Effects of different methods of simulating the shim rods in the 'as-built' configuration in the Cernavoda NPP Unit 1

Shim rods play an important role in the frame of CANDU reactor regulating and control systems. The geometry and effective size of these rods change following the results obtained from AECL computer simulations. For the as-built rods of the Cernavoda NPP Unit 1, the computations reported in this work were carried out using both the classical AECL procedure and an alternating multigroup transport methodology developed in our institute, based on WIMS and PIJXYZ codes. Subsequently, by means of DIREN code (developed in INR Pitesti), single and group shim rod reactivities were computed for two types reactor cores (fresh and burned for 600 days). From a comparison of the results obtained with the two methodologies the following three conclusions could be drawn. 1. The reactivities estimated for the 'as-built' shim rods are close to those specified in the AECL project. Although, all types of 'as-built' shim rods are slightly more absorbent than the project ones, the differences are minor and practically insignificant for NPP's operation. 2. The rather small differences between the shim rod efficiencies as measured in Cernavoda NPP 'PIF' conditions and those resulting from Westcott classical methodology indicate a correct utilization of an assembly of computer codes and working networks, able to simulate the usual reactor core of the NPP, with a rather high level of precision. These relative differences are within the range (-6.77, + 9.56)% , with an range width of 16.33%. 3. The alternate simulation methodology, now in progress at INR Pitesti, proved to be satisfactory from the very beginning. The relative differences between computation and measurements are comprised within the range (-14.25, +5,01)%. The range width of 19.26% is only 2.93% higher than in the case of Westcott methodology. The fact that the computed reactivities are close to the measured ones is important since the INR Pitesti provides a local information source extremely rich, as for instance, for reaction rates for regions accurately represented, multigroup neutron fluxes in a hyper-fine distribution, etc. In addition, the applied methodology allows easy use of new nuclear data, no empirical relationship and data, computational relation or result fitting being involved