Detailed Analysis of Hazards Identified during a PHA

A Process Hazards Analysis (PHA) is a tool used to identify failures that can defeat or bypass safeguards, leading to process safety hazards. Most methods of PHA are qualitative, and involve a team of subject matter experts with diverse technical backgrounds to identify how these safety issues can arise. Qualitative methods, such as Hazard and Operability (HazOp) method, require the team to systematically evaluate thousands or even tens of thousands of potential scenarios. From some of these scenarios come recommendations to either further analyze a particular problem, or to mitigate risk to a perceived hazard.

 

This presentation will focus on more detailed methods of analysis which allow the user to quantify some of the hazards identified during the PHA, and thus to decide whether they may be deemed acceptable based on a particular risk criteria, or otherwise require mitigation. Case studies for each method will be presented and discussed.

 

The first case study analyzes the impact of blast hazards over a critical process vessel. Once a potential blast scenario has been identified and modeled using a Vapor Cloud dispersion method, dynamic structural loads expressed as overpressures and impulse are applied to either a Finite Element Analysis (FEA) or a simplified Single Degree of Freedom (SDOF) model of the vessel, and the structural response of the equipment is compared to a set of acceptable parameters based on accident investigation, modeling and full scale testing to determine the condition of the vessel after the postulated blast.

 

The second case study deals with the response of a critical process reactor to a major seismic event. Most typical building code based analysis are based on simplified response spectrum methods which do not accurately reproduce the behavior of the structure, while insurance based estimates are often based on highly simplified statistics in lieu of actual analysis: in this case, actual accelerometer data from a documented seismic event was loaded into a Finite Element Analysis to obtain a realistic response and damage estimate.

 

Finally, a third case study discusses optimized placement of detectors using fire & gas modeling tools. Upon identification of potential flammable and toxic release scenarios, dispersion clouds of those were imported into a fire & gas modeling software. The software models the location of existing flammable and toxic detectors placed within the unit and determine which clouds encounter specific detectors. This data is integrated to determine the probability that a release from a particular source, in any wind direction, will be seen by a detector. From these results, the placement of detectors is optimized to ensure maximum coverage for a release.