ABSTRACT
A great deal of effort has been expended in order to generate a database which can be used to validate primary circuit aerosol transport codes and to highlight their 1 imitations, however certain aspects are only poorly represented. Several missing phenomena and uncertainties arising from particular modelling aspects have been identified. Most of the individual models for aerosol deposition and agglomeration contained in the transport codes are supported by small-scale, single-effects experiments. Large-scale experiments have confirmed that the methodology used is basically sound but that several phenomena not modelled in the codes may be important under certain accident conditions.
Typically, missing phenomena include aerosol resuspension, the behaviour of thick deposits of aerosols, deposition in bends in pipework, deposition between adjacent volumes, chemical interactions between vapours and aerosols, and possible interactions between different mechanisms. Uncertainties in the results of calculations performed using these codes also arise from limitations in the databases supporting each individual model, the paucity of suitable thermodynamic data to identify the chemical species being transported, modelling compromises made in order to simplify calculations. Uncertainties in the thermal hydraulics codes may play a significant role in limiting the accuracy of the transport calculations, and in particular the modelling of aerosol nucleation is sensitive to predictions of temperature and temperature gradient. In addition, the accuracy of aerosol codes may be dependent on the numerical schemes incorporated into each code, and exhaustive testing and comparisons are required to ensure that numerical effects do not dominate the physics of each particular calculational problem.
392A significant amount of work remains to be done to address the outstanding problems and many of the items discussed are under investigation at present. Some consideration is given to the priorities to be attached to each of these items. Developments of thermal hydraulic modelling may also be required if this proves to be the major source of uncertainty. Care must be taken to ensure that the codes, and all their inherent assumptions and simplifications, are not taken for granted and become accepted as “tablets of stone”; new numerical methods and relevant data should be exploited wherever this proves to be of benefit to the accuracy and reliability of the codes.
