Prony’s Method for BWR Stability Analysis

Power uprates are becoming a quite viable option both for improved performance and safety of nuclear power plants. Part of the success in power uprates are new designs of fuel assemblies. Nevertheless, power uprates in BWRs has made mandatory new stability analyses since operational and startup maneuvers are not necessarily independent of the performance of such new fuel assembly designs. For example, it has been inferred that certain operational areas, within the before-to-uprate power-flow map, could be reached for the new expected operational conditions, especially during startup, because certain fuel assembly designs have shown faster response to certain neutronic perturbations and their lower coolant flow area. As a consequence, stability monitors are being taken into account as a great help for operation in such special maneuvering conditions. Currently, power oscillations during startup are of not of much concern since boiling boundary length, decay ratio, and other similar parameters are continuously followed by power stability monitors. However, there still exist certain transients, as a recirculation pump trip suddenly occurring at rated operation, in which the instability range can be difficult to determine in real time. One alternative, for a fast analysis of fast events, is Prony’s Method, which is mostly employed in power and energy transport systems problems. This analysis method provides information about stability through a short-time series, in quasi-steady conditions, which may be expected in certain BWR transients. The method computes the complex dampening coefficient, which is the most instable pole in an autoregressive analysis of the time series. However, not many applications in BWR operation have reported and supported to establish the scope of using such analysis for actual events. This work presents the response method to the impulse of an autoregressive model and its relation to the damping coefficient of Prony’s method. It is found a second order behavior with respect to the decay ratio. To support such result, three different noise signals are analyzed where BWR transient events in which low coolant flow and high power conditions are assumed.