Conceptual Design for a Fast Reactor Safety Test Facility

In May 1975 Los Alamos Scientific Laboratory issued a preliminai.. report on a study of Fast Reactor Safety Test Facilities. The study addressed itself to three closely related tasks. 1. A review of the current understanding of fast reactor safety with the aim of identifying important areas of uncertainty which cannot be adequately resolved using analysis, out of pile and/or existing in-pile facilities. 2. Conceptual design studies of one or more new in-pile facilities having characteristics identified in (1) above. 3. An examination of advanced data acquisition techniques for possible incorporation in the new facilities. The work reported in this document is an extension of the earlier work in task area (2) above. Sased largely on conclusions drawn from the earlier work the scope of the current effort has been narrowed to the design study of a Type A facility operating in the Class III mode, i.e., a facility capable of accommodating up to 37 test pins and capable of imposing a burst on top of a high steady state power level. I. SUMMARY AND CONCLUSIONS The Safety Test Facility (STF) conceptual studies carried out at LASL earlier this year provided a basic understanding of the strongly interdependent aspects of STF designs, namely the neutronic, thermalhydraulic, material properties and safety, and indicated the various trade-offs which must be considered in desigi. optimization. Conclusions drawn from those studies provided the starting point for the work reported here. In brief these conclusions were: 1. Of the three fuel types studied, UO2-Cr,(UC-ZrC)-graphite (ROVER), and BeO/ UOj, the last one showed far superior potential for the driver fuel in an STF operating in the Class III mode. 2. Neutronic and thermal-hydraulic considerations heavily favor a liquid metal. cooled driver over one cooled by gas. 3. For a Type B facility with a BeO/ U02 driver, a converter region will be necessary in <̂ rder that the ratio of maximumto-minimum power across the test pin bundle be less than the upper limit of 1.15 which has been specified for 217-pin FTR or larger subassemblies. A facility employing a converter tends to be large with high heat removal requirements, amounting to several hundred MW for Class III operation, even with highly enriched test pins. The capability to test normal FTR fuel pins would involve reactor powers approaching 1000 MW. 4. By virtue of its smaller test bundle size the design problems of a Type A facility are significantly less than those of a Type B facility. In particular, a converter region is not likely to be required in a BeO/UOdriven Type A facility because of different experimental objectives. Also, potential positive reactivity effects due to test fuel relocation after the burst are much smaller in Type A. In this recent study the emphasis has been placed upon developing a reference design for a Type A, Class III facility cooled with sodium rather than the general studies of tt.e preliminary work. The feasibility of using PuO2-UO2 as a driver fuel was considered while ROVEK fuel and UO-j-C. were removed from further consideration. Improved resonance self-shielding methods were used and a slightly more realistic treatment of the test fuel cross sections was employed. More -detailed transient heat transfer calculations were undertaken including investigations of the effect of nonprototypical power gradients across an individual test pin. A coupled transient heat transfer-neutronic feedback code was developed and used to study the burst kinetics. Two-dimensional transport calculations were employed to look at test fuel relocation reactivity effects. Initial attempts have been made to translate the reference physics design into a semi-detailed engineering design. Although the general aspects of the safety of the reference design have been discussed, the costs, construction schedule and siting have not been considered. The following broad conclusions were drawn from the study. 1) Both BeO/UO2 and PuO2~UO2 have good potential as STF driver fuels. The former has the advantage that burst energy densities in excess of 2000 J/g above steady state can be deposited in FTR-fuel test pins. On the other hand PuO2-UO2 drivers cannot deposit the desired energy density unless the uranium in the FTR fuel is enriched with U. A major disadvantage of the BeO/UO fuel is the large power depression it creates across Type B sized test pin bundles. In view of the different experimental requirements for facility types A and R it is apparent that PuO -uo., would be the better driver fuel for a Type B facility, vhere a flat power profile (without tes! fuel enrichment grading) in an important criterion, but in a Type A facility, where the capability to test normal FTR fuel is desirable and graded enrichment acceptable, BeO/UO, is the preferred fuel. 2) A BeO/UO2 driven Type A Clciss III facility operating with FTR fuel in the test region will have heat removal requirsn.ents in the range 125-185 MW when the driver radius is fixed at 40 cm. With highly enriched FTR size test pins the newer is arourd 50 MW and the burst energy density capability is much greater. 3) Fairly moderate reactivity insertions (up to 2$) and ramp rates {up to 5OS/ s) will give the required range of burst energy densities in bursts lasting 10 ms or so. The Doppler coefficients calculated for the reference design would be sufficient to turn the burst around and reduce the system to about 50C below prompt critical. However, there is an FOM* value below which the post-burst power level plateau would damage the driver unless a negative reactivity ramp is initiated within a certain time after the burst. Although the coupled neutronics-transient heat transfer calculations performed in the study were preliminary, the indications are that with normal FTR fuel such a shutdown mechanism will be needed. 4) In the range of driver core sizes being considered, 40-to 50-cm radius, it appears that adequate control for burn-up, power coefficients, burst initiation and post-burst shutdown can be achieved by control mechanisms placed in the reflector. To obtain an adequate shutdown margin (about $10) for loading/unloading operations it may be necessary to insert poison rods into the driver region. If these tods can be inserted into the diagnostic instrument slots •Figure of Merit defined as the ratio of the minimum power density in the test fuel to the maximum power density in the driver fuel, the impact on the system performance would be minimal, providing the rods were inserted only after collection of experimental data was complete. I