Development of EMRALD Services in a Fully-Integrated RISMC Platform
This report describes how FPoli fulfilled its vision for the “Development of EMRALD Services in a Fully-Integrated RISMC Platform”, the objective of SBIR Grant DE-SC0018755. The motivation behind the services were business goals and quality attributes envisioned for the platform developed with this project and were documented in the final commercialization plan, FPS-2019-5, Revision 1 (April 2022). Key functionalities were added to the EMRALD software to enable integration into FPoli-AAP service platform and significant knowledge transfer occurred from INL/BEA to FPoli for FPoli to continue development of EMRALD as needed beyond the completion of the SBIR project. The EMRALD frontend UI is now an integral part of the FPoli-AAP platform. EMRALD models are saved to the FPoli-AAP SQL database. EMRALD systems and components are linked to the SSCs datastore in FPoli-AAP. Models can then be simulated as a backend service from the stack. The results are then be saved to the FPoli-AAP SQL database, viewable to the analyst on the frontend. Integrated post-processing of the results creates probability distributions that automatically feed dynamic event trees managed by RAVEN, etc. From a product definition the EMRALD service is an integral part of the Risk-Informed System Engineering (RISE) application, FPoli’s go-to-market strategic tool. RISE was designed to orchestrate the complexity of the NEI-18-04 (10 CFR Part 53) process. A high-level flowchart of the RISE workflow is presented below. The initial go-to-market target for the RISE application is the advanced reactor developers and regulators. However, considering that NEI-18-04 is really an evolution of 10 CFR 50.69, the applicability to SSCs safety re-classification is a natural extension of the use case. Note that RISE was also architected to enable a more graded approach to maximal PRA usage, a calibrated solution that was found to be more attractive to the industry. Within the context of the envisioned DPRA analyses, EMRALD will be used to translate the existing scope of static PRA inputs into dynamic component interaction models and to perform simulations with these models. The output of these simulations is essentially the failure propagation information of the DPRA analysis, effectively complementing or taking the place of traditional fault trees. In practice, this output is the distributions of system failures and timing used to inform the event progression, specifically dynamic event trees. The event progression is simulated separately with RAVEN and RELAP5-3D or other tools as needed. The output of the event progression simulations using RAVEN and RELAP5-3D can be the core damage frequency (CDF) estimation, similar to that obtained from an event tree in a traditional PRA analysis. When positioned within the RISE framework for the NEI-18-04 process, the output is expected to be a consequence (i.e., dose) distribution. RISE enables a graded approach to Part 53. Ultimately a reactor designer of a design with high margin to safety may simply choose to continue using the bounding/representative approach and use a detailed PRA analysis as a confirmatory step of their safety case. The vision for the platform is a process that allows a user to incorporate updated state of knowledge on an ongoing basis. In the future this could evolve in creating a Bayesian framework for this type of analysis. You could consider initial event sequences, combine them in surrogate metrics (the ESFs) and update your prior as you gather data. This can in principle be framed in a mathematically coherent framework via automation. FPoli will continue development of the EMRALD enterprise system after the completion of the SBIR Phase-II project. Focus will be given to ease of use which will be a combination of improving/guiding workflows, online tutorials, videos, improved views, etc. It is anticipated that this DPRA analysis will likely replace, modify, or augment event sequences in a plant’s PRA analysis; ideally event trees which contribute significantly to the overall risk. The vision is for RISE/EMRALD to become the infrastructure for ‘exploring’ the space of possibilities to identify and screen hazards. This is particularly important in the early stages of a new design which is iterative in nature. This can complement other hazard identification and assessment tools used in the industry and ultimately facilitate the adoption of risk-informed approach in defining the safety case for the next generation of nuclear power plants.
- Research Organization:
- FPoliSolutions 450 William Pitt Way, Pittsburgh, PA 15238
- Sponsoring Organization:
- USDOE Office of Nuclear Energy (NE)
- DOE Contract Number:
- SC0018755
- OSTI ID:
- 1870853
- Type / Phase:
- SBIR (Phase II)
- Report Number(s):
- DOE-FPOLI-DE SC0018755
- Country of Publication:
- United States
- Language:
- English
Similar Records
RISK-INFORMED EXTERNAL HAZARDS ANALYSIS BY EEVE TOOLKIT
RAVEN. Dynamic Event Tree Approach Level III Milestone