ADAPT – Dynamic Event Tree Generation and Analysis
ADAPT is a flexible dynamic event tree generation and analysis platform
Overview
Traditional Probabilistic Risk Assessment (PRA) requires analysts to assume order of events, while Dynamic PRA (DPRA) is driven by time-resolving models of the relevant phenomena. DPRA can give additional insights into complex event progressions including impacts from physical parameters and the timing of human interactions.
ADAPT can accommodate diverse physical systems
- Limited only by availability of appropriate simulators
- Includes MELCOR, RELAP5, RADTRAD, RADTRAN, and SAS4A/SASSYS-1
ADAPT is open source and easily adaptable to various computation environments
- No proprietary software dependencies
- Either SSH/SCP connections on a local cluster or High Performance Computer (HPC) submissions using the Slurm job scheduler
Extensible data analysis tools
- Dynamic importance measure platform
- Reduction of Dynamic Event Trees (DETs) according to user-input slicing rules
- Scalable from hundreds to 1M+ branches
ADAPT can be applied in any case where:
- There is a system analysis problem that may be represented by an event tree,
- There is a dynamic computer code that can capture the anticipated progression of events, and
- That computer code may be made to stop after designated events occur and restart the analysis with changed input.
What is PRA?
PRA is an analytic tool that seeks answers to three questions:
- What failure events might affect the system?
- What are the possible consequences of an initiating failure event?
- What is the likelihood of each set of consequences?
PRA uses Event-Tree/Fault-Tree (ET/FT) analysis to evaluate the impacts of internal or external initiating events on a complex system. In a Fault Tree (FT), basic failure events are assembled using primarily AND/OR logic to determine the combinations of individual component failures that lead to failure of the system as a whole (a “top event”). Fault trees are generated for top events that represent a system failing to meet one or more of its design goals. These fault trees are stored in a database and can be used for a variety of analyses.