Industrial PhD project - Modelling premixed combustion with non-premixed liquid substructures

The main goal of the PhD project is to implement and validate a combustion model for aerosols in the computational fluid dynamics (CFD) software FLACS. Whereas the phenomena involved in release, dispersion and combustion of gaseous fuels have been investigated quite extensively, this is not the case for aerosol releases and flame propagation in the resulting mists and sprays. An aerosol is a cloud of liquid droplets dispersed in air, and the size of the droplets is of primary concern with respect to the rate of combustion, and hence the safety level in the facility where such clouds can be generated. The work in 2019 focused on two aspects: the writing of the thesis and development of the combustion model. An improved combustion model that is compatible with the FLACS code has been developed and tested. In order to explore the limitations of the model and its range of use, the model has been tested against results from 6 different scientific papers. The integration of the combustion model into the FLACS solver will now be undertaken, and the code will be ready for simulating validation cases in early 2020. The writing of the thesis has progressed according to the plan, and only the section related to the CFD modelling remains to be written in 2020.

-

Some large-scale accidents thought to be vapour cloud explosions in the petrochemical industry, are believed to be or to be combined with aerosol explosions. To cite a few with devastating consequences, there are Flixborough (England, 1974), Piper Alpha ( North Sea, 1988) and Buncefield (England, 2005). Moreover, aerosols are also encountered in major transformer accidents as Tonstad (Norway, 1973) or Roncovalgrande (Italy, 1988). FLACS is a CFD tool able to handle gas explosions. For risk assessment purp oses, gas is used to represent aerosols in the current version of the software. This solution works but is not optimal. The differences in physics between explosions in aerosols and gas explosions regarding ignition, dispersion and combustion are not acco unted for in the modeling. While release of gases is extensively investigated in the literature, liquid releases (aerosols) is far less investigated in terms of explosion risk and underestimated. In order to get a better knowledge of aerosol explosions, G exCon AS has performed series of experiments through the programs LICOREFLA and MEXOS for risk assessments offshore, and SEBK for transformer risks. Recent work points out that FLACS the necessity of a tool for predicting explosion hazards of two-phase e xplosions. The proposed doctoral work is part of the scheme for the future development of FLACS regarding two-phase media and establishment of a numerical model for combustion of non-premixed substructures with focus on mist/spray explosions that could be extended to hybrid and dust explosions.