CALPUFF Modeling of Accidental Releases

The Lagrangian non-steady state model CALPUFF has the capacity to model short term accidental releases. Sub-hourly time steps allow for an accurate portrayal of highly transient source terms and the availability of ASOS 1-minute data can be used to provide high resolution for meteorological data fields. For emergency planning purposes it is often desirable to predict the impact of a potential future accidental release which could happen at any time. Non-steady state models such as CALPUFF carry over emissions from the prior hour allowing for stagnation and re-circulation effects to be modeled. This requires special data handling techniques in order to analyze each potential release in isolation. Post processors have been developed to allow for this type of analysis. These post processors have been applied to flaring events and are equally applicable to a variety of accidental releases.

In this study we examine two accidental release scenarios using the non-steady state Lagrangian model CALPUFF. We are using sub-hourly time steps to model a short term accidental release. The accidental release is modeled as a volume source in order to represent a range of low to high level releases from a complex source. The multiple sources are staggered in time to cover the possibility of the accidental release starting at any point of the day. The variable emission files are prepared so that each start time of the accidental release is treated as a separate source. The impacts from each source are independently processed using the post-processors in the CALPUFF suite of processing tools.