STATUS OF HIGHLY LOADED URANIUM-ZIRCONIUM HYDRIDE LOW ENRICHED URANIUM (LEU) FUEL PROGRAMS. G. B. West GA Technologies Inc.

Transcription

1 STATUS OF HIGHLY LOADED URANIUM-ZIRCONIUM HYDRIDE LOW ENRICHED URANIUM (LEU) FUEL PROGRAMS G. B. West GA Technologies Inc. San Diego, CA The long term test program for U-ZrH LEU (less than 20$ enrichment) fuels is nearing completion and fabrication of this versatile, rugged, long lasting fuel has been completed for several operating reactors and continues for new orders. In early 1976, GA undertook the development of fuels containing up to 45 wt-$ uranium (3.7 gm U/cc) in order to allow the use of LEU to replace highly enriched fuels while maintaining long core life. The 45 wt-$ fuel contains a relatively modest 20 volume percent of uranium. At GA these highly loaded LEU fuels were subjected to water quench tests from 1200 C, in-core tests of over 2000 thermal cycles where the reactor was operated from shutdown to powers of 1 to 1.5 MW, and pulsing tests to 725 C (reactor P >2000 MW). The pulsing tests were performed with fuel rods with 0.5 inch diameter as well as 1.5 inches O.D. Demonstration of the U-ZrH LEU fuel is presently being performed by a full-scale, long-term fuel burnup test at the 30 MW Oak Ridge Research Reactor (ORR) as part of the U.S. Department of Energy RERTR program managed by Argonne National Laboratory (ANL). A 16-rod LEU cluster containing fuel rods with nominal 0.5 inch O.D. is being tested. The uranium loadings have included 20, 30 and 45 wt-$ and the active fuel length is 22 inches. The initial test objectives were to reach burnup values of 35, 40 and 50$ of contained U-235 respectively in these fuels. The test began in December 1979 and these objectives have been successfully met. Since June 1982, only 45 wt-$ fuel has been in-core. In November 1983 (after 901 full power days of irradiation) one 2-1

2 of the 45 wt-$ fuel rods failed by what appears to be steam pressure resulting from "water logging". Based on fission product analysis performed by ORNL, the failed rod has about 55$ burnup of the initial 55 gm loading of U-235. This corresponds to about 26 MWD of burnup. Most of the other fuel rods have more burnup than the failed rod - up to 20$ more - and no abnormalities are apparent in the other fuel rods. Hot cell examination of the failed rod is still in progress. One important result evident from the ORR tests is the excellent dimensional stability of the fuel rods. The individual rods and the cluster were designed with extensive bending restraint, but only very minor bending has been experienced. Several reactor facilities have already begun conversion to TRIGA LEU fuel. Currently, U-ZrH LEU fuel is in use in a mixed plate and rod fuel configuration in the 1.0 MW reactor at the National Tsing Hua University in Taipei, Taiwan. A complete core of 20 wt-$ LEU fuel has been fabricated for the 3 MW forced flow TRIGA Mark II reactor system with pulsing capability scheduled to be completed in 1984 in Bangladesh. Also, a complete core of 20 wt-$ 4-rod shrouded cluster fuel has been fabricated for conversion and upgrading of the PRR-1 reactor in the Philippines. The converted reactor will operate at 3 MW with forced cooling and have pulsing capability. Startup is now scheduled for early U-ZrH LEU fuel has also been delivered to existing TRIGA reactor facilities in Malaysia, Thailand and Yugoslavia and is being put to use as new fuel is needed to meet burnup requirements. For stepwise conversions of higher power reactors (>5 MW) where flow distribution and neutron spectrum effects are of greater importance, a detailed analysis and study is a prerequisite. A joint program involving GA and ANL is now in progress for analysis of a single U-ZrH LEU fuel test cluster operating in a 5 MW HEU plate-type core. Inserting this one cluster would be the initial step in a step-wise conversion of the HEU plate-type core to U-ZrH LEU fuel. 2-2

