DOI 10.1007/s10512-016-0064-4 Atomic Energy, Vol. 119, No. 5, March, 2016 (Russian Original Vol. 119, No. 5, November, 2015)

DEVELOPMENT OF FUEL FOR RESEARCH REACTORS

A. V. Vatulin, V. B. Suprun, and G. V. Kulakov

UDC 621.039.546:621.039.542.34

The different stages in the development of fuel for research reactors are analyzed. Validation is given for high-density fuel based on the alloy U–9Mo for the fuel elements of swimming-pool type research reactors. Two effective methods of increasing the radiation resistance of the fuel composition, comprised of highdensity uranium-molybdenum fuel in an aluminum matrix, have been developed and experimentally corroborated. Reactor and post-reactor studies of mini fuel elements and full-size fuel assemblies with uranium-molybdenum fuel have been completed and show this fuel to be serviceable.

The development of dispersion fuel elements at the Bochvar All-Russia Research Institute for Inorganic Materials (VNIINM) began in 1953 along two directions: • fuel elements for commercial and research reactors; • fuel elements for propulsion reactors. G. A. Meerson and A. G. Samoilov stood at the sources of these two directions. Subsequently, their work was continued by Ya. D. Pakhomov and M. I. Solonin [1]. The distinguishing features of the directions founded by Meerson were: 1) focus on uranium dioxide as nuclear fuel; 2) choice of aluminum and its alloys as the material for the matrix and cladding; 3) use of powder metallurgy methods and pressure treatment to obtain kernel blanks; and 4) use of extrusion to obtain ready kernels and combined extrusion of kernel and cladding blanks for fabrication of ready fuel elements. Fuel elements of the KVK type for the first commercial heavy-water reactors for producing tritium and isotopes, uranium-graphite reactors of the type ADE, and others were developed in a short time on the basis of this technology. In order to use aluminum as a structural material for dispersion fuel elements, it was necessary to develop new aluminum-based radiation and corrosion resistant alloys. A team of specialists headed by E. S. Sarkisov and L. I. Kolobneva under the supervision of A. A. Bochvar created a special class of alloys AMSN, B1T, AZhN, and others [1]. A great achievement by VNIINM was the development of dispersion fuel elements for the water cooled and moderated reactor Ruslan (1979) and the heavy water reactor Lyudmila (1985). From 1979 to 1986, all domestic research reactors of the type MR, IRT-M, IVV-2M, VVR-M, TVR-S, and TVR-M as well as experimental reactors abroad were switched to the fuel elements developed at VNIINM. This made it possible to increase their operational parameters and safety considerably. Specialists at VNIINM working jointly with the Novosibirsk Chemical Concentrates Plant organized the production of an extensive list of fuel elements for commercial and research reactors. The dispersion fuel elements developed at VNIINM with a fuel kernel made of uranium dioxide in an aluminum matrix are structurally and technologically different from their foreign analogs, surpassing the latter in terms of some Bochvar All-Russia Research Institute for Inorganic Materials (VNIINM), Moscow, Russia. Translated from Atomnaya Énergiya, Vol. 119, No. 5, pp. 249–255, November, 2015

304

1063-4258/16/11905-0304 ©2016 Springer Science+Business Media New York

Fig. 1. Centrifugal spraying setup: 1) feeder; 2) electrode; 3) vacuum chamber; 4) rotating water-cooling crucible; 5) electric motor; 6) storage hopper.

Fig. 2. Uranium-molybdenum powder produced at VNIINM.

parameters. The high level of their design and technology was confirmed by many years of successful operation of commercial and research reactors. RERTR Program. An international program for lowering the fuel enrichment of research reactors to 19.75% 235U was accepted in 1978. VNIINM became the coordinator of this work in our country. Fuel elements with higher uranium density in a fuel kernel but lower enrichment developed as part of this program. The fuel elements and assemblies satisfied the following primary requirements: 1) the geometric dimensions of the fuel elements and assemblies and the thickness of the fuel kernel remained unchanged; 2) the residual reactivity and burnup of the fuel in the extracted fuel assemblies and the reactor power remained at the previous level; and 3) the higher cost of an assembly was due to only the increase in the mass of the extracted 235U. The fuel elements developed at VNIINM had technological reserves owing to which the 235U enrichment in the first stage decreased to 36% [2]. Subsequent reductions of the enrichment could have been accomplished by replacing the uranium oxide with more uranium-intensive fuel or by modernizing the fuel elements and assemblies. High-Density Fuel. High-density uranium alloys were picked on the basis of pre-reactor studies and reactor tests [2–4]. The fuel powder is fabricated by means of thermo-centrifugal spraying in a laboratory setup operating at VNIINM since 2001 (Fig. 1). The fractional composition of the powder made from uranium-molybdenum alloy falls into the range 100–140 μm (Fig. 2) [3]. 305

Fig. 3. Irradiation setup.

