A mathematically guided strategy for risk assessment and management.

Conference · Tue Mar 01 04:00:00 EST 2005

OSTI ID:949017

Cooper, James Arlin

Strategies for risk assessment and management of high consequence operations are often based on factors such as physical analysis, analysis of software and other logical processing, and analysis of statistically determined human actions. Conventional analysis methods work well for processing objective information. However, in practical situations, much or most of the data available are subjective. Also, there are potential resultant pitfalls where conventional analysis might be unrealistic, such as improperly using event tree and fault tree failure descriptions where failures or events are soft (partial) rather than crisp (binary), neglecting or misinterpreting dependence (positive, negative, correlation), and aggregating nonlinear contributions linearly. There are also personnel issues that transcend basic human factors statistics. For example, sustained productivity and safety in critical operations can depend on the morale of involved personnel. In addition, motivation is significantly influenced by 'latent effects', which are pre-occurring influences. This paper addresses these challenges and proposes techniques for subjective risk analysis, latent effects risk analysis and a hybrid analysis that also includes objective risk analysis. The goal is an improved strategy for risk management.

• Research Organization: Sandia National Laboratories
• Sponsoring Organization: USDOE
• DOE Contract Number: AC04-94AL85000
• OSTI ID: 949017
• Report Number(s): SAND2005-1895C
• Country of Publication: United States
• Language: English

Related Subjects

99 GENERAL AND MISCELLANEOUS
COMPUTER-AIDED DESIGN
HUMAN FACTORS
OCCUPATIONAL SAFETY
PERSONNEL MANAGEMENT
PRODUCTIVITY
RISK ASSESSMENT
STATISTICS