3 INTRQPUCXIQN During the natural development of uranium zirconium-hydride research reactor fuel technology, reasonable core reactivity lifetimes for reactors operating at higher power levels and at high duty cycles were achieved by increasing the uranium-235 enrichment to 70$ and 93$ and by incorporation of erbium in the fuel. These high enriched uranium (HEU) fuels preserved all of the unique safety features of the TRIGA fuel system - prompt negative temperature coefficient of reactivity, high fission product retentivity, chemical stability when quenched from high temperatures in water, and dimensional stability over a large range of operating temperatures. Following the adoption of policies by the U.S. Government and other fuel supplier nations to limit, with few exceptions, export of fuels to those enriched to less than 20$ in U-235, GA undertook, starting in 1976, a rigorous program to develop uranium-zirconium hydride fuels with higher uranium concentrations (up to 3.7 grams U/cc) while limiting the enrichment to less than 20% in U-235. By increasing the uranium concentration from the nominal 8.5$ wt-$ up to 45 wt-$ and by including erbium as a burnable poison, the long reactivity lifetime of the HEU cores has been preserved. These LEU fuels have been extensively tested out-of-pile and in limited in-pile pulse testing, and are now completing long term burnup testing in the Oak Ridge Reactor. The significant results of fuel characterization, out-of-pile testing, and in-pile testing of the uranium-zirconium hydride LEU fuels will be summarized below and the current status of performance verification presented. 1. Fuel Characterization and Testing A summary of the various types of TRIGA fuels, developed over the past 25 years, is shown in Figure 1.* It can be seen from this table that GA has produced uranium-zirconium hydride fuel with uranium concentrations up to 12 weight percent for the TRIGA-ACPR reactors, the first of which was Figures appear at the end of the report. 2-3

4 placed in operation in Also work on the SNAP reactor program included the development of uranium-zirconium hydride fuels with uranium concentrations up to 20 weight percent. Under the LEU fuel development program at GA, the performance of uranium-zirconium hydride fuel with a uranium concentration of 45 weight percent has been demonstrated by evaluation of the parameters listed on Figure 2. Analytical performance verification includes evaluation of the parameters listed on Figure 3. Fabricability verification was conducted by the production of about 100 pieces of uranium-zirconium hydride fuel by the standard production techniques of induction melting and furnace hydriding. No difficulties were experienced with the fabrication processes although none were anticipated since the uranium volume is still less than 20 percent in the 45 weight percent alloy. The results of metallographic and microprobe tests indicated that the grain structure, phase distribution, and homogeneity were well within normal and acceptable limits (Figure 4). Based upon the completely satisfactory results of the fabricability testing and metallurgical examination, macrolevel and physical property testing was conducted, including: Fission Production Release Physical Property Determination Hydrogen Pressure Quench Testing The fission product release testing of 45 weight percent uranium fuel showed some scatter about the correlation curve (figure 5), but basically was consistent with the data obtained for lower uranium concentrations. This characteristic continues to be a significant element in contributing to the safety of the TRIGA fuel - moderator system. Thermal diffusivity, a, from which the thermal conductivity is derived, is a physical property of considerable interest and shows little 2-4

5 variation with uranium concentration (Figure 6). Similarly, quench testing of fuel samples from 1200 C in water (Figures 7 and 8) indicated completely satisfactory performance of the LEU uranium-zirconium hydride fuel. Hydrogen pressure buildup was as expected - not influenced by the uranium concentration - as deduced from examination of hydrogen parameters recorded during the hydriding process. The culmination of characterization testing was thermal cycling tests to demonstrate dimensional stability. In these tests, samples were cycled through the uranium phase change temperature of about 680 C for over 100 cycles in the laboratory and 40 pulses in-core. The results of the latter tests are shown in Figure 9«These tests were on 1/2 in. diameter fuel rods with no intermediate support between the top and bottom of the element. 2. Performance Characteristics The success of the TRIGA research reactor over the past 26 years can be closely related to the prompt negative temperature coefficient of reactivity and related safety characteristics. Thus a high priority item in evaluation of the LEU fuel was determination of the temperature coefficient. Nuclear design and analytical studies have shown that the prompt negative temperature coefficient, for the 20 weight percent uranium fuel is essentially the same as that for standard fuels over the temperature range of interest (20 to 700OC). The prompt negative temperature coefficient for the more highly loaded LEU fuel shows a temperature dependence, whereas the coefficient is relatively constant for standard fuel. The value of the prompt negative temperature coefficient of reactivity is slightly lower for the 45 weight percent uranium fuel; however it is still large and significantly higher than the prompt negative temperature coefficients for any other type of reactor fuel (Figure 10). The thermal neutron flux of a core loaded with TRIGA-LEU fuel is depressed, as would be predicted, compared to the flux attained by a comparable 8.5 weight percent uranium fueled core. However, the flux in 2-5

