Removal of Volatile Iodine from Gas Bubbles Rising in Water Pools: Review and Assessment of Pool Scrubbing Codes

During a hypothetical nuclear reactor accident with core damage the fission products released from the degrading fuel bundles often pass through aqueous beds before entering the containment, mitigating in part the source term. Several computer codes have been developed for predicting the fission product and aerosols removal in pool scrubbing scenarios. In addition to particle removal, these codes simulate the retention of some volatile iodine compounds. Nonetheless, experimental data on the matter are rather scarce and further validation remains to be done. In this work a review of volatile iodine removal models included in SPARC and BUSCA codes is presented. Besides, the results and discussions of a validation of both codes against the available experimental data are summarized. SPARC and BUSCA codes modell the diffusion of iodine towards the bubble interface by using the film penetration theory, which assumes a double layer gas-liquid at the interface. However there are some differences between the two models, mainly related to the boundary conditions in the aqueous volume for the diffusion of molecular iodine. In SPARC, a set of fast reactions in the liquid phase control both the molecular iodine concentration in the pool and the partition coefficient of iodine at the interface. Thus, the aqueous chemistry plays an important role in the boundary conditions for the diffusion proccess. On the contrary, the BUSCA model has no chemical considerations at all, and assumes a null iodine concentration in the water bulk. Several sensitivity studies have been made in order to weigh the effect of these differences. The variables examined in these studies were the pool temperature and the incoming iodine concentration in the pool. Additionaly, sensitivity studies focused on the steam mass fraction of the injected gas were performed to study the effect of the different approach of both models for the condensation proccess. The results showed a different sensitivity of SPARC and BUSCA to the incoming concentration. In SPARC, at low iodine concentrations in the pool (10" M) and depending on the pH conditions the hydrolysis considerably reduces the concentration of iodine in aqueous phase and also leads to an increase of the total partition coefficient at the interface, predicting as a consequence a high decontamination factor. In the case of high incoming concentrations (10" M), however, the amount of iodine remaining in equilibrium is substantial, leading to a greater accumulation of iodine in the gaseous side of the interface and to a lower decontamination factor. On the contrary, BUSCA predictions do not show any sensitivity to the incoming concentration, due to the absence of chemical interactions in the water volume. 1) CIEMAT, Nuclear Technology Institute, Avda Complutense 22, 28040 Madrid, Spain