skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: LAVA (Los Alamos Vulnerability and Risk Assessment Methodology): An expert system framework for risk analysis

Conference ·
OSTI ID:5091023

The Los Alamos National Laboratory has developed an original methodology for performing risk analyses on subject systems characterized by a general set of asset categories, a general spectrum of threats, a definable system-specific set of protecting the assets from the threats, and a general set of outcoming resulting from threats exploiting weaknesses in the safeguards system. Applications of this methodology can be tailored to address a wide variety of subject systems' risk analysis requirements. The Los Alamos Vulnerability and Risk Assessment Methodology (LAVA) models complex systems having large amounts of ''soft'' information about both the subject system itself and occurrences related to the system. Its structure lends itself well to automation on a portable computer, making it possible to analyze numerous similar but geographically separated installations consistently and in as much depth as the subject system warrants. LAVA is based on hierarchical systems theory, event trees, fuzzy sets, natural-language processing, decision theory, and utility theory. LAVA's framework is a hierarchical set of fuzzy event trees relating the results of several embedded (or sub-) analyses: a vulnerability assessment providing information about the presence and efficacy of system safeguards, a threat analysis providing information about static (background) and dynamic (changing) threat components coupled with an analysis of asset ''attractiveness'' to the dynamic threat, and a consequence analysis providing information about the outcome spectrum's severity measures and impact values. Each sub-analysis can be simplified or made complex, depending on the sensitivity and relative worth of the subject system. 34 refs., 8 figs.

Research Organization:
Los Alamos National Lab., NM (USA)
OSTI ID:
5091023
Report Number(s):
LA-UR-88-1385; CONF-880593-1; ON: DE88009112
Resource Relation:
Conference: Federal analysis workshop, Denver, CO, USA, 24 May 1988
Country of Publication:
United States
Language:
English