6 experiment locations (reflector and in-core) is very similar to the reactor with standard fuel, and the flux performance is quite comparable to that predicted for a plate-type fueled core with LEU fuel (Figure 11). Inclusion of erbium burnable poison in the TRIGA LEU fuel has enabled core lifetimes of up to 7000 MWD to be predicted for the 45 weight percent uranium fuel. This corresponds to about 4000 MWD for a core configured for operation at 10 MW (Figure 12). This long reactivity lifetime results in the lowest predicted cost per MWD of any of the LEU-fuels currently commercially available (Figure 13). 3. In-Pile LOOP Testing The final demonstration of the TRIGA-LEU fuel is being performed by a full-scale long-term fuel burnup test which is underway at the 30 MW Oak Ridge Research Reactor (Figure 14) as part of the U.S. Department of Energy RERTR program managed by Argonne National Laboratory. A standard geometry 16-rod TRIGA LEU cluster (Figure 15) containing uranium loadings of 20, 30, and 45 weight percent uranium (Figure 16) was placed under test beginning December 1979 and testing is still in progress. Some of the fuel test parameters are shown in Figure 17. The fuel rods containing 20 and 30 weight percent uranium have exceeded target burnups (-50 percent and -47 percent, respectively) and were removed from the test cluster in May of During 1983, one fuel rod of each test type - shown in Figure 18 - was moved to the ORNL hot cell for post-irradiation examination. Measurements and metallographic examinations have shown fuel swelling to be as predicted and fuel characteristics appear normal. Fission product evaluations have allowed a preliminary determination of burnup vs time for each fuel rod. The curves for the 45 wt-j rods are shown in Figure 19. The irradiation of the 9 rods with 45 wt-$ U h s continued since May 1982 with the configuration shown in Figure

7 One important result evident from the ORE tests is the excellent dimensional stability of the fuel rods. The individual rods and the cluster were designed with extensive bending restraint, but only very minor bending has been experienced. In November 1983 (after 901 full power days of irradiation) one of the 45 wt-$ fuel rods (location Z1) failed by what appears to be steam pressure resulting from water gaining entrance inside the fueled region. Photos of the rod and failure area are shown in Figures 21 and 22. The failed rod had about 55$ burnup of the initial 55 gm loading of U-235. This corresponds to about 26 MWD of burnup. It does not appear that the failure is generic, because, as shown in Figure 19, most of the other fuel rods have more burnup than the failed rod - up to 20? more - and no abnormalities are apparent in the other fuel rods. Hot cell examination of the failed rod is still in progress, CONCLUSION The LEU uranium-zirconium hydride fuel has successfully completed all essential development testing and is offered commercially in both the 20 weight percent and 45 weight percent uranium concentrations for existing TRIGA reactors and as a conversion fuel for plate-type reactors. Several reactor facilities have already begun conversion to LEU uranium-zirconium hydride fuel. Currently, U-ZrH LEU fuel is in use in a mixed plate and rod fuel configuration in the 1.0 MW reactor at the National Tsing Hua University in Taipei, Taiwan. A complete core of 20 wt-$ LEU fuel has been fabricated for the 3 MW forced flow TRIGA Mark II reactor system with pulsing capability scheduled to be completed in 1984 in Bangladesh. Also, a complete core of 20 vt-% 4-rod shrouded cluster fuel has been fabricated and shipped for conversion and upgrading of the PRR-1 reactor in the Philippines. The converted reactor will operate at 3 MW with forced cooling and have pulsing capability. Startup is now scheduled for early U-ZrH LEU fuel has also 2-7

8 been delivered to existing TRIGA reactor facilities in Malaysia, Thailand and Yugoslavia and is being put to use as new fuel is needed to meet burnup requirements. For stepwise conversions of higher power reactors (>5 MW), where flow distribution and neutron spectrum effects are of greater importance, a detailed analysis and study is a prerequisite. A joint program involving GA, ANL, and the Cekmece Center is now in progress for analysis of a single U-ZrH LEU fuel test cluster operating in the 5 MW HEU plate-type core of the TR-2 reactor at the Cekmece Nuclear Research and Training Center in Istanbul, Turkey. Inserting this one cluster would be the initial step in a stepwise conversion of the HEU plate-type core to U-ZrH LEU fuel. glbuo-gfiaphy 1. "TRIGA LEU Shrouded Fuel Cluster Design for Operation Between 2.0 MW and 10 MW (Thermal)", GA Technologies Report, UZR-14A. 2. Simnad, M. T., "The U-ZrH x Alloy: Its Properties and Use in TRIGA Fuel", GA Technologies Report E , February Baldwin, N. L., et_..al., "Fission Product Release from TRIGA-LEU Reactor Fuels", GA Technologies Report GA-A16287» November

