Burlington, MA, United States
Burlington, MA, United States

Time filter

Source Type

Qi R.,Texas A&M University | Raj P.K.,Technology and Management Systems Inc. | Mannan M.S.,Texas A&M University
Journal of Loss Prevention in the Process Industries | Year: 2011

An underwater LNG release test was conducted to understand the phenomena that occur when LNG is released underwater and to determine the characteristic of the vapor emanating from the water surface. Another objective of the test was to determine if an LNG liquid pool formed on the water surface, spread and evaporated in a manner similar to that from an on-the-surface release of LNG. A pit of dimensions 10.06 m × 6.4 m and 1.22 m depth filled with water to 1.14 m depth was used. A vertically upward shooting LNG jet was released from a pipe of 2.54 cm diameter at a depth of 0.71 m below the water surface. LNG was released over 5.5-min duration, with a flow rate of 0.675 ± 0.223 L/s. The wind speed varied between 2 m/s and 4 m/s during the test. Data were collected as a function of time at a number of locations. These data included LNG flow rate, meteorological conditions, temperatures at a number of locations within the water column, and vapor temperatures and concentrations in air at different downwind locations and heights. Concentration measurements were made with instruments on poles located at 3.05 m, 6.1 m and 9.14 m from the downwind edge of the pit and at heights 0.46 m, 1.22 m, and 2.13 m. The phenomena occurring underwater were recorded with an underwater video camera. Water surface and in-air phenomena including the dispersion of the vapor emanating from the water surface were captured on three land-based video cameras. The lowest temperature recorded for the vapor emanating from the water surface was -1 °C indicating that the vapor emitted into air was buoyant. In general the maximum concentration observed at each instrument pole was progressively at higher and higher elevations as one traveled downwind, indicating that the vapor cloud was rising. These findings from the instrument recorded data were supported by the visual record showing the "white" cloud rising, more or less vertically, in air. No LNG pool was observed on the surface of water. Discussions are provided on the test findings and comparison with predictions from a previously published theoretical model. © 2011 Elsevier Ltd.


Raj P.K.,Technology and Management Systems Inc. | Bowdoin L.A.,Weavers Cove Energy LLC.
Journal of Loss Prevention in the Process Industries | Year: 2010

One of the scenarios of concern in assessing the safety issues related to transportation of LNG in a marine environment (ship or underwater pipeline) is the release of LNG underwater. This scenario has not been given the same level of scientific attention in the literature compared to surface releases and assessment of consequences therefrom. This paper addresses questions like, (1) does an LNG spill underwater form a pool on the water surface and subsequently evaporate like an LNG spill "on the surface" producing cold, heavier than air vapors?, and (2) what is the range of expected temperatures of the vapor, generated by LNG release due to heat transfer within the water column, when it emanates from the water surface?Very limited data from two field tests of LNG underwater release are reviewed. Also presented are the results from tests conducted in other related industries (metal casting, nuclear fission and fusion, chemical processing, and alternative fuel vehicles) where a hot (or cold) liquid is injected into a bulk cold (or hot) liquid at different depths.A mathematical model is described which calculates the temperature of vapor emanating at the water surface, and the liquid fraction of released LNG that surfaces, if any, to form a pool on the water surface. The model includes such variables as the LNG release rate, diameter of the jet at release, depth of release and water body temperature. Results obtained from the model for postulated release conditions are presented. Comparison of predicted results with available LNG underwater release test data is also provided. © 2010 Elsevier Ltd.


Raj P.K.,Technology and Management Systems Inc.
Journal of Loss Prevention in the Process Industries | Year: 2011

Liquefied Natural Gas (LNG) storage facilities generally include channels to convey potential spills of the liquid to an impoundment. There is increasing concern that dispersion of vapors generated by flow of LNG in a channel may lead to higher than limit vapor concentrations for safety at site boundary from channels that may be close to the dike walls. This issue is of recent concern to regulatory agencies, because the calculation of vapor hazard distance(s) from LNG flow in a channel is not required under existing LNG facility siting standards or regulations.An important parameter that directly affects the calculated LNG vapor dispersion distance is the source strength (i.e., the rate of vaporization of LNG flow from the wetted channel surfaces, as a function of spatial position and time). In this paper a model is presented which considers the variation of the depth of the flowing LNG with spatial location and time, and calculates the spatial and temporal dependence of the mass rate of vapor generation. Self similar profiles for the spatial variation of the thermal boundary layer in the liquid wetted wall and liquid depth variation are assumed. The variation with time of the location of the liquid spread front and the evaporation rate are calculated for the case of a constant LNG spill rate into a rectangular channel. The effects of two different channel slopes are evaluated. Details of the results and their impact on dispersion distances are discussed. © 2011 Elsevier Ltd.


Raj P.K.,Technology and Management Systems Inc.
Fire Technology | Year: 2010

The flames over a burning liquid fuel are observed to spill over the downwind edge of the liquid pool in a wind. Empirical correlations in the literature relate the total base dimension of the fire (diameter + the spill over) with the wind Froude number. This leads to erroneous and physically incorrect (negative) value for the flame spillover at low wind speeds and/or in large diameter fires. The data from laboratory scale (0.1-0.6 m) to field scale (up to 35 m) fires of different hydrocarbon fuels on the wind induced flame "drag" or "spillover" were re-examined. The ratio of the flame spillover distance with the pool diameter is seen to vary in direct proportion to the square root of the Froude number but with different proportionality constants for different fuels. A physical model was developed to analyze the phenomena that occur at the base of a pool fire subject to a wind. This model indicates that the non-dimensional downwind flame spillover distance is directly proportional to the square root of the Froude number, inversely proportional to the square root of the dimensionless heat of combustion of the fuel and directly proportional to the 1/4th power of the ratio of vapor density to air density. Available experimental data are synthesized into a single correlation when plotted on the basis of the non-dimensional parameters from the model. This correlation includes the Froude number, the Damkohler number (dimensionless heat of combustion of the fuel), the wind flow Reynolds number and the fuel vapor-to-air density ratio. © 2009 Springer Science+Business Media, LLC.

Loading Technology and Management Systems Inc. collaborators
Loading Technology and Management Systems Inc. collaborators