(111a) A Risk-Based Approach for Predicting Domino Effects Due to Fires - Combining Exceedance Curves with Thermal Stress Dynamic Analysis | AIChE

(111a) A Risk-Based Approach for Predicting Domino Effects Due to Fires - Combining Exceedance Curves with Thermal Stress Dynamic Analysis

Authors 

Dunjo, J. - Presenter, Bright Calcs LLC
This paper proposes a risk-based method for characterization of domino effects and potential escalation for process equipment affected by thermal radiation (i.e. fires). This approach intends to answer two key questions: (1) which process equipment is impacted by a heat flux capable of resulting in escalation due to equipment failure?; and (2) how long until a process equipment is expected to fail; i.e., Time To Failure (TTF)?.

The first phase consists of developing dedicated heat flux exceedance curves for a given location of interest, using a risk-based approach and considering a cumulative frequency threshold based on worldwide recognized risk tolerability criteria. The exceedance curve approach is valuable to graphically identify which heat flux impacts the process equipment at the given frequency of interest.

The second phase involves the prediction of the TTF due to fires (e.g., pool fires, jet fires) impacting the equipment identified in step one. A two-step approach is proposed: (1) vessel wall segmentation to accurately predict heat transfer from fire to defined wall segments, from wall segment to wall segment, and from wall segments to vessel contents. This dynamic simulation is conducted to determine how the Ultimate Tensile Strength (UTS) of the material decreases as a function of temperature. The UTS is then compared with the internal stress (i.e., Hoop stress) by considering the equipment internal pressure combined with the installed overpressure protection performance. This paper defines step-by-step how to conduct a risk-based quantitative assessment and determine the TTF using a case study. It demonstrates the applicability and accuracy of this approach, which can result in a less expensive procedure than, for example, Computation Fluid Dynamics. Furthermore, it helps the decision-making process on how potential mitigation measures can be implemented.