Fig. 4. Appearance of irradiated mini fuel elements.

Fig. 5. Macrostructure of transverse sections.

A special collapsible setup making it possible to test 32 mini fuel elements simultaneously was developed jointly with NIIAR in order to irradiate the fuel samples [2]. Four such setups can be installed in one channel (Fig. 3). The mini fuel elements can be examined during irradiation and the failed ones can be replaced. In 2003, two such setups were inserted into the MIR reactor for irradiation. The tests were conducted under conditions that are typical for swimming-pool type reactors. The tests were completed in February 2005. In all, 85 mini fuel elements were tested. Twenty nine of these elements attained maximum burnup 70%. Under irradiation, the cladding of one mini fuel element depressurized at average burnup 63%. Post-reactor studies showed that the mini fuel elements retained their shape; no mechanical or corrosion damage to cladding was found; tight cladding-kernel contact was maintained; there was no cladding-kernel interaction; a 10 μm thick interaction layer formed between the fuel particles and the matrix (Figs. 4–6). All types of nuclear fuel showed satisfactory radiation resistance under the conditions of swimming pool type reactors. The alloy U–9Mo was chosen for subsequent studies. Post-reactor studies of a depressurized mini fuel element showed that failure starts from the inner surface of the cladding along the ribbing. Large numbers of through microcracks were found at these locations (Fig. 7). Analysis of 306

Fig. 6. Microstructure of the kernel at burnup 65%: a, b) U–9Mo, γ-phase, density in terms of uranium 6 g/cm3; b) fractography of fracture; c, d) U–6.5Mo, (γ + α), density in terms of uranium 6 g/cm3; c) initial stage of pore formation; e, f) U–1.5Mo, αʹ-phase, density in terms of uranium 4 g/cm3; ƒ) structure of the interaction layer.

the stress-strain state showed that for a fuel element of this form stresses and strains are observed to concentrate along the ribbing at sites coinciding with crack nucleation in the cladding (Fig. 8). At the same time the particularities of the fabrication technology result in the formation of residual stresses at the same sites. In the aluminum alloy SAV-1 because of residual stresses, secondary phases precipitate along grain boundaries at the start of irradiation; this results in embrittlement of these regions of the cladding. This phenomenon is typical for such alloys and has been observed in other fuel elements in the past. For this reason, the relaxation of stresses that are due to kernel swelling occurs owing to the formation of microcracks along grain boundaries and not because of plastic deformation. Annealing was performed at the final stage in order to eliminate residual technological stresses. The final annealing of a fuel element after extrusion was introduced in the technological process of fabricating all subsequent rod-shaped fuel elements. In addition, other well-known aluminum alloys, such as 99 (Al–Ni–Fe–Cu) and SAV-6 (Al–Mg–Si–Cu), were used to fabricate mini fuel elements with modified fuel for heat-stressed research reactors. Increase of the Radiation Resistance of Uranium-Molybdenum Fuel. In the fuel composition UMo+Al, the alloy interacts with the aluminum matrix. To create a reliable high-density fuel for research reactors with heightened heat flux from the surface of the fuel elements, this interaction must be eliminated or greatly reduced. For this, the matrix aluminum alloyed with silicon in the range 2–13% and a protective coating was applied to the fuel particles. Several variants of mini fuel elements were fabricated at VNIINM [5, 6]. In accordance with the operating-life reactor tests, the parameters must meet the following specifications: the maximum temperature of a fuel kernel to 250°C and average 235U burnup 80%. The tests were stopped at average fuel burnup in mini fuel elements 75% (the average burnup in the most stressed mini fuel element is 78%) because the indications of the seal-tightness monitoring system were exceeded. 307

Fig. 7. Transverse section of mini fuel element with microcracks in the corners of the cladding after 50% burnup (a), typical form of microcracks at high magnification (b–d).

