IL H DESIGN FOR KRYPTON-85 ENRICHMENT BY THERMAL DIFFUSION Roger A. Schwind and William M. Rutherford Monsanto Research Corporation Mound Laboratory* Miamisburg, Ohio 45342 Substantial quantities*of krypton having a krypton-85 concentration of less than 10% will become available if nuclear fuel-processing plants are required to collect the gaseous fission products rather than releasing them into the atmosphere. A modular thermal diffusion unit was designed for the enrichment of the krypton-85 to useful concentrations of greater than 45%. The design emphasizes reliability and integrity by incorporating no moving parts within the unit. The modular design also offers flexibility in the size of the enrichment facility that need be constructed at any time. *Mound Laboratory is operated by Monsanto Research Corporation for the U. S. Atomic Energy Commission under Contract No. AT-33-1-GEN-53. This document is ILICLY RELEASABLE 1 f OJDQQJS
INTRODUCTION Krypton-85 is of value in scientific and technological applications. It is also a potential source of environmental contamination so that the recovery and retention of this noble gas may be required in the near future.»»» The nuclear power industry is expanding at a rapid rate, and the nuclear fuel reprocessing industry will also expand. A direct result of this will be an increased production rate of krypton-85 and its possible release to the environment during nuclear fuel reprocessing. Although krypton-85 release rates at reprocessing plants are presently deemed acceptable, release rates from larger plants under design will probably have to be controlled. In addition, krypton-85 has several uses based upon the fact that it gives off a 0.72 MeV beta particle, a 0.54 MeV gamma, and has a half-life of approximately 10 years. It has been used in radiation stimulated light sources, leak detection equipment and thickness gages. It has also been incorporated into solid materials by preparation of kryptonates. ' Higher 2
isotopic concentrations of krypton-85 than are available from nuclear fuel reprocessing would be useful for preparing kryptonates with greater specific activity. At concentrations of 40-50%, krypton-85 could possibly be used as a heat source fuel. This paper describes a modular thermal diffusion system for the isotopic enrichment of krypton-85 emanating from nuclear fuel reprocessing plants. CASCADE DESIGN Krypton available from fuel reprocessing is expected to have a Kr concentration of less than 10%. For the purposes of this discussion, it will be assumed that other techniques (such as adsorption) will be applied to separate krypton from other gaseous fission products. The exact concentration of the fission product gas will vary depending on the neutron flux of the reactor from which the fuel was recovered and the time elapsed since the fuel was removed from the reactor. In order to be consistent with a previous krypton-85 cascade design prepared by ORNL? and also in order not to be too optimistic with regard to the krypton-85 concentration in the feed gas, the following isotope composition 3
was assumed as typical for the fission product krypton gas. Volume /o Krypton-86 51 Krypton-85 4 Krypton-84 30 Krypton-83 15 Small variations from these concentrations would not affect the results or conclusions of this article. A double cascade of thermal diffusion columns was designed to enrich the krypton-85 to an enrichment of greater than 45%. In the first cascade the components lighter than J Kr are removed; in the second cascade the heavier components are removed. (Alternately, the roles of the two cascades can be reversed). This design problem was approached in two steps. The first step was the application of multicomponent cascade theory to determine the size and shape of the two idealized cascades. The sepond step was to approximate these smooth cascade shapes with three different types of thermal diffusion columns. 4
The performance of the squared-off cascades was then calculated by a computer program which describes the steady-state behavior of each cascade, respectively. The key weight or M* cascade design method originally described Q by De la Garza, 1963, was used to obtain the ideal double cascade size and shape required to obtain 45% krypton-85. Krypton-85 can be enriched in either of two separate double cascade systems. In one double cascade system the krypton-85 would be enriched at the top of the second cascade while in the other type the krypton-85 is enriched at the bottom of the second cascade. It was decided to use the double cascade system in which the krypton-85 is enriched at the top of the second cascade because the highly enriched stream would be physically more accessible and, therefore, easier operation would result. The other type of double cascade system ideally requires slightly less (2-3 per cent) separation capacity than the first arrangement. However, it is believed that the operational convenience outweighs the slightly smaller equipment that might be required in the other system. 5
