(44c) The Use of CFD to Evaluate the Interactions Between Multiple-Leak Sources and to Assess Effectiveness of Integral Modeling Techniques in a Multiple-Leak Scenario

In many
practical applications within the chemical industry, it is of vital importance
to know the concentration profile that may occur from the dispersion of a
material. For example, effect footprints for the release of a toxic material
are needed to properly assess risk to both a facility and its surrounding
communities and are useful for emergency response planning. Likewise, knowing
the concentration gradient of a flammable gas can affect, among many other
things, the layout of a facility to prevent or lessen the consequence of
explosion and fire scenarios or for the placement of detection systems.

There are two main tools for the modeling of dispersion scenarios.
Computational fluid dynamics (CFD) is a useful tool for accurately assessing
the dispersion of materials, whether toxic, flammable, pollutant, or otherwise.
It has found extensive use in academic as well as industrial settings when
scenarios and geometries are well-defined because it takes into account the
surroundings in the release environment. However, it is relatively slow and
computationally expensive to use.

Also used extensively in the chemical industry are integral
consequence modeling tools (such as Phast) which are generally more
user-friendly and have the advantage of delivering results in a more
time-efficient manner, but cannot take into account complex geometry
considerations. A particularly interesting problem to consider in practice is
that of multiple simultaneous release sources. This could be caused by a common
failure of two units causing concurrent release, a domino effect where an event
occurs that causes another event leading to multiple releases, or even an
intentional event such as a terrorist attack. While CFD has the capability to
resolve the possible interactions between the multiple release sources,
integral modeling does not.

A simple current approach to the problem of multiple-source
scenarios is to model them independently and superimpose the results upon one
another. This may be a good approximation in some cases, but it does not take
into account the complex interactions that may take place between the sources. The
practical question then becomes whether it is acceptable in certain cases to
model a certain set of simultaneous releases as an independent set of sources
or whether it could be possible to model them as a single combined source.

This study aims to utilize CFD modeling of a set of varying
scenarios to classify multiple-release dispersion scenarios into one of three
different categories:

·              
Releases that are
independent and may be modeled as such

·              
Releases that may be
combined and modeled as one source

·              
Releases that must be
considered together due to inherent interactions between them

Based upon criteria such as material, distance between sources,
dispersion rate and duration, weather conditions, and geometry considerations,
a general methodology for classifying multiple-release scenarios into these
three different categories will be proposed so that these releases can
ultimately be modeled in the appropriate manner.