Fig. 8. Region of concentration of stresses (a) and strains (b).

Post-reactor studies [5, 6] showed that the thickness of the interaction layers of different fuel compositions is different. The presence of silicon in the aluminum matrix results in a significant reduction of the layer thickness; the zirconium nitride ZrN barrier coating almost completely prevents the matrix–cladding interaction and formation of an interaction layer. The protective properties of the oxide layer deposited on the grains to prevent interaction were not made clear (Fig. 9). Measurements of the thickness of the oxide layer on the surface of the fuel elements (in each section 5–15 measurements were performed on the boundaries and ribbing) showed that the corrosion of the alloy 99 cladding is somewhat higher. The maximum thickness of the film on the fuel elements is 82 μm in the case of cladding made from this alloy and 67 μm for the alloy SAV-6. Development of a Unified Fuel Element. At present, tubular fuel elements are used in research reactors. For each type of reactor, the fuel elements and assemblies are being developed individually (Fig. 10). The specialists at VNIINM proposed the concept of unified fuel elements. Any assembly of a research reactor can be comprised of such fuel elements [2]. Tubular fuel elements confer high operating efficiency of research reactors, but they come in a large number of sizes and configurations (about 40 types) and their fabrication process is highly complex. This makes it much more difficult to automate production and increases labor-intensiveness and fuel costs. At the same time, the technology of rod-shaped fuel 308

Fig. 9. Microstructure of mini fuel elements with average burnup 75%: a) PA-4 + U–9Mo (base variant); b) PA-4 + U–9Mo in the two-phase state; c) PA-4 + oxide coating; d, e) PA-4 + U–9Mo with a ZrN coating; ƒ) Al + 2%Si + U–9Mo; g) Al + + 5%Si + U–9Mo; h) Al + 13%Si + U–9Mo.

Fig. 10. Diagram of the transverse section of fuel assemblies [1] with standard tubular (right-hand side) and unified fuel elements (left-hand side).

elements is much simpler and can be automated. To fabricate fuel-element rods, it is necessary to use equipment with approximately 2.5–5-fold lower power. There is no machining of kernels or calibration drawing giving the required dimensions and shape. For this reason, complex and expensive instruments are not needed. 309

Fig. 11. Model of nondestructive testing setup: a) overall view; b) controlling head.

Fig. 12. Transverse section of fuel assembly with tubular (a) and rod (b) fuel elements.

Fig. 13. Design of experimental channel: 1) inlet tube; 2) tightening nut; 3) sealing ring; 4) aluminum case; 5) displacer; 6) fuel assembly; 7) end piece; 8) dimensions of rod fuel element.

The most complicated fuel-assembly design with rod-shaped fuel elements is one with a square shape – fuel assemblies of the type IRT [5, 6]. The specialists at the Novosibirsk Chemical Concentrates Plant performed a great deal of work on the design of rod-shaped fuel assemblies, perfecting the technology for fabricating full-scale fuel elements and assembling fuel assemblies. Hydraulic tests performed on one of them at VNIINM showed heat-engineering reliability in the most stressed reactor. In the course of the hydraulic tests, it was also shown that fuel elements do not vibrate at the working velocity of the coolant; the ribbing does not abrade; and, transport of an assembly does not lead to inadmissible changes in its state. A model of a setup for performing nondestructive monitoring of rod-shaped fuel elements was developed at VNIINM and tested at the Novosibirsk Chemical Concentrates Plant (Fig. 11) [4]. Reactor Tests of Full-Scale Fuel Assemblies. Four full-scale fuel assemblies of the IRT type with U–9Mo fuel (two regular with tubular and two with unified rod-shaped fuel elements) were fabricated at the Novosibirsk Chemical Concentrates Plant (Fig. 12) [5, 6]. Reactor tests of full-scale fuel assemblies were conducted in the working channels of the MIR reactor. 310

Fig. 14. Appearance of irradiated assembly of the type IRT after removal of the case (a), parts of fuel elements (b), bottom (c), and top (d) end.

Fig. 15. Thin section of finished articles.