9 TRIGA FUELS Designation Cladding U Content (WT %) Original Aluminum 8.0 Standard Stainless steel 8.5 ACPR Stainless steel 12.0 FLIP Stainless steel 8.5 High power Incoloy LEU-FLIP Stainless steel 20.0 LEU - high power Incoloy Fig. 1

10 DEMONSTRATION OF TRIGA LEU FUEL Fabricability Metallurgical examination Fission product release Physical properties Hydrogen pressure Quench testing Cycling/pulse testing In-pile testing Fig. 2

11 ANALYTICAL VERIFICATION OF TRIGA LEU FUEL I Prompt negative temperature coefficient Flux performance Reactivity lifetime Fig. 3

12 %*«<# <#-* >: ^ *..Jf ^UlJi «&* ' ^w^/y/mw/>fm Microstructure of 45 wt % LEU fuel rod Fig

13 1 - URANIUM 10"' r 1 ^^ WT% ENR 10 "2 m R D167-R D167-R5-2 A 45 0 D168-R1 A 45 0 D168-R1-1 / E451-R1-A J A 10 "3 TRIGA FISSION PRODUCT / RELEASE CORRELATION / A 10-4 I A/ 6 A A ' *R/B = RATIO OF FISSION GAS RELEASE RATE TO BIRTH RATE IN THE FUEL(UNCORRECTED FOR CONTAINER OR CLADDING EFFECTS) i IRRADIATION TEMPERATURE ( C) Temperature dependence of fission gas release from TRIGA fuel Fig

14 0.08 CJ LU CO u > > % 0.06 u_ U- o _l < 45 WT- 30 WT-% U-Zr Hj'e 8.5WT-% U-Zr Hi' ALL DATA I E TEMPERATURE ( C) Thermal diffusivity versus temperature Fig. 6

15 Photograph of fuel sample 9 before quench from 1200 C Fig

16 Photograph of fuel sample 9 after quench from 1200 C Fig

17 BOW DATA FOR LEU-1 AND LEU-2 TC LEU-1 (1/2-in. diameter) LEU-2 TC (1/2-in. diameter) Cycles (a) A Bend (in.) Bow (b) (in.) A Bend (in.) Bow (b) (in.) pulses pulses pulses Rotated Power cycles (up to 1500 kw) plus indicated number of transient pulses. Bow is approximately given by (A Bend)/2. Fig. 9

18 CALCULATED BEGINNING OF LIFE PROMPT NEGATIVE TEMPERATURE COEFFICIENT (a) AND CORE LIFETIME TRIGA FUEL TYPE DIAMETER (IN.) WT% LENGTH (IN.) U Er URANIUM ENRICHMENT (%) x 10~ 5 = a AVERAGE ( C) CORE LIFETIME ( a ) (MW DAYS) ORIGINAL LEU LEU FLIP 10 MW 10 MW-LEU 14 MW 14 MW-LEU (a) BEFORE INITIAL RELOAD G-453(5) Fig. 10

19 10 15 TRIGA-LEU CORE PLATE-TYPE LEU CORE PLATE-TYPE FUEL LU LU o CM UJ THERMAL FLUX AT MIDPLANE S A,B C, PJ E, F,SS DISTANCE (CM) Mid-plane thermal flux (<0.625 ev) at 10 MW for reactors with TRIGA-LEU fuel and plate-type LEU fuel Fig

20 MV ID FOR 4.3% RE AC TIVITY LO SS TIME FC R INITIAL R ELOADSTEP ' COREBURNUP-MWD K f f as a function of core burnup for reference design - 10 MW Fig

21 COMPARISON OF HEU AND LEU CORES FOR THE IAEA 'TYPICAL' 10-MW RESEARCH REACTOR FUEL TYPE REACTOR PROPERTIES REFERENCE ALUMINIDE ALUMINIDE OXIDE CARAMEL ZIRCONIUM HYDRIDE ENRICHMENT 93% 20% 20% 6.5% 20% GEOMETRY PLATES PLATES PLATES PLATES RODS PLATES/ROD PER ELEMENT URANIUM DENSITY G/CM FUEL MEAT WATER CHANNEL THICKNESS OR DIAMETER, MM 0.51/ / / / U IN FRESH STD. ELEMENTS AVERAGE DISCHARGE BURNUP, %U NUMBER OF EQUIVALENT ELEMENTS DISCHARGE IN ONE YEAR

