Neutronic design of instrumentation for thermal effects measurement on VR-1 reactor

At the Czech Technical University (CTU) in Prague a new instrumentation for measurement of thermal effects is currently designed for the use on the training reactor VR-1 “Sparrow”. The instrumentation is designed to serve both for the purposes of education and training and R&D; in the latter case especially for the examination of possible precision and accuracy measure of calculation of such effects. VR-1 is a light water zero power pool type reactor which operates at Dept. of Nuclear Reactors (DNR) CTU in Prague. The core is assembled on a grid of 8 × 8 positions (each position possessing 71.5 mm × 71.5 mm) and is composed of three various tubular fuel assembly types: four-, six- and eight-tube fuel assemblies of IRT-4M type. This fuel consists of UO
₂
dispersed in Al with the enrichment on U-235 of 19.7%. As at the nominal power of ca. 1 kW the reactor operates at an ambient temperature and there is almost no heat generated, the instrumentation has to be designed as a closed loop possessing some external system for water heating. It is assumed it could cover 1 to 4 positions of the reactor grid. Thus, the quantity to be measured with the instrumentation will be the isothermal temperature coefficient of reactivity (ITC) of the used module. It is assumed that the measuring range should start at an ambient temperature going up to some 70°C, the upper boundary being chosen in such way to stay sufficiently safely below design limits of used fuel assemblies. There are two types of nuclear material under consideration to be potentially used in the instrumentation that are in the inventory at DNR: Besides IRT-4M fuel assemblies which are used for standard operation of the reactor also EK-10 pins could be used. These pins use as a fuel meat Mg-UO
₂
enriched on 10% of U-235 and at DNR is their utilisation restricted to experimental purposes only. With both fuel types, an optimization of the instrumentation was carried out. It comprised the calculations of isothermal temperature coefficients of reactivity for various cases of considered fuel assemblies in infinite lattice in the temperature range of 20 to 70°C. In the analysis all types of IRT-4M fuel assemblies (4-, 6- and 8-tube fuel assemblies) were included; for 4- and 6-tube FA additionally also the case with inner displacer of varying width for modifying fuel to moderator ratio was assumed. As to EK-10 fuel pins, infinite lattices with varying pin to pin distance both for square and triangular geometry were studied. From the results, several modules covering 1 to 4 positions of reactor grid were proposed and the corresponding achievable changes of reactivity for typical core of the reactor containing these module variations were calculated. In the case of modules with EK-10 fuel pins also a layer isolating the instrumentation from the rest of the core was designed (Isolation of modules with IRT-4M fuel is not feasible due to their and grid geometry). Finally, the results were intercompared and the optimal design was found and discussed with respect of design limitation of the particular fuel types, possible reactivity change achievable by given module (and possibility of measuring such reactivity changes at the reactor), possibility of precise modelling of the operational states of the module (for example the impossibility of isolating the module from the rest of the core in the case of IRT-4M fuel assemblies and corresponding uncertainty in reactivity changes prediction connected with heat transfer to the rest of the core handicap this type to some degree against EK-10 variations) and the possibility of unaffecting of the rest of the core by module operation. The calculation analysis was carried out in MCNP5 code with the utilization of detailed MCNP model of the VR-1 reactor. With regard to relatively small absolute values of reactivity changes, great care was paid to nuclear data used and to their processing. The data from ENDFB7 nuclear data library were used; they were processed by NJOY 99.259 code using well described and validated data processing sequence which was used at NEA for processing JEFF 3.1 library for use in MCNP code. The most significant cases were calculated also with JEFF 3.1 and JENDL 3.3 data libraries for comparison purposes using the same data processing sequence. Thermal scattering law data (which are standardly available only for a small set of temperature values) were for desired temperatures obtained by interpolation between the available cross sections data.