The use of natural gas has been on the increase for the past three decades and current consumption is expected to double by the year 2020. A rise in the number of transoceanic shipments will, logically, increase the risk of accidental spills, especially at terminal facilities. An LNG spill will generate a potentially dangerous cryogen vapour cloud. Consequently, the study of the rate of vapourization of this cryogen over water surfaces is of primary importance. Experimental measurements on confined LNG spills indicate the presence of high vapourization rates at early times after spillage. The present work introduces a vapourization model that has not been explored in previous studies. This model proposes the existence of a thin liquid layer at the cryogen-water interface that is not in thermodynamic equilibrium with the bulk of cryogen liquid. The implementation of the proposed model in a computer program allowed for the simulation of the boil-off process, and the estimation of the initial vapourization rates and heat transfer coefficients in confined and unconfined spills. The simulated cryogen spills offered a description of the vapourization rates that correlated well with experimental boil-off data. The simulations indicated the existence of early vapourization rates in the range of 0.03 to 0.06 gcm_2s_1P>, and an average initial heat transfer coefficient of 1576 Wm_2K_1 for a typical LNG mixture.
Previous efforts to describe the rate of vapourization of LNG include a study to determine the effect of its chemical composition on the vapourization rate (1). The results of this study indicated that the vapourization rate of LNG differs from that of pure methane, especially in the late stages of a spill. The differences were attributed to the ethane-enrichment of the bulk liquid phase during vapourization, which dramatically decreased the rates during the last stages of film boiling, followed by increasing rates after the collapse of the vapour film layer. The vapourization rate, therefore, exhibited a minimum as a function of spill time.
The alluded study, however, suggested a mechanism of vapourization that would not allow for the high vapourization rates measured experimentally. A major assumption of this previous work was that the bulk of liquid was in thermodynamic equilibrium with the vapour film throughout the vapourization process ("bulk model"). Some authors (2)(3)(4) have questioned the equilibrium hypothesis. The authors argue that the preferential vapourization of the more volatile component will cause the liquid layer of the cryogen in direct contact with the film of vapour to be richer in the heavier component than the bulk of liquid. Valencia-Chavez and Reid (3) have suggested the existence of a repetitive mechanism by which the thin layer of LNG is constantly being depleted of its light component (methane). The methane is preferentially vapourised and carried in the bubbles through the liquid layer.
