The third phase of a joint industry project (JIP) on flashing liquid jets and two-phase droplet dispersion was recently kicked off by DNV Software. The aim of the project is to increase the under-standing of the behaviour of flashing liquid jets, and to improve the prediction of droplet atomisation and rainout.

Project management, theory and software development will be carried out by DNV Software, while experimental work will be carried out by Cardiff University (smallscale experiments) and INERIS (largescale butane experiments).

Background
Many accidents involve two-phase flashing releases of hazardous chemicals into the atmosphere. Rainout results in reduced concentrations in the remaining cloud, but can also lead to extended cloud duration because of re-evaporation of the rained-out liquid. For accurate hazard assessment, one must accurately predict both the amount of rainout and re-evaporation of the cloud.

Modelling of droplet dispersion, rainout and re-evaporation in Phast and Safeti

Previous first and second phases of JIP
The first phase of the JIP (ref. 1) was sponsored by DNV Software, HSE, ExxonMobil and ICI Eutech. It involved a detailed literature review on flashing liquid jets and two-phase droplet dispersion, and served to establish the state-of-the-art and provide recommendations. The review considered models and validation data for the sub-processes of droplet atomisation (including orifice characteristics), atmospheric expansion to ambient pressure, two-phase dispersion, rainout, pool formation and reevaporation.

The second phase of the JIP (ref. 2, 3) was sponsored by DNV Software, Arkema, Gaz de France and RIVM. It served to imple-ment and further validate/test the recommendations from Phase I. This involved the performance at Cardiff University of scaled two-phase experiments with water for conditions representing low to high superheat. The experiments measured velocity and droplet size distributions close to the orifice (hence post-expansion data). A criterion was derived for the transition between ‘low’ and ‘high’ superheat (non-flashing jets and fully flashing jets), and droplet size correlations were proposed in the regimes of mechanical breakup, transition to flashing and fully-flashing. Recommendations were made for the droplet size distribution (in all regimes), and the critical size of the droplet below which no rainout occurs. The new formulation was compared against a range of other droplet size and rainout correlations published in the literature, in conjunction with validation against an extensive set of experiments.

Third phase of JIP – additional experiments and further improved modelling
The objectives and confirmed scope of the third phase are as follows:

• Additional scaled water experiments at Cardiff University (Wales, UK) including measurements of dense sprays across the full relevant range of superheats.
• Large-scale butane experiments by INERIS (France). These will ensure more realistic scenarios and that derived droplet size correlations are accurate.
• Model validation and model improvements by DNV Software. This will include added validation from the above scaled and large-scale experiments and further refinement of droplet size correlations. It would also include simultaneously modelling of a range of droplet sizes (droplet parcels) and distributed rainout (rather than
  rainout at one point) to further improve the rainout prediction. It also includes further improvements to pool and dispersion modelling after rainout.
• Scoping and planning for potential further work.

The sponsors currently confirmed for the above scope of work are DNV Software, Gaz de France, Hydro, RIVM, and Statoil. The work on this commenced in September 2006.

Additional sponsors are currently being sought for an extension of the above work. This would include additional scaled experiments for butane, hexane, and possibly also LNG and gasoline, to ensure that the derived droplet size correlations are also valid for other chemicals than water (with fine-tuning of the correlation as necessary).

Phase III sponsors would have immediate access to the Phase II results. This includes complete documentation and access to the new versions of the DISC, ATEX, UDM models including improved droplet and rainout modelling. Following the Phase III work, it is the intention to implement the latest version of these models in a future version of Phast and Safeti.

Please contact the author in case you wish further information or are interested in sponsoring this work.