Evaluation of the CFD model FLACS for LNG safety

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner
Tuesday, 19 January 2010: 2:45 PM
B308 (GWCC)
Olav R. Hansen, Gexcon US, Bethesda, MD; and M. Ichard, S. Davis, and J. Melheim

An increasing amount of liquefied natural gas (LNG) is transported by ship and handled at terminals which are often located near densely populated regions. Vapor dispersion models are used for the analysis of safety features for LNG facilities. A Model Evaluation Protocol (MEP) and Model Assessment Protocol (MAP) have been developed on behalf of NFPA by the Health & Safety Laboratory, the research agency of the United Kingdom Health & Safety Executive, to help regulatory agencies to make decisions on the appropriateness of dispersion models for predicting hazard ranges for large LNG spills. The MEP provides guidelines for assessment of atmospheric dispersion models and the MAP provides guidelines for source term model assessment. The MEP and MAP have been applied to the CFD model FLACS. For the MEP, a model validation database has been established and repertories 33 test configurations. The model validation database is composed of unobstructed and obstructed field trials and wind tunnel tests. In the MEP Statistical Performance Measures (SPM) are used to quantify the performance of a model. FLACS is shown to meet the SPM quantitative assessment criteria required to pass the MEP. The main difference between the MEP and the MAP is that the latter does not include quantitative assessment criteria. The MAP identifies datasets for spills on land and water that should be used in a preliminary source term model assessment. These tests are simulated with FLACS and results for pool spreading and vaporization are compared with experimental data and other model predictions when available.