The shape and size of the working channels were not suitable for IRT square fuel assemblies, so that a special collapsible experimental channel was developed (Fig. 13). According to the irradiation program, for two fuel assemblies (tubular and rod-shaped) the burnup was supposed to reach 40% and for the other two 60%. The tests confirmed the radiation resistance of the fuel composition U–9Mo+Al in swimming-pool type reactors. The serviceability of IRT type fuel assemblies with rod-shaped fuel elements was also confirmed. A fuel-element bundle retained its shape and no curving or displacement of the fuel elements was observed (Fig. 14) [5, 6]. The state of the surface of tubular and rod-shaped elements is good. The surface has a silvery color. The thickness of the oxide layer does not exceed 15 μm. Irradiation of both types of fuel elements showed that fission products migrated into the coolant. The increase of the radioactivity did not affect the attainment of the prescribed burnup. Post-reactor studies of depressurized rod-shaped and tubular fuel elements revealed technological defects, which show that individual fuel particles become embedded in the cladding. It was concluded on the basis of the results of these studies that the experimental technology and methods of monitoring must be improved. Experimental tubular fuel elements of several sizes were fabricated and investigated using the modified technology. The investigations showed that the fuel elements satisfy the design and technical specification for the IRT-3M experimental fuel assemblies with low-enrichment uranium-molybdenum fuel (Fig. 15) [7]. Conclusions. The high-density fuel U–9Mo for the fuel elements of swimming-pool type research reactors was substantiated. Two effective methods for increasing the radiation resistance of a fuel composition comprised of the high-density fuel U–9Mo in an aluminum matrix were developed and experimentally confirmed: alloying the aluminum matrix with silicon to 13% and coating the fuel particles with zirconium nitride. 311

The design and fabrication technology for the unified rod-shaped fuel element and fuel assembly based on it were developed. Reactor tests confirmed the performance of such fuel assemblies. Unified fuel elements make it possible to automate production and reduce the fuel fabrication costs. Successful reactor tests of full-scale IRT type fuel assemblies with high-density uranium-molybdenum fuel were conducted. The experimental technology can make it possible to fabricate tubular and rod-shaped fuel elements with high-density fuel U–9Mo in accordance with the technical requirements. Reactor experiments showed that these fuel elements can give nominal burnup to 60%. Post-reactor studies showed that the depressurization of individual fuel elements is due to technological defects in the form of fuel particles embedded in the cladding. Two experimental tubular fuel assemblies were fabricated by means of the optimized technology for reactor tests for purposes of licensing high-density fuel.

REFERENCES 1. 2. 3. 4.

5. 6. 7.

312

A. V. Vatulin, “State of and prospects for development of dispersion fuel elements for various water moderated and cooled reactors,” VNIINM – 50th Anniversary, Moscow (1999), Vol. 3, pp. 90–99. V. A. Starkov, V. V. Pimenov, N. G. Gataullin, et al., “Validation of the conditions of the in-pile tests of high-density low enrichment fuel for research reactors,” 7th Int. Meeting RRFM-2003, France, March 2003, pp. 33–38. A. Vatulin, A. Morozov, V. Suprun, et al., “Main results and status of the development of LEU fuel for Russian research reactors,” 9th Int. Meeting RRFM-2005, Hungary, April 2005, pp. 76–82. A. V. Vatulin, A. V. Morozov, V. B. Suprun, et al., “Development of IRT-type fuel assembly with pin-type fuel elements for LEU conversion of WWR-SM research reactor in Uzbekistan,” 8th Int. Meeting RRFM-2004, Germany, March 2004, pp. 184–187. A. V. Vatulin, I. V. Dobrikova, V. B. Suprun, et al., “Current status of the development of high-density LEU fuel for Russian research reactors,” Int. Meeting RERTR-2009, China, Nov. 2009, pp. 1–11. A. Savchenko, A. Vatulin, I. Dobrikova, et al., “Specific features of UMo fuel – Al matrix interaction,” 10th Int. Meeting RRFM-2006, Bulgaria, May 2006, pp. 121–125. A. V. Vatulin, I. V. Dobrikova, V. B. Suprun, et al., “Development of high-density fuel for research reactors,” Vopr. At. Nauki Tekhn. Ser. Materialoved. Nov. Mater., No. 1(74), 39–49 (2013).