Comparative study on gas dispersion Page i 0. SUMMARY The Directorate for Civil Protection and Emergency Planning (DSB) and Scandpower have jointly performed a comparative study on gas dispersion. Both integral tools (Phast and Trace) and CFD tools (FLACS and KFX) has been used in the comparison. The following have been concluded from the comparative study: - Simulated gas dispersion distances must be regarded as somewhat uncertain, independent of what tool that is used. However, the spread in results from the different tools are considered to be moderate, and none of the tools are disregarded as applicable for gas dispersion calculations. - It is also seen that different project teams may end up with slightly different results, even if they are using the same tool. This also contributes to a spread in the results. The uncertainty linked to the user is larger for CFD tools than for integral tools since there are more parameters that need to be user specified in the CFD tools (e.g. the grid resolution, the release conditions and the boundary conditions). - It is seen that low momentum leaks (e.g. pool leaks) are likely to lead to larger hazardous distance than high momentum leaks (jet leaks). This is contributed the fact that high momentum leaks will give larger turbulence and hence better mixing of air and gas. - It is seen that when obstructions and terrain is not regarded a driver for the results, then both integral tools and CFD tools can be applied with similar accuracy. - When obstructions and terrain are considered important for the hazardous distances, only CFD tools can be applied. This will typically be when there are large buildings near the release, when the leaks are in highly congested areas (much process equipment), and when there is potential for release of heavy gas in a sloping and terrain. - Details of the comparative study are summarized in Chapter 2.6. Typical safety distances for 13 different land facility plant types (named DSB cases in this report) has also been calculated using Phast. Hazardous distances for the 10-4 to 10-7 events have been calculated and presented. The following has been concluded from the study of the DSB cases: - It is generally the activity levels that are dominating the safety distances, and not the amount of storage. Typical activities that are dominating the safety distances are loading and offloading. Typical parameters that are contributing to the severity of a leak are large cross sections and high pressures and large segment volumes. - Liquid phase releases (LNG/LPG) will normally give significantly longer safety distances than gaseous leaks (typically a factor 2-3 times longer). - There is often a significant difference in the safety zone referring to ½ LFL and LFL. The safety zone may be more than 2 times larger (longer) if ½ LFL is used as maximum safe gas concentration. 101368_r1_final 24 January 2012 Scandpower is a member of the Lloyd's Register Group
Last ned PDF-fil
Arkiv