoint file.