Calibration of in-reactor detectors and determination of the computational error of the RBMK-1000 SKALA-mikro information measuring system

Abstract

Several methods are used to calibrate in-reactor detectors or to determine their sensitivity to the power of the fuel assemblies at zero burnup. For a long time, the most widely used method was axial scanning of the fuel assemblies of assembly 49 with a dry hollow casing by detector Dt.6 or Dt.4 with an HfO 2 detector and long sensitive part, whose length equals the core height [1]. The power of a fuel channel with an in-reactor detector, reconstructed from the scan results and calculated using the signals from the detector, was used to determine its sensitivity taking account of the burnup of the fuel assembly and detector. The rms deviation of the error in this calibration is 3‐4%. Subsequently, an adjacent reference sample was used to calibrate the sensitive part of the detector, since the amount of HfO 2 per unit length of the sensitive part varies very little along the sensor cable. The rms deviation of the ratio of the HfO 2 amount to the sensitive part of the in-reactor detector and two reference samples, placed at the ends of the detector, is about 1%. The sensitivity of each reference sample was determined in the IRT reactor at the Moscow Engineering Physics Institute and extended to the corresponding sensitive part of the in-reactor detector. However, this method was time-consuming and expensive. It was improved by determining the dependence of the sensitivity of the calibrated reference samples on HfO 2 content of each per unit length. The amount of HfO 2 was determined by measuring the resistivity and length of emitter shell of the reference sample, making it possible to determine the mass of the emitter shell, and weighing the emitter in the inner shell after the outer shell and sensor cable insulation were removed. This calibration requires about 10 min. The sensitivity of an in-reactor detector is calculated from the relation η = ƒ(m), (1) where m is the mass of the emitter material. The unit (current) chosen for the sensitivity to the number of neutrons captured by a hafnium nucleus in 1 sec makes it possible to eliminate the effect of the neutron spectrum on it. The ratio of the rms deviation of the measurements of the sensitivity of the reference samples and the calculation with the (1) fit was obtained on the basis of the expressions in [2]. The result was the rms error in determining the sensitivity using Eq. (1), equal to about 1%. The procedure described also made it possible to determine the mass of the outer shell and insulation of the emitter, having included them in the expression for the sensitivity calculation. Since the positive gain in the rms error of calibration is negligible (it is 1.1 times less with 1.3‐1.5 times larger costs), dependence (1) is used in practice. The procedure described is termed the factory calibration.