Development of a Noise-Based Method for the Determination of the Moderator Temperature Coefficient of Reactivity (MTC) in Pressurized Water Reactors (PWRs)

Doctoral thesis, 2002

The Moderator Temperature Coefficient of reactivity (MTC) is an important safety parameter of Pressurized Water Reactors (PWRs). In most countries, the so-called at-power MTC has to be measured a few months before the reactor outage, in order to determine if the MTC will not become too negative. Usually, the at-power MTC is determined by inducing a change in the moderator temperature, which has to be compensated for by other means, such as a change in the boron concentration. An MTC measurement using the boron dilution method is analysed in this thesis. It is demonstrated that the uncertainty of such a measurement technique is so large, that the measured MTC could become more negative than what the Technical Specifications allow. Furthermore, this technique incurs a disturbance of the plant operation. For this reason, another technique relying on noise analysis was proposed a few years ago. In this technique, the MTC is inferred from the neutron noise measured inside the core and the moderator temperature noise measured at the core-exit, in the same or in a neighbouring fuel assembly. This technique does not require any perturbation of the reactor operation, but was nevertheless proven to underestimate the MTC by a factor of 2 to 5. In this thesis, it is shown, both theoretically and experimentally, that the reason of the MTC underestimation by noise analysis is the radially loosely coupled character of the moderator temperature noise throughout the core. A new MTC noise estimator, accounting for this radially non-homogeneous moderator temperature noise is proposed and demonstrated to give the correct MTC value. This new MTC noise estimator relies on the neutron noise measured in a single point of the reactor and the radially averaged moderator temperature noise measured inside the core. In the case of the Ringhals-2 PWR in Sweden, Gamma-Thermometers (GTs) offer such a possibility since in dynamic mode they measure the moderator temperature noise, whereas in static mode they measure the spatial distribution of the neutron flux. Both of these are required to estimate the core average moderator temperature noise. There are 12 radial positions where GTs are installed, which makes it possible to approximate averages over the horizontal cross-section of the core quite well.