Three different size columns were used to fit the smooth curve of the M* cascade shapes. The column dimensions, operating conditions, and calculated transport coefficients for these columns are given in Table 1. A schematic of the results of this squaring off of the cascades is shown in Figure 1. The steady-state flow rate and concentrations are also presented in Figure 1. The logarithm of the separation factor used in the calculations is equal to 80 percent of the theoretical value. It can be seen that 0.042 liter/day of krypton containing greater than 46% krypton-85 can be produced in a 16-column arrangement. The total power input to this 16-column system would be 61 kilowatts. EQUIPMENT CONCEPTS As described in the section on cascade design, an efficient separation system for krypton-85 necessarily requires two coupled multiple column cascades. Each cascade must be tapered so that it has a high transport capacity at the feed point and tapers to a low capacity and low holdup at the product ends. It has been normal practice at Mound Laboratory and elsewhere to assemble such cascades from a large number of identical thermal diffusion columns. Tapering is accomplished by variation. 6
of the number of columns in parallel along the length of the cascade. In principle this approach is effective; however, in practice parasitic flows occur in the parallel sections unless special, complex, timed valving arrangements are installed to interrupt the flows. Such cascades have been successfully operated at Mound Laboratory for a number of stable isotope separations, including krypton. Theoretical, computational and experimental techniques are now well established;» and the performance of these units can be predicted with confidence. In dealing with radioactive materials, it is undesirable to Q use complicated valving systems. In addition, it is undesirable to have any components such as interstage circulation pumps which would be subject to failure or which would require periodic service. A system which meets these severe limitations has been successfully worked out. The proposed system has these major features: 1. Effective taper of the cascades is accomplished by variation of the cold wall diameter. There are no columns in parallel. 7
2. The dual cascade is a single modular unit approximately 30 cm in diameter and 10 m long. (A schematic of the modular arrangement is shown in Figure 2.) 3. There are no moving parts within the unit. Interstage circulation is established by thermosiphon convection. 4. Production capacity can be increased indefinitely by the addition of modular units. 5. The unit can be completely sealed by welding or brazing with the exception of valve outlets for feed, product waste, and samples. 6. For shielding, each unit can be placed in a small diameter cased hole in the ground. The proposed column design is quite similar to that used routinely 12 at Mound Laboratory. The hot wall is the outer surface of a 4-mm OD tubular electric heater with a 7.3 m long active section. The cold wall is a water cooled stainless steel tube, the inside diameter of which depends on position of the column in the cascade. In keeping with standard practice at Mound Laboratory, 8
spacers are spot welded to the heater at 40 cm intervals to maintain axial alignment of the heater. The columns comprising a cascade are connected in series by thermosiphons. The use of thermosiphons for this purpose is well known. The principal drawback of such devices, high gas holdup, is offset by the advantage that there are no moving parts and no seals. Therefore, maintenance is minimized. The columns of the two cascades are placed in a circle around a common thermosiphon heater, and the assembly is enclosed in a single water jacket. With reasonable care in layout, it should be possible to insert the assembled system inside an 18-in diameter cased hole in the ground. Detailed operational tests with non-radioactive krypton would be necessary to verify design calculations. The non-radioactive experiments would verify calculated column parameters and interstage transport rates. With this experimental information at hand, subsequent performance with krypton-85 could be quite 13 accurately predicted. 9
At the completion of this testing program, a prototype unit would be available for installation in a suitably shielded and ventilated facility for production of krypton-85. This design is applicable to a wide range of production rates since the choice of the number of modules in the facility is arbitrary. SUMMARY The above modular thermal diffusion unit is a reliable and flexible design. Reliability is a necessity to minimize the probability of any release of krypton-85 into the environment. Flexibility is desirable to allow for future increases in production. 10
Table 1. Column Parameters Column A Column B Column C Effective Length, L, cm 731.5 731.5 731.5 Cold Wall Radius, cm. 0.687 0.629 0.549 Hot Wall Radius, cm. 0.4 0.4 0.4 Cold Wall, Temperature C 44 44 44 Hot Wall, Temperature C 750 750 700 Operating Pressure, Torr 3146 3146 3146 Initial Transport Coeff. 2.959 x 10~ 5 1.389 x 10" 5 0.338 x 10" 5 H 0, gram/sec. Convective Remixing Coeff. 0.03189 5.861 x 10" 3 2.816 x 10-4 K c, gram cm/sec. Diffusive Remixing Coeff. 6.217 x 10' 4 4.752 x 10~ 4 2.816 x 10~ 4 &d> gram cm/sec K c /K d 51.3 12.3 1.0 11
CASCADE I CASCADE II Top Product I _ Top Product 0.042 liter/dayf II R B 0.37% Kr-86 1.44% Kr-85 63.50% Kr-84 34.69% Kr-83 C B 7.69% Kr-86 46.42% Kr-85 45.38% Kr-84 0.51% Kr-83 A B Feed I 0.737 liter/day,- 51% Kr-86 4% Kr-85 30% Kr-84 15% Kr-83 A A A B Bottom Product I, Feed II 0.419 liter/day/ 89.45% Kr-86 5.94% Kr-85 4.56% Kr.84 0.05% Kr-83 B A A B * B B ^ Bottom Product 0.377 liter/day r- II 98.47% Kr-86 1.47% Kr-85 0.06% Kr-84 i- 1 Figure 1 - Double cascade design and steady-state performance for krypton-85 enrichment. A, B, and C designate three thermal diffusion columns having different operating characteristics.
