Automated Methods in Sizing of Relief Vent Stacks Based on Proximity to Occupied Structures

Quantifying the impact of a relief event to personnel is critical to site safety and is mandated by PSM. Within sites and chemical parks containing several occupied structures, analyzing the impact to all affected personnel from a stack release is highly labor intensive. Performing comparisons between several potential locations can be both cost and time inhibitive.

This paper details a methodology for utilizing automated scripting to run several dispersion models to determine the potential number of significant injuries to building occupants from a stack release. The count of significant injuries is used as a cost that is minimized through optimization of stack height and stack diameter. The model is iterated in all potential wind directions, with the worst case wind direction used in setting stack size.

The paper will also discuss methods of both assessing safe relief rates through existing vent stacks, and how to determine stack sizing as well as safe relief rate if no information is available.

Significant injury in this model is calculated based on dose-response methodology. A dose-response methodology returns a probabilistic score for potential impact to personnel inside of an occupied structure.

This paper will discuss the required inputs to the model, detail on the flow of the automated methods, how to interpret the results including some caveats noted around the model behavior during its development.