ENVIRONMENT for all four spills even though the wind speed and atmospheric conditions varied considerably. Figure 5 shows measurements made with the Anarad infrared sensor, the only instrument in 150 100 propane. Theseare thefirst field measurements ever made of differential boiloff of the various constituents of LNG. Differential boiloff tO | & ° a 2 2 50 e 10 6 Ob 8 qr | 0 | ° 5,11 is important becauseit affects the detonability of the vapor cloud. Detonating pure methaneis difficult; detonating methane mixed with ethane or propane is much easier. (b) 50s N *? B 0 among methane, ethane, and | (a) 10s E the array that could distinguish | 50 100 | 150 Source 200 0 Distance — m 150 | (c) 90s 1 50 | 100 | 150 200 Distance — m - ] 9 Pig. a tld . Dot plots of a particle-in-cel!l calculation of the plumefrom the second 5-m? experi- ment (a) at 108, (b)at 50s, and (c) at 90s. The large, numbered dots are the instrument stations shown in Fig. 1. A shift in wind direction (possibly influenced by a e 100 [— small hill) created a persistent kink in the plume. | is] o Cc a S sok 0 0 50 100 Distance — m | 150 200 | | : , con 6-8 o07 vwliU UO } 33 : _