Zeolite membranes for the separation of krypton and xenon from spent nuclear fuel reprocessing off-gas
Crawford, Phillip Grant
MetadataShow full item record
The goal of this research was to identify and fabricate zeolitic membranes that can separate radioisotope krypton-85 (half-life 10.72 years) and xenon gas released during spent nuclear fuel reprocessing. In spent nuclear fuel reprocessing, fissionable plutonium and uranium are recovered from spent nuclear fuel and recycled. During the process, krypton-85 and xenon are released from the spent nuclear fuel as process off-gas. The off-gas also contains NO, NO2, 129I, 85Kr, 14CO2, tritium (as 3H2O), and air and is usually vented to the atmosphere as waste without removing many of the radioactive components, such as 85Kr. Currently, the US does not reprocess spent nuclear fuel. However, as a member of the International Framework for Nuclear Energy Cooperation (IFNEC, formerly the Global Nuclear Energy Partnership), the United States has partnered with the international nuclear community to develop a “closed” nuclear fuel cycle that efficiently recycles all used nuclear fuel and safely disposes all radioactive waste byproducts. This research supports this initiative through the development of zeolitic membranes that can separate 85Kr from nuclear reprocessing off-gas for capture and long-term storage as nuclear waste. The implementation of an 85Kr/Xe separation step in the nuclear fuel cycle yields two main advantages. The primary advantage is reducing the volume of 85Kr contaminated gas that must be stored as radioactive waste. A secondary advantage is possible revenue generated from the sale of purified Xe. This research proposed to use a zeolitic membrane-based separation because of their molecular sieving properties, resistance to radiation degradation, and lower energy requirements compared to distillation-based separations. Currently, the only commercial process used to separate Kr and Xe is cryogenic distillation. However, cryogenic distillation is very energy intensive because the boiling points of Kr and Xe are -153 °C and -108 °C, respectively. The 85Kr/Xe separation step was envisioned to run as a continuous cross-flow filtration process (at room temperature using a transmembrane pressure of about 1 bar) with a zeolite membrane separating krypton-85 into the filtrate stream and concentrating xenon into the retentate stream. To measure process feasibility, zeolite membranes were synthesized on porous α-alumina support discs and permeation tested in dead-end filtration mode to measure single-gas permeance and selectivity of CO2, CH4, N2, H2, He, Ar, Xe, Kr, and SF6. Since the kinetic diameter of krypton is 3.6 Å and xenon is 3.96 Å, zeolites SAPO-34 (pore size 3.8 Å) and DDR (pore size 3.6 Å) were studied because their pore sizes are between or equal to the kinetic diameters of krypton and xenon; therefore, Kr and Xe could be separated by size-exclusion. Also, zeolite MFI (average pore size 5.5 Å) permeance and selectivity were evaluated to produce a baseline for comparison, and amorphous carbon membranes (pore size < 5 Å) were evaluated for Kr/Xe separation as well. After permeation testing, MFI, DDR, and amorphous carbon membranes did not separate Kr and Xe with high selectivity and high Kr permeance. However, SAPO-34 zeolite membranes were able to separate Kr and Xe with an average Kr/Xe ideal selectivity of 11.8 and an average Kr permeance of 19.4 GPU at ambient temperature and a 1 atm feed pressure. Also, an analysis of the SAPO-34 membrane defect permeance determined that the average Kr/Xe selectivity decreased by 53% at room temperature due to unselective defect permeance by Knudsen diffusion. However, sealing the membrane defects with polydimethylsiloxane increased Kr/Xe selectivity by 32.8% to 16.2 and retained a high Kr membrane permeance of 10.2 GPU at ambient temperature. Overall, this research has shown that high quality SAPO-34 membranes can be consistently fabricated to achieve a Kr/Xe ideal selectivity >10 and Kr permeance >10 GPU at ambient temperature and 1 atm feed pressure. Furthermore, a scale-up analysis based on the experimental results determined that a cross-flow SAPO-34 membrane with a Kr/Xe selectivity of 11.8 and an area of 4.2 m2 would recover 99.5% of the Kr from a 1 L/min feed stream containing 0.09% Kr and 0.91% Xe at ambient temperature and 1 atm feed pressure. Also, the membrane would produce a retentate stream containing 99.9% Xe. Based on the SAPO-34 membrane analysis results, further research is warranted to develop SAPO-34 membranes for separating 85Kr and Xe.