Conditional Dependence of Individual Events and Outcome Probabilities in Chemical Processes
Loading...
Authors
Mott, Timothy Cruz
Issue Date
2020
Type
Thesis
Language
Keywords
Alternative Title
Abstract
Accounting for the dependence between events in chemical process safety is important because assuming that events are statistically independent does not accurately represent the probability of an outcome such as disaster, process shutdown, or continuing operations. The approach presented in this work is based on event tree analysis and quantitative risk analysis to calculate conditional probabilities of outcomes considering the dependence between events. Using the approach developed, probabilities of disaster were calculated using conditional probabilities and assuming statistical independence. Three scenarios were analyzed: a simple alarm �" operator interaction, and two examples that involve a chemical reactor, operator, and multiple layers of protection. Results demonstrate that conditional probabilities need to be considered for accurate calculations of disaster probabilities, regardless of the number of independent layers of protection. The approach shown can be incorporated in chemical process safety education.
Description
Citation
Publisher
License
Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 United States