No abstract

No abstract

This report is issued as the first revision to FCRD-MRWFD-2016-000297.Zirconium may be recovered from the Zircaloy® cladding of used nuclear fuel (UNF) for recycle or to reduce the quantities of high-level waste destined for a geologic repository. Recovery of zirconium using a chlorination process is currently under development at the Oak Ridge National Laboratory. The approach is to treat the cladding with chlorine gas to convert the zirconium in the alloy (~98 wt % of the alloy mass) to zirconium tetrachloride.A significant fraction of the tritium (0-96%) produced in nuclear fuel during irradiation may be found in zirconium-based cladding and could be released from the cladding when the solid matrix is destroyed by the chlorination reaction. To prevent uncontrolled release of radioactive tritium to other parts of the plant or to the environment, a method to recover the tritium may be required.The focus of this effort was to (1) identify potential methods for the recovery of tritium from the off-gas of the zirconium recycle process, (2) perform scoping tests on selected recovery methods using nonradioactive gas simulants, and (3) select a process design appropriate for testing on radioactive gas streams generated by the engineering-scale zirconium recycle demonstrations on radioactive used cladding.It was determined that the off-gas from the zirconium recycle process is most likely to contain tritium as TCl in very dilute concentrations. This off-gas will also contain Ar and Cl 2 . An analysis of potential recovery methods for HCl/TCl identified a chemical conversion process and a liquid scrubbing process using sodium hydroxide solution as holding the most promise for the recovery of dilute HCl from an Ar/Cl 2 gas stream. Proof-of-principle experiments were designed to demonstrate HCl recovery with both potential processes using a non-radioactive simulant off-gas stream containing deuterium chloride (DCl). Use of DCl as a non-radioactive surrogate allows analysis of the liquids associated with each process by mass spectrometry. The chemical conversion process was designed such that the simulant gas stream would pass through two heated solid beds, with the first bed containing finely divided copper shot, or steel wool, to convert DCl to D 2 , and the second bed containing CuO pellets to convert D 2 to D 2 O. This then allowed recovery of D 2 O by condensation for analysis. The liquid scrubbing process was designed to absorb the DCl from the gas stream using NaOH solution, which could be analyzed to determine the amount of D that had transferred into the solution.Both methods were shown to recover D from the simulant gas stream. However, the recovery was not quantitative. The conversion and condensation method recovered up to 66% of the D based on mass recoveries, and up to 45% of the D based on instrumental analysis of the D/H ratios. An additional experimental test with DCl-containing simulant was performed in an effort to quantify the recovery of DCl by a liquid scrubbing system. Recovery of D in this test was ...

B.J. Riley, (a) J. McFarlane, (b) G.D. DelCul, (b) J.D. Vienna, (a) C.I. Contescu, (b) L.M. Hay, (a) A.V. Savino, (a) H.E. Adkins, (a) (a) Pacific Northwest National Laboratory

No abstract