Derivation and experimental demonstration of the perturbed reactivity method for the determination of subcriticality

The perturbed reactivity method is a general technique for the estimation of reactivity. It is particularly suited to the determination of a reactor's initial degree of subcriticality and was developed to facilitate the automated startup of both spacecraft and multi-modular reactors using model-based control laws. It entails perturbing a shutdown reactor by the insertion of reactivity at a known rate and then estimating the initial degree of subcriticality from observation of the resulting reactor period. While similar to inverse kinetics, the perturbed reactivity method differs in that the net reactivity present in the core is treated as two separate entities. The first is that associated with the known perturbation. This quantity, together with the observed period and the reactor's describing parameters, are the inputs to the method's implementing algorithm. The second entity, which is the algorithm;s output, is the sum of all other reactivities including those resulting from inherent feedback and the initial degree of subcriticality. During an automated startup, feedback effects will be minimal. Hence, when applied to a shutdown reactor, the output of the perturbed reactivity method will be a constant that is equal to the initial degree of subcriticality. This is a major advantage because repeated estimates can be made of this one quantity and signal smoothing techniques can be applied to enhance accuracy. In addition to describing the theoretical basis for the perturbed reactivity method, factors involved in its implementation such as the movement of control devices other than those used to create the perturbation, source estimation, and techniques for data smoothing are presented