Application of a ZED-2 benchmark experiment to CANDU coolant void reactivity validation

Neutron transport codes are used to calculate the reactivity coefficients needed to predict and model reactor behaviour. Computed reactivity coefficients, such as coolant void reactivity, must be carefully validated against experimental data to ensure their accuracy. This work describes an application of the TSUNAMI nuclear data-based validation methodology to a CANDU Coolant Void Reactivity (CVR) application using experimental data from the Zero Energy Deuterium 2 (ZED-2) reactor. Eight highly correlated measurements from a ZED-2 reactor flooding experiment were used to predict the computational bias of CVR predictions for a CANDU bundle using the TSURFER nuclear data adjustment code. Experimental uncertainties and correlations were determined by evaluating the effects of perturbations to experimental parameters. The post-adjustment χ ν 2 value for the set of experiments was determined to be 1.1. A computational bias due to nuclear data between 0.32 mk and 0.36 mk, depending on bundle burn up, was determined with an average nuclear data uncertainty reduction of 19% relative to the reference data. This work serves as a simple example of how the TSUNAMI framework can be applied to reactivity coefficient validation. • The nuclear data adjustment code TSURFER was applied to a CANDU CVR application. • Data from a ZED-2 flooding experiment yielded a 0.32–0.36 mk CANDU CVR bias. • The ZED-2 data reduced CVR nuclear data uncertainty by 19% . • The applicability of TSURFER to reactivity validation is discussed.