Dispersion dense gas

A dense gas is defined as any gas whose density is greater than the density of the ambient air through which it is being dispersed. This result can be due to a gas with a molecular weight greater than that of air, or a gas with a low temperature due to auto-refrigeration during release, or other processes. 

The importance of dense gas dispersion has been recognized for some time. Early field experiments (Koopman et al., 1984 Puttock et al., 1982 Van Ulden, 1974) have confirmed that the mechanisms of dense gas dispersion differ markedly from neutrally buoyant clouds. When dense gases are initially released, these gases slump toward the grotmd and move both upwind and downwind. Furthermore, the mechanisms for mixing with air are completely different from neutrally buoyant releases. 

Reviews of dense gas dispersion and modeling are given by AIChE/CCPS (1987a, 1995b, 1996a), Goyal and Al-Jurashi (1990), Blackmore et al. (1982), Britter andMcQuaid (1988), Havens (1987), Lees (1986,1996), Raman (1986), and Wheatley and Webber (1984). 

Three distinct modeling approaches have been attempted for dense gas dispersion mathematical, dimensional and physical. 

The most common mathematical approach has been the box model (also known as top-hat or slab model), which estimates overall features of the cloud such as mean radius, mean height, and mean cloud temperature without calculating detailed features of the cloud in any spatial dimension. Some models of this class impose a Gaussian distribution equating to the average condition. 

There are cases in which the dispersion of a gas cloud is influenced by the fact that the density of the gas-air mixture differs from that of air. Such differences can result from differences in molar masses or from temperature differences between the air and the released gas. However, these differences only have an impact if the concentration of the gas is large enough. 

Many of the gases with relevance for accidents like hydrocarbons, chlorine, ammonia and oxygen can form clouds which are heavier than air. Whether a gas behaves as a dense gas or not depends on the following factors  

In [27] the following criteria are used to decide whether as gas is treated as dense or not  

A dense gas cloud behaves differently from a cloud of neutral density. It is not only dispersed in the direction of the wind, but also against this direction. It is flatter than a cloud of neutral density and the mechanism of mixing with the surrounding air is different. 

In its initial phase a dense gas cloud spreads less in the vertical direction than a cloud of neutral density. Yet, the belief that a dense gas cloud therefore migrates further than one of neutral density is not correct. The different mechanism of mixing with air leads to faster spreading especially under stable weather conditions. In the long run the density of a dense gas cloud becomes practically neutral due to mixing with air. A phase of passive dispersion, whose modelling was explained in the preceding section, ensues. 

The Britter and McQuaid10 model was developed by performing a dimensional analysis and correlating existing data on dense cloud dispersion. The model is best suited for instantaneous or continuous ground-level releases of dense gases. The release is assumed to occur at ambient temperature and without aerosol or liquid droplet formation. Atmospheric stability was found to have little effect on the results and is not a part of the model. Most of the data came from dispersion tests in remote rural areas on mostly flat terrain. Thus the results are not applicable to areas where terrain effects are significant. 

The model requires a specification of the initial cloud volume, the initial plume volume flux, the duration of release, and the initial gas density. Also required is the wind speed at a height of 10 m, the distance downwind, and the ambient gas density. 

The first step is to determine whether the dense gas model is applicable. The initial cloud buoyancy is defined as 

Britter and J. McQuaid, Workbook on the Dispersion of Dense Gases (Sheffield, United Kingdom Health and Safety Executive, 1988). 

A characteristic source dimension, dependent on the type of release, can also be defined. For continuous releases 

---

Initial plume volume flux for dense gas dispersion, voliime/time Continuous release rate of material, mass/time Instantaneous release of material, mass Release duration, time T Absolute temperature, K... 

A complete analysis of dense gas dispersion is much beyond the scope of this treatise. More detailed references are available (Britter and McQuaid, Workbook on the Dispersion of Dense Gases, Health and Safety Executive Report No. 17/1988, England, 1988 Lees, 1986, pp. 455 61 Hanna and Drivas, 1987 Workbook of Test Cases for Vapor Cloud Source Dispersion Models, AlChE, 1989 Guidelines for Chemical Process Quantitative Risk Analysis, 1989, pp. 96-103). 

