期刊
JOURNAL OF LOSS PREVENTION IN THE PROCESS INDUSTRIES
卷 26, 期 1, 页码 117-128出版社
ELSEVIER SCI LTD
DOI: 10.1016/j.jlp.2012.10.002
关键词
LNG; CFD; Dense gas dispersion; Risk assessment; NFPA 59A; Impoundment
资金
- Key Laboratory of LNG Cryogenic Technology of Guangdong High Education Institute [39000-3211101]
- SYSU-BP LNG Centre [99103-9390001]
- Key Laboratory of Fire Science and Technology of Guangdong Province [2010A060801010]
- Australia-China council scholarship
Computational fluid dynamics (CFD) simulations have been conducted for dense gas dispersion of liquefied natural gas (LNG). The simulations have taken into account the effects of gravity, time-dependent downwind and crosswind dispersion, and terrain. Experimental data from the Burro series field tests, and results from integral model (DEGADIS) have been used to assess the validity of simulation results, which were found to compare better with experimental data than the commonly used integral model DEGADIS. The average relative error in maximum downwind gas concentration between CFD predictions and experimental data was 19.62%. The validated CFD model was then used to perform risk assessment for most-likely-spill scenario at LNG stations as described in the standard of NFPA 59A (2009) Standard for the Production, Storage and Handling of Liquefied Natural Gas. Simulations were conducted to calculate the gas dispersion behaviour in the presence of obstacles (dikes walls). Interestingly for spill at a higher elevation, e.g., tank top, the effect of impounding dikes on the affected area was minimal. However, the impoundment zone did affect the wind velocity field in general, and generated a swirl inside it, which then played an important function in confining the dispersion cloud inside the dike. For most cases, almost 75% of the dispersed vapour was retained inside the impoundment zone. The finding and analysis presented here will provide an important tool for designing LNG plant layout and site selection. (C) 2012 Elsevier Ltd. All rights reserved.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据