Feed I Common, central heated chamber for thermosiphon lines Bottom product I Top product I» Cooling water Bottom product II Top product II Feed II Cooling water \ mm. MM,nji,n n., o.ti-r o_ Cooling water Water jacket (thermosiphon return lines are inside jacket) irai Figure 2 - Modular thermal diffusion unit for krypton-85 enrichment. (Top and frontal views are shown.) 13
REFERENCES 1. C. A. Rohrmann, "Fission Product Xenon and Krypton - An Opportunity for Large-Scale Utilization", Isotop. Radiat. Technol., 8(3): 253-260, (Spring 1971). 2. M. J. Stephenson, J. R. Merriman, and D. I. Dunthorn, Application of the Selective Absorption Process to the Removal of Krypton and Xenon from Reactor Off-Gas, Union Carbide Corporation, Nuclear Division, Oak Ridge Gaseous Diffusion Plant, February 14, 1972 (K-L-6288). 3. 0. 0. Yarbro, J. P. Nichols, W. E. Unger, "Environmental Protection During Fuel Processing", Oak Ridge National Laboratory, paper presented at 72nd National AIChE Meeting, St. Louis, May, 1972. 4. C. L. Bendixsen, G. F. Offutt, B. R. Wheeler, "Rare Gas Recovery Facility at the Idaho Chemical Processing Plant", Idaho Nuclear Corporation, paper presented at winter meeting of American Nuclear Society, San Francisco, November 30- December 4, 1969. 14
REFERENCES (Contd.) 5. J. E. Carden, "Preparation, Properties, and Uses of Kryptonates in Chemical Analyses", Isotop. Radiat. Techno1. 3(3): 206-14 (Spring 1966). 6. J. E. Carden, "Applications of the Kryptonates in Materials Research", Isotop. Radiat. Technol. 3(4): 318-28 (Summer 1966). 7. K. H. Lin, A Conceptual Design of a Continuous Thermal Diffusion Plant for 5 Kr Enrichment, Union Carbide Corporation, Nuclear Division, Oak Ridge National Laboratory, February 1969, (ORNL-4372). 8. A. De la Garza, The Variable Key Weight Cascade for Multi-Component Isotope Separation, Union Carbide Corporation, Nuclear Division, Oak Ridge Gaseous Diffusion Plant, 1963, (K-1571). 9. Stable Gaseous Isotope Separation and Purification: April- June 1972, MLM-1943 (August 18, 1972), pp. 9-10. 15
10. Stable Gaseous Isotope Separation and Purification: October-December 1969, MLM-1614 (May 8, 1970), pp. 42-50. 11. Stable Gaseous Isotope Separation and Purification: January-March 1972, MLM-1904 (July 10, 1972), pp. 8-15. 12. W. M. Rutherford, F. W. Weyler, and C. F. Eck, "Apparatus for the Thermal Diffusion Separation of Stable Gaseous Isotopes", Rev. Sci. Inst., 39 (1), pp. 94-100, (1968). 13. W. J. Roos and W. M. Rutherford, "Experimental Verification, with Krypton, of the Theory of the Thermal-Diffusion Column for Multicomponent Systems", J. Chem. Phys., 50(1), pp. 424-429, (1969). 16