Alloy M5 in action3 September 2002
The use of M5 for fuel rod cladding and guide tubes provides significant improvements in oxidation, hydrogen absorption, axial growth, and creep compared to Zircaloy-4. By G L Garner and J P Mardon
Increases in fuel assembly burnup are limited by the irradiation behaviour of its materials. Corrosion, hydriding, growth and creep performance of current zirconium alloys does not permit the more aggressive operating and environmental conditions required to achieve higher duty and longer cycles.
Alloy M5, developed by Framatome ANP, for fuel rod cladding, guide tubes and spacer grids, exhibits superior performance in high burnup, high duty PWR applications and in a wide variety of PWRs worldwide. M5 is a fully recrystallised, ternary Zr-Nb-O alloy with the composition shown in Table 1.
Four advanced lead assemblies with M5 fuel rod cladding and guide tubes recently completed three cycles in Dominion Generation's North Anna 1 reactor. A post irradiation inspection (PIE) campaign, funded by EPRI as part of the Robust Fuel Programme and performed by Framatome ANP during and after the cycle 15 refuelling outage, served two purposes. The first was to assess the performance of the advanced lead assemblies at normal discharge burnups for three representative cycles of irradiation. The second was to verify that the assemblies had sufficient margin so that one could be safely irradiated for a fourth cycle to a peak rod burnup greater than 70MWd/kgU.
Each of the advanced lead assemblies consists of 264 fuel rods, 24 guide tubes and one instrument tube in a 17x17 square array. Each assembly utilises mid-span mixing grids in the upper three spans to provide increased thermal margins. Mixing vanes also are provided on the second through sixth intermediate grids numbered from top to bottom.
A seventh intermediate grid (grid number 7) does not have mixing vanes.
Fuel cycle operation
A summary of the operation of the three lead assembly
irradiation cycles in North Anna 1 is shown in Table 2.
Fuel assemblies in unit 1 are typically shuffled in an in-in-out pattern. The power history for the lead assemblies is given in Table 3 in terms of radial power distribution.
Each cycle followed a coordinated lithium-boron chemistry control programme, in which the pH at the average operating temperature was held constant at a value of approximately 6.9 throughout the cycle.
Fuel examination results
All of the examination data was within the design ranges and models used for the advanced lead assembly design. No adverse trends were indicated, and the performance of the M5 lead assemblies was consistent with or superior to that seen in the existing Framatome ANP worldwide database. The advanced lead assemblies at North Anna 1 easily possess sufficient margin for a fourth cycle of irradiation. The current schedule calls for one assembly to be inserted in North Anna 2 later this year.
More than 15 individual fuel assembly attributes were evaluated on the lead assemblies in the course of the cycle
15 PIE examination. The results of the most significant examinations are summarised here.
Visual and leak test
Full-length visual examinations were performed on each lead assembly at the end of each cycle of operation. The appearance of the assemblies was typical and no visual abnormalities were observed. The fuel rod cladding was seen to be clean and smooth throughout their active length. The visual exams were followed by in-mast sipping to determine if there were any failed fuel rods. In their third cycle, the lead assemblies were in locations on the core baffles that had exhibited fuel rod failures in other fuel designs. This examination confirmed that all four lead assemblies were leak-free.
Fuel assembly length
The length of each lead assembly was measured using the same method and tooling used in the characterisation of the assemblies prior to their initial insertion in unit 1. The results for all three cycles showed essentially zero growth for the four lead assemblies after three cycles of operation. (Two of the lead assemblies with a "floating" upper end grid design actually experienced a slightly negative growth.)
The growth of all four lead assemblies was significantly less (approximately 10 times less after three cycles) than that of conventional fuel assemblies with Zircaloy-4 SRA cladding, RXA guide tubes and 3-leaf holddown spring systems. The low growth of alloy M5 is well understood and the use of M5 fuel rod cladding and guide tubes certainly played a role in the low growth of the lead assemblies. However, the large forces imparted by the Advanced Mk-BW lead assembly holddown springs may also have been a significant contributor to the reduction of growth. These large holddown forces were specific to the design of the North Anna lead assemblies and the forces will not necessarily be as high on reload batches of Mk-BW fuel.
Fuel rod growth
A plot of fuel rod growth as a function of rod burnup is shown in Figure 1. The M5 database and the nominal Zircaloy-4 growth curve are included for comparison. The saturation growth behaviour of recrystallised M5 is evident and accounts for the lower growth of M5 compared to Zr-4 particularly at the higher burnups.
Fuel assembly distortion
Three examinations were performed to evaluate fuel assembly and fuel rod distortion: RCCA drag tests (withdrawal and insertion), fuel assembly bow measurements and water channel gap measurements. The results from all measurements confirmed that the lead assembly structure and fuel rod distortions were low and well within the normal experience database for these measurements.
Fuel rod oxidation
One of the widely recognised advantages of zirconium-niobium alloys is their excellent corrosion performance in the PWR environment. This optimum alloy composition and the low temperature fabrication process for M5 result in a microstructure that is highly stable.
Oxide thickness measurements were made in the hottest span of the North Anna 1 lead assemblies (second from top) on peripheral and interior rods. The results are plotted in Figure 2 along with the European and American Zr-4 and M5 databases. The Zr-4 data shown in this plot is from optimised, low tin cladding. It is evident that the behaviour of M5 and Zr-4 is quite different. Firstly, Zr-4 oxidation accelerates at burnups greater than 40MWd/kgU. Secondly, the data scatter for Zr-4 becomes increasingly wider with increasing burnup. This widening of the data scatter is evidence of the sensitivity to reactor duty parameters (such as temperature and power) of Zr-4. The alloy clearly behaves differently in different temperature and power history ranges. The oxidation of M5, on the other hand, exhibits a linear evolution with burnup in a wide range of operating conditions and burnups greater than 70MWd/kgU with no widening of the data scatter band.
Oxide thickness was also measured on the M5 guide tubes after cycles 14 and 15. The maximum measured guide tube oxide thickness on the inside surface was less than 5 micrometres after three irradiation cycles (52,472MWd/tU fuel assembly burnup). Oxide thicknesses measured on the RXA Zr-4 spacer grids (which historically tend to be similar to guide tube values) were of the order of 30 micrometres in the hottest span.
The hydrogen uptake in alloy M5 can logically be expected to be much lower than Zircaloy because of its lower oxidation rate. Simply put, less hydrogen is available for absorption into the alloy because of the lower oxidation kinetic. Another contributor to lower hydrogen absorption in M5 is the material's lower pick-up fraction. Hot cell examinations of M5 fuel rod cladding confirm this assessment and can be seen graphically in Figure 3.
As with the oxidation performance of M5, the hydrogen content increases linearly with little scatter and very low values. End of life hydrogen content for M5 fuel rod cladding will certainly be less than 100ppm.
TablesTable 1: M5 composition Table 2: Lead assembly cycles at North Anna 1 Table 3: lead assembly history