22 COMPARISON OF HEU AND LEU CORES FOR THE IAEA 'TYPICAL' 10-MW RESEARCH REACTOR (CONTINUED) FUEL TYPE YEARLY FUEL CYCLE COSTS (THOUSANDS OF DOLLARS) REFERENCE ALUMINIDE ALUMINIDE OXIDE CARAMEL ZIRCONIUM HYDRIDE URANIUM ENRICHMENT FABRICATION / FRESH FUEL SHIPPING SPENT FUEL SHIPPING REPROCESSING URANIUM CREDIT TOTAL / S/MWD /

23 LOCATION OF TRIGA LEU FUEL TEST BUNDLE IN ORR c j H CM INTERVALS^ PER REGION i WEST REGION NUMBERS 44 C O N T R O L R O Ds , /B fill = II 30 I1TI = U i Min » r< c* 1 b. u n a a a 9» en c o e a ex > a 1 b j -*? 3 ^« * *» - n ea e. i IN. Q t IN. N n til 20 flij R a 3 CT i K. ) CI o c O N a c< J u i ce» C*. > ««j EAST b 3 C b i b k ex T H ^TRIGA LEU TEST BUNDLE _fc^ V ' < G-45314) Fig

24 GENERAL LAYOUT OF 16-PIN FUEL CLUSTER CM (3.189 IN.) CM (3.135 IN.) CM (2.679 IN.) FUEL CLUSTER DIMENSION INCLUDING CLEARANCE (SAME AS CENTER TO CENTER CLUSTER SPACING) I ho FUEL CLUSTER SHROUD IN. WALL CM CM CM (2.679 IN.) (2.981 IN.) (3.035 IN.) FUEL PIN CM (0.643 IN.) G-45313) IN. WALL CM (0.375 IN.) EL 2686 Fig. 15

25 w TC TC X WT % Uranium - Calculated rod power Y Z 30 TC Power 691 kw As-built fuel pin configuration of TRIGA-LEU fuel cluster - ORR in-pile testing Fig

26 TRIGA-LEU FUEL ORR IN PILE TESTING 20 WT-% U 30 WT-% U 45 WT-% U CONTAINED U235 PER 22 IN. FUEL ROD (GM) VOL % U (19.7% ENRICHED) MAX CALC ROD POWER GENERATION (KW) INITIAL CONFIGURATION FULL CLUSTER CONFIGURATION WT-% ONLY CONFIGURATION 74 TIME AT POWER (FPD) INITIAL CONFIG (DEC 79 NOV 80) FULL CLUSTER CONFIG (MAY 81-MAY 82) WT-% ONLY CONFIG (JULY 82 OCT 83) CONTINUING TARGET BURNUP OF U 2 35 (%) BURNUP STATUS (%) Fig. 17

27 TRIGA LEU FUEL AFTER IRRADIATION IN THE OAK RIDGE REACTOR (ORR) to i ho 20 WEIGHT-% 35% BURNUP 45 WEIGHT % 45% BURNUP 30 WEIGHT % 40% BURNUP G-45318) Fig. 18

28 *TC = THERMOCOUPLE ELEMENT *FPD = FULL POWER DAYS 39.3 i m CO ON = 50 ROD 1093 FAILED AT 901 FPD ~ 53% U-235 BU E I TC ROD 1088 REMOVED FOR ANALYSIS AT 721 FPD -47% U-235 BU PRESENT IN-CORE EXPOSURE FULL POWER DAYS (ORR) NOTE: NORMALIZED TO ORNL FISSION PRODUCT ANALYSIS ON RODS 1082, 1088, 1093, AND 1098 Fig. 19. Calculated Burn-Up Versus Full Power Days for 45 Wt-% TRIGA-LEU Fuel in the ORR 2-28

29 CURRENT CONFIGURATION, TRIGA-LEU FUEL CLUSTER ORR IN PILE TESTING ss WT % URANIUM - CALCULATED ROD POWER 45TC H 2 OINC 0 H 2 OINC 0 H 2 OINC H 2 OINC H 2 OINC 0 H 2 OINC 0 SS = STAINLESS STEEL ROD H 2 0 INC = WATER-FILLED INCOLOY CLAD TUBE EPOWER 564 KW Fig. 20

30 I O! U&& Fig. 21

31 ho I