Testing of an Integrated Iodine Scrubber and Polishing Bed System
In 2016 an initial engineering evaluation and design of the off-gas abatement systems required for a hypothetical 1000 t/yr used nuclear fuel (UNF) reprocessing facility treating 5 yr-cooled, 60 GWd/tIHM UNF was completed. In this study two types of silver-based iodine sorbents were considered. The first was a silver-exchanged mordenite (AgZ), and the second was a silver-functionalized silica aerogel. In the initial design of the dissolver off-gas (DOG) iodine capture beds, the sorbent capacity
... he sorbent capacity was assumed to be 40 mg I/g sorbent. The total iodine released from the fuel, used as input for the design of the DOG system, was 1.635 kg/day. Tramp halogens increased this to ~5.1 kg/day on an iodine equivalent basis. The daily consumption of iodine sorbent was ~166 kg (including the mass transfer zones that have lower iodine loading). The iodine capture system was designed to utilize five beds in parallel. In this design each bed was ~1 m in diameter and 1.5 m long. This design allowed the beds to remain on line for 20 days, but at that point all five beds would need to be replaced, resulting in considerable material handling. It was proposed that the bed life could be extended and the number of parallel beds reduced if an aqueous iodine scrubber preceded the solid sorbent iodine capture beds. The testing described in this report was intended to provide an evaluation of an integrated system in which bulk iodine would first be removed by an aqueous NaOH scrubber with a target iodine decontamination factor (DF) of 10 to 50. This would reduce the sorbent usage by 90 to 95% and significantly reduce the associated frequent remote handling of the iodine filter beds. The effluent from the scrubber would then be passed through an AgZ sorbent bed to remove residual iodine. The integration of these unit operations was intended to demonstrate an iodine DF of >1000, which would facilitate regulatory compliance with regards to total volatile radionuclide emission from a reprocessing facility. Initial baseline performance testing of the scrubber system includes the characterization of CO 2 removal by the aqueous NaOH scrubber as a function of temperature, caustic concentration, packing length, liquid-to-gas mass ratio, and liquid flow rate. It was found that when the column was well wetted, very high CO2 DFs were achieved (>650 and potentially as high as 15,000), preventing the ability to elucidate the effects of each specific variable tested. The L/G ratio in this system was not representative of typical L/G ratios found in industrial-scale scrubber systems. By significantly lowering the liquid recirculation rate to the point that column wetting was marginal, some significant variations in the CO2 DF or CO2 effluent concentrations could be observed. The additional tests conducted under these more adverse conditions still showed CO2 DFs >100. The only operational parameter with significant effects was determined to be column length, with increased column length resulting in higher CO2 DFs. Following these CO2 tests, a single iodine scrubber test was performed deliberately at a condition indicated by the CO2 demonstration to result in poor column performance. This was intended to establish a lower performance bound for iodine removal by the aqueous scrubber. The result showed that even under the adverse operating conditions, an iodine DF of >100 could be achieved. Finally, the aqueous scrubber and a AgZ polishing bed were integrated. This simple caustic scrubber system achieved an iodine DF of >12,000 for just the scrubber portion. Using this iodine DF, the iodine loading on the polishing bed was reduced by 99.993%. The performance of that integrated system was compared to a system including only a AgZ sorbent bed. The system without the scrubber also showed iodine DFs >3000 until breakthrough was observed. The loading on the saturated portion of the bed averaged 87.2 mg I2/gm AgZ. Comparing the penetration of iodine into the sorbent beds for these two systems showed a marked difference; the penetration with the scrubber was only 0.65 cm at a maximum loading of 0.11 mg I2/gm AgZ, whereas for the case without the caustic scrubber, the bed was saturated for the first 7 cm and breakthrough of the 13.3 cm bed occurred. This page is intentionally left blank.