Derivation of Uncertainty Distributions for Channel Flow Rate and Fuel Critical Heat Flux Predictions for Best-Estimate Plus Uncertainty Analysis of Slow Loss-of–Reactor Power Regulation Accidents in CANDU Stations

Uncertainty in the figure of merit (FOM) parameters is a central feature of the best-estimate plus uncertainty (BEPU) method, which provides insight into the analysis margins not available from other analysis methods. The FOM uncertainty distributions are formed from propagation of the variations and uncertainty distributions in the operational and modeling parameters used in simulations of a design-basis accident (DBA) scenario for a nuclear power plant.To compute an accurate FOM uncertainty distribution, it is critical to accurately quantify and account for the input parameter prediction uncertainties. The coolant flow rate through fuel channels, or more precisely, the hydraulic resistance, including the impact of two-phase flow and its distribution in the primary heat transport system and the critical heat flux (CHF) of the fuel, are two key parameters for the limiting postulated accident scenarios in a CANDU reactor for various DBAs.Prediction uncertainty distributions for these parameters can be derived by directly validating code predictions against in-reactor measurements of flow rate and experimental measurements of CHF, respectively. Such code validation circumvents the convoluted and complex approach of decomposing computer models of physical phenomena into microscopic parameters, such as interfacial mass, momentum, and heat transfer correlations, and propagation of their uncertainty distributions to obtain an overall parameter uncertainty distribution of interest. Uncertainties associated with predictions of the coolant flow rate and CHF arise due to temporal and spatial variations and uncertainties in reactor conditions, limitations of physical models and their implementation in the codes, and in the case of CHF, measurement uncertainties associated with full-scale experiments.Careful assessment of key uncertainties, specifically their magnitudes, is important for ensuring uncertainty magnitudes are not unnecessarily over- or underestimated. These uncertainties also need to be characterized properly, e.g., whether uncertainties are common to a group of reactor fuel channels or vary independently for each fuel channel. Inadequate identification or incorrect classification or characterization of uncertainties would result in an inaccurate FOM uncertainty distribution.One important focus area for this study is the distinction between apparent prediction uncertainty (the difference between code prediction and measurement) and actual prediction uncertainty (the difference between code prediction and the true value). The actual code prediction uncertainty can be calculated from the apparent code uncertainty, provided there is adequate knowledge about the measurement uncertainty. The uncertainty models developed using this approach will be used as part of the BEPU analysis for slow loss-of–reactor power regulation accidents.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.