(137a) LNG Spills on Water: A Comparison of the Shallow Water Model to Experiments

The sudden increase in the supply of shale gas and the demand for gas in both Europe and Asia have made it attractive to liquefy natural gas for exportation in large LNG tankers.  With this new source of gas, its price is now lower than diesel or gasoline on an energy basis. This has made LNG viable as a transportation fuel in North America.  Multiple LNG fueled transportation projects have been financed and are now being developed in trucking, shipping and rail transportation.  This new wide scale application of LNG as a transportation fuel has raised a broad range of safety and environmental questions.  This paper focuses on the manner in which a spill of LNG behaves on water.  LNG is lighter than water, and floats on it.  Its low temperature causes LNG to evaporate in the film boiling regime.  Even though a thin vapor layer separates the LNG from the water, the interface is not flat and smooth.  It is wavy and causes additional drag.  The engineering quantities that control how a pool spreads on water are the bluff-body drag that the leading edge of the pool experiences and the friction between the spreading pool and the water below.

For purposes of consequence analyses, it is often necessary to estimate the vapor cloud size resulting from a spill of LNG either on water or on the ground.  Since the total rate of evaporation of LNG depends on the size of the pool, the accuracy of the analysis depends on the accuracy of the pool spread model. 

Very few experiments have been carried out that were sufficiently instrumented to help quantify the phenomena involved in a spill of LNG on water.  One such series of tests was performed at MIT by Chang.  Those tests were done in a linear channel and monitored the position of the leading edge following the sudden release of an initially stationary pool of LNG.  This paper simulates these tests using a shallow water model, and quantifies the phenomena that control the growth of the LNG pool.