Puttock, J. S. (ed.), "Stably Stratified Flow and Dense Gas Dispersion." Oxford University-Press, New York, 1988. 

EMGRESP is overly conservative for passive gas dispersion applications. No time-varying releases may be modeled. Dense gas dispersion may be computed for only "instantaneous" releases conditions. 

Colenbrander, G. W. and J. S. Puttock, 1983, Dense Gas Dispersion Behavior Experimental Observations and Model Developments, International Symposium on Loss Prevention and Safety Promotion in the Process Industries, Harrogate, England, September. 

Dc is the characteristic source dimension for continuous releases of dense gases (length), q0 is the initial plume volume flux for dense gas dispersion (volume/time), and u is the wind speed at 10 m elevation (length/time). 

Use the Britter-McQuaid dense gas dispersion model to determine the distance to the 1% concentration for a release of chlorine gas. Assume that the release occurs over a duration of 500 s with a volumetric release rate of 1 m3/s. The wind speed at 10 m height is 10 m/s. The boiling point for the chlorine is —34°C, and the density of the liquid at the boiling point is 1470 kg/m3. Assume ambient conditions of 298 K and 1 atm. 

LNG Vapor Dispersion Prediction with the DEGADIS Dense Gas Dispersion... 

LNG Vapor Dispersion Prediction with the DEGADIS Dense Gas Dispersion Model, Report GRI 0242, Gas Research Institute, Chicago, 111. 

Hanna, S.R., and Chang, J.C. (2001) Use of the Kit Fox field data to analyze dense gas dispersion modeling issues, Atmospheric Environment 35,2231-2242. 

Vol. 15 Toluene, the Xylenes and their Industrial Derivatives (Hancock, editor) Vol. 16 Dense Gas Dispersion (Britter and Griffiths, editors)... 

HAVENS, J.A., SPICER, T., LNG Vapor Dispersion Prediction with the DEGADIS Dense Gas Dispersion Model, Topical Report (April 1988 - July 1990), University of Arkansas, Fayetteville, USA (1990). 

The dispersion is passive or airborne if the released material is lighter than air. If it is heavier than air, on the other hand, we are dealing with dense gas dispersion. A combination of both types is possible. For example, refrigerated ammonia is dispersed as a dense gas in the first place. After being warmed up by the surrounding air its dispersion behaviour becomes passive. 

In what follows in the first place the airborne and thereafter the dense gas dispersions are treated. 

The model for dense gas dispersion in [27] is based on experimental results and similitude relations. It is to be preferred to the simple model presented in the next paragraph. 

In general one can state that the modelling of both airborne and dense gas dispersion comprises numerous problems which are stiU to be solved. This is particularly true for near field and if obstacles like buddings or industrial structures must be accounted for, which is the usual situation for releases from process plants. 

In what follows the simple model of Van Ulden [2, 30] is described. It gives an impression of the mechanisms of dense gas dispersion. In any case the model according to [27] should be preferred for practical apphcations. 

The differences of the results underline the modelling uncertainties still existent in dense gas dispersion. ... 

Mohan M, Panwar TS, Singh MP (1995) Development of dense gas dispersion model for emergency preparedness. Atmos Environ 29(16) 2075-2087... 

Dispersion of gases VDI models for airborne and dense gas dispersion (vid. Sect. 10.5)... 

Britter, R.E. and R.F. Griffiths (Editors), Dense Gas Dispersion, Elsevier Scientific Pub. Co., Amsterdam, 1982. 

In modeling the behavior of gas clouds, it is very important to select between passive/ buoyant and dense gas dispersion, as appropriate for the situation. 

Spicer, T., and J. Havens. 1989. User s Guide for DEGADIS 2.1 Dense Gas Dispersion Model, EPA Report EPA 450/4-89-019, U.S. Environmental Protection Agency, Cincinnati, OH. 

---