Comparative study on toxic gas infiltration in a non-process area using CFD and multi-zone models
Conference Paper
Overview
Additional Document Info
View All
Overview
abstract
2016 IChemE. Although the majority of incidents involving toxic gas release in process industries occurs outdoors, nearby buildings and indoor environments are also at high risk. Particularly, non-process areas such as administration buildings and central utility plants are often the least protected, even though they are in the vicinity of potential sources. Ingress of contaminants depends on various factors such as meteorology, building ventilation, building envelope, among others. In literature, indoor exposure modelling techniques range from simple statistical regression and mass balance approaches to more complex models such as multi-zone and computational fluid dynamics (CFD) (Milner et al. 2011). Therefore, to study toxic gas infiltration, a proper selection of models is required. Despite the significant risk posed by such events in process facilities, there is still a lack of data and comparative studies concerning the appropriate models and mitigation methods. This paper investigates a realistic pipeline leak in a natural gas facility and the subsequent H2S exposure of the nearby administration building. A comparative study is performed by utilizing a dispersion model (SLAB), a multi-zone model (CONTAM) and a CFD model (Quick Urban and Industrial Complex - QUIC). The influence of ventilation network, nature of openings, wind speed, direction and pressure on toxic gas ingress is examined. Furthermore, we also study the sensitivity of wind pressure calculation on the toxic gas infiltration rate by using American Society of Heating, Refrigerating, and Air-Conditioning Engineers (ASHRAE) handbook and CFD modelling. Indoor toxic levels are attained using multiple combinations of the above mentioned models. Results on indoor toxic levels indicated high sensitivity to wind characteristics which led to varying risks and conclusions. A detailed description of studied scenarios and findings is also presented.
